First published in 2012 by Gloucester Publishers Limited, Northburgh House, 10 Northburgh Street, London EC1V 0AT Copyright © 2012 John Watson The right of John Watson to be identified as the author of this work has been asserted in accordance with the Copyrights, Designs and Patents Act 1988. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, electrostatic, magnetic tape, photocopying, recording or otherwise, without prior permission of the publisher. British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library. ISBN: 978 1 85744 939 6 Distributed in North America by The Globe Pequot Press, P.O Box 480, 246 Goose Lane, Guilford, CT 06437-0480. All other sales enquiries should be directed to Everyman Chess, Northburgh House, 10 Northburgh Street, London EC1V 0AT tel: 020 7253 7887 fax: 020 7490 3708
email:
[email protected]; website:www.everymanchess.com Everyman is the registered trade mark of Random House Inc. and is used in this work under licence from Random House Inc. To
Gerry
Everyman Chief Commissioning Assistant Typeset Cover
Giles,
favourite
teacher
and
father-in-law,
for
his
kindness
Chess
and
Contents
my
advisor: editor: editor: edited by design
First by
Byron John Richard Rank Horatio
Publishing,
and
generosity. Series Jacobs Emms Palliser Brighton. Monteverde.
Bibliography Introduction
1 Advance Variation: Introduction and Traditional Lines 2 Advance Variation: 6 a3 3 Advance Variation with 5 ... Nh6 4 Tarrasch Variation: Introduction to 3 ... Nf6 with 5 Bd3 5 Tarrasch Variation: Main Lines with 5 Bd3 6 Tarrasch Variation with 5 f4 7 Winawer Variation: Fourth Moves 8 Winawer Variation: Fifth Moves 9 Winawer Variation: Poisoned Pawn Variation 10 Winawer Variation: Positional Main Lines 11 Winawer Variation: Portisch-Hook Variation 12 Exchange Variation 13 Classical Variation with 4 e5 14 MacCutcheon: Introduction 15 MacCutcheon: Main Lines with 6 Bd2 16 King’s Indian Attack 17 Uncommon Early Moves
Bibliography
Rather than present a nearly endless list of sources, I’ve chosen the most influential titles which have appeared since the previous edition, as well as important periodicals. You should note that, although I credit authors and annotators for significant analysis, I haven’t cited one- or two-move suggestions for the most part, especially not if they’reRybka’s orHoudini’s (or Fill-inthe-Blank’s) immediate choice! Books and eBooks Attacking Chess: The French, Simon Williams (Everyman 2011) Chess Explained: The French, Viacheslav Eingorn & Valentin Bogdanov (Gambit 2008) Dangerous Weapons: The French(akaDWF), John Watson (Everyman 2007) French Advance(2nd edition), Sam Collins (Everyman 2006) French: Advance and Other Lines, Steffen Pedersen (Everyman 2005) French Defence: 3.Nc3 Bb4, Lev Psakhis (Batsford 2003) French Defence Advance Variation,Volumes 1&2, Evgeny Sveshnikov; (Olms 2007) French Defence: Steinitz, Classical, and Other Systems, Lev Psakhis (Batsford 2004) How to Beat the French Defence, Andreas Tzermiadanos (Everyman 2008) How to Play Against 1 e4, Neil McDonald (Everyman 2008) Opening Repertoire According to Anand,Volumes 6&7, Alexander Khalifman (Chess Stars 2006) Play the French(1st edition, akaPTF1), John Watson (Everyman 1984) Play the French(2nd edition, akaPTF2), John Watson; (Everyman 1996) Play the French(3rd edition, akaPTF3), John Watson (Everyman 2003) Starting Out: 1e4!, Neil McDonald (Everyman 2006) The Flexible French, Viktor Moskalenko (New in Chess 2008) The Wonderful Winawer, Viktor Moskalenko (New in Chess 2010) The French Defence: A Complete Black Repertoire, Nikita Vitiugov (Chess Stars 2010) The French Defence Reloaded, Nikita Vitiugov (Chess Stars 2012) CDs and 1 ... e6: A Solid Repertoire, Nigel Davies (ChessBase Beating the French, Volumes 1&2, Rustam Kasimzhdanov (DVD, ChessBase The French Defence, Ari Ziegler (DVD, ChessBase Fit for the French, Viktor Bologan (ChessBase Killer French Defence(Parts 1 and 2), Simon Williams (DVD, Ginger GM Encyclopaedia of Chess Openings, Volume C(akaECO) (DVD, Chess Informant Opening Encyclopaedia 2011(DVD), (ChessBase Französisch Vol.1-2, Thomas Luther (ChessBase Periodicals & Secrets of Opening Surprises,Jeroen Bosch, Editor ChessBase Magazine(DVD), New in Chess Yearbook(New Kaissiber, Stefan Bücker, Chess Informant(Chess New in Chess Magazine(New ChessPublishing.com, Tony Kosten, Webmaster The Week in Chess, Major UltraCorr3a, Correspondence Megabase Downloads
Tim Database fromThe
Introduction
Week
in
(New
in
in
in and Mark
Harding 2011(ChessBase 2012(ChessBase Chessand
the
Internet
Chess
DVDs 2009) 2007) 2006) 2010) 2010) 2007) 2011) 2005) Websites Chess) ChessBase Chess) Editor Informant) Chess) Editor Crowther Databases (2010) 2011) 2012) Club
Why another edition ofPlay the French? Over the past two or three years, after all, numerous books on the French Defence have appeared, most of them written by very competent and insightful authors. As is the wont with modern opening works, these books usually centre their recommended variations around an instructive and/or entertaining game, without great depth but with sufficient detail to show the main branches and explain basic ideas. This is absolutely legitimate and is in fact the direction towards which I have gravitated in some of my own works. But it leaves room for a book on the French with fewer generalities and more nitty-gritty chess substance, one in which you have a better chance of finding precisely how your latest adventure with 1 ... e6 corresponds to a set of moves in the book, whether or not those moves are flawed.Play the French 4is first of all a repertoire book, with a wider-than-normal choice of lines; but it is also as a reference work which tries to anticipate as many of your opponents’ logical moves as possible.
The repertoire itself is highly flexible, consisting of both mainstream variations and experimental forays. At most key junctures, the reader is offered two or more distinct variations to choose from, providing sound alternatives should some line of play prove unsatisfactory or in need of repair. Of course, a book that tries to cover so many byways will necessarily be dense and in some spots difficult to read. Therefore you may want to skim over chapters before settling in to a closer examination. But I would urge you to pay particular attention to the alternatives I’ve given to what are (temporarily) the ‘main lines’. This both prepares you to meet moves that are likely to be played, but also informs you of key ideas that may not be explicitly described elsewhere. Notice, too, that I’ve let many subvariations speak for themselves. Not thatPlay the Frenchwas ever excessively chatty, but I have often sacrificed verbiage to make room for specifics. I’ve also devoted much more space to original analysis, with less resort to bare game excerpts. What’s new in this edition? It’s been nine years sincePlay the French 3, seemingly a lifetime given the rapidity with which theory changes in such a popular opening. I have therefore updated and re-analysed the great majority of variations, often in depth. I’ve also presented new solutions. For example, in the Advance Variation (3 e5), apart from changing some fundamental ways of playing 5 ... Qb6 in the main line, I have added a chapter on 5 ... Nh6. Versus the Tarrasch (3 Nd2), I have returned to 3 ... Nf6, not only on the basis of proven merit but also because very little, if anything, has changed in a fundamental sense in the 3 ... c5 lines. In the Winawer (3 Nc3 Bb4), you will see a new chapter on 6 ... Qa5 (which I call the ‘Portisch-Hook Variation’), and a return to the Winawer Poisoned Pawn in the main 7 Qg4 line. There are two new chapters on the MacCutcheon Variation (3 Nc3 Nf6 4 Bg5 Bb4, replacing 4 ... dxe4), which is an active approach in the spirit of the Winawer; and you’ll also find a new main line in the Classical System (3 Nc3 Nf6 4 e5). These variations and similarly new approaches in lesser systems by no means indicate the failure of corresponding solutions from previous editions; on the contrary, most of them are still fully playable. But they do show how wonderfully flexible the French Defence is, and how you needn’t despair of it should a particular system fail to satisfy your needs. Finally, in terms of coverage, I have tried to pay deeper and more respectful attention to variations which are not currently in fashion and/or which have a strong following at lower levels of play. Another change concerns the natural bias in such a book. The ‘author-advocate’ form is a difficult one; for this volume, at the harsh insistence of computer engines, I’ve spent more time modifying or rejecting my own previous suggestions than in any book that I’ve previously written. In the past, moreover, I’ve made the assumption that the reader will understand how I have tended to choose excerpts to illustrate how Black can do well (hence the use of ‘e.g.’ or ‘for example’). That is still true, but this time I have attempted to be clearer with regard to lines that are truly balanced, not by writing some sort of objective thesis but by indicating when the French player is looking to dig up chances in an equal position, rather than trying to prove any theoretical advantage. In other words, I have tried not to claim too much. In a related way, I’ve often undertaken to identify who has the easier position to play in a theoretically equal situation. It’s difficult these days not to over-rely upon computer engines for analysis, and I’ve undoubtedly been guilty of this in some cases. But I have also found that engines are capable of systematic bias in certain pawn structures, as well as with some unusual piece configurations. So when you use a computer to analyse the material from this and other books, be a little careful. If you disagree with my or the computer’s assessment of a position (or are simply curious), don’t just set up the position and let the engine run. Instead, suggest moves for the side you think has a better position than the computer believes. Give the engine a lot of time to respond, and keep playing against the engine with the moves that you think correct, until either you change your assessment or it does. Although the machine will probably prove correct more often than not, you’ll be surprised how often it defers to your judgment. Finally, I want to thank just a few of the many people who have made suggestions or otherwise contributed to this book, in particular John Hartmann, Franz Steenbekkers, and Joachim Wintzer. Thanks also to the many contributors to the ChessPublishing Forum, not to mention its columnists Neil McDonald and Goh Wei Ming, and of course the owner and webmaster Tony Kosten. A special nod to Viktor Moskalenko for writing the most inspiring French books in recent years, easily the biggest influence on this edition. The publicationsKaissiber(Stefan Bücker, editor) andSecrets of Opening Surprises(Jeroen Bosch) are remarkable for their contributions to the theory of lesser-known lines. Most importantly, thanks to the insights and hard work of my editor Jonathan Tait.
Chapter One
Advance
Variation:
Introduction
and
Traditional
Lines
1 e4 e6 2 d4 d5 3 e5
The Advance Variation was initially one of the most popular responses to the French Defence, and today 3 e5 is still a major system, not much behind 3 Nd2 in mass appeal (whereas 3 Nc3 is very much the leading choice). With 3 e5, White establishes a wedge in the centre which serves two major functions: 1. It hinders the natural development of the opponent’s pieces (for example, Black’s knight can’t go to f6 and his bishop is denied access to d6); 2. It stakes out territory on the kingside, where White’s pieces will have more room to move about. Normally in a pawn-chain structure like this, White would have the goal of attacking on the kingside with f4-f5; but in the Advance Variation he seldom achieves that, mainly because Nf3 must be played early on, blocking the f-pawn. Nevertheless, restricting Black’s movement on that side of the board can assist in attacking there, and you will also see White expand on the queenside (by a3 and b4) to further deny Black room to manoeuvre. The outlines of Black’s conventional opening strategy are fairly simple to describe. He wishes to break up White’s central pawns, usually with the moves ... c5 and ... f6. After 3 ... c5, White will usually support the pawn by 4 c3 (be sure to take note the important exceptions which follow). This leads to one of two situations: (a) Black exchanges pawns on d4, then develops so as to put pressure upon the d-pawn (with knights on c6 and f5, for example, and/or a queen on b6). At some point Black tries to break down White’s centre by means of ... f6 (trying to establish a central majority, which can be a powerful asset); or exchanges on e5 and targets the remaining pawn there with pieces. If White successfully defends the pawn on d4, Black’s pressure on it might serve to tie down White’s defending pieces so that other attacks such as ... g5-g4 succeed. (b) Black pressures but does not capture on d4, trying to force White to surrender his control by dxc5 or, again, to distract White’s forces so that an alternate attack is successful. Occasionally, the move ... c4 can be useful. Obviously, since 3 e5 has been used in tens of thousands of master games, the implementation of these basic ideas is enormously complex, and actual play is full of exceptions to these strategies, as well as alternate approaches. Nevertheless, to be successful as Black, you will almost always do well to concentrate upon breaking up the d4/e5 pawn centre. 3 ... c5 Here White has four primary approaches, the last of which dominates master practice: 1.1 1.2 1.3 1.4
4 4 4 4
dxc5 Qg4 Nf3 c3
I don’t want to spend excessive time on other alternatives to 4 c3, which have the serious drawback of letting Black break up White’s pawn chain. You should know that they all allow for multiple solutions by Black, so if you’re not satisfied with my proposed remedies, others can be found, either in the literature or by means of your own analysis: (a) 4 b4?! is a modified Wing Gambit (compare Chapter 17) 4 ... cxd4 5 a3 and now: (a1) 5 ... a5 is a good solution; e.g. 6 b5 Nd7 7 Nf3 (7 f4 Nh6 8 Nf3 Nf5 9 Bd3 Ne3 and ... Bc5) 7 ... Qc7 8 Bf4 Ne7 9 Be2 Ng6 10 Bg3 Bc5-/+ Cherkasov-Nikolenko, Czestochowa 1992.
(a2) 5 ... Nc6 6 f4 (6 Nf3 Qc7 7 Bf4 Nge7 intends ... Ng6, when 8 b5 Ng6 is advantageous, but 8 ... g5! is at least as strong) 6 ... Nh6 (or 6 ... a5 7 b5 Nce7 8 Nf3 Nf5 9 Bd3 Ngh6=/+ intending ... Bc5 and ... Ne3, Frenkel-Rao, Boston 1988) 7 Nf3 a5 8 b5 Ne7 9 Nxd4 Nef5-/+ with multiple ideas, including ... Qb6 and ... Bc5. (b) 4 f4 shores up e5,
but Black takes a large lead in development and activity following 4 ... Nc6 (or 4 ... cxd4 5 Nf3 Nc6 with a promising game after 6 Nd4 Qb6 or 6 Nbd2 Nh6 7 Bd3 Ng4! 8 Nb3 Bb4+) 5 c3 (equivalent to 4 c3 Nc6 5 f4) 5 ... Qb6 6 Nf3 Nh6 7 b3 (7 Qb3 cxd4 8 Qxb6 axb6 9 cxd4 Nf5 10 Bb5 Bd7 11 Bxc6 bxc6 of Pulnikov-Iljin, Togliatty 2001, leaves Black with a pair of bishops versus White’s very bad one; while 7 Bd3 Bd7 8 Qb3 is well met by 8 ... Qc7!, threatening ... c4, when 9 Qc2 cxd4 10 cxd4 Qb6! wins material) 7 ... cxd4 8 cxd4 Bb4+ 9 Kf2 Ng4+! 10 Kg3 (not 10 Kg1?! Be1! with the idea 11 Qxe1? Nxd4 12 Be3 Nxf3+ 13 gxf3 Nxe3–+) 10 ... h5 11 h3? (but 11 Bd3 Bd7 and 11 Be2 Nh6 clearly favour Black) 11 ... h4+! 12 Nxh4 (12 Kxg4 f5+ 13 exf6 e5+) 12 ... Rxh4! 13 Be2 (13 Kxh4 Nf2 14 Qh5 Be7+ 15 Kg3 Nxh1+ 16 Kh2 Qxd4) 13 ... Nf2! 14 Kxf2 Nxd4 15 Be3 Bc5 and White has no satisfactory move. (c) 4 Bb5+ develops, but trading White’s ‘good’ bishop for Black’s ‘bad’ one is generally frowned upon: 4 ... Bd7 (or 4 ... Nc6; e.g. 5 Nf3 cxd4 6 Qxd4 Ne7) 5 Bxd7+ Nxd7 6 Nf3 (6 c3 Qb6 7 Nf3 Ne7 8 0-0 Nc6 leaves White on the defensive) 6 ... cxd4 (6 ... Ne7) 7 Qxd4 Ne7 8 0-0 Nc6 9 Qa4 g6!? 10 Bf4 Bg7 11 Re1 0-0 12 Na3 (12 Qb5 f6! 13 exf6 Qxf6) 12 ... f6! 13 exf6 Qxf6 14 Bd6 Rf7 and Black intends 15 ... e5, even after 15 Rad1. (d) 4 Be3 cxd4 (4 ... Nc6 5 c3 transposes to 4 c3 Nc6 5 Be3) 5 Bxd4 leaves White’s bishop subject to attack: 5 ... Nc6 6 Nf3 Nge7 7 c3 (or 7 Bb5 Nf5 8 0-0 Bd7 9 Bxc6 bxc6 10 b3 c5 11 Bb2 Be7-/+ with ... 0-0 and ... Bc6 cementing the bishop-pair advantage) 7 ... Nf5 8 Bd3 Nfxd4 9 cxd4 Qb6 10 Qd2 Bd7 11 Be2, and one instructive sequence (because it comes up so often) is 11 ... g5!? (11 ... Be7 12 Nc3 f6=/+ is calmer and objectively preferable) 12 h3 Nxd4! 13 Nxd4 Bc5 14 Nb3! Bxf2+ 15 Kd1 (15 Kf1 Bg3 16 Bd3 Rf8!) 15 ... a5! with a powerful attack and sufficient compensation for the piece. 1.1
4
dxc5
Rather than defend d4, White shifts his attention to bolstering e5 with pieces, so that after ... f6, he will be able to occupy the outpost on e5. This capture is typical in several lines of the French, but in this case ceding d4 so quickly seems too co-operative. Black gains a small lead in development and will target e5 next. Although the variation isn’t that important, the way in which it develops is quite instructive. 4 ... Nc6 Avoiding 4 ... Bxc5 5 Qg4, when 5 ... Ne7 6 Qxg7 Ng6 allows White to draw by 7 Bd3 Bf8 8 Qf6 Be7 9 Qg7 Bf8=. 5 Nf3! 5 Bf4?! Bxc5 is already miserable for White, who not only needs to defend b2 and f2 against ... Qb6, but can’t prevent the dissolution of his centre: (a) 6 c3 f6 (easiest, although 6 ... Nge7 7 Nf3 Ng6 8 Bg3 f6! 9 exf6 Qxf6 is also strong; e.g. 10 Bd3 Nf4 11 0-0 0-0 12 Bh4 Qh6 13 Bg5 Qh5 14 Bxf4 Rxf4) 7 Nf3 (7 Bb5 Bxf2+!) 7 ... Qb6 (or 7 ... fxe5 8 Bxe5 Nf6) 8 Qe2 g5!? (8 ... fxe5 9 Nxe5 Nf6=/+) 9 Bg3?! (9 Bc1 g4 10 Nfd2 Nxe5) 9 ... g4 10 Nh4 Nxe5 11 Bxe5? fxe5 12 b4 Bd6 13 Qxg4 Nf6 14 Qe2 0-0 and White is lost. (b) 6 Qg4 Nge7! (6 ... Qb6 7 Qxg7 Bxf2+ 8 Kd1 Qxb2 9 Qxh8 Nce7) intending 7 Qxg7? (7 Bd3? Qb6; 7 Nc3 Ng6 8 Qg3 0-0 9 Nf3 Nxf4 10 Qxf4 Qb6-/+) 7 ... Rg8 8 Qxh7 Qb6 and wins. 5 ... Bxc5 6 Bd3 White needs to get castled. (a) 6 Bf4?! Qb6 7 Bg3 Qxb2 8 Nbd2 Nge7 leaves White with no compensation for the pawn. (b) 6 a3 is slow after 6 ... f6 (or 6 ... Qb6 with the idea 7 Qe2 f6 8 b4 Be7 or 7 Qd2 Nd4!=/+) 7 b4 Bb6! 8 b5!? (8 exf6 Nxf6 9 Be2 0-0 10 0-0 e5! 11 b5 Nd4-/+) 8 ... Nxe5 9 Nxe5 fxe5 10 Qh5+ Kf8 11 Qxe5 Qf6 (or 11 ... Bc7 12 Qe3 Qf6 13 Ra2 Ne7 14 Bb2 e5=/+) 12 Qxf6+ Nxf6 13 Bb2 e5! 14 Bxe5 Kf7 with an attack; e.g. 15 Bxf6 Re8+ 16 Be2 Kxf6. 6 ... f6 The clearest course, since it liquidates the centre. For those wishing for a game of a more closed nature, 6 ... Nge7 is a sound alternative.
Briefly: (a) 7 0-0 Ng6 8 Qe2 (8 Re1 0-0 9 c3 f6) 8 ... 0-0 9 c3 Bd7 (or 9 ... Qc7 10 Re1 f6) 10 Re1 (10 Be3 Be7!) 10 ... f6 11 Bxg6 hxg6 12 exf6 gxf6 13 Be3 Bxe3 14 Qxe3 e5!=/+ intending 15 Qh6 Ne7 16 c4 Rf7! 17 cxd5 Nxd5 18 Qxg6+ Rg7 19 Qc2 Nf4 20 g3 Bc6 21 Nbd2 Qd7 with a strong attack. (b) 7 Nc3 Ng6 8 Bxg6 fxg6!? 9 0-0 0-0 yields a common structure in which the bishops and f-file co-ordinate; e.g. 10 a3 Rf5 11 b4 Bb6 12 Re1 Bd7 13 Bb2 g5! 14 h3 h5 with the initiative. (c) 7 Bf4 Qb6 (7 ... Ng6 is also logical) 8 0-0 Ng6 9 Bg3 (Keres suggests 9 Qc1 Nxf4 10 Qxf4, when 10 ... Qxb2 11 Nbd2 f5! is good) 9 ... Qxb2 10 Nbd2 Ngxe5! 11 Nxe5 Nxe5 12 Rb1 Qc3 13 Rb3 Qd4 14 Bxe5 (14 Bb5+ Nc6-/+) 14 ... Qxe5 15 Bb5+, T.Johansson-E.Berg, Bergen 2001, and 15 ... Ke7 was fine, but 15 ... Kf8! is safer; e.g. 16 Re1 Qc7 17 Qh5 g6 18 Qh6+ Kg8 and Black has a decisive advantage with two extra pawns, the bishop pair, and mobile central pawns. Returning to our main line with 6 ... f6:
7 Qe2! 7 Bb5? Qb6 intending 8 Qe2 Bxf2+ is weak, and the others are fun for Black: (a) 7 exf6?! Nxf6 8 0-0 0-0 9 c4 (9 Re1 e5!; 9 Qe2 e5! 10 Nxe5 Nxe5 11 Qxe5 Ng4 12 Qh5 g6-/+) 9 ... Qd6 (9 ... e5!? 10 cxd5 e4 11 dxc6 Qxd3 12 Qxd3 exd3=/+ Van Scheltinga-Van der Tol, Netherlands 1946) 10 Nc3 a6=/+ 11 Bg5? Ng4! intending ... Rxf3, Camilleri-Debarnot, Malta Olympiad 1980. (b) 7 Bf4? is also weak: 7 ... fxe5 (or (7 ... Qb6) 8 Bxe5 (8 Nxe5? Qf6-/+; e.g. 9 Nxc6 Qxf4 10 Qh5+ g6 11 Bxg6+ Kf8 12 Ne5 Nf6 13 Qe2 Qc1+ 14 Qd1 Qxd1+ 15 Kxd1 Bd4–+ Rohel-Keitlinghaus, Badenweiler 1990) 8 ... Nxe5 9 Nxe5 Qg5 10 Bb5+ Ke7 11
Nf3 Qxg2 12 Rf1 Nf6 with an extra pawn. 7 ... fxe5 7 ... Qc7!? 8 Bf4 g5!? 9 Bg3 g4 is a dynamic alternative:
Black wins a key central pawn at the cost of exposing his king; for example, 10 Nfd2! (10 Nh4 fxe5 11 Qxg4 Nf6 12 Qg5 Qf7 favours Black’s mobile centre) 10 ... Nxe5 11 Nc3 (11 Bb5+ Kf7!) 11 ... Bd7 12 0-0-0 (12 h3 gxh3 13 Rxh3 0-0-0) 12 ... Nxd3+ 13 Qxd3 Qc6 with mutual chances. 8 Nxe5 You’ll see a variety of positions like this in the French. White endeavours to blockade the centre and doom Black’s lightsquared bishop to passivity, but Black’s central majority is a plus, and his active piece play can thwart White’s plans.
8 ... Nxe5 Leading to a degree of simplification, but there remains an imbalance in the position; in particular, White won’t be able to maintain his control of e5. Instead, 8 ... Nf6 can be played for simplification, or even for a win if Black is willing to take some risks. Black threatens ... Nxe5 and ... Bxf2+; e.g. 9 Bf4 (9 0-0 Nxe5 10 Qxe5 transposes to the main line; 9 Nxc6 bxc6 10 Nd2 0-0 11 Nf3 can also be met by 11 ... e5! 12 Nxe5 Re8 intending 13 0-0 Bg4 or 13 h3 Qd6) 9 ... 0-0 (9 ... Qb6!? is more ambitious and practically forces White to gambit; for example, 10 0-0 Qxb2 11 Nxc6 bxc6 12 Nd2 Qa3 13 Nf3 0-0 14 Be5 with some compensation for the pawn) 10 0-0 Qe8!? (intending ... Nh5 or in some cases ... Qh5; otherwise 10 ... Bd4 equalizes following 11 Nc3! Ne4 or 11 ... Bd7; while 10 ... Bb6 and 10 ... Qe7 are good alternatives) 11 Nxc6 (11 Nd2 Nd7! 12 Nxd7 Bxd7 13 Bg3 e5) 11 ... bxc6 12 Nd2 Nd7!? 13 Bg3 e5 14
Rae1 Qf7 with a dynamically balanced game, whereas 14 ... e4 15 Nxe4 dxe4 16 Bxe4 Kh8 is obscure. 9 Qxe5
9 ... Nf6 9 ... Qf6 tries to get to an ending that will be favourable by virtue of Black’s central majority. White can probably hold the balance with 10 Bb5+ (10 Qh5+ g6 11 Bb5+ Kf7 transposes to 10 Bb5+; instead, 10 Qc7?! is speculative after 10 ... Qxf2+ 11 Kd1 Bd7! 12 Rf1 Rc8; 10 Qxf6 Nxf6 threatens ... Ng4 and ... e5; 10 Bf4 Ne7 11 Qxf6 gxf6 intending ... e5 is at least equal; 10 Qg3 Ne7 11 Nc3 0-0 12 0-0 Nf5 13 Qh3, Heidsiek-T.Schmidt, corr 1988, when 13 ... Bd7 is solid and good) 10 ... Kf7! 11 Qh5+! (11 Qg3 Ne7 12 0-0 Rd8 intending ... Bd6; 11 Qe2 Ne7 12 Nd2 Rf8 13 Nf3 Kg8) 11 ... g6 12 Qe2 Ne7 13 Nd2 a6 14 Bd3 Nc6=. 10 0-0 White can’t be happy with 10 Bf4? Bxf2+ or 10 Bb5+?! Kf7, when ... Qb6 and/or ... Rf8, and ... Kg8 will follow; for example, 11 0-0 Qb6 12 Be2 Rf8 13 Nd2 Kg8 14 Qg3 e5!. 10 ... 0-0 Or 10 ... Bd6 11 Qg5 0-0 12 Nc3 e5 13 f3 Be6 with a central advantage. 11 c4 Ng4! Or 11 ... Qb6, again threatening ... Bxf2+! and answering 12 Qe2 with 12 ... e5!, threatening both ... e4 and ... Ng4. 12 Qh5 Rxf2 13 Rxf2 Bxf2+ 14 Kh1 Nf6 15 Qe2 Bb6 And White has some, but not fully sufficient, compensation for the pawn. 1.2
4
Qg4
Nimzowitsch’s radical idea. White abandons d4 and intends to overprotect e5 by means of, for example, Nf3, Qg3, and later Bf4 and/or Re1. Bringing the queen out before the other pieces is not necessarily bad; but if she loses more time and fails to create weaknesses in the enemy camp, as happens here, such a foray is unlikely to succeed. For my main line, I will show a less tested and generally simpler response that solves Black’s problems and creates ones for White. It was played by Botvinnik as far back as 1937 and has been revived to good effect in recent times: 4 ... Qa5+ Meeting a queen move with a queen move! Black’s idea is to be sure that White’s queen will lose time to ... Nh6 by diverting the c1-bishop from a potential capture on h6. At the right moment ... f5 can also be useful. Alternatively, 4 ... Nc6 affords Black several routes to equality and apparently some advantage. I’ll just indicate a few of several satisfactory approaches for Black (see alsoPTF3): 5 Nf3 cxd4 (note that here 5 ... Qa5+ 6 Bd2 Qb6 attacks b2, and otherwise 6 c3 cxd4 transposes to our 4 ... Qa5+ main line; finally, Moskalenko speaks favourably of 5 ... Nge7 6 c3 Nf5, with the main line 7 Bd3 cxd4 8 0-0 Qb6 or 8 ... g6, and here 6 ... cxd4 7 cxd4 Nf5 may be better still) 6 Bd3 and now two systems are worth mentioning: (a) 6 ... Qc7
This was analysed in detail in the previous editions, and you can employ databases to flesh out a repertoire. Very generally, this is how it breaks down: (a1) Moskalenko recommends 7 Bf4 for White, perhaps correctly, although I don’t think that Black has anything to fear.
As outlined inPTF3, the reply 7 ... Nb4! poses White the most problems. Briefly: 8 0-0 (8 Na3?! Nxd3+ 9 cxd3 Qb6 10 Bc1 Nh6! 11 Qxd4 Bc5 12 Qa4+ Bd7 13 Qc2 Bb4+!? 14 Bd2 Rc8, Cornelison-Dirr, email 1994, when 15 Qb3! Bxd2+ 16 Nxd2 avoids disaster, but 16 ... Qa5 retains a solid advantage; 8 Nxd4?! Nxd3+ 9 cxd3 Qb6 10 Nb3 Qb4+, and now not 11 N1d2?? g5!, winning a piece, but 11 Ke2 Ne7 12 h3! Nc6 13 Nc3 Bd7=/+) 8 ... Nxd3 9 cxd3 Qc2 10 Nxd4 Qxb2 11 Nb3, and myPTF3recommendation was 11 ... h5! (somewhat more accurate than Moskalenko’s 11 ... Bd7, I think), when the critical line is 12 Qf3! Bd7 13 Rc1 Rc8 14 Rxc8+ Bxc8 15 Nc3 Bd7; e.g. 16 Bd2 Qa3 17 Rb1 Bb4 18 Qg3 Ne7 19 Qxg7 Rg8 20 Qh7 Bxc3 21 Bxc3 d4!? 22 Bxd4 Bc6 23 g3 Qxa2 24 Rb2 Qa4 intending ... Bd5 with good prospects. (a2) 7 Qg3 f6! 8 exf6 (8 Bxh7?! Nxe5 9 Bg6+ Kd7!) 8 ... Qxg3 9 f7+ Kxf7 10 hxg3 Nf6=/+; e.g. 11 Ng5+ Kg8 12 f4 h6 13 Nf3 Ne4. (a3) 7 0-0 f6!? (Black can also play for a win with 7 ... Nxe5 8 Nxe5 Qxe5 9 Bf4 Nf6 10 Qg3 Qh5) 8 Bxh7 (8 Bf4 fxe5, and instead of 9 Nxe5?! Nf6!, White can virtually equalize after 9 Qh5+ Qf7 10 Qxf7+ Kxf7 11 Nxe5+ Nxe5 12 Bxe5 Ne7 with the idea of ... Nc6 and ... Bc5) 8 ... Nxe5 9 Nxe5 fxe5 10 Bg6+ (10 Re1 Nf6 11 Bg6+ Kd7!) 10 ... Kd8 11 h3 (11 Qg3 Bd6) 11 ... Nf6 12 Qd1, Smyslov-Lisitsin, Moscow 1942, and after 12 ... Bc5!=/+, Black threatens to mobilize with ... e4, whereas 13 Nd2 Bd7 14 Nb3 Bd6 15 Bd2 Be8 doesn’t help White. (b) Another traditional line that is looking strong is 6 ... Nge7 7 0-0 Ng6 8 Re1 Qc7 9 Qg3 f6! (a standard theme in this variation; it’s worth giving the pawn back to exchange queens) 10 Bxg6+ hxg6 11 Qxg6+ (11 exf6 Qxg3 12 f7+ Kxf7 13 hxg3 Be7 14 Bf4 Bf6-/+ – Moskalenko) 11 ... Qf7 12 Qxf7+ (12 Qg3 Bd7! 13 c3 Bc5=/+ – Moskalenko; and here 12 ... Bc5 was not bad either) 12 ... Kxf7 13 Bf4? (13 Nbd2 Rh5!) 13 ... g5 14 Bg3 g4 15 Nfd2 fxe5–+.
5 c3 The most enterprising reply. 4 ... Qa5+ is a pragmatic move that counts upon the drawbacks of each response: (a) 5 Nd2?! Nc6 (5 ... Nh6 is also good) 6 Ngf3 (6 dxc5 Qxc5 hits c2 and e5) 6 ... Nh6
This exploits the fact that the c1-bishop no longer covers h6; the knight will often come to f5 (an alternative is 6 ... Nb4! 7 Kd1 c4! 8 a3 Bd7 – compare what follows): 7 Qf4 (7 Qg5 Nb4) 7 ... Nb4! (forcing the king to move) 8 Kd1 (8 Nb3? Qa4) 8 ... c4 (now White has to defend against attacks on c2) 9 c3 (9 a3 Bd7 10 b3 Rc8!; 9 Nb1?! Nxa2! and White has no way to exploit the pin, while Black threatens to play ... Nc3+) 9 ... Nc6 (intending ... b5-b4; 9 ... Qa4+ 10 b3 cxb3 11 Nxb3 Nxa2 12 Kc2 Bd7 13 Nfd2 Rc8 gives Black a small advantage after complications) 10 Be2 b5 11 Ke1 b4 12 Nb1 Be7-/+ intending ... 0-0 and ... f6 in combination with ... Rb8. (b) 5 Bd2 is objectively best: 5 ... Nh6
(5 ... Qb6 6 Nc3! Nh6 7 dxc5! Nxg4 8 cxb6 Bc5 9 f4 Bf2+ 10 Ke2 Bxb6 11 h3 Nf2 12 Rh2 with approximate equality) 6 Qxe6+ (6 Bxa5 Nxg4 7 h3 Nh6 8 Nf3 Nc6 9 Bc3 Nf5=/+; 6 Qg3 Qb6 7 Qb3 Nc6 8 Nf3 cxd4 9 Qxb6 axb6=/+) 6 ... fxe6 7 Bxa5 Bd7!? (7 ... Nf5!? 8 dxc5 Nc6 9 Bc3! Bxc5 10 Nf3 0-0) 8 Nf3 Nc6 9 Bc3 Rc8=/+. These 5 Bd2 lines are close to equal, but Black has slightly better chances. 5 ... cxd4 6 Nf3!? Gambiting a pawn may be best. The alternative 6 Qxd4 Nc6 is easy for Black to play. The stem game for 4 ... Qa5+ is Rabinovich-Botvinnik, Moscow 1937, which went 7 Qf4 Nge7 8 Bd3 Ng6 (here 8 ... d4 9 Nf3 Nd5 is also interesting) 9 Bxg6 hxg6 10 Nf3 Qa6 (Moskalenko suggests the thematic 10 ... Qc7 11 0-0 f6!; then 12 Qg3 Nxe5 13 Nxe5 Qxe5 14 Qxg6+ Ke7 would win the centre and intend ... Qh5) 11 Ng5 Nd8 12 Nd2 Qd3!? 13 Nb3, and while 13 ... b6 was mildly advantageous, 13 ... Be7! with the idea of ... f6, and in some cases ... Rh5 looks better still; e.g. 14 Qd4 Qxd4 15 Nxd4 f6 16 Ngf3 Nf7 17 Nb5 Kd7, after which Black will soon achieve a mobile central majority to go with his bishops.
6 ... Nc6 7 Nxd4 7 Bd3 is poorly timed due to 7 ... dxc3 (or 7 ... f5!? 8 Qg3 dxc3 9 bxc3 Nh6!) 8 Nxc3 (8 bxc3 d4 or simply 8 ... b6 9 0-0 Ba6) 8 ... f5! (8 ... d4!? 9 Nxd4 Qxe5+ 10 Be3 Nf6 is also good, if less clear) 9 Qf4 Nge7 10 0-0 Ng6 11 Qg3 Be7 and it’s hard to see White’s compensation for the pawn. 7 ... Qc7 I analysed 7 ... Nxe5!? 8 Qg3 f6 and 8 ... Ng6 inPTF3, but the text is more straightforward. 8 Bf4!? 8 Nxc6 bxc6 strengthens Black’s centre; e.g. 9 Qg3 Ne7 10 Bd3 Ng6 11 f4 Be7 12 0-0 Qb6+ 13 Kh1 Ba6. 8 ... Nxd4 8 ... Nh6 is also worth considering, since after 9 Bxh6 gxh6 10 Nxc6 bxc6 11 Qg3, then 11 ... Rb8 12 b3 f6! 13 exf6 Bd6 grabs the initiative; e.g. 14 Qh4 0-0 15 Qxh6 Bf4 16 Qg7+ Qxg7 17 fxg7 Kxg7 followed by ... e5. After 8 ... Nxd4 9 cxd4, Black again grabs the initiative with 9 ... Qb6! (hitting b2 and d4) 10 Nc3 (if 10 Bc1, 10 ... Nh6 11 Qf4 Nf5 wins the d-pawn) 10 ... Bd7 11 0-0-0 Rc8 with queenside pressure ( ... Ba3 is a threat) and easy development to follow. 1.3
4
4 ... Nc6 4 ... cxd4 5 Bd3 Nc6 transposes.
Nf3
5 Bd3 5 c3 is 1.4 (the main line), and 5 dxc5 transposes to 4 dxc5 (1.1). With the text move, White sacrifices the d4-pawn in order to shore up e5 with pieces. This was also his philosophy with 4 dxc5 and 4 Qg4, but one could argue that Nf3 and Bd3 are the optimal combination if quick development is the goal. 5 Bb5 is generally a mediocre move in Advance Variation structures (barring White having an outpost on e5), because it is White’s good bishop that will be exchanged if Bxc6 follows, and in most cases that strengthens Black’s centre after ... bxc6. This position is no exception, and Black has more than one way to try for a minor positional advantage, including 5 ... Qb6 and: (a) 5 ... Bd7 (threatening 6 ... Nxe5) 6 Bxc6 bxc6 7 0-0 cxd4 8 Qxd4 c5 9 Qg4 Ne7 (or 9 ... h5) 10 Nc3 h5 11 Qh3 Qb8 12 Rb1 Nf5=/+. (b) 5 ... cxd4 6 Qxd4 (6 Nxd4 Nge7 7 0-0 Bd7 8 Re1 Nxd4 9 Bxd7+ Qxd7 10 Qxd4 Nf5 11 Qd3 Bc5=/+; 6 0-0 Nge7) 6 ... Nge7 7 0-0 Bd7 8 Bxc6 Nxc6 9 Qf4 Nb4 10 Na3 Rc8=/+. 5 ... cxd4 6 0-0 f6 I’m putting this as the leading move because it’s thematic and leads to relatively forced lines (if White is to avoid disadvantage), thus minimizing study time. 4 Nf3 Nc6 5 Bd3 cxd4 6 0-0 is a line against which Black has numerous good answers. In the two previous additions, I’ve suggested 6 ... g6, and 6 ... Bc5 has done respectably well over the years. Here are some thoughts about 6 ... Nge7, a well-regarded and practical solution:
(a) 7 Re1 Ng6 8 Nbd2 Qc7 9 Bxg6 fxg6!? (9 ... hxg6 10 Nb3 Be7 11 Nbxd4=; 10 ... f6!? 11 Bf4 g5 12 Bg3 Qb6 13 Qd3 Kf7 is complex) 10 Nb3 Bd7 11 Nbxd4 Nxd4 12 Nxd4 Bc5 13 c3 0-0 with ... Rf7 to come. (b) 7 Bf4 is by far the most popular continuation, because White wants to get Bf4-g3 in before ... Ng6 prevents him. There follows 7 ... Ng6 8 Bg3
and now 8 ... Bc5 and 8 ... Qc7 are known main lines, or Black can play with a twist: (b1) 8 ... Be7 9 Nbd2 is played in most games, and now a radical approach is 9 ... f5!? (9 ... Bd7 10 Nb3 Rc8= is less ambitious) 10 exf6! (Black is planning ... 0-0 and ... f4, while a cruel trick is 10 Re1?? f4! 11 Bxg6+ hxg6 12 Bxf4 g5 13 Bg3 g4 and White’s knight is trapped; 10 Qe2 avoids that, but 10 ... 0-0 11 h3 Bd7 12 Nb3 Qc7 contemplates ... f4 forcing favourable simplification: play might continue, for example, 13 Rfe1 Bb4 14 Rf1 f4 15 Bxg6 fxg3 16 Bd3 gxf2+ 17 Rxf2 Qb6 18 Raf1 h6 19 a3 Be7 20 Qd2 Rf7, when even ... Rxf3 followed by Nxe5 becomes a theme) 10 ... gxf6 (Black plays shamelessly for a central pawn roller) 11 Nh4!, Spraggett-M.Gurevich, Havana 1986. This is thought to be very strong for White, but 11 ... f5! maintains Black’s extra pawn and at least dynamic equality; for example, 12 Nxg6 hxg6 13 Bb5! (‘±’ Spraggett, but:) 13 ... 0-0 14 Bxc6 bxc6 15 Be5 (not 15 Nf3? f4) 15 ... c5 16 Nf3, and Black can allow a draw with 16 ... g5 17 Nxg5! Bxg5 18 Qh5 Bf6 19 Qg6+ Kh8 20 Qh6+ etc, or keep some pieces on with 16 ... Rf7!? intending ... g5-g4 and ... Rh7. (b2) Having seen that, the idea of 8 ... f5 9 exf6 gxf6 looks more effective here (than after 8 ... Be7 9 Nbd2) in the key line 10 Nh4!? (10 Nxd4 Nxd4 11 Bxg6+ hxg6 12 Qxd4 Kf7=/+; 10 Re1 Be7 11 Nh4 f5 12 Nxg6 hxg6 13 Bb5 Qb6) 10 ... f5! 11 Nxg6 (11 Qh5 Qf6 12 Nd2 Bd7=/+ and ... 0-0-0) 11 ... hxg6 12 Bb5 Bg7 13 Re1 0-0 14 Bxc6 bxc6 15 Be5 Bxe5 16 Rxe5 c5 17 Nd2 Rb8 18 b3 g5 with the idea 19 Qh5 Rb7.
7 Qe2 Instead: (a) 7 Re1? loses the centre to 7 ... fxe5 8 Nxe5 Nxe5 9 Rxe5 Nf6 (or 9 ... Bd6) and: (a1) 10 Bg5?! Bd6 11 Bb5+ Kf7 12 Re1 Qb6!? (12 ... h6 13 Bh4 a6 14 Be2 e5=/+) 13 Bxf6 gxf6 14 Qh5+ Kg7 15 Qg4+ Kf8 16
Bd3 Rg8 17 Qh4 Rg7 18 Qxf6+ Kg8 19 Nd2 Bd7 intending ... Rf8. (a2) 10 Bb5+ Kf7 11 Qxd4 Bd6 12 Re1 (12 Re3 e5 13 Qd1 Rf8) 12 ... Qc7 (12 ... a6 13 Bf1 Qc7 14 h3 e5=/+) 13 h3 Rf8=/+. (b) 7 Bf4? g5! (7 ... fxe5 8 Nxe5 Nf6 9 Qe2 transposes to 7 Qe2) 8 Bg3 g4 9 Nh4 (9 Nxd4!? Nxd4 10 Qxg4 Nc6 11 Qh5+ Kd7 12 Nc3 Qe8 and White has some attack, but not enough for a piece) 9 ... fxe5 (or 9 ... f5=/+ intending ... Be7 and ... f4) 10 Qxg4 Nf6-/+; Black has a pawn and huge centre, and intends moves like ... Bd6, ... Qb6, and ... e4. (c) 7 Bb5 is White’s safest move: it contests e5 but also costs a full tempo, so Black has no problems: 7 ... Bd7 (7 ... Qc7!? is an unusual move that keeps a more open position on the board; e.g., 8 Qxd4 fxe5 9 Qxe5 Bd6 10 Qh5+ g6 11 Qg5 Bd7, intending ... Nge7 and ... 0-0-0) 8 Bxc6! (8 Nxd4 Nxd4 9 Bxd7+ Qxd7 10 Qxd4 Rc8! with ideas of ... Rxc2, ... Bc5, ... fxe5, and/or ... Rc4) 8 ... bxc6 9 Qxd4
9 ... f5 (9 ... c5 and 10 Qf4 Qb8 or 10 Qh4 f5 11 Bg5 Be7 12 c4 h6 13 Qh5+ Kf8 14 Bxe7+ Nxe7=; 9 ... fxe5 is the most direct move: 10 Nxe5 Nf6 11 Bf4 Be7 12 Nd2 0-0=; or 9 ... Qb6 10 Qh4 h5!? with the idea of ... g5, when there might follow 11 h3 Kf7 12 Qf4 Qc7 13 Ng5+ Ke8=) 10 Qf4 Nh6 11 Ng5 Be7 12 Qg3 Nf7! 13 Nxf7 Kxf7 14 c4?! d4 15 Nd2 c5 16 Nf3 h6=/+ with two bishops and moves like ... Rb8, ... Bc6 and ... g5 in store, Gamback-Kiriakov, Skellefteå 2001. 7 ... Qc7 I use this as the main line because it is a dynamic way to play for a win and also because there is less theory than after 7 ... fxe5 8 Nxe5, when I’ve analysed both the safe 8 ... Nxe5 and the more complex 8 ... Nf6 in previous editions. The latter can lead to 9 Bf4 (9 Bg5 Nxe5 10 Qxe5 Bd6 11 Bb5+!? Kf7 12 Qxd4 Qc7=/+) 9 ... Bd6 10 Nd2 0-0 11 Ndf3.
Here Black has a few ways to play, among them:
(a) 11 ... Qe8!? 12 Bg3 (12 g3 Nh5 13 Bd2 Rxf3!? 14 Nxf3 e5 is a promising exchange sacrifice) 12 ... Nh5 13 Nxc6 Nxg3 14 fxg3 bxc6 15 Ne5 Rb8 16 Rxf8+ Bxf8 17 b3 g6 with the ideas ... Bd6, ... c5, and ... Rb7-c7. This is unclear. (b) 11 ... Nxe5 12 Bxe5 (12 Nxe5 Ne4 with the idea 13 Bxe4 Rxf4 14 Bd3 Qf6 15 Rae1 Qxe5 16 Qxe5 Bxe5 17 Rxe5 Bd7) 12 ... Ng4 13 Bxd6 Qxd6 14 Bxh7+ Kxh7 15 Ng5+ Kg8 16 Qxg4 e5 17 Qg3 (17 Qh5?! Bf5) 17 ... Bd7 18 Nf3 Rae8 19 Nxd4 Rf6=/+. (c) 11 ... Qc7 12 Bg3 (a cute miniature followed 12 Rae1 Ne4 13 Nxc6 Bxf4 14 Ncxd4 e5 15 Nb3 Bg4 16 Kh1?? Ng5 0-1 in Aagaard-N.McDonald, Budapest 1996) 12 ... Bd7=, or 12 ... g6 intending ... Nh5. 8 Bf4!? 8 Bb5 controls e5 but again at the cost of a tempo and activity. Black has an easy time of it after 8 ... fxe5 9 Nxe5 (9 Qxe5 Bd6 10 Qxd4 Nf6=/+) 9 ... Nf6 (9 ... Bd6!? 10 Qh5+ g6 11 Nxg6 Qf7 12 Nf4 Nf6=/+) 10 Nd2 (10 Bg5 Bd6 11 Bxf6 0-0! is a nice trick; if White undertakes to win material by 12 Bxg7 Qxg7 13 Nxc6 bxc6 14 Bxc6, then the two bishops, two open files and huge centre dominate following 14 ... Rb8; e.g. 15 Bb5 Rb6 16 Nd2 e5) 10 ... Bd6 11 f4 (11 Ndf3 0-0 and Black wins the centre while White hasn’t even won his pawn back) 11 ... 0-0 12 Ndf3 Ne4=/+ or 12 ... Nxe5 13 fxe5 Bc5. 8 ... g5! 9 Bg3 g4 10 Nh4 10 Nfd2 fxe5 11 Qxg4 (11 Bb5 Bg7 12 Qxg4 Nh6! 13 Qxd4 Nf5 14 Qc3 0-0 is good for Black) 11 ... Qf7 12 Bb5 e4 13 Nb3 Bg7 14 Qd1 Nge7 15 Nxd4 0-0-/+.
10 ... fxe5! A true central majority (four pawns to none!). 11 Qxg4 Or 11 Bb5 Bd6 12 Qxg4 Nf6 13 Qg5 Kf7-/+. 11 ... Nf6 12 Qe2 Bd6 13 Bb5 0-0 14 Nd2 e4 Black has a significant advantage, and 14 ... Na5 is also strong. As in so many lines of the Advance Variation, Black’s unopposed central pawn mass is the most important factor on the board. 1.4 4 This is easily White’s most popular choice; he supports d4 before committing his pieces. 4 ... Nc6
c3
5 Nf3 Here we have the main line of the Advance Variation. White has some irregular options at this point: (a) 5 Qg4? cxd4 6 cxd4 Qb6 7 Nf3 Nh6! 8 Qf4 (8 Bxh6? Qxb2) 8 ... Nf5-/+ and Black wins a pawn. (b) 5 f4 Qb6 (or 5 ... Nh6) 6 Nf3 Nh6 transposes to 4 f4. (c) 5 a3!? could transpose, after 5 ... Qb6 6 Nf3, to 6 a3 in the next chapter. Black has a few options such as 5 ... Nge7 and the thematic 5 ... f6, which gives chances for an advantage following 6 Nf3 fxe5 (or 6 ... Qc7 7 Bf4 Nge7 with the idea 8 Bg3 Nf5 9 exf6 Nxg3 10 fxg7!? Bxg7 11 hxg3 cxd4 12 cxd4 Nxd4! 13 Nxd4 Qe5+=/+) 7 Nxe5 (7 dxe5 Qc7 8 Bb5 Nge7 9 Bf4 Ng6 10 Bg3 c4!?=/+) 7 ... Nf6 (7 ... Nxe5 8 dxe5 Ne7=) 8 Bb5 Qb6 9 Qa4 Bd7 10 Nxd7 Nxd7 11 0-0 Bd6 12 Re1 0-0! with the idea 13 Rxe6 Qc7 (13 ... Rae8!?) 14 g3 cxd4 15 Qd1! Nc5 16 Re1 Qf7 17 f4 Rfe8=/+. (d) 5 Ne2 Qb6 (or 5 ... f6 6 f4 Nh6!) 6 g3 f6 7 exf6 (7 Nf4 Qc7) 7 ... Nxf6 8 Bg2, Neretljak-Brynell, Rodeby 1998; 8 ... Bd6 9 0-0 0-0 10 Qb3 Qa6!=/+. (e) 5 Be3 is Kupreichik’s move, the most appealing of these alternatives to 5 Nf3.
By getting the bishop into play, White is able to meet ... Qb6 with Qd2. On the other hand, the bishop is rather passive on e3 and Black can continue with his plan to break up the centre by various means: (e1) 5 ... Nh6 is a standard idea in the Advance Variation; here Bxh6 isn’t advisable as it constitutes a loss of time for White. Play can go 6 Bd3 (6 Nf3 Nf5 7 Bd3 Nxe3 8 fxe3 Qb6 9 Qd2 f6!?) 6 ... Qb6 7 Qd2 (this has also arisen via 5 ... Qb6 6 Qd2 Nh6 7 Bd3) 7 ... Ng4 8 Nf3 Nxe3 9 fxe3 Be7 10 0-0 0-0 11 Kh1 (11 Qc2 g6 12 Nbd2 Bd7 13 Rf2 Rac8), Kupreichik-Vasiljevic, Cetinje 1993; when 11 ... g6 was fine, but 11 ... f6 12 Qc2 (12 exf6 Rxf6) 12 ... f5! 13 Nbd2 Bd7 and ... Rac8 gives Black
immediate queenside pressure, whereas 14 Rg1 Rac8 15 g4?! is too slow after 15 ... cxd4 16 gxf5 exf5 17 exd4 Nb4 18 Qb1 Nxd3 19 Qxd3 f4!. (e2) Williams suggests 5 ... f6 6 exf6 (6 Nf3, intending 7 Bb5, can be met by 6 ... Qb6=) 6 ... Nxf6 7 dxc5 e5 with sufficient central play for the pawn. (e3) The normal move is 5 ... Qb6 6 Qd2 with a variety of ideas:
(e31) 6 ... cxd4 7 cxd4 Bd7 8 Nc3 Rc8 (8 ... Bb4!? 9 Nf3 Rc8) 9 Nf3 Na5 is a simple solution; e.g. 10 Rc1 Bb4 11 Bd3 Nc4 12 Bxc4 Rxc4 with a comfortable game, Starostits-Glek, Baden-Baden 2001. (e32) 6 ... Bd7 7 Nf3 Rc8 (or 7 ... f6; e.g. 8 Bd3 fxe5 9 Nxe5 Nxe5 10 dxe5 Nh6!? 11 0-0 Nf7 12 f4 g5! unclear) 8 Bd3?! (play is level after 8 dxc5 Bxc5 9 Bxc5 Qxc5; or 8 Be2 cxd4 9 cxd4 Bb4 10 Nc3 Na5) 8 ... cxd4 9 cxd4?? (9 Bxd4 Nxd4 10 Nxd4 f6!) 9 ... Nb4 10 b3? (there’s no good defence) 10 ... Rc1+! (W.Conrad). (e33) 6 ... f6 7 Nf3 (Black left his king in the centre but won all the key squares after 7 f4 Nh6!? 8 exf6 gxf6 9 Be2 Nf5! 10 Bh5+ Kd8, Filimonov-Shorokhov, Smolensk 2001) 7 ... Bd7 (7 ... fxe5 8 Nxe5 Nf6 9 Bd3?! Nxe5 10 dxe5 Ng4) 8 Bd3 fxe5 9 dxe5 Nh6 10 0-0 Nf7 11 Bf4 Be7 12 Re1 0-0-0=/+ Westerinen-Ulibin, Benidorm 1993. (f) 5 g3!?, Murey’s move, has the original idea of Bh3 to exchange a knight that might get to f5. As far as I can tell, 5 ... Qa5! hasn’t been played at all, but it looks awfully good, because it’s hard to prevent ... cxd4 (6 Bd2 Qb6), and 6 dxc5 Qc7=/+ can’t appeal to White. So 6 Nf3 may be as good as anything, with a speculative gambit following 6 ... cxd4 7 Bd2, or 6 ... Nxd4 7 Nxd4 cxd4 8 Qxd4 Ne7, and Black’s advantage is limited. Nevertheless, is 5 ... Qa5 a practical refutation of 5 g3 – Black shouldn’t have an advantage on move 5 of any opening! If need be, 5 ... Qb6 6 Ne2 f6 and 6 ... Nge7 are also fine. 5 ... Qb6
This is the traditional and most direct move, putting pressure on d4 and, by eyeing b2, preventing White from moving his queen’s bishop just yet. In this chapter we will look at everything except the main line 6 a3, which is the subject of the next chapter: 1.41 1.42
6 6
Be2 Bd3
Others are almost never seen, but 6 Na3 is of interest: (a) 6 b3? cxd4 7 cxd4 Bb4+ 8 Ke2 Bd7 9 a3 Be7-/+. (b) 6 g3 f6 (6 ... Bd7) 7 Bg2 (7 Bh3 cxd4 8 cxd4 Bb4+ 9 Bd2 Bxd2+ 10 Qxd2 fxe5 11 dxe5 Nge7 12 0-0 0-0=/+ Hazai-Farago, Hungarian Ch. 1979; 7 Nh4 g6!; 7 Bd3 fxe5 8 dxe5 Nh6=) 7 ... cxd4 8 cxd4 fxe5 9 dxe5 Nh6 10 0-0 Nf7 11 Nc3 Be7 12 Qe2 0-0=. (c) 6 dxc5?! Bxc5 7 Qc2 (or 7 Qe2 Qc7 8 b4 Bb6 with the idea 9 a4 Nd4! 10 Nxd4 Bxd4 11 cxd4 Qxc1+ 12 Qd1 Qb2 13 Bb5+ Bd7 14 Bxd7+ Kxd7=/+) 7 ... Qc7 8 Bf4 Nge7 9 Bd3, Okhotnik-Zaid, Kiev 1986; here after 9 ... f6!, White can only protect e5 by 10 Qe2 (10 Bxh7 Nxe5 11 Bxe5 fxe5; 10 Bg3 Nxe5 11 Nxe5 fxe5 12 0-0 Bd7 13 Re1 Bd6-/+) 10 ... fxe5 11 Bxe5 Nxe5 12 Qxe5 Qxe5+ 13 Nxe5 0-0 14 0-0 Nc6 15 Nxc6 bxc6 with ... e5 soon to come. (d) 6 Na3 has been played by the creative Zvjaginsev and seems a good choice if one is seeking original positions.
Then 6 ... Bd7 7 Nc2 f6 is thematic, as is the immediate 6 ... f6; e.g. 7 Bb5!? (7 Qb3 Qa5 8 Qb5 Qd8!) 7 ... cxd4 8 Nxd4 Qc7
9 Bf4 a6 10 Bxc6+ bxc6 11 Qh5+ g6 12 Qe2 c5 13 Nf3 Bh6!?=, or here 13 ... g5 14 Bg3 g4 15 Nd2 f5!. These ... f6 plans look satisfactory, but by far the most popular choice has been 6 ... cxd4 7 cxd4 Bb4+ 8 Bd2 and now, instead of 8 ... Nxd4?? 9 Nxd4 Qxd4 10 Qa4+, Black has options:
(d1) 8 ... Nge7 9 Nb5 (9 Nc2 Bxd2+ 10 Qxd2 Qxb2 improves slightly over ‘d2’ below; 9 Bc3 0-0 10 Bd3 f6! has the idea 11 exf6 gxf6) 9 ... Bxd2+ 10 Qxd2 0-0 11 Be2 f6 12 exf6 Rxf6= Haskamp-Heyken, German League 1991. (d2) 8 ... Bd7 9 Nc2 Bxd2+ 10 Qxd2 f6!? (10 ... Qxb2 – ‘obligatory’, according to Vitiugov – 11 Bd3 Nge7 12 0-0 Qb6 13 Qg5!? Ng6, when ‘White has compensation, Black has an extra pawn’ – Vitiugov; then 14 Rab1 Qc7 15 Qg3 0-0 is unclear) 11 Bd3 (11 exf6 Nxf6 12 Bd3 0-0 13 0-0 Nh5 14 Rfe1!? Rxf3! 15 gxf3 Rf8 and ... Nf4=) 11 ... fxe5 12 dxe5 Qxb2!? (12 ... Nh6 13 0-0 00 14 Rae1 Rxf3!? 15 gxf3 Rf8 16 f4 Qd8 17 b4 Qh4 gives sufficient compensation for the exchange, but no more) 13 0-0 Nxe5 (risky, but two pawns is a lot); e.g. 14 Nfd4 (14 Nxe5 Qxe5 15 Rfe1 Qd6 16 Qg5 Ne7 17 Qxg7 0-0-0) 14 ... Qb6 15 Rab1 (15 Qg5 Nf7!) 15 ... Qd6 16 Rxb7 Nf6 17 Nb5! Bxb5 18 Bxb5+ Ned7 19 Ne3! 0-0 20 Nc4 Qc5 21 Bxd7 Nxd7 22 Rxd7 Qxc4 23 Rc1 Qb5 24 Rxg7+ with a draw. 1.41 6 Be2 This is the traditional move, which has been slowly replaced by 6 a3, but still has high-level adherents. I will present two solutions out of several available ones: 1.411 1.412
6 6
... ...
cxd4 f6
I featured 6 ... Nh6 in past editions, but it requires more study and is rather space-consuming. These are simpler approaches. Nevertheless, just to show that there’s something for everyone, you can find two suggestions in the main line 6 ... Nh6 7 Bxh6 gxh6 8 Qd2 Bg7 9 0-0 0-0 10 Na3 cxd4 11 cxd4 Bd7 12 Nc2 in Chapter 3, section 3.4, the note on 7 ... Qb6. 1.411 6 ... cxd4 7 cxd4 Nh6
This is one of the oldest lines in the Advance Variation. It’s safe and established, and thus provides a counterpoint to 6 ... f6. At this point, in response to the threat of ... Nf5 (winning the d-pawn), White has traditionally played one of two moves: 1.4111 1.4112
8 8
Nc3 b3
But the alternatives, are important, especially 8 Bd3: (a) 8 Bxh6, often queried, is better than its reputation; although that primarily means that White has a number of drawing possibilities after 8 ... Qxb2 and:
(a1) 9 Nc3 Nxd4!? (a good move which, however, gives White a drawing option; 9 ... Qxc3+ 10 Bd2 Qa3 is hardly clear but keeps winning chances – then with 11 0-0 a6! Black intends to develop and consolidate; instead, after 11 ... Be7 12 Qc2, Mantovani-Yemelin, Kalithea 2008, Vitiugov gives 12 ... 0-0!? 13 Rab1 and calls the situation ‘very complicated’) 10 Qxd4! (10 Qc1 Qxa1!? is ultimately unclear, so perhaps 10 ... Qxc1+ 11 Bxc1 Nc2+ 12 Kd2 Nxa1 13 Bb2 Bd7 14 Bxa1 a6, intending ... b5 and/or ... Bc5, is the best way to search for an advantage) 10 ... Qxa1+ 11 Bd1 gxh6 12 0-0 Qb2 13 Ba4+ Kd8 14 Rb1 Qa3 15 Nxd5 exd5 16 Qxd5+ Kc7 ½-½ Maslik-Turcan, corr 2001, in view of 17 Rxb7+ Bxb7 18 Qd7+ Kb8 19 Qe8+ etc. (a2) 9 Nbd2 gxh6 10 0-0 (10 Rb1 Qxa2 11 0-0 seems well answered by 11 ... Qa5. Vitiugov recommends 11 ... Nb4!?, but 12 Bb5+ Bd7 13 Bxd7+ Kxd7 14 Ne1, intending f4 with Rf2 and/or f5, looks dangerous) 10 ... Nxd4 (10 ... a6!?) 11 Rb1 Nxe2+ 12
Qxe2 Qc3 (I think I prefer 12 ... Qa3!? with the idea 13 Qb5+ Bd7 14 Qxb7 Rc8!?; e.g. 15 Nb3 Bg7 16 Rfc1 Ke7) 13 Rfc1 Qa5 14 Nd4 a6 15 N2b3 (and here 15 Nf1! with the idea Ng3-h5 is an interesting try, maybe ‘dynamically equal’?) 15 ... Qd8 16 Qe3 Rg8. According to Vitiugov, the position ‘is quite reliable for Black’ and White’s compensation is ‘hardly sufficient for two missing pawns’. Hmmm, perhaps. At any rate, Black has as many chances as White, and can’t complain about drawishness! (b) 8 Bd3 has been played relatively often at master levels, especially over the past decade.
It is a strange-looking move, seemingly wasting a tempo (Be2-d3) to prevent ... Nf5, which results in an exact position from the Milner-Barry Gambit (see 6 Bd3 below) with the extra move ... Nh6 for Black. But now that White has gained c3 for his knight, Black’s own commitment of a knight to h6 helps White in some ways: (b1) 8 ... Nf5 9 Bxf5 exf5 10 Nc3 Be6 is a standard type of position which arises in several variations (it is also the same as 8 ... Bd7 9 Bc2 Nf5 10 Bxf5 exf5 11 Nc3 Be6, although in that case Black has the extra possibility of 11 ... Bb4 – see ‘b23’).
White has extra space with a strong centre that is difficult to assail. Black has two bishops (versus White’s bad one), while his doubled f-pawns seriously reduce White’s chances of doing any damage on the kingside; in fact, he has to pay attention to Black’s potential expansion with ... g5. Sveshnikov has hopes for a White advantage in this type of position, whereas Moskalenko says that ‘White has nothing in this structure’. I would lean towards the latter, at least in the sense that Black’s position should never be worse. Remarkably, Moskalenko has even written an article promoting Black’s position a full tempo down after 6 Bd3 (see the Milner-Barry) 6 ... Nh6!? 7 0-0 cxd4 8 cxd4 Nf5 9 Bxf5 exf5 10 Nc3 Be6. Furthermore, he claims that when White has b4 and Bb2 in, the Bxf5 exchange ‘favours Black in almost every case’; White has gained valuable space, but used time and created weaknesses, so that
has to be worked out case-by-case. Overall, I’d say that these Bxf5/ ... exf5 positions are objectively equal. In practice, however, White seems more often to be lacking in positive chances. Since this structure comes up extremely often, I’ll give several examples: (b11) 11 Ne2 h6 12 h4 (keeping the rook usefully on h1, at the cost of a clumsy king) 12 ... 0-0-0!? (ambitious and doubleedged – Black wants to play on the kingside with ... g5 and/or ... f4, although he reserves the idea of queenside play; instead, 12 ... Be7 13 Kf1 0-0 is a sound set-up, intending moves such as ... Rac8, ... Nb4, and ... Qa6) 13 Kf1 Kb8 (or 13 ... Be7 first, to discourage Nf4 without the need for h5) 14 Kg1 (14 Nf4!? can be met in similar fashion by 14 ... Be7 15 h5 Rd7, or by 14 ... g6 15 Kg1 Be7, contemplating a timely ... g5) 14 ... Be7 15 Rb1 (15 a3 Na5 16 b4 Nc4 17 Nf4 g6 18 Rh3 Qc6 19 Qb3 Rdg8 and ... g5 tends to follow – sometimes as a pawn sacrifice, or after ... Qe8; 15 Nf4?! g5 16 hxg5 hxg5 17 Rxh8 Rxh8 18 Nxe6 fxe6 19 Bxg5 Bxg5 20 Nxg5 Nxd4=/+) 15 ... f4! (15 ... Rc8 is safer) 16 Nxf4 (16 Bxf4 can be met by 16 ... Ka8, with the ideas of ... Bf5 or ... Bg4, or by 16 ... Bf5! 17 Rc1 Qxb2 18 e6+ Bd6 19 exf7 Rhf8=/+) 16 ... Bf5 17 Ra1 Be4 (17 ... Bg4! 18 Be3 Qxb2=/+) 18 Ne2! f6 19 Rh3! Rhf8 (19 ... Ka8!) 20 Bf4 Ka8 21 exf6 Bxf6 22 Ne5 Bxe5 23 dxe5, Benjamin-Shaked, Kona 1998; and here Black can try the active solution 23 ... Nxe5 24 Be3 d4!? 25 Bxd4 Bxg2! 26 Kxg2 Qg6+ 27 Kf1 Qf5 28 Ra3 Nc6 with an ongoing attack which likely ends in repetition. (b12) 11 Kf1 h6 12 h4 Be7 13 Ne2 0-0-0 (13 ... 0-0 is also logical, intending moves such as ... Rac8 and ... Qa6; e.g. 14 Nf4 Rac8 with the idea ... Nb4) 14 Kg1 Kb8 transposes to 11 Ne2. (b13) 11 0-0 subjects White to a potential kingside attack after ... 0-0-0 or ... Kd7 in conjunction with ... g5, but playing safely is at least as good; e.g. 11 ... Rc8 (or 11 ... h6 12 h4 Be7 13 Ne2 Rc8 14 b3 g6 15 Bf4 Kf8 16 Qd2 Kg7= Movsesian-Vallejo Pons, San Sebastian 2009) 12 Rb1 Be7 13 Ne2 h6 14 h4, McShane-Harikrishna, Goa 2002; and Black’s simplest course is 14 ... 00, intending to double on the c-file while protecting his kingside; if White attacks h6 by 15 Be3 Qa6 16 Qd2, then 16 ... Rfe8=/+ makes room for defence by ... Bf8 if necessary. I’ll revisit the important topic of ... Nf5 and Bxf5 in Chapter 3. (b2) 8 ... Bd7 9 Bc2 (9 Nc3 doesn’t produce quite enough compensation; e.g. 9 ... Nxd4 10 0-0 a6 11 Rb1 Bc6=/+)
and here the violent 9 ... g5!? of Vandervorst-Lundgren, corr 1994, and later games stands unrefuted. That’s worth investigating, but these moves are steadier and fully satisfactory: (b21) 9 ... g6!?, intending to recapture on f5 with the g-pawn, is an attractive way to create original play; for example, 10 Nc3 Nf5 11 Bxf5 (11 Na4 Bb4+ 12 Bd2 Qb5 is unclear) 11 ... gxf5 12 0-0 Rg8 13 Rb1 Rc8 14 Bf4 Rg6 15 Qd2 Na5=. (b22) 9 ... f6 10 exf6 (10 Bxh6 gxh6 hits b2, and 11 0-0 fxe5 12 dxe5 Qxb2 13 Nbd2 Qa3 looks to favour Black, as does 12 ... 0-0-0) 10 ... gxf6 (10 ... Nf7 11 Nc3! gxf6 transposes) 11 Nc3 Nf7 12 0-0 Bd6 13 Re1 0-0-0 14 Na4 Qc7 15 Nc5 (15 a3 Rdg8 16 Be3 f5!? 17 Rc1 f4 18 Bd2 Ng5 with mutual dynamic chances) 15 ... Bxc5 16 dxc5 Kb8 17 a3 Rhg8 18 g3 (18 Bxh7? Rg7 19 Bc2 e5/+) 18 ... e5 19 Qxd5 Bg4 20 Qb3 Bxf3 21 Qxf3 Nd4 22 Qc3 f5 with a strong attack. (b23) 9 ... Nf5 10 Bxf5 exf5 11 Nc3 and now 11 ... Be6 transposes to 8 ... Nf5 9 Bxf5 exf5 10 Nc3 Be6 above (with an added move for both sides). Black can also play 11 ... Bb4 12 Kf1 0-0 13 Kg1 Bxc3 14 bxc3 Rfc8 15 h4 Nd8!? (15 ... Na5=/+) 16 Bd2, Benjamin-Gretarsson, Bermuda 1999; and 16 ... Ne6 was best, since 17 Rb1 Qa6 18 Qb3 Rc7! 19 Qxd5 Bc6 20 Qb3 Rd8 21 Rh3 Bd5 22 Qd1 f4 gives Black full compensation. (c) 8 Na3!? can be answered in various ways, the most direct of which is the capture 8 ... Bxa3!? (8 ... Nf5 9 Nc2 Be7 is also possible) 9 bxa3 Nf5 10 Be3 Qa5+ 11 Qd2 (11 Bd2!? Qxa3 12 Bc1 Qb4+ 13 Bd2 Qe7=/+) 11 ... Bd7!? 12 Rb1 b6 with a structural advantage which outweighs the bishops. 4111
8
Nc3
8 ... Nf5 9 Na4 9 Kf1 Nfxd4! 10 Be3 (10 Na4 Qb4 11 Bd2 Qe7 12 Bg5 f6 13 exf6 gxf6 14 Nxd4 fxg5=/+ 15 Bh5+ Kd8 16 Rc1 Nxd4 17 Qxd4 Bg7-/+ – Boey) 10 ... Qxb2 11 Nxd5 Qxe2+! 12 Qxe2 Nxe2 13 Nc7+ Kd7 14 Nxa8 Nc3-/+ Schümans-Claesen, Belgian Ch. 1987. 9 ... Qa5+
9 ... Bb4+ 10 Kf1 (10 Bd2 Qa5 transposes to the main line) 10 ... Qd8 11 Bg5 Be7 12 Bxe7 Qxe7=, while 11 g4!? is a little loosening after 11 ... Nh4 12 Nxh4 Qxh4. Now 13 Bf4 has been recommended and played, but 13 ... h5! looks like a complete reply. For the significance of 9 ... Bb4+, see note ‘b’ to 10 Bd2 below. 10 Bd2 Instead: (a) 10 Kf1 attempts to preserve White’s space advantage, but leaving his king in the centre is quite a price to pay:
(a1) 10 ... b5 11 Nc3 b4 12 Nb5 Ba6!? (or 12 ... a6! 13 g4 axb5 14 gxf5 exf5=/+) 13 Nc7+ Qxc7 14 Bxa6 Qa5 15 Be2 h5=. (a2) 10 ... Bd7! indirectly attacks a4; e.g. 11 Bd2 Qd8 (to answer 11 g4 with 11 ... Nh4) 12 Bg5 Be7 13 Bxe7 Qxe7 14 h4 f6 15 g4 (15 exf6 Qxf6 wins the d-pawn) 15 ... Nfxd4! 16 Nxd4 Nxd4 17 Qxd4 fxe5 18 Qd1 0-0=/+. Black has two pawns for the piece and intends ... d4 and ... Bc6. (b) 10 Nc3 Qb6 11 Na4 Qa5+ 12 Nc3 repeats, so if Black wants to win, he should play the 9 ... Bb4+ order. 10 ... Bb4 11 Bc3 b5 11 ... Bxc3+ 12 Nxc3 Qb6 leads to a balanced game: 13 Bb5 (virtually forced) 13 ... Bd7 14 Bxc6 Bxc6 15 Qd2 Bb5 16 Nxb5 Qxb5 with equality or perhaps slightly better prospects for Black. 12 a3 Bxc3+ 13 Nxc3 b4 14 axb4 Alternatively: (a) 14 Bb5 Bd7 15 axb4 Qxb4 16 0-0 (16 Qa4 Qxb2 – see 14 axb4 Qxb4 15 Qa4 Bd7 16 Bb5 Qxb2 below; 16 Bxc6 Bxc6 transposes to the main line) 16 ... Nxe5 17 dxe5 Bxb5 18 Nxb5 Qxb5 19 g4 Ne7 20 Nd4 Qb4=/+ (Sveshnikov). (b) 14 Nb5 bxa3+ 15 Kf1 (Soltis) isn’t intimidating after 15 ... 0-0 16 Rxa3 Qb6 17 g4 (what else?) 17 ... Nfe7 with ideas of ... f6, ... Ng6 or simply ... Rb8; Black has the chances. 14 ... Qxb4
15 Bb5 The principal line. White has two serious alternatives: (a) 15 Qa4 Bd7 16 Bb5!? (16 Qxb4 Nxb4 is equal after 17 Kd2 0-0 or 17 0-0 Nc2 18 Rad1 0-0 19 Rd2 Nb4 20 Ra1 Rfb8=) 16
... Qxb2!? (or simply 16 ... Qxa4 17 Rxa4 0-0 18 Kd2 Rfb8 19 g4 Nfe7 20 Rha1 Rb7=) 17 Nxd5 (or 17 Ra3!? Nfe7 18 0-0 Qb4, when 19 Rb1 Qxa4 20 Nxa4 0-0 21 Nc5 Rfd8 gives White enough for a pawn, but no more) 17 ... exd5 18 0-0 Nfxd4 19 Rfb1 Nxf3+ 20 gxf3 Qc3 21 Rc1 Qb4 22 Bxc6 Qxa4 23 Rxa4 Bxc6 24 Rxc6 Ke7 with a very small edge, Kupreichik-Ulibin, Moscow 1989. (b) 15 0-0!? is an enterprising but speculative gambit: 15 ... Nfxd4 16 Nxd4 Nxd4 17 Bd3 (17 Ra4 Nxe2+ 18 Qxe2 Qb6=/+) 17 ... Bd7 (versus Ra4) 18 Qg4 g6 19 Qg5, Kuzmin-Lempert, St Petersburg 1993; and instead of 19 ... Nc6!?, giving Black a small edge after 20 Qh6 Qf8 21 Qf4 Qg7, 19 ... 0-0!-/+ was safe enough; e.g. 20 Rae1 Nc6 21 Re3 Qe7 22 Qh6 f5 23 exf6 Qxf6. 15 ... Bd7 16 Bxc6 16 Qa4 transposes to 15 Qa4; not 16 0-0? Nxe5! 17 Ra4 Qxb2 18 Bxd7+ Nxd7 19 Nxd5!? 0-0! and Black stays a pawn up. 16 ... Bxc6 17 0-0 17 Qd2!? Bb5!? (17 ... 0-0 18 0-0 Rfb8 transposes to the main line) 18 Nxb5 Qxb5 19 Ra5 Qb6 20 0-0 0-0 21 Rc1 Rab8 22 Rc2, Nunn-Schmittdiel, Dortmund 1991; and 22 ... Rfc8 23 Rac5 Ne7 is quite level. 17 ... 0-0 17 ... Qxb2 is safe but potentially drawish: (a) 18 Qd3 Qb7 19 Rfb1 Qd7 20 g4 Ne7 21 Nb5?! (21 Nd2 h5=) 21 ... Bxb5 22 Qxb5 Qxb5 23 Rxb5 h5! 24 h3 hxg4 25 hxg4 Nc6 with the idea 26 Ra6 Rh3!=/+. (b) 18 Na4! Qb4 19 Nc5 (19 Qc2 Qc4!) 19 ... 0-0 20 Rb1 Qc4 21 Rc1 Qb4 22 Rb1=. 18 Qd2 Rfb8 19 Rab1
The starting point of many contests over the years. With the b- and d-pawns as targets, Black stands at any rate not worse. 19 ... Qe7 Black has good alternatives here, such as 19 ... a5, intending ... a4, and 19 ... Rb6; e.g. 20 Rfc1 Rab8 21 Rc2 h6 22 g3 a5 with slightly the better side of equality, Shirov-Ivanchuk, Monaco (blindfold rapid) 2005. 20 g4 Against 20 b4, 20 ... a6 is Bardwick’s suggestion, intending simply ... Rb6 and ... Rab8. This is at least equal, and 20 Rfc1 Rb3 21 Ne2 Bb5 22 Rc3 Bc4 also gives nothing. 20 ... Nh4 This is one point of ... Qe7, the other being to cover b4. Now play can go 21 Ne1 Bb5 (21 ... f5!? 22 exf6 Qxf6 and 21 ... Rb3 are also good) 22 Nxb5 Rxb5 23 f4?! (23 f3 Rb4=/+) 23 ... h5 24 h3 Rb3 with real pressure. 1.4112 8 White will defend the d-pawn from b2. 8 ... Nf5 9 Bb2 Bb4+ Or 9 ... Be7 10 0-0 Bd7 11 g4? Nh4 12 Nxh4 Bxh4 13 Na3 0-0 14 f4 f6=/+ Kupreichik-Hertneck, Debrecen 1992. 10 Kf1 0-0
b3
The theory of this variation doesn’t seem to have changed sincePTF3, so I’ll take the opportunity to repeat what I wrote last time, with a few tweaks here and there. The position after 10 Kf1 has been extremely favourable to Black. For example, a typical line after 10 ... Be7 is 11 Nc3 Qd8 12 g3 f6 13 Kg2 fxe5 14 dxe5 0-0 15 Rc1 Bd7 16 Bd3, Sveshnikov-L.Portisch, Biel Interzonal 1993, and now Sveshnikov gives 16 ... Be8!=/+; e.g. 17 Ne2 Bh5 18 Bxf5 Rxf5 19 Nf4 Bxf3+ 20 Qxf3 Qd7 with a distinctly better game. 10 ... Bd7 is also effective. An example with characteristic tactics is 11 g3 (11 g4 Nh6 12 Rg1 0-0 13 a3 Be7 14 b4 f6 15 exf6 Rxf6=/+ Damljanovic-Velimirovic, Belgrade 1997) 11 ... 0-0 12 Kg2 f6 13 a3 Be7 and the white king still isn’t safe: 14 b4 fxe5 15 dxe5 Bc5! 16 g4 Nfe7 17 Nc3
17 ... Bxf2! 18 Na4 Qe3 19 Rf1 Ng6 20 Rxf2 Nf4+ 21 Kf1 Nh3! 22 Qe1 Nxe5 23 Bxe5 Bxa4 24 Bd4 Qe4 25 Qd2 Qg6? (25 ... Nxf2 26 Kxf2 Rxf3+ 27 Bxf3 Rf8 with the idea 28 Qe3 Rxf3+ 29 Qxf3 Qxd4+) 26 Qe3? 0-1 Sl.Martinovic-M.Pavlovic, Yugoslav Team Ch. 1994. White is lost after 26 ... Nxf2. 11 Nc3 This developing move avoids weaknesses while defending the d-pawn indirectly (11 ... Ncxd4?? 12 Na4). Alternatives: (a) 11 Bd3 f6! 12 Bxf5 fxe5!? (or 12 ... exf5!-/+) 13 Bc2? (13 Bh3 e4 14 Ng5 e3!; 13 Bxh7+ Kxh7 14 dxe5 Bd7=/+) 13 ... e4 14 Ne1 e5 15 a3 Bd6 16 b4 exd4 17 Bb3 Be6 18 f3 exf3 19 Nxf3 Rxf3+!–+ Nikoleris-Tsouros, Athens 1973. (b) 11 g4 is a traditional line which used to be critical, but Black’s successes have caused it almost to disappear.
For that reason, I’m going to cut way down on my previous coverage, though still hit a few critical lines: After 11 ... Nh6, White has: (b1) 12 Rg1 f6 (or 12 ... Bd7!?) 13 exf6 Rxf6 14 g5 (14 Nc3 Rf4! 15 h3 Bd7=/+) 14 ... Rxf3 and: (b11) 15 gxh6?! Rf7 16 Bd3 (16 Qd3 e5!) 16 ... e5 17 Qh5 e4 18 Be2, Botterill-Ligterink, London 1978, and here Botterill gives 18 ... g6!, winning. (b12) 15 Bxf3 Nf5 16 Rg4 Bd7 17 Nc3 Rf8 18 Na4 Qc7!? 19 Kg1 Be8 20 Nc5! Qe7! 21 Nd3 (21 Rc1 Bh5-/+) 21 ... Bd6!=/+ with ideas of ... Bh5 and/or ... Bg6. White’s position is loose and a pawn will fall to begin with. (b2) 12 h3 (best) 12 ... f6 (a rare recent game with this line saw 12 ... Bd7 13 Kg2 f6 14 a3 Be7 15 Nbd2 fxe5 16 dxe5, Lorenzo de la Riva-Moskalenko, Ampolla 2008; and here was a good place for the wonderful manoeuvre 16 ... Nf7 17 Rf1 Nh8!, intending ... Ng6-f4, and if 18 Bd3, 18 ... Be8! with ... Ng6 next) 13 exf6 Rxf6 14 a3 (for 14 g5 Rxf3 15 Bxf3 Nf5, compare 12 Rg1) 14 ... Bd6 15 g5? Rxf3! 16 Bxf3 Nf5-/+ Gutierrez Castillo-Matamoros Franco, Malaga 2001. (c) 11 a3 Be7 12 b4 f6 13 g3 fxe5 14 dxe5 a5! 15 b5 Bc5 16 g4 Nfd4–+ Campora Perez-Matamoros Franco, Seville 2001. 11 ... f6 12 Na4 12 g4 Nfe7!=/+ intending ... Ng6-f4. 12 ... Qd8 13 h4 fxe5 14 dxe5 Be7 White’s king is still stuck and he has no attack, so Black stands better. A sample line is 15 Bd3 Bd7! intending ... Be8-h5. This is advantageous, since 16 g4? Nxh4 only exposes White, and if 16 Rc1, then 16 ... Be8 17 h5 Bg5 activates Black’s pieces. 1.412
6
...
f6
It’s always fun when a long-neglected move like this appears playable. US Master Dan Heisman alerted me to 6 ... f6 based upon a game of his student and his own favourable impressions. Certainly there’s nothing unusual about attacking both d4 and e5. As I analysed it, I kept reaching sound positions with plenty of content, while even in fairly irrational lines Black’s position seemed resilient enough. So, since the classical variation 6 ... cxd4 7 cxd4 Nh6 from section 1.411 is a conservative one, I thought I’d put 6 ... f6 out there as a dynamic alternative to avoid established theory. Perhaps White can eke out a small edge somewhere, but I haven’t found it, and I doubt that he can achieve anything out of the ordinary. In the meantime, you’ll see that many lines with obvious moves actually favour Black. Be aware, though, that there is always more risk when Black is neglecting development in favour of pawn moves; hopefully the fun of being the first in some new territory will compensate you for that worry. Oh yes, I forgot to mention that the engines are happy with Black’s position and the succeeding play: 7 0-0 My databases have 131 games with this move, and 43 with assorted seventh moves. It’s worth looking at some ideas involving the latter: (a) 7 exf6 Nxf6 is co-operative of White, but has to be taken seriously. At least he doesn’t have a knight on d2 clogging up the works, as he does when he makes this move in the Tarrasch Variation. However, Black’s knight has come to f6 directly, as opposed to the three moves involved in the line 3 Nd2 Nf6 4 e5 Nfd7, in which Black later plays ... f6 and, after exf6, ... Nxf6. Incidentally, the only game I have with a 2600+ player on either side of 6 ... f6 (in this case 2668 at the time) involves 7 exf6 – see Beliavsky’s win in ‘a22’: (a1) 8 a3?! Bd6 (or 8 ... cxd4 9 cxd4 Bd6 10 Nc3 0-0 11 0-0 Bd7) 9 0-0 (9 dxc5 Bxc5 10 0-0 0-0 is already promising for Black; e.g. 11 Qe1 Ng4! 12 c4 e5! 13 cxd5 Nxf2 14 Rxf2 Nd4 will net material) 9 ... 0-0 and White needs a plan because ... cxd4 and ... e5 is threatened, but 10 dxc5? Bxc5 threatens 11 ... Ng4 or 11 ... Bf2+, and 11 Nbd2 (11 b4? Bxf2+ 12 Rxf2 Ng4 13 Qf1 e5 is horrible) 11 ... Bxf2+ 12 Rxf2 Ng4 13 Qf1 e5! 14 h3 Nxf2 15 Qxf2 Qxf2+ 16 Kxf2 e4 leaves Black the exchange ahead. (a2) 8 0-0 cxd4 9 cxd4 Bd6 10 Nc3 0-0
Again, in the Tarrasch Variation with 3 ... Nf6, White’s knights take four moves instead of two to get to their posts, but so do Black’s knights! The bishop on e2 is presumably not as well placed as on d3, at least if White has any hopes of a theoretical advantage. Here’s some analysis and an actual game(!): (a21) 11 h3 Bd7 12 a3 (12 Be3 Qxb2!? 13 Nb5 Be7 14 Rb1 Qxa2 15 Ra1 Qb2=; if Black wants more, he should play 12 ... Ne7 with the idea ... Nf5) 12 ... Rac8 13 Be3 Ne7 14 Rb1 Nf5 (this is standard ... Nf6 Tarrasch manoeuvring) 15 Bd3 Nxe3 16 fxe3 g6!? (intending ... Nh5; here 16 ... h6 17 Nh4!? Be8 18 Ng6 Bxg6 19 Bxg6 Bb8 and 16 ... Bg3!? are good alternatives) 17 Qe1! (17 Qc2 Kg7) 17 ... Nh5 18 Qh4 Rf7 with roughly equal chances. Now 19 g4!? might lead to 19 ... Bg3! 20 Qg5 Bf4! 21 exf4 Nxf4 22 Qh6! Nxd3 23 Qe3 Nxb2 24 Ng5 Rff8 25 Rf2 Nc4 26 Rxf8+ Rxf8 27 Rxb6 Nxe3 28 Rxb7 h6=. a22) 11 b3 Bd7 12 Nb5 Be7 13 Bf4!? Rac8 14 Nd6 Rcd8 15 Nb5 Ne8! 16 Bg3? Bf6 17 Be5 Nxe5 18 dxe5 Bxb5 19 Bxb5 Qxb5-/+ 20 Nd4 Qb6 21 exf6 Nxf6 22 f4? Ne4 23 Kh1 e5 0-1 T.Johansson-Beliavsky, Copenhagen 2003. (b) 7 Bd3 is a logical reaction to ... f6, even if it wastes a little time: 7 ... cxd4 (or 7 ... fxe5 8 dxe5, when the standard 8 ... Nh6 9 0-0 Nf7 should be fine for Black) 8 cxd4 Bb4+ with:
(b1) 9 Kf1 fxe5 10 dxe5 Nge7 (10 ... Nh6 11 Bxh6 gxh6 12 Nc3 Bf8!? 13 Na4 Qa5 14 Rc1 Bg7 15 a3 0-0 is unclear) 11 a3 Bc5 12 Nc3 Bd4!; for example, 13 Na4 Qc7 14 Nxd4!? Nxd4 15 Bf4 0-0=/+ 16 Bxh7+?! Kxh7 17 Qxd4 Qc2! with the idea 18 Rc1?? Nc6. (b2) 9 Nbd2!? Nxd4 10 0-0 Nc6! 11 a3 Be7 12 Qe2 (12 Re1 Bc5) 12 ... fxe5 (or 12 ... Nh6) 13 Nxe5 Nxe5 14 Qxe5 Bf6 15 Qh5+ g6 16 Qe2 e5! 17 Bb5+ Kf7=/+. (b3) 9 Nc3 fxe5 10 dxe5! (10 Nxe5 Qxd4 11 Nxc6 Bxc3+ 12 bxc3 Qxc3+ 13 Qd2 Qxd2+ 14 Kxd2 bxc6 15 Bb2 Nf6 16 Rac1
Bd7 17 Be5 with real pressure, but two extra pawns are worth some suffering) 10 ... d4 11 a3 dxc3 12 axb4 cxb2 13 Bxb2 Qxb4+ 14 Qd2 Qxd2+ 15 Nxd2 Nge7 16 Ne4 0-0 17 0-0 a5 18 Nd6 Bd7 with mutual chances; White has compensation, and it looks pretty balanced.
7 ... fxe5! For some reason 7 ... cxd4? has been the usual response in the limited practice of 6 ... f6, with very poor results (it unnecessarily presents White with c3 for his knight). The text move, by contrast, achieves a standard position in which there’s no reason Black shouldn’t stand perfectly well. Finally, 7 ... Bd7 isn’t as pointed a move as 7 ... fxe5, but it seems playable. Movsesian-Socko, Warsaw 2004, continued 8 Re1 (8 Bd3! may be best; e.g. 8 ... fxe5 9 Nxe5 Nxe5 10 dxe5 and 10 ... Ne7 is probably better than 10 ... 0-0-0 11 Nd2 Ne7 12 Nf3 Nc6 13 b4! Qc7!? 14 bxc5 Bxc5 15 Qe2+/=, but Black is in the game) 8 ... fxe5 9 Nxe5 Nxe5 10 dxe5, and here Black can develop smoothly by 10 ... c4; for example, 11 Qc2 Bc5 12 Bh5+ g6 13 Bg4 Ne7 14 Bh6 Nf5 15 Bxf5 gxf5 with mutual chances. After 7 ... fxe5, White chooses between recaptures: 1.4121 1.4122
8 8
Nxe5 dxe5
1.4121 8 Nxe5 Now 8 ... Nxe5 9 fxe5 Bd7 10 c4 d4 11 Nd2+/= is ancient analysis by Zaitsev, and the only 6 ... f6 line cited by many sources. But Black has two good alternatives: 8 ... Nf6 8 ... cxd4 9 Nxc6 bxc6 10 cxd4 also looks adequate after: (a) 10 ... Nf6 11 Nc3 Bd6 12 g3 (12 f4=) 12 ... 0-0 13 Na4 Qc7 14 f4! a5 15 Be3 Ne4 16 Rc1 Ba6 and White’s edge, if any, is tiny. (b) The surprising 10 ... Bd6! looks best, not yet committing the knight: 11 b3 (11 Nc3 Ne7! 12 Na4?! Qc7 13 h3 0-0 14 Bd2 e5 and Black stands better; 11 Nd2 Nf6! 12 Nf3 0-0 13 Qc2 Ba6!?= or 13 ... c5) 11 ... Qc7 12 h3 (12 g3 e5) 12 ... Ne7!? (or 12 ... e5=) 13 Bb2 Ng6 14 Nd2 0-0=; e.g. 15 Nf3 Nf4 16 Rc1 a5!? 17 Re1 a4 18 bxa4 Rf6 19 Bf1 Ba6=. 9 dxc5 Surrendering the centre appears unlikely to bring an advantage. Two others: (a) 9 Bh5+ g6 10 Bf3 (or 10 Be2 cxd4 11 Nxc6 bxc6 12 cxd4 Bd6 13 Nc3 0-0=, a sample line being 14 Bh6!? Rf7 15 Na4 Qc7 16 f4 Ne4 17 g3 Be7! intending ... Bf6) and we get some typical structures: 10 ... Nxe5 (Pizzuto-Handel, corr 1994, went 10 ... cxd4 11 Nxc6 bxc6 12 cxd4 Bd6=; e.g. 13 Nc3 0-0 14 Bh6 Rf7 15 Qd2, and 15 ... Bd7 was equal, but 15 ... e5 16 dxe5 Bxe5 gets there more directly) 11 dxe5 Nd7 12 c4 (12 Na3?! Bg7) 12 ... d4 13 Qa4 Bg7 14 Bg4 Bxe5 (or 14 ... 0-0 15 f4 g5! 16 g3 gxf4 17 gxf4 Bh6=/+ intending ... Kh8) 15 f4 Bg7 16 Re1 Qb4 (16 ... e5!? 17 fxe5 Qb4 18 Qxb4 cxb4=) 17 Qxb4 (17 Rxe6+? Kf7-/+) 17 ... cxb4 18 Bxe6 Kd8 19 f5 Re8 20 Bg5+ (20 Nd2 Nc5) 20 ... Bf6 21 Bxf6+ Nxf6, followed by ... gxf5 with equality or just a tad more. (b) 9 Nf3 cxd4 (or 9 ... Bd6 10 dxc5 Bxc5 11 b4 Bd6 with ample opportunities in a dynamically equal position) 10 cxd4 Bd6 11 Nc3 0-0
White has played good moves, but this is a safe version for Black of a 3 Nd2 Nf6 Tarrasch main line with 7 ... Qb6. He should have comfortable equality following 12 Be3 Ng4!? (12 ... Qxb2 13 Na4 Qa3 14 Bc1=, or here 13 Nb5 Be7 14 Rb1 Qxa2 15 Ra1 Qb2=; while 12 ... Bd7 13 Ne5 Bxe5!? 14 dxe5 Qxb2 15 exf6! Qxc3 16 fxg7 Qxg7 17 Qd2 Rf6, with the idea 18 f4 Be8, leaves White with about a pawn’s worth of compensation) 13 Na4 Qc7 14 h3 Nxe3 15 fxe3 Qe7=. 9 ... Bxc5 10 Nd3 Not 10 b4? Bxf2+! or 10 Nxc6?! bxc6 11 b4 Bd6 12 Be3 Qc7 13 h3 a5=/+. 10 ... Bd6 11 Be3 Qc7 12 f4 0-0 13 Nd2 e5 13 ... Bd7! is more ambitious. 14 fxe5 Nxe5 15 Nxe5 Bxe5 16 Nf3 Bd6 17 Kh1 17 Bd4 Bg4! with a compensating attack after 18 Bxf6!? Rxf6 19 Qxd5+ Be6 20 Qe4 Kh8. 17 ... Bg4 17 ... Bd7 18 Bd4 Ne4!=. 18 Bg5 18 Nd4 Rae8 19 Nb5 Bxe2 20 Qxe2 Qe7 21 Nxd6 Qxd6 22 Qd3 b6=. 18 ... Kh8 19 Bxf6 Rxf6 20 Nd4 Bxe2 20 ... Rxf1+ 21 Qxf1 Rf8 22 Qd1 Bd7 with some chances. 21 Qxe2 Bxh2 22 Rxf6 gxf6 23 Qe6= Wieweg-Dysing, Sweden 1993. 1.4122
8
dxe5
Nh6
This is my favourite Advance Variation move. Instead, 8 ... Nge7 has been played several times with success, but I think that White should get something from 9 c4! d4 10 Nbd2 Ng6 11 Ne4! Ncxe5 12 Nxe5 Nxe5 13 Bg5! with the idea of Bh5+. Perhaps this isn’t much, but at any rate Black is on the defensive. 9 Re1 Probably the most important move, aiming to protect e5. Let’s look at a few of the more obvious alternatives: (a) 9 Bxh6
9 ... gxh6 (9 ... Qxb2?? 10 Qd2 Qxa1 11 Be3 d4 12 cxd4 cxd4 13 Bxd4 Bb4 14 Bc3+–) 10 b3 (10 c4!? allows 10 ... Qxb2, which is risky, or 10 ... d4 intending 11 Nbd2 Bd7!? and ... 0-0-0) 10 ... Bg7 11 Bd3 Bd7 keeps open the choice of ... 0-0-0 or ... 00; alternatively, 11 ... 0-0 12 Qe2 Bd7 with the idea ... Be8 is level, whereas here 12 Qc2?! Nxe5 13 Bxh7+ Kh8 simply favours Black. (b) 9 Qc2 Nf7 10 Re1 g6 (or 10 ... Bd7, again considering ... 0-0-0) 11 Bd3 (11 Na3 Bd7, to meet Nb5 with ... Ncxe5; 11 Be3! Be7 12 Nbd2 0-0 13 b4 Ncxe5 14 Nxe5 Nxe5 15 Nb3 Qc7 16 Bxc5 Bxc5 17 Nxc5 a5!) 11 ... Bg7 12 Bf4 0-0 13 Nbd2 (13 Na3 Qc7 14 Qe2 a6 or 14 ... c4 15 Bc2 Bd7) 13 ... c4 14 Bf1 Qc7 15 Bg3 Bd7 and I like Black’s game. (c) 9 b3 Nf7 10 Bf4 and one of several possibilities is 10 ... Be7 with the idea of ... g5 and ... h5 if allowed (or 10 ... g6!? 11 Bd3 Bg7 12 Qe2 0-0 13 Nbd2 Bd7 14 h3 Nd6!? 15 Bh2 Nf5=), so 11 Qd2 Bd7 12 Na3 a6 is logical, or here 12 ... 0-0 13 Nc2 Be8=.
9 ... Nf7 10 Na3 10 b3 g6 11 Bf4 Bg7 12 Bd3 0-0 resembles 9 b3, and here it’s not clear whether the rook should be on e1. 10 ... Be7 Or 10 ... Bd7. 11 Bf1 a6 11 ... Qc7 is also fine; e.g. 12 Nb5 (12 Bf4 a6) 12 ... Qb8 13 Bf4 (13 h3 a6 14 Na3 0-0 15 Bf4 Qc7 16 Qd2 Bd7=) 13 ... a6 14 Na3 b5 (14 ... g5!? 15 Bg3 h5; 14 ... Qc7) 15 Nc2 0-0 16 Qd2 a5 and Black can advance on the queenside: ... a4/ ... Na5-c4 could follow, or simply ... Ba6. 12 Nc2 0-0 13 Rb1 Qc7 14 Qe2 14 Bf4 g5. 14 ... Bd7= Black stands very flexibly here: he can play on the queenside with ... b5 or reorganize the kingside by ... Be8 and either ... Nfd8/ ... Bg6 or ... Nh8/ ... Bg6. 1.42 6 Bd3 This usually leads into the Milner-Barry Gambit, in which White gives up a pawn for rapid development. It has never gained respectability, but has some fascinating ramifications. 6 ... cxd4 7 cxd4 7 0-0 can answered by 7 ... Bd7, transposing to the main line after 8 cxd4 Nxd4. Also possible is 7 ... dxc3 8 Nxc3 Nge7. 7 ... Bd7
Not 7 ... Nxd4?? 8 Nxd4 Qxd4?? 9 Bb5+. 8 0-0 The normal move, because the alternatives can be rather dispiriting: (a) 8 Be2?! is a full tempo down on older 6 Be2 lines after 8 ... Nh6! 9 b3 Nf5 10 Bb2 Bb4+=/+ 11 Kf1 0-0 12 Nc3 f6 and Black will wreak havoc on the f-file. (b) 8 Bc2!? is slow. After 8 ... Nb4, White has lost any chance for advantage and done poorly: 9 Nc3 (not 9 Ba4? Qa6! 10 Bxd7+ Kxd7 and ... Nd3+ follows; or 9 Bb3? Qa6; while 9 0-0 Nxc2 10 Qxc2 Rc8 11 Nc3 transposes to 9 Nc3) 9 ... Nxc2+ 10 Qxc2 Rc8 (or 10 ... Ne7 11 Be3 Nc6 12 0-0 Be7 intending ... 0-0 and ... f6) 11 0-0 Ne7 12 Be3 (12 Qd3 Nf5) 12 ... Nc6 (12 ... Nf5!) 13 Qd2 (13 a3 Na5) 13 ... Bb4!? 14 a3, Stadler-Ilchmann, corr 1958; 14 ... Bxc3! 15 bxc3 Na5=/+ (Heemsoth). (c) 8 Nc3 Nxd4
and now 9 Nxd4 Qxd4 10 0-0 is the main line gambit. White also has: (c1) 9 Ng5!? resembles the 8 0-0 Nxd4 9 Ng5 gambit line below. But this move order is flawed due to 9 ... Nc6!, since Re1 is not available; e.g. 10 0-0 (10 Qe2? d4! 11 Nb5 Nb4!) 10 ... Nxe5 11 Re1?! (11 Nxd5! Qd6!=/+) 11 ... Nxd3 12 Qxd3 Nf6-/+ with the idea 13 Nxd5 Nxd5 14 Qxd5 Be7-/+. (c2) 9 0-0 Nxf3+ 10 Qxf3 Ne7 11 Qg3 Bc6!? (11 ... Rc8) 12 a4, Sobolevsky-Enders, Mülhausen 2000; and 12 ... a6=/+ was called for, stopping Nb5. (d) 8 Nbd2 Nxd4 9 Nxd4 Qxd4 10 Nf3 is arguably premature due to 10 ... Bb4+ 11 Kf1 (11 Ke2 Qb6 12 Be3 Bc5) 11 ... Qb6 12 Be3 (12 Ng5 Bb5) 12 ... Bc5 13 Bxc5 Qxc5=/+.
8 ... Nxd4 9 Nxd4 White has other moves to take seriously: (a) 9 Nbd2 is Nun’s move, White’s best way, in my opinion, to play for genuine chances in the Milner-Barry. Now 9 ... Nxf3+ and 9 ... Ne7 have been tried, along with these critical lines: (a1) 9 ... Nc6 10 Nb3! Nge7 11 Be3 (11 Bf4 Ng6 12 Bg3 a5 13 a4 Be7-/+ intending ... 0-0 and ... f5) 11 ... Qc7 12 Rc1 Ng6 13 Nc5! (13 Re1 Ngxe5!). White has definite compensation, it’s just not clear how much; for example, 13 ... Rc8 (13 ... Bxc5!? 14 Bxc5 Ncxe5 can enter into a lengthy drawing line: 15 Nxe5 Qxe5 16 Qb3 Bc6 17 Rfe1 Qg5 18 Bb5! Rd8 19 Bxc6+ bxc6 20 Qb7 Nf4 21 Qxc6+ Rd7 22 Qc8+=) 14 Qd2! (14 b4?! Bxc5 15 Bxc5 Ncxe5 16 Nxe5 and 16 ... Nxe5 gives Black a meaningful advantage, as does 16 ... Qxe5 of Polivanov-Ponkratov, Moscow 2011) 14 ... Be7!? 15 b4 a6 16 a4 Ncxe5!? 17 Nxe5 Qxe5 18 Nxd7 Kxd7 19 f4 Qb8 and everything is still up for grabs. White’s attack is real, yet all Black needs is some simplification for his two extra pawns to come into play. (a2) 9 ... Bc5! 10 Nxd4 (10 Rb1 Bb5!=/+; 10 b4 Qxb4 11 Rb1 Qa4!=/+ Pap-Gleizerov, Predeal 2006, or here 11 Nxd4 Bxd4 12 Rb1 Qc3 13 Nf3 Bb6=/+; 10 ... Nxf3+ 11 Nxf3 Qxb4 is also possible) 10 ... Bxd4 11 Nf3 Ne7 12 Nxd4 (12 Rb1 Nc6 13 Bf4, Soloviev-Zhukov, St Petersburg 2008; and 13 ... a5! looks good, or 13 ... Rc8 14 b4 Bc3!; while 12 Bxh7!? can be met by 12 ... Nc6!? or 12 ... Bb5! 13 Bd3 Bxd3 14 Qxd3 Nc6 with easy play) 12 ... Qxd4 13 Re1 Nc6 (13 ... Ng6) 14 Be3 Qxe5 15 Qd2, SmerdonV.Akopian, Khanty-Mansiysk 2010; here Moskalenko suggests 15 ... Qd6=/+. (b) 9 Nc3!? might transpose after 9 ... a6, but Black also has 9 ... Nxf3+! 10 Qxf3 Ne7; e.g. 11 Qg3 Bc6 (11 ... Rc8!) 12 a3 a6 13 b4 Ng6=/+ Maderna-Najdorf, Mar del Plata 1942. (c) 9 Ng5 is Sorensen’s Gambit, tricky but perhaps too speculative: 9 ... Nc6! is the simplest course, fighting for the initiative by attacking e5: (c1) 10 Re1 Bc5 (this is not the only solution, but it’s relatively forcing) 11 Qf3 0-0-0 (11 ... Nh6 12 Nc3 Nd4 is also good) 12 Nc3 (12 Nxf7?? Rf8–+) 12 ... f6! 13 exf6 (not 13 Na4?? Qa5 or 13 Nf7 Nxe5! 14 Rxe5 fxe5 15 Nxh8 e4) 13 ... Nxf6 14 Nf7 Ng4! (14 ... Rhf8 and 14 ... Nb4 are also good) 15 Qxg4 Bxf2+ 16 Kf1 Bxe1 17 Kxe1 Nb4! 18 Ne5 Rhf8 (18 ... Qg1+ 19 Ke2 Nxd3 20 Kxd3 d4) 19 Ne2 Nxd3+ 20 Nxd3 e5 with too strong an attack. (c2) 10 Bf4? Qxb2 (or 10 ... Nge7) 11 Nd2 (11 Qd2 Qxa1 12 Nc3 Qxf1+ 13 Kxf1 Bb4–+) 11 ... Be7! 12 Ngf3 Nb4!-/+. (c3) 10 Qe2 Nb4 wins the bishop pair and remains a pawn up, unless White tries 11 Bxh7, when, among others, 11 ... Nh6! is strong (threatening ... Bb5 and ... g6, trapping the bishop): 12 Rd1 (12 Na3 g6 13 Be3 Qd8-/+) 12 ... Be7 13 Be3 (13 Bd3 Nxd3 14 Qxd3 Rc8 15 Nc3 Ng4! 16 Qg3 Nxf2!) 13 ... Qa6 14 Qxa6 bxa6 15 a3 Nc6-/+. 9 ... Qxd4
We have reached the traditional Milner-Barry main line. 10 Nc3 Instead: (a) 10 Qe2 is a little slow: 10 ... Ne7 (10 ... f6 is also playable, but it’s better to develop) 11 Kh1 (11 Nc3 Nc6 12 Bb5 a6 13 Rd1 Qh4!? 14 Bxc6 bxc6 favours Black; 11 Nd2 Nc6 12 Nf3 Qb6 13 Be3 Bc5 14 Bxc5 Qxc5 15 Rfc1 with compensation, according to Moskalenko, but White doesn’t seem to have enough for a pawn after 15 ... Qe7; e.g. 16 a3 0-0 17 Qe3 f6 etc) 11 ... Nc6 (11 ... Qh4 is also good) 12 f4 Bc5!? (or 12 ... Nb4 13 Bb5 0-0-0-/+) 13 a3 a6 14 Ra2 0-0 15 b4 Ba7 16 Nd2 Qe3-/+ J.Brown-Hummel, Las Vegas 1997. (b) 10 Re1 removes protection of f2 and reduces support for the f4-f5 attack:
10 ... Ne7 11 Nc3 (11 Na3!? Qa4! 12 b3 Qa5 13 Nc2 Ng6 or 13 Qe2 a6 14 Nc2 Ng6) 11 ... a6 12 Be3! (the best try; 12 Qf3 Ng6 13 Qg3?! Be7!) 12 ... Qxe5 13 Qf3 (or 13 Qg4 h5! 14 Qb4 Nc6 15 Qxb7 Rb8 16 Qxa6 Nb4 17 Qa7 Nxd3 18 Bd4, Vazquez Torres-Moreno Ruiz, Madrid 2010; and instead of 18 ... Qd6, good but messy, Moskalenko points out the simple 18 ... Bc5! 19 Rxe5 Bxa7 20 Bxa7 Rb7-/+) 13 ... Bc6!. It’s hard to believe that White has enough for two pawns (and 13 ... d4 14 Bf4 Bc6! is probably good as well). 10 ... a6 This frequently-played move rules out Nb5 and clears the a7-square for the dark-squared bishop. 10 ... Qxe5 is perfectly good, but players have complained, probably correctly, that certain lines don’t yield enough winning chances.
11 Qe2 11 Re1 Ne7 transposes to 10 Re1 above. Others: (a) 11 Kh1 Qxe5 12 f4 (12 Re1 Qd6 13 Qf3 Bc6-/+ 14 Bf4 Qd7 15 Qg3 Ne7) 12 ... Qd6 13 Be3 Be7 14 Qf3 Bc6 (14 ... d4 15 Ne4 Qb6-/+) 15 Bd4, Eley-Whiteley, British Ch., Ayr 1978; and 15 ... Bf6! is a simple solution; e.g. 16 Be5 Qe7 17 Qg3 Nh6 18 f5 Bxe5 19 Qxe5 Qg5-/+. (b) 11 Qf3 Ne7 (11 ... Qxe5 12 Bf4 Qf6 13 Qg3 Ne7=/+) 12 Rd1 Qxe5 13 Bf4 Qf6 14 Re1 (14 Qg3 Ng6-/+) 14 ... g5! 15 Be5 Qxf3 16 gxf3 Rg8 17 Bxh7 Rg7! 18 Bd3 f5 19 Bxg7 Bxg7-/+; for the exchange, Black has a pawn, the bishop pair, and ideal outposts for his knight on f4 and h4. 11 ... Ne7 This has been the overwhelming favourite for years, though 11 ... Qh4!? has also done well ( ... Nh6 is a theme), and
Moskalenko likes 11 ... Rc8, preparing ... Bc5; he cites Startseva-Esebua, Mureck 1998: 12 Rd1 Qb6 13 Kh1 Bc5 14 Qg4 g6 with a solid position.
12 Kh1 The older 12 Rd1 doesn’t equalize after 12 ... Nc6 (12 ... Qh4 is also good) 13 Bxa6 (13 Bb5 Qb6=/+ 14 Ba4?! Bc5!-/+; or 13 ... Qh4) 13 ... Qxe5 14 Qxe5?! (14 Bxb7 Qxe2 15 Nxe2 and here 15 ... Ra7 16 Bxc6 Bxc6 is slightly better for Black, but his bishop pair and centre are better-suited to the position after 15 ... Rb8 16 Bxc6 Bxc6 17 Nd4 Bd7; e.g. 18 a4 f6 19 a5 e5 20 Nc2 d4 21 a6 Bc5 22 b4 Ba7=/+) 14 ... Nxe5 15 Bxb7 Ra7 16 Bxd5 exd5 17 Re1 f6 18 f4 Bc5+ 19 Kh1 d4-/+ Mnatsakanian-Monin, USSR 1979 (and several later games). 12 ... Nc6 13 f4 13 Be3 looks unsound after 13 ... Qxe5 (or 13 ... Qh4, threatening ... d4; e.g. 14 g3 Qd8 15 Bf4 Be7-/+) 14 f4 Qd6 15 f5 (15 Rad1 Be7 16 f5 Qe5) 15 ... Qe5!-/+. 13 ... Nb4 This is the introduction to an apparently effective solution which has heretofore been rather neglected. InPTF3I gave a lengthy exegesis on the virtues of 13 ... Bc5!?, which develops, prevents Be3, and allows for ... Qf2 after White’s Rd1. That move is still holding up very well, with the only serious threat provided by the line 14 a3 Na5 (14 ... Ba7=/+) 15 Qf3!? (15 b4 and 15 Bd2 were examined there) 15 ... Ba7 16 f5 Nc4 17 Ne2 Qb6 18 Qg3 of Rodriguez Vila-Ramirez Alvarez, Calvia Olympiad 2004; when instead of 18 ... exf5?, 18 ... 0-0-0! looks sound and promising. 14 Rd1 Now 14 ... Nxd3 15 Rxd3 has a long history and great body of theory behind it (verdict: White has some compensation). Instead, 14 ... Bc5! may make this unnecessary and in fact be a refutation of the Milner-Barry main line.
The first point is that Black threatens 15 ... Nxd3 16 Qxd3 (16 Rxd3?? Qg1 mate) 16 ... Qxd3 with a clear pawn up. As far as I can tell, White can never equalize. 15 Bxa6 White regains his pawn; this is probably the move which comes closest to levelling the chances. Otherwise: (a) 15 Bb5 Qf2! 16 Bxd7+ Kxd7 17 Qg4?? Nd3 threatens smothered mate with ... Qg1+, and 18 h4 Qg1+ 19 Rxg1 Nf2+ 20 Kh2 Nxg4+ 21 Kg3 Bxg1 isn’t comforting. (b) 15 Bxh7 Qf2 16 Qxf2 (16 Bb1?? Rxh2+) 16 ... Bxf2 and now:
(b1) 17 Bb1 0-0-0 (with the idea ... Rxh2+!; 17 ... Nc6 intending ... Rc8 also favours Black) 18 h3 Kb8 and Black stands better (Ziegler and Har Zvi, whose analysis directed my attention to 14 ... Bc5). Black’s initial advantage consists of his active pieces and passed pawn; he also has the prospect of ... f6, giving him a 2:0 central majority if White has to play exf6. This applies to several of the positions after both 15 Bxa6 and 16 Bxh7. (b2) 17 Rf1 Bg3!? (17 ... Rxh7 18 Rxf2 d4 19 Ne4 Bc6 20 Nd6+ Ke7=/+) 18 Rf3 Bxh2 (18 ... Be1!? 19 Bb1 Rc8 20 a3 Nc6=/+) 19 Rh3 Bg3 20 Kg1 Be1 and Black emerges a pawn up. (c) 15 Bb1 doesn’t even win the pawn back, and if White answers 15 ... Qf2 with 16 Qg4, 16 ... h5! keeps a very large advantage: (c1) 17 Qxg7? 0-0-0 18 Qg3 Qxg3 19 hxg3 Rdg8 with the idea 20 a3 (20 Kh2 Rg4! 21 Kh3 Rhg8 22 Ne2 Bb5; 20 Ne2 Rg4 21 Ng1 h4 22 gxh4 Bc6! and large material gain will follow ... d4 and ... Rhxh4+) 20 ... Rg4!! (threatening ... h4; not that 20 ... Nc6 isn’t good) 21 axb4 Bf2! 22 Rd3 h4 23 gxh4 Rhxh4+ 24 Rh3 Rxh3+ 25 gxh3 Rg1+ 26 Kh2 Rxc1 and ... d4-d3-d2 follows.
(c2) 17 Qh3 Rc8 (or 17 ... 0-0-0, or 17 ... Ba7 18 a3 Nc2 19 Bxc2 Qxc2) 18 Bd2 Ba7 19 a3 Nc6-/+. 15 ... Qf2 16 Qxf2 16 Bb5 Qxe2 17 Bxe2 Nc2 18 Rb1 0-0 19 Bd2 Ne3 20 Rdc1 Bb6 21 a3 f6!=/+. 16 ... Bxf2
17 Bb5! 17 Be2 0-0 (or 17 ... Bc6 18 a3 Nc2 19 Rb1 Nd4) 18 Bd2 (18 a3 Na6!? intending ... Nc5) 18 ... Nc2 19 Rab1 Ne3 20 Rdc1 Bc6+/= 21 Bf3? Nc4–+ Milesi-Apicella, Cap d’Agde 2003; Black will be a pawn ahead with much the better position. 17 ... Bxb5 17 ... Bc6 keeps only a small edge after 18 a3. 18 Nxb5 Rxa2 19 Rxa2 An instructive example went 19 Bd2 Rxa1 20 Rxa1 Na6 21 Nd6+ Kd7 (21 ... Ke7!?) 22 Nxb7?! (22 Nxf7 Rc8; 22 g3) 22 ... Rb8 23 Rxa6 Rxb7 24 b4 (Black has the better bishop and king) 24 ... Bb6 (or 24 ... Ra7! with the idea 25 Rxa7+ Bxa7 26 g3 Kc6 27 Kg2 Kb5 28 Kf3 Kc4) 25 g4 Kc6 26 Kg2, Chua-Dokoutchaev, Dallas 2002; and here 26 ... Ra7! would again be the clearest solution. 19 ... Nxa2
With an extra pawn, although it’s not easy to convert. 20 Bd2 Apparently trapping Black’s knight.
20 ... Kd7 21 Ra1 Rc8! This tactic is necessary. 22 Na3 22 Rxa2 Rc2. 22 ... Bc5! 23 Rxa2 Bb4 24 Be3
Or 24 Bc3 d4 25 Nb5 dxc3 26 bxc3, Toufighi-J.Gonzalez Garcia, Khanty-Mansiysk 2010; here 26 ... Bxc3 gives good winning chances, since 27 Nd6 Rc7 28 Nxf7? will lose to 28 ... b5 and the pawn is too fast. After 24 Be3, Aryanejad-Potkin, Abu Dhabi 2005, continued 24 ... d4! 25 Bxd4 Rc1+ 26 Bg1 Bc5 27 h3?! (27 h4 Rxg1+ 28 Kh2=/+) 27 ... Rxg1+ 28 Kh2 Rf1 29 g3 Rf2+ 30 Kh1 Rf1+ 31 Kh2 Rf2+ 32 Kh1 Kc6! 33 Nc4 Kd5 (33 ... b5! 34 Nd6 b4) 34 Na5 b6 35 Nb7 Bd4 36 Ra8 Rxb2 37 Nd6 f6! and Black won shortly.
Chapter
Two
Advance
Variation:
6
a3
1 e4 e6 2 d4 d5 3 e5 c5 4 c3 Nc6 5 Nf3 Qb6 6 a3
This has become the very main line of the Advance Variation. White prepares an expansion on the queenside and counts upon his space advantage in that sector (and in the centre) to keep Black’s pieces cramped. Sometimes, later, he expands on the kingside for good measure. The drawback to this strategy is that Black gains a nice lead in development and therefore has quick pressure on the centre. I recommend two variations: 2.1 2.2
6 6
... ...
Bd7 Nh6
Black has a wide variety of alternatives. 6 ... c4 is the oldest move (recommended inPTF1) and is apparently holding up well; players who enjoy closed positions may want to investigate it. Let me add some superficial comments on a couple of other moves that you may already be playing: (a) 6 ... f6
This has been a suggestion from the first edition onward, and still strikes me as sound. Black can (like his opponent) play on both wings here: on the queenside with ... Bd7, ... a5, and ... Rc8; and on the kingside with ideas like ... fxe5, ... Nh6-f7, and ... Be7 or ... g6/ ... Bg7. Here’s a very abbreviated set of ideas to review: (a1) 7 b4 fxe5!? 8 Nxe5 (8 dxe5!? Nh6! plans 9 Bxh6 gxh6 10 Bd3 Bg7 11 0-0 0-0=/+ Guigonis-Renaudin, Paris 2000; Black also achieves good play after 8 bxc5 Qa5 and 8 dxc5 Qc7) 8 ... Nxe5 9 dxe5 Ne7 (or 9 ... Qc7, when 10 Bb5+ Bd7 11 Bxd7+ Qxd7 12 0-0 a5 might follow) 10 Bd3 Qc7 11 Qe2 (Collins). Now 11 ... Ng6!? and 11 ... c4 12 Bc2 Ng6 are equal, but easiest is 11 ... a5! 12 b5 Nf5 with the idea 13 Bxf5 exf5 14 0-0 Be6=. (a2) 7 Bd3 is the most common reply: 7 ... fxe5 (7 ... c4 8 Bc2 g6!? is also playable) 8 dxe5 (8 Nxe5 Nf6 9 0-0 Bd6) 8 ... c4! 9 Bc2 Nh6 (or 9 ... g6) 10 0-0 Nf7 11 b3 (11 Qe2 g6!? 12 Ra2 Bg7 13 Bf4 0-0 14 Nbd2 Qc7 15 Re1 b5 with complications) 11 ... cxb3 12 Bxb3 Be7, Dvoirys-Murey, Dieren 2000. Here Collins suggests 13 Be3 Qc7 14 Bd4; Black has more than enough play following 14 ... 0-0; e.g. 15 Re1 Nh6 16 Nbd2 (or 16 Bc2 Nf5 17 Qd3 g6 18 Nbd2 Bd7 19 Nb3 Na5 20 Nxa5 Qxa5 with easy play) 16 ... Nf5 17 Nf1 Bd7 18 Ne3 Nfxd4 (or 18 ... Ncxd4 19 cxd4 Nxe3 20 fxe3 Rac8) 19 cxd4 Be8 intending ... Bh5. (b) 6 ... Nge7 can be used to avoid Bxh6 ideas (e.g. 6 ... Nh6 7 b4 cxd4 8 Bxh6), although that isn’t necessary. Then 7 b4 cxd4 8 cxd4 Nf5 is our main line 2.21 below. White has two ways to deviate: 7 dxc5 (7 Bd3!? can also be unique, since 7 ... cxd4 8 cxd4 Bd7 9 Nbd2!? Nxd4 10 Nxd4 Qxd4 11 Nf3 Qb6 12 Be3 can’t be met by ... Bc5, as is the case with Black’s knight on h6; nevertheless, White will have to work hard if he is to prove full compensation for the pawn) 7 ... Qc7! (7 ... Qxc5 8 b4 Qb6 9 c4!+/=)
8 Bb5 (Sveshnikov prefers this pin; 8 b4 can be met by 8 ... a5, 8 ... b6!? or 8 ... Ng6 9 Be3 Be7; Zlotnik’s pre-computer solution was 8 Nd4 Qxe5+ 9 Be2 ‘with initiative’; then one idea is 9 ... Ng6 10 b4 Nf4, or here 10 ... a5 11 b5 Nb8; finally, the attempt to protect e5 by 8 Bf4!? is uninspiring: 8 ... Ng6 9 Bg3 Bxc5 10 b4 Be7 11 Qe2 Qb6 – or here 11 ... h5 12 h4 Bd7 – 12 Qc2 Qc7 13 Qe2 Qb6 14 Qc2 f6!? 15 exf6 Bxf6 16 Bd3 Nce7 17 0-0 0-0=/+ with a central majority, Hertneck-Psakhis, BadenBaden 1992) 8 ... Ng6 (or 8 ... Bd7 9 0-0 Nxe5 10 Bxd7+ Nxd7 11 b4 b6 12 cxb6 axb6=) 9 Qd4 Be7 10 0-0 0-0 11 Bxc6, EndersT.Meister, German League 1996; and although 11 ... bxc6 12 b4 a5 should be alright, 11 ... Qxc6 is more interesting; e.g. 12 Be3 (12 b4 f6) 12 ... b6! 13 cxb6 axb6 with the idea 14 Qxb6 Qxb6 15 Bxb6 Ba6 16 Re1 Rfb8=. 2.16 ... Bd7 In this system Black brings a rook to c8 to attack the queenside. The main idea is to exploit the light-square weaknesses caused by a3 (and b4).
7 b4 7 Be2 develops soundly, yet the early commitment of the bishop allows for easy equality: 7 ... Nh6!? (7 ... Nge7 is simpler, avoiding Bxh6 and intending 8 dxc5 Qc7) 8 b4 cxd4 9 Bxh6 (9 cxd4 Nf5) 9 ... gxh6 10 cxd4 Rg8! compares well with other ... Rg8 lines because Black gains a tempo and can get to g4 with attack on d4; for example, 11 0-0 (11 g3 Rg4 12 h3 Rxd4 13 Nxd4 Nxd4 14 Nc3 a5=/+) 11 ... Rg4 12 h3 Rf4 13 g3 Rxf3 14 Bxf3 Qxd4 15 Qxd4 Nxd4 16 Bh5 a5 (16 ... Bg7!? 17 f4 Ke7 18 Nc3 f6 19 exf6+ Bxf6=) 17 Nc3 Rc8=. 7 ... cxd4 8 cxd4 Rc8
This system for Black is arguably riskier for him than the one with 6 ... Nh6, because in the latter case Black develops his kingside pieces rapidly and easily get to safety by ... 0-0 and ... f6. On the other hand, the play after 8 ... Rc8 tends to be rather forcing, and full of ways for White to go wrong. Assuming that you’re prepared, you’re unlikely to run into a surprise. Here White has three logical moves: 2.11 2.12 2.13 2.11
9 9 9 9
Be2 Bb2 Be3 Be2
Safe, but a little committal, since White might prefer the bishop on d3 in some lines.
9 ... Nge7! 9 ... a5!? 10 b5? runs into the standard idea 10 ... Nxd4! 11 Nxd4 Rxc1 12 Qxc1 Qxd4 and 13 Nd2 Qxe5 14 Nf3 Qb8 or 13 Qc3 Bc5 14 0-0 Ne7 15 Ra2 Qxc3 16 Nxc3 Bd4 17 Na4 Bxe5, Komliakov-Nevednichy, Kishinev 1993; when Black’s two central pawns and bishop pair for the exchange give him the advantage. However, I think that the gambit 10 0-0! offers White enough and probably more than enough compensation for a pawn, as you can verify with analysis. Amazingly, the move 10 0-0 hasn’t been played in any master games from this position. Because of this, I think that 9 ... Nge7 is best. 10 0-0 Nf5 11 Bb2 Be7
12 Qd2 Others: (a) 12 Bd3!? Ncxd4!? (12 ... g6) 13 Nxd4 Nxd4 14 Qg4 Nb3 15 Qxg7 Rf8 16 Ra2 h5!? or 16 ... a5=. (b) 12 Kh1 h5 13 Nc3 Na5! 14 Na4 Qc6 15 Rc1 Nc4 with good play; compare lines with 9 Bb2. (c) 12 Qd3 Na5 13 Nbd2 Nc4 14 Nxc4 dxc4 prepares ... Bc6 or ... Qc6 with light-square control. 12 ... 0-0 A traditional approach. The modern 12 ... h5 is seen by transposition in 2.211, note ‘d2’ to 10 g4. 13 Rd1 13 Bd3 activates the bishop; there might follow 13 ... h6!? 14 Bxf5 exf5 15 Nc3 Be6=.
13 ... f6 14 Nc3 14 g4 can be answered by 14 ... Nh6 15 exf6 Rxf6 16 g5 Rxf3! 17 Bxf3 Rf8 with full counterplay; but 14 ... Nh4!? 15 Nxh4 fxe5 16 Nf3 e4, with the idea 17 Ng5 Ne5 or 17 Ne1 Bh4, is more fun and dynamically balanced. 14 ... fxe5 15 dxe5 Be8! 16 Rac1 Bh5= Sveshnikov-Lputian, USSR Team Ch. 1990. 2.12 9 Bb2 This is the main line of our 6 ... Bd7 variation and has loads of theory behind it. Here 9 ... Na5 is the most complex line. It’s fully satisfactory, but I’ll give two simpler options so that you can pick something with which you feel confident: 2.121 2.122 2.121
9 9 9
... ... ...
Nge710
Nge7 Nh6 Nc3
Na5!
11 Nd2 Obviously 11 bxa5? Qxb2 is weak. McDonald likes 11 Na4, giving 11 ... Qc6 12 Nc5 Nc4 13 Qb3 and now 13 ... b5(?) allows 14 a4. But why not 13 ... b6! 14 Nxd7 Qxd7, or even 13 ... Ng6 ? I think Black has equal chances in those cases. Another alternative is 11 Rc1 Nc4 12 Bxc4 dxc4 13 0-0 Nd5= 14 Nxd5?! exd5 15 Bc3 Be7 16 Ne1, Shirov-Socko, Ajaccio 2005; and 16 ... Bf5 is the simplest of several moves that give Black a nice advantage. 11 ... Nc4 12 Bxc4 dxc4 13 0-0 Nd5 14 Nxd5?! 14 Nde4 Be7 15 Qg4 g6 16 Rae1 h5 17 Qg3 0-0 intending ... Bc6. 14 ... exd5=/+ Polovnikova-Radziewicz, Istanbul 2003. 2.122 9 ... Nh6 This is a sound developing move illustrating some characteristic themes: 10 Nc3 10 Bd3 Na5! 11 0-0 Nc4 12 Bxc4 Rxc4 13 Nc3 (13 Nbd2 Rc8 14 Rc1 Rxc1 15 Qxc1 Be7=/+ with the bishop pair and a solid position) 13 ... Qa6 (13 ... Nf5!=/+) 14 Rc1 Be7 15 Nd2?!, Korchnoi-Kotsur, Istanbul Olympiad 2000; and here Sveshnikov analyses the straightforward 15 ... Rxd4 16 Ne2 Rh4 17 g3 (17 Nd4 Ba4 18 Qf3 0-0) 17 ... Ba4 18 Qe1 Bb5 19 Rc8+ Bd8 20 gxh4 Bxe2 21 Qc1 0-0 ‘and it is White who is struggling for equality’. 10 ... Na5!
This is a standard type of position in lines with Bb2, whether after 6 ... Bd7 or 6 ... Nh6. Black continues his light-square quest, eyeing c4 and a4. If, after ... Nc4, White captures and Black recaptures with a pawn (often forced), he will gain the d5square and wonderful a8-h1 diagonal. From White’s point of view, that recapture grants access to e4 for his pieces (a knight can be powerful there) and potentially allows the move d5, opening up the centre. 11 Na4 White achieves nothing after 11 Bd3 Nc4 12 Qb3 Nf5. 11 ... Qc6
12 Nc5 12 Rc1 Nc4 13 Bxc4 dxc4 14 Nc3 Be7!? (or 14 ... Nf5! 15 0-0 b5=) 15 0-0 0-0 (15 ... Nf5!?) 16 d5!? (16 b5 Qb6 17 a4 a6 18 d5 Bc5=) 16 ... exd5 17 Nd4 Qg6 18 Nxd5, Dur-Damjanovic, Graz 1979; 18 ... Bg5! 19 h4 (19 f4 Bd8=/+) 19 ... Bd8=. 12 ... Nc4 13 Bxc4 dxc4 14 0-0 Qd5! 14 ... b6 15 d5!? Qxd5 16 Qxd5 exd5 17 Nxd7 Kxd7. 15 Qe2?! 15 Na4 Bxa4!? 16 Qxa4+ Qd7 with the idea 17 Qxa7 c3 18 Bc1 Nf5. 15 ... Bc6 16 Rfe1 Be7 17 Rac1 Korchnoi-Iruzubieta, Oviedo 1992. Here 17 ... 0-0 was good; but 17 ... b6! 18 Ne4 0-0 was better still.
2.139 The bishop is more actively placed here than on b2, but it is also subject to attack from a knight on f5 or g4.
Be3
9 ... Nh6 Alternatively,9 ... a5 10 b5 Nce7 clears the c-file and prepares to secure occupation of f5: 11 Bd3 (11 Qa4 Nf5 12 Bd2 f6! with the idea 13 Qxa5?? Rc1+ 14 Ke2 Bxb5+ or 13 Bxa5 Qa7) 11 ... Nf5 12 0-0 (12 Bxf5 exf5 13 a4 Bb4+ develops quickly and promises a long-term queenside attack) 12 ... Ngh6 (12 ... Nxe3 13 fxe3 Be7 14 Qd2 Nh6=) 13 Ra2 Be7 14 h3?! (versus ... Ng4, but it’s unnecessary), Stevic-Korchnoi, Novi Sad 2009; and 14 ... Nxe3! 15 fxe3 0-0 followed by ... f6 or ... f5 would be effective. 10 Bd3 Ng4 11 0-0 Nxe3 11 ... Be7 might gain a tempo if White plays h3, although he needn’t do so: 12 Nbd2 (12 Ra2 Nxe3 13 fxe3 0-0 14 h4?! f6!=/+ Fedoseev-Goganov, St Petersburg 2011) 12 ... Nxe3 13 fxe3 Nb8!? (intending ... Bb5, but 13 ... Nd8! with the same idea looks more precise) 14 Qe2 (14 Qb1 Bb5! 15 e4 Nc6! 16 Bxb5 Qxb5 17 Rd1 0-0 18 exd5 Qxd5=; 14 Nb3 Bb5 15 Nc5 0-0 16 Rc1 h6 with a minimal edge for White) 14 ... Ba4 and now: (a) 15 e4 Nc6!? 16 Qe3 0-0 17 Rab1, Motylev-Rychagov, Kolontaevo 1999; and instead of 17 ... a5, Black should play directly: for example, 17 ... g6= intending ... f6. (b) 15 Rac1 was Areshchenko-Pähtz, Gibraltar 2008; Black might do well continue 15 ... 0-0 16 Nb1! Rc6!? 17 Nc3 Qc7 18 Qd2 Rc8 with dynamic equality. 12 fxe3 Be7 13 Nbd2
13 Qe2 0-0 14 Nbd2 f6 15 b5 Nd8!? 16 exf6 Bxf6 leads to equality, according to Sveshnikov, but 15 ... Na5! looks even better. 13 ... Nb8! The idea is to play ... Bb5; also, White won’t be able to answer ... a5 with b5 in some lines. 14 Ne1 Others: (a) 14 Nb3 Ba4 15 Qb1 Bxb3 (15 ... Nd7! 16 Nc5 Nxc5 17 bxc5 Qc6 and ... b6) 16 Qxb3 0-0 ½-½ Stevic-Socko, Novi Sad 2009. (b) 14 Qe2 Rc3!? (with the idea ... Rxd3) 15 Rfb1 (15 Rfc1 Rxc1+ 16 Rxc1 a5!) 15 ... a6 16 Nb3 Bb5 17 Nc5 (17 Bxb5+ Qxb5 18 Qd2 Rc7 19 Rc1 Nc6 20 Nc5 0-0=) 17 ... Bxd3 18 Nxd3 Qb5 19 Rc1 Qxd3 20 Qxd3 Rxd3 21 Rc8+ Bd8 22 Rxb8 b6=. 14 ... 0-0 15 Rf3 Or 15 Qh5 g6 16 Qh6 f5 17 exf6 Bxf6=. After 15 Rf3, Arngrimsson-I.Johannesson, Reykjavik 2009, went 15 ... f5! 16 exf6 Rxf6!? (good enough, but 16 ... Bxf6! is safe and more thematic) 17 Nb3 Ba4 18 Qb1 Bxb3 19 Qxb3, and here 19 ... Qc7 is the easiest equalizer. 2.2
6
...
Nh6
There is an undeniable logic to developing rapidly versus the slow 6 a3. Black threatens ... cxd4 and ... Nf5, winning a pawn. He also plans straightforward development ( ... Bd7, ... Be7, ... Rc8), with ... f6 at some point to break up White’s centre. 7 b4 Instead: (a) 7 Bxh6? Qxb2 8 Bc1 Qxa1 9 Qc2 is unsound: 9 ... c4! intending ... Na5, ... Bd7, and ... Nb3. (b) 7 Bd3 is a rare move which loses time, but gives White the option of defending the d-pawn without needing to play b4 and either Bb2 or Be3: 7 ... cxd4 8 cxd4 Bd7 (for those who aren’t interested in learning the lines that follow, 8 ... Nf5 9 Bxf5 exf5 10 Nc3 Be6 is a standard position which we will look at briefly under the order 6 Be2 cxd4 7 cxd4 Nh6 8 Bd3 Nf5 9 Bxf5 etc; the difference here is the move a3, which somewhat weakens the light squares and probably hurts rather than helps White).
Now Black threatens the d-pawn. White has two tries: (b1) 9 Bc2 (protecting d4) 9 ... g6!? (Black wants to recapture with his g-pawn after ... Nf5 and Bxf5; instead, 9 ... Nf5 10 Bxf5 exf5 11 Nc3 Be6= is the position arrived at by 8 ... Nf5 in the previous note, with an extra move for each side) 10 Nc3 Nf5 11 Ne2 (11 Na4 Qc7 12 Bxf5 gxf5 13 Bf4 Rg8 14 g3 Rc8 15 Rc1 Qa5+ 16 b4 Qa6=) 11 ... Rc8 12 0-0 (12 Bxf5 gxf5 13 Nf4 Qb5 with ideas of ... Rg8, ... Na5, and ... Be7) 12 ... Be7 (12 ... Na5!?) 13 Bxf5!? gxf5 14 Nf4 Rg8 with double-edged play; e.g. 15 Re1 (15 b4 a5) 15 ... Na5 16 Re3 Bb5!? 17 Bd2 Nc4 18 Rc3 Kd7=, a familiar spot for Black’s king in the French Defence. (b2) 9 0-0!? (a strange form of Milner-Barry Gambit, trading the only moderately useful a3 for the developing move ... Nh6; this should favour Black, although he needs to be careful that the knight on h6 is not captured to White’s advantage) 9 ... Nxd4 10 Nxd4 Qxd4 11 Nc3 a6 12 Qe2?! (White has some compensation, probably not enough, after 12 Be3! Qh4 13 f3 Rc8) 12 ... Qg4!? (12 ... Qh4!=/+) 13 f3 Bc5+ 14 Kh1 Qh4=/+ 15 g3 Qh3 16 Qe1?! Bc6 17 Bf4 0-0-/+ with an extra pawn, Dvoirys-Lputian, Istanbul 2003. 7 ... cxd4
Now White chooses between two moves, both the subject of considerable theoretical investigation: 2.21 2.22
8 8
cxd4 Bxh6
2.218 And 2.211 2.212
before
we
even
get
cxd4 started, 9 9
a
final
parting
of
the
Nf5 ways: Bb2 Be3
2.211 9 Bb2 Bd7 There are some move order issues here. The text prevents 10 Bd3 because Black would capture on d4. Although 9 ... Be7 has been played by strong GMs, it allows Bd3, and some players prefer to have the e7-square for the knight to retreat to after White’s g4. Still, 9 ... Be7 has its own advantages and my choice here is to some extent a matter of taste.
10 g4 All the leading players and theoreticians treat this move as the most important one. Nevertheless, White’s options have a lot of practical value, and from them we can learn the typical themes associated with this variation: (a) 10 b5?! Na5 gives up the c4-square too easily: 11 Nc3 (11 a4 a6 12 Nc3 Bb4 13 bxa6 bxa6 14 Be2 Rc8-/+ HitzkerKaland, Pinneberg 1992) 11 ... Rc8 12 Na4 Qd8 13 Rc1 Rxc1 14 Qxc1 a6 (14 ... Nc4! 15 Bxc4 Qa5+=/+) 15 Nc5? (15 Nc3=/+) 15 ... Bxb5 16 Bxb5+ axb5 17 0-0 Nc4-/+ Marszalek-Beyen, Leningrad 1960. (b) 10 Qd3 Rc8! threatens ... Bxb4. The tactics tend to favour Black; e.g. 11 Bc3 a5!? (continuing to develop by 11 ... Be7 also secures the better game after 12 Be2 0-0 13 0-0 a6=/+, when White should avoid the natural 14 Nbd2? Nxb4! intending ... Bb5; 11 ... a6 has similar ideas) 12 b5
12 ... Ncxd4! (Black’s lead in development lets him get away with this) 13 Bxd4 (13 Nxd4 Rxc3 14 Nxc3 Qxd4 15 Qxd4 Nxd4=/+; e.g. 16 Bd3 Be7 17 0-0 Nb3 18 Ra2 Nc5 19 Rd1 f6) 13 ... Nxd4 14 Qxd4 Bc5 15 Qb2 Bxf2+ 16 Kd1, and a pretty line goes 16 ... f6! 17 exf6 0-0 18 fxg7? Rxf3 19 gxf3 Qe3 20 Be2 Bxb5!, winning. (c) 10 h4 is a modern treatment.
The idea is to play h5 and g4 to establish a large space advantage on the kingside. In the meantime, Black knight’s is denied access to g6 after a retreat to e7. Nevertheless, Black has well-placed pieces and good options: (c1) 10 ... h5 stops g4 and h5 while preparing queenside play; for example, 11 g3 (11 Be2 Rc8 12 0-0 Be7 13 g3 Na5) 11 ... Rc8 (intending ... Na5; 11 ... Be7 12 Bh3 g6 is also fine) 12 Qd2 (12 Bh3 a5 13 Bxf5 exf5 14 Nc3 axb4 with the idea 15 Nxd5?? Qb5 16 Nf4 bxa3 17 Bc1 Nb4) 12 ... Be7 (12 ... g6 13 Be2 Bh6 14 Qd3 Na5=/+) 13 Bh3 a5 14 Bxf5 exf5 and the b-pawn will fall, in view of 15 bxa5?? Nxa5 16 Bc3 Nb3 17 Qb2 Qb5 18 Ra2 Qd3 19 Qxb3 Qe4+ 20 Kf1 Qxf3 21 Kg1 f4 with an overwhelming position. (c2) 10 ... Rc8 (again, quick development is a reliable course) 11 g4 Nh6 (or 11 ... Nfe7 12 Nc3 Na5! 13 Na4 Qc6 14 Nc5 Nc4 15 Bc1, Movsesian-Haimovich, Panormo 2001, when 15 ... b6! 16 Nxd7 Qxd7 was a good way to establish an advantage) 12 Rg1 Na5!? 13 Nfd2?! (13 Nbd2 Nc4 14 Nxc4 dxc4 15 Rc1 a5=) 13 ... Be7 14 Bc3 Nc4 15 Nxc4 dxc4 16 Nd2 0-0!, Kozak-Antonio, Las Vegas 1999, with advantage due to 17 Nxc4? Rxc4! 18 Bxc4 Qc7-/+. (d) 10 Be2
White frequently plays this. Now 10 ... h5 has been played, but we’ll look at two equally good moves: (d1) 10 ... Be7 is solid; e.g. 11 0-0 0-0 and: (d11) 12 Qd2 f6 13 g4 (13 exf6 Bxf6 14 Ne5 Be8!-/+ S.Lalic-N.Berry, British League 2001) 13 ... Nh6 14 exf6 Rxf6 15 g5 Rxf3 16 Bxf3 Nf5 wins the d-pawn with a powerful attack against White’s weakened kingside. (d12) 12 Bd3 g6!= intending to recapture on f5 with the g-pawn; this is particularly safe when White’s bishop is offside on b2. (d2) 10 ... Rc8 11 0-0 Be7 transposes into a position we looked at in Section 2.11. After 12 Qd2, Black played the solid move 12 ... 0-0. But it’s also possible to play aggressively with 12 ... h5, intending to drive White back by ... g5-g4; e.g. 13 Rd1 (13 Rc1 g5!? 14 Nxg5 Ncxd4 15 Rxc8+ Bxc8=)
13 ... g5! 14 Ne1(14 Nxg5? Ncxd4 15 Bxd4 Nxd4 16 Qxd4 Qxd4 17 Rxd4 Bxg5-/+ 18 f4 Bd8!, threatening ... Bb6, when Black has the two bishops and control of the c-file; 14 Nc3?! g4 15 Ne1 Grosar-Tukmakov, Bled 2001, and now simply 15 ... Ncxd4 16 Nxd5 Nxe2+ 17 Qxe2 exd5 was strong, since White nearly has to enter the variation 18 e6 Qxe6 19 Qxe6 Bxe6 20 Bxh8 f6 21 Rac1 Rb8! followed by ... Kf7 trapping the bishop; 14 Bc3 a6 15 Ne1 f6! 16 Nc2 Sevillano-Clawitter, Los Angeles 2012, 16 ... fxe5 17 dxe5 g4 with good prospects) 14 ... g4 15 Nc2 (15 Bc3 a6 16 Nc2 Rg8 17 a4 Bg5=/+ Savic-Kosic, Vrsac 2000) 15 ... Na5! 16 Ra2 Nc4 17 Bxc4 Rxc4=/+ Gruzmann-Goczo, Budapest 1998. (e) 10 Nc3?! Nfxd4 11 Nxd4 Qxd4 (11 ... Nxd4 12 Nxd5) 12 Qxd4 Nxd4 13 Nxd5 Nc2+ 14 Kd2 Rc8 15 Rc1 Nxa3 favours Black; e.g. 16 Rxc8+ (16 Nc7+ Kd8 17 Nxe6+ Bxe6 18 Rxc8+ Kxc8 19 Bxa3 Be7=/+ with the idea ... a5/ ... Kb8 and ... Rc8 or ... Rd8 – Vitiugov) 16 ... Bxc8 17 Bxa3 exd5 18 Bd3 Be7=/+.
Returning to 10 g4, we arrive at a key position for theory. Both retreats are considered good. 10 ... Nh6
With this move, Black not only targets g4 and wins time, but keeps a diagonal open for his f8-bishop. A central counterattack with ... f6 can increase in force because the h6-knight can come to f7, with pressure on e5, and gain access to the influential square d6 should the e5-pawn be exchanged on f6. There are other positive features, too. The two main disadvantages of 10 ... Nh6, as opposed to 10 ... Nge7, are that the knight doesn’t control a central square (in many variations the d5-pawn disappears and a knight would like to go there), and White’s weaknesses on f4 and h4 are not targeted (as they would be after ... Nge7-g6). Fortunately, 10 ... Nh6 doesn’t require a great deal of theory to play well, although, as in all openings, knowing some theoretical positions is useful. If for some reason 10 ... Nh6 doesn’t work out, you should know that 10 ... Nge7 is also fine. Black intends moves such as ... h5 (to regain the f5-square) and ... Ng6 (probing the weakened f4 and preparing ... Be7/ ... 0-0/ ... f6). Black has a number of other ideas in the air such as ... Na5-c4 (upon a capture on c4, the e7-knight can go to d5), ... Nc8-b6, and ... a5. The main problem is that 10 ... Nge7 has become theoretically dense and requires more memorization. 11 h3 White loses a pawn after 11 Bh3? a5!. Instead, 11 h3 has been the main move, although in the last couple of years, several grandmasters have played 11 Rg1.
That includes Sveshnikov (the king of 6 a3 theory and practice), who says: ‘I do not think that 11 h3 is in any way better than this move’. The logic is that White retains the move h4, prepares g5 in case of an ... f6 break, and can swing the rook into
action via g3. The price for these opportunities is the loss of castling (a handy escape measure) and the fact that h4 is less well defended. Black can simply develop with 11 ... Be7 and ... Rc8, before deciding upon the best response to White’s set-up. But here are two more direct and interesting challenges: (a) 11 ... Rc8 (Sveshnikov likes this and gives it an ‘!’) 12 Nc3 Na5 13 Na4 (Sveshnikov gives 13 Nd2 Be7 14 Na4 Bxa4 15 Qxa4+ Nc6; this looks equal) 13 ... Qc6!
14 Nc5 (14 Rc1 Nc4 15 Bxc4 dxc4 16 Nc5 b6 17 Nxd7 Qxd7 and ... b5 with good prospects, Sigfusson-Sigurjonsson, Reykjavik 2004) 14 ... Nc4 15 Bc1 (15 Bc3 b6 16 Nxd7 Qxd7 17 Bd3 is unclear, according to Sveshnikov; Black has time for ... Ng8-e7, ... g6/ ... Bg7 and ... 0-0) 15 ... Ng8 16 Bd3 (16 Rg3 – Sveshnikov; then 16 ... Bxc5! 17 bxc5 b6=/+ retains Black’s structural advantage) 16 ... Bxc5 17 dxc5 (17 bxc5 b6 18 cxb6 axb6=/+) 17 ... b6 (or 17 ... Ne7! 18 Kf1 b6) 18 Bxc4 ½-½ Sveshnikov-G.Dizdar, Slovenian Team Ch. 2005; Black has a comfortable advantage following 18 ... dxc4 19 Be3 Qb7! 20 Kf1 Ne7, when White is suffering on the light squares; ... h5, ... Bc6, and ... Nd5 are among the ways Black can make progress. (b) 11 ... f6 12 exf6 gxf6 13 Nc3 Nf7
This popular position is unbalanced and unclear. Black has a central majority and apparently easy development, but White has more territory and queenside pressure beginning with Na4-c5. A couple of snippets from this much-discussed position: (b1) 14 Rc1 Bh6 15 Na4 (15 Rc2 Ne7 16 h4 Bf4!? 17 Bd3 h6 18 Re2, Lautier-C.Bauer, French Ch. 2002, shortly drawn; Black could have spiced thing up with 18 ... 0-0-0!) 15 ... Qd8 16 Rc3 (16 g5 – Sveshnikov; but he doesn’t look at the simple 16 ... fxg5!) 16 ... b6 17 h4 Ne7! 18 g5 (18 b5 0-0 19 g5 fxg5 20 hxg5 Bg7) 18 ... fxg5 19 Nxg5 Nf5!.
(b2) 14 Na4 has been tested a lot: 14 ... Qc7 15 Rc1 Qf4! 16 Nc5 (Li Shilong-Goh Wei Ming, Singapore 2006, went 16 Rc3 and 16 ... b6! was equal, but Black speculated by 16 ... Nxb4!? 17 axb4 Bxb4 18 Nc5!? – 18 Be2+/= – 18 ... Rc8 19 Be2, and now Li Shilong suggested 19 ... b6! 20 Nd3 Bxc3+ 21 Bxc3 Qc7!=) 16 ... Bxc5.
Both of White’s recaptures have merit: (b21) 17 Rxc5 Ng5 18 Nxg5 (18 Be2=) 18 ... fxg5 19 Bc1?! (19 Qd3 0-0 20 Qe3= Wei Ming) 19 ... Qxd4 20 Qxd4 Nxd4 21 Rg3, Sveshnikov-Malakhatko, Moscow 2007; and 21 ... Rf8 22 Bg2 b6! 23 Rc7 Nb5 24 Rc2 a5=/+ was one way to secure a limited advantage. (b22) 17 dxc5 Nfe5!? (17 ... Nce5! is somewhat more accurate, when 18 Rg3 Nxg4 is equal, as opposed to 18 Nxe5 of Short-Lputian, Batumi 1999, when McDonald notes that 18 ... fxe5! is good; 17 ... Rf8!? should also be considered) 18 Nxe5?! (18 Rg3! Nxg4!? 19 h3 Nge5 20 b5!+/=) 18 ... fxe5 19 Qd2 (19 Rg3? 0-0 20 Rc2 Nd4! 21 Bxd4 exd4, Haznedaroglu-Malakhatko, Panevezys 2008) 19 ... Qxd2+ 20 Kxd2 0-0 21 Ke1, Sveshnikov-Berend, Turin Olympiad 2006, and 21 ... Nd4! 22 Bxd4 exd4 followed shortly by ... e5 affords Black the better prospects. 11 ... f6 Black can also use piece play to make progress: 11 ... Rc8 12 Nc3 Na5 13 Na4 Qc6 14 Rc1 (14 Nc5 Nc4 15 Bc1 Ng8 16 Bd3 Bxc5 17 dxc5 b6=/+ Haba-G.Dizdar, German League 1991) 14 ... Nc4 15 Bxc4 (15 Rxc4?! dxc4 16 Nc3 Qb6 17 Bg2 a5-/+ ShirovSadvakasov, Astana 2001) 15 ... dxc4 16 Nc5 Ng8!? (16 ... b6 17 Nxd7 Qxd7=) 17 Nxd7 Qxd7 18 Nd2 Qd5 19 0-0 h5 20 gxh5 g6! with excellent squares for Black’s pieces.
12 Nc3
Others: (a) 12 Bd3 fxe5 (12 ... Nf7 13 Nbd2 a5! 14 b5 Ne7 15 a4 Ng6 16 Bxg6 hxg6 was unclear in Wiersma-Malakhatko, Belgian Team Ch. 2005, but I think at least equal for Black) 13 Nxe5?! (13 dxe5 Nf7 14 0-0 Be7 15 Nbd2 Rc8 16 Nb3 0-0 17 Rc1 a5 18 Nxa5 Nxa5 19 Rxc8 Rxc8=, or here 18 b5 a4! 19 bxc6 bxc6) 13 ... Nxe5 14 dxe5 Nf7 15 0-0, Hollrigl-Van der Weide, Lichtenstein 1999; and Black should get some advantage from 15 ... Bc5!? (with the ideas ... Bd4 or ... Ng5) 16 Be2 Ng5 17 Kg2 0-0 18 h4 Ne4=/+. (b) 12 exf6 gxf6 13 Bd3 (13 Nc3 Nf7 transposes to the next note) 13 ... Nf7 14 Nc3! (14 Nbd2?! a5 15 b5 Ne7 16 a4 Ng6=/+) 14 ... Nxd4 (or 14 ... h5 15 gxh5 Rxh5 16 Na4 Qd8=) 15 Nxd5! Nxf3+ 16 Qxf3 exd5 17 Bxf6 Be7 18 Bxh8 Qe6+! 19 Qe2 Qxe2+ 20 Kxe2 Nxh8 21 Bxh7 Kf7 is unclear, with Black’s strong bishops balancing out White’s rook and passed pawns; e.g. 22 f4 Ng6 23 Bxg6+ Kxg6 24 Kf3 d4!? 25 Kg3 Ba4 26 h4 Rc8 27 h5+ Kh7 28 g5 Rc3+ 29 Kg4 d3.
12 ... fxe5 12 ... Nf7 13 exf6 (13 Na4 Qd8 14 Bd3 fxe5 15 dxe5 Ng5!=) 13 ... gxf6 14 Na4 Qc7 (to control the weak square f4 in conjunction with ... Bd6) 15 Nc5 Bxc5 16 dxc5 (16 bxc5 0-0-0 intending ... e5) 16 ... Qf4
17 Be2 Nfe5 (with mutual chances: White’s bishop pair is offset by Black’s 2:0 central majority) 18 Nxe5 (18 Rg1 a5! 19 b5 Nxf3+ 20 Bxf3 Ne5=) 18 ... fxe5 19 Bc1! Qf6 (19 ... Qd4 20 Bh6! Qc3+ 21 Kf1 is absolutely unclear; the bishops-versus-centre battle continues) 20 0-0!? (risky-looking, but Black has no immediate way to open lines) 20 ... Nd4 21 f4 0-0!? with complications. 13 dxe5
13 Na4 Qd8 14 Nxe5 Nxe5 15 dxe5 Be7 gives Black good dark-square control. 13 ... Be7 Or 13 ... Nf7 14 Na4 Qd8. 14 Na4 Qd8 15 Rc1 0-0 16 Nc5 Bxc5 Or 16 ... Nf7; e.g. 17 Bg2 Bxc5 18 Rxc5 Qb6. 17 Rxc5 Nf7 17 ... Qe7 intends ... a5 or ... Nf7-g5. Then, for example, 18 Rc1 Rf4=, or 18 ... Nxb4!? 19 g5! Nf5 20 axb4 Nh4! is unclear. 18 Bg2 b6 Or 18 ... Qe7 19 Rc1 Nh8! with the idea ... Ng6-f4. 19 Rc1 Likewise 19 Rc2 Ne7 20 Qd2 Nh8 intending ... Neg6-f4/ ... Nhg6. 19 ... Ne7 20 Qd2 Ng6 21 h4?! Rather slow, although 21 0-0 Nfh8! followed by ... Nf4 and ... Nhg6 is hardly desirable for White. 21 h4 is Sveshnikov-Lputian, Sochi 1993, which continued 21 ... Nh6! 22 g5 Ng4 23 Bh3 Rxf3!? (after 23 ... Rf4! Black has the advantage because of his outposts) 24 Bxg4 Rf4 25 f3, and Black would still have the better of it after 25 ... Rf8. 2.212 9 Be3 White tries to shore up his centre at the cost of the bishop pair. This is less important than 9 Bb2, but is popular enough to warrant close attention.
9 ... f6 The most irritating move for White, because he has to choose between a line in which Black has a drawing option, which is not an inspiring result, or try to get more, although that can easily lead to a disadvantage. Black has legitimate alternatives at this point: (a) 9 ... Nxe3 10 fxe3 g6 is a natural approach, still thinking about ... f6 but with an eye on the queenside as well; for example, 11 Bd3 Bh6!? (I actually prefer 11 ... Bd7 first, waiting until White castles so that h4-h5 has less effect – this will usually transpose, but 12 Qe2 Rc8 13 0-0 Bh6=, say, is a good move order) 12 Qe2 0-0 13 g4!? (a critical attempt to cramp Black on the kingside; 13 0-0 is easily met by 13 ... Ne7 14 Nc3 Bd7=, whereas in Berg Jensen-V.Potkin, Chalkidiki 2000, Black went for the more ambitious 13 ... f6!? 14 exf6 Rxf6 15 Kh1 Bd7 16 Nc3, when the simple reorganization 16 ... Qc7 17 Rac1 Qd6 keeps the position level; 13 h4! is dangerous and best – it’s easiest to avoid this with the 11 ... Bd7 order) 13 ... f6!? 14 exf6 e5! 15 g5 (15 Nxe5 Nxe5 16 dxe5 Bxe3-/+) 15 ... e4 16 gxh6 exd3 17 Qxd3 (17 Qg2! Rxf6 18 Nc3 Qd8=) 17 ... Rxf6 18 0-0 Bh3 19 Rf2 Raf8 20 Nbd2 Nd8!=/+ intending ... Nf7. (b) To create a more positional contest, Black may want to try to block the kingside by ... f5, or at least tailor the ... f6/ ... f5 decision to White’s set-up: 9 ... Be7, for example, and in White’s array of replies we see many of the characteristic ideas of this line.
(b1) 10 b5?! Na5 intends a quick ... Bd7, ... Nc4, and ... Rc8. (b2) 10 g4?! allows 10 ... Nxe3 11 fxe3 f6!=/+. (b3) 10 Nc3 f6!? 11 Na4 Qc7 12 exf6 Bxf6 13 Rc1 Qd6 14 Bd3 Nxe3 15 fxe3 0-0 16 b5 Ne7=. (b4) 10 Nbd2 Nxe3 11 fxe3 f5 and now: (b41) 12 exf6 Bxf6 13 Bd3 (13 Ng5? Nxd4! 14 exd4 Qxd4, winning) 13 ... e5! 14 b5! (14 dxe5 Nxe5 15 Nxe5 Qxe3+ 16 Qe2 Qxe5-/+) 14 ... Nd8 15 dxe5 Qxe3+ 16 Qe2 Qxe2+ 17 Bxe2 Be7 18 0-0 Ne6 with the bishop pair and good play. (b42) 12 Bd3 0-0 13 Nb3 (13 0-0 a5 14 b5 Nd8 15 a4 Bd7 16 Nb3 Nf7 17 Qe2 Rac8 18 Rfc1 Bb4) 13 ... a5 14 b5 a4! 15 Nbd2 Nd8 16 0-0 Bd7 17 Qb1 Nf7 18 Rc1 g5!, intending ... g4 and ... Ng5 or ... h5-h4. (b5) 10 Bd3 Nxe3 11 fxe3 f5!? (or 11 ... Bd7 12 Nc3 Rc8 13 Na4 Qd8) 12 exf6 (12 0-0 a5 13 b5 Nd8 14 Nc3 Nf7=) 12 ... Bxf6 13 Nc3 (13 Ng5 Bxd4! 14 exd4 Qxd4 15 Qh5+ Ke7! 16 Qf7+ Kd6–+; 13 0-0 a5! with the idea 14 b5? Nxd4!) 13 ... 0-0 14 0-0 g6 15 Qc2, M.Carlsen-Lahlum, Gausdal 2001; and 15 ... Ne7! followed by ... Bd7 and ... Rac8 is equal.
10 Bd3 A direct move which virtually forces Black into sharp lines, though it turns out that the alternatives are also extremely complex: (a) 10 b5 Nxe5! 11 dxe5 Nxe3 12 fxe3 Qxe3+ (12 ... fxe5!? can be used to avoid the draw, with the idea 13 Nxe5?? Qxe3+ 14 Qe2 Qc1+ 15 Qd1 Qb2 – this is obviously risky and objectively ought to be a bit better for White, although Black can use his bishops and centre to generate play; e.g. 13 e4! Bc5 14 Qd3! Bf2+ 15 Kd1 0-0 16 Nbd2 Rd8 intending ... Bd7-e8-g6; here 13 ... dxe4 14 Nfd2 Bc5 15 Nxe4 Bd4 is another way to speculate) 13 Qe2 (13 Be2? fxe5 is bad, leaving Black with two bishops and the
centre). Now Black can either accede to a draw or take chances:
(a11) 13 ... Qc1+ is a draw following 14 Qd1 Qe3+ 15 Qe2 Qc1+ etc. In this line, Cumbers has suggested 14 ... Qb2 15 Nbd2 fxe5 (the quick-developing 15 ... Be7!? may be better) 16 Rb1 Qxa3 17 Nxe5, when to avoid a draw Black might play 17 ... Qc3 18 Nd3! (18 Nef3 Bd6 19 Be2 Qc5!) 18 ... Bd6!? (or 18 ... g6) 19 Qh5+ (19 Qf3 Rf8 20 Qh5+ g6 21 Qxh7 Qf6 22 Qh6 e5=) 19 ... g6 20 Qh6 Qf6 about equal. (a12) 13 ... Qxe2+!? 14 Bxe2 fxe5 15 Nxe5 Bd6 gets two pawns for the piece and prepares an ... e5 advance. Again, this would probably be played only if Black wants to avoid a draw. There might follow 16 Nf3 0-0 17 0-0 Bd7 18 Nbd2 Rac8, intending ... e5. Black only has two pawns for the piece but they are mobile passed central pawns and his bishop pair can be effective. (a2) 10 exf6 gxf6 gives up White’s central advantage in return for quick development. Now not 11 Qd2? Nxb4!, but: (a21) 11 Bd3 Nxe3 12 fxe3 Bh6 (12 ... e5 13 Nc3 Be6 is playable, as is 12 ... Bd7 intending ... Rc8) 13 Qe2 a5 (13 ... 0-0 14 Nc3 Ne7!) 14 b5 Ne7 15 Nc3 0-0 16 0-0 e5= Maximov-Sengupta, Pardubice 2011. (a22) 11 Nc3 Nxe3 12 fxe3 Bh6 has been seen a fair amount; for example, 13 Qd3 (13 Qd2 0-0 14 Bd3 Ne7 15 Na4!? – to avoid 15 0-0 Nxd4! – 15 ... Qc6= Nilsson-Fernandez Romero, Dos Hermanas 2002; 13 Kf2 0-0 14 Bd3?! Ne5!=/+) 13 ... 0-0 (or 13 ... a5 14 Na4 Qd8 15 b5 Ne7 16 Kf2 0-0 17 g3, Velitshko-Vyoshin, Zaporozhie 2010; and now 17 ... e5!) 14 Be2 Bd7 15 Na4 (15 00 Ne7!) 15 ... Ne5! 16 Nxb6 Nxd3+ 17 Bxd3 axb6 18 Ke2, Gruic-Arkhipov, Belgrade 1990, and 18 ... e5=/+ is simplest. 10 ... Nxe3 11 fxe3
11 ... fxe5 11 ... Bd7!? is a sound option: 12 exf6 gxf6 13 Nc3 Bh6 14 Qe2 Ne7 (14 ... 0-0 is simpler) 15 0-0 (15 Ne5 fxe5 16 Qh5+ Kd8 17 Qxh6 Rc8 was fine for Black in Jonkman-San Segundo Carrillo, Mondariz 2000) 15 ... 0-0 16 Kh1 Rac8 17 Nd1?! Bg7 18 Nf2 e5 (or 18 ... Rc3!) 19 dxe5 fxe5 20 e4 d4=/+ Morozevich-Bareev, Monaco (blindfold rapid) 2002. 12 b5 Nxd4!? Oddly, 12 ... e4 is generally queried, but looks satisfactory after 13 bxc6 exf3! 14 Qxf3! bxc6 15 Nd2 Qa5 16 Ra2 Ba6, which is unclear but balanced. 13 exd4 e4 14 Bxe4! Rather than 14 Ne5? Qxd4, or 14 0-0? exf3 15 Qxf3 Qxd4+ or 15 Rf2 Bd6 with a winning position. 14 ... dxe4 15 Ne5
15 ... Bd7 Again Black has alternatives that equalize: (a) 15 ... g6 16 0-0 (16 Nc3 Bg7 17 Nxe4 0-0 and White can’t castle, while 18 Nf3 Qxb5 means trouble) 16 ... Bg7 17 Kh1 Bxe5 18 dxe5 Qxb5 19 Nc3 Qxe5 20 Qb3 (20 Qc2 Rf8 21 Nxe4 Bd7 22 Rfe1 Bc6 is dynamically equal) 20 ... b6 21 Rad1, HurleyT.Clarke, Irish Ch. 1996; and now Cumbers and McDonald analyse 21 ... Bb7, when I think White should bail out with 22 Qa4+! b5 23 Nxb5 Bc6=; e.g. 24 Nc7+! Qxc7 25 Qc4 Qc8 26 Qd4 Rf8 27 Rxf8+ Kxf8 28 Qh8+ Kf7 29 Qxh7+ Ke8, which leads to a perpetual. (b) 15 ... e3 16 Nc3 e2! is less tested, when White went wrong by 17 Qd3?! Bd6 18 Kxe2 Bxe5 19 dxe5 Bd7=/+ in
Filimonov-Eraschenkov, Sukhumi 2007. Better was 17 Nxe2! Qa5+ 18 Qd2 Qxd2+ 19 Kxd2 Bd7= with ... Bd6 next. 16 0-0?! Not 16 Qh5+? g6 17 Nxg6 0-0-0! 18 Nxh8 Qxd4 19 Ra2 Bxb5! and wins. But 16 Nc3! Rc8 17 Qh5+ g6 18 Nxg6 hxg6 19 Qxg6+ draws after 19 ... Ke7 20 Qg5+ Ke8 21 Qg6+ Ke7 22 Qg5+. 16 ... 0-0-0 17 Nc3 This is Potkin-Filippov, Togliatti 2003. The game continued 17 ... Bc5?! 18 Na4 Bxd4+ 19 Kh1, leading to equality. Instead, 17 ... Be8! secures Black’s advantage; e.g. 18 Ne2 Kb8 19 Kh1 Bd6 20 Nf7 Bxf7 21 Rxf7 e5! (or 21 ... Rhf8 22 Rxg7 e5!-/+) 22 d5 (22 Rxg7? exd4 is killing) 22 ... Rhf8! (or 22 ... e3 23 Rxg7 Rhf8 followed by ... Rf2) 23 Rxg7 Rf2 and White has insufficient resources; for example, 24 a4 e3 25 Rxh7?! (25 a5 Qxb5 26 a6! Ka8! 27 Rxb7 Qc4 28 Ng3 Rdf8 with ... e2 and/or ... Qxd5) 25 ... e4 (25 ... Bc5, threatening ... Qg6, also wins) 26 Rc1 Be5 27 d6 Qxd6 28 Qxd6+ Bxd6 29 Nc3 e2 30 Rg1 Bb4, winning. 2.22
8
Bxh6
gxh6
9
cxd4
This is a fascinating line in which White fractures the black kingside, and does so by exchanging his bad dark-squared bishop, which can be a problem piece. On the other hand, that piece can be an important defender of d4 and e5, exemplifying the maxim that ‘bad bishops protect good pawns’. Black’s position has its own trade-offs: he has to deal with weaknesses on f6, h6, and h5, but he gets the bishop pair and is able to develop actively. His initial plan is to play ... Bg7, ... 0-0, and ... f6, contesting those dark squares which can no longer count upon the support of a white bishop. Moskalenko calls 8 Bxh6 ‘weak’, but Sveshnikov likes it. In the end, I suspect that this capture and 8 cxd4 are worth about the same. 9 ... Bd7 I like the fact that this move prevents 10 Bd3 (due to 10 ... Nxd4) and it probably needs to be played anyway. 9 ... Bd7 also clears the c-file, so that 10 Nc3 can be met by 10 ... Nxb4! (see below, and note that the bishop hasn’t gone to g7 in this case). 10 Be2 Instead: (a) The natural move 10 Nc3? allows 10 ... Nxb4! 11 axb4 Bxb4
when Black’s attack on c3 has no adequate answer, even against a heroic defence, showing the value of delaying the f8bishop’s development. One line goes 12 Qb3 Rc8 13 Rc1 Qa5 14 Kd2 0-0 (threatening ... Rc7 and ... Rfc8, as well as ... f6 and ... fxe5 if needed) 15 Bd3 Ba4 16 Qb2 Rc7 17 Ng1 f6! 18 exf6 (18 f4 fxe5 19 fxe5 Qb6!) 18 ... Rxf6 19 Nge2 (19 Nf3 e5! 20 dxe5 Rb6) 19 ... e5! 20 f3 exd4 21 Nxd4 Qc5, winning; e.g. 22 Nde2 Re6. (b) Sveshnikov is enthusiastic about 10 Ra2 with the idea of playing Rc2 and Nc3 (partly to defang ... Nxb4), as well as defending f2 versus Black’s idea of ... a5/ ... Nxd4/ ... Bc5 (see below). White’s development is retarded, however, and several reactions look promising, including: (b1) 10 ... Rg8 is the somewhat offbeat approach, but it looks fine: 11 h3 (versus ... Rg4) 11 ... h5!. Black holds down White’s kingside pawns and frees h6 for his bishop: 12 g3 (12 Rc2 Nxb4!) 12 ... h4!? (12 ... a6 13 Be2 h4 14 g4 h5 15 g5 Qd8! 16 Nc3 Be7 17 Rg1 Bxg5 18 Qb1 Rg7 with unclear complications) 13 g4 (or 13 gxh4 h5 14 Be2 Ne7; e.g. 15 Nc3 Nf5 16 Na4 Bxa4! 17 Qxa4+ Kd8 or 15 Bd3 a5 16 b5 a4! 17 Nc3 Ng6) 13 ... h5 14 gxh5 Bh6!? (14 ... Ne7) 15 Qd3! Ne7 16 Qh7 Bf8 17 Nc3 a5=. (b2) 10 ... a6 11 Rc2 Na7 isn’t ambitious, but simplification gets rid of Black’s bad bishop; for example, 12 Nc3 a5 13 Be2 Nb5!? 14 Nxb5 Bxb5 15 Bxb5+ Qxb5 16 Qe2 Qxe2+ 17 Kxe2 axb4 18 axb4 Be7!? 19 Rc7 Ra7 20 b5 Bd8 21 Rc2 Kd7=. (b3) 10 ... Rc811 Rc2 is complex: 11 ... a5 12 b5 Na7 13 Rxc8+ Nxc8 14 Qc1 (14 Qb3 Qc7 15 Bd3 Qc1+ 16 Qd1 Qxd1+ 17 Kxd1 Nb6 18 Ke2 ½-½ SveshnikovMalakhatko, Warsaw 2010) 14 ... Rg8!? 15 h3 (versus ... Rg4). Sveshnikov calls this ‘+/=’; but Black has nothing to fear and obtains good chances by getting his queen to the c-file (and perhaps a knight to c4) with 15 ... Kd8! 16 Nc3 (16 Qb2 Qc7 17 g3 Nb6 18 Be2 Nc4=/+) 16 ... Qc7 17 g3 Nb6 (intending ... Na4) 18 Qb2 h5 19 Nd2 f6! 20 exf6 h4! 21 g4 Rg6=/+ with the idea ... Rxf6 and ... Qf4. 10 ... Rc8 Black again retains the ... Nxb4 idea while making a useful move. Instead, 10 ... Rg8 hasn’t been played much, but whenever Black can attack the g-pawn with tempo and thus get to g4, it should definitely be considered. Here Black comes out with full-fledged play following 11 0-0 Rg4! (threatening d4) 12 h3 (12 Nbd2 Nxd4 13 Bd3 Rc8! 14 h3 Rf4 – Sveshnikov) 12 ... Rf4 13 g3 Rxf3 14 Bxf3 Qxd4 15 Qxd4 Nxd4 16 Bh5 (16 Bd1 a5! 17 bxa5 Rxa5) 16 ... a5 ‘with compensation’ – Sveshnikov. The game is balanced after 17 Nc3 Rc8 18 Ne2! Nc6! 19 Rfc1 axb4 20 axb4 Rb8 21 Nf4! Nxe5 22 Nxd5 exd5 23 Re1 Bd6 24 f4 Be6 25 fxe5 Bxb4=. If Black is worried, he always has ... d4 and ... Bc3. Thus 10 ... Rg8 might prove to be a relatively easy solution to this line. 11 0-0 Bg7
12 Qd2!? Keeping an eye on the dark squares. Since moves by the b1-knight lose the d-pawn, here are the alternatives: (a) 12 Qd3 isn’t quite what White wants after 12 ... f6! 13 exf6 (13 b5 Nxe5! 14 dxe5 fxe5=/+ and ... e4 follows, R.DiazOrsini, Buenos Aires 2006) 13 ... Bxf6 14 Rd1 Nxd4! (or 14 ... 0-0 15 b5 Ne7 16 Nc3 Nf5 17 Rac1 Qd6 with easy play) 15 Nxd4 Ba4 16 Nc3 (16 Rd2? Rc4) 16 ... Bxd1 17 Nxd5! exd5 18 Qe3+ Kd8!? (18 ... Kf7 19 Rxd1 Rhd8 may give Black a small edge) 19 Ne6+ Kd7 20 Nc5+ Qxc5 21 bxc5 Bxe2 22 Qh3+ Kd8 23 Qg3! Bxa1 24 Qd6+= (Sveshnikov). (b) 12 b5 is White’s safest move. Then 12 ... Ne7 is playable, but seizing c4 by 12 ... Na5 is more aggressive; e.g. 13 a4 0-0 14 Nbd2 f6 15 Bd3 (15 Rc1 fxe5 16 Nxe5 Bxe5 17 dxe5 Qd4 18 Nf3 Qxd1 19 Rfxd1 Nb3! 20 Rxc8 Rxc8 21 Kf1 Be8= intending ... Bg6) 15 ... Rc3 16 Qb1 is assessed as better for White by Sveshnikov, but Black holds his own after 16 ... f5 17 Rc1 Rxc1+ (or 17 ... Rfc8) 18 Qxc1 Rc8 19 Qd1 Qc7=. 12 ... 0-0
This plans ... f6 and reserves tactics such as ... Nxd4 for later. The immediate 12 ... Nxd4!? is also interesting: 13 Nxd4! (13 Qxd4 Qxd4 14 Nxd4 Bxe5 15 Nb3 Bxa1 16 Nxa1 Ba4 17 Nd2 Ke7-/+) 13 ... Bxe5 14 Rd1 Ba4 15 Qe3! Bg7 16 Nc3 Bxd1 17 Rxd1 00. The material isn’t bad, but White’s minor pieces give him a small edge in a wide open game. 13 Ra2 This time 13 Nc3 Nxd4! works: 14 Nxd4 Bxe5 15 Rad1 Qxd4! (15 ... Bxd4 16 Nxd5 Bxf2+ 17 Rxf2 exd5 18 Qxd5 Ba4 was roughly equal in Fressinet-Korchnoi, Cannes 1996) 16 Nxd5 Qxd5 17 Qxd5 exd5 18 Rxd5 Bxh2+ 19 Kxh2 Rc7 20 Rfd1 Be6 21 R5d3=.
13 ... f6 14 b5 Ne7 Planning ... Ng6 with an eye on both e5 and f4. 15 exf6 15 Bd3 is well met by 15 ... Ng6 or 15 ... fxe5 16 Nxe5 Be8!. 15 ... Bxf6 Or 15 ... Rxf6 16 Nc3 Be8 (16 ... Rxf3!? 17 Bxf3 Qxd4) 17 Rc1 Rf8! with chances for both sides. 16 Bd3 Bg7 17 Nc3! 17 Ne5?! Be8 18 a4 Ng6 (or 18 ... Nf5 19 Bxf5 Rxf5 20 Re1 Bxe5 21 dxe5 Bg6=/+), Bonay Toscas-Rodriguez Amezqueta, ICCF server 2005; and here 19 Bxg6 hxg6 20 Qe3 Rf5=/+ was best. 17 ... Rxf3!? 18 gxf3 Qxd4 19 Ne2 Qh4 This is objectively equal; for example, 20 Rc1 (20 Qb4 Ng6) 20 ... Rxc1+ 21 Qxc1 Ng6 22 Qc7 (22 Bxg6 hxg6 23 a4 Be5 24 f4 Qg4+=) 22 ... Ne5 23 Qb8+ Kf7 24 Bxh7 Nxf3+ 25 Kg2 Qg4+ 26 Qg3 Nh4+ 27 Kf1 Nf3! 28 Kg2 Nh4+=.
Chapter
Three
Advance
Variation
with
5
...
Nh6
1 e4 e6 2 d4 d5 3 e5 c5 4 c3 Nc6 5 Nf3 Nh6
A major point of 3 e5 is that Black’s king’s knight is prevented from going to its optimal square f6, and if it goes to e7, the knight obstructs Black’s king’s bishop. With 5 ... Nh6, Black keeps a path for that bishop open and plays flexibly, not wishing to commit his queen to b6 or his bishop to d7 before it’s clear where those pieces are best placed. Right away the question arises: what about 6 Bxh6 (or Bxh6 on one of the next few moves)? After all, this leaves Black with weak doubled h-pawns as well as specific weaknesses on h5 and f6. It must have been this possibility that for decades prevented 5 ... Nh6 from being played at master level, or only very rarely. Today, however, allowing doubled rook pawns has become much more common. Without minimizing the problems they entail, we find that in many variations of the French they are compensated by other advantages, namely: 1. Black gets the two bishops, a long-term advantage; 2. Without his darksquared bishop, White will find it difficult to attack Black’s king, and he will have one less defender of the key squares d4 and e5; 3. The move ... f6 gains in value, especially because Black can play ... Bg7 without loss of tempo and put more pressure on e5; 4. The open g-file can be useful for one of Black’s rooks and, since ... f6 will generally open the f-file, Black can get a nice set of rooks staring down at White’s kingside. Assuming that White refrains from capturing the knight on h6, the main advantage that ... Nh6 has over ... Nge7 is that Black now has the idea of ... Nf7. Specifically, the sequence ... f6, ... fxe5, ... Nf7, attacking e5, is a concern for White. Furthermore, there will be lines in which White plays Qh5+ and Black renders it harmless by responding with ... Nf7. Finally, the knight will sometimes go to g4; for example, to attack a pawn on e5 or bishop on e3. White’s logical replies are: 3.1 3.2 3.3 3.4 3.5 3.6
6 6 6 6 6 6
Be2 Bd3 a3 Bxh6 Na3 dxc5
3.1
6
Be2
This isn’t very challenging, but it serves as an introduction to a few of Black’s ideas. 6 ... Nf5 This is safe, effective, and arguably best. Other moves: (a) 6 ... cxd4 7 Bxh6 (7 cxd4 Qb6 or 6 ... Nf5 7 Nc3 Qb6 transposes to 5 ... Qb6 6 Be2 in the previous chapter; 7 ... f6!? would be a venturesome alternative) 7 ... gxh6 8 cxd4 f6!? (8 ... Bg7 9 Nbd2 0-0 10 Nb3 f6 11 exf6 Qxf6 12 0-0) 9 0-0 (9 exf6 Qxf6 10 0-0 Bg7 transposes; here 10 ... Bd6 is an alternative) 9 ... Bg7 10 exf6 Qxf6 11 Bb5 0-0 12 Bxc6 bxc6 13 Nc3 c5! 14 dxc5 Rb8 15 Rb1 Rb4=; e.g. 16 Re1 Qg6! 17 Kh1 Bb7 intending ... d4 and ... e5. (b) 6 ... f6 is a forthright attempt to smash up White’s centre. The first point is that Black aims to play ... fxe5 and ... Nf7 in conjunction with ... Qc7. White should react quickly; for example, 7 0-0 (7 exf6 Qxf6 8 Bg5 Qg6 9 0-0 cxd4 10 cxd4 Bd6 11 Nc3 00=; 7 Bxh6 gxh6 8 0-0 cxd4 9 cxd4 Bg7) 7 ... Nf7 8 exf6 Qxf6 9 Re1 cxd4 10 cxd4 Bd6 11 Nc3 0-0 with double-edged play.
7 g4!? Not 7 0-0? cxd4 8 cxd4 Qb6. Otherwise: (a) 7 b3!? cxd4 8 cxd4 Bb4+ 9 Kf1 is not attractive; e.g. 9 ... 0-0 (9 ... Be7 10 Bb2 0-0 11 Na3 f6=/+ Aliabiev-Mesropov, Moscow 2001) 10 g4 Nh4 11 Bg5 Be7 12 Bxh4 Bxh4 13 Kg2 f6! 14 Nxh4 fxe5 15 Nf3 exd4 and ... e5 with advantage. (b) 7 Bd3!? cxd4 8 Bxf5 exf5 9 cxd4 Be7 10 Nc3 Be6 is a basic Advance Variation position discussed in the previous chapter; here White has burned up a tempo with Be2-d3 and Black should be happy with his position. (c) 7 Na3 cxd4 8 cxd4 Bb4+! 9 Kf1 (9 Bd2?! Bxa3 10 bxa3 Nfxd4) 9 ... Be7 10 Nc2 0-0; e.g. 11 Ne3 Nh4 12 Nxh4 Bxh4 13 g3 Be7 14 Kg2 f6 15 f4 Bd7=/+.
7 ... Nh4 8 Nxh4 Qxh4 9 Be3 Bd7 10 0-0 Or 10 Nd2 cxd4 11 Nf3 Qd8 (11 ... Qxg4 12 Rg1 Qxg1+! 13 Nxg1 dxe3=/+ – Moskalenko; this would be a serious practical advantage at any rate) 12 cxd4 h5 (12 ... Qb6!? – Moskalenko) 13 gxh5 Rxh5 14 h4 Ne7 15 Ng5, Kupreichik-Moskalenko, Lvov 1988; and 15 ... Rh6 16 Bd3 Nf5 17 Bxf5 exf5 should favour Black. 10 ... Qd8 Here 10 ... h5!? 11 g5 f6!? and 10 ... f6 11 f4 cxd4 12 cxd4 Be7 13 Nc3 0-0 are sound alternatives. After 10 ... Qd8, Sax-Glek, German League 1994, continued 11 Nd2 Qb6 12 Rb1 (12 Qb3=) 12 ... cxd4 13 cxd4 a5 14 Kh1 a4!? 15 f4 h5!? (15 ... Bb4 16 Nf3 h5! – Glek) 16 gxh5 Ne7 17 Bf2 Bb5 (17 ... Nf5 18 b3 a3 19 Bg4 Nh6 20 Be2 Bb5=) 18 b3 Bxe2 19 Qxe2 a3 20 b4 Nf5=. 3.2
6
Bd3
This is a move you’ll see a lot and it deserves serious study. 6 Bd3 targets the kingside, makes ... f6 look riskier, and allows e5 to be defended directly by Re1 or Qe2. White is also ready to capture a knight on f5 in one move (as opposed to Be2d3xf5). On the negative side, the bishop on d3 interferes with White’s defence of d4. In addition, it may be forced to lose a tempo in positions where Black plays ... Nb4 or ... Nxe5. Most of what follows is my own analysis. I’ll look at two approaches: 3.21 3.22
6 6
... ...
cxd4 f6
These are the two moves which lead to unique types of positions. But Black can also consider 6 ... Nf5, trying to transpose to the 6 ... cxd4/7 ... Nf5 lines without allowing Bxh6, and also not ceding White the c3-square until he has committed to 0-0. The key line is 7 0-0 (7 Bxf5 exf5 8 0-0 h6!? has the idea that 9 dxc5?! Bxc5 10 Nd4 0-0 helps Black more than White; 7 dxc5 Bxc5 8 b4 Bb6 9 b5 Nce7 10 0-0 Ng6 11 a4 0-0 also seems balanced) 7 ... cxd4 8 Bxf5 (the gambit 8 a3?! can be met by simply 8 ... Nh4! 9 Nxh4 Qxh4 10 g3 Qd8) 8 ... exf5 9 Nxd4 (9 cxd4 Be6 10 Nc3 is the main line, where Black might like the fact that White has committed to 0-0). Pedersen claims an advantage for White after 9 Nxd4, but that’s not apparent after 9 ... Be7 10 f4 Nxd4 11 cxd4 0-0 12 Nc3 Be6, when Black has as many prospects on the queenside as White. 3.21 6 ... cxd4 7 7 Bxh6!? gxh6 8 cxd4 has won some games, but Black should be fine and has one particularly inspiring solution: (a) 8 ... Qb6 9 Qd2 Bd7 (threatening ... Nxd4) 10 Be2 Nxd4!?
cxd4
This comes out of nowhere! Black’s sacrifice seems to be sound, and whoever is more familiar with the analysis is likely to do well: 11 Nxd4 (11 Qxd4 Bc5 12 Qd2 Bxf2+ 13 Kf1 Be3 14 Qd1 0-0 intending ... f6, even in the case of 15 Qb3 f6! 16 Qxb6 axb6!, threatening both ... fxe5 and ... Bc1) 11 ... Bc5 12 Nc2 (12 Nb3 Bxf2+ 13 Kf1 Be3 14 Qd3 0-0 15 Nc3 f6 16 exf6 Rxf6+ 17 Ke1 Raf8 18 Nd1 Bg5 with an ongoing attack; e.g. ... Be8-g6 or ... a6 and ... Bb5 could follow) 12 ... Bxf2+ 13 Kf1. Here I like 13 ... Rc8 (13 ... Rf8 14 Qb4 is an issue) 14 Nba3 Bc5 15 b4 Be7 16 g3 (16 Qxh6 a5!=/+ with the idea 17 b5 Rc3) 16 ... Qc7 17 Qe3 Bg5 18 Qd4 f6!=; e.g. 19 exf6 (19 Bh5+ Ke7; 19 Kg2! fxe5 20 Qxa7 0-0 21 Rhf1 d4 22 b5 Be7 23 Rxf8+ Rxf8 24 Rf1 Be8! 25 Rxf8+ Bxf8 unclear, with the ideas ... Bg6 and ... Qd7-d5) 19 ... 0-0 20 Kg2 Bxf6 21 Qg4+ Kh8 22 Rac1 e5 23 Qh5 Qd6 24 Rhf1 d4 25 Qxh6 Bc6+! (25 ... Qd5+ 26 Bf3 e4 27 Bg4 Bxg4 28 Rxf6 e3+ 29 Kg1=) 26 Bf3 e4 27 Bg4 e3+ 28 Kg1 (28 Bf3?? e2) 28 ... e2! 29 Bxe2 d3/+. (b) Calmer is 8 ... Bg7 9 Nc3 f6
and now: (b1) 10 Bb5 0-0 11 Bxc6 bxc6 12 exf6 Qxf6 13 0-0 Rb8 14 Qe2 (or 14 Na4 e5!?, when 15 dxe5? Qg6! is already better for Black, whereas 15 Nxe5! Ba6 16 Nd7 Qf5 17 Nxb8 Bxf1 18 Qxf1 Rxb8 is unbalanced but equal) 14 ... Qg6 (14 ... c5!? 15 dxc5 Qe7 16 Nxd5 Qxc5 17 Nc3 Qb6=, planning ... Ba6 or ... Bb7 with hyperactive bishops) 15 Rac1 Rb6 16 Rfe1 Ba6! 17 Qxe6+ Qxe6 18 Rxe6 Bc8 19 Re7 Rxb2 20 Rxa7 Bg4 with sufficient counterplay. (b2) 10 0-0 0-0 (10 ... fxe5 11 Nxe5 0-0) 11 Re1 (11 Qc2?! fxe5 12 Bxh7+? Kh8 13 dxe5 Rxf3! 14 gxf3 Qh4 15 Kg2 Bd7!-/+ and White will have to give a lot back) 11 ... fxe5 12 dxe5 Bd7 13 Bc2 (13 Nb5 Qb6 14 Nbd4 Nxd4 15 Nxd4 Rf4! 16 Nf3 Raf8=) 13 ... Be8 14 Nd4 Qe7 15 Qg4 Nxd4 16 Qxd4 Bg6=.
7 ... f6!? This is an enterprising way to go, opening lines, although Black risks a minor positional disadvantage. As in similar positions, Black can play 7 ... Nf5, which is a safe route to equality: 8 Bxf5 (if White actually prefers the black queen on b6 rather than d8, he can play 8 Bc2!? Qb6 9 Bxf5 exf5 10 Nc3 Be6, and then can castle or play 11 Ne2 h6 12 h4; an instructive example, Benjamin-Shaked, Kona 1998, went 12 ... 0-0-0!? 13 Kf1 Kb8 14 Kg1 Be7 15 Rb1, and Black found the dynamic and sound sacrifice 15 ... f4! 16 Nxf4 Bf5 17 Ra1, when his most accurate path to advantage was 17 ... Bg4! 18 b3 Bxf3 19 gxf3 Qxd4) 8 ... exf5. In this important position, White has two fundamentally different ways to set up, with and without castling: (a) 9 Nc3 Be6
Moskalenko rates this ‘=/+’, based upon his lengthy experience, and says: ‘A well-known type of position, White has nothing in this structure.’ I think this set-up is objectively equal, but White will have more difficulty than Black digging up positive chances. You will be well served to study what follows, since the basic position often arises: (a1) 10 Ne2 Be7 11 h4 h6!? (this standard move prevents White from exchanging a knight or bishop on g5; Black also plans to play ... g5 if allowed) 12 Nf4 Qb6 13 Rh3 (13 h5 is generally answered by ... 0-0, since it’s no longer easy to play ... g5 after ... 0-0-0 – then, since Black’s kingside is easy to defend, the action takes place on the queenside) 13 ... g6! (the idea is that after 13 ... 0-0-0 14 h5, ... g5 would be hard to organize) 14 Kf1 0-0-0 15 Kg1 g5!? (15 ... Rdg8! and 15 ... Kb8 are also good preparatory moves for ... g5; here it’s playable right away, albeit perhaps without achieving a meaningful advantage) 16 Nxe6 fxe6 17 hxg5 hxg5 18 Nxg5?! (18 Rxh8! Rxh8 19 Bxg5 Bxg5 20 Nxg5 Qxd4 21 Qxd4 Nxd4 22 g3! Kd7 23 Rd1 Nc6 24 f4 equalizes; instead, Black’s structure is slightly superior after 20 ... Nxd4 21 Qd2 Rh5 22 Rd1 Ne2+ 23 Qxe2 Rxg5, though it’s not much either) 18 ... Qxd4 19 Qxd4 Nxd4 20 Rd3 Ne2+ 21 Kf1 Nxc1 22 Rxc1+ Kd7 23 Nh3 Rc8 24 Rdc3 Rc6! 25 Ke2 Rh4=/+ NunMoskalenko, Pardubice 1994. Black has the superior minor piece and a potentially important passed central pawn. (a2) 10 h4 is the other order by which to set up a blockade, when a sample of the play is 10 ... h6 11 h5 Be7 12 Ne2 (12 Bf4 Qb6 13 Qd2 Rc8 intending ... Na5 is easy for Black to play) 12 ... Qb6 (12 ... Qa5+ 13 Kf1 0-0) 13 Nf4 (13 Rh3 f4! 14 Rh1 Bg4 15 Kf1 0-0 16 Nxf4 Rfd8 and Black’s activity gives him the edge) 13 ... 0-0 14 Rh3 Rfc8 15 Nxe6 (15 Rg3 Nb4 16 Kf1 Rc4) 15 ... fxe6 16 Rg3 Kf7 17 Kf1 Nb4 18 Bd2 Rc4! 19 b3 Rc6=/+. (b) 9 0-0 Be6 10 Nc3 Be7
The lines in which White plays 0-0 have grown in popularity by comparison with those involving h4. Naturally, he wants to develop quickly and connect rooks, but it’s also the case that Black can now throw his kingside pawns forward in an assault that can be hard to counter. Here are some examples: (b1) 11 Ne1 is frequently played; as with 11 Ne2, White wants to strengthen his centre with f4 and help restrain ... g5: 11 ... Qb6 12 Nc2 Rc8 13 f4 g6 14 Kh1 h5 (Black prevents g4; now he can play on the queenside) 15 Qd3 a5 16 Na4!? Qa6! 17 Qxa6 bxa6 18 Nc3 0-0 19 a3 Rb8 20 Rd1 Rfd8 21 Ne3, Kabanov-Poghosyan, Moscow 2011; and Black might well open a new front with 21 ... f6! 22 exf6 (22 g3? fxe5 23 fxe5 Nxe5 24 dxe5 d4; 22 Ne2 a4! intending ... Na5, when 23 Nc3? fxe5 24 fxe5 f4 25 Nf1 Bg4 is strong) 22 ... Bxf6 23 Nc2 Re8 with a nice plus. (b2) 11 Qb3 Qd7 12 Ne2 h6 is fine: 13 Nf4 (13 h4 a5!? 14 Nf4 a4 15 Qd3 0-0 has the idea of ... Nb4 and ... a3) 13 ... g5 14 Nxe6 fxe6 with kingside chances and the better bishop. (b3) 11 a3 can be met by 11 ... g5! (or 11 ... Rc8 12 Na4 b6 13 b4 g5! 14 Qd3 f4 15 Bb2 Qd7=/+ Hartikainen-Soffer, Cracow 2011) 12 Ne1 Qb6 13 Nc2 f4! 14 b4, Kalashnikov-Manukyan, St Petersburg 2009; and here 14 ... Rc8 15 Qd3 Qc7! 16 Bb2 Qd7! prepares ... 0-0 and ... Bf5. (b4) 11 Ne2 can be answered slowly, or again by 11 ... g5!?; e.g. 12 Ne1 (12 Qb3 Rb8 13 Rd1 f4 14 Ne1 Qb6!? 15 Qxb6 axb6 16 Bd2 f6=/+ intending ... Kf7 and ... Bf5, Degraeve-Ulibin, Mamaia 1991) 12 ... f4 13 f3? Qb6 14 Kh1 Bf5 15 b3 Qa6 16 Rf2 0-0-0 17 Bb2 f6 18 exf6 Bxf6-/+ Teschner-L.Portisch, Monte Carlo 1969. 8 Bxh6 gxh6
9 exf6
White’s centre is broken up anyway after 9 0-0 Bg7 10 exf6 (10 Re1 0-0 11 Nbd2 fxe5 12 dxe5 Bd7 13 Nf1 Be8! 14 Ng3 Bg6= Finegold-Bercys, ICC 2010) 10 ... Qxf6. Then 11 Nc3! 0-0 12 Ne2 shores up d4. After 12 ... Bd7, Black has ideas of ... Be8h5/g6, so White does well to control e5 by 13 Bb5! (13 Qb1? Be8 14 Bxh7+ Kh8 and ... Bh5), when a sample line pitting bishop versus knight is 13 ... Qg6 14 Bxc6 bxc6 15 Ne5 Bxe5 16 dxe5 c5 with mutual chances; e.g. 17 f4 h5!? 18 Rc1 Rac8 19 Rc3 h4! (19 ... Kh8 20 Rg3 Qe4) 20 Rh3 Qg4 21 Qe1 Bb5! 22 Rff3 Bxe2 23 Rxh4 Bxf3 24 Rxg4+ Bxg4 25 Qg3 Kh8 26 Qxg4 c4 27 Qxe6 d4 28 Qd5 c3 29 e6 c2 30 Qe5+ Kg8 31 Qg5+=. 9 ... Qxf6 10 Bb5! White increases his grip on d4 and e5. 10 0-0 Bd6 11 Nc3 0-0 transposes after 12 Bb5; otherwise Black combines pressure on d4 and e5 with rook play on his open kingside files. 10 ... Bd6 11 0-0 0-0 12 Nc3 Now 12 ... Kh8 is plausible, but more interesting is: 12 ... Qg7!?
This intends 13 ... Rf4!, attacking the d-pawn and strengthening Black’s position on the f-file. The play is roughly balanced. I’m going to give a lengthy analysis in order to demonstrate the typical ideas and tactics. First, White tries to take over e5: 13 Bxc6 bxc6 14 Re1 Or 14 Qe2 a5, intending ... Ba6, ... Raf8. 14 ... Rf6 Covering e6 and preparing to double rooks in some lines. The alternatives 14 ... Rf4, 14 ... Rf5, and 14 ... Rb8 are all noteworthy. It’s useful to hold out the possibility of ... Ba6 for a while, rather than commit to ... Bd7. 15 Na4 Others: (a) 15 Rc1 Ba6 (15 ... Bd7 16 Rc2 Raf8 17 Ne5 Bxe5 18 Rxe5 c5! with the idea 19 dxc5 Rg6) 16 Na4 Raf8 17 Kh1 Rxf3 18 gxf3 Bxh2! intending 19 Kxh2? Bf1! 20 Rxf1 Rf4 and mate. (b) Upon 15 Ne5, Black can counterattack by 15 ... Ba6 16 Qa4 Raf8!; e.g. 17 Qxa6 Rxf2 18 g3 Bxe5 19 Rxe5 Qg4 20 Qxc6! Qf3 (20 ... R8f6!? 21 Nxd5! Qxd4 22 Nxf6+ Rxf6+ 23 Kh1 Qxe5=) 21 Qxe6+ Kh8 22 Qh3 Rxb2 23 Rf1 Rf2 24 Rxf2 Qxf2+ 25 Kh1 Qf3+ 26 Qg2 Qxc3=. 15 ... Ba6! 15 ... Bd7 16 Nc5 Bxc5 17 dxc5 Raf8 18 Qd4! keeps Black tied down. 16 Rc1 This leads to dramatic play. Tactics also arise from 16 Nc5 Bxc5 17 dxc5 Raf8 18 Re3 Rg6 19 Nh4 (19 g3 Rgf6) 19 ... Rg5 20 Rxe6 Bc8!? 21 Rxc6 Bh3 22 Kh1! (22 g3? Qxb2) 22 ... Rxf2 23 gxh3, which draws: 23 ... Qe5 (23 ... Qxb2 24 Ng2 Rfxg2 25 Rc8+ Kg7 26 Rc7+=) 24 Rc8+ (24 Nf3 Rgg2=) 24 ... Kf7 25 Nf3 Qe4 26 Rc7+ Kf8 27 Rc8+=. 16 ... Raf8 17 Kh1 There’s a nice symmetry in 17 Re3 Bf4 and 17 Rc3 Bb4!. 17 ... Rxf3! 17 ... Kh8!? 18 Rxc6? (this loses beautifully; instead, 18 Rg1! yields a small advantage) 18 ... Rxf3! 19 gxf3 (19 Rxd6 loses to 19 ... Rxf2 20 Rg1 Bf1!, as does 19 Rxa6 Rxf2 20 Rg1 Rg8 21 g3 Rxh2+! 22 Kxh2 Bxg3+ and ... Qg5) 19 ... Bxh2!
Now ... Rg8 is threatened and there is no good defence: 20 Rexe6 (20 f4 Rg8 21 Qf3 Be2!! 22 Qh3 Bxf4) 20 ... Bd3! (20 ... Rg8 21 Rg6!) 21 Kxh2 (21 Rg6 Bxg6 22 Kxh2 Be8 23 Rc3 Rf4) 21 ... Rg8 22 Qh1 Qg5 23 Rg6 Bxg6 24 Qg2 Qh4+ 25 Qh3 Qxf2+ 26 Qg2 Qxd4 and wins. 18 gxf3 Bxh2! The idea of ... Rf4-h4 produces a level game. 19 f4 19 Rxc6? Kh8-/+ would be mistaken, but 19 Nc5 also keeps the balance after 19 ... Kh8 20 f4, transposing to 19 f4; though not here 20 Nxe6?! Qe7 21 Nxf8 Qh4 22 Kg2 Bf4 23 Ng6+ hxg6 24 Re8+ Kg7 25 Qg1 Qg5+ 26 Kh1 Bxc1 27 Qe1 Qh4+ 28 Kg2 Bf4! with excellent winning chances. After 19 f4, best play is 19 ... Kh8! 20 Nc5 Rg8 21 Qf3 Qf6 22 Kxh2 (or 22 Qh5 Bxf4 23 Nxa6 Bxc1 24 Rxc1 Qxf2 25 Qe5+ Rg7 26 Qb8+ Rg8 27 Qe5+=) 22 ... Qh4+ 23 Qh3 Qxf2+ 24 Kh1 Rg3! 25 Qxg3 Qxg3 26 Nxa6 Qh3+=. 3.22
6
...
f6
Black attacks the centre directly, intending ... fxe5 and ... Nf7. The retreat to f7 can also come in handy if White plays Qh5+. 7 0-0 Maintaining the tension poses the most problems. Instead:
(a) 7 Bxh6 gxh6
(a1) 8 Bb5 strengthens the e5 point, albeit with loss of tempo: 8 ... Bg7 9 exf6 (9 Nbd2 cxd4 10 Bxc6+ bxc6 11 cxd4 c5!? 12 dxc5 0-0= 13 Qe2?! Rb8 14 Nb3 fxe5 15 Nxe5 Qg5 16 Nd3 e5!=/+) 9 ... Qxf6 10 0-0 (10 dxc5 wins time but cedes the centre: 10 ... Qe7 11 b4?! 0-0 – threatening ... Nxb4 – 12 Bxc6 bxc6 13 0-0 a5 etc; after 11 0-0 Qxc5 instead, Black has no worries) 10 ... 0-0 (threatening the d-pawn) 11 Bxc6 bxc6 12 Nbd2 (12 dxc5?! e5=/+ Eklund-M.Johansson, corr 1989) 12 ... cxd4 13 cxd4 c5 14 dxc5 Qxb2 15 Qa4 Qb7 16 Rab1 Qc7 with ... e5 to come. (a2) 8 0-0 cxd4 9 Nxd4 (9 cxd4 Bg7 transposes to 6 ... cxd4 7 cxd4 f6) 9 ... Nxe5 (9 ... fxe5!? is more forcing; e.g. 10 Nxc6 bxc6 11 Qh5+ Kd7 12 Qxe5 Rg8 13 Re1 Qg5! 14 Qxe6+ Kc7 15 Qe5+ Qxe5 16 Rxe5 Rb8 17 b3 Bh3 18 g3 Bg7 19 Re7+ Kb6=) 10 Qh5+ Kd7 11 Bb5+!? (11 Bc2 Kc7 12 Nd2 Nc6 and ... e5 gives just enough play; e.g. 13 N2f3 e5 14 Nf5 Be6 15 Rfd1 Qe8) 11 ... Kc7 12 f4 Ng6 13 f5 a6 14 Be2 exf5 15 Nxf5 Qe8!. (b) 7 exf6 gives up the centre for development: 7 ... Qxf6 8 Bg5 (8 0-0 cxd4 9 Bg5 Qf7 10 cxd4 Bd6 11 Nc3 0-0) 8 ... Qf7 9 0-0 cxd4 (9 ... c4!? 10 Bc2 Bd6 11 Nbd2 0-0 12 Bxh6 gxh6 13 Qe2+/=) 10 cxd4 Nf5!? (or 10 ... Bd6 11 Nc3 0-0) 11 Ne5 (11 Bxf5 Qxf5 12 Nc3 Bd6 13 Bh4 0-0 14 Bg3 Bf4 15 Re1 Bd7 16 Bxf4 Qxf4, roughly equal; e.g. 17 Rc1 Be8! 18 Rxe6 Bh5 19 Re3 Qh6! with persistent pressure) 11 ... Nxe5 12 dxe5 Bc5 with equality; for example, 13 Qc2 (13 Nc3 0-0 14 Rc1 b6) 13 ... Bb6 14 Bxf5 exf5 15 Nc3 0-0 16 Bf4 Bd7=. 7 ... cxd4 7 ... Nf7 8 Re1 (8 exf6 Qxf6 9 Re1 Bd6 10 Be3! c4 11 Bc2 0-0 12 Nbd2 b5 13 Nf1 Rb8 14 Ng3 g6=; e.g. 15 h4 b4 16 h5 g5 17 Nh2 h6 18 Ng4 Qg7 with mutual prospects) 8 ... fxe5 9 Nxe5, and here Wei Ming analyses 9 ... Nfxe5! 10 dxe5 g6 11 Nd2 Bg7 12 Nf3 0-0=. 8 cxd4 Nf7
9 Re1 9 Nc3!? fxe5 10 Nxe5 Nfxe5 11 dxe5 g6! (11 ... Nxe5 leads to the pretty idea 12 Qh5+ Nf7 13 Bb5+ Bd7 14 Re1! and after the e-pawn falls Black’s king remains in the centre) 12 Bf4 Bg7 13 Re1 0-0 14 Bg3 Bd7=. 9 ... fxe5 10 dxe5 g5! 11 h3 Bg7 12 Bc2 12 Bd2 Qe7 13 a3 Bd7 with the idea 14 Bc3 h5!. 12 ... Qe7 Black stands well. Play might go 13 Nc3 h6 (threatening e5) 14 Bg6!? 0-0 15 Bxf7+ Rxf7 with the bishop pair and active play. The 6 ... f6 variations are undoubtedly looser than those with 6 ... cxd4 (or 6 ... Nf5), and you should probably have exact lines memorized. Notice, too, the opportunity to enter into different versions of the positions with ... Nf5; that’s a safe option which may appeal to someone nervous about piece sacrifices and kings in the centre. 3.3
6
a3
You will probably encounter this move sometimes, if only because so many players use 6 a3 versus 5 ... Qb6. There are significant differences, however, in that Black can make moves that are more dynamic than 6 ... Qb6. 6 ... f6 Once again, this thematic break creates original play. Instead, 6 ... Qb6 is a direct transposition to the main line of
Chapter 2. 6 ... Nf5 is a sound alternative. A recent game Zelcic-Kovacevic, Oberwart 2011, went 7 dxc5?! (7 b4 cxd4 8 cxd4 Qb6 again transposes to Chapter 2, though 8 ... Bd7 is a good alternative, planning ... Rc8 and, depending upon the reply, ... Qb6, ... Be7, ... 0-0, and/or ... f6; 7 Bd3 cxd4 8 Bxf5 exf5 9 cxd4 Be6= is the standard position from the last two sections; then the move a3 proves to be double-edged, since it creates potential weaknesses on the vulnerable light squares) 7 ... Bxc5 8 Bd3 Nh4! (removing the central defender) 9 Nxh4 Qxh4 10 Qe2 0-0 (or 10 ... f6 11 exf6 0-0!, because now 12 Nd2 can be met by 12 ... e5!, or 12 0-0 Rxf6 13 Be3 Bxe3 14 Qxe3 e5) 11 Nd2 f6 12 Nf3 Qh5 (12 ... Qxf2+!? 13 Qxf2 Bxf2+ 14 Kxf2 fxe5 15 Re1 e4 16 Bxe4 dxe4 17 Rxe4 e5 18 Kg1 Bf5 19 Re1 Bd3=/+) 13 exf6 Rxf6 14 Be3 Bxe3 15 fxe3 e5 16 e4 Bg4 with a limited advantage.
7 b4!? This is the natural follow-up move, but others deserve respect: (a) 7 exf6 Qxf6 8 Bg5 Qf7 (8 ... Qg6 may be more accurate, intending 9 Bb5 Bd6 or 9 Bd3 Qh5) 9 Bb5! Bd6 10 dxc5 Bxc5 11 0-0 0-0 12 Qe2?! (12 Bxh6! gxh6 13 Nbd2 Ne7!? 14 Ne5 Qg7 15 Ndf3 Ng6=; e.g. 16 Nxg6 hxg6 17 Qe2 Bd6 18 Rae1 g5) 12 ... Qh5 13 Nbd2 Bd6! 14 Bh4 Nf5 15 Bg3 Nxg3 16 fxg3 e5 17 Nb3, Samuelsen-Jepson, Torshavn 2003, and here simply 17 ... a6 18 Bxc6 bxc6 leaves Black with a big centre and two active bishops. (b) 7 Bb5 Nf7 8 Bf4 (8 exf6 Qxf6 9 0-0 cxd4 10 cxd4 Bd6 and ... g5 is a theme) 8 ... g5! (8 ... Bd7) 9 Be3 g4 10 Ng1 cxd4 11 cxd4 fxe5 12 Qxg4 exd4 13 Bxd4 e5 14 Qe2 Bd6-/+; Black’s centre dominates. (c) 7 Bd3 Nf7 8 Qe2 (8 exf6 Qxf6 9 Be3 c4 10 Bc2 Bd6 at the least equalizes) 8 ... cxd4 9 cxd4 fxe5 (or 9 ... Qc7=/+ threatening ... Nxd4) 10 dxe5 g6 11 0-0 Bg7 12 Bf4, Ho-Ding, Nashville 2009; 12 ... g5! 13 Bg3 h5 14 h3 g4, winning the e-pawn. (d) 7 Bxh6 gxh6 8 Bb5 Bg7 9 exf6 Qxf6 10 0-0 0-0 11 dxc5 Qe7=, in view of 12 b4? a5.
7 ... cxd4!? Two worthwhile alternatives: (a) 7 ... fxe5 looks fine, intending 8 b5 (8 Bxh6 gxh6 9 dxe5 c4) 8 ... Nxd4! 9 cxd4 exd4 with three pawns and a huge centre for the piece; ‘=’ seems fair. (b) 7 ... Nf7!?, as I gave inDWF, isn’t necessary, but I should mention the remarkable idea 8 exf6 cxd4!? (8 ... Qxf6 was my recommendation and a little safer) 9 b5 Nce5! 10 Nxe5 Nxe5 11 fxg7 Bxg7 12 cxd4 Ng6 with ideas of ... e5, ... Qf6 or ... Qh4, ... Nh4 and ... 0-0, giving considerable and entertaining compensation. 8 cxd4 fxe5
9 Bxh6! Others: (a) 9 dxe5 Nf7 10 Bb2 a6 intends either ... g6 or ... g5. (b) 9 Nxe5 Nxe5 10 dxe5 Nf7 11 Bb2 a5! 12 Qd4! (12 b5 allows 12 ... Bc5, with smooth development by ... 0-0 and perhaps ... Qg5 or ... Qh4, which is hardly fun for White) 12 ... Qe7!? (ambitious; whereas 12 ... axb4 13 axb4 Rxa1 14 Bxa1 Qc7 15 Bc3 Be7 is equal) 13 Bd3 axb4 14 0-0 Nd8 and ... Nc6. 9 ... gxh6 10 Nxe5 Nxe5 11 dxe5 Qc7 11 ... Qg5!? 12 Bb5+ Kf7 (or 12 ... Kd8 13 0-0 Qxe5 14 Nd2 Bd6=/+) 13 0-0 Qxe5 14 Nd2 Bd6 15 g3 Rf8 gives White unclear compensation for the pawn. 12 Qh5+ Kd8 13 Nd2 Qc3 14 Rd1 Qxa3
Here 14 ... Bd7 is also possible. White’s compensation looks real and not easily assessed; nevertheless, given the choice, I’d take Black. On the positive side, he may end up having two extra passed pawns with which to keep his opponent busy. 3.4
6
Bxh6
gxh6
White’s immediate capture has become rare, and Moskalenko even calls it a ‘mistake’. That’s going a bit far, as some lines are merely equal, but there’s enough imbalance to keep things interesting. 7 Be2 The solid move. (a) 7 Bd3 Qb6 (7 ... Bd7! might be still more accurate, threatening to win a pawn by ... Qb6, when 8 Qd2 now gives Black the option of 8 ... f6!) 8 Qd2 Bd7 (8 ... cxd4 9 cxd4 Bd7 10 Be2 Bg7 11 Nc3 0-0 12 0-0 f6 13 exf6 Rxf6 is a fairly standard and equal position, in which Black has ideas of ... Raf8 and ... Be8-h5, but also simply ... Rxf3 and ... Nxd4 at the right moment) 9 dxc5 Qc7 10 Qf4 Bg7 11 Qg3 (11 Bb5) 11 ... 0-0! ‘and Black’s pieces are excellently positioned’ (Moskalenko). (b) 7 dxc5?! Bxc5 (7 ... Qc7! – Moskalenko) 8 Bd3 (8 b4 Bf8! followed by ... Bg7 is ideal for Black, both attacking e5 and defending the king; compare this with 6 dxc5 Bxc5 7 b4) 8 ... f6 9 exf6 Qxf6 10 0-0 0-0 11 Bc2 Rf7! 12 Nbd2 Bd7 13 Nb3 Bb6 14 Qe2 e5 and Black already had a clear plus in Short-Glek, Cap d’Agde (rapid) 1996. (c) 7 Bb5 is normally a harmless move in the absence of ... f6, because White can’t put a piece on e5.
This position is nevertheless important and interesting, because there will arise two-bishop-versus-two-knight positions
in which both sides’ minor pieces have sufficient activity and yet neither’s are ideally placed. Here are a few sample lines out of many more possibilities: (c1) 7 ... Qb6 8 Qe2 a6 (or 8 ... Bg7 9 0-0 0-0 10 Rd1 Bd7; e.g. 11 dxc5 Qxc5 12 a4 f6 13 exf6 Rxf6 14 Nbd2 Raf8 15 Nb3 Qd6 16 a5, and 16 ... a6 17 Bxc6 Bxc6 equalizes, although 16 ... Rxf3!? 17 gxf3 Ne5 is more fun) 9 Bxc6+ bxc6 10 0-0 a5 11 Qc2 Ba6 (11 ... Rb8 12 b3 Ba6) 12 Rc1 (12 Rd1 Be7 13 Nbd2 cxd4 14 cxd4 0-0= with ideas of ... c5, ... Rab8, and/or ... Be2) 12 ... Be7 13 Nbd2 0-0 14 Nb3 (14 Nf1 Rfb8 15 Rab1 Qb5! 16 Ne3 Qd3=) 14 ... cxd4 15 cxd4 Rfb8 16 Rab1 Bc4 (16 ... Qb5 17 Qc3 Bb4 18 Qe3 Qe2 19 Qxh6 Bd3!=) 17 Nc5 Bxa2 18 Ra1 Bc4! 19 Nd7 Qxb2 20 Nxb8 Qxb8 21 Qd2 a4! 22 Qxh6 Bd3 with a complex and equal game. (c2) 7 ... Qa5 is interesting because the move ... cxd4 will keep the position unbalanced; e.g. 8 Qe2 (8 a4 cxd4 9 Nxd4 a6!) 8 ... cxd4 9 Nxd4 Bd7 10 0-0 Bg7 11 f4 0-0 and, for example, 12 Na3 f6 13 exf6 Rxf6 14 Nac2 Qb6 15 Kh1 Raf8=. (c3) 7 ... Bd7 threatens ... Nxe5. It is slower than the options but quite playable; e.g. 8 Qe2 a6 (8 ... Qb6 is normal) 9 Bxc6 Bxc6!? (an unusual recapture, but the possibility of ... Bb5 creates unique possibilities) 10 a4 b5 (10 ... Qd7 11 b3 b5!?) 11 axb5 Bxb5 12 Qd1! (12 Qc2?! Rg8 13 g3 Rg4!) 12 ... Rg8 13 g3 Qb8! (intending ... cxd4) 14 Na3 Bd7 15 Rb1 cxd4 16 cxd4 Bb5!? (16 ... Ra7 17 0-0 Rb7 18 Qc2 Rg6) 17 Nxb5 Qxb5 18 Qe2 Bb4+ 19 Kf1 Kd7 with equal chances.
7 ... cxd4 What we see now is Black opening the position to activate his bishop pair, and White trying to exploit holes and weaknesses in his opponent’s camp. Other moves employ the same ideas, and are equally significant: (a) 7 ... f6 8 0-0 cxd4 9 cxd4 Bg7 is normal; e.g. 10 exf6 Qxf6 11 Bb5 0-0 12 Bxc6 bxc6 13 Nc3 Rb8 14 Qe2 Qg6 (or 14 ... Rb6) 15 Kh1 (15 Rfe1 Rb4 16 Rad1 e5!) 15 ... c5 16 dxc5 Bb7 intending d4; for example, 17 Rae1 d4 18 Ne4 Qxe4 19 Qxe4 Bxe4 20 Rxe4 Rxb2=. (b) 7 ... Qb6 8 Qd2 Bg7 9 0-0 0-0 10 Na3 Bd7 11 Nc2 cxd4 12 cxd4 is one of the old main lines fromPTF3.
This is extremely complex, as described there, but in order to be brief let me mention two continuations that don’t need a great deal of study to implement: (b1) 12 ... a5 is a solid, conservative move: 13 Ne3 (this is played most often; 13 Rab1 f6 14 exf6 Rxf6 15 b4 axb4 16 Nxb4 Qa5 is satisfactory for Black; e.g. 17 Ne5 Nxe5 18 dxe5 Rff8 – planning ... Be8-g6 – 19 f4 Be8 20 Bg4 Bf7 21 Kh1 Rae8 22 Rbc1 h5 23 Bh3 Bg6=) 13 ... h5! (preventing Ng4) 14 Rad1 (14 a4 f6 15 exf6 Rxf6, intending ... Bh6 or the usual ... Be8, or 15 Bb5 fxe5 16 Bxc6 bxc6 17 Nxe5 Be8! 18 f4 Rb8 19 Rf2 c5=/+ D.Fernando-Brynell, Bled Olympiad 2002) 14 ... f6 (14 ... Kh8 15 b3 Rac8 16 g3 f6 ½-½ Thipsay-Ganguly, Atul 2006) 15 exf6 Bxf6!? (15 ... Rxf6!=) 16 Nc2 Rae8 17 b3 Re7 18 Qh6 Be8 19 Bd3 Rg7 20 Rfe1 Nd8, reaching an obscure position with chances for both sides, Savic-Miljanic, Neum 2002. (b2) 12 ... f6 is more ambitious: 13 exf6 Rxf6 14 b4 Be8 (I like Moskalenko’s treatment of the position; 14 ... Raf8 15 b5 Ne7 16 Ne5 Be8 is the traditional line) 15 b5 Ne7 16 Ne5 Ng6 17 Ng4!? Rf8 18 Nxh6+ Kh8 19 Ng4 Nf4 20 a4 (20 Rfe1 Bg6=) 20 ... Bg6 (with excess compensation for the pawn) 21 Nge3?!, Roy Chowdhury-Moskalenko, Montcada 2008; after 21 ... Rac8 Black had only a moderate advantage, but there was also 21 ... e5! with a clearly superior game, in view of 22 dxe5? d4 23 Nc4 Qc5 and ... d3. 8 cxd4 Bg7 9 0-0 0-0 9 ... f6 is note ‘a’ to 7 ... cxd4. 10 Nc3? 10 Na3! f6 11 exf6 Qxf6 12 Nc2 Bd7 13 b4 Be8!= 14 b5 Ne7 15 Ne3!? Ng6 intending ... Nf4 is probably equal, but it’s easier for Black to play, Nakamura-Antal, Budapest 2002. 10 ... f6 11 exf6 11 Re1 is better but still not equal following 11 ... fxe5 12 dxe5 Nxe5 13 Nxe5 Bxe5 14 Bf3! Bd6. 11 ... Qxf6 Winning the crucial d-pawn; White’s subsequent recovery on h7 is hardly equivalent. 12 Qc2 Nxd4 13 Nxd4 Qxd4 14 Bd3 Bd7 15 Bxh7+ Kh8 16 Bd3 Rac8 17 Qe2 Djuric-Larino Nieto, Herceg Novi 2005; here 17 ... Qe5!, with a clear advantage, was even better than 17 ... e5 intending ... Bg4. To be clear, I don’t believe that Black gets the superior game in this line by force, and it’s even easy to fall into a limited positional disadvantage with a careless move or two. But that needn’t happen, and in fact White himself has a tendency to fall victim to tactics or powerful central advances. 3.5 6 Na3 Another knight goes to a rook’s file! White plans Nc2 to secure d4. 6 ... cxd4 A general consideration in the French pawn-chain structure: when White’s knight can no longer go to c3, the main drawback to ... cxd4 has disappeared. 7 cxd4
7 ... Bxa3!? A safe choice is 7 ... Nf5 8 Nc2, which is a traditional main line, since it also arises from 5 ... Nge7 6 Na3 Nf5 7 Nc2 cxd4 8 cxd4.
This has been heavily analysed, but to get you started, here are some suggestions: (a) 8 ... Qb6 9 Bd3 Bb4+!? (9 ... Bd7 is ‘b2’) 10 Kf1 (10 Bd2 Bxd2+ 11 Qxd2 Bd7 12 Bxf5 exf5 13 0-0 0-0 14 Ne3 Be6= 15 g3 f4! 16 gxf4 f6! with the idea 17 f5 fxe5 18 fxe6 Rxf3 19 dxe5 d4 20 Qe2 Rff8 21 Nd5 Qc5 22 e7 Nxe7 23 Nxe7+ Qxe7 24 f4 Rad8=) 10 ... Be7 11 g3 (11 h4 h5 12 g3 Bd7 13 Kg2 Rc8 14 Bxf5 exf5 15 Ne3 Be6 16 Qd3 Nb4 17 Qb3 Bd7!= intending 18 a3 Qa6) 11 ... Bd7 12 Kg2 Rc8 (not 12 ... Nfxd4?? 13 Ncxd4 Nxd4 14 Be3 Bc5 15 Rc1+–) 13 Bxf5 (13 Rb1 a5; or 13 g4 Nh4+ 14 Nxh4 Bxh4 15 g5 h6 16 gxh6 Nxe5!, with the idea 17 hxg7 Rg8 18 Bh6 Nxd3 19 Qxd3 f5!, or 17 dxe5 Qxf2+ 18 Kh3 Bf6! 19 Qe2 Qxe2 20 Bxe2 Bxe5 with full compensation for the piece) 13 ... exf5 14 b3 0-0 15 Bg5? (15 h4 h6 16 Qd3 Be6=/+) 15 ... Bxg5 16 Nxg5 f4! 17 gxf4? (17 Qd3 g6 and ... Bf5) 17 ... Ne7! (intending ... h6 and ... Qg6; White’s light squares are fatally weak) 18 Ne3, Short-Vaganian, Montpellier Candidates 1985; and 18 ... h6! 19 Nf3 Qg6+ 20 Kf1 Bb5+ 21 Ke1 Qe4 22 Rg1 Rc3 is resignable. (b) 8 ... Bd7 is a flexible and popular move:
(b1) 9 Be2 Qb6 (or 9 ... Nb4; e.g. 10 Ne3 Nxe3 11 fxe3 Be7 12 0-0 0-0, or 10 Nxb4 Bxb4+ 11 Bd2 Qb6 12 a3 Bxd2+ 13 Qxd2 0-0 14 0-0 Bb5=; 9 ... Be7 10 0-0 Qb6 11 Bd3 a5=; while 9 ... Rc8 is a useful move in any line) 10 0-0 (10 g4!? Nfe7 11 0-0 h5! 12 h3 hxg4 13 hxg4, Shirov-Van Wely, Wijk aan Zee 2001; and McDonald suggests 13 ... Ng6, when 14 Bd3 Be7! 15 Bxg6 fxg6 16 Qd3 0-0! threatens ... Nb4 and ... Bb5, meeting 17 a4 with 17 ... Qa6! 18 Qxa6 bxa6 19 Kg2 Rab8) 10 ... Rc8 11 g4 Nfe7 12 Nh4 (12 Rb1 h5 13 h3 hxg4 14 hxg4 Ng6=/+) 12 ... Nb4 13 Nxb4 Qxb4 14 f4 Nc6 15 Be3 Be7 16 Ng2, Sveshnikov-Gleizerov, 1992; now Black could play 16 ... h6!? (or 16 ... Qxb2) 17 f5 Bg5! 18 Nf4! exf5 19 gxf5 Bxf5 20 Nxd5 Bxe3+ 21 Nxe3 Bg6 22 Bc4 0-0 23 Qg4 Kh7 and White’s position is a bit looser than Black’s, but it’s about equal. (b2) 9 Bd3 Qb6 (or 9 ... Be7 10 h4 Qb6 11 Kf1!?, and the d-pawn is immune due to 11 ... Ncxd4?? 12 Nfxd4 Nxd4 13 Be3 Bc5 14 b4, but Black stands better after 11 ... f6! 12 g4!? fxe5! 13 gxf5 e4 14 fxe6 Bxe6 15 Ng5 Bxg5 16 hxg5 exd3 17 Qxd3 0-0-0 intending to play a rook to f8 and then ... Bf5) 10 Bxf5 (10 0-0 a5 11 Bxf5 exf5 12 Be3 Nd8! 13 Rb1 h6 14 h4 Be7 15 h5 Ne6 16 g3 g5, Jonkman-M.Gurevich, Vlissingen 1997, and now 17 hxg6 fxg6 threatens ... f4! with decisive activity) 10 ... exf5 11 0-0 h6 (11 ... Be7 12 Bg5) 12 a3 a5 13 Re1!? (13 Nfe1 Rc8 14 Nd3 g5! 15 b4 f4=) 13 ... Be6 14 h4?! Be7 15 g3 Qd8! (to support ... g5) 16 h5 g5 17 hxg6 fxg6 18 Kg2 g5 19 Rh1 Kd7! 20 Qd3 Qb6! (the a8-rook joins the attack) 21 Bd2 f4! 22 gxf4 Raf8!? (intending ... Bf5e4, but 22 ... gxf4 was more direct), Blees-M.Gurevich, Gent 1997. Here 23 a4! Na7! (23 ... Bf5? 24 Qb5) 24 Qe2 would have defended, although 24 ... Bf5! 25 e6+ Kc8 26 Rhc1 Be4 27 Kh2 Kb8 still looks nice for Black. 8 bxa3
Black’s capture on a3 has several positive points, but also involves some risk. White gains the bishop pair, and his ‘bad’ dark-squared bishop has no counterpart, opening the possibility of its causing damage. These are meaningful advantages.
Nevertheless, Black is slightly ahead in development (and on the move), with chances of simplifying. Because of White’s weak a-pawns, Black’s knights have great potential on the queenside; for instance, there are outposts on c4 and a4, squares which are also accessible by his light-squared bishop. Finally, Black has no meaningful weaknesses, since d6 won’t be accessible to anything but a bishop, which can be driven away. 8 ... Nf5 The most direct attempt. The alternative 8 ... Qa5+ is also satisfactory: 9 Qd2 (9 Bd2 Qxa3 10 Bxh6!? gxh6 11 Be2 Qa5+ 12 Kf1 b6 intending ... Ba6; e.g. 13 Qc1 Ne7, and now 14 g3 Ba6 15 Bxa6 Qxa6+ 16 Kg2 Nf5 17 Qf4! Rc8 18 Rfc1= is better than 14 Qxh6 Ba6 15 Bxa6 Nf5 16 Qd2 Qxa6+ 17 Qe2 Qa4!=/+), when Black has a wealth of options to consider:
(a) 9 ... Nf5! (now a knight gets to a5 and c4, ensuring at least equality) 10 Bb2 Qxd2+ 11 Kxd2 Na5! 12 Rc1 Bd7 (after ... Bc6 and ... Nc4, White’s capture on c4 would give Black’s bishop a terrific diagonal in contrast to his opponent’s dark-squared counterpart) 13 Bd3 (13 Rc7 Kd8 14 Rc3 h6 15 Bd3 b5!? 16 Rb1 Nc4+ 17 Ke2 a5) 13 ... Ne7 14 h4 Bc6 (or 14 ... b5) 15 h5 h6 16 Rh4 Nc4+! 17 Bxc4 dxc4 18 Rxc4 0-0 19 Bc3 a5 20 Rh1 Nd5 (or 20 ... Bd5 21 Rc5 Bxf3 22 gxf3 b6 23 Rc4 Nd5 – the only question is if Black can win) 21 Rb1 f6! 22 a4 fxe5 23 dxe5 Nf4=/+. (b) 9 ... Qa4!? 10 Rb1 a6 11 Bd3 Nf5 12 Bc2 Qa5 13 Bxf5!? exf5 14 0-0 0-0, Vasicek-Gdovin, Czech League 1995, is equal. (c) 9 ... Qxd2+ 10 Bxd2 Nf5 11 Bc3 0-0 12 Bd3 was Mora Manez-Gomez Gomez, Spanish Team Ch. 1994. Generally, when you have two knights versus two bishops but lack outposts, you will need to open lines in order to create them; so instead of the game’s 12 ... Bd7, Black does well to play 12 ... f6!; for example, 13 exf6 gxf6 14 Bxf5 (14 0-0 Nd6!) 14 ... exf5 15 Kd2 Re8 16 Nh4! f4 17 Rhe1 Rxe1 18 Rxe1 Kf7=. Let’s return to 8 ... Nf5:
9 Bb2 Black can be happy with positions like 9 Qd2 0-0 10 Be2 Na5 11 0-0 Nc4, as well as 9 Be3 Qa5+ 10 Qd2 Qxa3. I.Zaitsev-Mesropov, Moscow 1996, continued 9 Bd3 Qa5+ (not 9 ... Nfxd4? 10 Nxd4 Nxd4 11 Qg4; but 9 ... 0-0 is safe and equal) 10 Bd2 (10 Qd2 Qxd2+ 11 Bxd2 Nfxd4) 10 ... Qxa3 11 Qb1. Here Mesropov’s 11 ... a6 was fine, but Black can also prevent Bb4 and threaten to capture on d4 by 11 ... a5 (11 ... Nfxd4 12 Nxd4 Nxd4 13 Bb4 Qa4 14 0-0 b6 is risky); e.g. 12 Bxf5 exf5 13 00 0-0 14 Qb5 h6! with the idea 15 Qxd5?! Be6 16 Qb5 Rfd8 and ... Bd5. 9 ... Bd7 With the bishop on b2, 9 ... Qa5+ 10 Qd2 Qxd2+ 11 Kxd2 0-0 is possible. Black has the c4 outpost to play with, supported by an open file. Only the potential strength of the bishop pair makes this an unclear situation. 10 Bd3 10 h4 0-0 11 g4?! Nh6 12 Rg1 f6! exploits White’s f-file issues; for example, 13 exf6 Qxf6 14 g5 Qxf3 15 Qxf3 Rxf3 16 gxh6 Rf7, when after an exchange on g7 White has five isolated pawns remaining! 10 ... Nh4!? Black wants to eliminate a potential attacking piece, but this is probably not the best move. Two good alternatives are 10 ... 0-0, intending 11 h4 Na5, and 10 ... Qa5+. 11 Nxh4 Or 11 0-0 0-0 12 Rb1 f5!?+/=. 11 ... Qxh4 12 0-0 Rc8 13 Rc1 0-0 14 Qd2 f5 15 exf6!? 15 f4! keeps Black’s bishop under restraint with a slight edge. 15 ... Rxf6 16 g3 Mossakowski-Blees, Wichern 1995. Although 16 ... Qh5!? worked out well in the game, 16 ... Qh3 would be a more pointed move, with a possible continuation 17 f4!? (or 17 f3 Be8 18 Rfe1!? Rxf3 19 Bf1 Qh5 with the idea 20 Be2 Rf2!) 17 ... Be8! 18 Qg2 Qxg2+ 19 Kxg2 Bg6 20 Bxg6 Rxg6 21 Bc3! (preventing ... Na5 and intending Bb4-d6; 21 f5 is answered by 21 ... Rf6) 21 ... Kf7 22 Bb4 Rd8 with equal chances. 3.6 6 Sveshnikov, the French Advance Variation expert, considers this the critical line. 6 ... Bxc5 6 ... Ng4? 7 Qa4! hits g4 while protecting e5.
dxc5
7 b4 7 Bxh6 gxh6 8 b4?! is an inferior move order because of 8 ... Bf8!, transferring the bishop to an ideal post on g7 (this can be avoided by playing 7 b4 first, as in the main line): 9 b5 (9 Bd3 f6 10 b5 Nxe5 11 Nxe5 fxe5 12 Qh5+ Kd7 13 Qxe5 Rg8 14 g3 Bd6 15 Qe2 Qf6=/+ with the bishop pair and centre, Khalifman-Kaidanov, Moscow 1987) 9 ... Ne7 10 Bd3 Bg7 11 0-0 Ng6 (not only attacking e5 but eyeing f4; for example, 12 Re1 0-0 13 Nbd2 f6 14 exf6 Qxf6 15 Rc1 Nf4) 12 Bxg6!? hxg6 13 c4 0-0 14 cxd5 exd5 15 Nc3 d4! 16 Qxd4 Qxd4 17 Nxd4 Bxe5 with two bishops on an open board, Sveshnikov-Glek, Moscow 1991. 7 ... Bb6 While not as logical, the extraordinary 7 ... Bf8!? still leaves Black with more pieces out and planning to play ... f6:
(a) Sveshnikov gives 8 Bd3!±, but 8 ... Ng4! initiates strong counterplay against e5; for example, 9 Bf4 (9 Bb5 f6! 10 h3 Ngxe5 11 Nxe5 fxe5 12 Qh5+ Kd7! 13 Qxe5 Qf6=/+; or 9 Qe2 Qc7 10 Bf4 f6 11 Bb5 a5!?=/+) 9 ... f6 10 exf6 Qxf6 11 Bg5 Qf7 12 0-0 Bd6 with nice central control and a modest edge. (b) Presumably 8 b5 is better, although Black has satisfactory play after 8 ... Ne7 9 Bd3 (9 Ba3 Bd7 10 Bd3 Qb6 11 Qb3 Ng6 12 Bxf8 Nf4! 13 Bf1 Rxf8=/+ with the idea 14 g3 Ng4 15 Nd4 Ng6 16 f4 f6!) 9 ... Qc7 10 0-0 Ng6 11 Re1 Bd7 (threatening ... Ng4) 12 h3 Bc5 13 Bxh6 gxh6 14 Qd2 0-0-0! 15 Qxh6 Qb6 with counterplay. 8 b5 This takes the pressure off of e5, but it also drives Black’s knight to the kingside where it will play a role in both defence and attack. After 8 Bxh6 gxh6 9 Qd2, attacking h6, Black counterattacks e5 with 9 ... Qc7:
(a) 10 Bb5 Bd7 11 Bxc6 Bxc6 12 Qxh6? is met by 12 ... d4!, and 12 0-0 Rg8 13 g3 0-0-0 14 Re1 d4!? 15 Nxd4 Rg4! brings every black piece into play. (b) After 10 Qf4, 10 ... f5! sets up a later attack on the kingside. Then 11 Qxh6?! Rg8! puts White too far behind in development; e.g. 12 Qh5+ (12 a4 a5! 13 bxa5 Rxa5) 12 ... Kf8! 13 Qh6+ Rg7 14 Qf4 Rg4 15 Qh6+ Kg8-/+. 8 ... Ne7 Moskalenko, always on the alert for creative ideas, notes 8 ... Ng4!? 9 bxc6 Nxf2 10 cxb7 Bxb7 11 Qa4+ Kf8 ‘with an extremely complicated position’. Remarkable. I don’t see anything better for White than 12 Qf4 Rc8!? (the utterly crazy 12 ... Nxh1 13 Ng5 Qd7 14 Bb5! Nf2! 15 Kd2! Bd8! 16 Bxd7 Bxg5 17 Qxg5 Ne4+ 18 Kc2 Nxg5 19 Bxg5 favours White with his powerful bishops; but 12 ... h6 is only mildly better for White following 13 Rg1 g5 14 Qb4+! Kg7 15 Nd4 Ng4 or 15 ... Ne4) 13 Ng5 Qd7 14 Rg1! (14 Bb5 Bc6 15 Ba6 Bb7 16 Bb5 Bc6=) 14 ... Rc4! 15 Bxc4 dxc4 16 Ba3+ Ke8 17 Qxf7+ Qxf7 18 Nxf7 Nd3+ 19 Kd2! Kxf7 20 Rf1+ Kg6 21 Rf8! Rxf8 22 Bxf8 Bxg2 23 Na3 Bd5 24 Bd6 with a slight edge.
After 3.61 3.62 3.61 9 Bxh6 gxh6
8
...
Ne7,
White 9 9
chooses
between: Bxh6 Bd3
This is the main continuation but, again, White pays a price for the weaknesses created in terms of Black’s open lines and his active pieces. 10 Bd3 White’s pawn grab by 10 Qd2 Ng6! 11 Qxh6 Qc7 12 Qg5 is too slow. Among others, Black has 12 ... Qc5 13 Qd2 Bd7 14 a4 Qc7 15 Qg5 h6 16 Qg3 0-0-0 17 Bd3 Rdg8 18 0-0 f5! 19 Ra2 f4 and ... Nxe5. 10 ... Ng6 11 0-0 Qc7! 12 Re1 It’s not smart to surrender the e-pawn when ... 0-0-0 is still an option: 12 Qd2 Nxe5 13 Nxe5 Qxe5 14 Qxh6 Bd7 15 a4 00-0 favours Black’s bishops and centre. 12 ... 0-0 As far as I know, 12 ... f6 hasn’t been suggested or tested, but it looks promising. Briefly, play might go 13 exf6 (13 Bxg6+ hxg6 14 Qd3 f5) 13 ... 0-0 14 c4 (or 14 Qb3 Nf4 15 Bf1 Nh5! and 16 ... Nxf6=, which is simpler than 15 ... Rxf6) 14 ... Rxf6 (or 14 ... Nf4 15 Bf1 Rxf6, transposing) 15 Nbd2 Nf4 16 Bf1 Qc5 17 g3! e5! (beginning a lengthy, semi-forced continuation) 18 Kh1! (18 cxd5 Nh3+) 18 ... Nh3 19 Bxh3 Bxh3 20 Ng1 Be6 21 Rxe5 Qc7 22 f4 (22 Re2 dxc4 23 Ne4 Rf7 24 f3 Rd8) 22 ... Bd4 23 Rxe6 Rxe6 24 Rc1 Qd7 (or 24 ... Qf7!? 25 Ndf3 Be3) 25 Ndf3 Bg7 26 Qxd5 Qxd5 27 cxd5 Rd6=/+. 13 a4
A ‘theoretical’ position, to the extent that we get such a thing after 5 ... Nh6. Now Black has a pleasant choice: (a) 13 ... f5!? 14 Ra2!? (14 Qc1) 14 ... Qg7 15 Qc1 Bd7= 16 c4?! Ba5! 17 Ree2 dxc4 18 Bxc4 Kh8 19 Rac2 Rac8 20 Na3 Bc7 21 Qe3 Bb6 22 Qc1 Rg8! (threatening ... Nf4) 23 g3 f4! 24 Kh1 fxg3 25 fxg3 Rgf8 26 Rc3 Rf5, winning the e-pawn with an ongoing attack, Morosi-Luppi, corr 2002. (b) 13 ... Bd7 14 Ra2 f6 15 Bxg6?! hxg6 16 Qd3 f5 (somewhat in Black’s favour, I think) 17 Nbd2?! Rfc8 18 Rc2 a6 19 Rec1 axb5 20 axb5 Ra3=/+ Sveshnikov-Nikolaev, Moscow 1992. (c) 13 ... f6 is simple and thematic. Play might go 14 Bxg6 hxg6 15 Qd3 g5 (or 15 ... Kg7) 16 Ra2! (16 Qg6+ Qg7 17 Qxg7+ Kxg7 18 exf6+ Rxf6=/+ with multiple ideas including ... Bd7, ... Raf8 and ... g4) 16 ... a6! 17 a5 Bc5 18 b6 Qg7 19 exf6 Qxf6 20 h3 Bd7 21 Nbd2 Rac8 with superior activity. 3.62 9 Bd3 9 ... Ng6 10 0-0 Ng4 transposes. Here 10 ... 0-0 is more common and theoretically equal, but requires more preparation. 10 0-0 Ng6 If Black prefers, 10 ... Qc7 11 Qe2 Ng6 12 Bxg6 hxg6 transposes to the following note.
Ng4
11 Bxg6 11 Qe2 Qc7 12 Bxg6 hxg6 13 Ba3 Bd7 14 Bd6 Qc8, intending ... a6, is complex and dynamically balanced; e.g. 15 a4 a6 16 Qb2! (16 bxa6 Rxa6 17 Qb2 Bc7) 16 ... Qc4!? (16 ... axb5 can become wild after 17 h3 Nh6 18 Na3! Rxa4 19 Nxb5 Ra6 20 Ba3 Nf5 21 Nfd4 Bxb5 22 Nxb5 Rh4! 23 g4? Qc4 24 gxf5?? Qf4–+ intending ... Qg3+) 17 h3 (17 Nbd2? Qe2!) 17 ... Qf4! 18 Qd2 (18 Nbd2? Rxh3!) 18 ... Qxd2 19 Nbxd2 Nh6=. 11 ... hxg6! 11 ... fxg6?! is the normal recapture in the French Defence, but here White is in a good position to exploit it. The following high-level game is quite instructive anyway: 12 h3 (Collins suggests 12 Ba3!, which should favour White; e.g. 12 ... a6 13 bxa6 Rxa6 14 Qe2 Nh6 15 c4!+/=) 12 ... Nh6 13 Bxh6!? (now 13 Ba3 a6 14 bxa6 Rxa6 and ... Nf5 is not convincing) 13 ... gxh6 14 Nbd2 0-0 15 c4 a6! 16 Qe2 axb5 17 cxd5 exd5!? (17 ... Qxd5 is also unclear after 18 Ne4 Qc4 19 Nf6+ Kg7 20 Qb2) 18 Qxb5 Ra5 19 Qb4 Bc5 20 Qc3 Ba7, Grischuk-Bareev, Cannes 2001. Black’s rooks are better than White’s and his bishops are becoming active. Nevertheless, White can maintain equality by bringing a knight to the d4 outpost; e.g. 21 Nd4 Ra4 22 N2f3 Qe7 23 Rfd1 Rc4 24 Qe1=. 12 h3 12 Qa4 Nh6 13 Bf4 Bd7 14 Qb3 a6=/+. 12 ... Nh6 13 Bxh6 13 g4?! f5! 14 exf6 gxf6 15 Qd3 Kf7. 13 ... gxh6 14 c4! 14 Nbd2 a6 15 bxa6 Rxa6 16 Nb3 0-0 17 Qd2 Kg7=/+.
14 ... dxc4 15 Na3 Or 15 Nbd2 c3 16 Nc4 Qxd1 17 Rfxd1 Ke7. 15 ... Qxd1 16 Rfxd1 c3 17 Rac1 Ke7 18 Rxc3 Rd8 19 Rxd8 Bxd8 This is probably equal, though the bishop-versus-knight imbalance keeps it interesting; for instance, 20 Nc4 Bd7 21 Nd6 Bb6! 22 a4 a6 23 Rc2 (23 bxa6 Rxa6 24 Nxb7 Rxa4 intending ... Ra2 or ... Ra1+; the bishops ensure a small edge) 23 ... axb5 24 axb5 (the play is still balanced, though White has to play carefully) 24 ... f5!? 25 Rb2!? (25 Nxb7! Ra1+ 26 Kh2 Rf1 27 Nd6 Rxf2 28 Rc8!? seems to hold for White after the forcing sequence 28 ... Rxf3! 29 gxf3 Bd4 30 f4 Bxc8 31 Nxc8+ Kd7 32 Nd6 Bc5 33 Nc4 g5 34 Kg3) 25 ... g5 26 g3 (26 Nxb7? g4) 26 ... Ra5 27 Nxb7 Rxb5 28 Rxb5 Bxb5 29 Nd6 Ba6. Here the bishops are strong and, even with pawns only one side of the board, Black has some definite winning chances.
Chapter
Four
Tarrasch Variation: Introduction to 3 ... Nf6 With 5 Bd3 1 e4 e6 2 d4 d5 3 Nd2
3 Nd2 constitutes the Tarrasch Variation, which has the reputation of being a safe line for White that still provides an imbalance. It is the second most popular third move, ahead of 3 e5, but has been dwarfed by 3 Nc3, the most ambitious and thematic move. The advantages of 3 Nd2 are clear: it develops a piece and protects e4, maintaining an ideal centre. Unlike 3 Nc3, it does so without being subject to a pin on b4 and leaves the move c3 open. On the other hand, the knight blocks the c1bishop and interferes with the protection of White’s d-pawn by his queen. Those are temporary problems and not structural ones, both resolvable by moving the knight again. But apart from the loss of time that implies, the knight needs somewhere to go to. After White protects or advances his e-pawn, it can logically go to b3, a notoriously poor place for a knight (and indeed, it seldom goes there in the major lines), or to f3. On f3, however, it interferes with the natural development of White’s king’s knight, which might itself need two moves to get to a good square like c3 or f4 via e2. Because of these reorganizational difficulties, Black has great leeway in choosing his own pawn structure and piece placement. Thus even literally eccentric moves such as 3 ... a6 and 3 ... h6 are playable, as well as the mainstream 3 ... Be7, 3 ... Nc6 (hitting the d-pawn), and the most popular choice 3 ... Nf6. We’ll be examining the latter move in this and the next chapter. The theory of 3 ... c5, given in all earlierPTFeditions, hasn’t changed much and I’ve chosen to economize by omitting it in this one. 3 ... Nf6 The idea behind this move is to provoke e5, with themes similar to those of the Advance Variation. The subsequent loss of time by ... Nf6-d7 is balanced by a loss of tempo inherent in White’s Nd2. Black then lays siege to the pawn chain by ... c5 and ... f6. In response, White can either shore up the e5-pawn by f4 (see Chapter 6) or he can exchange on f6, after which the game revolves around his attempts to control e5. The key ideas are best explained by example. 4 e5 Note that 4 exd5 exd5 is already equal for Black, because White’s knight on d2 is poorly placed for an Exchange Variation and will lose time moving again. The main alternative is 4 Bd3 c5.
Now play can transpose to the main line after 5 e5 Nfd7 6 c3. Others: (a) 5 dxc5 gives up the centre too quickly: 5 ... dxe4 6 Nxe4 Nxe4 7 Bxe4 Qxd1+ 8 Kxd1 Bxc5 with active play and at least equality. The natural 9 Ke2, preparing Be3, can be met by 9 ... f5 10 Bf3 (10 Bd3 e5 11 Be3 Bxe3 12 Kxe3 Nc6) 10 ... Nd7 (or 10 ... Nc6) 11 Be3 Bxe3 12 Kxe3 Nf6 and ... e5 with an edge. (b) 5 exd5 Qxd5 (attacking g2 and d4; 5 ... exd5 is also fine, since White’s bishop stands better on e2 or b5 if an isolated queen’s pawn position arises) 6 Ngf3 cxd4 7 0-0 Nc6 is a better version for Black of the variation 3 ... c5 4 exd5 Qxd5 5 Ngf3 cxd4 6 Bc4 Qd6 7 0-0 Nf6 fromPTF3. (c) By far the most important alternative to 5 e5 is 5 c3, which introduces a plan where White aims to play with an isolated pawn and launch an attack. An effective way to interfere with White’s standard set-up is 5 ... cxd4 (5 ... Nc6 6 Ngf3 cxd4 7 cxd4 dxe4 8 Nxe4 Be7 9 0-0 0-0 10 Nc3 b6 is what White is after: a standard isolated queen’s pawn position, with a fullbodied game and equal chances, not that Black need avoid that) 6 cxd4 dxe4 7 Nxe4.
From this position, Black can either simplify or head into a more complex struggle: (c1) 7 ... Nxe4 is a relatively unambitious equalizer. After the capture 8 Bxe4, Black has: (c11) 8 ... Bb4+ is the simplest solution, frustrating White’s attacking ambitions. Contrary to what you may have read, exchanging two sets of minor pieces does not necessarily put the possessor of an isolated queen’s pawn at a disadvantage, but it’s usually enough to eliminate any attacking chances and secure an equal game: 9 Bd2 Bxd2+ (the surest way to play against the IQP; 9 ... Qa5 also suffices; e.g. 10 Ne2! 0-0 11 Nc3 Qb6 or 11 Bxb4 Qxb4+ 12 Qd2 Qxd2+ 13 Kxd2 Nd7 14 Rhc1 Nf6 15 Bf3 Rb8 16 Rc7 Rd8 17 Ke3 Ne8 18 Rc3 b6 19 Rac1 g5 20 h3 Kg7=) 10 Qxd2 Nd7 11 Ne2 (best; White risks coming out slightly worse
after 11 Nf3?! Nf6; e.g. 12 Bc2 0-0 13 0-0 b6! 14 Rfe1 Bb7 15 Ne5 Qd5!? 16 f3 Rfd8 17 Red1 Rac8=/+)
11 ... Nf6 (11 ... 0-0 12 0-0 Qb6 13 Bf3 Rd8 14 Rac1 Nf6 15 Rfd1 Bd7 16 Rc5 Rac8 17 Rdc1 Bc6=) 12 Bf3 0-0 13 0-0 Qb6 14 Rac1 (14 Rfc1 Rd8 intending ... e5; e.g. 15 Rc3 e5 16 Rb3 Qd6 17 Bxb7 Bxb7 18 Rxb7 exd4=), S.Pavlov-Fingerov, Kiev 2002; and the easiest course is 14 ... Rd8 (threatening ... e5) 15 Rfd1 Bd7; e.g. 16 Rc5!? Rac8 17 Rdc1 Bc6 18 b4!? (18 Bxc6 Rxc6 19 Rxc6 bxc6= is a well-known type of position) 18 ... Nd7 19 R5c2 Ne5 20 Bh5 Ra8!=, intending ... Ng6 and/or ... a5. (c12) 8 ... Nd7 (intending ... Nf6 and ... Bb4+ or ... Bd6) 9 Ne2 (9 Nf3 Nf6 10 Bc2 Bb4+ 11 Bd2 Qa5 12 0-0 Bxd2 13 Nxd2 00=) 9 ... Nf6 10 Bf3 Bd6 11 0-0 0-0 12 Nc3 (12 Bg5 h6 13 Bh4?? g5 14 Bg3 g4) 12 ... Bd7 13 Bg5 (13 Re1 h6 14 d5 e5!?=; here 14 Bxb7 is met by 14 ... Bxh2+) 13 ... h6 14 Bh4 (14 Bxf6 Qxf6 15 Ne4 Qe7 16 Nxd6 Qxd6=) 14 ... Rc8=. (c2) 7 ... Bd7 is a good unbalancing option.
For example, White should probably avoid 8 Nf3 Nxe4 9 Bxe4 Bb4+ 10 Bd2 Bxd2+ 11 Qxd2 Bc6; e.g. 12 Bxc6+ Nxc6 13 0-0 0-0 intending ... Qd5 or ... Qd6, when Black has a slight edge. If White tries 8 Nxf6+ Qxf6 9 Nf3, then 9 ... Bb4+ 10 Bd2 Bxd2+ 11 Qxd2 Bc6 is strong. So the safest move is 8 Nc3, when Black has a choice of the following: (c21) 8 ... Bc6 9 Nf3 Nbd7 gives a complex position with equal chances; for example, 10 0-0 Be7
11 Qe2! (11 Re1 0-0 12 Bc2 Qb6 13 Qd3 Rfd8 and White is rather at a loss for good moves) 11 ... 0-0 12 Bd2 Qb6, and if White plays Be3, both ... Nd5 and ... Qa5, intending ... Nd5, are fine, whereas 13 Rfd1 Rac8 14 a3 Rfd8 gives Black a comfortable position. (c22) 8 ... Nc6 9 Nf3 Qb6 threatens ... Nxd4. Then White might sacrifice the d-pawn by 10 0-0 (or 10 Bc2 Nb4 11 Bb1 Bc6 12 0-0 Rd8 13 Be3 Nbd5; note, too, that in this line, 10 ... Qa6!? 11 Bd3 Qb6 offers a repetition, while 11 Bg5 Nb4 has the idea 12 Bb1 Bc6) 10 ... Nxd4 11 Nxd4 Qxd4 12 Nb5 (12 Be3 Qe5) 12 ... Qb6 13 Be3 Qd8!? (safer is 13 ... Qa5 14 Bd2 Qb6 15 Be3=, or if Black has to win, then 14 ... Qd8 15 Rc1 Bc6 16 Bf4 Nd5 17 Be5 Qg5 18 Bg3 Be7=) 14 Rc1 Nd5 15 Bd4 a6 16 Nc3 Bc6 17 Re1 Nxc3 18 Bxc3 Be7!, which is at least equal. 4 ... Nfd7 5 Bd3 White has several options: (a) 5 c3 c5 will very likely transpose to main lines after 6 f4, 6 Bd3, or 6 Ngf3. (b) 5 Ngf3 has independent significance if White answers 5 ... c5 by 6 c4!? (6 Bb5 Nc6 7 0-0 cxd4 or 7 ... Qb6; 6 c3 Nc6 7 Be2!? Qb6 with the idea 8 Qb3 cxd4 9 Qxb6 axb6 10 cxd4 Nb4, or 8 Nb3 cxd4 9 cxd4 a5! 10 a4 Bb4+ 11 Kf1 0-0 12 g3 f6 13 exf6 Nxf6 14 Kg2, Chiburdanidze-E.Vladimirov, Baku 1979, and here 14 ... Bd6! is best).
After 6 c4, White has more space but his centre is vulnerable: (b1) 6 ... cxd4 7 cxd5 exd5 8 Bd3 (8 Nb3 Bb4+ 9 Bd2 Bxd2+ 10 Qxd2 Nc6) 8 ... Nc6 (or 8 ... Nc5 9 Nb3 Nc6) 9 0-0 Be7 (9 ... Nc5 10 Nb3 Bg4 11 Be2 Bxf3 12 Bxf3 Ne6 13 Bg4 Qb6 14 f4 0-0-0 is unclear, Winsnes-Arnason, Reykjavik 1990; here 11 Re1 Ne6 12 h3 Bh5 was equal in Kotronias-M.Gurevich, Manila Olympiad 1992) 10 Re1 Nc5 11 Bb1 Qb6=/+ Smagin-Knaak, Yerevan 1988.
(b2) 6 ... Nc6 7 cxd5 exd5 8 Bd3 (8 Bb5 Qb6 9 Ba4 Qa6 or 9 Qe2 Be7; also possible are 8 ... Be7 and 8 ... g6!?) 8 ... g6! (more enterprising than 8 ... Be7=) 9 0-0 Bg7 10 Re1 (10 Qe2 cxd4 11 e6 Nc5 12 exf7+ Kxf7=/+ Khalifman-Bareev, USSR Ch., Kiev 1986) 10 ... 0-0 11 dxc5 Ndxe5 12 Nxe5 Nxe5 13 Bf1 Be6 14 Nb3 Nc6 15 Bb5 Qf6=/+ with control over d4, which makes the dpawn a force to be reckoned with, Sasu Ducsoara-Radu, Rumania 1992. (c) 5 Ndf3 is an unusual move order: 5 ... c5 6 c3 (for 6 Ne2, see the Classical Variation of Chapter 13, via 3 Nc3 Nf6 4 e5 Nfd7 5 Nf3 c5 6 Ne2, or 5 Nce2 c5 6 Nf3) 6 ... Nc6 7 Nh3?! (7 Bd3 transposes to 7 Ndf3 of 4.1 below; while 7 Ne2 again transposes to Chapter 13; i.e. 3 Nc3 Nf6 4 e5 Nfd7 5 Nf3 c5 6 Ne2 Nc6 7 c3) 7 ... cxd4 (7 ... h6 or 7 ... Qa5!? 8 Bd2 Qb6=/+) 8 cxd4 h6! (intending ... g5) 9 Nf4 g5 10 Ne2 g4 11 Nd2 f6! with the idea 12 Qc2 fxe5! 13 Qg6+ Ke7=/+; for example, 14 Qxg4 Nf6 15 Qh4 Rg8 16 dxe5 Nxe5 17 Qd4!? Nc6 18 Qc3 Kf7 and ... e5. (d) 5 c4 dxc4 (5 ... c5 is normal, when 6 cxd5 exd5 7 Ngf3 Nc6 is note ‘b2’ above) 6 Nxc4 (6 Bxc4 c5 7 Ngf3 Nc6=) 6 ... b5!? 7 Ne3 a6 8 Nf3 Bb7= intending ... c5 next, Wallwork-Lumley, corr 1990. 5 ... c5 6 c3
White should not give up the centre by 6 Ngf3 cxd4 (6 ... Nc6 7 c3 transposes to 4.2) 7 0-0 (7 Nb3 Nc6 8 Qe2 Bb4+) 7 ... Nc6 8 Re1 Bb4!. This favours Black after 9 a3 Ba5 10 Re2 (10 b4 Bc7 11 Bb5 0-0 12 Bxc6 bxc6 13 Nb3, J.Stevenson-Orr, Edinburgh 1999; 13 ... f6! 14 exf6 Qxf6) 10 ... Bc7 11 Qe1 Bb8 (or 11 ... 0-0 12 b4 f6 13 exf6 gxf6!) 12 Nb3 0-0 13 c3?! (13 Bf4 f6; 13 Bb5 Qc7) 13 ... dxc3 14 Qxc3 f6 15 exf6 Qxf6 16 Qc2 h6 17 Bd2 e5-/+ Rybak-Thielen, ICCF email 2001. 6 ... Nc6 Here White makes a fundamental choice. In this chapter, we look at: 4.1 4.2 4.3
7 7 7
Ne2
Ndf3 Ngf3 (Introduction)
In Chapter 5, we’ll examine the main lines following 7 Ne2 cxd4 8 cxd4 f6 9 exf6 (9 Nf4 is included in this chapter). A little strangely, French Defence books neglect the move 7 Qg4.
ECOdoes offer 7 ... f5 (‘!’) 8 Qf4 Be7 from Zvarik-Jurcik, Bratislava 2002, but it’s not obvious why 9 Ne2! shouldn’t afford White good prospects. The best sequence seems to be 7 ... h5! 8 Qf4 cxd4 (8 ... Be7 9 Ne2!) 9 cxd4, and now 9 ... g6!? intending ... Bh6 looks very strong (9 ... Nb4 10 Bb1 Qc7 11 Ndf3! Nc2+ 12 Bxc2 Qxc2 13 Nh3 is unclear); for example, 10 Ngf3 Bh6 11 Qg3, but even in the seemingly advantageous line 11 ... h4!?, with the idea 12 Qg4 Nf6! 13 Qh3 (13 exf6 e5 traps the queen!) 13 ... Nh5!, here 14 g3! is miraculously okay for White after 14 ... hxg3 15 hxg3 Nb4 16 Bb1 Nxg3!? 17 Qxh6! Rxh6 18 Rxh6 Qc7 19 Nb3 Ne4 20 Bg5! with adequate counterplay. Ultimately, in this line, 11 ... Qb6 12 Qh4 Nb4 13 Bb1 Qb5 might be the best try for advantage, but obviously there is much to look at for both sides. 4.17
Ndf3!?
This is a tricky sequence which I suggested for White as a ‘dangerous weapon’ in my French book with that title (DWF). Against it, I’m hesitant to recommend a messy line which, whatever the objective assessment, gives White most of the attacking chances. Fortunately, there seems to be a fairly clear way for Black either to neutralize White’s play or bring him back to the main lines: 7 ... cxd4 Some readers may find one or two of the positions resulting from this exchange too simplified for their tastes. For such players, there are all sorts of fascinating lines beginning with 7 ... Qa5, after which White will either move his king by 8 Kf1!? (a concession which is not entirely one-sided), or gambit a pawn by 8 Bd2 Qb6 9 Ne2!? (after 9 Qc2, John Emms suggests 9 ...
Be7!, having the insight to understand that White can’t find very useful moves, since 10 Ne2?? loses a piece after 10 ... c4 11 Bxh7 Nf8, whereas 10 Nh3? lets Black capture his pawn on d4 without allowing White the usual compensation that follows from Nf3; lastly, 9 Qb3 cxd4 10 Qxb6 Nxb6 11 cxd4 Nc4 12 Bc3 b5 is promising). After 9 Ne2, Black can capture on b2 directly by 9 ... Qxb2 or by 9 ... c4 10 Bc2 Qxb2. In both cases, White castles with hard-to-assess attacking chances for a pawn. This may be the way to go if you want something more exciting than what follows. 8 cxd4 f6 9 exf6 9 a3!? gambits a pawn for activity:
(a) The modest 9 ... Be7 seems to equalize following 10 exf6 Bxf6 11 Ne2 e5 12 dxe5 Ndxe5 13 Nxe5 and 13 ... Bxe5 or 13 ... Nxe5. (b) Alternatively, 9 ... Qb6 10 exf6 Nxf6 11 Ne2 Bd6 12 0-0 0-0 transposes to the old main line 7 Ne2 Qb6 (see the note to 7 ... cxd4 in 4.3) with White committed to the move a3 – this is considered equal. (c) Black can also take up the challenge by 9 ... fxe5 10 Nxe5 Ndxe5 11 dxe5 Nxe5 (11 ... g6!?) 12 Qh5+ Nf7 13 Bxh7 Bd7, which is dynamically equal; for example, 14 Nf3 Bb5 15 Be3 (15 Ng5 Qf6!) 15 ... Qf6 (15 ... Bd3!? leads to a messy draw after 16 Ne5! g6 17 Bxg6 Qa5+ 18 Bd2 Rxh5 19 Bxf7+ Ke7 20 Bxh5 Qb5 21 Rc1!, and 21 ... Kf6 22 Ng4+ Ke7 23 Ne5= or 21 ... Bc4 22 Bg5+ Kd6 23 Bf4 Ke7 24 Bg5+) 16 Bd4 Qf4 (16 ... Qe7!? 17 Ne5 Rxh7 18 Qxh7 Nxe5 19 Bxe5 Qg5 unclear), and White should repeat with 17 Be3 Qf6 18 Bd4 etc, because 17 Rd1 Rc8 18 Qg6 Rc7 19 Ne5? (19 Be5 Qc4) 19 ... Rxh7! 20 Nxf7 Rxf7 21 Qxh7 e5 22 Be3 Qc4 23 Rd2 Qc1+ 24 Rd1 Qxb2 wins for Black. 9 ... Nxf6 10 Nh3!? This is the move I suggested inDWFto stir things up (thanks to John Emms and Jonathan Rowson). 10 Ne2 is the main line of 7 Ne2 (in Chapter 5). 10 ... Qa5+ In the book I concentrated upon the similar but less effective 10 ... Bb4+ 11 Bd2 Qb6. However, 10 ... Bb4+ 11 Bd2 Bxd2+! 12 Qxd2 Ne4 also equalizes; for example: 13 Qe3?! (13 Bxe4 is better: 13 ... dxe4 14 Ne5 0-0! 15 Rd1 Qd5=) 13 ... Qa5+ 14 Kf1 00 15 Bxe4?! (15 Nf4, but still, 15 ... Qc7 16 Ne2 e5 is rather depressing) 15 ... dxe4 16 Qxe4 e5! 17 dxe5 Qb5+ 18 Qe2 (18 Kg1? Bf5 19 Qe3 Bxh3 20 gxh3 Rf5–+) 18 ... Qxe2+ 19 Kxe2 Bxh3 20 gxh3 Rf5 (intending ... Raf8) and Black has the better of it, since 21 Rhg1?! Raf8 22 Rg3?? loses to 22 ... Rxf3! 23 Rxf3 Nd4+. 11 Bd2! After 11 Kf1 Bd6, the exchange 12 Bf4 Bxf4 13 Nxf4 0-0 leaves White’s king vis-à-vis Black’s rook on f8, and his own rook undeveloped on h1. 11 ... Bb4! 12 Bxb4 12 a3 has been seen several times over the past few years: 12 ... Bxd2+ 13 Qxd2 Qxd2+ 14 Kxd2 0-0 and the play is instructive in that it illustrates Black’s resources. Remarkably, White has great difficulties equalizing, if that is indeed possible:
(a) 15 Rhg1 h6!? (both 15 ... e5! and 15 ... Ng4 seem to yield slight advantages after all is said and done) 16 Nf4 Ne4+ 17 Ke3 Nxf2 18 Kxf2 Rxf4 19 Ke3 Rf8 20 Bb5 Ne7 21 Rac1 Nf5+ 22 Ke2 Nd6=/+ Djuraev-N.Nguyen, Al Ain 2008. (b) 15 Bb5 e5! is the real point behind the ... Qa5+/ ... Bb4 sequence: 16 Bxc6 Ne4+ 17 Ke3 exd4+ 18 Nxd4 bxc6=/+ 19 f3 Bxh3 20 gxh3 Ng5. (c) 15 b4 e5! 16 dxe5 Bxh3 17 gxh3, Soors-Gleizerov, Biel 2009; and 17 ... Nd7!? was effective, but 17 ... Nh5 and 17 ... Ne4+ were stronger still. (d) After 15 Rae1, 15 ... e5 still works: 16 Nxe5 (16 dxe5 Ne4+ 17 Bxe4 dxe4 18 Nfg5 Bf5!) 16 ... Bxh3 17 Nxc6 Bxg2 18 Ne7+ Kh8 19 Rhg1 Be4 with a small but definite advantage. (e) 15 Rhe1 is the most interesting try, but still falls a tad short following 15 ... e5! 16 dxe5 (16 Nxe5 Nxd4 has the double idea ... Nb3+ and ... Bxh3) 16 ... Bxh3 17 gxh3 (17 exf6 Rxf6 18 Ng5 Bxg2 19 Bxh7+ Kh8 20 Bd3 Raf8=/+) 17 ... Nd7 (17 ... Nh5 18 Be2!? Nf4 19 Rad1 Rae8=/+ Le Diouron-C.Andersson, Guingamp 2006) 18 Be2 (18 Ng5? h6 19 Ne6 Rxf2+ 20 Re2 Rxe2+ 21 Kxe2 Re8-/+) 18 ... Ncxe5 19 Nxe5 Nxe5 20 Rf1 Rf4 21 Rac1 Nc4+! (21 ... Rd4+ 22 Ke3 Nc6 23 Bg4 Re8+ 24 Kf3 Ne5+ 25 Kg3 Rd3+ 26 f3 Kf7=/+) 22 Bxc4 dxc4 23 Kc3 a6=/+. 12 ... Qxb4+ 13 Qd2
13 ... 0-0 13 ... Qxd2+ 14 Kxd2 0-0 is like the previous note, except that White hasn’t played a3. This ‘lost move’ helps White in the line 15 Rhe1 e5 16 Nxe5 Nxd4?, because ... Nb3+ isn’t threatened and 17 Nf4 retains an edge. But Black can play 16 ... Bxh3 17 Nxc6 bxc6 (17 ... Bxg2 18 Ne5 Nh5!?) 18 gxh3 Nh5 19 Re5 Rxf2+ (19 ... g6=) 20 Ke3 (20 Ke1 Rxh2 21 Rxh5 Rh1+ 22 Bf1 Rf8 23
Ke2 Rh2+ 24 Ke3 Rh1=) 20 ... Raf8 21 Be2 g6 22 Bxh5 gxh5 23 Rg1+ Kh8=. Alternatively, Black doesn’t need to play 15 ... e5; he can also continue with the calm 15 ... h6, or the dramatic 15 ... Ne4+ 16 Bxe4 dxe4, and 17 Rxe4?! e5! or 17 Ne5 Nxe5 18 dxe5 Bd7 19 Rac1 Bc6=. 14 Qxb4 Nxb4 15 Ne5! 15 Bb1 e5! 16 Nxe5 Bxh3 17 gxh3 Rae8 18 0-0 Nh5=/+ intending ... Nf4. 15 ... Nd7 15 ... Ng4!? 16 Ng5 g6 is unclear. After 15 ... Nd7, play could go 16 Ng5?! g6 17 Rd1 (17 Bb1 Nxe5 18 dxe5 Bd7 19 a3 Nc6 20 Nf3 Rxf3 21 gxf3 Nxe5=/+) 17 ... Nxe5 18 dxe5 Bd7 19 Bb1 (19 0-0? Nxd3 20 Rxd3 h6 21 Nf3 Bb5) 19 ... Rac8=/+ Black’s pieces are more active. But 16 Ke2! improves; e.g. 16 ... Nxe5 17 dxe5 h6!? (17 ... Bd7 18 Rhc1 Nc6 19 f4 h6=/+) 18 f4 Bd7 19 Rhc1 Rac8 20 Rxc8 Rxc8 21 Nf2 Kf7 22 g3 Nxd3 23 Kxd3 g5 and Black has only a slight edge. 4.27
Ngf3
The ‘Universal System’, whose widespread use is one of the most important developments in the Tarrasch Variation over the last decade. This is an aggressive and appealing system for those who are tired of slow development and the tiresome task of defending the centre without corresponding activity. White’s idea is to develop rapidly in classical style, leaving open several options to play with, say, 0-0/Re1/Nf1-g3; or to expand on the queenside by a3 and b4; or, finally, to play in Nimzowitsch overprotection style with dxc5 and subsequent control of e5 and d4. The main drawbacks to this system in general are that the knight on d2 lacks a definite future and White’s pawn on d4 can be hard to defend. The following moves look unassuming, but they each initiate surprisingly active counterplay: 4.21 4.22 4.23
7 7 7
... ... ...
g6 f6 h6
The first solution is highly theoretical. The others are rather unusual and may not appeal to everyone (or even stand the test of time), but they allow for more original play. What about other moves? (a) 7 ... Be7 transposes to a solution fromPTF3(via 3 ... Be7), which is also in good shape theoretically, but whose theory has vastly expanded and the space required by its analysis would preclude other options. (b) 7 ... Rb8, with the idea ... b5-b4, is reasonable and almost unexplored. A queenside attack has a better chance of success with fewer of White’s pieces able to swing quickly to the kingside (as opposed to the variation with knights on e2 and f3). (c) Finally, the most radical solution is 7 ... g5!?, which has only recently been taken seriously. The key idea is that, although White hasn’t castled and can play 8 h3, Black can still aim for ... g4, supported by ... h5 and ... Rg8 if necessary. I’ll leave that for enterprising players to investigate on their own. 4.217 ... g6 This is a ‘mainstream’ solution, to which I will devote the most space and time. Black’s desired set-up is ... Bg7/ ... 0-0/ ... cxd4/
...
f6,
dissolving
White’s
centre.
4.211 4.212
White
has
two
8 8
logical
continuations: 0-0 h4
There aren’t many useful waiting moves. 8 Bc2 Bg7 (8 ... b6!?) usually transposes following an early 0-0. White can also play uniquely; e.g. 9 a3 (9 Nb3 b6 10 Bg5 Qc7 11 0-0 a5 and ... Ba6; 9 h4 cxd4 10 cxd4 f6!) 9 ... cxd4 10 cxd4 f6! 11 exf6 Qxf6 12 Nb3 e5! gives Black the initiative, and White can hardly afford to lose more time by 13 dxe5 Ndxe5 14 Qxd5 Be6 15 Qe4. If nothing else, 15 ... Nxf3+ 16 Qxf3 Qxf3 17 gxf3 Bc4! establishes an easy advantage. 4.211
8
0-0
Bg7
9 Re1 Otherwise Black can pursue the above-mentioned plan. For example: (a) 9 b3 0-0 10 Bb2 (10 Ba3 f6 11 exf6 Qxf6=) 10 ... f6 11 exf6 Qxf6 12 c4?! cxd4 13 cxd5 exd5 14 Bb5 Qd6 15 Bxc6 bxc6 16 Bxd4 Bxd4 17 Nxd4 Ne5! and it’s already hard for White to bring his pieces into play, Vogt-Bareev, Budapest 1988. (b) 9 Bb5 isn’t normally a worrisome move in these structures and this is no exception. One solution is 9 ... 0-0 10 Re1 Qb6 11 Qa4 (11 Bxc6 bxc6 12 Nb3 cxd4 13 cxd4 a5) 11 ... Qc7 (to bring the knight to b6 with tempo) 12 Nf1 a6 13 Bxc6 (13 Bd3 cxd4 14 cxd4 f6) 13 ... bxc6 14 Bf4 a5, planning moves such as ... Rb8 and ... cxd4 followed by ... c5.
9 ... 0-0 9 ... Qb6 shows Black’s cards, but the point is to commit White to a strategy. The most independent line is 10 dxc5 Qc7 11 Nb3 Ndxe5 12 Bf4 (12 Nxe5 Bxe5 13 Bh6!? Bxh2+ 14 Kh1 Bf4 15 Bg7 Rg8 16 Bf6, Forster-Hertneck, Biel 1997; in this obscure position, Black can play for simplification by 16 ... Be5 and upon 17 Bg5, repeat with 17 ... Bf4 18 Bf6 Be5, or try for more by 17 ... f6 and ... Kf7) 12 ... 0-0 13 Nbd4 Nxf3+ 14 Qxf3 e5 (14 ... Nxd4 15 cxd4 Qa5) 15 Bg3 Rd8 16 Qe2, Anagnostopoulos-Harley, London 1994; and 16 ... a6 is safe and equal. 10 Nf1 Instead: (a) 10 h4 is a bit strange after White has castled. Although the h-pawn clamps down on ... g5, it weakens the kingside and Black can get active play by 10 ... cxd4 11 cxd4 f6 12 exf6 Nxf6 (12 ... Qxf6 is also good; e.g. 13 Nb3 e5! 14 dxe5 Ndxe5 15 Be2 d4!?=/+ Roiz-Psakhis, Ramat Aviv 2000) 13 Nb3 Ng4! and White has to deal with both ... Qb6 (when f2 is vulnerable) and ... Qd6, aiming at h2 and preparing ... e5. (b) 10 Bc2 is a popular semi-waiting move which performs some useful functions. It gives the d-pawn protection by the queen on d1 (when White moves his knight), and it permits Nb3 because ... c4 won’t be a fork. Black’s easiest course is to stick to his plan: 10 ... cxd4 (although 10 ... a5 is useful and discourages Nb3) 11 cxd4 f6
12 exf6 Qxf6 (attacking the d-pawn and using the open f-file; 12 ... Nxf6 is also playable) 13 Nb3 e5! 14 dxe5 (14 Bg5 Qd6 15 dxe5 Ndxe5 16 Be4 Be6) 14 ... Ndxe5 15 Qxd5+ Be6 and Black has a strong initiative; e.g. 16 Qd1 (16 Qe4 Bg4!) 16 ... Rad8! 17 Nbd2 Nd4-/+.
(c) 10 Qe2 Qb6 (or 10 ... cxd4 11 cxd4 f6 12 exf6 Qxf6 13 Bb5 Re8! 14 Nb3 e5 15 Bg5 Qf5=) 11 Qe3? (innocent-looking, but 11 dxc5 Qd8! 12 Nb3 Ndxe5= was necessary) 11 ... cxd4 12 cxd4 f6 (suddenly things are collapsing) 13 exf6 Nxf6 14 Nb3 Ng4-/+, when Eschmann-Plesec, Swiss Team Ch. 2004, continued 15 Qe2 e5 16 dxe5 Ncxe5 17 Nxe5 Rxf2 18 Qxg4 Re2+ 19 Qd4 Rxe1+ 20 Kf2 Qxd4+ 21 Nxd4 Rxe5–+. 10 ... cxd4 11 cxd4 Qb6 The other move, about equally good, is 11 ... f6. Then after 12 exf6, 12 ... Nxf6! is best: (b1) 13 Be3 Ng4 14 h3?! (but 14 Qb3 Qb6! 15 Qxb6 axb6 will leave Black with the bishop pair under favourable circumstances) 14 ... Nxe3 15 fxe3 e5 with a significant advantage, Coote-Dhar Barua, British Ch., Torquay 2002. (b2) 13 Bb5 Qb6 (or 13 ... Ne4=) 14 Bxc6 bxc6 15 Qc2 Ne4 16 Be3 c5 17 dxc5 Qxb2 18 Qxb2 Bxb2 19 Rab1 Bg7=. (b3) 13 b3 Nh5 14 Bb2 Qb6
This is the sort of piece co-ordination that Black is looking for in the ... g6/ ... Bg7 system. White has done nothing obviously wrong but he has to take care about his d-pawn and kingside. After 15 Bc2 Bd7 16 Ne3, Milosevic-Züger, Swiss Team Ch. 1991, the thematic 16 ... Rxf3! 17 Qxf3 (17 gxf3? Nf4-/+) 17 ... Rf8 18 Qd1 Nxd4 19 Rb1 Nf4 puts considerable pressure upon White’s position, and Black’s centre gives him something a bit better than equality. 12 Bc2 Others: (a) The pawn sacrifice 12 Be3 Qxb2 13 h4 (13 Rb1 Qxa2 14 Rb3 f6 15 Bc1 Qa5-/+ Balinas-J.Watson, San Francisco 1995) has some sting. Black must play resolutely: 13 ... f6! 14 exf6!? (14 h5 fxe5 15 hxg6 hxg6 16 Bxg6 e4 17 Ng5 Nf6 18 Ng3 Ne7 19 Bh5 Nf5=) 14 ... Nxf6 15 Ng3 A.Olsson-Brynell, Gothenburg 2005, and now 15 ... Ng4!; e.g. 16 Rb1 Qxa2!? 17 h5 Nxe3 18 fxe3 Ne7 and Black seems to be defending. (b) 12 Bb1?! f6 13 exf6 Nxf6 14 b3?! Ng4 15 Bb2 e5! (or 15 ... Nxf2 16 Kxf2 Nxd4) 16 Ne3 Nxe3 17 fxe3 e4 18 Nd2 Be6 with an obvious advantage. 12 ... f6 13 exf6 Nxf6
This position has done very poorly for White, since Black is so active. 14 Ba4 Trying to eliminate the pressure on d4. It is particularly worthwhile to play through the notes to this move, as they illustrate a substantial number of the ideas in the entire Bd3 Tarrasch: (a) 14 Ng3 Ng4 (or 14 ... Ne8, attacking d4, when 15 Be3 Qxb2 16 h4?, Eslon-Czerniak, Biel 1976, ran into 16 ... Rxf3!; e.g. 17 Rb1 Qc3 18 Ne2 Rxe3! 19 Nxc3 Rxc3 with the better prospects) 15 h3?? (15 Ne2 Nce5! 16 Nxe5 Bxe5 17 f4 Bf6=/+) 15 ... Bxd4 16 Nxd4 Nxf2 winning, Krayz-Shaked, Berlin 1997. (b) 14 Rb1 Bd7 (14 ... Ng4 15 Ne3 Nh6 16 Rf1? Nf7 won the d-pawn in Zimina-Socko, Istanbul 2003; better 16 Nf1 Nf7=/+) 15 h3 (Black is for choice after 15 Ng3?! Ng4!=/+ or 15 Be3!? Ng4 16 Ba4 Nxe3 17 Nxe3 Rf4) 15 ... Nb4 16 Bb3 Bb5 17 Ne5? (17 Be3 Nd3 18 Re2 Ne4=/+) 17 ... Ne4 18 Be3 Nxf2! 19 Bxf2 Rxf2 20 Kxf2 Bxe5 21 Kg1 Bxd4+ 22 Kh2 Nd3 23 Re2 Bg1+–+ BenjaminM.Gurevich, Cap d’Agde (rapid) 2000. (c) 14 Bb3 Bd7 (14 ... Ng4 15 h3 Nh6=; 14 ... a5!=/+) 15 h3!? Ne4 16 Be3 a5! 17 Rb1 Nb4 (17 ... Rxf3!? 18 Qxf3 Rf8 19 Qd1 Bxd4 – M.Gurevich; 20 f3!=) 18 Ne5 Be8 19 a4 Nc6 20 f3 Nd6 21 g4 Nf7 22 Nxc6 Bxc6 23 Bc2 Qc7 24 Kg2 e5=/+ BabaevGharamian, Cappelle la Grande 2008. 14 ... Ne4 15 Bxc6 bxc6 16 Ng3
16 Qc2 c5! 17 Ng3 cxd4 18 Nxe4 dxe4 19 Qxe4 Bb7 20 Qxe6+ Qxe6 21 Rxe6 Bxf3 22 gxf3 Rac8=/+ Benjamin-Shaked, US Ch., Parsippany 1996. 16 ... Nxf2!?
Alternatively, there is 16 ... e5!? 17 Nxe4 dxe4 18 Nxe5!, Mahjoob-Soln, Bled Olympiad 2002; then Black should play 18 ... Bxe5! 19 dxe5 Qxf2+ 20 Kh1 Be6 21 Bh6 Rf5 22 Qc1! Bd5=. Moskalenko mentions 16 ... Nd6 17 b3 Nb5 with pressure; there might follow 18 Be3 Rxf3! 19 Qxf3 Nxd4 20 Qf4 c5 with dynamic equality. After 16 ... Nxf2, Smirin-J.Watson, San Francisco 1999, went 17 Kxf2 Bxd4+ 18 Be3 Bxe3+ (avoiding 18 ... Qxb2+? 19 Kg1 Qxa1 20 Qxd4 Qxd4 21 Bxd4, when White dominates the dark squares) 19 Rxe3 e5 20 Qb3 e4 21 Nxe4 Qxb3 22 axb3 dxe4 23 Rxe4=. 4.212 8 White goes for the unsubtle attack by h5 and hxg6. If nothing else, this discourages Black from playing ... f6 in many lines. 8 ... h6!?
h4
Black will meet h5 with ... g5, and will also respond to 0-0 with ... g5. There are many alternatives here. I’ll mention a couple, along with some original ideas: (a) 8 ... Qb6 9 h5 (9 dxc5 Qc7 10 h5?! Nxc5 11 Bc2 Bg7 and Black stands well), and now 9 ... g5 10 Nxg5 cxd4 11 cxd4 is a well-known position in which, after many tests, 11 ... Nxd4 12 Ndf3 keeps an edge for White, but 11 ... Qxd4 12 Ndf3 Qb4+ is still viable at the time of this writing. Instead, Romero Holmes once suggested the astonishing move 9 ... Rg8(!), which he calls an interesting alternative.
‘White should do something quickly,’ he says, and gives the moves 10 hxg6 hxg6 11 Qa4. I don’t think this should bother
Black after 11 ... g5 12 g4 cxd4 13 cxd4 f6! (the rook does something on g8 after all, preventing Bg6+). Probably White should consider 10 Kf1, having in mind 10 ... cxd4 11 cxd4 Nxd4?! 12 hxg6 hxg6 13 Nxd4 Qxd4 14 Nf3 with wonderful compensation. Black should try something slower; e.g. 10 ... a5 11 hxg6 (or 11 dxc5 Nxc5 12 Bc2 Bg7) 11 ... hxg6 12 dxc5 Nxc5 13 Nb3 Nxd3 14 Qxd3 Bg7, which is very close to equal. (b) 8 ... Be7, with the idea 9 h5 g5, is one of Black’s most respectable options. I’m not going to go into the details, but did want to point out something in the widely recommended line for White: 9 Kf1 0-0 10 Kg1 and here Sebag-V.Popov, Cappelle la Grande 2006, went 10 ... f6 11 exf6 Bxf6 12 Ng5!?, when 12 ... Qe7 13 Ndf3 cxd4 14 h5! is hard to assess. Instead, Black should opt for 10 ... cxd4! 11 cxd4 f6 12 exf6 Bxf6, and if White plays as in the Sebag game with 13 Ng5, then 13 ... Qe7! 14 Ndf3 Nxd4 15 Nxd4 Bxd4 16 Nf3 is virtually forced, and White barely escapes after 16 ... Bc5 17 h5 (17 Bh6 Rf7 18 h5 Ne5!) 17 ... Ne5! 18 hxg6 Ng4! 19 Rxh7! Nxf2 20 Qe2! Nxd3+ 21 Kf1 Qd6 22 Bh6! Qg3 23 Rg7+ Kh8 24 Rh7+ Kg8 25 Rg7+=. 9 0-0 In some sense this is unnatural, since h4 doesn’t seem to go with castling, yet it works to discourage ... g5. At this key juncture, White has many other reasonable ways to try to use his space advantage: (a) 9 a3, intending b4, is an important alternative, although it’s a tad slow:
9 ... Qb6 (or 9 ... a5 10 0-0 g5! 11 hxg5 hxg5 12 g4 Qb6 and White’s king is a little loose; e.g. 13 Qa4 Be7 14 Re1 Kf8! 15 Kg2 f6! 16 exf6 Nxf6 17 Ne5 Nxe5 18 dxe5 c4! with the idea 19 exf6?? Rh2+; instead, 19 Bg6! Rh2+ 20 Kxh2 Qxf2+ 21 Kh3 Qh4+ draws) 10 0-0 (10 b4!? cxd4 11 cxd4 Nxd4 12 Bb2 Nxf3+ 13 Nxf3, Solovjov-Se.Ivanov, St Petersburg 2009, and a direct solution is 13 ... Nc5 intending 14 ... a5, even after 14 Bxg6!?) 10 ... g5 11 hxg5 (11 b4 cxd4 12 cxd4 g4 13 Nh2 Qxd4 14 Nb3 Qxe5 15 Nxg4 – Postny; then 15 ... Qg7 16 Qe2 Rg8 17 f3 Qc3!? 18 Rb1 Nce5 is at least satisfactory for Black) 11 ... hxg5 12 Nxg5 cxd4 13 cxd4 Qxd4! 14 Ndf3 Qg4 15 Nxf7! Rg8! 16 N7g5 Ndxe5 17 Be2 Nxf3+ 18 Bxf3 Qh4 19 g3? (19 Bh5+ Kd7 20 Qg4=) 19 ... Rxg5 20 Bxg5 Qxg5 with a winning game, Perunovic-E.Berg, Gothenburg 2005. (b) 9 Bc2 is consistent with a strategy of waiting for Black to commit and has the benefit of allowing direct defence of the d-pawn after a move by the knight from d2. The drawback is that White uses a tempo and also yields the f1-a6 diagonal: 9 ... b6!? (intending ... Ba6 directly; Black can also play conventionally with 9 ... Bg7, which is probably best met by 10 Nf1, when 10 ... Qa5 11 Bd2 Qb6 might follow, or 10 a3, inviting 10 ... a5!? 11 Nf1 b5)
10 0-0 (10 c4 Ba6! has the tactical point 11 cxd5 exd5 12 h5?? cxd4! 13 hxg6 Qe7 14 gxf7+ Kxf7–+) 10 ... g5! (10 ... Ba6 11 Re1 cxd4 12 cxd4 Nb4 is complex and playable) 11 hxg5 (11 Ba4 Bb7 12 h5 Rg8!? 13 g4 a6!? planning queenside expansion) 11 ... hxg5 12 g4 a5 13 Re1 Be7 14 Nf1 Ba6 15 Ng3 b5 with dynamic prospects for both sides. (c) 9 Nf1 is another standard idea: 9 ... Qa5 (Black could try to call White’s bluff with 9 ... cxd4 10 cxd4 Qb6 11 Bc2 Bb4+ 12 Ke2 Qb5+!? and 13 Qd3 Qa5, thinking about ... b6 and ... Ba6, or 13 Bd3 Qb6 14 Bc2 Qb5+) 10 Ke2!? (10 Bd2 Qb6 11 Qb3 Qxb3 12 axb3 cxd4 13 cxd4=) 10 ... b5! 11 a3 b4 12 cxb4 cxb4 13 Be3 Ba6 14 Bxa6 Qxa6+ 15 Qd3 Qxd3+ 16 Kxd3 bxa3 17 bxa3 Nb6 18 N1d2 ½-½ Smirin-Gleizerov, Biel 2008. (d) 9 Bb1 resembles 9 Bc2, so just a peek: 9 ... b5 (9 ... b6; 9 ... Rg8!? intending ... g5) 10 a3 a5 11 Nf1, Smirin-Abbasov, Baku 2011, when 11 ... Ba6! 12 h5 g5 13 Ne3 b4 looks fine; e.g. 14 Ng4 bxa3 15 bxa3 cxd4 16 cxd4 Qc7=. (e) 9 Nb1 stops ... g5, although it allows 9 ... Qb6, when White really can’t gambit the pawn and must lose a little time. How that works out isn’t obvious, but in Monroy-Moskalenko, Illes Medes 2006/07, Black chose 9 ... Bg7 10 Be3 a6 (‘!?’ – Moskalenko; if Black wants to expand, 10 ... Rb8! looks appropriate; e.g. 11 Nbd2 b5 12 h5 g5 13 Nh2? cxd4 14 cxd4 Ndxe5! 15 dxe5 d4) 11 Nbd2 b5 12 h5 (12 0-0 b4!?) 12 ... g5 13 Nh2!? 0-0 14 Ng4 (14 0-0 f6 15 exf6 Nxf6= with the idea 16 dxc5 e5 17 Nb3 b4), and now Moskalenko suggests 14 ... cxd4 15 cxd4 f5 16 exf6 Nxf6 17 Nxf6+ Qxf6 18 Nf3 e5!. 9 ... g5 Here the anti- ... g5 recipe 10 dxc5 fails because Black’s queen hasn’t developed to b6.
9 ... Qb6 is playable anyway, if not inspiring; e.g. 10 c4 has a certain logic, now that Black has loosened his kingside and especially g6, but it also allows White’s centre to be destroyed by 10 ... cxd4 11 cxd5 exd5 12 Nb3 Bg7, Rendle-E.Berg, Liverpool
2007; and instead of 13 Bf4 0-0 14 Re1 Re8-/+, here 13 e6! fxe6 14 Bxg6+ Kd8 should have been tried, although Black’s impressive central mass and extra pawn should at the least make up for his king position. 10 hxg5 Others: (a) 10 h5 Qb6 11 c4 cxd4 12 cxd5 exd5 13 Nb3 Bg7 14 Nfxd4! (a pseudo-sacrifice; White has to move quickly or his centre will falls as the king scurries to safety; now Nf5 is threatened) 14 ... Nxd4 15 Be3 Bxe5 16 Re1 and White recovers the piece (by, say, Bf1) with apparent equality, Zhang Pengxiang-M.Gurevich, Hoogeveen 2004. (b) 10 c4 g4!? (safer moves are 10 ... Be7!? and 10 ... Nb4 11 Bb1 g4 12 Nh2 h5 – Moskalenko) 11 cxd5, Gopal-Ganguly, Dhaka 2007; and the conservative 11 ... exd5 12 e6! fxe6 13 Bg6+ Ke7 14 Nh2 h5 with balanced chances is probably best. 10 ... hxg5 11 g4 Qb6 12 Re1!? White uses the familiar gambit idea. 12 ... cxd4 12 ... Be7= is the safe move. 13 cxd4 Nxd4 14 Nxd4 Qxd4 15 Nb3 Qa4!? Black is a pawn to the good and has prepared a convincing answer to 16 Bxg5 Rg8 17 f4: 17 ... Rxg5! 18 fxg5 Qf4 with a strong attack on both the king and White’s weak pawns; for example, 19 g6 (19 Kg2 Nxe5 20 g6 f6! 21 Rf1 Qg5 22 Bb5+ Bd7 23 Bxd7+ Kxd7 and White’s king is so exposed that Black will probably end up with three passed pawns!) 19 ... Qg3+ 20 Kf1 Bg7! 21 gxf7+ Kxf7 22 Qd2 Qh3+ 23 Kf2 Qh4+ 24 Kf1 Kg8=/+ and ... Nxe5 follows. 4.22
7
...
f6
There are several valid but obscure ideas versus the Universal System (e.g. 7 ... g5!?), and this is one of the strangest, in the sense that French books either don’t mention it or don’t take it seriously. Moskalenko is as always an exception: he even plays it! I don’t think any abstract explanation of 7 ... f6 is necessary: Black simply wants to rid himself of White’s pawn chain in the time-tested ... c5/ ... f6 fashion. The resulting positions are not simple, but they are well defined and very little memorization of variations is required. Before continuing, note that 7 ... cxd4 8 cxd4 f6 is similar, when 9 exf6 Nxf6 transposes. The most important deviation is 9 0-0!? (9 Ng5? is flawed due to 9 ... fxg5 10 Qh5+ g6! 11 Bxg6+ hxg6 12 Qxg6+ Ke7 with a winning advantage; e.g. 13 Nf3 Ndxe5! 14 Bxg5+ Kd7 or 13 Ne4 Ndxe5 14 Bxg5+ Kd7 15 Nf6+ Kc7 16 dxe5 Nxe5)
9 ... fxe5 10 Nxe5 Ndxe5 11 dxe5 Nxe5 12 Qh5+ Nf7 13 Re1, when 13 ... Bc5 is perhaps best; this is a move that isn’t available in the 7 ... f6 8 0-0 fxe5 order. The problem with 13 ... Bd6 is that 14 Nf3! g6 15 Qg4! makes it hard for Black to develop, and the idea of Qd4 is annoying. This could provide some incentive for delaying 7 ... cxd4 in favour of 7 ... f6, depending on whether you like 13 ... Bc5. 8 exf6 Instead: (a) As above, 8 Ng5? loses to 8 ... fxg5 (various books and annotated games give 8 ... Ndxe5 ‘!’, but it is itself a mistake) 9 Qh5+ g6 10 Bxg6+ hxg6 11 Qxg6+ Ke7 12 Ne4 (12 Nf3 Ndxe5!, and 13 dxe5 Kd7 or 13 Nxe5 Nxe5 14 Bxg5+ Kd7 15 dxe5 Qe8–+) 12 ... Ndxe5! 13 Bxg5+ Kd7 14 Nf6+ Kc7 15 Ne8+ Qxe8! 16 Qxe8 Bg7 and the queen is trapped, so 17 Bd8+ Kb8 18 Be7 b6 19 Qxh8 Nd3+ 20 Kd2 Bxh8 21 Bd6+ Kb7 22 Kxd3 cxd4 might follow. (b) This time the gambit 8 0-0 fxe5 9 Nxe5 Ndxe5 10 dxe5 Nxe5 11 Qh5+ Nf7 12 Re1 (not 12 Bxh7?? g6! 13 Qxg6 Qh4–+) has a different character after 12 ... Bd6! (12 ... g6 13 Bb5+ Bd7 14 Rxe6+ Be7 15 Qe2, Pitra-Marret, Kuala Lumpur 2011; and now 15 ... Bxb5! 16 Qxb5+ Qd7 17 Qe2 Nd8 would have been about equal)
Now we get some fascinating lines: (a) 13 Nf3 g6 14 Qg4 e5!. Compare 7 ... cxd4 8 cxd4 f6 above: at this point, Qd4! was strong, since the c-pawns were gone; as it is, White has precious little compensation. (b) 13 Bb5+ Kf8! 14 Qe2 g6 15 Nf3 e5. (c) 13 g3!? threatens Qxd5: 13 ... Kf8 (13 ... Qe7 is safer) 14 c4 g6 15 Qe2 Kg7 16 b4! Be5!? 17 Rb1 b5!? unclear.
(d) 13 f4 c4 (13 ... Qe7 14 Bb5+ Kf8=/+) 14 Bc2 Bxf4 and now: (d1) 15 Ba4+ Kf8 16 Rf1 e5 17 g3 b5 18 Bd1 (18 gxf4 g6 19 Qe2 bxa4 20 fxe5 Kg7=/+) 18 ... Qg5=/+. (d2) 15 Rf1 g6 16 Qf3 e5 17 Nxc4 (17 g3 Ng5 18 Qe2 0-0 19 gxf4 exf4 with three pawns and big attack for the piece) 17 ... 0-0 18 Bxf4 exf4 19 Rad1 Qc7 20 Qxd5 Bg4=/+. (e) 13 c4 d4 14 Bxh7 (14 Ne4 Be7; 14 Nf3 g6 15 Qg4 Qf6=/+) 14 ... g6 15 Qxg6 Qh4, Chouraqui-B.Nilson, ICCF server 2009; and instead of 16 Ne4 Qxh2+ 17 Kf1 Be7=/+, White had 16 Bg8! Qxh2+ 17 Kf1 Qh1+ 18 Ke2 Qh5+ 19 Qxh5 Rxh5 20 Ne4 Be7=. 8 ... Nxf6 9 0-0 cxd4 Black avoids 9 ... Bd6 10 dxc5 Bxc5, when both 11 b4 and 11 c4 seem to offer White a small advantage. 10 cxd4 10 Nxd4 requires some care: 10 ... Bd6 11 Re1 (11 N2f3 e5! 12 Nxc6 bxc6 13 Nxe5! Bxe5 14 Re1 Ng4! 15 Bf4 0-0 16 Bxe5 Nxf2 17 Qc2 Nxd3 18 Qxd3 Bf5 19 Qg3 Qd7=) 11 ... e5 12 Nf5 0-0 (12 ... e4!?) 13 Nxd6 Qxd6 14 Ne4! (upon slower moves, Black consolidates his centre) 14 ... Nxe4 15 Bxe4 Be6 and Black’s centre matches the bishop pair in worth. A radical continuation would be 16 c4!? Nd4!? 17 cxd5 Bxd5 18 Qg4 Bxe4 19 Qxe4 Rac8 20 Be3 Nc2 21 Rad1 Qc6 22 Qxc6 Rxc6 23 Re2 Nxe3 24 fxe3 Kf7=. 10 ... Bd6
Moskalenko feels that this position is equal due to the passivity of White’s knight on d2, and the limited practice so far has been fairly promising. Even if White has a technical advantage (which is not at all clear), the variation is at any rate unbalanced. Someone who understands how to handle his pieces in this pawn structure should be able to retain a good deal of play on the board without ceding ground. 11 Re1 Easily the most popular move. Its only drawback is that White takes a defender away from f2. Instead: (a) 11 b3 0-0 12 Bb2 is solid but the bishop has no mobility.
Two good examples: (a1) 12 ... Bd7 13 Rc1 (after 13 Ne5, Black should consider 13 ... Be8, with ideas of ... Bxe5, ... Qb6, and/or ... Bh5) 13 ... Nb4 14 Bb1 Bb5 15 Re1 Nd3 16 Bxd3 Bxd3 17 Ne5 Ba6 18 a3 Qe7 19 b4, S.B.Hansen-M.Richter, Gothenburg 2011; 19 ... Nd7! 20 Ndf3 (20 Nxd7 Qxd7 21 Nf3 Rf5! 22 Ne5 Qe7) 20 ... Rf5 intending ... Raf8 with an edge. (a2) 12 ... Qb6 13 Rc1 (13 Ne5 a5!? 14 Ndf3 a4) 13 ... Bd7 14 Ne5 Be8 15 Bb1? (after 15 Ndf3 Moskalenko suggests 15 ... Nb4, probably intending 16 Bb1 Bh5) 15 ... Bh5 16 Qe1 (16 Ndf3? fails to 16 ... Bxe5 17 dxe5 Ne4) 16 ... Nxd4 17 Bxd4 Qxd4 18 Nef3 Qb6 ½-½ Lastin-Kosyrev, Moscow 2006. Moskalenko likes Black, giving 19 Qxe6+ Kh8 20 Qh3 Rae8-/+. Black’s pieces are very active and he isn’t even down any material. (b) The territorial grab 11 a3 0-0 12 b4 can be met by the central counter 12 ... e5!.
Then White should play 13 Qb3! (13 dxe5?! Nxe5 14 Nxe5? Bxe5 15 Ra2 Bxh2+!) 13 ... Kh8 14 dxe5 Nxe5 15 Nxe5 (15 Bb2 Nxd3 16 Qxd3 a5 17 b5 Ng4!? 18 h3 Bf5 19 Qd4 Nf6=) 15 ... Bxe5 16 Bb2, and Black has the traditional Greek sacrifice 16 ... Bxh2+!? (for the record, 16 ... Ng4 17 Bxe5 Nxe5 should equalize) 17 Kxh2 Ng4+ leading to a series of mostly forced moves: 18 Kg3 Qd6+ 19 f4 Qh6 20 Rae1 Bd7! 21 f5! (21 Re7 d4) 21 ... Ne3 22 Nf3 Bxf5 23 Rxe3! Qxe3 24 Bc4 Qh6 25 Bxd5 Rae8 26 Kf2! Bg4 27 Kg1 Bxf3 28 Bxf3 Qb6+ 29 Kh2 Qh6+=. (c) 11 Nb3 0-0 and: (c1) After 12 Qe2, I like 12 ... a5! 13 a4!? Qb6 14 Bc2 Qc7= with ideas of ... Nb4/ ... b6/ ... Ba6, and a well-timed ... Ng4 or ... Ne4 is still in the air. (c2) 12 Bg5 brings up the most common position besides 11 Re1.
Apart from the obvious 12 ... Bd7 and 12 ... Qc7, Black can take more active approaches, including: (c21) 12 ... a5 13 Rc1 h6 (or 13 ... Qb6 14 a4 Bd7) 14 Bh4 (14 Bd2 is well met by 14 ... Ne4!) 14 ... a4 15 Nc5 a3 16 b3 Bf4 17 Rc3 e5 18 dxe5 Nxe5 19 Be2! Nc6 20 Re1 Ra5; this is equal; 21 Bxf6 Qxf6 22 Qxd5+ Kh8 threatens ... b6 as well as other irritants. (c22) 12 ... Qe8 is a manoeuvre worth noting: Black might play ... Qh5, or else ... Nh5, when the knight is protected against discovered attacks.
11 ... 0-0 12 Nb3 The most logical move. Some characteristic and instructive themes arise after the alternatives: (a) 12 a3 Qb6 13 Nf1?! (not 13 Nb3?! Ng4!; 13 h3 is perhaps best, since 13 ... Nxd4?! 14 Nxd4 Qxd4 15 Nc4! gives White excellent compensation) 13 ... Kh8 (13 ... Bd7 14 Bc2 Rac8) 14 Bc2 Bd7 15 Bg5, Garcia Luque-De la Riva Aguado, Spanish Ch. 1993; and now 15 ... Rac8 was fine, intending to meet 16 Qd3 with 16 ... Rf7=. (b) 12 Nf1 Qc7 (12 ... Qb6!) 13 Bg5 Bd7 14 Rc1 Rae8 (or 14 ... h6 15 Bh4 Qb6 16 Ne5 Rad8=) 15 Bb1 (15 Ne5!+/=) 15 ... Qb6 16 Re2, Vorobiov-Rychagov, Moscow 1997; and 16 ... h6 17 Bh4 Nh5! is equal, intending ... Nf4. 12 ... Qc7 Or 12 ... a5!, limiting White’s options. 13 Qe2 a5!
14 a4! The most obvious move. Instead: (a) Findlay-Carton, British Ch., Blackpool 1988, went 14 Bd2 a4 15 Nc1 Bd7= (15 ... a3!=/+). (b) 14 Bg5?! can likewise be answered by 14 ... a4, because 15 Nc5 (15 Nbd2 Ng4 16 h3? Nxd4!) 15 ... Bxc5! gives Black a powerful centre following 16 Bxf6 (16 dxc5? Ng4 17 g3 e5 18 h3 e4) 16 ... gxf6 17 dxc5 e5=/+. 14 ... Nb4 14 ... Qb6 is a candidate here. 15 Bb1 b6 The sequence of ... a5/ ... Nb4/ ... b6/ ... Ba6/ ... Rac8, or some part of it, arises frequently in this structure when there’s a knight on b3. Play might continue 16 Bd2 Qc4 17 Bxb4 Qxe2 18 Rxe2 axb4 19 a5 bxa5 20 Rxa5 Bd7. White has a very slight advantage in theory because of his minor pieces, but Black can use his bishops and open files to minimize any problems. In general, this variation can be used in order to avoid the memorization that is necessary to play variations based upon ... g5. 4.237
...
h6
This position can arise via 3 Nd2 h6, a specialty of the creative Viacheslav Eingorn, who has contributed some brilliant ideas to its theory. I analyse 3 ... h6 inDWF, when the follow-up 4 Ngf3 Nf6 5 e5 Nfd7 6 Bd3 c5 7 c3 Nc6 transposes to the line before us. Why would Black waste a move on this seemingly inessential advance? Basically, this is an exotic way to play a safer
... g5 (the pawn is defended) without having immediately to deal with the weaknesses that a more committal move creates. White has to declare himself – by contrast, for example, 7 ... g6 can be answered by 8 h4, whereas after 7 ... h6, 8 h4 makes less sense. Also, once Black plays ... g5, he has more options than he does after a line with ... Be7 and ... g5, because his bishop can go to g7 (hitting e5 and supporting ... f6), or in response to dxc5, Black might play ... Bxc5 and save a tempo over lines with 7 ... Be7 and a later ... Bxc5. On the downside, 7 ... h6 is a slow move and makes ... f6 less likely to work because of Bg6+ (or Nh4). Additionally, White can sometimes consider the move c4, because a central breakdown can favour his more quickly developed pieces. Let’s see how this plays out: 8 0-0 One of the problems for White in the Ngf3 Universal System is that he can get stuck without a clear way of proceeding. Usually he finds himself reacting to a committal move by Black ( ... g5 with dxc5 or ... g6 with h4, for example), and this explains why Black can get away with slower moves such as 7 ... a5, 7 ... Be7, and the move in front of you, 7 ... h6. White is in some sense co-operating by playing 8 0-0 and inviting 8 ... g5, so other moves invite consideration: (a) 8 Nf1?! prepares to meet ... g5 with Ng3-h5, but the d-pawn is too vulnerable to support this: 8 ... cxd4 (8 ... Qb6 is at least as good, but it’s more complicated) 9 cxd4 Qb6 10 Qa4 (10 Bc2 Bb4+, forcing 11 Ke2, when 11 ... Be7 will favour Black, who has ... f6 or ... Nb4 at his disposal) 10 ... Bb4+!? (10 ... Be7 11 a3 0-0 12 Ng3 f6 13 exf6 Nxf6=/+) 11 Bd2 Bxd2+ 12 N1xd2 Qxb2 13 Rb1 Qc3 14 Ke2 Qa5 and White has insufficient compensation for the pawn. (b) 8 a3 is a clever approach. It prevents ... Bb4+ and ... Nb4, and tries to cut down on Black’s options. When White doesn’t castle, it’s often good to delay ... g5, so here are two other strategies: (b1) 8 ... Qb6 9 0-0 a5 brings us to a standard position, but with a3 and ... h6 thrown in.
I feel that this is about even. White has to find moves that leave him with compensation if the d-pawn falls. Here are some sample continuations: (b11) 10 dxc5 Bxc5 11 Nb3 Ncxe5 12 Nxe5 Nxe5 13 Nxc5 Qxc5 14 Be3 (14 Bc2 0-0) 14 ... Qc6 15 Bc2 0-0 16 f4 (16 Qh5 f5) 16 ... Nc4 17 Bd4 f6 18 Qd3 f5= with the idea ... Nd6-e4. The bishop pair more or less makes up for the extra pawn. (b12) 10 Qa4 cxd4 11 cxd4 Be7 12 Re1 0-0=, intending ... f6; for example, 13 h3 f6 14 exf6 Bxf6 15 Rxe6 (15 Nb3 Qc7=/+) 15 ... Bxd4 16 Nxd4 Qxd4 17 Qxd4 Nxd4 18 Rd6 Nc5 19 Rxd5 Nxd3 20 Rxd4 Bf5 with an initiative for a pawn; e.g. 21 Rc4 (21 Nf3?! Nc5! or 21 Nc4?! Nc5) 21 ... Rad8 22 Nf1 Ne5 23 Rc5 Rd1 24 Rxe5 Bd3. (b13) 10 Nb1 g5!? 11 h3 cxd4 (11 ... Bg7) 12 cxd4 Nxd4 13 Nxd4 Qxd4 14 Nc3 (this is a standard gambit set-up for White) 14 ... Bc5 15 Nb5 Qxe5 16 Re1 Qf6!? 17 Nc7+ (17 Qc2 0-0) 17 ... Kd8 18 Nxa8 Qxf2+ 19 Kh1 Qg3 intending ... Bd6. In combination with two pawns and the knight trapped on a8, Black has satisfactory play for the rook and probably somewhat better chances. (b2) 8 ... a5!? is another good choice; Black waits for White to commit, while planning ... a4 and preparing for Nb3: 9 0-0 (or 9 Nb1 Be7 10 Bc2 b5!? 11 0-0 Ba6 12 Re1 b4!, when White’s attack is not yet visible; here 9 ... Qb6 10 0-0 transposes to ‘b1’) 9 ... g5! 10 dxc5 Nxc5 11 Bc2 Bg7 12 Re1 Nd7!? (12 ... g4 13 Nd4 Nxe5 is unclear) 13 Nb3 Ndxe5 14 Nxe5 Nxe5 is balanced; e.g. 15 Qh5 a4 16 Nd4 Qf6 or 15 f4!? Nc4 16 Nd4 Qc7 17 Qf3 Bd7! 18 Qxd5 0-0-0 19 Qf3 Kb8. 8 ... g5
This position can be compared to the popular one after 7 ... Be7 8 0-0 g5. By comparison, Black has one fewer piece out, but his pawn on g5 is well defended and his bishop can still go to g7 or e7 or c5, depending upon events. 9 dxc5! A good move which clears d4 for a knight in the event of ... g4. On the other hand, it weakens the centre and helps Black to develop. Alternatives: (a) 9 h3 is perhaps the most natural move: 9 ... h5
(9 ... Qb6!? 10 dxc5 Nxc5 11 Bc2 Nd7 12 Re1 Bg7= has been played a couple of times) and: (a1) 10 dxc5? g4 11 Nd4 Ncxe5 attacks the bishop on d3 and threatens ... gxh3 as well as moves such as ... Bxc5, and ... Qg5 or ... Qh4. (a2) 10 Nb1?! g4 11 hxg4 (11 Ng5 Qb6 12 hxg4 cxd4 13 gxh5 Ndxe5=/+) 11 ... hxg4 12 Ng5 cxd4 13 cxd4 Nxd4 14 Qxg4 Nc6! wins the e-pawn, and 15 Nxf7 Ndxe5! 16 Nxe5 Nxe5 17 Bb5+ Kf7 18 Qf4+ Qf6 will leave Black with the better queenless middlegame. (a3) The natural 10 g4 hxg4 11 hxg4 runs into 11 ... Qb6!, when d4 is hard to defend; for example, 12 Re1 cxd4 (12 ... Be7 is also fine) 13 cxd4 Nxd4 14 Nxd4 Qxd4 15 Nb3 Qa4=/+. If White tries to recover his pawn by 16 Bxg5, Black has the counterblow 16 ... Rg8 17 f4 Rxg5! 18 fxg5 Qf4 with an attack, which after 19 g6! Qg3+ 20 Kf1 Bg7!? 21 gxf7+ Kxf7 ultimately gives him a powerful centre and small advantage. (a4) 10 Nb3!? is wild. Play might continue 10 ... g4!? (10 ... c4 11 Bxg5 Be7 12 Bxe7 Nxe7 looks safer, and roughly equal after 13 Bxc4 dxc4 14 Nbd2 b5 15 a4 b4, or here 14 ... Nb6 15 Ne4 Kf8!) 11 Ng5 Be7 12 Nxe6 fxe6 13 Bg6+ Kf8 14 hxg4 hxg4 15
Qxg4 Rh4 16 Qf3+ Kg7! 17 g3! Qf8! 18 Bf4 Bg5 19 gxh4 Qxf4 and the chances are balanced (entirely computer analysis). (b) 9 c4!? is a thematic move when Black has played the weakening ... g5, and has the idea of exchanging on d5, followed by e6. Still, White’s central support disappears: 9 ... Bg7! (giving the king some room) 10 cxd5 exd5 11 e6 (11 Re1 0-0 12 Bb1 Re8! 13 Qc2 Nf8-/+; both ... cxd4 and trapping the knight with ... g4 are threatened) 11 ... fxe6 12 Bg6+ Kf8 and White hasn’t sufficient compensation.
9 ... Bxc5 Thus Black has saved a move on the conventional lines in which he plays ... Be7 first and then ... Bxc5, even if the extra tempo is the modest ... h6. He has other reasonable continuations: (a) 9 ... Ndxe5 10 Nxe5 Nxe5 11 Nb3! (with the idea f4) 11 ... Nxd3 12 Qxd3 Bg7 13 f4 0-0 (Black’s bishop pair compensates for his weaknesses) 14 fxg5 hxg5 15 Qg3 f6 16 Nd4 (16 h4 e5! 17 hxg5 f5! intending ... f4, even following 18 g6 f4 19 Qh2 Rf6! 20 Qh7+ Kf8 and ... Qe8) 16 ... b6!? 17 cxb6 Qxb6 with a hard-to-assess position. (b) 9 ... Nxc5 10 Bc2 Bg7 11 Nb3 (or 11 Re1 Nd7 12 Nb3) 11 ... Nd7! 12 Nfd4 (12 Re1 transposes to the next note; i.e. 10 ... Bf8 11 Re1 etc) 12 ... Ndxe5 13 f4 Nxd4 14 Nxd4 Nc6 is safe enough; White probably has enough pressure for a pawn, but no more. 10 Nb3
10 ... Bb6 A natural move. Strange to say, having left his knights attacking the e-pawn, Black can also re-route to g7 (with the same tempi as after 9 ... Nxc5) by 10 ... Bf8 with his share of the play; e.g. 11 Bb5 (11 c4?! Ndxe5 12 cxd5 Qxd5 13 Nxe5 Nxe5 14 Be2
Bg7=/+; 11 Re1 Bg7 12 Bc2 Ndxe5 13 Nxe5 Nxe5 and White has barely sufficient compensation; 11 Nbd4 Ndxe5 12 Nxe5 Nxe5 13 Bc2 Bg7 14 f4 Nc6! with dynamic equality) 11 ... Bg7 12 Re1 Qc7 13 Qe2?! (13 Be3!? 0-0 14 Bxc6 bxc6=) 13 ... a6 14 Bxc6 bxc6=/+. 11 Re1 g4
12 Nfd4! 12 Nfd2? Qh4! 13 g3 Bxf2+! 14 Kxf2 Qxh2+ 15 Kf1 Qxg3 gives Black three pawns for the pawn, not to mention the direct attack which follows 16 Bb5 Rg8! 17 Re2 Rg5-/+. 12 ... Ncxe5 13 Bf4 White’s best chance is 13 Bb5! Qf6 14 Qe2 Bc7!? (14 ... Kf8 would aim for more) 15 Nc5 0-0 16 Bxd7 Bxd7 17 Nxd7 Nxd7 18 Qxg4+ Kh7, recovering the pawn. Black is still more active and will use the g-file, but his king is more exposed and the game is equal. 13 ... Qf6 14 Bg3 h5 15 Bb5 15 Qe2 Nxd3 16 Qxd3 h4 17 Bd6 e5 18 Ba3 Rh5 19 Re3 Kd8. 15 ... Ng6 And White has to prove that he has enough for his pawn. 4.37
Ne2
White guards his d-pawn and leaves room on f3 for his other knight. This is the classical and still most popular way of playing. I’ll introduce it here and then examine the most frequently-played variation in the next chapter. 7 ... cxd4 In the first two editions, I recommended 7 ... Qb6, with the main line going 8 Nf3 cxd4 9 cxd4 f6 10 exf6 Nxf6 11 0-0 Bd6 – which, incidentally, can be arrived at via 7 ... cxd4 8 cxd4 f6 9 exf6 Nxf6 10 0-0 Bd6 11 Nf3 Qb6. That was back in 1996, and with time master players largely abandoned the old ... Qb6 system in favour of 7 ... cxd4 and 8 ... f6 without an early ... Qb6. I’ve always kept an eye upon the older move, however, and it bears commenting upon. There have been various challenges to this system (including recent ones), but Black seems to be able to cope with most of them fairly easily. The most often-cited disincentive to ... Qb6 (and the one about which I am most asked) is the move 12 b3 with the idea Bf4. Let me take some time to show how Black might meet this in new ways:
(a) 12 ... 0-0 13 Bf4, when the question is whether Black can take advantage of the move b3, or at least render it meaningless. This is tested in both the following lines: (a1) After 13 ... Bxf4 14 Nxf4, White has generally preserved a modest but one-sided positional advantage based upon control of the dark squares. The trade-off of weaknesses on d4 and e6 is a standard and often balanced one, as is shown in various modern main lines, but chess is a specific game and in general White has been for preference. I should say, however, that the move 14 ... g6! promises more for Black than conventional responses.
This threatens ... Nxd4 and opens up ideas such as ... Nh5 and ... Ng4 (otherwise those moves fail to Bxh7+). For example:
(a11) 15 Bxg6?! hxg6 16 Nxg6 Rf7 17 Qd2 Bd7 18 Qh6 Rh7. (a12) 15 Re1 Nxd4 (15 ... Ng4!?) 16 Nxd4 Qxd4 17 Nxe6 Bxe6 18 Bf5! Qb4! 19 Bxe6+ Kg7=. (a13) 15 Bc2! Bd7 16 g3 (16 Qd2 Qc7= 17 g3?! Nh5!; or 16 Re1 Rac8!? with the idea 17 Nxe6?! Rfe8 18 Nc5 Bg4!; here 17 g3 is better, but not particularly frightening after, say, 17 ... Rf7) 16 ... Nxd4!? (or 16 ... Rf7! 17 Ne5 Re7 intending ... Rf8) 17 Qxd4 Qxd4 18 Nxd4 e5 19 Nfe6 Rfc8 20 Rfe1! exd4 21 Nxd4 Kf7 and White has only a nominal advantage. (a2) At some point it occurred to me that, come what may, Black might try to avoid the exchange of bishops and exploit the minor weaknesses created by b3. 13 ... Be7 proved insufficient for this task, but I think the strange-looking move 13 ... Ba3!? is worthy of closer examination.
Black simply wants to develop normally ( ... Bd7 and pressure down the e-file), and in the meantime has the new idea of ... Bb2, pressuring the d-pawn, and even ... Nb4 at the right moment. These are important benefits, but the cost is also significant: loss of control over the critical square e5. Needless to say, ‘my’ idea 13 ... Ba3 had been tried, most notably by the FIDE Master Heiko Kummerow, who has considerable experience with it. An overview: (a21) 14 Qb1 Bd7 15 Be5 h6!? (15 ... Kh8! with the idea 16 Bxf6 and 16 ... gxf6 or 16 ... Rxf6!, when 17 Bxh7? runs into 17 ... Rxf3! 18 gxf3 Nxd4 19 Nxd4 Qxd4 with too strong an attack) 16 Nf4 Be8 17 Re1 Nxe5 (or 17 ... Nd7!) 18 Rxe5 Bd7 19 Ng6, Mietzner-Kummerow, Bad Wiessee 2001; and the best way to play for a small advantage would be to take over the c-file by 19 ... Rfc8 20 Qd1 Rc7 and ... Rac8. (a22) 14 Nc3 Ne4! 15 Na4 Qd8 (or 15 ... Qa5=) 16 Be3 b6! (this keeps White’s knight out of c5; also possible is 16 ... Bd6 17 Bxe4 dxe4 18 Ng5 Bf4 with the idea 19 Nxe4 Bxh2+) 17 Qb1 (17 Bxe4 dxe4 18 Ng5 h6 19 Nxe4 Ba6 20 Re1 Qh4 21 Nac3 Nb4!) 17 ... Rxf3! 18 gxf3 (18 Bxe4 dxe4 19 gxf3 exf3 20 Qe4 Qd5 21 Qxd5 exd5) 18 ... Ng5 19 f4? Nf3+ 20 Kg2 Qh4! 21 Kxf3 e5!/+. (a23) 14 Ne5
14 ... Bd7 (14 ... Nxe5 is a solid move; e.g. 15 Bxe5 Rf7 16 Qd2 Ng4 17 Bg3 Bd7 or 15 dxe5 Ne4 16 Qc2 Bd7! 17 Bxe4 dxe4 18 Qxe4? Bc6 19 Qe3 Bb5) 15 Rb1 (15 Qd2 Nb4 16 Bb1 Bb5 17 Re1 Nc6=/+ Litwak-Kummerow, Senden 2005; 15 Bc1 Bd6 16 Nxc6 Bxc6 17 Bf4 Ba3 18 Bc1= Ringoir-Kummerow, Bad Zwesten 2006, or here 16 ... bxc6 intending ... c5) 15 ... Nb4 (15 ... a5 16 b4! axb4 17 Qd2 Na5!? intending ... Bb5 is unclear, probably level) 16 Bc1! Bxc1 17 Rxc1 g6 (threatening ... Nxa2) 18 Qd2 Nxd3 19 Nxd7 Nxd7 20 Qxd3 Rae8, planning ... Nb8-c6. (a24) 14 Rb1 has the idea 15 b4, trapping the bishop: 14 ... Nb4 15 Ne5 a5!? (or 15 ... a6 16 Nc3 Nd7!) 16 Re1 Nxd3 17 Qxd3 Bd7 (17 ... Qa6!?) 18 Nxd7 Nxd7 19 Bd2 e5 20 dxe5 Qxf2+ 21 Kh1 Nxe5 22 Qxd5+ Nf7 23 Qxb7 Rae8 24 Qf3 Ne5 25 Qd5+ ½-½ P.Lund-Nilson, ICCF server 2009. (b) 12 ... Qc7 seems a little strange but is equivalent to the main line of the next chapter (7 ... cxd4 8 cxd4 f6 9 exf6 Nxf6 10 Nf3 Bd6 11 0-0 Qc7), except that Black has lost a tempo ( ... Qb6-c7).
Specifically, White’s extra tempo is b3. Is that a useful, productive move, a meaningless one, or even one that can have disadvantages? My feeling is that b3 doesn’t do much good in the ‘normal’ ... Qc7 lines where White plays some combination of Bg5/Be3, Nc3, and Rc1. So an important question is whether b3 and Bb2 is a set-up that Black should be worried about. And while putting the bishop on b2 instead does assist White’s fight for e5, it has drawbacks in terms of activity and kingside defence. Whether Black, a tempo down, can secure the full equality he normally gets isn’t clear but, even if not, the resulting positions are nevertheless playable. In general, White’s inability to bring his queen to b3 or a4 is a drawback to the inclusion of b3; however, that same move saves White a tempo in lines where Black returns a queen to b6 and finds the b-pawn protected
(i.e. rather than hanging on b2). I’ll throw around a few ideas and quote a game or two, but you need to investigate this to see if it’s to your taste: (b1) White can follow the main line with 13 Bg5 0-0 14 Bh4 Nh5!? 15 Qc2 h6 16 Bh7+ Kh8 17 Bg6 Rxf3 18 gxf3 (18 Bxh5 Rfa8 doesn’t seem very different, nor does 18 ... Bxh2+ 19 Kh1 Rf5 20 Bg4 Bd6 21 Bxf5 exf5) 18 ... Bxh2+, whose assessment doesn’t seem to change with the inclusion of b3. (b2) 13 h3 0-0 14 Be3 Bd7 15 Nc3 a6 16 Rc1 Be8= 17 Ng5?! (17 Na4) 17 ... Qe7 and Black stands slightly better; e.g. 18 f4?! Ba3 19 Rc2? Nb4 20 Re2, Miroslaw-Kumor, Krakow 2008; and now 20 ... Rc8 was good, but 20 ... h6! is better still: 21 Nf3 Rc8 and the c3-knight will soon have to retreat or be exchanged, allowing ... Nxd3 and ... Bb5. (b3) 13 g3 0-0 14 Bf4 Ng4 15 Bxd6 (15 Rc1 Bxf4 16 Nxf4 Rxf4! 17 gxf4 Qxf4 is a known type of position with equal chances) 15 ... Qxd6 16 Nf4 (16 Nc3 Qe7 17 Nh4 Nh6 18 f4 Qb4=) 16 ... Rxf4! 17 gxf4 Qxf4 18 Be2! e5! 19 dxe5 Ncxe5 20 Qxd5+ Be6! 21 Qxe6+ Kh8 22 Rfc1 Nxf3+ 23 Kf1 Rd8!? 24 Rc8 Nd2+ 25 Ke1 Nf3+ 26 Kf1 Nd2+ with perpetual check. The moves after 16 ... Rxf4! look forced. (b4) I suspect that 13 Bb2 yields a small edge somewhere, but Black gets plenty of play. Although he is essentially a tempo down on a normal line, sometimes a tempo less can have its benefits. One that might help Black here is that, without a bishop on d7, there are some lines in which his queen can more easily slip away from c-file pressure to transfer to useful squares e7 or f7. 13 ... 0-0 might lead to: (b41) 14 Rc1 Ng4 15 Ng3 (15 h3 Rxf3! 16 hxg4 Rf8 is fine for Black), and an interesting line is 15 ... g6!? (15 ... Bf4 is the usual move) with the idea 16 Ng5!? Qe7! 17 Qxg4 Rf4 18 Qd1 Qxg5=. (b42) 14 Nc3 is perhaps best. Then 14 ... Qd8 asks White what he’s doing, and contemplates ... Qe8-h5; this should be somewhat in White’s favour, but fully playable. Similarly, 14 ... Qe7 is reasonable. 8 cxd4 f6
Black strikes back in the centre. Now, due to the attack on e5, White has only two satisfactory moves. I’ll deal with 9 Nf4 here, and 9 exf6 in the next chapter. 9 Nf4 White threatens Qh5+. For some time this appeared to be harmless, but today the resulting positions tend to be unclear. A third try, 9 f4?!, is too loosening: 9 ... fxe5 10 dxe5 (10 fxe5? Nxd4! with the idea 11 Nxd4 Qh4+ or 11 0-0 Qb6 12 Kh1 Nxe5 13 Nf4 g6 14 Nb3 Nxb3 15 axb3 Bd6 16 Nh5 gxh5 17 Qxh5+ Kd7 18 Bxh7 –ECO, intending Bg5/Rf7 – 18 ... Qd4! 19 Bg5 Qe4! 20 Rf7+ Kc6 21 Rc1+ Kb5 and wins) 10 ... Nc5 (10 ... Qb6 11 Nf3 Nc5 12 Bc2 Be7) 11 Bb1 Qb6 12 a3 (12 Nf3 Bd7 13 Ned4 Ne4) 12 ... a5=/+; White is tied up. 9 ... Nxd4 Destroying White’s centre at the cost of the exchange. 10 Qh5+ Ke7
11 exf6+! 11 Ng6+ hxg6 12 exf6+ lets Black transpose by 12 ... Nxf6, but gives him two further options, namely 12 ... Kxf6!? and 12 ... gxf6 13 Qxh8 Ne5. 11 ... Nxf6 Now 11 ... Kxf6?? 12 Qh4+ g5 13 Nh5+ Kf7 14 Qxd4 is killing; and I don’t like 11 ... gxf6!? due to 12 Nb3 (12 Ng6+ hxg6 13 Qxh8 Ne5 is messy) 12 ... Nxb3 13 axb3; e.g. 13 ... Qe8 14 Bg6!? hxg6 15 Qxh8 Qf7 16 Bd2 e5 17 Bb4+. 12 Ng6+ hxg6 13 Qxh8 Kf7
This is the critical position for 9 Nf4. White has an extra exchange for a pawn. Black doesn’t have many pieces developed, but his knight on d4 is strong and he is ready for ... e5, which not only opens a line for his bishop out but establishes a mobile centre. Now White has an important choice: 4.31 4.32
14 14
Qh4 0-0
Others: (a) 14 f4? Nc6! (14 ... Nh5 is also good, with the trap 15 0-0?? Ne2+ 16 Kh1 Qh4, winning; Barbitskij-L.Smirnov, St Petersburg 2000, went 15 Qh7? Qh4+ 16 g3 Qg4 17 0-0 Bd7!–+ threatening ... Nxg3) 15 Qh4 e5! 16 fxe5 Nxe5 17 Be2 Qb6 18
Qg3 (18 Nf3 Nxf3+ 19 Bxf3 Bf5–+) 18 ... Bg4!-/+ De la Paz-Matamoros, Santa Clara 1996. (b) Be aware of 14 Nf3?? Bb4+, winning the queen. 4.31
14
Qh4
e5
You
might 15
get
lucky! Nf3
This has been played hundreds of times over the years. Note that 15 Nb3 Bb4+ 16 Kf1 Nxb3 17 axb3 e4 gets Black’s pawns moving quickly and must be promising for him. 15 0-0? Be7! leaves White’s queen exposed with few retreats; then 16 Qg3 e4 wins a piece. After 15 Nf3, I’ve chosen to find paths of least resistance rather than enter mainline theory which sometimes goes to move 30: 4.311 4.312
15 15
... ...
Nxf3+ e4
4.311 15 ... Nxf3+ 16 gxf3Bf5 16 ... Qb6 isn’t common but has the idea ... Bb4+ and looks reasonable; e.g. 17 0-0!? Bf5 18 Bxf5 gxf5 19 Bg5 d4!. 17 Bxf5 gxf5 18 Bg5 Otherwise comes ... f4. At this point, 18 ... Qa5+ is the move which has received the lion’s share of attention. Although the current evidence indicates that Black can hold his own in that case, let me point to two fresher moves requiring less memorization:
18 ... Bc5 18 ... d4 is relatively new and hasn’t drawn much attention. It does away with the usual preliminaries and mobilizes the centre. Play can go in many original ways. The key game so far is Rasulov-Volkov, Nakhchivan 2011: 19 0-0 (19 0-0-0!? Qd5 20 Kb1 Qxf3!? 21 Bxf6 gxf6 22 Qh7+ Ke6 23 Qg8+ Kd6; here 19 ... Be7 was possible, or 20 ... Rd8 intending ... Rd6) 19 ... Qd5 20 Qh3 (20 Bxf6 gxf6 21 Qh7+ Ke6 is a typical position for this line, generally good enough for Black; e.g. 22 Qg8+ Kd6 23 Qg6 Bh6! 24 Qxf6+ Qe6 25 Qxe6+ Kxe6 and the bishop on h6 both watches over c1 and can go to f4 with attack – this should be equal) 20 ... Qe6 21 Kh1 (21 Rfd1 Bd6 22 Rac1 Kg8=) 21 ... Rd8 (21 ... Nd7!?) 22 Rad1 Bc5 (22 ... f4! 23 Qxe6+ Kxe6 24 Rc1 Rd7) 23 Rg1 d3 24 Be3 Bxe3 25 fxe3 e4 26 Qg3 Ne8! 27 Qf4 exf3!? (27 ... g6 28 fxe4 Qxe4+ 29 Qxe4 fxe4 and Black has all the chances) 28 Rg3?! Nf6 29 Rxf3 g6 30 Kg1 Qxa2?! (30 ... Qc6!) 31 e4 Qb3=/+ 32 Rd2 Qb6+ 33 Kh1 Rd4 34 Rfxd3 Rxe4 with a strong initiative. 19 0-0-0 This is considered the most challenging move. Others: (a) 19 Rc1 Bb6 (19 ... Qa5+!? 20 Kf1 Bd6=) 20 0-0 (20 Rg1 Qh8 21 Bxf6 gxf6 22 Qa4 Ke6 23 Ke2, De Gleria-Hintze, German League 2009; and instead of 23 ... Qh2?? 24 Rc6+!, leading to mate, Black could secure a nice advantage with 23 ... e4!) 20 ... d4 21 Qh3 Qd7 22 Rfd1 Rd8=. (b) 19 Bxf6 Qa5+!? (19 ... Qxf6) 20 Kf1 Qb5+ 21 Kg2 gxf6=. 19 ... d4!? 19 ... Qb6 is more direct, when Moskalenko gives 20 Bxf6 Bxf2 21 Qa4 gxf6 as unclear; but here either 21 ... Kxf6 or 21 ... Qc5+ is better, the latter continuing 22 Kb1 Kxf6 23 Qd7 Bd4! 24 Qxb7 Qb6! 25 Qxb6+ axb6 with full equality. After 19 ... d4, Yandemirov-Moskalenko, Kemerovo 1985, went 20 Qh3 (20 Kb1 Qb6 21 Rd3 Qe6 22 Re1 Re8 23 Rb3 b6= M.Schneider-Bigg, Swiss Ch. 2009; or 20 ... Qh8) 20 ... Qd7 21 Bxf6. At this point, Moskalenko suggests 21 ... gxf6! 22 f4?! e4 23 Qb3+ Ke7 24 Rhg1 Kd6!. Black’s centre protects him; e.g. 25 f3 (25 Rg6 Rf8 26 Kb1) 25 ... a5! 26 Rge1 Re8 27 Kb1 Kc6!? (27 ... Kc7 28 Qc4 b6) 28 Qa4+ b5 29 Qxa5 Kb7, and Black’s threat to advance his centre pawns secures an even game. 4.312 15 ... e4 Although this has been the usual move order for the idea which follows, I should mention that 15 ... Bb4+! is more accurate, since you can get to the same position without worrying about 17 Bd2 below. That is, 16 Kf1 e4 17 Nxd4 transposes, and here 16 Bd2?? fails to 16 ... Nf3+ 17 gxf3 Bxd2+ with the idea 18 Kxd2 Ne4+!. 16 Nxd4 Bb4+ 17 Kf1 17 Bd2 has been tried, but it’s not relevant because Black should use the 15 ... Bb4+! transposition. 17 ... exd3
18 Bg5 Instead: (a) After 18 Be3, Oreopoulos on ChessPublishing analyses 18 ... Bc5 (or 18 ... Be7) 19 Re1 Qb6! 20 Nf3 Bxe3 21 Rxe3 Kg8 22 Qd4 Bf5 and again, Black has good compensation after 23 h3 Qxd4 24 Nxd4 Rc8!. (b) 18 Nf3 Qe7 19 Bg5 (19 Be3 Bf5 20 Rc1 Re8 21 Kg1, And.Horvath-Dobcsanyi, Budapest 1977; when Moskalenko suggests 21 ... Kg8 22 h3 b6, while the immediate 21 ... b6! may be better still) 19 ... Qe2+ 20 Kg1 Ne4 21 h3 Bf5 22 Kh2 Bd6+ 23 Bf4 Rh8 24 Ng5+ Nxg5 25 Qxg5 Qxf2 0-1 Sadykov-Polivanov, Samara 2011. (c) 18 Qf4?! Qa5!? (18 ... Qb6!? is also playable: 19 Nb3 Bd7 20 Be3 was a draw in Arzumanian-Cech, Brno 2004; but Moskalenko correctly pointed out that Black can continue playing for the win after 20 ... Qb5) 19 Be3 Bd7 20 Kg1 Kg8 (now Black is ready for ... Ne4) 21 Nb3 Qb5 22 Rd1 b6 23 Qg5 Bf5 and White is struggling to find a constructive plan. 18 ... Qb6 Alternatively, Black has 18 ... Bd7 19 Rd1 Qb6! with the idea 20 Bxf6 Re8!. 19 Bxf6 19 f3 Kg8! 20 Bxf6 gxf6-/+. 19 ... Qxf6 20 Qxf6+ gxf6 21 Rd1 b6 22 a3 Bc5
23 Nb5?! Bd7 This position stems from a game by Kosten and original analysis by Moskalenko. Vysochin-Shchukin, St Petersburg 2011, continued 24 Nc3 d4 25 Ne4 Ba4! 26 Rd2 Rc8 27 g3 Be7 28 f3 f5 29 Nf2 Bg5! 30
Rxd3 Be3 with advantage. White can improve somewhat with 23 Nf3, which is double-edged; e.g. 23 ... Ba6 24 g3 Re8 25 Kg2 Re2 26 Rd2 Bc4 is unclear. 23 Nb3 Ba6!? is given a lengthy analysis by Oreopoulos on the ChessPublishing forum, suggesting that the play is equal. A cute line after 24 h4 is 24 ... Re8 25 Rd2 f5! (hindering 26 g3 due 26 ... f4!) 26 Nc1 Rc8 27 Nxd3 Bb4!! 28 Ke1 Re8+ 29 Kd1 Bxd2 30 Kxd2 Re4 ‘and Black’s active pieces more than compensate for his slightly inferior pawn structure.’ 4.32 14 This isn’t objectively better than 14 Qh4, but is arguably harder for both sides to play. 14 ... e5
0-0
15 Nb3 Others: (a) 15 Qh4? transposes to 14 Qh4 e5 15 0-0? (where 15 ... Be7! is strong). (b) 15 Nf3 Nxf3+ 16 gxf3 is a major line, when 16 ... Bf5 17 Bxf5 gxf5 18 Bg5 ultimately leads to a complex equality (e.g. after 18 ... Bc5 and 18 ... d4). If you absolutely have to win, you may want to put some time into these fascinating lines. But I think the more direct 16 ... e4! is both practical and good:
(b1) 17 fxe4 Nh5! 18 exd5! (other moves are weaker; e.g. 18 e5 Qh4 19 Bxg6+ Kxg6 20 Qxf8 Bh3=/+; or 18 Be3? Qh4-/+ 19 f3 Bh3 20 Qh7 Bc5 21 Bxc5 Qg5+, as in Darnstädt-Schäfer, Berlin 1993) 18 ... Qh4 19 Bxg6+ Kxg6 20 Qxf8 Bh3 (or 20 ... Bf5 21
Qa3! Qg4+ 22 Kh1 Re8! 23 Rg1 Re1! 24 Rxe1 Be4+ 25 Rxe4 Qxe4+=) 21 Qf3 Rf8! 22 Qd3+ Rf5 23 f3 Bxf1 24 Qxf1! (24 Kxf1 Qxh2 is more difficult for White), and there is nothing better that 25 ... Qd1+ 26 Qf1 Qd4+=. (b2) 17 Be2!? Nh5 is risky; e.g. 18 Be3 Qh4 19 Rfc1! Bd7! (19 ... Bd6 20 f4 Bh3! 21 Qxa8 Nxf4 22 Qxb7+ is ultimately drawn) 20 Rc7 Rd8 with Black advantage. Instead, 18 Rd1 Be6 19 f4 leads to the forcing line 19 ... Qh4 20 f5! Bd6 21 fxg6+ Kxg6 22 Qxh5+ Qxh5 23 Bxh5+ Kxh5. It’s close, but I’d rather be Black. 15 ... Nxb3 15 ... Ne6 is also possible, counting on the powerful pawn centre even after the exchange of queens; for example: (a) Williams gives 16 Bd2 Bd7 17 Rfe1 e4!? (17 ... Bd6 18 Qxd8 Rxd8 looks satisfactory) 18 Be2 Qb8!? ‘when Black is planning to play ... Qe5 and ... Bd6.’ (b) 16 Be3 b6!? (16 ... d4 17 Rfd1 a5 is complex and dynamically balanced) 17 Rad1 (17 Rac1) 17 ... Bb7 18 Qh4 Bd6 19 f3 (19 Rfe1 Qh8! 20 Qg3 Qh5), Vallejo Pons-Larino Nieto, El Sauzal 2010; and here 19 ... Qh8! 20 Qg3 Qh5= was correct. 16 axb3 Be6!? I like this move, which simplifies things. Moskalenko’s 16 ... e4!? 17 Be2 Qc7 is interesting, and probably slightly better for the first player after 18 Qh4. The traditional variation is terribly complex: 16 ... Bf5 17 Bxf5 gxf5 18 Bg5 (18 Qh3 g6 19 Bg5 Bc5 about equal) 18 ... Bc5 19 Qh3 (the current main line). Then 19 ... Kg6!?, 19 ... Ke6, and 19 ... g6 have all been played, an example of the latter going 20 b4!? Bb6 21 Qb3 Qd6 22 Be3 Rd8 23 Bc5 Qe6 24 Bxb6 Qxb6= M.Richter-E.Berg, Gothenburg 2011. 19 ... Qd7 is the most popular and ultimately equal, but the analysis goes on and on.
17 Bg5 17 Qh4 isn’t frightening; e.g. 17 ... Qb6 (17 ... e4!?) 18 Bg5 e4 19 Bc2 (19 Be2 Qxb3 is unclear) 19 ... Rc8 20 Rac1 Rc6!, intending ... Bd6 in several lines. After 17 Bg5, Pähtz-Kipper, German League 2010, continued 17 ... e4 18 Be2 Qb6 19 Rfc1; and instead of 19 ... Rd8?!, 19 ... Qxb3 wins a pawn and intends ... Qxb2. Williams gives 20 Rc7+ Nd7 21 Rac1 Qxb2 22 Rxd7+!? Bxd7 23 Rc7 Ke8 ‘with an unclear game’. Then 24 Qh7! avoids what would otherwise be a disadvantage, leading to the nice line 24 ... Qb1+ 25 Bf1 e3! (25 ... Qb6 26 Rxb7! Qd6 27 h4!) 26 Bxe3 Rc8 27 Rxb7! Qxb7 28 Qxg6+ Kd8 29 Bg5+ Be7 30 Bxe7+ Kxe7 31 Qxg7+ with a geometrically pleasing perpetual check.
Chapter
Five
Tarrasch
Variation:
Main
Lines
with
5
Bd3
1 e4 e6 2 d4 d5 3 Nd2 Nf6 4 e5 Nfd7 5 Bd3 c5 6 c3 Nc6 7 Ne2 cxd4 8 cxd4 f6 9 exf6 In Chapter 4 we saw 9 Nf4 here. 9 ... Nxf6 10 Nf3
This position has been played regularly and argued over by masters for at least 55 years with thousands and thousands of games. Forming a respectable repertoire from this mass of experience is possible, however, because so much is understood about which variations are worth contesting from both players’ points of view. Here 10 0-0 Bd6 11 Nf3 transposes. A rare option is 11 f4 0-0 12 Nf3; e.g. 12 ... Qb6 13 Kh1 Bd7 14 a3 Rac8 (14 ... Na5 looks like an interesting try, with ideas of ... Bb5, ... Nb3/c4, and ... Rac8; e.g. 15 Ne5! Rac8 16 Nc3 Be8 with the idea ... Bh5) 15 Rb1 (15 Ne5 Be8=/+), Ljubojevic-Timman, Tilburg 1986; 15 ... a6 (or 15 ... Ne7 with the ideas ... Bb5 and ... Nf5) 16 Nc3 Qc7 17 Qe2 Kh8=. 10 ... Bd6 Sometimes Black plays 10 ... Qc7 first, which usually transposes to the main line after 11 0-0 Bd6. But this isn’t a sure way to avoid White’s Bf4 ideas, because White can deviate: (a) 11 Bf4 Bb4+ 12 Kf1! Qb6 (12 ... Bd6 13 g3 e5 14 dxe5 Bh3+ 15 Kg1 Nxe5 16 Bxe5 Bxe5 17 Bb5++/=) 13 g3 0-0 14 Kg2 and Black can potentially become a bit cramped. (b) 11 Nc3!? a6 will transpose in most cases, but 12 Bg5 is of interest: after 12 ... Bd6, 13 Bxf6 (13 Bh4 0-0 14 Bg3) 13 ... gxf6 14 0-0 can be tried. Even though Black has equal chances, he may not want to play this kind of position. 11 0-0 White can still try to frustrate Black with the important move 11 Bf4, before Black can prevent this by ... Qc7. In response, 11 ... Qa5+! interferes with the co-ordination of White’s pieces. (Actually, the more accurate move, strictly speaking, is 11 ... Bxf4! 12 Nxf4 Qa5+!, when 13 Qd2 transposes to 11 ... Qa5+ 12 Qd2; and here 13 Kf1 0-0 is 11 ... Qa5+ 12 Kf1 Bxf4 etc. This bypasses any recapture on d2 with the bishop on f4 – see below; but note that Black shouldn’t play 11 ... Bb4+?! on account of 12 Nc3 Qa5 13 0-0!.) After 11 ... Qa5+: (a) 12 Bd2 Qc7 is the main position of the 11 ... Qc7 variation except that White’s Bd2 has been substituted for 0-0, hardly an improvement. Here 12 ... Qb6 13 0-0 0-0 14 Bc3 Bd7= is also fine – the bishop isn’t well placed on c3. (b) 12 Kf1 Bxf4! (12 ... Qc7 13 Bxd6 Qxd6 14 g3 0-0 is unclear) 13 Nxf4 0-0
Black’s stands better, based upon White’s king position on the open f-file; but it’s instructive to see how this can play out: 14 g3?! (14 h4 prevents ... g5 and keeps Black’s advantage small after 14 ... Bd7 15 h5!? Qb6 16 Qb3) 14 ... Qb6 15 Bc2 (15 Qd2 Nxd4 16 Nxd4 e5-/+) 15 ... g5! 16 Nxg5 Nxd4 (intending ... e5) 17 Bxh7+ (17 Nxh7 e5!) 17 ... Nxh7 18 Nxh7 e5! 19 Nxf8 (19 Nxd5?? Bh3+ 20 Kg1 Qxb2) 19 ... exf4 20 Qh5 Bf5 21 Nd7 Qb5+ 22 Kg1 Ne2+ 23 Kg2 Qxd7, with the idea 24 Qg5+ (24 Qxe2?? f3+! wins the queen) 24 ... Kh8 25 Qf6+ (25 Rhe1 Re8 26 f3 d4) 25 ... Kh7 26 Qh4+ Kg6 27 f3 Bc2! 28 g4 Re8 29 Qh5+ Kg7 30 Qg5+ Bg6-/+. (c) The best White can do is 12 Qd2 Qxd2+ (12 ... Bb4 13 Nc3 0-0 14 0-0 Ne4 15 Qe3!+/=) 13 Kxd2 (13 Bxd2 could have been bypassed by playing 11 ... Bb4+ first, although 13 Bxd2 0-0 is equal anyway after 14 Rc1 Bd7) 13 ... Bxf4+ 14 Nxf4 0-0
This queenless middlegame is of some significance if Black aims for a ... Bd6/ ... Qc7 set-up. Each White attempt raises issues: (c1) 15 g3 Ne4+ 16 Ke3 g5! 17 Ne2 Nxf2! 18 Kxf2 g4 19 Nf4 gxf3 20 Ke3 e5!=/+. (c2) 15 Ne2 Bd7 (or 15 ... Ne4+ 16 Ke3 Nb4!?; while 15 ... Ng4 16 Rhf1 transposes to 15 Rhf1) 16 Nc3 Nh5!? (or 16 ... a6; or 16 ... Rac8) 17 Ne2 (17 Rac1 Nf4=/+) 17 ... Nf4 18 Nxf4 Rxf4 19 Ke3 Raf8=/+ with the ideas ... h6 and ... g5, and in some cases ... Be8. (c3) 15 Rhf1 Ng4!? (or 15 ... Bd7=) 16 g3 (16 Ne2 e5 17 dxe5 Ngxe5 18 Nxe5 Nxe5 is also equal, because the natural 19 Nd4!? runs into 19 ... b6!, intending ... Nxd3, and White is forced to move the bishop, when 20 Be2 Rf4! follows) 16 ... e5 17 dxe5 Ngxe5 18 Nxe5 (18 Be2 Nc4+) 18 ... Nxe5 19 f3 (19 Nxd5?? Rd8; 19 b3? g5! 20 Ne2 Rf3! 21 Bc2 Bh3) 19 ... Bd7 (or 19 ... Bf5 20 Bxf5 Rxf5, intending ... g5, or 21 h4 h6! 22 h5 Nc4+=) 20 Rae1 Rae8 21 h4 Bc6 22 Be2 d4=.
(c4) 15 Rhe1 is arguably the most important: 15 ... Ne4+ 16 Bxe4 Rxf4 17 Bd3 Bd7!? (this is the solution which leaves the most play on the board; instead, 17 ... Nxd4 18 Ke3! Nxf3 19 Kxf4 Nxe1 20 Rxe1 Bd7 21 Ke5! is an ending from the famous game Tal-Timman, Montpellier Candidates 1985, won by White; then 21 ... Bc6 and 21 ... Kf7 are both equal)
Between Black’s f-file and White’s weak d-pawn, the first player is under pressure: (c41) 18 Ke3 Raf8 19 Re2?! Be8!? (a strong alternative was 19 ... h6!, which intends ... Be8, but without allowing Ng5, and also prepares ... g5) 20 Rd2 (20 Ng5! Nxd4 21 Nxh7! Nxe2 22 Nxf8 Kxf8 23 Kxe2=), Sorokin-Dreev, Borzhomi 1988, and here Black had 20 ... e5! 21 Nxe5 (21 dxe5 Bh5 22 Be2 Nb4! threatens mate with 23 ... Re4) 21 ... Nxe5 22 dxe5 d4+ 23 Ke2 Rxf2+ 24 Ke1 Rxd2 25 Bc4+ Bf7 26 Bxf7+ Kxf7 27 Kxd2 Ke6=/+. (c42) Tal suggested that 18 Bb5 Raf8 19 Bxc6 Bxc6 20 h3 favours White (and not 20 Rxe6 Rg4!=/+), but 20 ... Be8 21 Rxe6 (21 Re5 Bg6 22 Rc1 Be4=/+) 21 ... Bh5 22 Re7 Bxf3 23 gxf3 Rxf3 24 Rf1 R3f7 25 Rxf7 Rxf7 gives Black a tiny edge. Finally, we return to 11 0-0:
With the move 11 0-0 we have reached a fundamental position. Since the same structure arises in so many French Defence variations, I should say a few words about it. To start with, the play is remarkably flexible: both sides can play in all three sectors of the board, although Black is more likely to concentrate upon the centre and kingside. Structurally, the layout is fairly simple: Black has a backward e-pawn, a hole on e5, and bad bishop on c8. White has an isolated pawn on d4. This sounds like a pretty good trade-off for the first player, but there are some compensating considerations. First, a pawn on one’s third rank is easier to defend than one on the fourth rank, and it turns out that Black’s e-pawn is seldom seriously threatened, especially in comparison to Black’s threats against White’s d-pawn. Secondly, Black has a central pawn majority which covers
more central squares, and the potential freeing move ... e5 requires White’s constant supervision. Notice that in an Open Sicilian in which Black has played ... e5 (such as the Najdorf), White has a similar outpost on d5, and Black’s bad bishop on e7 plays the same role as the one on c8 in the French: protecting the backward pawn (initially, at any rate). In the Sicilian, the freeing move ... d5, supported by a knight on f6 and bishop on e6, corresponds to ... e5 as supported by Black’s knight on c6 and bishop on d6. In both cases, an intruder on the outpost will often be captured immediately so as to close off the open file. On a less abstract level, White has many possible set-ups and all his pieces can go to various spots, but a few key ideas are: trading off Black’s dark-squared bishop by Bf4 or Bg5-h4-g3 (thus weakening his hold on e5); playing on the queenside with Nc3, Rc1, and Na4; and, when given the opportunity, aiming his pieces kingside with moves like Ng3 or Nf4, Qc2, and perhaps even Re1-e3. For his part, Black will often try to pressure the d-pawn with a queen on b6, knights on c6 and f5, and/or a bishop on g7 or f6. He uses the f-file to assist in a kingside attack, doubling on it or, surprisingly often, sacrificing a rook on f3. His f6knight may also participate from e4 or g4. Another manoeuvre that we see in just about every variation is ... Bd7-e8-g6/h5. It goes without saying that the fight for e5 is always in the background. After 11 0-0, Black has two main moves: 5.1 5.2
11 11
... ...
Qc7 0-0
As in many openings, the main lines are the ones with the most interesting and challenging content, arguably giving you the best chance to defeat a strong opponent. Their study will certainly increase your chess understanding the most. These are represented by 11 ... Qc7, which, as you will see, is associated with some complex material including some unavoidable tactical mêlées. If you simply don’t want to put time into this much material, however fascinating, I’ve offered a much simpler option stemming from 11 ... 0-0. That inevitably lacks the heft that the main lines carry, but it might be exactly what you need to get out of the opening in reasonable shape. I should say that all of the themes which are expressed in the 11 ... 0-0 option are represented in the 11 ... Qc7 repertoire as well, so the study of either variation will inform your study of the other. 5.111
...
Qc7
Black prevents Bf4. This is one of the most important lines in the Tarrasch Variation. I like it because both sides enjoy flexible positions in which they can choose from diverse strategies. The other main lines with 11 ... 0-0 12 Bf4 Bxf4 13 Nxf4 Ne4, while playable, tend to be forcing and theoretically heavy in the extreme; they also contain some hard-to-avoid draws. In fact, I’ve chosen to deviate at an early stage with my recommendation for Black in the 11 ... 0-0 variation to avoid those problems (see 5.2). With 11 ... Qc7, Black’s play will come from his active development, open f-file, and the threat of playing ... e5 at the right moment, a freeing move whose prevention requires White to tie up resources. The usual trade-off exists between Black’s weak pawn on e6 and White’s on d4. Now the play splits into four main moves: 5.11 5.12 5.13 5.14
12 12 12 12
Bg5 Nc3 g3 Bd2
Naturally there are other legitimate moves, but they aren’t used much by stronger players, except in the context of transpositions to main lines. The main thing is to understand Black’s strategy from other variations and apply it here: (a) 12 h3, to play Be3 without being harassed by ... Ng4, is sound but allows Black freedom to develop: 12 ... 0-0 13 Be3 Bd7 (with White’s pieces passively placed to prevent ... e5, the manoeuvre ... Be8-h5/g6 is logical; another example is 13 ... a6 14 Rc1 Bd7 15 a3 Kh8 16 Nc3 Be8 17 Ng5 Qe7 18 f4 h6 19 Nf3 Nh5 20 Ne5 Ng3 21 Re1 Bxe5 22 fxe5 Nf5 23 Bxf5 Rxf5= DaniliukChuprikov, Voronezh 1991) 14 Rc1 Be8 15 Ng5 (15 Ne5 Bxe5 16 dxe5 Qxe5 17 Bc5 Rf7 18 Re1 Nd7!, or here 17 Re1 Bh5=) 15 ... Qd7!? (15 ... Qe7!) 16 f4 (16 Nf4!? Bxf4 17 Bxf4 h6 18 Nf3 Bh5) 16 ... Ne7 17 Qc2 Bg6 18 Bxg6 Nxg6= Daniliuk-Mesropov, Voronezh 1991; with the light-squared bishops off, Black has no problems. (b) 12 Re1 will usually transpose to another line, but generally White doesn’t want to commit to this move too early since it can be dangerous to leave f2 unprotected. After 12 ... 0-0, the rook move is sometimes combined with 13 Ng3, either by this order or 12 Ng3 0-0 13 Re1. Then 13 ... Qb6 puts a little pressure on the centre (13 ... Bd7 is naturally playable too; and even 13 ... e5 14 dxe5 Nxe5 15 Nxe5 Bxe5 is fully equal, as several games have shown); for example, 14 b3 (to develop the c1-bishop) 14 ... Ng4! 15 Re2 (15 h3? Nxf2! 16 Kxf2 Nxd4-/+; 15 Bxh7+?! Kxh7 16 Ng5+ Kg8 17 Qxg4 Qxd4=/+) 15 ... Bd7 16 Bb2 g6 17 Bc2 Rf6 and ... Raf8 with active play. (c) The immediate 12 Ng3 0-0 13 Bd2 is an ultra-conservative development which is sound but easy to meet; e.g. 13 ... Bd7 14 Rc1 Qb6 15 Bc3 a5!? (15 ... Bf4 16 Rc2 Rac8=; 15 ... Rac8) 16 Re1 Bf4= Guseinov-Zarubin, Kherson 1990, with the idea 17 Rc2 Nb4 18 Bxb4 axb4=/+. (d) 12 b3 is slow, and I doubt that the bishop belongs on b2 anyway: 12 ... 0-0 (12 ... e5!?) 13 Bb2 Ng4!? (13 ... Bd7 14 h3 Nh5 15 Nc3 a6 16 Ng5! Nf6 17 Re1 Rae8 and moves like ... Bf4 followed by ... e5, or ... Qb6 in conjunction with ... e5, make Black’s game easier to play) 14 h3 Rxf3 15 hxg4 Rf7! 16 Rc1 Bd7 17 Qc2 g6 18 Qd2 Raf8; if anyone stands better here, it’s Black. 5.1112
Bg5
This is easily the most popular move and the main line of the Bd3 Tarrasch with ... Qc7. 12 Bg5 has the logical idea of Bh4-g3 to exchange off Black’s good bishop. Having achieved that, White can start thinking about working on the queenside with Rc1, or occupying the outpost on e5 by means of Re1. If Black plays ... e5 first, he gets saddled with an isolated queen’s pawn; remember that isolated pawns tend to be somewhat weaker after simplification. That’s the general strategy, but of course it takes time. Black tends to rely on piece play in these lines, so he’ll usually aim his forces towards White’s king to provoke weaknesses, and then strongly consider ... e5 in order to bring his other pieces into the attack. If need be, he can also develop slowly by ... Bd7 before undertaking further action. 12 ... 0-0 Now we get a split into the two main and extremely popular moves: 5.111 5.112
13 13
Bh4 Rc1
5.11113 Bh4 13 Qc2 h6 14 Bh4 Nh5 transposes, but allows 13 ... Nb4. This normal order with 13 Bh4 does let Black play 13 ... e5, but extensive theory shows that it comes up short for the second player. 13 ... Nh5 14 Qc2 The only really pointed move. 14 Nc3 a6 would transpose to 5.12; while after 14 Bg3 Nxg3 15 hxg3, Black has equalized
with 15 ... Qb6, 15 ... Bd7, and 15 ... a6 followed by ... g6 and ... Qg7. 14 ... h6
Now 5.1111 5.1112
comes
the
last 15 15
important
division: Bg6 Bh7+
Alas, I will use up a lot of paper on these two moves, which have been the centre of theory and practice for years, and are properly considered the current main line of the 11 ... Qc7 Tarrasch. ‘Current’ is the operative word, however. I believe that both variations work out to Black’s satisfaction, and that eventually they won’t attract White enough to justify their frequent employment. Unfortunately, just as is the case with mainline Sicilians, King’s Indians and other complex tactical openings, you need to know how to meet the most critical moves (or at least have a good reference for them), so what follows is unavoidable. Before getting to those, here are two interesting options: (a) The logical 15 Rac1 can arise from the order 13 Rac1 h6 14 Bh4 Nh5 15 Qc2. Black can reply by 15 ... Bd7 with a satisfactory game; but in general White tends to avoid this line because of a move/strategy that you’ll get used to: 15 ... g5!.
In modern practice we see the advance ... g5 in French Defence variations of all types, from irregular to main lines. Black grabs space and threatens ... g4 in many positions, weakening White’s defence of the d4-pawn. In the ... Qc7 Tarrasch, a major
idea is ... Qg7, targeting that same d4-pawn and defending Black’s king, whose defences have been weakened. Sometimes it’s easier to play ... Nh5-f4 or ... Be8-g6. This particular version of ... g5 also allows Black to exchange White’s influential bishop for his knight and establish equality: 16 Bg3 (16 Bg6?! Nf4 17 Nxf4 Bxf4 18 Bg3 Qg7 and ... d4 is vulnerable) 16 ... Nxg3 17 Nxg3! (17 hxg3 Qg7 18 Qc3 a6 stops Bb5 and intends simply ... Bd7 and ... Rac8) 17 ... Qg7 (17 ... a6, versus Bb5, helps secure e5 and is a good move; e.g. 18 Rfe1 Rxf3!? 19 gxf3 Bf4 20 Rcd1 Nxd4!) 18 Nh5! Qe7 19 Rfe1 Bd7 20 Bb5! Be8!? (20 ... Qf7 21 Bxc6 Bxc6 22 Ng3 Bb5! gives balanced play) 21 Bxc6 Rc8 22 Ng3 Rxc6=. (b) 15 Bg3 simply cuts down on White’s options after 15 ... Nxg3: (b1) The recapture 16 Nxg3? leaves d4 unprotected and makes 16 ... Rxf3 effective: 17 gxf3 Nxd4 (or 17 ... Qb6, if you want to leave the queens on; White can discourage that by 18 Qa4, but Black will still gain excess compensation with 18 ... Bd7 19 Bb5! Rf8 20 Bxc6 Bxc6 21 Qd1 Rf4 22 Qe2 Bd7! 23 Rad1 Rxd4 24 Rxd4 Qxd4=/+) 18 Qxc7 Bxc7 19 Kg2 (19 Rac1 Bf4) 19 ... Bd7 20 Rad1 Rf8! 21 Bh7+ Kxh7 22 Rxd4 Bc6 and ... e5 will soon follow. (b2) Alternatively, 16 hxg3 g5 sets up Black’s standard idea of ... Qg7 (Williams notes that 16 ... a6!? stops White from playing Bb5 and prepares ... g5 anyway). Then 17 Bb5 could be answered by 17 ... Ne5!? 18 Qxc7 Nxf3+ 19 gxf3 Bxc7 and the bishops will serve Black well. 5.111115 Bg6 Rxf3! Unfortunately, we are now forced to enter into high theory full of critical lines. 16 gxf3 White can play more calmly with 16 Bxh5, when the main line is 16 ... Bxh2+ 17 Kh1 Rf8!? (for 17 ... Rf5!? 18 Bg6 Bd6 19 Bxf5 exf5, as in Andreikin-Kosyrev, Moscow 2005, compare the next section with 15 Bh7+ instead of 15 Bg6 – with the bishop pair and an extra pawn, Black has enough for the exchange, particularly because his king is better placed on g8 than h8 in some lines; here 18 g4!? Rxh5 19 gxh5 Qf7 is fine for Black – Williams) 18 g3 g5 19 Bxg5 (Kotronias-S.Williams, Cork 2005, saw 19 Qg6+ Qg7 20 Kxh2!? Qxg6 21 Bxg6 e5! 22 f4?! Bg4 23 Nc3 Nxd4 24 fxg5 hxg5 25 Bxg5 Kg7=/+) 19 ... hxg5 20 Kxh2 Qg7 with equality; e.g. 21 Kg2 e5! 22 dxe5 Bf5 23 Qb3 Rad8=. Again, compare the next section with 15 Bh7+ Kh8 16 Bg6. 16 ... Bxh2+ 17 Kh1 Nf4!
18 Ng3 One of the most-analysed positions in the ... Qc7 Tarrasch. It’s pretty much worked out by now, with equality as the verdict, though as Williams points out, it’s White who is usually under pressure. Alas, you absolutely can’t get away from memorizing here, but try to pick out a single path rather than absorb everything. White has numerous other moves, all except one advantageous for Black: (a) 18 Kxh2 Nxg6+ 19 Bg3 Qf7 gives Black excellent chances; e.g. 20 f4?! (20 Kg2 Bd7=/+; 20 Rh1 e5 21 dxe5 Ngxe5) 20 ... Nge7 21 Rh1 Nf5=/+ 22 Rad1?!, Jankovskis-Szczepanski, Porz 1993; and now 22 ... h5!-/+ was very strong. (b) 18 Bg3 Bxg3 19 fxg3 Nxe2 20 Qxe2 Qxg3=/+. (c) 18 Nxf4 Qxf4 19 Bg3 Bxg3 20 fxg3 Qxg3=/+ is pretty much the same story. (d) 18 Rad1! Nxe2 19 Qxe2 Qf4 20 Bg3 Bxg3 21 fxg3 Qxg3 results in equality with chances for both sides.
18 ... e5 Black has a choice here. The text is the most straightforward move, threatening the bishop on g6. It should be said that 18 ... Qb6, formerly an important move, is inferior. But 18 ... Qd6 19 Rad1 e5 transposes to the main line. The alternative 18 ... Bxg3 should equalize if you ever need a back-up.
I’ll give some brief but essential details: 19 Bxg3 (19 fxg3 Nxg6 20 Qxg6 Nxd4 21 Rac1 Nc6 22 f4 Qf7=; 19 Bh7+ Kh8 20 fxg3 Qf7! 21 gxf4 Qh5=) 19 ... Bd7 20 Qd2 Rf8 21 Rg1 (21 Rac1?! Qd6! 22 Bc2 e5 23 dxe5 Qxe5-/+; while on 21 Rad1, Black has 21 ... Qd6 with ... Be8 or 21 ... Qd8 22 Bxf4 Qf6 23 Bc2 Qxf4 24 Qxf4 Rxf4 25 Ba4 g5 26 Rfe1 Kf7=) 21 ... Qd6 22 Rg2! Rf6 23 Rh2! Rxg6 24 Bxf4 Qe7; White has at best a tiny edge. 19 Rad1 Others: (a) After 19 Kxh2 Qd6 20 Bh7+ Kh8, White should transpose by 21 Rad1 to 21 Kxh2 in the main line because after, for example, 21 dxe5?! Nxe5 22 Bf5 Bxf5 23 Nxf5 Qe6! Black, a rook down, is winning! In Ulibin-E.Berg, Bajada de la Virgen 2005, play continued 24 Bg3 Rc8 25 Qxc8+ (25 Qb1? Nxf3+ 26 Kh1 Nd2) 25 ... Qxc8 26 Bxf4, and 26 ... Nxf3+! 27 Kg3 Qxf5 28 Kxf3 g5 was easiest. (b) 19 Rfe1 Bh3! 20 Rad1 (a nice mate follows 20 Kxh2? Nxd4! 21 Qxc7 Nxf3+ 22 Kh1 Bg2 mate; 20 Ne2! is best, and equal, after 20 ... Nxe2 21 Rxe2 e4 22 Re3 Qf4 23 Qb3 Be6 24 Rxe4 Nxd4 25 Rxf4 Nxb3) 20 ... Nxd4! 21 Rxd4 Qxc2 22 Bxc2 exd4 23 Kxh2 d3 24 Bb1 g5=/+ Biti-Gleizerov, Zadar 2005. 19 ... Qd6
Apparently 19 ... e4! is even better: 20 Bxe4?! (but White stands worse anyway) 20 ... dxe4 21 fxe4 Bh3 22 Qb3+ Kh8 23 Kxh2 g5-/+ Groetz-E.Berg, Tromsø 2009. Very convincing! However, since White had an opportunity to transpose to a 19 ... Qd6 line (see the note about 19 Kxh2 above), we’ll continue on: 20 Bh7+! Kh8
21 Kxh2 The most relevant move, given the previous note. Others are instructive: (a) 21 Nf5?! Qe6! gains a large advantage after 22 dxe5 Qf7! 23 Bg3 Bxg3 24 fxg3 Kxh7 25 Nd6+ (25 Ne7+ Ng6) 25 ... Qg6. (b) 21 dxe5 Nxe5 22 Be4! (22 Rfe1 Bh3) 22 ... Bh3!? 23 Kxh2 (a cute mate follows 23 Bxd5 Qxd5 24 Rxd5 Bg2+ 25 Kxh2 Nxf3 mate) 23 ... Qe6 24 Nh5! dxe4 25 Nxf4 Nxf3+ 26 Kh1 Qf5 led to a forced draw in Kotronias-M.Gurevich, Kusadasi 2006. (c) 21 Rfe1 has a ton of theory associated with it. In brief, White’s has to work for the equality which results from both 21 ... g6 and 21 ... g5!?. An example of the latter move is 22 dxe5 Qe7 23 Kxh2! Qxh7 24 Qxh7+ Kxh7 25 Ne2 Nxe2 26 Rxe2 gxh4 27 e6 Ne7 28 Rd4! (Williams); and 28 ... b6 or 28 ... Kg7 29 Rf4 Ng6 30 Rf7+ Kg8 31 Rf6 Kg7=. (d) 21 Bf5?! Bxf5 22 Qxf5 (Black wins following 22 Nxf5?? Qg6 23 Ne3 Qh5) 22 ... Nxd4! (or 22 ... Rf8 23 Qg4 Nxd4-/+) 23 Rxd4 (23 Qg4 Rf8 24 Kxh2 Ng6-/+) 23 ... exd4 24 Kxh2 Rf8 25 Qg4 (25 Qb1?? Qe6) (Williams) and here 25 ... Ng6 26 Qh5 (26 Qxd4 Rf4 27 Be7 Qe6) 26 ... Ne5 wins. 21 ... g6! 21 ... g5 is ultimately balanced following 22 dxe5! Nxe5 23 Bf5! Bxf5 24 Nxf5 Qg6 25 Rde1! gxh4 26 Rg1 Qc6 27 Qb3 Re8 28 Qe3! Ng4+ 29 Rxg4 Rxe3 30 Rxe3 Ng6 31 Rc3 Qe6 32 Rxg6! Qxg6 33 Rc8+=. After 21 ... g6, 22 Ne2! is apparently the only equalizer, and barely so; for example, 22 ... Kxh7 23 dxe5 Qxe5 24 Nxf4 Qxf4+ 25 Bg3 Qxf3 26 Qd3 Qh5+ (26 ... Qxd3 27 Rxd3 d4 is also equal, but it’s a winning try) 27 Kg1 Bh3 28 Qxd5 Qxd5 29 Rxd5 Bxf1 30 Rd7+ Kg8 31 Kxf1 and the ending will be drawn. 5.111215 Bh7+ White’s current favourite. 15 ... Kh8 16 Bg6 The interpolation of Bh7+ before playing Bg6 leads to critical new problems, as is typical of the tactically deep theoretical lines of major openings. 16 ... Rxf3! Still an effective sacrifice, and characteristic of the French as a whole.
17 Bxh5! This has become the main line of this variation and, for now, the most important one following 11 ... Qc7. In contrast to the previous section, Black’s king is one square further from the centre (h8 as opposed f8), which makes it harder to fend off rook incursions. But it’s not so simple, because in some lines, the king is better placed defensively, for example against captures on d5 (which are no longer with check) and the rare attack along the g-file. As in the section with 15 Bg6, here 17 gxf3 Bxh2+ 18 Kh1 (18 Kg2 Nf4+ 19 Nxf4 Qxf4 20 Bg3 Bxg3 21 fxg3 Qxd4) 18 ... Nf4 is equal, but trickier for White to handle: (a) 19 Kxh2 Nxg6+ 20 Bg3 Qf7 with good chances, much like the last section; ... Nge7-f5 was played there, and ... Bd7 and ... Rf8 is a normal plan. (b) 19 Nxf4?! is worse because of 19 ... Qxf4! 20 Bg3 Bxg3 21 fxg3 Qxg3. (c) 19 Ng3? (frequently played) 19 ... Qd6! threatens ... e5 and ... Nxd4. White is virtually lost because, unlike after 15 Bg6, he lacks the resource Bh7 (which was check in that case): 20 Kxh2 (20 Rad1 e5! 21 Bf5 transposes into the previous section, where Bg6/Bh7+/Bf5 occurred instead of Bh7+/Bg6/Bf5; suffice it to say that Black gets a very big advantage) 20 ... Nxd4 21 Qd1 Nxg6 (or 21 ... e5!) 22 Qxd4 e5! (Williams) 23 Nf5 Bxf5 24 Bg3 Qe6 with two pawns for the exchange and a winning position. (d) The best move is apparently 19 Rad1, when Black gets only an equal game by either 19 ... Bd7 20 Kxh2 Nxe2+ 21 Bg3 Nf4 22 Rfe1! Qd8! 23 Bxf4 Qh4+ 24 Kg2 Qxf4 25 Qd2 Qf6=, or the simpler 19 ... Nxe2 20 Qxe2 Qf4 21 Bg3 Bxg3 22 fxg3 Qxg3=; then Black’s passed pawns are compensated for by White’s active major pieces.
17 ... Bxh2+
Instead: (a) Most players who don’t sacrifice the exchange play 17 ... Rf8 as the solid approach (although 17 ... Rf5, with the idea 18 Bg6 Rf8!, may be even better – see ‘b’).
At first it seems as though White, with control of e5, will have an easy time establishing an edge; it turns out, however, that Black can stay reasonably active with familiar ideas: pressure on the d-pawn, a potential ... e5, and exchange or activation of his bad bishop via ... Be8. There follows 18 Bg3 Bd7. Now I’ll concentrate upon the best moves, and in a few lines the position can admittedly become dull. But they’re instructive, and it’s worth remembering things will likely become more double-edged in an over-the-board game: (a1) 19 Rad1 is common, covering the d-pawn.
Here the thematic 19 ... Rf6 is fine, with ideas of ... Be8 and/or ... Raf8. But these are more instructive: (a11) Swapping light-squared bishops by 19 ... Be8 is logical: 20 Bg4 (20 Bxe8 Raxe8 21 Rfe1 e5!? isn’t necessary, but results in pieces active enough to work with the isolated pawn after 22 dxe5 Bxe5 23 Qd2 Qb6 24 Nf4 Bxf4 25 Bxf4 Qc5) 20 ... Rf6 21 b4 a6 (21 ... Bg6 22 Qd2 a6) 22 a3 Bg6 23 Qc3, Kotronias-Ulibin, Stockholm 2006; and 23 ... Bf5!? was approximately equal, while 23 ... Re8 is the patient option, with few if any problems. (a12) 19 ... Rac8 is a useful move. White will have difficulties generating pressure; e.g. 20 Qd2 Bxg3!? (this gives White’s knight f4; we’ve seen the safer 20 ... Be8! and 20 ... Rf6) 21 hxg3 Rf6 22 Nf4 g5!, Burlacu-Berescu, Rumanian Team Ch. 2007; and 23 Ng6+ isn’t effective because of 23 ... Kg7. Stoica gives 23 Ne2 Be8 24 Bxe8 Rxe8 25 f4 Qg7 26 Rf3 Ne7, intending ... Nf5. A potential advantage to having Black’s queen on g7 is that ... g4, in combination with ... Nf5 and ... h5-h4, becomes possible.
(a2) 19 Rac1 is more active, although leaving d4 less protected:
(a21) 19 ... Rf6 is instructive, in part because of its use in other lines: 20 Bxd6 (20 Bg6? was played in a high-level game but it is the wrong square for the bishop: 20 ... Bxg3 21 hxg3, and now 21 ... Qb6! has multiple attacks: 22 Rfd1 Raf8 23 f4 e5!; while Black might meet 20 b4 with 20 ... Bxg3 21 hxg3 Qb6 22 Qc5 g6 23 Bg4 h5 24 Bf3 Qxb4 25 Qxb4 Nxb4 26 Rb1 a5 27 a3 Nc2 28 Rxb7 Rd8=) 20 ... Qxd6 21 f4, and 21 ... Be8 or 21 ... Rc8 is slightly more accurate than 21 ... Raf8 as in J.MohammadSatyapragyan, Bhubaneswar 2011. (a22) 19 ... Be8 frees Black’s game more quickly: 20 Bxe8 (20 Bg4 Rf6 21 a3 Bg6 22 Qb3 Raf8 with the idea 23 Bxd6 Qxd6 24 Qxb7 e5!) 20 ... Raxe8 21 Qb3 Bxg3 22 hxg3 Qf7= Fargere-Gleizerov, Marrakesh 2010. (b) 17 ... Rf5! has been wrongly neglected and may be the best way to generate positive prospects.
It can transpose to our main line, but Black has an good opportunity to deviate: (b1) The conventional 18 Bg6 Rf8 (18 ... Bxh2+ 19 Kh1 Bd6 transposes to our main line) 19 Bg3 also differs from 17 ... Rf8 in that White’s bishop on g6 can be a target: (b11) 19 ... Bd7 20 Rac1 Rf6! 21 Bh5 (making Black’s point; 21 Bd3!?) 21 ... Be8! 22 Bf3 (22 Bxe8 Rxe8 is easy for Black) 22 ... Bg6 23 Qb3 Bxg3 24 hxg3, Egin-Barsov, Tashkent 2011; and this was a good time to consolidate with 24 ... Qb6!. (b12) 19 ... Rf6 is also sound. Then 20 Rad1 (20 Bxd6 Qxd6 21 Bh5 e5; 20 Bh4 Rf8 leaves White nothing better than 21 Bg3) 20 ... Bd7 21 Rfe1 Rc8 is equal. (b2) From White’s point of view, 18 Bg4 is more likely to create chances. Black plays 18 ... Rf8 19 Bg3 and an interesting try is 19 ... e5!? (19 ... Qe7 20 Rae1 Bd7 is solid) 20 dxe5 Bxe5 21 Bf3 and now 21 ... Be6 or even 21 ... Rxf3!? 22 gxf3 Qf7!; e.g.
23 Rfe1 (otherwise White might stand worse) 23 ... Bh3 24 Qb3 Na5 25 Qd3 Bf5 26 Qb5 Nc6 27 Rad1 Rd8 with an obscure position in which, between the d-pawn and White’s kingside structure, Black has more compensation than might be immediately obvious. 18 Kh1 Rf5! This time 18 ... Rf8? loses material to 19 f4 Bxf4 20 Nxf4 Rxf4 21 Bg3±, because 21 ... Rxf1+? (21 ... e5 22 Bxf4 exf4 23 Rae1 Bd7 24 Qc5!) 22 Rxf1 Qd8 23 Rf7 is killing. 19 Bg6 Bd6 20 Bxf5 exf5
Here’s another theoretically hot position of the 11 ... Qc7 variation (although 17 ... Rf8 or 17 ... Rf5 are simpler solutions). Black is down an exchange for a pawn but has the bishop pair; other factors being equal, that is materially more-or-less even. He also threatens ... g5 and ... f4, burying White’s bishop, while ... Qf7-h5 can launch a very serious attack. On the other hand, White goes into this position willingly because of several factors. He is ahead in development, and has a good chance to develop his rooks actively and quickly along the e- and/or c-files. Additionally, Black’s pawn on f5 blocks in the bishop on c8. 21 Rae1 This move, whose worth is enhanced by prospects of entry to the 7th or 8th ranks, is considered best at the moment (along with 21 f4). Moves like 21 f3? and 21 Nc3?! are met by 21 ... Qf7 and ... Qh5, but several alternatives are instructive or even playable: (a) 21 Qd3?! is not very inspiring. Black can play the obvious 21 ... Qf7 22 Bg3 f4 23 Bh2 (23 Qf3 g5) 23 ... Bf5 24 Qd2, Ismagambetov-Kosyrev, St Petersburg 2006; and instead of 24 ... f3?, here 24 ... g5! puts White in a bind; e.g. 25 f3 Nb4! 26 Nc3 Nc2 27 Rac1 Ne3 28 Rfe1 Qh5 with an attack. (b) 21 Bg3?! f4 22 Bh2 f3! 23 Bxd6 fxg2+ 24 Kxg2 Qxd6 with a pawn and an attack against White’s exposed king. (c) 21 f4 has been played many times, mutually imprisoning Black’s light-squared bishop and White’s dark-squared bishop, with the benefit of preventing Black’s kingside pawn storm. But exposing White’s king and ceding e4 has its drawbacks: 21 ... Qf7! 22 Rf3 (or 22 Kg1 Bd7 23 Qd2 Re8 24 Nc3 Qh5 25 Bg3 Ne7!? 26 Rae1 Rc8 27 Re3 Qf7, Rasulov-Timman, Baku 2008; as Williams points out, White has no entry squares) 22 ... Bd7 23 Bf2 (23 Rd1?! is well met by 23 ... Nb4, since 24 Qb3?? loses to 24 ... Qh5, and 24 Qd2 Nxa2 favours Black) 23 ... Re8 (after something like 23 ... Na5 24 b3 Nc6, it’s difficult for either side to come up with a productive plan; but there are more entertaining ideas) 24 Nc3
This position has occurred in at least three games; for example, 24 ... Ne7 (an interesting if optional reorganization; instead, 24 ... Re4!? was tried in two games, but it seems to me that 24 ... g5! establishes equality right away; e.g. 25 Ne2! Nb4 26 Qd2 Bb5 27 Re1 and 27 ... g4 or simply 27 ... Bxe2 28 Rxe2 Rxe2 29 Qxe2 Bxf4 with equal chances; in this line 27 Nc3!? Bxf4! 28 Rxf4 gxf4 29 Nxb5 Qh5+ 30 Kg1 f3 intending ... Rg8 achieves at least a draw) 25 Qb3 Bc6 26 Re1 Ng8 (the point) 27 Rxe8 Qxe8 28 Qc2 (28 Nxd5 b6!?= intending ... Qa8 or ... Qh5+) 28 ... Qh5+ 29 Kg1 Nf6 30 Nd1 Ne4= Levushkina-A.Heinz, German League 2007. (d) Most of these positions, while theoretically equal, tend to be harder for White to play than Black; for example, 21 Qd2 f4 (21 ... g5 22 Bg3 f4 23 Bh2 Bf5=) 22 f3 Be6 23 Nc3 Qf7 24 Rfe1 Re8 25 Bf2 Ne7 26 Kg1 Nf5 27 Ne2 g5 28 Rac1 Qh5 (suddenly Black has a clear advantage) 29 Qd3?? Ng3 is simply winning, Medina-Nodirjanova, Khanty-Mansiysk 2010. 21 ... Qf7! The main reason to give this an ‘!’ is that it took Black a while to realize that other moves were worse. 22 Bg3 Bb4!
Although this temporarily cedes central control, it gains a tempo, retains the idea of ... f4, and has in mind attacking White’s only pawn weakness on d4. 22 ... f4? fails to 23 Nxf4 Bxf4 24 Bxf4 Nxd4 25 Qc7. 23 Rc1 23 Rd1 emphasizes safe protection of the d4-pawn. Most of the ideas are similar: 23 ... Bd7 24 a3! (24 Rd3?! f4! 25 Nxf4 Bf5 regains the exchange) 24 ... Ba5 (this improves upon 24 ... Be7?! 25 Rfe1 Bf6 26 Qb3!, Antal-Michielsen, Pardubice 2007, when 26 ... b6? 27 Nf4 Bxd4 28 Qxd5 was winning; 26 ... Re8 is better, but even 27 Kg1 shows how thankless Black’s task is) 25
b4 Bb6 26 Rfe1 (Williams thinks 26 Qa2 is White’s best, and calls 26 ... Re8 27 Nf4 Ne7, intending ... g5 and then ... f4, ‘dynamically equal’; 26 ... g5 can also be considered) 26 ... Re8 27 Kg1, and here simply 27 ... Kg8! preserves the ideas of ... Re4, ... g5, and ... f4; a verdict of dynamic equality seems fair here as well. 23 ... Bd7 24 a3 Ba5 According to plan. 24 ... Bf8 of Giri-Brandenburg, Leeuwarden 2009, is passive. 25 b4 Bb6
26 Qd2 Instead: (a) 26 b5?! Nxd4 27 Nxd4 f4! 28 Bh2 Bxd4 29 Qd2 Bb6 30 Qxf4 Qxf4 31 Bxf4 Bxb5 32 Rfd1 Bc4=/+ Parligras-Michiels, German League 2009. (b) 26 Rfe1 Re8 (26 ... Nxd4!?) is extremely similar to the position that we saw in the note about 23 Rd1, except that White has a rook on c1. Play might go 27 Qa2 (27 b5!? Na5!? 28 a4 Kg8 intending ... Nc4) 27 ... g5!? (27 ... Nxd4 28 Nxd4 Rxe1+ 29 Rxe1 Bxd4 is slightly worse but Black should hold) 28 Bd6! (28 b5 Nd8! 29 Nc3 Rxe1+ 30 Rxe1 f4 31 Bh2 Ba5 32 Qb3 Ne6 33 Rd1 g4! with counterplay) 28 ... Re4 29 Kg1 Qe6 30 Bc5 Bc7 31 f3 Re3 32 Kf2 f4 33 Nc3 g4!, again counterattacking. 26 ... Re8 This isn’t much different from lines above. 26 ... g5 27 b5 (27 f4!?) 27 ... Na5 28 Be5+ Kg8 29 Qd3 Nc4 is unclear. 27 f3 Here 27 Rfe1 Re4! wins the d-pawn in view of 28 f3?? Qh5+. Amazingly, in Xiong-Mabe, USCL 2011, the opponents, with an average rating of 2104, got all the way to the position after 27 Rfe1. The game continued 27 ... a6? 28 a4? Bd8? and ended in a draw. After 27 f3, a slow move such as 27 ... a6 is possible; but I like the aggressive 27 ... g5!; for example, 28 Kg1 f4 29 Bf2 Rg8 30 Rfe1 Ne7 and Black can attack via ... g4 or play for the outpost by ... Nf5-d6-c4. 5.112 13 Rc1 This is the most important strategic set-up today, trying to speed up White’s development and develop queenside play at the same time. If my analysis is correct, then the variations above in 5.111 ultimately work out well for Black; which suggests that 13 Rc1 and other options may be what you’ll be seeing over-the-board in a couple of years, if not before.
split
These 13 Rc1 lines wonderfully illustrate the positional underpinnings of Black’s play in this variation, justifying a final (with apologies):
5.1121 5.1122
13 13
... ...
Bd7 h6
13 ... Ng4 is a valid option, albeit double-edged: 14 Ng3 (14 h3 Rxf3! 15 hxg4 Rf8=) 14 ... Qb6!? attacks b2 and keeps an eye on d4, yet somewhat weakens Black’s kingside. Now Tzermiadanos’ recommendation is 15 Nh4 (adorned with ‘!’; otherwise there can follow 15 Qc2 h6 16 Be3 e5! 17 dxe5 Nxe3 18 fxe3 Nxe5 with at least equality) 15 ... Nf6 16 Ne4! 17 Bxe4 dxe4 18 Bh6!? (18 Qg4 e5 19 Qg3 Nxd4 apparently favours Black), and here Tzermiadanos gives 18 ... Rf7 (18 ... gxh6!? 19 Qg4+ Kf7 20 Qxe4!; I think Black is okay, but it’s probably not worth more than a draw) 19 Bxg7 Rxg7 20 Nxg7 Kxg7 21 Qg4+ Kh8 22 Qh5 and calls it ‘unclear’. Black stands at least equal following 22 ... Bd7; e.g. 23 Ng6+ Kg8, and either 24 Rfd1 Be8 or 24 Nf4 Qd8!. 5.1121 13 ... Bd7 Black develops before undertaking action. This is his most solid continuation, completely equal in my opinion. While apparently peaceful, the position contains a latent dynamism which expresses itself in several lines.
14 Re1 Tzermiadanos’ recommendation. Others are also complicated: (a) 14 Bh4 Nh5 (or simply 14 ... Qb6 15 Qd2 Rae8 with ideas of ... e5 or ... Nh5) 15 Nc3 (15 Bb1 can be met by 15 ... g6 or the fancier 15 ... Rxf3!? 16 gxf3 Bxh2+ 17 Kg2 Bf4! 18 Rc3! g6 19 Ng3 Ng7, which is unclear; Tzermiadanos gives 15 Qc2 h6 16 Bh7+ Kh8 17 Bg6 Be8!=) 15 ... a6 16 Ne5!? (16 Qc2? Rxf3!) 16 ... Nf4 17 Nxc6 (17 Nxd7 Qxd7 18 Bb1 g5! 19 Bxg5 Qg7 20 Bxf4 Bxf4 with the idea 21 Rc2 Nxd4) 17 ... Bxc6 with a comfortable position; e.g. 18 Bb1 Rae8! 19 Qc2 g6=/+. (b) 14 Ng3 has attacking ideas, although it takes away control of d4 and f4, which shows up in lines such as 14 ... Qb6!? (14 ... h6 15 Bd2 Qb6 16 Bc3 Bf4 17 Rc2 Rac8=) 15 Bxf6 Rxf6 (15 ... gxf6!? 16 Qd2 Rf7) 16 Bxh7+ Kxh7 17 Ng5+ Kh8! 18 Qh5+ Rh6 19 Nf7+ Kh7 20 Nxh6 Be8 21 Nf7+ Kg8 22 Nh6+ Kh7=. (c) 14 Qc2 Be8! 15 Be3 (15 Bh4 Qb6 16 Bg3 Bxg3 17 hxg3 Nb4; 15 Rfe1 h6 16 Bh4 Bh5; 15 Nc3 Bh5 16 Nb5 Qb8 17 Nxd6 Qxd6) 15 ... h6 16 Ng3 Rc8=. (d) 14 a3 Be8 15 Ng3 (versus ... Bh5) 15 ... h6 16 Bd2 Qb6 17 Bc3 Nh5= Oleksienko-Karthikeyan, Chennai 2012. 14 ... Rac8 This move isn’t mentioned by Tzermiadanos, but it almost always comes in handy. 15 Bb1 15 Qb3 hinders ... Qb6. Black might reorganize by 15 ... Qb8 16 Ng3 Ng4, ready to answer h3 with ... Nxf2; then 17 Bb1 Bc7!? 18 Qd3 g6 intends ... Bb6, hitting d4. Here 16 ... a6 is another option, considering ... Qa7 or ... b5. 15 ... Qb6 16 Bf4! 16 b3?! Ba3 forces White’s pieces to poor squares, unless he chooses to cede the bishop pair by 17 Bxf6 Rxf6 18 Qd3 g6 19 Rcd1 Rcf8=/+. 16 ... Be7 Black intends to play ... Ne4. 17 Qd2! Ne4! 18 Bxe4 dxe4
19 Ng5 Or 19 Ne5 Be8 20 Be3 Rd8 21 Qc2 Nxe5 22 dxe5 Qa5 23 Qc7 Bb4! 24 Qxa5 Bxa5=, an entertaining line going 25 Nc3 Bc6!? 26 Bxa7 Rd2 27 Bc5 Rf5 28 b4?! Bxb4! 29 Bxb4 Rfxf2=/+. 19 ... Bxg5 20 Bxg5 Nb4! 21 Be7! Nd3 22 Bxf8 Rxf8 23 Rb1 Here Black could play 23 ... Nxe1 24 Rxe1 Bc6=, but this leaves him few positive chances. The more enterprising 23 ... e5!? 24 Rf1 exd4 produces a dynamic balance. Black’s advanced central pawns and 6th-rank knight give him compensation; e.g. 25 Ng3 Qg6 26 Qc2 Be6 27 Qa4 Nf4 28 Rbe1 Bd5! 29 Qxd4 h5! 30 Rxe4 (30 h4 e3!=/+) 30 ... Bxe4 31 Qxe4 Qg5 32 Re1 h4 33 Nh1 Rf6! (threatening ... Re6) 34 Qxb7 Re6 35 Rxe6 Nh3+ 36 Kf1 Qc1+ 37 Ke2 Nf4+ 38 Kf3 Nxe6=. 5.112213
...
h6!?
This popular move resolves the position of White’s bishop at the cost of a slight weakening of the kingside. Unlike other lines following 13 Rc1, Black retains the familiar possibility of throwing his kingside pawns forward, putting a queen on g7, and/or sacrificing a rook on f3. There are many games here, so I’ll abbreviate dramatically: 14 Bh4 This is supposed to be a particularly appropriate moment for 14 Bd2, preserving the bishop and preparing b4. Black can still try to attack, but it’s more likely that he will play for sound development: 14 ... Bd7 15 b4 a6 16 a4!? (consistent; 16 Re1 b5!? is ambitious, or here 16 ... Rac8 17 Qb3 Be8 18 Ng3 Bf7 19 Ne5 Qb6! with counterplay against d4) 16 ... Qb6! 17 b5 axb5 18 axb5
Now 18 ... Ne7 looks best (although 18 ... Nb4 19 Bxb4 Bxb4 20 Ne5 Ba3 21 Rb1 Be8 22 Nf4 Nd7! come very close to equality); e.g. 19 Ne5 Be8! 20 Nf4 (20 Qc2 Nd7! 21 Nxd7 Bxd7 keeps White out of the e5 outpost, except in the case of 21 Be3 Nf5 22 Bxf5 Rxf5=) 20 ... Ne4 21 Bxe4 (21 Be3 Nxf2! 22 Nxe6 Bxe5 23 Bxf2 Qxe6 24 dxe5 Bg6!=; 21 Nxe6?! Bxe5 22 Nxf8 Bf6 and the knight is trapped) 21 ... Qxd4! with complications and equal chances. 14 ... Nh5 The standard move, though I’m not sure that 14 ... Qb6 isn’t satisfactory, especially with the bishop away from the queenside on h4. Tzermiadanos gives 15 b3 Bd7 16 Bg3, but then 16 ... Ba3! is satisfactory for Black: 17 Rc2 (17 Rb1 Nb4 18 Ne5 Rac8) 17 ... Rad8 (17 ... Nb4?? 18 Bc7) with the idea 18 Ne5?! Nb4!.
15 Bg3 Instead: (a) 15 Qc2 simply encourages Black to play 15 ... g5! 16 Bg3 Nxg3 and 17 hxg3 Qg7 or 17 Nxg3 Rxf3! 18 gxf3 Qf7 with ample compensation. (Compare the order 13 Bh4 Nh5 14 Qc2 h6 15 Rac1.) (b) 15 Bg6 Rxf3! 16 gxf3 Bxh2+ is a sound exchange sacrifice: (b1) 17 Kh1 Nf4 18 Be8!? Nxe2 19 Qxe2 Qf4 (19 ... Bf4 20 Bg3 Qe7 21 Bxf4 Qxe8 tries for more, with some risk) 20 Bg3 Bxg3 21 fxg3 Nxd4! 22 Qf2 Qf5!=. (b2) 17 Kg2 Nf4+ 18 Nxf4 Bxf4 19 Bg3 (19 Rh1 Qd6!?= with ideas of ... e5, ... Rf8 and ... Qb4) 19 ... Bd7 20 Rh1!, as recommended by Tzermiadanos. But 20 ... Qd6! gives sufficient play; e.g. 21 Bh5! (21 Bb1 Qf8! intends ... Qf6: 22 Rc3 Qf6 23 Qd3 Nxd4 24 Rd1 Nf5 25 Bxf4 Nh4+ ½-½ Slinger-Rodríguez Amezqueta, ICCF server 2008; 21 Qd3 Bxg3! 22 fxg3 e5=) 21 ... Qf8, and as an illustration only: 22 Rc3 Rd8!? 23 Rh4 Bg5 24 Rg4 Be8 25 Bxe8 Rxe8 26 Be5 Qf5 27 f4 Nxe5 28 fxe5 Rf8 29 Qe2 h5 30 Rgg3 Bf4 with equalizing counterplay in view of 31 Rh3 Qg6+ 32 Kf1 Qb1+ 33 Qe1 Qxb2=. (c) 15 Bb1 g5 16 Bg3 Nxg3 has been seen in many games.
Play can go in almost any direction, but in general Black’s attack is sufficient to make up for his weaknesses. For example: (c1) 17 fxg3 Qg7 18 g4 Bd7 19 Qd3 Rac8 20 Ng3 Be8=. (c2) 17 hxg3 Qg7 (or 17 ... Qf7) 18 g4 (18 Qd3 Bd7 19 a3 Na5!? 20 b3 a6 is a unique approach) 18 ... Bd7 19 Ng3!? Rxf3!? (19 ... Be8) 20 Qxf3 (20 gxf3 Nxd4 followed by ... Rf8) 20 ... Nxd4 21 Qd1 Rf8. Black has sufficient compensation for the exchange, although probably no more.
(c3) 17 Nxg3Qf7 18 Qd3 Bd7 19 a3 (preventing ... Nb4; a unique Black attacking scheme came about in Bruned-Bhat, Andorra la Vella 2006: 19 Rce1 a6 20 a3 Kh8 21 Re2 Rg8!? 22 Rfe1 Rg7 23 b4 Rf8 with excellent attacking prospects) 19 ... Rae8 and now: (c31) 20 Rfe1 Re7 21 Kh1 Be8 22 Re2, C.Balogh-Moor, Zürich 2005, and here Black could have played 22 ... g4! 23 Ne5 Qf4 24 Rce1 Nxe5 25 dxe5 Bc5-/+. (c32) 20 Rcd1 Re7 21 Rd2 Be8 22 Re1 Qg7 23 Rde2?! (23 h3 h5) 23 ... g4 24 Ne5 Nxe5 25 dxe5 Bc5 26 Rc1 Bb6 27 Rc8, S.Brady-Ulibin, Isle of Man 2002; and now the easiest course was 27 ... Ref7, winning material. 15 ... Nxg3 16 hxg3 16 Nxg3 (16 fxg3 Qb6) 16 ... Qf7!? (16 ... a6!? puts a stop to Bb5) 17 Qe2 (17 Bb5 Bd7 18 Re1 wins e5, but 18 ... a6 19 Bxc6 Bxc6 is fine; e.g. 20 Rc2 Rac8 21 Ne5 Bxe5 22 Rxe5 Bd7) 17 ... Bd7 18 Bb1 g5 19 Rcd1 Bf4 20 Ne5? Nxe5 21 dxe5 Be8! prevents Nh5 and prepares ... h5-h4, Kotronias-E.Berg, Heraklio 2007. 16 ... a6 16 ... Qf7! is flexible, making room for ... Bd7 and preparing to go to h5 or f6 as appropriate. If White isn’t careful, ... e5 can be played because of the extra support given to d5. 17 Bb1 g5 18 Re1!? 18 g4, with the idea of holding up Black’s kingside and at some point playing Ng3, is a good alternative, when 18 ... Qg7 keeps an eye on d4 and prepares ... Bd7. After 18 Re1, Schönberger-Pirrot, Bad Wiessee 2010, continued 18 ... Qg7 (here 18 ... Qf7! indirectly hits f2; then 19 g4 can be answered by 19 ... h5!) 19 Nc3 Bd7 20 Na4! g4 21 Nh2 Qxd4 22 Qxg4+ Qxg4 23 Nxg4 Bb4 24 Nb6!? (24 Red1 Kg7) 24 ... Bxe1 25 Rxe1 Rad8 26 Nxd7 Rxd7 27 Rxe6 Re7=. 5.1212 White shifts to the queenside and threatens Nb5. 12 ... a6
Nc3
13 Bg5 Instead: (a) 13 h3 is sometimes played, to prevent ... Ng4, and this resembles 12 h3 (note ‘a’ at the beginning of 5.1); e.g. 13 ... 00 14 Be3 Bd7 15 Rc1 and here 15 ... Rae8 and 15 ... Rac8 are perfectly playable, but a thematic solution is 15 ... Be8! 16 Ng5 (after 16 Re1, Black might activate his bishop by 16 ... Bh5, when 17 g4 Nxg4! 18 hxg4 Bxg4 19 Be2 Qe7 begins a dangerous attack) 16 ... Qe7 17 f4 h6 18 Nf3 Nh5 19 Ne5 Ng3 20 Rf2 Nf5 21 Bxf5 Rxf5 22 Ne2 Rf8 23 Rf1 Rd8! 24 Qb3 Bxe5 25 fxe5 (25 dxe5 d4=/+) 25 ... Bg6=. (b) 13 Be3 (rarely seen) 13 ... 0-0 (or 13 ... Bd7, intending 14 Rc1 Ng4) 14 Rc1 Bd7 15 Na4, Dvoirys-Levin, St Petersburg 2011; and here 15 ... Be8 could be played, anticipating 16 Ne5 (16 Nc5?! Bxc5 17 Rxc5 Bh5) 16 ... Qe7 with a solid position. 13 ... 0-0 This can arise via 12 Bg5 0-0 13 Nc3 a6 as well. 14 Bh4 14 Rc1 (14 Re1? Ng4 15 g3 Qf7 16 Be2 Qh5 with an attack) 14 ... Ng4!? (14 ... Bd7) 15 h3 Nh2 16 Nxh2 Bxh2+ 17 Kh1 Bf4 18 Bxf4 Qxf4 19 Ne2 Qh4 20 f4 Bd7= Kubsch-G.Gutierrez, corr 1987. 14 ... Nh5
This position closely resembles the main line with 12 Bg5 0-0 13 Bh4 Nh5 (5.111), with the moves Nc3 and ... a6 thrown in, and in fact 14 Nc3 a6 would transpose exactly. But it’s good to think of these positions with Nc3 separately in strategic terms because, instead of focusing on the centre and kingside (with 14 Qc2 or 14 Bg3, for example), the Nc3 lines tend to be based upon Rc1/Na4 with queenside pressure. Black, for his part, has the idea of playing ... g6 and ... Qg7, drifting towards the kingside and putting pressure on d4. 15 Re1 A traditional move, simply contesting e5 and aiming at the backwards pawn on e6. This can hardly be bad, but does allow Black an optimal reorganization of his pieces. The alternatives are equally important: (a) 15 Bg3 Nxg3 16 hxg3 g6 17 Rc1 Qg7 18 Bb1 g5! has the idea ... g4 and ... Nxd4, which is another benefit of ... Qg7. This position has arisen repeatedly over the years with positive results for Black; for example, 19 Ne2 (19 Re1 transposes into a 15 Re1 line) 19 ... g4 (19 ... Bd7 is often played) 20 Ne1 Bd7 21 Nc2 Rf7 22 Qd2 Raf8 23 Ne3 Kh8 24 Rcd1 Bb8 25 Nc1 Ba7 26 Nb3, Buervenich-Hedman, Stockholm 2001, when it was simplest to open lines with 26 ... e5! 27 dxe5 Qxe5; e.g. 28 Qxd5 Rxf2! 29 Qxe5+ Nxe5 30 Rxf2 Bxe3 31 Rdd2 Nc4 32 Rde2 Rxf2 33 Rxf2 Nxb2 with an extra pawn. (b) After 15 Qc2, Black doesn’t even have to defend his kingside and can play 15 ... Rxf3! 16 Bxh7+ Kh8 17 gxf3 Nxd4 with a strong attack and advantage. (c) White’s 15 Rc1 g6 16 Na4 is supposed to discourage Black’s ... g6/ ... Qg7 plan (rather than 16 Bb1 Qg7 17 Na4 Rxf3! 18 gxf3, Guseinov-Se.Ivanov, USSR 1985, when Black should play 18 ... Nf4! 19 Bg3 Bd7 20 Nb6 Rd8 21 Nxd7 Rxd7=/+ intending ... Rf7 and ... Nxd4; or 16 Be2 Bf4! 17 Rb1 h6 18 Na4 g5 19 Bg3 Qg7; instead, 16 Bg5!? is more interesting: 16 ... Qg7 17 Be3 Nf4 18 Bb1 Bd7 19 Na4 g5 20 Nb6 Rad8= intending ... Be8)
16 ... b5! (the most adventurous move, and better than the often-played 16 ... Qg7?! 17 Nb6 Rb8 18 Bxa6!+/=) 17 Nc5 (17 Nc3 Bb7!? renews ideas of ... Qg7 and ... e5; White can reroute to protect his centre by 18 Bg5 Qg7 19 Be3, but 19 ... Nf4 leaves Black’s pieces better placed) 17 ... Rxf3! (the standard exchange sacrifice) 18 Qxf3 (18 gxf3? Bxh2+ 19 Kg2 Bf4) 18 ... Bxh2+ 19 Kh1 Bf4 with plenty of compensation, Mamedov-Ulibin, Moscow 2004. There might follow 20 Nb3 Nxd4 (or the unclear 20 ... Bxc1 21 Rxc1 Bd7 22 Bg5) 21 Rxc7 Nxf3 22 Rxc8+ Rxc8 23 gxf3 Kf7, intending ... g5 or ... Be5. 15 ... g6 Protecting the king and intending ... Qg7, Black’s standard resource. 16 Rc1 Not 16 Bc2? Rxf3!. 16 ... Qg7 Attacking d4. The options are 16 ... Bf4 first, or simply 16 ... Bd7. 17 Bb1 17 Bf1 doesn’t contribute much: 17 ... Bf4 (17 ... h6!?) 18 Rc2 (18 Rb1 Bd7=/+) 18 ... g5 19 Bg3 Nxg3 20 hxg3 Bc7=/+ with the ideas ... Bb6 and ... g4. 17 Bb1 is Kudrin-Korchnoi, Beersheba 1984 (and several later games) when, instead of the normal 17 ... Bd7 (18 Bg5!?), 17 ... Bf4! is a simpler solution: 18 Rc2 (18 Bg5?! Nxd4!) 18 ... g5 19 Bg3 Nxg3 20 hxg3 Bb8! intending ... g4 and ... Ba7. Then best is 21 Na4 g4 22 Ne5! Nxe5 23 dxe5 Ba7 24 Nc5 Rf7=. 5.13 12 White sets up to exchange bishops on f4. In return, he weakens his kingside light squares. 12 ... 0-0 13 Bf4 Ng4 This active and ambitious move covers e5. 13 ... Bd7 is a common alternative, solid but not much fun.
g3
14 Bxd6 Probably best, though these are logical as well: (a) 14 Qd2 e5! 15 dxe5 Ngxe5 16 Nxe5 Nxe5 17 Bxe5 Bxe5 18 Rac1 Qb6 19 Nc3 Be6 20 Qe2 (20 Bc2 Rad8 21 Bb3 d4!) 20 ... Bxc3 (20 ... Bf6 21 Na4 Qd6 22 Nc5 Rae8=) 21 bxc3 (21 Rxc3 d4=) 21 ... Rae8=. (b) A typical exchange sacrifice follows 14 Rc1 Bxf4 15 Nxf4 Rxf4! 16 gxf4 Qxf4
and: (b1) 17 Be2 Nf6! 18 Kh1 Ne4!=/+ Azarov-Berescu, Djakovo 2005; Black prepares ... Bd7 and ... Rf8 with an attack, and sometimes the immediate ... e5 will work. (b2) 17 Bb5 Bd7 18 Bxc6 Bxc6 19 Rc3 Rf8 20 Qc1 Qd6! 21 Qg5 Rxf3 22 Qxg4 Rf4 23 Qg3 h5! 24 f3 (Williams) 24 ... h4! 25 Qf2 e5 26 dxe5 Qxe5=/+ intending ... d4. (b3) 17 h3 e5! 18 hxg4 Bxg4 19 Be2 Bh3 20 Nxe5 (Williams) 20 ... Nxd4! 21 Re1 Re8 22 Bg4 Rxe5 23 Rc8+ Kf7 24 Bh5+ g6 25 Rxe5 Qxe5 26 Bg4 Bxg4 27 Qxg4 Ne6 28 Qf3+ Kg7=. 14 ... Qxd6 15 Nc3! This threatens Bxh7+. 15 Nf4?! allows the standard sacrifice 15 ... Rxf4! 16 gxf4 Qxf4=/+.
15 ... e5!? It’s hard to resist this active solution, which gives excellent tactical chances and only a small disadvantage versus accurate defence. Nevertheless, a solid alternative like 15 ... Nh6 may be objectively better; and I like 15 ... Qe7!, intending ... Qf6, when the queen will be well placed attacking d4, f3 and, indirectly, f2. For example, 16 Rc1 (16 h3 Nh6 17 Kg2 Bd7 18 Re1 Qf6 19 Bb5, Wharrier-Corfield, ICCF email 2009; and here 19 ... Rae8 is a straightforward move; e.g. 20 Bxc6 bxc6 21 Qe2 Nf7 22 Rad1 e5 23 dxe5 Nxe5 24 Nxe5 Rxe5 25 Qxe5 Qxf2+ 26 Kh1 Qf3+ with a perpetual) 16 ... Qf6 17 Ne1 Nh6 18 Ne2 a6 (or 18 ... Bd7 19 Bb5 Rad8) 19 f4 Ng4 20 Nc2 (20 Qd2 Nxd4 21 Bxh7+ Kxh7 22 Nxd4 e5=) 20 ... Qh6 21 h4 Qf6 22 a3 Bd7 23 Nc3 Nh6=. 16 dxe5 Qh6!? The alternative is 16 ... Ncxe5; e.g. 17 Be4!? d4! 18 Nxd4 Nxh2! 19 Kxh2 Qh6+ 20 Kg1 Ng4 21 Nf3 Rxf3 22 Qd8+ Rf8 23 Bd5+ Be6 24 Qh4=. But 17 Nxe5 Qxe5 18 Be2 gives White an edge. 17 h4 Ngxe5 17 ... Ncxe5 18 Be2 Be6 19 Ng5 Rad8, Idani-D.Petrosian, Yerevan 2012; and 20 Qb3! Kh8 21 Rad1 Bg8 offers White some advantage, although this is playable (Black later won). 18 Ng5! The only try for advantage, attacking both h7 and d5; whereas 18 Nxe5 Nxe5 cedes plenty of play against White’s weakened kingside squares: 19 Be2 Be6 20 Nb5 Rad8 21 Nd4 Bh3 22 Re1, Meier-Navara, Mainz (rapid) 2009; and 22 ... Qb6 23 Qd2 Bf5!=/+ was best. 18 ... d4! This is Vachier Lagrave-E.Berg, Liverpool 2008; and here the greedy 19 Bxh7+! apparently favours White: 19 ... Kh8 20 f4! (20 Nce4 Bg4 21 Qa4 Be2 22 Rfe1 Qh5! with sufficient counterplay) 20 ... Bg4 21 Qc2! with the idea 21 ... dxc3 22 fxe5 Nd4 23 Qe4 cxb2 24 Rab1 Nf3+ 25 Nxf3 Bxf3 26 Rxf3 Rxf3 27 Qxf3 Kxh7, although White’s advantage is limited by his weak pawns. Still, Black’s 15th move alternatives look theoretically preferable. 5.14 12 Bd2 A calm move, meant to bring a rook to c1 quickly and support a queenside advance with b4. This is also a reaction to the troubles this bishop often runs into after 12 Bg5. 12 ... 0-0 13 Rc1
13 ... Bd7 Against slow moves such as 12 Bd2, I counsel patience. Another set-up is 13 ... a6 14 Ng3 (14 Bc3 Bd7 15 h3 Nh5!? with the idea 16 Qc2? Rxf3! 17 gxf3 Qd8!) 14 ... Bd7 15 Re1 Qb6 16 Bg5?! (16 Bc3 Bf4 17 Rb1 Rac8=) 16 ... Ng4!? (16 ... h6 is fine) 17 Bb1 g6 (17 ... Qxb2!) 18 Rc2 Rf7 19 Nf1 Raf8 20 h3? Rxf3! 21 hxg4 (21 gxf3 Nxd4–+), Castaneda-Bhat, Budapest 2001; and the simplest move was 21 ... Nxd4!-/+. 14 Ng3 14 h3 is also logical; e.g. 14 ... Qb6 (14 ... Rac8; 14 ... Be8) 15 Bc3 Rac8 and now: (a) 16 Qd2 Nb4 17 Bb1 Bb5 18 Ng5! h6 19 Nxe6 Rce8! 20 a4! Ba6 21 a5! Qxa5 22 N2f4 Bxf1 23 Nxf8 Bxg2 24 Kxg2 Rxf8 with sufficient counterplay. (b) 16 Ne5 Nxe5! 17 dxe5 Rxc3 is an interesting exchange sacrifice: 18 bxc3 (18 Rxc3 Bxe5 19 Rb3 Qc7 20 Qd2 Bh2+ 21 Kh1 Bd6 with compensation based on the bishop pair and centre) 18 ... Bxe5 19 c4 Bc7!? (19 ... Bb8 20 cxd5 exd5 21 Rb1 Qc7) 20 cxd5 Qd6 21 Ng3 exd5.
14 ... Qb6 Black has good alternatives; for example, 14 ... a6 transposes to the note on 13 ... a6 above, and 14 ... Rae8 prepares ... e5, which can’t be stopped without other concessions; e.g. 15 Re1 Qb6 16 Bc3 and:
(a) 16 ... e5!? 17 dxe5 Ng4 18 Bxh7+ (18 exd6!? Nxf2 19 Bxh7+ Kh8= 20 Bd4?! Rxe1+ 21 Qxe1 Nh3+ 22 gxh3 Nxd4 23 Nxd4 Qxd4+ 24 Kg2 Kxh7=/+ Almasi-E.Berg, Mülheim 2011) 18 ... Kh8 19 Rf1 Bc5 20 Kh1 (20 Rc2 Bxf2+ 21 Rcxf2 Nxf2 22 Rxf2 Kxh7 23 Qxd5 Be6=) 20 ... Nxf2+ 21 Rxf2 Bxf2 22 e6 Rxe6 23 Qxd5 Rxf3 24 gxf3 Qe3 25 Bd2 Qd4 26 Qxd4 Nxd4 and equality becomes evident in a few more moves. (b) 16 ... Bf4 is solid: 17 Rc2 g6 18 Rce2 (18 Ne5 Nb4!; 18 a3 Ng4! intending ... e5, and using the standard trick 19 h3? Nxf2 19 Kxf2 e5; better 19 Rce2, arguably with an edge, but it’s hard to find a plan) 18 ... Rc8= 19 a3, and one legitimate idea is the strange-looking manoeuvre 19 ... Ne8 20 Qc2 Rf7 21 Nf1 Nd6 22 Ne3 Nb5. 15 Bc3 Bf4 This is standard when White’s bishop is on c3; for one thing, Ne5 can often be answered by ... Nxe5 and no fork results. 16 Ra1 Avoiding 16 Rc2 Nb4= or 16 Rb1 Nb4=. 16 ... g6 After 16 ... Rac8, it would be interesting to see what either side’s plan is. 17 a3 Rae8 18 Ne5 To prevent ... e5. 18 Re1 e5 19 dxe5 Ng4!=/+ hits f2 and e5. 18 ... Nxd4! Or Black can wait with 18 ... Rd8. After 18 ... Nxd4, Hou Yifan-Muminova, Guangzhou 2010, went 19 Nxd7 Nxd7 20 Bxg6! hxg6 21 Bxd4 Qd6 (Black has a mobile centre with active pieces and a passed d-pawn) 22 Bc3 b6?! (22 ... Rf7 23 Qd3 e5! 24 Rad1 Nb6) 23 Qd3. Now 23 ... Kh7 is satisfactory, whereas the game’s 23 ... e5? 24 Rad1 was good for White. 5.211 ... 0-0 Our second solution is a relatively rare line which is characterized by safety and a somewhat slower game. It involves fewer tactics and tests the positional skills of both players.
12 Bf4 This is played the vast majority of the time and is really the point of the Ne2 system. Otherwise Black can develop freely with ... Qb6 or ... Qc7 and moves like ... Ng4 and ... Nh5 take on more strength. Finally, the simple ... e5 will tend to equalize. These two moves merit a brief look: (a) 12 Nc3 Bd7 (this is safe and sound, contemplating ... Be8-h5 or simply ... Qb6; another option is the direct 12 ... e5!? 13 dxe5 Nxe5 14 Nxe5 Bxe5). Now the innocent-looking 13 Re1?! allows 13 ... Ng4!=/+, when the bothersome knight is awkward to dislodge.
For example, 14 Nb5 (14 h3? Nxf2! 15 Kxf2 Qh4+ 16 Kf1 Rxf3+–+; 14 Be2 Qc7 15 g3 a6=/+) 14 ... Nxf2! 15 Bxh7+! (15 Kxf2 Qh4+ 16 Kg1 Rxf3 17 Bxh7+! Qxh7 18 Nxd6 Rd3 19 Qg4 Rxd4-/+) 15 ... Kxh7 16 Qc2+ Kg8 17 Qxf2 (17 Nxd6 Nh3+ 18 Kh1 Qb8 19 Nb5 Nb4 20 Qe2 Nd3!) 17 ... Bb8 18 Be3 a6 19 Nc3 e5=/+, or 19 ... Be8=/+. (b) 12 Bg5 Bd7 (12 ... Qb6= is a well-established line from the old 7 ... Qb6 variation) 13 Re1 (13 Nc3?! h6 14 Bh4 Qb6!; 13 a3 Be8 14 Ng3 h6 15 Be3 Nh5!? 16 Re1 Nxg3 17 hxg3 Bh5 18 Rc1 Qf6= Hracek-Kindermann, Nussloch 1996; 13 Rc1 Qb6 14 Qd2 Ne4 15 Bxe4 dxe4 16 Ne5 Bb4! 17 Nc3 Nxe5 18 dxe5 Bc6=) 13 ... Qb6 (or 13 ... Rc8=, or 13 ... h6 14 Bh4 Qb6=) 14 Nc3 Rae8!? 15 a3 a6 16 Bh4 Kh8 17 Na4!? Qa7 18 Bc2?, Rogic-Volkov, ICC 2002; and 18 ... e5!=/+ would be awkward for White to meet: Black’s bishop pair offers some chances after 19 Bxf6 Rxf6 20 Nc3! Nxd4 21 Nxd4 Qxd4 22 Qxd4 exd4 23 Nxd5 Rff8.
After 12 Bf4, the main line is 12 ... Bxf4 13 Nxf4 Ne4, which has been tested in hundreds of grandmaster games. As of this writing, that sequence appears perfectly good for Black, with a verdict of equality; objectively it may even be the ‘correct’ move. But not only does 12 ... Bxf4 require mastery of a huge body of theory, it also includes several lines that are well and truly drawn, which may not be to your taste. 12 ... Nh5 I became interested in this move while reading Moskalenko’s bookThe Flexible French; he has been a long-term practitioner of the move, and has played it off-and-on with success for about 20 years, including a game played within weeks of this writing. Nor is he alone among his grandmaster colleagues. The idea of 12 ... Nh5 is obvious: Black forces the exchange on d6, thus increasing control over e5 and focusing three pieces on the f4-square, which tends to be an important one in this structure. In addition, the move ... Rxf3 becomes a motif, one that White has to monitor at the potential cost of committing his pieces to defence. Naturally 12 ... Nh5 has drawbacks. For one thing, Black can often play an effective ... Ne4 or ... Ng4 in this structure, which moves are no longer possible, and his h7 and d7 are left undefended. Temporarily, Black’s freeing move ... e5 can become less effective (or impossible) because his pawn on d5 will be attacked and captured; obviously a knight on f6 helps in that respect. This means that White can play for some dull positions in which Black’s light-squared bishop remains defensive. The good news for Black is that, even when White does get an edge, it’s usually of minor proportions and quite defensible. 13 Bxd6 Qxd6 Here the most frequently-played moves are:
5.21 5.22
14 14
Qd2 Ng3
Others: (a) 14 Re1 Bd7 15 Rc1
15 ... a6!? (15 ... Rf6 16 Ng5!? h6 17 Nh7 Rf7 18 Bg6 Re7 19 Bxh5 Kxh7 20 f4 Be8=; it’s remarkable that so many players go in for the type of position after 15 ... Rxf3!? 16 gxf3 Rf8, as in Savchenko-Ponomariov, Moscow 2010, and I think that White should try 17 Bb5! Nf4 18 Bxc6 bxc6 19 Nxf4 Qxf4 20 Kh1! Rf5 21 Rg1+/=; even if this isn’t clear, it’s interesting that a super-GM would play the exchange sacrifice in this form – you can see why Moskalenko prefaces ... Rxf3 with ... a6) 16 Bb1 (16 Ng3 Nf4 17 Bb1 Be8, intending ... Bg6 in the near future) 16 ... Rxf3! 17 gxf3 Nf4 18 Qd2 (18 Nxf4 Qxf4 19 Qd3 g6 20 Rcd1 Rf8) 18 ... Rf8 19 Qe3 Nxe2+ (19 ... Nh3+ 20 Kh1 e5) 20 Rxe2 Rf4=, threatening both the d-pawn and ... Rh4. (b) 14 Rc1 a6 (14 ... Bd7) 15 Re1 Rxf3!? 16 gxf3 Bd7 17 Ng3 Nf4 gave strong compensation in Maratonac-CapNemo, playchess.com 2007. (c) 14 Bb5 Bd7 (threatening ... Rxf3 in some lines; 14 ... Nf6!? 15 Ng3 Bd7 is roughly equal; perhaps White should bring his bishop back to d3 at some point) 15 Rc1 (15 Ne5 Nf6! 16 f4 Nxe5! 17 fxe5 Qb6=) 15 ... Rxf3!? (an enterprising move, albeit a speculative one; the calm 15 ... Rac8 and 15 ... Nf6 are sounder) 16 gxf3 Rf8 17 Bxc6 Bxc6+/= with practical compensation. If White’s knight moves, ... Nf4 tends to be extremely strong, while ... Bb5 and ... e5 are moves to watch for. (d) 14 Qc2 is one of White’s better replies:
(d1) 14 ... h6 15 Bh7+ (15 Ng3 Nf4=/+ 16 a3? Nxg2! 17 Kxg2 Rxf3!, Milos-CapNemo, playchess.com 2008; 15 Bb5!? with the idea 15 ... Bd7 16 Bxc6 is the kind of position in which White has an abstract edge but Black stands solidly) 15 ... Kh8 16 Nh4 Ne7 17 Rac1 Nf4 18 Bd3 (18 Nxf4 Qxf4 19 Nf3 Nc6=; 18 Ng6+ Nexg6 19 Bxg6 Nxe2+ 20 Qxe2 e5=) 18 ... Nxd3 19 Qxd3 Bd7 or 19 ... Kh8 verges on equality. (d2) 14 ... Nf6 looks best; for example, 15 a3 (15 Qd2 Bd7 16 h3 Rac8 or 16 ... Rae8 intending ... e5; 15 Bb5 Bd7 16 Bxc6 Bxc6 17 Ne5 Nd7=) 15 ... Bd7 16 b4 (16 Rfe1 Rac8 17 Qd1 Rce8 18 h3 e5 19 dxe5 Nxe5 20 Nxe5 Rxe5 21 Nd4 Rxe1+ 22 Qxe1 Nh5!=) 16 ... Rac8 (16 ... e5 17 dxe5 Nxe5 18 Nxe5 Qxe5=) 17 Qb2?! Ng4 18 Ng3 e5 19 b5? Nxd4 20 Nxd4 exd4 21 Qxd4 Rf4 22 Qxa7 b6 23 Be2 Nxf2! 24 Bf3 Nd3-/+ Ponizil-De la Villa Garcia, Pardubice 2011. 5.2114
Qd2
This logical move covers f4 and connects rooks. 14 ... Bd7 Instead: (a) Moskalenko has played 14 ... a6 three times, preferring to stop Bb5: 15 Qe3 (15 Rac1 can be met by 15 ... Bd7 or 15 ... Nf4; Garrido Dominguez-Moskalenko, Roquetas Mar 2006, went 15 ... Rxf3!? 16 gxf3 Bd7 17 Qg5 g6 18 Kh1 Rf8 19 Rg1 Nf4 with complications; 15 Qg5 Nf4 16 Nxf4 Qxf4 was Perez Mitjans-Moskalenko, San Marti 2010, when 17 Qxf4 Rxf4 would have been equal) 15 ... Nf4 16 Nxf4 Qxf4 17 Rac1 Bd7= Perez Mitjans-Moskalenko, Barcelona 2011. (b) 14 ... Rxf3 ‘is extremely interesting’
(Moskalenko): 15 gxf3 Bd7 (15 ... e5!?) 16 Qg5 (16 Kh1 Rf8 17 Qe3 e5!?) 16 ... g6 17 Rad1 Rf8 18 f4? (18 Qe3) 18 ... Nxf4 19 Nxf4 Nxd4!-/+ Basas Gamazo-Moskalenko, Catalonia 2006. 15 Qe3 15 Rad1 Rxf3!? (15 ... a6 16 Rfe1 Rxf3 17 gxf3 Rf8 18 Qg5 g6! with unclear complications, Dzhakaev-Visser, Hoogeveen 2000) 16 gxf3 Rf8 17 Qe3 e5! 18 dxe5 (18 Bc2!? exd4 19 Nxd4 Nf4) 18 ... Nxe5 19 f4 Nxd3 20 Qxd3 Nxf4 21 Nxf4 Rxf4 with compensation, Wolff-Moskalenko, Wijk aan Zee 1992 ( ... d4 and ... Bc6 are ideas). 15 ... a6!? 15 ... Nf4! is a little dull, but exchanging knights can ease Black’s defence; e.g. 16 Nxf4 (16 Bb5 Nxe2+ 17 Qxe2 Rxf3 18 Bxc6 Rf4 19 Bxd7 Qxd7 20 Qd2 Raf8 21 f3 Qd6=) 16 ... Qxf4 (or 16 ... Rxf4) 17 Rac1 (17 Bb5 Qxe3 18 fxe3 Nb8 19 Bd3 Nc6 20 a3 a6; basically all these lines are equal or extremely close to it) 17 ... Rac8 (or 17 ... a6) 18 Bb5 Qxe3 19 fxe3 Ne5! 20 Nxe5 (20 Bxd7 Nxd7=) 20 ... Rxf1+ 21 Bxf1!? Rxc1 22 Nxd7 Rc2 23 b4! Rxa2 24 Nc5 b6 25 Nxe6 a5 26 bxa5 bxa5 and the game will likely end in a draw. 16 Ne5 White has a little pull here. Cruz Ravina-Brynell, Barbera del Valles 2011, continued 16 ... Nb4 17 g3 (17 Bb1 Rac8 18 a4 – versus ... Bb5 – 18 ... Nc6 19 f4 Nf6+/=) 17 ... Nxd3 18 Qxd3 Bb5 19 Qd2 (19 Qe3!+/=, but it’s nothing serious) 19 ... Nf6 20 f3 Nd7 21 Nxd7 Qxd7=. 5.22 14 This is recommended by Tarrasch expert Tzermiadanos and has been played quite often: 14 ... Nf4
Ng3
15 Bb5 This is given ‘!’ by Tzermiadanos, but Black should be happy enough. The alternative is 15 Re1 a6 (15 ... Bd7 16 a3 a6 17 Rc1 Rae8 18 Bf1 e5= Lyrberg-Moskalenko, Helsinki 1992) 16 Rc1 Bd7 17 Ne5?! (17 Bb1 Rae8 18 Re3 e5=) 17 ... Nxe5 18 Rxe5, Azarov-Moskalenko, playchess.com 2006; 18 ... Qb4! 19 Rc7 Rad8 20 b3 Nxd3 21 Qxd3 Rf4!=/+. 15 ... Bd7!? The safe 15 ... Ne7, intending ... a6, has also been played. But 15 ... Qb4! is more ambitious; for example, 16 Qb3 (16 Bxc6 bxc6 17 Qc2 Ba6 18 Rfe1 Bd3) 16 ... Nh3+ 17 gxh3 Qxb3 18 axb3 Rxf3 19 Bxc6 bxc6 20 Rfc1 Rxb3 21 Rxc6 Rxb2 with a significant advantage. 16 Bxc6 Bxc6 17 Qb3 a5 18 Ne5
Nisipeanu-Visser, Groningen 1997; Tzermiadanos likes this for White, but Black can play 18 ... a4! (the game’s 18 ... Qb4 19 Qe3 Qxb2 20 Rfb1 Qc2 wasn’t bad either) 19 Qe3 (19 Qc2 a3 20 b3 Bb5 21 Rfe1 Qb4 favours Black slightly, since the d-pawn is awkward to defend) 19 ... Bb5 20 Rfc1 b6! (versus Rc5; 20 ... Ba6 is also playable) 21 Rc3 (21 Qc3? h5! with the idea 22 h4 g5!) 21 ... Qe7 22 Rac1 h5! 23 Rc7 Qf6 24 Rb7 Rab8 25 Rxb8 Rxb8 26 Rc7 Rf8=.
Chapter
Six
Tarrasch
Variation
with
5
f4
1 e4 e6 2 d4 d5 3 Nd2 Nf6 4 e5 Nfd7 5 f4
With 5 f4, White establishes his space advantage on the kingside and bolsters e5. Ideally, White would like to develop a bishop on d3, a knight on e2 and the other on f3, and eventually engineer a pawn break with f5. 5 f4 has two problems: it exposes the white king and central underbelly should d4 fall, and the c1-bishop is blocked from play. In terms of immediate issues, White takes another move without developing a piece and he will have to use more time to get the knight on d2 to a useful spot. 5 ... c5 As usual, Black attacks the pawn chain.
6 c3 Rather than expend more energy defending the pawn chain, White might want to develop and try to control d4 with pieces. There are two ways to pursue this: (a) 6 Ndf3 normally transposes with an early c3. Black usually plays 6 ... Nc6, and the most independent line is 7 Be3 (7 Bd3?! Qb6! forces 8 dxc5 Bxc5=/+, since 8 Ne2?? c4 and 8 c3 cxd4 9 cxd4 Nxd4 are weak; 7 Ne2 transposes to a sideline of the Classical System in Chapter 13, via the order 3 Nc3 Nf6 4 e5 Nfd7 5 f4 c5 6 Nf3 Nc6 7 Ne2 or, equivalently, 5 Nce2 c5 6 f4 Nc6 7
Ndf3; I will discuss that position there – see the introductory notes to 13.2) 7 ... Qb6 8 Rb1 f6!
Play might go 9 c3 (9 Ne2 cxd4 10 Nexd4 fxe5 11 fxe5 Bc5 12 c3 0-0 13 b4 Bxd4 14 Bxd4 Qc7 15 Bd3 Ndxe5 16 Nxe5 Nxe5 17 Bxh7+ Kxh7 18 Qh5+ Kg8 19 Qxe5 Qxe5+ 20 Bxe5 Rf5 with an edge, although it’s unlikely Black can make anything of it) 9 ... cxd4 10 Bxd4 (10 cxd4 fxe5 11 fxe5 Ndxe5; 10 Nxd4 fxe5 11 fxe5 Ndxe5) 10 ... Nxd4 and now: (a1) 11 cxd4 Bb4+ 12 Kf2 g5! gives a large advantage, threatening ... g4 and, after the knight moves, ... fxe5; this is difficult to do anything about: 13 h3 (13 exf6 Nxf6 is obviously bad, as is 13 a3 Be7 14 g3 g4 15 exf6 Bxf6) 13 ... gxf4 (13 ... 0-0 is also strong) 14 exf6 Nxf6 15 Ne2 (15 Bd3 e5) 15 ... Bd6 16 Nc3 Bd7 with an extra pawn and a great game. (a2) 11 Nxd4 fxe5 12 fxe5 g6! 13 Ngf3 Bg7 14 Bb5 0-0 15 Qe2 (15 Bxd7 Bxd7 16 0-0 Rf4! 17 Qd2 Raf8=/+) 15 ... Bh6 (versus Qe3) 16 0-0 Nc5=/+. (b) 6 Ngf3 is more challenging. After 6 ... Nc6, White has three logical moves: (b1) 7 c3 may well transpose. An independent line is 7 ... Be7 8 Bd3 0-0 9 Qe2 cxd4 (9 ... Qb6 10 Qe3!?) 10 cxd4 Nb4 11 Bb1 b6, with the idea 12 a3 Ba6 13 Qe3 Rc8! 14 Bxh7+! Kxh7 15 axb4 Bb5=. (b2) 7 Bd3!? is very ambitious, but it doesn’t help White’s centre.
Black has several respectable replies, among them: (b21) 7 ... Be7 is conservative and sound: 8 0-0 (8 c3 Qb6) 8 ... Qb6 9 c4!? 0-0 10 cxd5 exd5 11 Kh1 Nxd4 (11 ... a5!? is more interesting) 12 Nxd4 cxd4 13 Nb3 g6 14 Be2 Nc5 15 Nxd4 Ne4=. (b22) 7 ... cxd4 8 Nb3 Be7 (or 8 ... Bc5! 9 0-0 Bb6 10 Ng5 g6=/+; or 8 ... Qb6 9 0-0 a5 10 a4 Nc5) 9 0-0 0-0 10 Kh1 f6 11 Nfxd4 Nxd4 12 Nxd4 Nc5 13 Be3 fxe5 14 fxe5 Rxf1+ 15 Qxf1, Donev-Remensberger, Bad Ragaz 1993; and here 15 ... Bd7
intending ... Qc7 is slightly better for Black. (b23) 7 ... Qb6! tries for an immediate initiative. Briefly, play might go 8 dxc5 (8 0-0? c4! wins material; 8 c4 is the reason that 7 Bd3 was temporarily revived a few years back, but 8 ... cxd4! 9 cxd5 Nc5! 10 Bb1 exd5 11 0-0 d3 12 Kh1 Bf5 13 Nb3 Nxb3 14 axb3 Nb4 15 Ne1 Qd4 or 15 ... Qg6 gives Black a comfortable advantage) 8 ... Bxc5 (or 8 ... Nxc5) 9 Nb3 Bf2+ (9 ... Nb4 has the idea 10 Nxc5?! Nxc5 11 Qe2 Bd7! intending 12 Be3 Nbxd3+ 13 cxd3 Qxb2!) 10 Ke2!? (Moskalenko thinks 10 Kf1 is better; then one simple course is 10 ... Nc5! 11 Qe2 Nxd3 12 cxd3 Bd4 13 Nbxd4 Nxd4 14 Nxd4 Qxd4 15 Be3 Qb4= intending ... b6) 10 ... f6!? (10 ... 0-0 11 Rf1 Bc5= is simpler) 11 Rf1 Bc5 12 Nxc5 Qxc5 13 Be3 Qe7 14 exf6 Nxf6= Cabrera-Moskalenko, Tenerife (rapid) 2006. (b3) 7 Nb3 is the most important of White’s deviations. Then White has a fairly safe game, but he isn’t able to generate much positive action, which probably accounts for the line’s lack of overall popularity. Black has a number of satisfactory replies, including 7 ... Qb6 and these three: (b31) 7 ... cxd4 8 Nbxd4 is safe, although it lacks ambition:
This position might interest players who use other French Defence move orders; for example (among others), 3 Nc3 Nf6 4 e5 Nfd7 5 f4 c5 6 Nce2 (or 6 Nf3 cxd4 7 Nb5 Nc6 8 Nbxd4) 6 ... Nc6 7 Nf3 cxd4 8 Nexd4. Or, if you’re into the Alekhine Defence, try 1 e4 Nf6 2 Nc3 d5 3 e5 Nfd7 4 d4 e6 5 f4 c5 6 Nf3 Nc6 7 Ne2 cxd4 8 Nexd4 ! One solution goes 8 ... Nxd4 (or 8 ... Qb6 9 c3 Nc5, or 8 ... Be7 9 c3 Nc5 10 Be3 Nxd4 11 Nxd4 0-0 12 Be2 f6 13 exf6 Bxf6 14 0-0 Qd6=) 9 Nxd4 Bc5 (9 ... Nc5 10 Be3 Bd7) 10 c3 Qb6 11 a4! a5 12 Bb5 0-0 13 Be3 f6!
An interesting and complex position from Smagin-Moskalenko, Moscow 1995, when 14 Qg4 fxe5! 15 Nxe6 Rf7 16 Nxc5
Nxc5 17 Qf3 exf4 18 0-0 Be6 19 Bd4 Qd6! actually favours Black slightly. Instead, the game continued 14 Qh5!?, and White would have had no advantage after 14 ... g6! 15 Qe2 fxe5 16 fxe5 Rf7!= with full counterplay. (b32) 7 ... f6 is the move Black plays in so many of these pawn structures.
A quick overview indicates that the position is equal: 8 exf6 (8 c3!? is very solid; maybe this is one of those moments at which 8 ... c4 9 Nbd2 b5 should be tried, intending to castle and then play ... Rb8 and ... b4, because White will be so slow in getting any kingside advance co-ordinated; the alternative 8 c4!? could become wide open with 8 ... fxe5 9 dxe5 Nb6 10 cxd5 exd5 11 Bb5 Bf5!? 12 0-0 Be7 13 Be3 d4 14 Bf2 0-0 15 Rc1 Nd7 16 Bxc6 bxc6 17 Qe2 Be6 18 Nbxd4! cxd4 19 Nxd4 Bd5 20 Nxc6 Qe8!?, when White has enough for the piece but no more) 8 ... Qxf6 9 Be3 Bd6 10 g3 cxd4 11 Nbxd4 e5! 12 fxe5 Ndxe5 (Williams); then play might go 13 Bg2 0-0 14 0-0 Ng4! 15 Ng5! Qe5 16 Rxf8+ Bxf8 17 Bf4 Qxd4+ 18 Qxd4 Nxd4 19 Bxd5+ Kh8 20 Nf7+ Kg8 21 Nh6+ Kh8 22 Nf7+=. (b33) With 7 ... c4 8 Nbd2 b5, Black looks as though he’s asking to be attacked on the kingside. But Sigma on the ChessPublishing forum points to Cabrera Trujillo-Garza Marco, El Sauzal 2010, which took a different tack with 8 ... h6!? 9 Be2 (9 b3 b5 doesn’t seem to be a problem, and 9 c3 g5 is similar) 9 ... g5!?, proving that this advance, so ubiquitous in the French, can be played anywhere! The game went 10 Nf1 (10 c3 g4 11 Ng1 Rg8!?) 10 ... g4 11 Ng1 Rg8 12 Ne3 f5! 13 g3 b5 14 h3 h5 15 c3 b4, and White was cramped on both sides of the board. You should take a look at some early alternatives on moves 9 and 10 before jumping into this, but it looks quite reasonable to me, and perhaps more fun than some of the others! 6 ... Nc6 6 ... cxd4 7 cxd4 Nc6 is a purely transpositional attempt to avoid potentially dull lines where White tries to recapture with a knight on d4. At first sight, this seems to cede the promising c3-square for White’s g1-knight, or the b3-square to his d2knight; but it turns out that neither means much. The player wishing to use this order should do some research, but here’s a few notes to guide you: 8 Ndf3 (8 Ngf3 f6! 9 Bd3 fxe5 10 fxe5 Nxd4! transposes to 7 Ngf3 below, without some of the complications) 8 ... Qb6 (8 ... Be7 9 Ne2 0-0 10 Nc3 f6 11 Bd3 Qb6 is the note on 9 Ne2; while 10 g3 leads to main lines) 9 h4 (9 g3 transposes directly to the main line; 9 Ne2 f6 10 Nc3!? is the obvious way to avoid main lines, but then 10 ... Be7 11 Bd3 0-0 12 Bb1!? – to cover d4 and prepare Qd3 – 12 ... fxe5 13 fxe5 Rxf3! throws a wrench in the works: 14 gxf3 Bh4+ 15 Kf1 Ndxe5 16 Qc2 Ng6-/+) 9 ... Be7 or 9 ... Bb4+, and we are back to the main 8 h4 lines below. Every one of these transpositions leads to lively and double-edged play. 7 Ndf3 This natural move introduces the main line of the f4 Tarrasch system. 7 Ngf3!? has never caught on among masters.
Black can play any number of moves such as 7 ... Be7, 7 ... Qb6, and of course 7 ... cxd4. Here are two other moves which keep things complicated: (a) 7 ... a5 discourages Nb3 and grabs territory on the queenside: 8 Bb5 (8 Be2 Qb6 9 dxc5 Bxc5 10 Nb3 Bf2+ 11 Kf1 Nc5!, Serper-Glek, Frunze 1988; 8 a4 cxd4 9 cxd4 Bb4!?) 8 ... Qb6 9 Qa4 cxd4 10 cxd4 Ndb8! 11 Qb3 Bd7 12 Bxc6 Qxc6 with the bishop pair and a comfortable game, Smagin-Kindermann, German League 1997. (b) 7 ... f6!? can be a bit wild. Then the critical move is 8 Bd3!? (8 Nb3 transposes to 6 Ngf3 Nc6 7 Nb3 f6 8 c3 above; after 8 g3 Qb6 9 Nb3 cxd4 10 Nbxd4, the most exciting line is 10 ... fxe5 11 Nxe6! Nc5! 12 Nxf8 Bg4 13 h3 Ne4 14 Rh2 Bxf3 15 Qxf3 Rxf8 with dynamic equality) 8 ... Qb6 (threatening d4) 9 a4! (9 Qa4?! cxd4 10 cxd4 g5! destroys White’s centre and gives Black the advantage; 9 Qb3 Qxb3 10 axb3 cxd4 11 cxd4 g5! and Black wins the centre or material; 9 Nh4 cxd4 10 Qh5+ Kd8 with the idea 11 Ng6 hxg6 12 Qxh8 dxc3-/+) 9 ... fxe5 (or 9 ... Be7) 10 Nxe5 (10 fxe5 Be7 11 0-0 0-0) 10 ... Ndxe5 11 fxe5!? cxd4 12 Qh5+ Kd8 and Black, a pawn up, is relatively safe; e.g. 13 Nf3 Kc7 14 Rb1!? dxc3 15 bxc3 Qc5 and White has limited compensation. 7 ... Qb6
Our repertoire move, which has always been the main line. Black pressures d4, ties the c1-bishop to defence of b2, and indirectly discourages White from castling due to potential captures on d4 and e5. His main asset is a lead in development, which can give him time for the breaks ... f6 and ... g5; expansion by ... a5-a4 is also common. Finally, White often isn’t able to castle and has to settle for Kf1 or Kf2. For his part, White has the massive, advanced centre, which confines Black’s pieces. In particular, Black’s knight on d7 has little future without a sacrifice on e5 or another violent breakthrough; barring that, the knight has to take precious time to
redeploy (to b6, for example) while White catches up in development. In addition, White has the flank advances h4 and a3/b4 to cramp Black’s position further, and potential kingside attacking prospects based upon f5. Ultimately, it proves almost impossible to achieve this kind of advance without permitting Black successful central counterplay. A couple of respectable alternatives to 7 ... Qb6 bear mention: (a) 7 ... cxd4 (7 ... f5 has the same idea) 8 cxd4 f5 is one of the more popular secondary set-ups, the point of which is shown by 9 Bd3 Be7 10 Ne2 0-0 11 0-0 a5 (or ... Nb6 first), with the typical follow-up ... a4, ... Nb6, ... Bd7 and a queenside attack, counting upon the move ... f5 and the presence of White’s king on the kingside to neutralize an attack by White in that sector. (b) 7 ... Be7 is flexible and may transpose after 8 g3 or 8 Ne2, but it allows 8 Bd3, since White’s d-pawn doesn’t fall as it does after 7 ... Qb6. Then 8 ... Qa5 is an independent idea, pinning the c-pawn and thinking about ... b5-b4:
White has several responses, one simply to develop: 9 Ne2 (if White wants to recapture with a pawn on d4, he should move his king: 9 Kf1 b5! 10 dxc5 b4 with good counterplay, Ripper-Manigk, corr 1989; 9 Kf2?! isn’t very attractive in the face of 9 ... b5, or 9 ... Qb6 10 Ne2 f6 11 Kg3 g5!, Adorjan-Vaganian, Teesside 1974, with the idea 12 exf6?! gxf4+ 13 Bxf4 Nxf6=/+) 9 ... b5 10 0-0 (10 dxc5 Nxc5 11 Bc2 b4=) 10 ... g6 11 g4!? (11 Bd2 Qb6 12 a3 a5 is complex), Ermenkov-Ravikumar, Copenhagen 1982; and here 11 ... h5! (11 ... b4 12 f5!) 12 h3 b4 looks satisfactory, since 13 f5 hxg4 14 hxg4 gxf5 15 gxf5 exf5 16 Bxf5 Nf8! heads for the blockade square e6 while preparing to bring the queen rook to the kingside. After 7 ... Qb6, White has four main moves: 6.1 6.2 6.3 6.4
8 8 8 8
Ne2 g3 h4 a3
8 Nh3 neglects the centre; e.g. 8 ... f6 9 exf6 Nxf6 10 Bd3 cxd4 11 cxd4 Bb4+ 12 Ke2 0-0 13 a3 Bd6 14 Re1, Shirazi-J.Watson, Los Angeles 1981; and 14 ... Ne4 was good, but 14 ... Nh5! 15 g3 e5! is virtually winning. 6.18 Ne2 This was out of fashion for a while but has seen a revival. In part, that’s because, in the Classical Variation, there are two ways to reach this position: 3 Nc3 Nf6 4 e5 Nfd7 5 Nce2 c5 6 c3 Nc6 7 f4 Qb6 8 Ngf3, or here 5 f4 c5 6 Ngf3 Nc6 7 Ne2 Qb6 8 c3, both re-investigated in face of the difficulty of winning against other lines in the Classical. The background to those two move orders is explained in Chapter 13, but I’ve shifted the main lines here so that everything stemming from 7 ... Qb6 is in one place. 8 ... f6 If you’d like to keep things under control a bit longer, the popular 8 ... Be7 is a good alternative, often transposing but bypassing some wild lines. For example, 9 g3 will transpose to 8 g3 lines (e.g. 9 ... cxd4 10 cxd4 f6 11 Bh3 is 6.21), and 9 a3 0-0 (9 ... c4!? is interesting) tends to transpose if White plays g3 or h4.
Otherwise: (a) 10 b4 cxd4 11 cxd4 a5!? 12 b5! Qxb5 13 Nc3 Qb6 14 Rb1 Qd8 15 Bd3 gives White compensation for the pawn. Here 11 ... a6! is a good alternative. (b) 10 Ng3?! (this is seldom good when Black is so far ahead in development, because it takes a tempo and abandons control of d4) 10 ... cxd4 11 cxd4 f6 12 Bd3 fxe5 13 fxe5 and 13 ... Rxf3! 14 gxf3 Nxd4 gives Black more than enough for the exchange, while a safer version of this is 13 ... g6!=/+ intending 14 ... Nxd4 or 14 ... Rxf3! next, since Bxh7+/Qh5 ideas are eliminated. After 8 ... f6,
White 6.11 6.12
has
two
major
traditional
options 9 9
and
a
host
of
minor
ones: g3 a3
Others: (a) 9 exf6 Nxf6 10 g3 is solid-looking but passive: 10 ... Bd6 11 Bg2 0-0 12 0-0 Bd7 (12 ... Ne7!? 13 Ne5 Nf5 is also good) 13 Ne5!? Be8 with a little edge, intending ... Bh5. (b) 9 Qb3 Qc7!? (arguing that the queen isn’t effective on b3; 9 ... Be7= is the sound move, because White has to show
that he has a plan and 10 Qxb6 axb6 doesn’t impress) 10 g3 cxd4 11 cxd4 Bb4+ 12 Kf2 0-0 13 Kg2 Ba5!? (13 ... Be7!=) 14 Be3 Bb6 15 Rc1, B.Lalic-G.Dizdar, Croatian Team Ch. 1998; and 15 ... Qd8 is a logical idea: Black can undertake ... a6/ ... Ba7/ ... Nb6/ ... Bd7. White may stand slightly better, but he doesn’t have an obvious plan. (c) 9 Ng3? is poorly timed: 9 ... cxd4 10 cxd4 Bb4+ 11 Ke2 g5! 12 exf6 (12 a3 g4!) 12 ... g4 13 f7+ Ke7 14 Nd2 Nxd4+ 15 Ke1 Qc7 and wins. (d) 9 h4 is a thematic move for the f4 lines, often transposing to 6.3 (8 h4). With this timing Black does have an extra option: 9 ... cxd4 (for 9 ... Be7 10 a3, see 6.12; here 10 h5 0-0 makes it difficult for White to develop) 10 cxd4 fxe5 (10 ... Be7) 11 fxe5 Bb4+ 12 Nc3 0-0 13 a3 Be7 14 Na4 Qd8=, intending 15 Bd3 Nxd4! 16 Nxd4 Bxh4+ 17 Kd2 Rf2+ 18 Be2 Nxe5 with good chances. 6.11 9 g3 cxd4 10 cxd4 10 Nexd4 is well answered by 10 ... fxe5 (or 10 ... Nxd4 11 cxd4 Bb4+ 12 Bd2 fxe5 13 fxe5 0-0 14 Bxb4 Qxb4+ 15 Qd2 Qxd2+ 16 Nxd2 Nb8!=) and now: (a) 11 Nxe6 is normally queried, although none of the given solutions are definitive. Play might go 11 ... Nf6 (or 11 ... e4 12 Nfd4! Nf6 13 a4!, and now 13 ... Bd7 will lead to equality, but 13 ... Nxd4 14 Nxd4 a6 15 a5 Qc7 16 Be3 Bc5 is more interesting) 12 fxe5 (12 Nxf8 Bg4 13 Be2 Bxf3 14 Bxf3 Rxf8=, or here 13 ... exf4!? 14 Bxf4 Ne4) 12 ... Bxe6! 13 exf6 0-0-0 with adequate compensation, Strobel-Piasecki, German League 1993. (b) 11 fxe5 Nc5 12 Bh3 (12 Nxc6?! bxc6 13 Nd4 g6 14 Qc2 Bg7 15 Bf4 0-0 16 Bh3 Ne4!) 12 ... Be7 13 0-0 0-0 14 Qe2 Nxd4 15 Nxd4 Rxf1+ 16 Kxf1 Bd7= Atlas-Schenk, Austrian Team Ch. 2003. 10 ... Bb4+
This is the older approach, which still works well. The other main set-up is to place the bishop on e7 without checking, which can occur before or after ... cxd4 or ... f6. At this point, 10 ... Be7 11 Bh3 transposes to 6.21 (8 g3 Be7 9 Bh3 cxd4 10 cxd4 f6 11 Ne2). 11 Nc3! Not 11 Bd2? fxe5 12 fxe5 (12 dxe5 Nc5 13 Bxb4 Qxb4+ 14 Kf2 Qxb2=/+) 12 ... 0-0 13 Bg2 Ndxe5! 14 dxe5 Nxe5, and White should avoid 15 Nf4? Qe3+ 16 Kf1 Bxd2 17 Qxd2 Qxd2 18 Nxd2 g5, as well as 15 Ned4? Nd3+ 16 Ke2 e5 17 Bxb4 Nxb4 18 a3 Qa6+–+. But 15 Bxb4 Qxb4+ 16 Kf2 Qe4 17 Nc3 Nd3+ 18 Kf1 Qc4 is also pretty miserable; Black’s centre is about to roll forward and White is likely to lose material one way or another. 11 ... 0-0!? 11 ... fxe5 12 fxe5 0-0 is well tested. It prevents exf6, but has the disadvantage of clearing the f4-square for White’s bishop. The play is dynamically balanced: 13 Bf4 (13 a3? Rxf3! 14 Qxf3 Nxd4 15 Qd1 Nxe5-/+; 13 Bg2?! Rxf3 14 Bxf3 Nxd4 or 14 ... Qxd4)
and now: (a) 13 ... Be7!? (attacking b2 and thinking about ... g5) 14 a3! (14 Qd2? g5! 15 Nxg5 Bxg5 16 Bxg5 Nxd4 17 Bg2 Nxe5–+ Dolmatov-Bareev, Elista 1997; 14 Na4 Qa5+ 15 Nc3 Qb6=, but much more fun is 14 ... Qb4+ 15 Bd2 Ndxe5! 16 Bxb4 Nxf3+ 17 Kf2 Bxb4 with full compensation; 14 Bh3? Qxb2 15 Bxe6+? Kh8 16 Rc1 fails to 16 ... Rxf4! 17 gxf4 Nxd4!-/+ intending 18 Nxd4 Bh4+) 14 ... Na5 (14 ... a6+/=) 15 Bh3 (15 b4 Qc6 16 Qc2 Nc4 17 Bd3 h6) 15 ... Re8 (15 ... Nb8 16 0-0 Nbc6+/=) 16 0-0 Nf8 17 Qd3 Bd7 18 Rf2 Nc4 or 18 ... Rac8; White has a limited edge in these lines, but they are double-edged and available if one needs to avoid the possibility of a draw in ‘b’. (b) 13 ... Ndxe5! is objectively best: (b1) 14 Nxe5?! Nxe5 15 Bxe5 Bxc3+ 16 bxc3 Qb2 17 Bf4 Qxc3+ 18 Kf2 e5 19 dxe5 g5-/+. (b2) 14 dxe5 appears risky for White, but the play is balanced: 14 ... Ba5 (14 ... d4 introduces a forcing sequence leading to equality: 15 a3 dxc3 16 axb4 cxb2 17 Rb1 Qxb4+ 18 Kf2 a5 19 Qe1 Qc5+ 20 Qe3 Qxe3+ 21 Kxe3 a4 22 Rxb2 a3 23 Ra2 Nb4 24 Bc4 b5! with equal chances) 15 Qd2! (15 Bd2? Nxe5 16 Be2 Bd7) 15 ... d4 16 Nxd4 Nxd4 17 0-0-0 Rd8!? (17 ... Nc6= is safer, with the idea 18 Qe1! Ne7!) 18 Bd3! (18 Kb1 Bd7! 19 Qxd4 Bc6!) 18 ... Qc7 19 Be3 g6 20 Rhf1 Qxe5 21 h4! Bd7, when White gets a clear draw from 22 Bxg6 (22 Bc4 Nf3!) 22 ... Bxc3 23 bxc3 Ne2+ 24 Qxe2 Qxc3+, but no more. (b3) 14 Bxe5 is the ‘main line’: 14 ... Nxe5 15 Nxe5 Bxc3+ 16 bxc3 Qb2 17 Qc1 (17 Nf3? Qxc3+ 18 Kf2 Rxf3+ 19 Qxf3 Qxa1) 17 ... Qf2+ 18 Kd1 Qxf1+ 19 Rxf1 Rxf1+ 20 Kd2 Rxc1 21 Rxc1 (21 Kxc1 Rb8 22 a4 Kf8 23 Kc2 b6 24 a5 bxa5 25 Rxa5 Rb7 26 Nc6 Bd7 27 Nxa7 Rb6=/+) 21 ... b6 22 Rf1 Rb8. This position has arisen in two games and is drawn, since Black can’t put his extra pawn to use in the face of White’s bind: 23 Rf7 (23 Ke3 Rb7 24 Kf4 Rc7 25 Rf3 Ba6 26 g4 Rc8 27 g5!, Kuular-Fadeev, Kyzyl 2007; White allows the rook exchange because after 27 ... Rf8+ 28 Ke3 Rxf3+ 29 Kxf3 Bc4 30 Nc6 a5 31 a3 Kf7, neither side can make progress) 23 ... Rb7 24 Rxb7! (½-½ Bok-Grover, Groningen 2010) 24 ... Bxb7 25 h4 Kf8 26 g4 Ke7, drawn shortly in R.RobsonG.Meier, Lubbock 2010.
12 a3! 12 exf6?! Nxf6 frees Black’s pieces, when White’s bad bishop on c1 balances Black’s on c8. 12 ... Be7
13 Bh3 Others: (a) 13 exf6 Nxf6 14 Bd3 a6 15 Bc2 Bd7 16 0-0 Be8!=/+ intending ... Bh5. (b) 13 Bd3 a6!? 14 b4 (14 Na4 Qa7 15 b4 fxe5 16 fxe5 g6! 17 Be2 Rxf3 18 Bxf3 Nxd4=/+ with the idea 19 Bb2 Nxe5!-/+) 14 ... fxe5 15 dxe5 Ndxe5 16 fxe5 Nxe5 17 Rf1 (or 17 Be2 Nxf3+ 18 Bxf3 Bd7!) 17 ... Nxf3+ 18 Rxf3 Bf6 19 Bb2 Qg1+ 20 Bf1 Qxh2 21 Qe2 Qh6 with unclear prospects. (c) 13 Na4 Qc7 and: (c1) 14 Bh3?! fxe5 15 Bxe6+ Kh8 16 dxe5 Ndxe5!. (c2) 14 Bd3 fxe5 (14 ... a6!? is a good option; in some cases ... b5 will follow) 15 dxe5 (15 fxe5? Rxf3! 16 Qxf3 Nxd4) 15 ... Nb6 16 Nc3! (16 Nxb6 Qxb6 17 Qc2 h6) 16 ... Nc4 17 Qc2 h6 18 b3 N4xe5! 19 Nxe5 Nxe5 20 fxe5 Qxe5+ with full compensation. (c3) 14 Be3 Nb6 (or 14 ... a6 15 Rc1 b5 16 exf6 Nxf6 17 Ne5 Bb7 18 Nc3 Nxe5! 19 fxe5 Nd7 intending 20 Nxd5 Qd8) 15 Rc1 Nxa4 16 Qxa4 Bd7 17 exf6 Bxf6 18 Ne5 Be8 19 Bh3 Qb6! with the idea 20 Bxe6+ Kh8 21 Nd7!? Bxd7 22 Bxd7 Nxd4 23 0-0 Qxb2=/+. 13 ... fxe5!? 13 ... Kh8 is playable; e.g. 14 Na4 Qa6! 15 Bf1 b5 16 Nc3 fxe5 followed by ... Qb6 regardless of which recapture, Sznapik-
Knaak, Bratislava 1983. 14 Bxe6+ 14 fxe5? fails as usual to 14 ... Rxf3 15 Qxf3 Nxd4-/+ and ... Nxe5. 14 ... Kh8 15 Nxd5 Qd8
This position has arisen in many games. Black is a pawn down but will exploit the shaky white centre with the help of tactics. 16 Bxd7 The best move, according to theory. Instead: (a) 16 fxe5? Ndxe5! 17 Nxe5 Bxe6 18 Nxe7 Qxe7 is actually winning, because White’s king is caught in the centre and light squares have fallen. (b) 16 dxe5 Ndxe5! is more complicated: 17 Nxe5! (17 Bxc8? Nxf3+ 18 Qxf3 Nd4! 19 Qe4 Bc5! threatens White’s bishop and ... Re8) 17 ... Bxe6 18 Nxe7! Qxe7 19 Nxc6 bxc6 20 0-0 and the opposite-coloured bishops in conjunction with White’s exposed king balances the extra pawn; for example, 20 ... Bd5 21 Re1 Qf6 22 Qc2 Rab8=. (c) 16 Nxe7 Qxe7 17 d5 e4! (17 ... exf4 18 Bxf4 Nf6=) 18 Ng5 Nc5! 19 Bxc8 (19 0-0 can be met by 19 ... Bxe6 20 dxe6 Nxe6 21 Nxe4 Nxf4= Se.Ivanov, or just 19 ... Nxe6=) 19 ... Nd3+ 20 Kf1 Raxc8 21 dxc6 Qxg5 22 cxb7 Rb8 23 Kg2 Qf6 24 Qe2 Rxb7=/+ Lukin-Se Ivanov, USSR 1984. 16 ... Bxd7
17 dxe5!
17 Nxe7 has a better reputation, but in spite of his two extra pawns, after 17 ... Qxe7 18 Nxe5 Nxe5 19 dxe5, White is under pressure and stands slightly worse: (a) 19 ... Bc6!? 20 Rf1 (20 0-0?? Qc5+ 21 Rf2 Rad8 and ... Qd5) 20 ... Rad8 21 Qe2 (21 Qc2 Bb5) 21 ... Rd5 and ... Bb5 is threatened along with moves such as ... Qe6 or ... Qc5. (b) 19 ... Rfd8! confers a surer advantage due to 20 Qe2 Qe8! or 20 Rf1 (20 Qe2 Qe8!) 20 ... Bh3 21 Qf3 Rac8! 22 Rf2 Qe6!. 17 ... Bg4! Black threatens ... Nxe5 and attacks White’s light-square weaknesses. 18 0-0!? White should return a pawn by 18 h3! Bxf3 19 Qxf3 Nxe5 20 Qe4!, when 20 ... Nd7! has the simple idea of ... Re8 and the clever one of ... Nf6; e.g. 21 0-0 (21 Nxe7 Nf6 22 Qxb7 Qd3!) 21 ... Re8 (or 21 ... Nf6 22 Nxf6 Bxf6 23 Re1! Qb6+ 24 Kg2 h6= intending ... Rfe8) 22 Nxe7 Nf6! 23 Qxb7 (23 Qf3 Qxe7 24 Bd2 Qd7 25 Bc3 Qxh3) 23 ... Rxe7 24 Qf3 Qd4+ 25 Qf2 Qd7 threatens ... Qxh3 and ... Rae8; hence 26 Be3! Qxh3 27 Bd4!, when 27 ... Rd8 is equal, though 27 ... Ne4 28 Qg2 Qxg3= is a simpler solution. 18 ... Bc5+ 19 Be3 19 Kg2 Nd4=/+. 19 ... Bxe3+ 20 Nxe3 Qb6 21 Qd2 Bxf3 22 Rxf3 Nd4 Lupulescu-Smerdon, Calvia Olympiad 2004. Black has threats on f3 and b3; Smagin suggests 23 Raf1, but then 23 ... Rad8 24 Qb4 Nxf3+ 25 Rxf3 Qxb4 26 axb4 g5!=/+, with the idea 27 f5? Rde8 28 e6 Rxe6 29 fxe6 Rxf3. 6.12
9
a3
9 ... Be7 9 ... cxd4 10 cxd4 Be7 is also playable, despite conceding c3 to a knight; for example, 11 Nc3 0-0 (or 11 ... a6, to preserve a7 for the queen) 12 Na4 Qd8 13 Bd3 fxe5 14 fxe5 Rxf3 15 Qxf3 Nxd4 16 Qh5 Nf5. 10 h4 A classic wing expansion which both prevents the move ... g5 and prepares to swing the rook to h3 for defence and attack. Instead, 10 b4 cxd4 gives White a choice of recaptures:
(a) 11 Nexd4 fxe5! (11 ... Nxd4 12 cxd4 0-0 is a playable alternative here) 12 Nxe6 (12 fxe5?! Ndxe5 13 Nxe5 Nxe5 14 Qh5+ Ng6! 15 Bd3 0-0 16 Bxg6? hxg6 17 Qxg6 Bf6 18 Be3 e5 19 Nc2 Bh4+ 0-1 F.Meyer-S.Pedersen, German League 1999) 12 ... Nf6 13 Nfg5! (13 Nxg7+? Kf8 14 b5 e4) 13 ... g6! 14 Qe2! h6 15 Ng7+ (15 fxe5 is met by 15 ... Nh5 16 Be3 d4) 15 ... Kf8 16 N5e6+ Bxe6 17 Nxe6+ Kf7 18 fxe5 Nxe5! 19 Qxe5 Bd6 20 Qe3 Rae8=/+. (b) 11 cxd4 0-0 12 Rb1 fxe5 (12 ... a5 13 b5 a4! was Shirov-M.Gurevich, Munich 1993) 13 fxe5 (13 dxe5 a5 14 b5 a4! 15 Nc3 Ncxe5! 16 fxe5 Nxe5 with two centre pawns and an attack against White’s underdeveloped position) 13 ... Rxf3! 14 gxf3 Bh4+ 15 Ng3 Nxd4
A position like this illustrates how Black’s compensation for the exchange is to a large extent structural, since he has no particular attack: 16 f4 Nf8 17 Bd3 Bd7 18 Bb2 Nf5 19 Bxf5 exf5 20 Bd4 Qb5 intending ... Ne6, and if anything it’s Black who stands better. 10 ... 0-0 11 Rh3 a5 Shipov suggests 11 ... cxd4 12 cxd4 Na5!, and we’ll see this idea in section 6.4 (8 a3). 12 b3 Another standard idea. It stops Black from holding down White’s queenside by ... a4 due to b4. 12 ... Qc7 13 Neg1!? Widely praised: White undevelops his pieces to hold the centre prophylactically and free his f1-bishop. On the other hand, his development is miserable. 13 Qc2 b6! 14 Bd2 Ba6 made for easy equality in Jamrich-Schneider Zinner, Budapest 2001. 13 ... a4!
Anand recommended 13 ... b6 intending ... Ba6, which was played later with satisfactory results. ElCreyente-MusiqueWand, ICC 2011, saw the immediate sacrifice 13 ... fxe5 14 fxe5 Ndxe5 15 dxe5 Nxe5 16 Nxe5!? Qxe5+ 17 Qe2 Qc7! with plenty of compensation: 18 Bg5 Bd6 19 Rf3 Bd7 20 a4 Bg3+!? 21 Kd1 h6 22 Bd2 Bxh4=/+. 14 b4 fxe5 Black ignores positional continuations and initiates the standard central demolition sacrifice. 15 fxe5 Ndxe5!? 16 dxe5 Nxe5 17 Nxe5 Qxe5+ 18 Qe2
Notice the defensive role of the rook on h3. What does Black have for his piece? Two extra, unopposed central pawns, a loose white structure to play against, and some active pieces (in contrast with White’s array on the first rank and vulnerable rook). 18 ... Bxh4+?! Black grabs a third pawn but in return has to accede to a queen exchange. Instead, in accordance with the advantages just outlined, he can profitably play for an attack by 18 ... Qc7! followed by ... e5 and/or ... Bf6. Black has the centre and White’s king is exposed. Of course, if White can develop he’ll have the upper hand, but on balance it will be extremely hard to defend and in fact the computer engines prefer Black. An immediate idea is ... cxb4 in combination with ... Bf6 (as well as ... e5 in due course), so play might go 19 Bg5! (almost certainly best; 19 Nf3? e5; 19 Qh5 cxb4 20 axb4 e5 21 Rg3 Bf5 22 Ne2 d4 23 Bd2 d3; 19 bxc5 e5 20 Rf3 Bf5) 19 ... Bd6! (19 ... Bxg5!? 20 hxg5 e5 21 Rf3! Bf5 is dynamically balanced) 20 Rf3 Rxf3 (20 ... Bd7 also looks slightly advantageous, as in the game ElCreyente-MusiqueWand, ICC 2001) 21 Qxf3 (21 Nxf3 cxb4 22 c4 b3!) 21 ... cxb4 22 axb4 (22 cxb4? Bxb4+ intending 23 axb4?? Qe5+) 22 ... Bxb4 23 Ne2 Bd6 and with a third pawn (including the dangerous passed apawn), Black is for choice, although White is certainly right in the game. 18 ... Bxh4+?! is Anand-Shirov, FIDE World Ch., Tehran 2000, which continued 19 Kd1 (19 Rxh4? Qg3+) 19 ... Qf6? 20 Nf3! Qxc3!? 21 Bb2 Qb3+ 22 Kc1 and White was winning. Instead of 19 ... Qf6, Black stood worse but still could have put his centre to use by 19 ... Qxe2+! 20 Bxe2 Bf6 (20 ... Be7!?) 21 bxc5! Ra5 22 Be3 Bxc3 23 Rb1 Bf6 24 Nf3 e5 25 Rh1 Rf7+/= with defensive prospects. 6.28 g3 With this move, White wants to develop, often by Bh3, and he creates a square for his king on g2 (we’ve seen the trouble it can get into on f2). He also shores up f4, in some cases lending support to a piece there. 8 ... Be7 This has become a standard modern approach in the 5 f4 systems. 8 ... Be7 is based upon the idea that ... Bb4+ tends to help the white king head towards sanctuary on the kingside, while wasting a tempo (the bishop almost always returns to e7 after ... Bb4+). As usual, the moves ... cxd4/cxd4 may have already taken place before ... Be7, but until dxc5 becomes a threat, it makes sense to keep the position flexible.
9 Bh3 The first thing to note about 8 ... Be7 (as opposed to 8 ... f6) is that 9 Bd3? simply loses the d-pawn. In fact, White’s need to commit to a plan is one of the advantages of the flexible 8 ... Be7: (a) 9 Ne2 (often coming from the important order 8 Ne2 Be7 9 g3) 9 ... cxd4 10 cxd4 f6 11 Bh3 is the main line with 11 Ne2 below (6.21). (b) 9 h4!? should be compared to 6.3 (8 h4); White’s g3 doesn’t fit in very well here, and one plan is 9 ... cxd4 10 cxd4 f6 11 Bd3 (11 a3 0-0 – or 11 ... Na5 – 12 b4 fxe5 13 dxe5!? a5! 14 b5 Ncxe5! 15 fxe5 Nxe5 with a huge attack, Sulskis-Saiboulativ, Cappelle la Grande 2005; here 14 ... Bc5! is also good 16 Ra2 Bc5 and White is already in some trouble) 11 ... 0-0! (11 ... fxe5 12 fxe5 0-0 13 Bf4! Qxb2 may be good but is less convincing) 12 Qc2 (12 Bxh7+? Kxh7 13 Ng5+ fxg5 14 hxg5+ Kg8 15 Qh5 Ndxe5! 16 dxe5 Qb4+ 17 Bd2 Qe4+ 18 Ne2 Nd4, winning) 12 ... h6 (or 12 ... f5 and White has ... Nb4 and ... Nxd4 to deal with) 13 Ne2 fxe5 14 dxe5 Nc5 15 Bh7+ Kh8 16 Qb1 d4 with a strong initiative. (c) 9 Kf2 cxd4 (or 9 ... f6) 10 cxd4 g5!? (there are many options; e.g. 10 ... 0-0 11 Kg2 f5 is a good version of an ... f5 plan, because White’s king gets in the way of a kingside attack; the follow-up would be ... a5-a4, ... Qd8, ... Nb6 etc) 11 h3 (11 fxg5 Ndxe5 12 Nxe5 Nxe5 13 Kg2 Nc6 with central pressure, Blumenfeld-Newman, corr 1987; 13 ... Nc4 14 b3 Nd6 is also possible) 11 ... gxf4 (11 ... f6!?) 12 gxf4, Beliavsky-Ree, Wijk aan Zee 1985; here both 12 ... f6 and 12 ... Rg8! are good moves. (d) 9 a3 a5 transposes to 8 a3 a5 9 g3 Be7 in the notes to 6.42; alternatively, Black has 9 ... cxd4 10 cxd4 Na5!? (10 ... f6 11 Bh3 0-0 is similar to the main lines, unless White plays 12 b3!?, which has various answers, a simple one being 12 ... Na5 13 Rb1 Nc4!? with the idea 14 Ra1 Nxa3!; 10 ... 0-0 11 Bd3 f6 is another course) 11 Bd3?! (11 Rb1 Bxa3; 11 b4 Nc4 12 Qb3 Qc7 intending ... a5) 11 ... Nb3 12 Rb1 Nxd4 13 Nxd4 Qxd4 14 Nf3 Qb6 15 Kf1 Nc5 16 Bc2 Ne4! 17 Ba4+ Bd7-/+ Glicksman-J.Watson, Buena Vista 1994. 9 ... cxd4 10 cxd4 f6!?
10 ... 0-0 is an obvious option. It transposes to 6.211 after 11 Ne2 f6, or to 6.22 after 11 Kf1 f6 12 Kg2 (or 11 Kf2 f6 12 Kg2). Here Black can also play 11 ... f5, or expand on the queenside by 11 ... a5 12 Kg2 a4 13 Ne2 (13 a3 Qa6=/+ – Gleizerov) 13 ... a3 (13 ... Qa6 and ... Nb6) 14 b3 Nb4! 15 g4 f6!, L.M.Perez-Gleizerov, Mondariz Balneario 2000. This is worth noting if you prefer a strategic struggle to the violent lines below. After 10 ... f6, White has a major crossroads: 6.21 6.22
11 11
Ne2 Kf1
11 Bxe6?! has the usual problem: 11 ... Ndxe5! 12 Bxc8 Nxf3+ 13 Nxf3 Rxc8 14 0-0 0-0=/+; White’s bad bishop and poor development count against him. 6.21
11
Ne2
This is a traditional main line, often beginning with 8 Ne2 9 g3 Be7 (or 9 ... cxd4 10 cxd4 Be7). The position is still debated, and calls for another split:
6.211 6.212
11 11
... ...
0-0 fxe5
6.211 11 ... 0-0 This is equivalent to 10 ... 0-0 11 Ne2 f6. 12 Bxe6+ White gets less than nothing from 12 0-0?! fxe5 13 fxe5? Ndxe5 or 13 Bxe6+ Kh8 14 Bxd7 Bxd7 15 fxe5 Bg4, or here 15 Bxd5 exd4 16 Bxc6 d3+ 17 Kg2 bxc6!. So 12 Rf1! is the main option: (a) 12 ... Nb4 is Williams’ original suggestion. He gives 13 Nc3 Qa6 14 Kf2 Nd3+ 15 Kg1 f5 intending ... Nb6 and ... Bd7. This is a good idea, although after 16 Qb3 the knight might want to be back on b4, so I’d suggest 14 ... f5 with approximate equality; e.g. 15 a3 (15 Kg1 Nb6 16 Rf2 Nd3! 17 Rc2 Nxc1) 15 ... Nd3+ 16 Kg1 Nxc1 17 Rxc1 (17 Qxc1 Qa5) 17 ... Nb6 18 Rf2 Bd7 19 Bf1 Qa5=. (b) I like the deceptively modest 12 ... Kh8!
Now Black is ready to blast open the centre: (b1) 13 Bxe6? Ndxe5 14 Bxd5 Nxf3+ 15 Bxf3 (15 Rxf3? Qa5+ 16 Nc3 Bg4) 15 ... Bh3-/+. (b2) 13 Rf2 g6! (13 ... Qa6 14 Kf1 Nb4 is interesting) 14 b3!? fxe5 15 fxe5 Ndxe5 16 Nxe5 (16 dxe5 Bb4+ 17 Kf1 Bc5! 18 Qe1 Bxf2 19 Qxf2 Qxf2+ 20 Kxf2 Nxe5 21 Bb2 Rxf3+ 22 Ke1 Re3=) 16 ... Rxf2 17 Nxc6! Rxh2 18 Nxe7 Rxh3 19 Kf2 e5 20 Nxc8 Qe6! 21 dxe5 Qf5+ 22 Kg2 Qh5! 23 Qg1 Rh2+ 24 Qxh2 Qxe2+ with perpetual. (b3) 13 Nc3 fxe5 (13 ... Qa6!?) 14 fxe5 Rxf3! 15 Qxf3 Nxd4 16 Qf2! (theory gives only 16 Qh5 Qd8! followed by ... Nc6 with a good game) 16 ... Nc6 (16 ... g5!? 17 Kd1! Qb4 18 Be3 Nc6 19 Bxe6 Ndxe5 20 Bxd5 Bh3 21 Kc2 Qd6 is probably a little better for White but conveniently messy) 17 Qxb6 Nxb6 18 Nb5! Bd7 19 Kd1 (19 Nc7 Rc8 20 Nxe6 Nxe5 21 Nf4 Bxh3 22 Nxh3 Ng4=; 19 Bf4 g5 20 Bd2 h5!?; e.g. 21 b3 g4 22 Bg2 Nxe5 23 Nd4 Rc8=) 19 ... Nxe5 (19 ... Nc4 20 Nc7 Rc8 21 Nxd5 Bc5!) 20 Nd4 Bf6 21 Bxe6 Be8! with a pawn for the exchange and pressure on White’s exposed king; for example, 22 g4 Ned7! 23 Ne2 Nc5 24 Bf5 Ba4+ 25 Bc2 Bb5 might follow, with a slight edge. (b4) 13 Qb3 Qa6! 14 Bd2 (14 Bxe6 Na5! 15 Qc3 Nxe5-/+; 14 Be3 Nb6=/+) 14 ... Nb6 15 Rf2 Nc4=/+ Malakhov-Ulibin, Russian Ch. 1995. 12 ... Kh8
13 Bxd5?! This is played most often. 13 exf6 is objectively correct and about equal. Black gets compensation for the pawn after 13 ... Bb4+ (or 13 ... Nxf6 14 Bxc8 Bb4+) and: (a) 14 Nc3 Nxf6 15 Bxc8 Re8+! 16 Kf2 Raxc8=. (b) 14 Bd2 Nxf6 15 Bxc8 Raxc8! (15 ... Bxd2+ 16 Qxd2 Ne4 17 Qd3 Raxc8 was slightly less active in Malakhov-Galinsky, Minsk 1997) 16 0-0 Rfe8 17 Bxb4 Qxb4 18 Rb1 Ng4 19 Qd3! Re3! 20 Qf5 Rf8 21 Qxg4 Rxe2 22 Rfd1 Qc4! with sufficient counterplay. (c) 14 Kf1 Nxf6 15 Bxc8 Raxc8 16 Kg2, Hübner-Mohr, German League 1988, 16 ... Qa6 (16 ... Ne4!?); e.g. 17 Ne5 Ne4 18 Qd3 Nxe5 19 Qxa6 bxa6 20 dxe5 Rc2 21 Kf3 Bc5 22 Rd1 Ng5+ 23 Kg4 Ne4 24 Kf3=. 13 ... fxe5 14 fxe5? This turns out to be a disastrous choice. The alternatives do not impress, but keep White’s disadvantage to manageable proportions: (a) 14 Bxc6 Qxc6 15 0-0 e4! 16 Ng5 (16 Ne5 Nxe5 17 fxe5 Bg4=/+) 16 ... h6 17 Nh3 Nf6 18 Nf2 b5! 19 Be3 Nd5 20 Qc1 Qxc1 21 Bxc1 Nb4 intending ... Nc2 and ... e3. (b) 14 dxe5 Ndxe5!? (or 14 ... Rd8! intending ... Nc5) 15 fxe5 Nxe5 (or 15 ... Bg4) 16 Nc3? (16 Nxe5? Bb4+; 16 Bf4? Nxf3+! 17 Bxf3 Bb4+ 18 Kf1 Bh3+ 19 Bg2 Qc6!–+ Lukianov-Gleizerov, USSR 1986; 16 Nf4 Bg4! 17 h3 Bxf3 18 Bxf3 g5-/+) 16 ... Bg4 17 Rf1 Bxf3 18 Qc2 (18 Bxf3 Nxf3+ 19 Rxf3 Qg1+) 18 ... Bxd5 19 Nxd5 Rxf1+ 20 Kxf1 Qb5+–+ Grill-Ihlenfeld, corr 1988. 14 ... Ndxe5! 15 Nxe5? Losing; but so does 15 dxe5 Nxe5 16 Nf4 Bg4, just a little more slowly. 15 ... Nxe5 16 Bf4 Bg4! 17 Bg2 Bb4+ 18 Kf1 Nc4 19 Qd3 Rae8and wins. 6.212 11 ... fxe5 12 fxe5 0-0 13 Rf1! The most challenging. 13 Bxe6+?! Kh8 will either transpose to 11 ... 0-0 or produce something like 14 Nf4 Ndxe5 15 dxe5 Rxf4! 16 Bxf4 Bxe6 and White stands poorly; or if 13 Nf4 Kh8 14 a3 then 14 ... g5! 15 Nxe6 Ndxe5! 16 dxe5 Bxe6 17 Bxe6 Rae8 (17 ... Nxe5!) 18 Qe2 Nxe5! 19 Nxe5 Qxe6 20 Rf1 Bf6=/+ E.Rosen-Zierk, St Louis 2010. 13 ... Kh8
14 Rf2 14 Nf4 Bb4+ 15 Bd2, W.Watson-Apicella, Ostend 1987; 15 ... Nxd4! 16 Nxd4 Qxd4 17 Bxb4 Qxb4+ 18 Qd2 Qe4+ 19 Qe2 Nxe5-/+. 14 ... Ncxe5 15 dxe5 Nxe5 This is Black’s standard attack. 16 Nfd4! 16 Ned4? Nxf3+ 17 Nxf3 e5 18 Bxc8 Bc5! 19 Rf1 Raxc8–+. 16 ... Bd7! 17 Bf4 17 Rxf8+ Rxf8 18 Qb3 Qa6 19 Nf4 g5-/+. 17 ... Nc4 18 Rf3! 18 a3 e5 19 Bxd7 exf4. These moves seems relatively forced. After 18 Rf3 play might continue 18 ... g5 (18 ... e5!? 19 Bxd7 exf4 20 gxf4 Rad8 gives Black a sufficient attack for the piece) 19 Bc1 Bc5 with more than enough compensation. 6.22
11
Kf1!?
White rushes his king to g2. 11 ... 0-0 12 Kg2 12 Bxe6+ Kh8 threatens 13 ... Ndxe5; hence 13 exf6! (13 Bxd5?? Qb5+; 13 Bh3 fxe5 14 dxe5 Ndxe5! 15 Bxc8 Nxf3 16 Bxb7
Qb5+ 17 Qe2 Qxb7 18 Nxf3 Rae8 with ideas of ... d4 and even ... g5 and a strong attack) 13 ... Nxf6 14 Bxc8 Raxc8 (now White has an extra pawn, but problems with delayed development, king position, and the e4 outpost) 15 Ne2 Ne4 (15 ... Nb4 16 Nc3! Ne4! 17 Kg2 Nxc3 18 bxc3 Rxc3) 16 Kg2 Bf6 17 Re1 Rfe8 with plenty of play; among other ideas are ... Nb4 and ... Nd6-f5. 12 ... Kh8 12 ... f5 is a less forcing move; e.g. 13 g4!? fxg4 14 Bxg4 Ndb8 15 Ne2 Na6 16 Ng5 Bxg5 17 fxg5 Bd7; this is hard to assess. 13 Ne2 13 b3!? f5 (13 ... a5 14 a4 Ndb8!? 15 Ne2 Na6, L.A.Schneider-M.Lindgren, Gausdal 1990; 16 exf6 Bxf6 17 Be3 Bd7=) 14 Be3 Ndb8! 15 Ne2 Bd7 16 g4 fxg4 17 Bxg4 Ba3 18 Rg1 Ne7=. 13 ... fxe5 14 fxe5 14 dxe5 Nc5!.
14 ... Ndxe5! Black is committed to this now. 15 dxe5 Nxe5 16 Ned4 The attack is dangerous and unclear after 16 Nfd4 g5!? (or 16 ... Nc4 17 b3 e5 18 Bxc8 Raxc8 19 bxc4 dxc4=) 17 g4 h5 18 Rg1! Nxg4 (18 ... hxg4?! 19 Kh1; 18 ... Bd7!? 19 gxh5 g4) 19 Bxg4 hxg4 20 Nc3 e5 21 Nxd5 Qh6 (or 21 ... Qc5 22 Nxe7 Qxe7 23 Re1 Qh7=) 22 h4! Bc5! 23 Bxg5 Qg6. This looks dynamically balanced; of course, there are many options. 16 ... Nxf3 17 Nxf3 e5 18 Qxd5 Bxh3+ 19 Kxh3 Qf2 20 Nxe5 Settling for equality because there is nothing better: 20 Qxb7 Rab8 21 Qxe7 Qxf3 22 Re1 Rf5 23 Bg5 h5 24 Rad1 Qg4+ 25 Kg2 Qf3+ 26 Kg1 Qf2+=; or 20 Bg5 Bxg5 21 Rhf1 Qe2 22 Qxe5 Qxe5 23 Nxe5 Bf6=. 20 Nxe5 is Shirov-Vallejo Pons, Spanish Team Ch. 2008, which continued 20 ... Qf5+ 21 Kg2 (21 g4? runs into 21 ... Qf2 22 Qd3? Qh4+ 23 Kg2 Rf2+ 24 Kg1 Raf8 25 Be3 R2f3! 26 Qe2 Rg3+ 27 hxg3 Qxg3+ 28 Qg2 Qxe3+ 29 Kh2 Bd6 – Vallejo Pons; here 22 g5 Rae8 is still no fun) 21 ... Qf2+ 22 Kh3 Qf5+ 23 Kg2 Qf2+ ½-½. 6.38
h4
some 6.31 6.32
A very popular idea. White plays to prevents ... g5 and prepares the rook lift Rh3 while supporting a kingside attack in lines. Black has several effective plans, of which I’ll present two active solutions: 8 8
... ...
Be7 cxd4
Other moves are 8 ... a5!?, to gain space on the queenside by ... a4 ( ... f5 will often follow); and 8 ... h5!? when, depending upon White’s reply, Black has ... Ne7-f5 or ... cxd4 and ... g6 with Leningrad System themes. 6.318 ... Be7 This can also be played after ... cxd4/cxd4; it is a flexible move which prevents the desirable 9 Bd3 because the d-pawn can be captured. 9 Rh3!? The alternatives are both important and very instructive: (a) 9 a3 is similar to 8 a3 after 9 ... cxd4 10 cxd4 Na5 (see 6.41), but the inclusion of h4 and ... Be7 seems a at least a fair trade-off for Black. Alternatively, 10 ... f6 puts Black one kingside attacking move ahead of the normal h4 lines, because ... Be7 has been traded for a3. I suspect that this makes kingside development by Ne2 or Bd3 riskier for White, whereas b4 can be met by ... a6, as in the 8 a3 lines, and with ... f5 if it becomes necessary. At any rate, 9 a3 cxd4 10 cxd4 Na5 can lead to:
(a1) 11 Rb1? Bxa3!. (a2) 11 b4?! Nc4 12 Bxc4 dxc4 13 d5? (but otherwise there comes ... Qc6 and ... Nb6) 13 ... exd5 14 Qxd5 a5!-/+ with the idea 15 b5 Nc5! 16 Qxc4 Qg6. (a3) 11 Ne2 Nb3 12 Rb1 Nb8!? (12 ... 0-0 leads to typically complex play following 13 Be3 f6; 12 ... Nxc1 13 Qxc1 0-0 14 Nc3 f6 is also unclear) 13 Be3 Nc6 14 Nc3 Bd7 15 Na4 Qa5+ 16 Nc3 Qb6 17 Be2?! Nca5 18 0-0 g6 19 h5 0-0-0 20 Bf2 Kb8 21 Bd3 and Black’s reorganization had succeeded in McShane-S.Williams, Reykjavik 2011; but he might have done well to open a line on the kingside with 21 ... gxh5 here. (a4) 11 Rh3 Nb3 12 Rb1, Yudasin-Gleizerov, USSR 1989. Now Black can play 12 ... a5 or 12 ... Nxc1 with confidence, but also 12 ... Nb8, to develop by ... Nc6 and ... Bd7 with a good game. (b) 9 Kf2 cxd4 10 cxd4 f6 11 Kg3 0-0 12 Ne2 fxe5 13 fxe5 Ndxe5! 14 dxe5 Nxe5=/+ 15 Qb3! (15 Qd4 Bc5!; 15 Ned4? Bd6; 15 Bf4?? Ng6!) 15 ... Qxb3 16 axb3 Nd3! 17 Kh2 Bd6+ 18 Kg1 Bc5+ 19 Kh2 Nf2-/+. (c) 9 h5!? is a bit slow: 9 ... cxd4 10 cxd4 f6 (ambitious; 10 ... h6 is a good alternative) 11 Bd3 (11 h6 could provoke 11 ... g5! 12 exf6 Nxf6 13 Nxg5 Bb4+ 14 Bd2 Qxd4 15 Bxb4 Qxf4!) 11 ... 0-0
12 a3! (12 h6 g6! attacks d4 and threatens the usual sacrifices on e5) 12 ... fxe5 13 fxe5 Nxd4! 14 Nxd4 Nxe5 with the usual ideas. Play might go 15 Bb1! Bd7 (to get to the c-file) 16 Nge2 Rac8 17 Be3 (17 b4 Bb5!; 17 Bf4 Ng4 18 Qd3 Rxf4! 19 Nxf4 Rc1+ 20 Kd2, Lane-Meesen, Leuven 1988; 20 ... Qxb2+! 21 Bc2 Rxc2+! 22 Nxc2 Nf2 23 Qf3 Ne4+ 24 Kd1 Ba4-/+) 17 ... Nc4 18 Qd3! Nxe3 and White should take a draw with 19 Qxh7+ Kf7 20 Qg6+ Kg8 21 Qh7+ (not 21 h6? Bh4+! 22 g3 Ng2+). (d) 9 Nh3?! decentralizes: 9 ... cxd4 10 cxd4 f6 11 Bd3 fxe5 12 fxe5 0-0 13 Nhg5 h6 14 Nxe6 Rxf3! 15 gxf3 Ndxe5 16 dxe5
Bxe6-/+; White has ... Bb4+, ... Nxe5, ... Nd4, ... Rf8, and even ... Bxh4+ to deal with. 9 ... cxd4 10 cxd4 f6 11 a3?! 11 Bd3 0-0 12 Ne2 fxe5 13 fxe5 Ndxe5! 14 dxe5 Nxe5 is familiar. After 15 Bc2!, 15 ... Bc5 is possible with good chances, or Black can bail out by 15 ... Bb4+ 16 Bd2 Ng4 17 Bxb4 Qf2+ 18 Kd2 Qe3+ with perpetual check. 11 ... 0-0 11 ... Na5 is also promising. 12 b4 fxe5 13 fxe5 Wollski-Eade, San Mateo 1995; and here 13 ... a5! (13 ... Ndxe5!? 14 dxe5 Nxe5 was played and should be slightly advantageous) 14 b5 Nxd4! 15 Nxd4 Nxe5 gives Black superb compensation. 6.32
8
...
cxd4
9
cxd4
Bb4+
10
Kf2
f6
Here Black directly undermines the white centre. 11 Be3 The most common move, although not necessarily best. True, 11 Bd3?? loses to 11 ... fxe5 12 fxe5 Ndxe5. Yet 11 Kg3 can allow Black his usual sacrificial motives and still achieve a passable position. Black plays 11 ... 0-0 and: (a) 12 Bd3? Nxd4! 13 Nxd4 (13 exf6 g6! – Gurevich) 13 ... fxe5 14 fxe5 Nxe5 15 Bc2 Ng6! 16 Bxg6 hxg6 is killing, intending moves such as ... Bd6, ... e5, and ... Qf2, Ljubojevic-M.Gurevich, Linares 1991. (b) 12 Nh3! fxe5 13 dxe5 (13 fxe5 Nxd4! 14 Nxd4 Nxe5 with a winning attack) 13 ... Be7 (13 ... Nc5!? 14 Kh2 Ne4 15 Bd3, Kharitonov-Berkes, Oropesa del Mar 1999; 15 ... h6 16 a3 Bc5 17 Qe2 Bd7 18 Bxe4 dxe4 19 Qxe4 Qb3! 20 Nd2 Qd5 21 Qxd5 exd5 with two active bishops and adequate compensation) 14 Kh2 Nc5! 15 Rb1!? Bd7 16 Be3 Qb4! with mutual chances. (c) 12 Kh3!?h6 (12 ... fxe5 13 fxe5 Nf6!? 14 exf6 e5+ – Knaak; and now 15 g4! is simply chaotic) 13 Bd3!? fxe5 14 fxe5 Nxd4! 15 Nxd4 Nxe5 with an attack.
11 ... 0-0! As opposed to 11 ... fxe5, this doesn’t give White the f4-square to put a piece on and block the attack. Yudasin-De la Villa Garcia, Leon 1992, saw the alternative 11 ... Be7 12 Qd2 0-0 13 Rd1 a5 14 Ne2 a4!? 15 a3 Na5 16 Nc3, when 16 ... Nb3! was best: 17 Qc2?! fxe5 18 fxe5 Nxe5! 19 dxe5 d4 20 Kg3!? dxe3 21 Bd3 h6 22 Bc4 (Yudasin), but now 22 ... Nd2!-/+. Finally, 11 ... fxe5 12 fxe5 0-0 will transpose to 11 ... 0-0, except in the case of 13 a3 Be7 14 b4 Qd8!? 15 Bd3 Ndxe5! 16 dxe5 Nxe5, Estrada-Watanabe, Bariloche 1991.
12 Bd3 12 Ne2 h6!? 13 a3 Be7 14 b4 Qd8 15 Kg1 Nb6=. 12 ... fxe5 13 fxe5 Nxd4! 14 Ne2
This is instructive, although I think that 14 Bxh7+! Kxh7 15 Bxd4 Bc5 16 Kg3! is objectively best. After 16 ... Bxd4 of S.Salov-Stijzhkin, USSR 1991, White should play 17 Qxd4! Qxd4 18 Nxd4 Nxe5 19 Ngf3 Nd3 20 Ng5+ Kg6 21 Rad1! Nxb2 22 Rb1 Nc4 23 Ngxe6 Bxe6 24 Nxe6 Rf7 25 Nf4+ and Black’s advantage is quite modest. 14 ... Bc5 Instead: (a) I like 14 ... Nxf3! 15 Bxb6 Nfxe5+ 16 Kg3 Nxb6=/+ (Kalinichenko), to be followed by ... Bd6. This is promising, to say the least: Black already has two pieces and two pawns for the queen, with the bishop pair, two passed central pawns, and a promising attack. (b) 14 ... Nxe5 is usually cited, but is slightly less convincing: 15 Bxd4 Ng4+ 16 Kg1! (16 Kg3 Qd6+ 17 Kxg4 e5+ 18 Kg3 exd4+ 19 Kf2 Bg4, Saltaev-M.Gurevich, Cappelle la Grande 2001; and perhaps here 20 a3! is the best move: 20 ... Ba5 21 b4 Bb6 22 Neg1, although Black still stands better here) 16 ... Bc5 17 Qb3! Qxb3 (17 ... Qd6!?) 18 axb3 Bxd4+ 19 Nexd4 e5 20 Nb5 e4 21 Be2 exf3 22 Bxf3 d4 23 Nxd4 Rd8, about equal. 15 Nexd4 Nxe5 16 Ke2 Zinchenko-Gavrjushin, Perm 2007. Now, although 16 ... Nxf3 17 gxf3 e5 was good for the advantage, 16 ... Ng4! 17 b4! (17 Qg1? e5; 17 Qb3 Qxb3 18 axb3 Nxe3 19 Kxe3 e5! 20 Nxe5 Re8–+) 17 ... Bxb4 18 Nc2 Bc5 19 Rb1 Qd6-/+ is simpler. 6.48
a3
Finally, we get to the currently most popular move at the top levels. White will very often play a3 anyway, and doesn’t want to commit his pieces until Black reveals what he’s doing. The immediate idea is b4, which will severely cut into the queenside play that we’ve seen Black achieve. We’ll look at three replies, all with different piece set-ups and different philosophies: 6.41 6.42 6.43
8 8 8
... ... ...
cxd4 a5 Be7
6.418 ... cxd4 9 cxd4 Na5
Aiming for b3 and c4. This only works because Black is ahead in development and can take a little time to reorganize his queen and knights to implement a light-square strategy. 10 Ne2 Obviously 10 b4!? Nc4 11 Bxc4 (but not 11 Bd3?! a5! and Black already stands better, since White’s structure fall apart, even after 12 Qa4 Qc7!, with the idea 13 b5? Ncxe5!) 11 ... dxc4 is important.
Play can go 12 Ne2 a5! (12 ... Qc6!? 13 d5! – otherwise Black takes over the light squares – 13 ... Qxd5 14 Qxd5 exd5 15
Nc3, Delchev-Kunosic, Bihac 1999; and Wells suggests 15 ... Nb6 16 Nb5 Kd7!?, when 17 Be3 Na4 18 Nxa7 Nb2! and ... Nd3 favours White very slightly) 13 Bd2! (13 b5 Qc7!? 14 Nc3 Nb6 15 Be3 Nd5 16 Nxd5 exd5 17 Qc2 g6! 18 0-0 Bf5=/+ D.HowellCapNemo, playchess.com 2009; here 13 ... Qxb5 14 Nc3 Qc6 loses a little time but looks good for Black) 13 ... axb4 14 Bxb4 and Black has 14 ... Qc6 with ... Nb6 to follow, which is equal, or 14 ... Bxb4+ 15 axb4 Rxa1 16 Qxa1 0-0!? 17 Qa5 Qc6 18 Nc3 Nb6, also equal after 19 Qc5! Nd5! 20 Qxc6 bxc6 21 Nxd5 exd5. 10 ... Nb3 This is given ‘!’ by Moskalenko. It isn’t played much, but perhaps it should be. I don’t think it serves Black well to let White avoid the exchange of queens or pieces by 10 ... Be7?! 11 Nc3 Qb3 12 Qd3. Instead, 10 ... Qb3!? is the favourite move here, when 11 Nd2 Qxd1+ 12 Kxd1 Nb6 or 12 ... b5 is double-edged; and 11 Nc3 Qxd1+ 12 Kxd1 Nb3 13 Rb1 a6 intending ... Nb6 looks balanced; but 11 Qd2! (hitting a5) creates the most problems, because Black has to justify his piece placement. Then 11 ... Nc4! 12 Qc3 Qxc3+ 13 Nxc3 gives White the easier game, but this is playable and probably the best way to go. 11 Rb1 Nxc1 12 Qxc1 Be7 13 Nc3 Here Moskalenko suggests simply 13 ... 0-0 14 Bd3 Nb8 intending ... Nc6, or Black can play: 13 ... Qd8 14 Bd3 Nb6 15 b3 0-0 16 0-0 Bd7 Intending ... Rc8. One’s first instinct might be that White stands better with an undisturbed centre and space advantage; but I don’t think he can create any immediate threats or establish an outpost. Generally, the bishop pair becomes more effective in the long term, and Black has a good share of the queenside action, so I suspect that this is equal or very close to it. 6.42
8
...
a5
Black both prevents b4 and at least hints at fixing a weakness on White’s queenside by ... a4. This comes at cost of a tempo, and since ... Bb4+ will no longer gain time, it’s natural to assume that White will stay close in terms of development. 9 b3 The leading move these days, designed to meet ... a4 with b4 as well as prepare Bb2 or even Ra2-f2. However, others may be just as good: (a) 9 g3 Be7 (9 ... f5 is a solid move; e.g. 10 Ne2 Be7 11 h3, Kindermann-Zsu.Polgar, Baden-Baden 1985, and 11 ... 0-0 12 g4 a4 13 Rg1 Kh8 14 gxf5 Rxf5 produces obscure play) 10 Bh3 cxd4 (10 ... a4 11 Ne2 Qa6 12 0-0 Nb6 appears sound enough; Black may even play ... g6, to meet g4 with ... h5) 11 cxd4 and now:
(a1) 11 ... 0-0 12 Ne2 f6!? 13 Bxe6+ Kh8 14 Bxd5 (14 exf6 Nxf6 15 Bxc8 Raxc8 16 0-0+/=; compare 8 g3: 13 exf6 in the notes to 6.211) 14 ... fxe5 15 fxe5 Ndxe5! 16 dxe5 Bg4!. This is Williams’ analysis; he gives a sample line 17 Rf1? Rad8 18 Nf4? Nxe5! 19 Qe2 Rde8 (winning), and points out that 17 Bxc6! Qxc6 18 Ned4 Qe4+ 19 Qe2 Qd5 20 0-0 Bc5 21 Be3 Bxd4 22 Bxd4 Bxf3 23 Qf2 may slightly favour White. Nevertheless, the opposite-coloured bishops make it double-edged. (a2) Again, Black might prefer not to ‘castle into it’ by 11 ... 0-0, and so play 11 ... a4 12 Ne2 Qa6 13 0-0 Nb6 14 Nc3 g6!? intending ... Nc4/ ... b5-b4. Some f5 sacrifice by White may be necessary, but it also might not work. (b) 9 Ne2 a4!? (9 ... g6 is a good prophylactic move) and: (b1) 10 g4 h5!? (10 ... Be7; 10 ... Qa5 11 Be3 Be7 12 Qc2 b6 13 h4 Ba6=) 11 gxh5 Rxh5 12 Ng3 Rh8 13 h4 Ndb8 14 Bd3 cxd4 15 cxd4 Na5! is unclear. (b2) 10 f5!? exf5 11 Nf4 cxd4 (or 11 ... Qa5 12 Bd2 c4!?, but then there’s no central pressure) 12 cxd4 Ra5 13 Bd2 Nxd4! 14 Nxd4 Qxd4 15 Bxa5 Qxf4 16 Qxd5 Bc5 17 Bd2 Qh4+! 18 g3 Qe7 with mutual chances.
9 ... cxd4 I prefer to wait with this, because it isn’t quite as flexible for Black; though for the popular h4 variation in this note, it doesn’t matter. The alternative is 9 ... Be7, when 10 Ra2! is non-committal, and defends f2 against all those ... Nxe5 sacrifices (10 h4 cxd4 11 cxd4 transposes to the main line). Then I like 10 ... g6!?, rather than the risky 10 ... 0-0 11 Bd3 f6 12 Bb1! (12 Ne2?! allows the familiar 12 ... cxd4 13 cxd4 fxe5 14 fxe5 Rxf3! 15 gxf3 Bh4+ 16 Kf1 Ndxe5=/+) 12 ... Qc7 (12 ... g6 13 h4 c4 14 Bc2) 13 Ne2 b6 14 Qc2 f5 15 g4! Ba6 16 gxf5 exf5 17 Ng3 g6 18 h4! with a very strong attack. 10 cxd4 Be7 11 h4
11 ... h5 I find this surprisingly effective, indefinitely delaying White’s attack. Its use by strong players is encouraging too, although there have only been a few games. An entirely different approach is 11 ... 0-0 12 Bd3 f5 (12 ... f6 may be playable, but is extremely risky)
Black challenges White to make progress and in the meantime plans to reorganize by ... Qd8/ ... Nb6/ ... Bd7/ ... a4!. White might defend his d-pawn by 13 Ne2 (to avoid 13 Bb2 a4 14 b4 Na5! 15 Qxa4!? Nc4! 16 Qxa8 Nxb2 17 Be2 Qc7, which is very risky for White; here 16 Qb3? Nxb2 17 Qxb2 Rxa3! is a very pretty shot) 13 ... Qd8 14 h5 Nb6 15 Rg1!? (15 h6 g6 16 0-0) 15 ... a4 16 b4 Qe8 17 g4 fxg4 18 h6! g6 19 Rxg4 Bd8!? (19 ... Bd7) 20 Ng3 Ne7. White’s attacking chances are real in this line, but Black has his own resources. 12 Nh3 This is the preference of masters at this time. In Saravana-Gleizerov, Gurgon 2009, White played 12 Ne2 Qd8 13 Nc3 Nb6 14 Bd3 g6!? (maybe Black should wait until f5 is a threat) 15 g3!? Bd7 16 Ra2 a4 17 b4 Na7, and here 18 Rc2 was a natural move to maintain the balance. 12 Nh3 is Sethuraman-Ulibin, New Delhi 2009 (and other games), where 12 ... f5!? was played; then 13 Be2! with the idea 13 ... Nf8 14 0-0! looks good for White. Instead, I like 12 ... Nf8!, when 13 Bd3 can be calmly met by 13 ... Bd7 14 Bc2 Rc8 or, after 13 Nhg5 Bd7 14 Be2 Ng6 15 Ra2, Black can safely castle kingside! Then he has the dual aim of arranging ... Bb5 and, when appropriate, bringing a knight to f5.
6.438 ... Be7 We’re used to this move, though at first it seems less effective because it doesn’t respond very quickly to b4. The main idea, however, is unique and apparently effective: 9 b4 9 g3 transposes to 8 g3 Be7 9 a3 in the notes to 6.2, or to those in 6.42 if Black plays 9 ... a5. 9 ... cxd4 10 cxd4 a6 Black prevents b5 and prepares his own queenside strategy, consisting of ... Qc7/d8, ... b5, ... Nb6, and at some point ... Nc4, often in conjunction with the break ... a5.
This is an old plan whose merits have only been widely recognized in the past few years. Of course, all those ... 0-0 and ... f6 set-ups are still possible, but after 9 b4 White can play Bd3 and, before any sacrifices can disturb his centre, he can resort to b5 if necessary. 11 h4 For some reason more popular than most other moves, although they all tend to get roughly the same, almost robotic (and strong!) treatment: (a) 11 Bb2 Qc7! (after 11 ... Qd8, I don’t like the fact that 12 Rc1 stops ... b5)
Here, surprisingly, White has serious trouble countering the primitive idea of ... b5 and ... Nb6: (a1) 12 Bd3 b5 13 Nd2 tries to match Black by playing Nb3-c5, but 13 ... Nb6 14 Nb3 Nc4 15 Bc3 a5! favours Black; e.g. 16
Nc5 axb4 17 axb4 Ne3! 18 Qe2 Rxa1+ 19 Bxa1 Qa7 20 Bc3 Nc4=/+. (a2) 12 Rc1 b5 13 Qc2 Bb7 neutralizes the c-file and ... Nb6-c4 follows. White might want to disturb the balance with 14 f5!? exf5 15 Qxf5, but 15 ... Nb6 proceeds with the plan, and White has no serious kingside attack; e.g. 16 Nh3 (16 Bd3 Bc8! 17 Qf4 Nc4) 16 ... Qc8 17 Qh5 h6 intending ... g6 or ... Nc4. (a3) 12 Ne2b5 13 Nc3 Nb6 14 Qc2 Bd7 15 Rc1 Qb7!? 16 Bd3 Nc4 17 f5?! (17 Bxc4 dxc4! 18 Ne4 a5) 17 ... Bxb4! 18 fxe6 Bxe6 19 Bxc4 Bxc3+ 20 Bxc3 bxc4-/+ Rudd-Dgebuadze, Liverpool 2006. (b) 11 Ne2 Qc7 12 Bd2 (12 Nc3 Nxb4!) 12 ... b5 13 Nc1 Nb6 14 Nb3 Nc4 leaves Black solidly placed, and ... a5 will soon follow. (c) 11 Be3 Qc7 12 Bd3 b5 13 Ne2 Nb6 14 0-0 Bd7 15 Bf2 Nc4 16 Bxc4 bxc4 17 Nc3 Na7 (17 ... 0-0!) 18 Bh4! Bxh4 19 Nxh4 g6? (19 ... Nb5 20 Nxb5 axb5 21 f5 exf5 22 Nxf5 Bxf5 23 Rxf5 0-0=/+) 20 f5! with a strong attack, Van der Wiel-Wilschut, Leiden 2010. (d) 11 Ra2 Qc7 12 Bd3 b5 13 Ne2 Nb6 14 0-0 Nc4 15 g4?! a5! 16 bxa5 Rxa5 17 Ng3 Qa7-/+ Rudd-Sengupta, Hastings 2010/11. 11 ... Qc7 11 ... Qd8 hits h4 and has some point here, with a stereotyped example going 12 Bd3 b5 13 Bb2 Nb6 14 Rc1 Nc4 15 Qb3 a5 16 Bc3 axb4 17 Bxb4 Nxb4 18 axb4 Ra3! 19 Qb1 and Black won quickly in Kuran-Gruber, Gmunden 2005; at this point 19 ... Ra4! would have been especially convincing.
12 Bd2 This may be getting dull, but the recapture on c4 with the d-pawn (normally not realistic) is noteworthy, because the long diagonal is unopposed: 12 Kf2 b5 13 Bd3 Nb6 14 Ne2 Nc4 15 Bxc4 dxc4 16 Nc3 Bb7 (16 ... Qb7! with the idea 17 d5 exd5 18 Nxd5?! Be6! 19 Nxe7 Nxe7) 17 d5 exd5 18 Nxd5 Qd8=/+ Papp-Dvirnyy, Forni di Sopra 2011. Or 12 Be3 b5 13 Bd3 Nb6 14 Ne2 Nc4 15 Bxc4 bxc4 (or again, 15 ... dxc4!?=/+) 16 Nc3 0-0 17 Qc2 f6 18 exf6 gxf6!? 19 f5!? Bd6 20 0-0 Ne7! 21 Bh6?, RuddV.Jones, Coulsdon 2011; 21 ... Nxf5! 22 Bxf8 Bxf8-/+ with ... Bh6 to come. 12 ... b5
13 Bd3 13 Ne2 Nb6 14 Nc1 Nc4 15 Nd3 a5 16 Nc5 0-0 17 Bd3 h6 18 Bc3 f6! illustrates that Black has play on the kingside as well once White is tied down on the queenside, D.Howell-R.Robson, Puerto Madryn 2009. 13 ... Nb6 14 Ne2 Nc4 15 Qc2 h6 16 0-0 0-0 17 h5 a5 18 Bxc4? dxc4! 19 bxa5 Nxa5 20 Rfb1 Nithander-M.Richter, Gothenburg 2011. Black went on to win after 20 ... Bd7, but an ideal reorganization would be 20 ... Qd7! and ... Bb7, with a nearly decisive position.
Chapter
Seven
Winawer
Variation:
Fourth
Moves
1 e4 e6 2 d4 d5 3 Nc3 Bb4
3 ... Bb4 is the Winawer Variation, one of the most fruitful and exciting openings in chess. Black puts pressure on e4 and prepares to double White’s c-pawns with consequent positional pressure. It is one of the premier chess variations because it embraces so many types of closed manoeuvring positions and wide-open tactical ones, often both within the same game. In general, 3 ... Nf6, discussed in Chapters 13-15, leads to more controlled positions, although the MacCutcheon Variation (Chapters 14-15) is more dynamic than the previous edition’s ... dxe4 system. After 3 ... Bb4, White’s main move is 4 e5. Then he has to suffer through a mighty counterattack on his centre and, in most cases, doubled pawns on c2 and c3. His rewards for this sacrifice are great in terms of both attacking chances and opportunities to dominate his opponent with his bishops, so 4 e5 continues to be easily the most popular choice. Nevertheless, some players prefer to avoid this kind of commitment and choose one of the alternatives in this chapter. None of them threatens the soundness of the Winawer, but each enjoys periodic popularity and the well-prepared French player must know something about them: 7.1 7.2 7.3 7.4 7.5 7.6 7.7
4 4 4 4 4 4 4
a3 Qg4 Bd2 Ne2 Bd3 exd5 Qd3
7.1 4 a3 In spite of the great popularity of 4 Ne2 and even 4 exd5, I feel that this is the most important of White’s fourth move deviations, at least in terms of creating the most difficulties for Black. In view of this, I want to offer a variety of solutions, the choice of which will like depend upon one’s taste for the resulting positions. 4 ... Bxc3+ 4 ... Ba5!? probably isn’t completely equal, but might appeal to those who want original play and/or don’t like 5 ... Ne7 in the next note. Here’s just a sample of what might happen: 5 b4 Bb6 6 e5 Ne7 7 Na4 Nf5 (7 ... 0-0 is more flexible; e.g. ... Nd7 and ... f6 or ... f5 might follow) 8 Nf3 0-0 9 Bg5 f6 10 exf6 gxf6 11 Nxb6 axb6 12 Bf4 Nc6 13 b5 (13 Bd3 e5! 14 dxe5 fxe5 15 Nxe5 Re8) 13 ... Nce7 intending ... Ng6 and perhaps ... e5. Probably White gets an edge somewhere with perfect play (the two bishops must be worth something), but that advantage should be very small and doesn’t negate the value of this idea for achieving a playable imbalance. 5 bxc3
In playing 4 a3, White hopes that his doubled c-pawns are compensated for by the bishop pair and Black’s dark-square vulnerability on the kingside. On the other hand, White loses a tempo and in the main line concedes Black a substantial lead in development. 5 ... dxe4 If one wants to avoid the preparation of a specialist in 4 a3, 5 ... Ne7 is a respectable option:
Then 6 e5 c5 is the main line of the Winawer. 6 Qg4 0-0 (intending ... e5) 7 e5 can be answered by 7 ... c5, which is another main line (6 ... Ne7 7 Qg4 0-0, as inPTF3); but here Black has some promising alternatives; e.g. 7 ... Nd7 intending ... f5, which secures the kingside before attacking via ... c5. Other tries for White are less tested: (a) 6 Nf3 dxe4 7 Ng5 c5!? (7 ... 0-0 8 Nxe4 e5! 9 Bg5?! exd4 10 cxd4 Qd5!) 8 Nxe4 cxd4 9 cxd4 Nbc6 10 c3 e5 11 Bc4 Bf5!? 12 Ng3 (12 Ng5? 0-0=/+) 12 ... exd4 13 0-0 0-0 14 Bg5?!, Banas-Herzog, Keszthely 1981; when the light squares would again give Black an edge after 14 ... Be6! 15 Bxe6 fxe6 16 cxd4 Qd5. (b) 6 exd5 exd5 7 Nf3 (7 Bd3 0-0 8 Ne2 Bf5=) 7 ... Bg4!? (7 ... Nbc6! 8 Be2 0-0 9 0-0 Bg4 10 h3 Bxf3 11 Bxf3 Na5 12 a4 Re8 and Black’s knights at least as good as White’s bishops, Kazantsev-Shchukin, St Petersburg 2000; 7 ... 0-0 8 Bd3 Bf5=) 8 Be2 0-0 9 0-0 Ng6!? (9 ... Nbc6 10 Rb1 b6=) 10 a4 (10 Rb1! b6 11 h3+/=) 10 ... Nc6 11 h3 Bf5 12 Ba3?! Re8 13 Qd2 Na5 14 Rfe1 f6 and White had some problems with his light-square weaknesses, O.Pedersen-Tonning, Gausdal 1997. (c) 6 Bd3 (probably White’s best non-transpositional move) 6 ... c5 7 exd5 (7 dxc5 has several adequate replies; e.g. 7 ... Nd7=, or 7 ... dxe4 and ... Qxd1+, or even 7 ... Qa5 8 exd5 Nxd5; the direct 7 Qg4 0-0 8 e5 is a main line and, instead, 7 ... c4 8 Be2 Qa5 9 Bd2 0-0 is equal; finally, 7 Nf3 is well met by 7 ... c4 8 Be2, when 8 ... Qa5!? 9 0-0 Nbc6 10 Bd2 f6 was Toshkov-
Chernin, St John 1988; alternatively, the straightforward 8 ... dxe4! 9 Ng5 Nd5 plans to meet 10 Bxc4 with 10 ... Qc7 or 10 ... e3 – here Wintzer analyses 10 Nxe4 f5 11 Bg5 Qc7 12 Bh5+ g6 13 Nf6+ Nxf6 14 Bxf6 0-0 15 Be5 Qa5 with good play) 7 ... exd5 8 dxc5 Nd7 (8 ... Qa5 9 Ne2 Nd7 10 Rb1 Nxc5!? 11 Bb5+ Bd7 12 Bxd7+ Nxd7 13 Rxb7 Nc5 14 Rb4 0-0 yields a position similar to some 7 Nf3 Positional Winawer lines; White may stand slightly better, but his extra pawn is largely balanced by his static weaknesses) 9 Be3 Qc7 intending 10 Qg4 (10 Ne2 0-0 11 0-0 Nxc5 12 Re1 Nxd3 13 cxd3 Re8) 10 ... Nxc5, when 11 Qxg7 Nxd3+ 12 cxd3 Rg8= equalizes; e.g. 13 Qf6 Rg6 14 Qh8+ Rg8 15 Qd4 Nf5!? (15 ... Rg4 16 Qf6 Rg6 17 Qh8+ Rg8=) 16 Qf6 Rxg2 17 Ne2 Qe7 18 Qxe7+ Kxe7=, in view of 19 Nf4 Nxe3 20 fxe3 Rc2 21 Nxd5+ Kd6 22 c4 Bg4=. After 5 ... dxe4, we will look at the main traditional move (6 Qg4) and a popular alternative:
7.11
6
7.12
f3 Qg4
6
The only other move to trouble Black would be 6 a4, but he has several good courses; e.g. 6 ... Nf6 (6 ... e5!?; 6 ... c5 7 Ba3 Qa5) 7 Ba3 Nbd7! 8 Nh3 b6 intending ... c5. 7.116 f3 This is the Winckelmann-Reimer Gambit. The idea is to get something resembling the Blackmar-Diemer (1 d4 d5 2 e4 dxe4 3 Nc3 Nf6 4 f3) with Black’s dark-squared bishop no longer available for defence. Here are two updated solutions to the previous edition: 7.111 7.112
6 6
... ...
Nd7 e5
6 ... c5 is generally thought to favour Black and I agree, but it gives White too many options and isn’t necessary. 7.111 6 ... Nd7 This is a very logical move because it gets a piece out, prepares central breaks ( ... c5 and ... e5), and also doesn’t commit to a fixed scheme of development.
7 Nh3 I think this is best because White has few other choices that get a piece out effectively. Nevertheless, a knight on h3 is not ideally placed for a Blackmar-Diemer sort of position (it belongs on f3 watching over e5), so many alternatives have been tried: (a) 7 fxe4? Qh4+ wins a pawn and forces White’s king to move. (b) 7 Be3 Ngf6 8 fxe4 (8 Bc4 Nb6 9 Bb3 Nbd5-/+) 8 ... Nxe4 9 Qg4 (9 Nf3 has several answers; the easiest is 9 ... Ndf6 10 Qd3 c5 11 dxc5 0-0=/+) 9 ... Ndf6 10 Qxg7 Rg8 11 Qh6 Rg6 12 Qf4 Nxc3-/+. (c) 7 a4 Ngf6 8 Ba3 (8 Qe2 exf3 9 Nxf3 0-0, or 8 ... 0-0 9 fxe4 e5!) 8 ... Nd5! (8 ... b6=/+) 9 Qd2 (9 Ne2 exf3 10 gxf3 Qh4+ 11 Ng3 f5-/+) 9 ... e3 10 Qd3 c5! 11 Ne2 (11 dxc5 0-0! 12 c6 bxc6 13 Bxf8 Qxf8-/+; 11 c4? Qa5+) 11 ... Qa5 (11 ... b6 12 c4 Ba6 13 Qb3 Ne7 14 Qxe3 Nf5) 12 Qb5 Qc7 13 dxc5 0-0 14 c4 Ne7=/+. (d) 7 g3 Ngf6 (or 7 ... exf3!? 8 Nxf3 b6) 8 Bg2 0-0! 9 fxe4 e5!=/+. (e) 7 Bf4 Ngf6 8 fxe4 (8 Qe2 Nd5!) 8 ... Nxe4 9 Qf3 f5!? (9 ... Ndf6 10 Be5 Qd5=/+) 10 Qh5+ (10 Bd3 Ndf6) 10 ... g6 11 Qh6 e5!? (or 11 ... Ndf6) 12 Qg7 (12 dxe5 Qe7 13 Nf3 Qc5) 12 ... Qf6 13 Qxf6 Nexf6 14 dxe5 Ne4 or 14 ... Nd5, both with a small but definite advantage for Black. (f) 7 Qe2 exf3 8 Nxf3 c5 9 Qd2 Ngf6 10 Bd3 b6 11 0-0 Bb7=/+. 7 ... Ngf6 8 fxe4 Almost always played. 8 Nf2 is definitely the sanest move, even if 8 ... exf3 9 Qxf3 doesn’t quite offer full compensation after 9 ... c5 10 Bd3 Qa5 (10 ... Qb6) 11 Bd2 0-0 12 Ne4 Nxe4 13 Qxe4 Nf6 14 Qh4 cxd4 15 cxd4 Qh5=/+. 8 ... Nxe4
9 Qg4 9 Qf3 Ndf6 (9 ... f5!?) 10 Bb2 (10 Be2 Bd7 11 0-0 Bc6) 10 ... e5!? (10 ... Bd7! 11 Bd3 Bc6=/+) 11 Nf2 (11 0-0-0 exd4 12 Rxd4 Qe7 13 Bd3 Nc5=/+) 11 ... Nxf2 12 Qxf2 0-0 13 0-0-0 Ng4! 14 Qd2 e4 15 Re1 e3 16 Qe2 Re8 (Black is winning) 17 h3 Nf2 18 Rh2 Qd6 or 18 ... Qg5!, Genis-Penzel, Frankfurt 2007. 9 ... Ndf6 10 Qh4 Black’s developmental advantage grows after 10 Qxg7 Rg8 11 Qh6 Nxc3!, hitting d4 as well as winning a pawn; this is even more convincing than 11 ... c5 with the idea ... Qa5, when Black retains a nice advantage. After 10 Qh4, Tahiri-J.Watson, Berlin 1997, went 10 ... Nxc3 11 Bd3 Nfe4 12 Qf4 f5 (or 12 ... Qxd4–+) 13 Be3 0-0 14 0-0 Qd6 15 Qh4 Bd7, with two extra pawns. 7.1126 ... e5 This is in many ways the ideal move. It frees the c8-bishop and therefore compels White to try to gain compensation quickly before Black develops.
7 Be3 Obviously 7 fxe4? Qh4+ is undesirable. Here are the logical alternatives: (a) 7 Bb5+ Nd7 (7 ... c6 8 Bc4 Qa5=/+; 7 ... Nc6) 8 Qe2 a6 9 Bc4 (9 Bxd7+ Bxd7 10 Qxe4 Ne7 11 dxe5 0-0 12 Ne2 Bc6=/+) 9 ... Ngf6! 10 fxe4 b5 11 Ba2 Bb7=/+.
(b) 7 a4 exd4 (7 ... Be6 is a good option) 8 cxd4 Nf6 (8 ... c5!? 9 fxe4! Qh4+ 10 g3 Qxe4+ 11 Qe2 Qxe2+ 12 Nxe2=; 8 ... Nc6! 9 Bb5 Bd7=/+) 9 fxe4 Nxe4 10 Nf3 0-0=/+. (c) 7 dxe5 Qxd1+ 8 Kxd1 Bf5 (or 8 ... exf3 9 Nxf3 Nc6 10 Bd3 Nge7=/+) 9 Ne2 (9 g4 Be6; 9 Bc4 Nc6 10 f4 f6!) 9 ... Nc6 (9 ... exf3 10 Nd4 Ne7) 10 Ng3 Bg6 11 f4 (11 Nxe4 Bxe4 12 fxe4 0-0-0+ 13 Bd3 Nxe5=/+) 11 ... h5 12 h4 0-0-0+ 13 Ke1 Nh6 14 Be2 Ne7! 15 Nxh5 Nhf5=/+. (d) 7 Bc4 Nc6 8 d5 Na5 9 Ba2 Nf6=/+. (e) 7 Qe2 exd4 (7 ... Nc6 8 dxe5 Nge7=/+ 9 Qxe4!? Bf5 10 Qe2 Qd5 11 f4 0-0-0) 8 Qxe4+ Ne7 9 Qxd4 0-0 10 Bf4, Grott-Van de Wynkele, desc-online.de 2004; and two good moves out of several are 10 ... Bf5 and 10 ... Qe8. 7 ... Nc6 Arguably the best move. (a) 7 ... Qe7 is almost unknown, but is also somewhat advantageous; e.g. 8 dxe5 (8 fxe4?! Qh4+ or even 8 ... Nf6) 8 ... Nc6 9 Qd5 (9 Bb5 Bd7; 9 f4 Nh6! 10 Qd5 Be6 11 Qxe4 0-0-0=/+ with ideas of ... Bd5, ... Ng4, ... f6 etc) 9 ... exf3 (9 ... Nxe5 10 Bd4 Nd7=/+) 10 Nxf3 Nf6 11 Qc5 Qxc5 12 Bxc5 Nd5 13 Bd4 Bf5 14 0-0-0 0-0=/+. (b) 7 ... exd4 has been played a lot, but doesn’t give Black much if anything; e.g. 8 Qxd4! (8 cxd4 Qh4+ 9 Bf2 Qe7) 8 ... Qxd4 9 Bxd4 Nf6 10 Bxf6 gxf6 11 fxe4 Be6 12 Ne2 Nd7 13 Nf4 0-0-0 with a tiny edge.
8 Bb5 Others: (a) 8 d5 Nb8 (8 ... Na5!) 9 Bc4 Nf6=/+. (b) 8 dxe5 Qh4+ (or 8 ... Bf5=/+) 9 g3 (9 Bf2 Qg5 10 Qe2 Qxe5 11 Qxe4 Qxe4+ 12 fxe4 Nf6 13 Bd3 0-0 14 Ne2 Re8 15 Ng3 Ng4=/+ Hanison-Killer, Kangasniemi 2001), Tinture-Jonckheere, ICCF email 2004; and 9 ... Qe7! 10 f4 f6 leaves Black with the upper hand. 8 ... Bd7 9 d5 9 Qe2 (9 dxe5 Qh4+ 10 Bf2 Qg5 11 Bg3 0-0-0=/+) 9 ... Nf6 10 Bg5 exd4 11 Rd1 0-0 12 fxe4 h6 13 Bxf6 Qxf6-/+ Titze-Killer, DESC email 2002. 9 ... Nce7 10 Bxd7+ Qxd7-/+ White hasn’t developed and his pawn structure is damaged. Oparaugo-Cech, Passau 1997, continued 11 fxe4 Nf6 12 Qd3 Qg4! 13 Qb5+ Nd7 14 Kf2 Qxe4 15 c4 Qf5+ 16 Kg3 Qxc2 and Black won shortly. 7.12
6
Qg4
Nf6
7
Qxg7
Rg8
8
Qh6
This rich and complex position features White’s bishops and dark-square play pitted against Black’s development, lightsquare play, and easier access to White’s weakness. There is a subtle interplay of strategies here, such that both sides can try a wide variety of set-ups which can be tailored to each other’s ideas. Properly played, these combine to produce a balanced struggle. Black has a few acceptable solutions, of which I’ll pick two that are both good and relatively safe, with a lessestablished one in the following note ‘b’: 7.121 7.122
8 8
... ...
c5 Rg6
Instead: (a) I have recommended 8 ... Nbd7 for several editions, but not only has a large mass of theory accumulated about it, I am also uncomfortable with a couple of lines involving a4 plans by White; they are about equal, but only with extremely accurate play by Black and putting no pressure on White. However, a little trick that Black can employ is 8 ... Nbd7 9 Ne2 (the problem move, with the idea Ng3 and a4; everything else has held up well for Black), and here instead of the normal 9 ... b6, which is marginal, play 9 ... c5, transposing to the line 8 ... c5 9 Ne2 Nbd7 in the notes to 7.121 below. (b) Another noteworthy move is 8 ... b6.
This doesn’t have a strong following, but theory doesn’t indicate any particularly effective way to attack it: (a) 9 a4 Rg6 10 Qe3 Nc6= stops a5 and intends ... e5. Ostojic recommends 10 ... Ba6, which also looks satisfactory. (b) 9 g3 Bb7 10 Bg2 Rg6 11 Qd2 Qd5! (Moskalenko); not only is Black better developed, 12 ... e3! is threatened. (c) 9 Nh3 Bb7 10 Bb5+ (10 Bg5 Rg6 11 Qh4 h6! with the idea 12 Bxh6?? Rg4) 10 ... c6 (10 ... Nbd7! looks better, since 11 Bxd7+ Kxd7 only assists Black’s light-square control) 11 Be2 Nbd7 12 Bg5 Qe7 13 0-0 0-0-0 14 f3?! (weakening) 14 ... Rg6 15 Qh4 Rdg8 16 Kh1 (16 fxe4 h6 17 Bxf6 Rxg2+ 18 Kh1 Nxf6) 16 ... h6!? (16 ... exf3!=/+; 16 ... c5 17 fxe4 Rxg5 18 Nxg5 Nxe4 19 Nf3 Qxh4 20 Nxh4 Nxc3 is complex – Moskalenko) 17 Bxf6 Qxf6 18 Qxf6 Nxf6, Hector-Brynell, Tastrup 1990, with a limited edge based upon 19 fxe4 Nxe4 20 Bd3 c5!. (d) 9 Ne2 Ba6 (9 ... Rg6 10 Qd2 Bb7 11 c4 c5 12 Bb2 Nc6 13 0-0-0 cxd4 14 Nxd4 Nxd4 15 Bxd4 Qe7= Ziatadinov-Rychagov, Vladimir 2004) 10 Bg5 Nbd7 11 Ng3 Bxf1 (when the light-squared bishops go off and White can’t castle, Black is almost always in the game) 12 Rxf1 Rg6 13 Qh4 h6 14 Bxf6 Nxf6 15 Nxe4, Bondarev-Yakimenko, St Petersburg 2011; and easiest was 15 ... Nd5! 16 Qxd8+ Rxd8 with the ideas ... f5 and ... Rxg2. (e) 9 Bg5
9 ... Rg6 (Moskalenko points out 9 ... Nbd7!, which is more flexible; then 10 Ne2 can be met by 10 ... Ba6) 10 Qh4 Bb7 (10 ... Nbd7 11 Ne2 Ba6? loses material to 12 Nf4; but 11 ... h6!? 12 Bxh6 Rg4 13 Qh3 Ba6 is interesting) 11 f3 (11 Ne2 h6 12 Bxh6 Rg4 13 Qh3 Nc6!? – Moskalenko; I suspect that this is slightly better for White, but within bounds and the reward is a complex and wide-open position; 11 Bb5+ Nbd7 12 Ne2= Knaak; then 12 ... h6! 13 Bxh6 e3!? 14 Bxe3 Rg4 15 Qh8+ Rg8 16 Qh4 Rg4= is a repetition, or perhaps 14 ... Bxg2 is preferable) 11 ... Nbd7 12 fxe4 h6! 13 Bxf6 (13 Bxh6 Rg4 14 Qh3 Rxe4+) 13 ... Qxf6 (13 ... Nxf6! 14 e5 Nd5=/+) 14 Qxf6 Nxf6 15 e5, Bronstein-Hort, Moscow 1971. The game was agreed drawn after 15 ... Nd5, but 15 ... Rxg2! 16 Bxg2 Bxg2 17 exf6 Bxh1 would be worth pursuing. 7.1218
...
c5
A logical and consistent move. Along with the threat to White’s d-pawn and normal development, Black has in mind ... Qa5, or else ... Qc7 with pressure down the c-file. By delaying ... Nbd7 (still a common idea), he keeps ... Nc6 as an option, and leaves open the manoeuvre ... Bd7-c6 (or ... Bd7-a4). Other features will be seen in the analysis. 9 Ne2 Easily the main line after 8 ... c5. Instead: (a) 9 Bb2 only works if White can successfully play c4: 9 ... Nbd7 (9 ... Bd7! intends ... Ba4, and if 10 a4, then 10 ... Bc6 11 Ne2 Nbd7 develops harmoniously) 10 Ne2 Nb6 11 Ng3 (11 dxc5 Nc4 12 Rd1 Qe7 13 Bc1 Rg6=) 11 ... cxd4 12 cxd4 Na4 13 Bb5+ (13 0-0-0 Bd7=/+) 13 ... Bd7 14 Bxd7+ Kxd7=/+. (b) After 9 Bg5, Black has several reasonable courses, including 9 ... Nbd7. Two other instructive fights between bishops and weaknesses:
(b1) Moles calls 9 ... Ng4!? 10 Bxd8 Nxh6 ‘=/+’. Probably not, but the play is balanced (compare 9 Ne2 Bd7 10 Bg5 Ng4 below); for example, 11 Bc7 (11 Bh4?! Nf5 12 Bg3 cxd4 13 cxd4 Nd7!? 14 c3 b6=/+ 15 Bb5 Bb7 16 Ne2 Ke7) 11 ... b6!? 12 Ne2 (12 dxc5 Ba6! 13 cxb6 axb6 14 Bxb6 Bxf1 15 Kxf1 Nd7 and the knights and rooks co-ordinate nicely against Black’s weaknesses; e.g. 16 Bc7 Ke7 17 Bf4 Nf5 18 Ne2 e5 19 Bc1 Nb6 20 Rb1 Nc4 21 Rb4 Nfd6 22 a4 f5 23 f3 f4! 24 fxe4 Rac8 and ... Nxe4) 12 ... Ba6 13 Ng3 Bxf1 14 Rxf1 cxd4 15 cxd4 Nf5 (or 15 ... Nc6 16 0-0-0 Rc8 17 Bf4 Nf5 18 d5 Ncd4=) 16 Nxf5 exf5 17 g3 Nc6 18 0-0-0 Rc8=. (b2) 9 ... Rg6 10 Bxf6 (10 Qh4 cxd4 11 cxd4 Qxd4 12 Rd1 Qc3+ 13 Bd2 Qxa3=/+) 10 ... Qa5!? (10 ... Qxf6 11 Qe3 Bd7 12 Ne2 Bc6 13 Nf4 Rg8) 11 Qe3 Rxf6 12 Ne2 cxd4 13 Nxd4 e5 14 Nb3 Qb6 15 Nd2 Qxe3+ 16 fxe3 Bf5 17 Rb1 Nd7! 18 Rxb7 Rb6 19
Rxb6 axb6=. (c) 9 dxc5 Nbd7 10 Ne2 (10 Bg5!? Ng4! 11 Bxd8 Nxh6 12 Bc7 Nxc5=/+) 10 ... Qc7 (or 10 ... Nxc5=) 11 Nd4 Qxc5=. After 9 Ne2, Black has a variety of satisfactory set-ups, arguably of roughly equal worth. 9 ... Bd7!?
I like this move. It was played quite a bit in the 1950s and 1960s, then largely abandoned in favour of other approaches. Black has the ideas of ... Ba4, ... Bc6, or even ... Bb5 under the right circumstances. With respect to the last move, I should note that the exchange of light-squared bishops is often worth roughly a pawn to Black’s benefit. At any rate, the most common reorganization will be ... Bc6 and ... Nbd7. There are several legitimate alternatives: (a) 9 ... Rg6 10 Qe3 transposes to 8 ... Rg6 9 Qe3 c5 10 Ne2, while 10 Qd2 Nbd7 is a well-known and double-edged line. (b) 9 ... cxd4 is rare but playable: 10 cxd4 (10 Nxd4 e5!? 11 Bb5+ Bd7 12 Bxd7+ Nbxd7 13 Nf5 Rg6 14 Nd6+ Ke7 15 Nf5+ Ke8=; 13 ... Rxg2 14 Nd6+ doesn’t change anything) 10 ... Nc6 11 Bb2 Bd7 12 Ng3 (12 d5 Nxd5 13 0-0-0 Qg5+! 14 Qxg5 Rxg5 15 c4 Nc7 16 Bf6 Rg6 17 Bh4 Na5 18 Nc3 Bc6 19 Bg3 Na6 and Black’s two effective knights on the edge of the board ensure him some advantage) 12 ... Qa5+ 13 Qd2 Qxd2+ 14 Kxd2 0-0-0!? (14 ... Na5 15 Re1 Bc6 16 d5 Nxd5 17 Rxe4 b5) 15 Re1 Ne7 16 Nxe4 (16 c4! Bc6 17 d5 Ne8!? 18 Rxe4 Ng6+/=) 16 ... Nxe4+ 17 Rxe4 Bc6 18 Rh4 Bxg2 19 Bxg2 Rxg2 20 Ke2 Rd5 21 Rxh7 Rf5 22 Rf1 Nd5 23 Bc1 b5 24 Bd2 Kd7= (light-square domination in return for a pawn) 25 Rh8 Kc6 26 Rc8+ Kb7 27 Rh8 Kc6 28 Be1 Nf6 29 Rh1 ½-½ Lu Shanglei-Cao, Kuala Lumpur 2011. (c) 9 ... Nbd7 is an interesting transpositional device.
Often this position arises via 8 ... Nbd7 9 Ne2 c5, but in that case 9 Ne2 is only one of several moves that Black needs to master. This may not be objectively the ideal 4 a3 move order for Black (in terms of theory, that is), but its good practical results and instructive qualities make it worth a look. (c1) 10 g3 b6 11 Bg2, Kir.Georgiev-Psakhis, Sarajevo 1986; and I like 11 ... Ba6! (11 ... Bb7 12 0-0 Rc8! is also equal) 12 Qd2 Qc7 13 0-0 Rc8=. (c2) 10 a4 b6 (10 ... Qc7 may not be best, but comes with the trick 11 dxc5 Qxc5 12 Ba3? Qxf2+!) 11 Ng3! (11 a5 Ba6!; 11 Ba3 Ba6!=/+) 11 ... Bb7 12 a5 (12 Bb5 Rg6 13 Qe3 a6 14 Bxd7+ Qxd7=) 12 ... Rc8 13 d5!? Rg6 14 Qh4 Bxd5 15 Ba6 Rb8 16 c4 Bc6=. (c3) 10 Ng3 Qc7
11 Bb5 (11 Qe3 b6!? 12 Nxe4 Nxe4 13 Qxe4 Bb7 looks a tad better for White with his bishop pair, although his pawns will be weak after ... Bxg2; 11 Bd2 Rg6 12 Qf4 Qxf4 13 Bxf4 Ke7 has the idea of ... Nd5 and ... f5, as well as simply ... b6, and 14 Rd1 Rg4 15 Be3 b6 16 Be2 Rg8 gives equal chances) 11 ... Rg6 12 Qd2 (12 Qf4 Qa5!), Bloch-Huss, Siegen Olympiad 1970, and Black can equalize with the tricky sequence 12 ... cxd4! 13 cxd4 (13 Qxd4! e5= or 13 ... a6 14 Bf1 e5=) 13 ... e3!! 14 Bxd7+ (14 Qb4 exf2+ 15 Kxf2 Qxc2+=/+; 14 fxe3? Rxg3! 15 hxg3 Ne4–+; 14 Qxe3?? Qa5+) 14 ... Bxd7 15 Qxe3 Bb5!=/+ or 15 ... Qxc2 16 0-0 Ng4=/+. 10 Bg5 The approved continuation/solution.
Others:
(a) 10 dxc5 is widely recommended: (a1) Psakhis gives 10 ... Bc6 11 Rb1 (‘!’) 11 ... Nbd7 12 Nd4 Qa5 13 Nb5! ‘with an extremely menacing initiative’; but 13 ... Bxb5 14 Bxb5 (14 Rxb5 Qxc3+ 15 Qd2 Qxd2+ 16 Bxd2 0-0-0) 14 ... Qxc3+! 15 Bd2 Qxc5 16 Qxf6 Qxb5! 17 Qxe6+ fxe6 18 Rxb5 b6 19 0-0 Rc8 poses Black no problems at all; e.g. 20 Rc1 Kf7 21 Rb4 Ne5! 22 Bf4 Nc4 23 Re1 Rgd8 24 g3 Rd4=. (a2) 10 ... Ba4 is forcing and looks reliable: 11 Nd4 (11 Bg5 Rg6 12 Bxf6 Rxh6 13 Bxd8 Kxd8=) 11 ... Nbd7! (11 ... Qa5 is playable) 12 Bb5 (12 Nb5 Nxc5 13 Be3 Nd3+!) 12 ... Bxb5! 13 Nxb5 Rxg2 14 Bg5! (14 Nd6+?! Ke7 with a significant advantage) 14 ... Qa5 15 Rb1 Ke7 16 Bf4 Rc8 (16 ... e5 17 Bg3 a6=) 17 Bd6+ Ke8 18 Qe3 Rg6=/+ intending ... a6. (b) 10 Ng3 Ba4 (hitting d4 and c2; 10 ... Bc6 is also satisfactory) 11 dxc5 Nbd7 (or 11 ... Rg6 12 Qe3 Bxc2 13 Bb5+ Nc6=) 12 Ra2?! (12 Be3?! Qa5; 12 Qe3 Qa5 13 Qd4 Bxc2) 12 ... Bc6 13 Be2 Rg6 14 Qe3 h5! 15 h4 Nd5! 16 Qd4 e5 17 Qc4 e3!-/+. 10 ... Rg6 10 ... Ng4 is more stable and yields equality, but offers fewer positive prospects: 11 Bxd8 Nxh6 12 Bc7 (12 Bh4 Nf5 13 Bg3 Ba4!) 12 ... cxd4 13 cxd4 (13 Nxd4 Nc6 14 0-0-0?! Nxd4 15 cxd4 Rc8 16 Bf4 Nf5=/+; here 14 Nb5 Nf5 15 0-0-0 a6 16 Nd6+ Nxd6 17 Bxd6 Ne7 is equal) 13 ... Nc6 14 Rb1 b6 15 Ng3 (15 c3!? Ne7! 16 c4 Rc8 17 Bf4 Nhf5 18 Ng3 f6 19 Nxe4 Kf7 and Black threatens ... Bc6; he has plenty for a pawn) 15 ... Nxd4 16 Nxe4 Ng4 17 Bd3 Bc6 (or 17 ... f5 18 Nd6+ Ke7 19 c3 Ne5 20 Nxf5+ Nxf5 21 Bxe5 Bc6=) 18 Rb4 Bxe4 19 Bxe4 Rc8 20 Bxb6 Nxc2+ 21 Bxc2 axb6 22 Ba4+ Ke7=. 11 Qh4 11 Bxf6 Rxh6! (giving more winning chances than 11 ... Qxf6 12 Qe3 Bc6) 12 Bxd8 Kxd8 13 g4 (13 dxc5?! Na6) 13 ... cxd4 (or 13 ... Rg6 intending ... Bc6) 14 cxd4 Nc6 15 h4 (15 Bg2 f5=/+) 15 ... e5! 16 g5 Rd6=/+. 11 ... Bc6
12 d5! This line has been given as the reason to reject 9 ... Bd7. Otherwise, 12 Rd1 (12 Nf4? Qa5!) 12 ... Nbd7 13 Nf4 Rxg5! 14 Qxg5 Qa5 15 Ne2 Qxa3 16 dxc5 a5 17 Qd2 a4 18 Nd4 Qxc5 19 Nxc6 bxc6 is at least equal; e.g. 20 Be2 a3 21 0-0 a2 22 Ra1 Nd5 23 c4 e3=/+. 12 ... exd5 13 Nf4 Rxg5! White’s point is that 13 ... Qa5? 14 Bxf6 covers c3. 14 Qxg5 Nbd7 15 Be2 Qa5 Black has a piece and a pawn for the exchange, with a powerful centre and chances to exploit White’s weak pawns. This is enough to claim equality, but no more, since White’s potentially active piece placement keeps it even. 16 Qg3 The alternative 16 0-0 Qxc3 already favours Black. 16 ... 0-0-0 17 0-0 Rg8 18 Qh3 Kb8 Or 18 ... Qc7 19 Nh5 and now 19 ... Kb8 is equal. 19 Nh5 Nxh5 20 Qxh5 20 Bxh5 Ne5 21 Rab1 (21 Qf5 Qxc3 22 Qxh7 Rg7 23 Qh8+ Kc7= with the bizarre threat of 24 ... Nf3+ 25 Kh1 Rxg2!) 21 ... Qd8 22 Qf5 f6=. 20 ... Qxc3 20 ... Rg7 is safer. After 20 ... Qxc3, a sample line would be 21 Qxf7?! (21 Qxh7 Qg7 22 Qxg7 Rxg7 23 f3 e3 24 g4 Nf8! intending ... Ne6 and/or ... f5) 21 ... Qg7! (even 21 ... Rg7 22 Qf4+ Qe5 23 Qxe5+ Nxe5 24 f4 Nd7! 25 g4 d4 is fully equal in spite of White’s dangerous pawns) 22 Qxg7 Rxg7 23 a4 (23 Rab1 d4 24 Bb5 Bd5!) 23 ... Kc7 24 Bb5 Ne5 25 Rfd1 d4 and Black has the upper hand.
7.122
8
...
Rg6
9 Qd2 Instead: (a) 9 Qh4 isn’t mentioned in most sources, perhaps because 9 ... Rg4 10 Qh6 repeats, and 10 Qh3 c5 looks fine, intending ... Nc6 and ... e5. Black has also played 9 ... c5, 9 ... Nbd7, and 9 ... b6. (b) 9 Qe3 is the main alternative:
(b1) 9 ... c5 10 Ne2 (this could also arise via 8 ... c5 9 Ne2 Rg6 10 Qe3) 10 ... Nc6 11 dxc5 (11 Bb2 Qa5 12 0-0-0 cxd4 13 cxd4 Ne7 14 Ng3 Bd7 15 Nxe4 Nxe4 16 Qxe4 Bc6=) 11 ... e5 (11 ... Qa5 12 Bd2!? e5= Moskalenko; there might follow 13 Ng3 Be6 14 a4 0-0-0 15 Bb5 Ne7 16 c4 Qc7=) 12 Ng3 (12 Bb2 Ng4!? 13 Qxe4 Qf6 14 Qf3 Qe7 15 Qg3 Rf6 16 f3 Ne3) 12 ... Be6 13 Bb5 Qa5 14 Bxc6+ bxc6 15 0-0 0-0-0 16 Bb2 Qb5 17 Rab1 Qc4 with a balanced game. (b2) 9 ... Nc6 10 Bb2 (10 Ne2 Ne7 11 Ng3 Bd7 12 c4 Ba4!? 13 Bb2 Nf5 14 Nxf5 exf5 15 d5 Qd6 16 0-0-0 b6 unclear) 10 ... Ne7 (10 ... b6 11 0-0-0 Bb7=) 11 0-0-0 b5!? 12 Qe1 (12 Bxb5+ c6 13 Bf1 Ned5-/+), Hector-L.Hansen, Graested 1990; and 12 ... Rb8 is the easiest way to obtain a favourable position. 9 ... b6 An interesting option is 9 ... Nc6, contemplating ... e5: 10 Ne2 (10 Bb2 e5 11 0-0-0 Be6 12 f3 exd4 13 cxd4 Qd6=, or here 10 ... Qd6) 10 ... b6 11 Nf4 Rg8 12 Bb5 Bb7 13 c4 a6 14 Ba4, Chiburdanidze-Levitina, Volgograd (m/2) 1984; 14 ... Qe7=. 10 Bb2
10 c4 Nc6 11 Bb2 Bb7 12 Ne2 Qd6=. 10 ... Bb7 10 ... Ba6 is a good solution: 11 Ne2 (11 c4?! c5) 11 ... Nd5 12 Nf4 (12 g3 Qd6) 12 ... Nxf4 13 Qxf4 Bxf1 14 Rxf1, EslonG.Garcia, Biel 1984; and simplest is 14 ... Rxg2 15 Qxe4 Qd5 16 Qxh7 Nd7!. 11 0-0-0 Qe7 Or 11 ... Nbd7 12 Ne2 Qe7 13 h3 0-0-0 14 c4 Ne8 15 Nf4 Rg8 16 Qe3 Nd6= Autran-Agdestein, Calvi 2005. 12 c4 Nbd7 13 Ne2 0-0-0 White is underdeveloped and Black’s activity guarantees at least an even game. Dukhin-M.Ustinov, Rybinsk 2001, continued 14 Nc3 (14 Qe1 Ng4! 15 h3 Qh4 16 g3 Qg5+ and ... e3) 14 ... Ng4! 15 Re1, and here 15 ... f5 16 h3 Ngf6 was the clearest way to a small advantage. 7.24
Qg4
The “Blitz Variation”. White tries to save a tempo on 4 a3 by bringing his queen out immediately. A drawback to this strategy is that he hasn’t bolstered his centre (in the Winawer, that often occurs by means of bxc3 in response to ... Bxc3), and Black can therefore counterattack more quickly in that sector. In fact, accurate play by Black makes it difficult for White even to equalize. 4 ... Nf6 5 Qxg7 Rg8 6 Qh6 c5! This move is better than the immediate 6 ... Rg6 7 Qe3 c5, when White can try 8 Bd2, though 8 a3 is more forcing: 8 ... Ba5 (8 ... cxd4 9 Qxd4 Nc6 10 Qd3 Bxc3+ 11 Qxc3 Nxe4 12 Qh8+ Ke7 13 Qxd8+ Kxd8 is unclear) 9 b4!? (9 Ne2) 9 ... cxb4 (9 ... cxd4 10 Qxd4 Bb6 11 Qd2=) 10 axb4 Bxb4 11 e5 Ne4 12 Ne2 with approximate equality. 7 Ne2 Supporting c3 and d4. This is supposed to be the most accurate move, because after 7 ... Rg6 8 Qe3, it transposes to a position that is also reached via 6 ... Rg6 7 Qe3 c5 8 Ne2, when 8 ... Nc6 9 Bd2 is the traditional main line of 4 Qg4. But 7 ... cxd4 disrupts this plan – see below. Here are White’s alternatives: (a) 7 Bg5? Rg6 8 Qh4 cxd4 9 e5 Rxg5! 10 exf6 Qxf6 11 Ne2 Nc6 12 a3 Ba5 13 b4 Nxb4! 14 0-0-0 Nxc2! 15 Nb1 Bd7! 16 Rd3 Ne1 0-1 Crowl-Harris, corr 1938. (b) 7 e5!? is critical: 7 ... cxd4 8 a3 (8 exf6 dxc3 9 b3 Nd7 10 Qxh7?! Nxf6 11 Qd3 Bd7 with a powerful centre and lead in development, Lanin-Sukhov, Eclipse email 2000; but it would be interesting to see 10 Nf3, which seems to be about equal after 10 ... Bf8 11 Qxh7 Nxf6 12 Qh4 or 10 ... Qf6 11 Qxh7 Qh8 12 Qxh8 Rxh8 – probably the latter ultimately favours Black’s central pawn mass over White’s passed h-pawn, but this untested idea may be the only hope for White to salvage equality) 8 ... Bf8 (8 ... Qa5!?) 9 Qxf6 (9 Qh4 dxc3 10 exf6 cxb2 11 Bxb2 Qa5+ 12 c3 Nd7-/+) 9 ... Qxf6 10 exf6 dxc3=/+
11 Ne2 (11 bxc3 Nd7 12 Bb2 Nxf6 13 c4 Bg7=/+ Y.Hernandez-S.Bjerke, Oropesa del Mar 1999) 11 ... Nd7!? (11 ... cxb2 12 Bxb2 b6=/+) 12 Nxc3 Bd6 13 g3 Nxf6 14 Bg2 Bd7 15 0-0 0-0-0 16 Be3 Kb8 17 h3 Bc6 18 Rad1, Antoniewski-Shaked, Zagan 1997; and Black can keep the better chances by 18 ... Be5 19 Bc5 Nd7 20 Bd4 Bc7=/+. (c) 7 a3 Rg6 8 Qe3 (8 Qh4?! Qa5 9 Bd2 cxd4 10 Nb1 Bxd2+ 11 Nxd2 dxe4 12 0-0-0 Qg5-/+ Walters-Sample, corr 1966) 8 ... Ba5!
At this point, three bad mistakes that have happened over the board are: 9 e5? Ng4 10 Qd3 cxd4, 9 Nf3? dxe4 10 Ne5 cxd4 11 Bb5+ Ke7!, and 9 exd5? Nxd5 10 Qd3 cxd4 11 Qxd4 e5! 12 Qd3 Rd6!. So: (c1) 9 Nh3 Nxe4 (9 ... e5!?; 9 ... Nc6 10 Bb5 dxe4-/+) 10 b4 Nxc3 11 Qxc3 cxd4 12 Qxd4 e5!-/+ Plasman-Steenbekkers, Dieren 1998. (c2) 9 Ne2 Nc6 (or 9 ... cxd4 10 Nxd4 Nxe4) 10 Bd2 cxd4 (10 ... Bb6!? 11 Nxd5!? Ng4=/+) 11 Nxd4 Ng4 12 Qd3 transposes to note ‘c33’. (c3) 9 Bd2 Nc6 puts great pressure on White’s centre:
(c31) 10 Nge2 Ng4! (or 10 ... e5!? 11 Nxd5 Bxd2+ 12 Kxd2 cxd4) 11 Qf4? (11 Qd3 transposes to 10 Nf3) 11 ... cxd4 12 Nb5 e5–+ Kolacz-Massarani, IECC 2000. (c32) 10 Bb5!? cxd4 11 Qxd4 Bd7! (‘Black dominates the game from a dynamic point of view’ – Moskalenko) 12 Bxc6 Bxc6 13 0-0-0 (13 exd5 Bb6 14 Qd3 Nxd5 15 Nxd5 Bxd5-/+) 13 ... Bb6 14 Qb4 d4 15 Be3 e5 16 Nf3 Nd7-/+ FoppeHG-ETC_Chess, playchess.com 2004. (c33) 10 Nf3 Ng4 (10 ... cxd4 11 Nxd4 Ng4 12 Qd3 transposes) 11 Qd3 cxd4 12 Nxd4 (12 Nb1 dxe4 13 Qxe4 f5! 14 Qe2 Bb6=/+; 12 Ne2 dxe4 13 Qxe4 f5 14 Qd3 e5-/+ Malm-Romano, IECC email 1999) 12 ... Nxf2 (12 ... Bb6! 13 Nxc6 Bxf2+ 14 Kd1 bxc6 is also good) 13 Kxf2 Nxd4! 14 Ke1 Bb6 15 Ne2 Qh4+ 16 g3 Qxe4 17 Qxe4 dxe4 18 Nxd4 Bxd4 19 Bb5+ Kf8 20 Bb4+ Kg7 with a meaningful advantage. (d) 7 Bd2!? cedes Black the advantage following 7 ... cxd4! 8 e5 (8 Nb5 Ng4! 9 Qxh7 Bxd2+ 10 Kxd2 Nf6!) 8 ... dxc3 9 bxc3 Qc7!=/+.
7 ... cxd4! Here’s another benefit of the early ... c5. Pedersen says that ‘it is obviously not favourable for [Black] to exchange so early on d4’. It’s true that 7 ... cxd4 frees White’s pieces, but it happens that there are concrete tactical reasons to favour the move. Instead, 7 ... Rg6 is roughly equal following 8 Qh4 Rg4 9 Qh3 cxd4 10 e5 dxc3 11 bxc3 with mutual chances. 8 Nxd4 e5 This looks very promising.
9 Bb5+ Bd7 10 Bxd7+ Nbxd7 11 Nf5 d4 12 Ng7+ 12 0-0 Bf8–+. 12 ... Rxg7 13 Qxg7 dxc3 14 0-0 cxb2 15 Bxb2 Qe7 15 ... Qb6! also prepares ... 0-0-0, and targets f2; e.g. 16 Rab1 (16 Qh8+ Bf8 17 Ba3 0-0-0) 16 ... Bf8 17 Qg3 Nxe4! 18 Qd3 Nxf2 19 Rxf2 Bc5 20 Qg3 Bxf2+ 21 Qxf2 Qxf2+ 22 Kxf2 Rc8-/+. 16 f4? 16 f3 0-0-0 17 Rab1!=/+ gives White some counterplay. 16 ... Nxe4 16 ... Bc5+! 17 Kh1 Nxe4. 17 Kh1 0-0-0? 17 ... Bc5! is pretty much winning. 17 ... 0-0-0 is Bartel-Szelag, Brzeg Dolny 2001; and instead of 18 fxe5?, which allows 18 ... Qh4, White could have kept his disadvantage down with 18 Qxh7 Nd2! 19 Rfd1 Qe6-/+. 7.3
4
Bd2
Here again White’s idea is Qg4, and play will resemble the main lines of 4 a3 and 4 Qg4. White develops rapidly and prepares 0-0-0, but he forfeits any pretensions to holding his centre and in fact loses the critical d-pawn. 4 ... dxe4
I won’t analyse alternative solutions, but they are worth mentioning: (a) InPTF3I analysed 4 ... Nc6!?, which can generate original play. (b) 4 ... Nf6 5 e5 Bxc3 6 bxc3 Ne4 7 Qg4 is a MacCutcheon Variation without ... h6 inserted! Now 7 ... Kf8! is the logical move, but unfortunately, I haven’t covered the ... Kf8 MacCutcheon systems in Chapter 15, which would be the appropriate comparison. Instead, 7 ... g6 weakens Black’s dark squares (more so than if ... h6 were in, as in the MacCutcheon). Then 8 Bh6 c5+/= might follow. (c) Finally, Simon Williams recommends 4 ... Ne7, which can hardly be too bad, although 5 Bd3 may lead to a minimal White advantage. 5 Qg4 5 Nxe4? Qxd4 fails to give White enough compensation following 6 Nc3 Nf6 7 Nf3 Qb6. 5 ... Nf6 6 Qxg7 Rg8 7 Qh6 Qxd4 I analysed 7 ... Rg6 at some length in previous editions. It offers prospects of equality or something very close, but has the disadvantage that Black’s positive chances aren’t evident if White knows a little theory, and there are a fair number of early draws. This might be acceptable, but 7 ... Qxd4 wins material, and if White doesn’t react quickly, he himself won’t equalize. 8 0-0-0 This traditional move is aggressive but currently looking shaky. White might want to choose 8 Nge2!? instead. Black has no particular problems after 8 ... Qe5, but the play is balanced.
In practical terms, White has to take a bit more care than his opponent: 9 0-0-0 (9 Bf4 Qf5=/+; then 10 Ng3 Qg6 11 Qxg6 Bxc3+! 12 bxc3 Rxg6 13 Bxc7 Nc6 offers Black some advantage, and Halldorsson-M.Schmitz, Reykjavik 2000, went 10 Bxc7 Qxf2+! 11 Kxf2 Ng4+ 12 Ke1 Nxh6 13 a3 Bxc3+ 14 Nxc3 f5 15 Nb5, when 15 ... a6! would have resulted in advantage) and: (a) 9 ... Nbd7 10 Ng3? (10 Bf4! Qa5=, or 10 ... Qf5 11 h3 Rg6 12 Qh4 Be7=) 10 ... Ng4! 11 Qxh7 Ndf6 12 Qh4 Be7 13 Re1? (13 f4 Qd4=/+) 13 ... Nxf2 14 Bf4 Qa5 (14 ... Nd7!) 15 Rg1 Bd7 16 Be2 0-0-0–+ Kasten-Kirwald, corr 1986; White’s queen is in danger. (b) 9 ... Rg6 10 Qh4 (10 Qf4 Qxf4 11 Bxf4 Ba5 12 h3 Bd7 13 g4! Bc6 14 g5 Nh5 15 Be3 f5!? 16 gxf6 e5 with equal chances) 10 ... Rg4!? (10 ... Be7 11 Bf4 Qa5 12 Kb1 Nc6=) 11 Qh3 Nc6!? (or 11 ... Bd7 12 f3 exf3 13 gxf3 Rg8=) 12 f3! exf3 13 Qxf3, Mineur-Abenius, corr 1988, and the natural 13 ... Bd7= equalizes.
8 ... Bf8 Already Black can force matters with 8 ... Rg6 9 Qf4 (9 Qh4 Rg4 10 Qh6 Rg6 repeats) 9 ... Bd6 10 Nge2?! (10 Qh4 Rg4=; can either side avoid this draw?) 10 ... Bxf4 11 Nxd4 Be5! and Black has somewhat the better of things. Alternatively, 8 ... Ng4!? is an idea of mine that I analysed at lengthPTF3. I believe that it is completely equal, but since what follows is so agreeable for the second player there’s no need to reprise that. 9 Qh4 Others: (a) 9 Qf4 Bd6 10 Nge2 Bxf4 11 Nxd4 Be5=/+ is similar to 8 ... Rg6 in the previous note. (b) 9 Qh3 Qxf2 wins a second pawn, and White has insufficient counterplay: 10 Nge2 Bd7! 11 Be3 (11 Ng3 e3 12 Be1 Qf4 13 Bd3 Qh6–+; 11 g4? Rxg4 12 Ng3 e3 13 Be1 Qf4-/+ Bezgodov-Kornev, Krasnoyarsk 2003) 11 ... Qf5 12 Qh4 Be7 13 Nd4 Qg6-/+. 9 ... Rg4 10 Qh3 Qxf2
11 Be2 The only way to try to justify White’s play to this point. Otherwise: (a) 11 Nb5? Na6 actually helps Black; e.g. 12 Qb3?! (12 Kb1 Bd7 13 Be3 Qf5 14 Nd4 Qg6 15 Nb3 Nb4-/+ BoleslavskyBronstein, Moscow 1950; 12 Be3 Qf5 clearly favours Black, as does 12 ... Qh4 with the idea 13 Qxh4 Rxh4 14 Bg5 Bh6) 12 ... Nd5! 13 Be2?! (13 c4 e3 14 Ba5 Nc5 15 Qa3 a6!; 13 Kb1 Bd7 is also hopeless) 13 ... Rxg2 14 Rf1 Rxg1 15 Rhxg1 Qxe2 0-1 RajnaVasiljevic, Pancevo 1987. (b) 11 Bb5+? c6 12 Rf1 Qxg2 13 Rxf6 cxb5-/+ Mossung-Vrbata, Prague 1989, in view of 14 Qxh7 Rg7 15 Qxe4 Bd7!.
(c) 11 Be3 isn’t so bad, but Black still comes out on top following 11 ... Qh4! 12 Qxh4? (12 g3 Qxh3 13 Nxh3 Rg6 14 Nf2 Bd7 15 Bg2 Bc6=/+ Jansen-Alger, email 1997) 12 ... Rxh4 13 g3 (13 Bg5? Bh6) 13 ... Rh5! 14 Be2 Bh6!–+ Cabanas JiminezMoreno Ruiz, Mostoles 2008.
11 ... Rh4 As I showed inPTF3, 11 ... Qxg2!? 12 Bxg4 Qxg4 13 Qxg4 Nxg4 14 Nxe4 Bd7 or 14 ... Nd7!? is equal. Black has two bishops, material equality and the two unopposed central pawns that he gets in the main line, although his pieces aren’t quite as well placed as in that case. 12 Qxh4! Qxh4 13 g3
Black’s queen is trapped but, as compensation for losing the exchange, he will have the bishop pair and strong central pawns. 13 ... Qh6! After this Black has no problems and should probably be able to find some advantage. InPTF3I looked (at some length) at the alternative 13 ... e3!? (threatening ... Qb4). This doesn’t give as many winning chances, but is good enough to equalize after 14 gxh4 exd2+ 15 Kxd2 Bh6+ or 15 Kb1 Bb4. 14 Bxh6 Bxh6+ 15 Kb1
Black has two pawns for the exchange (passed central pawns) as well as the bishop pair. On the negative side, his advanced e-pawn is vulnerable. Remarkably, no less than four moves are satisfactory here. 15 ... Nc6! ‘!’ for originality. With this almost unknown move, Black mobilizes as quickly as possible; in doing so, he offers his c-pawn (to a knight raid via b5, something prevented by 15 ... Bd7 and 15 ... a6); in return, he mobilizes his centre. According to theory, 15 ... e5 achieves equality, and there are two other ways to demonstrate the bishops’ effectiveness: (a) InPTF3I recommended the little-known 15 ... a6, which might already favour Black:
Black plans to develop quickly by ... b5 and ... Bb7, with ... b4 and ... Nbd7-e5 as supplementary ideas: (a1) 16 Nh3 b5 17 Rhf1 Nbd7 18 g4 b4 19 g5 (19 Na4 Be3 20 g5 Nd5) 19 ... Bxg5 20 Nxg5 bxc3 21 Rf4, and Wei Ming suggests 21 ... Bb7! 22 Nxf7! Nd5 23 Rf2 cxb2 24 Kxb2 e3 25 Rff1 Rb8! ‘with a strong initiative for Black’. (a2) 16 h4 b5 17 Nh3 (17 g4 b4 18 Na4 Be3 19 g5 Bd7 20 gxf6 Bxa4 21 Nh3 Nd7=/+) 17 ... b4 18 Rdf1 (18 Na4 Be3! 19 Rhf1 Ke7-/+; 18 Rhf1 Nbd7 19 Na4 Bf8 20 Nf4 Bb7=/+) 18 ... bxc3 19 Rxf6 Bg7 20 Rff1 f5 21 Bh5+ Ke7 22 g4 (22 b3 Nd7-/+) 22 ... cxb2 23 gxf5 exf5-/+ Korepanov-Skomorokhin, Podolsk 1993. (b) 15 ... Bd7 is at least equal after 16 Nh3! Be3! 17 Rhf1 Ke7 18 Nf4 (18 Nf2 Bc6 19 Bb5 Bd5! 20 Be2 Nbd7 21 Nb5 c6 22 Na3 b5 23 c4 bxc4 24 Nxc4 Bh6-/+ Lindquist-J.Watson, Long Beach 1996) 18 ... Be8!? (18 ... Bc6!=/+) 19 g4 h6 20 h4 Bc6= 21 g5 hxg5 22 hxg5 Ne8 23 Bh5?! Nd6=/+ Rasmussen-Czibulka, corr 1990.
16 Nh3 Trying to attack with Nf2 or block with Nf4. Others are worse: (a) 16 Nb5 Ke7 17 Nxc7 Rb8 leaves White trying to get his pieces developed, while ... a6 trapping the knight on c7 is also a concern; e.g. 18 Nb5 (instead 18 Nh3 is met by 18 ... e5 19 Ng1 Nd4 20 Nb5 Nxb5 21 Bxb5 Ng4!-/+) 18 ... Nd5-/+ 19 c3 f5 20 Na3 Ne3 21 Re1 e5 22 Nc2 Nxc2 23 Kxc2 Be6 24 b3 b5. (b) 16 Bb5 Ke7 17 Bxc6 bxc6 18 Nge2 e5 19 h3 (versus ... Bg4 or ... Bh3) 19 ... Be3 20 g4 h5! 21 gxh5 Bf2!-/+ with moves like ... Ba6, ... e3, ... Nxh5 and/or ... Rh8 to follow. 16 ... e5 17 Nf2 Nd4! 18 Nfxe4! Destroying Black’s centre. Otherwise the pressure mounts; e.g. 18 Rhe1 e3 19 Nd3 Bf5 or 18 Rhf1 Nxe2 19 Nxe2 e3 20 Nd3 Ne4. 18 ... Nxe4 19 Nxe4 Nxe2 20 Rhe1 Nd4 20 ... Kf8 was also possible: 21 Rxe2 Bg4 22 Ree1 Bxd1 23 Rxd1 Ke7. After 20 ... Nd4, Z.Ivekovic-Sa.Martinovic, Zagreb 2011, continued 21 Nf6+ Kf8 22 Rxe5 Ne6 23 Rf1 Bg7 24 Nxh7+ Kg8 25 Nf6+ Bxf6 26 Rxf6 Bd7 27 h4 Bc6 28 b3 Rd8 29 Ref5 Be8 30 Kc1 Rd4 31 Rf3 Rg4 and eventually Black prevailed. 15 ... Nc6 might be the final blow to White in the main line. Apparently 8 Nge2 is the best chance. 7.4
4
Ne2
This is still a popular alternative to 4 e5, mainly because White feels that it’s a safe move. The avoidance of doubled c-pawns has its price, however, in that White confines his own development and allows his centre to be challenged quickly. 7.41 7.42
4 4
... ...
dxe4 Nf6
There’s not enough room to cover three separate systems versus 4 Ne2, especially because Black seems to be doing well with most of what he plays. I think it’s fair to say thatPlay the Frenchhas had quite a bit to do with popularizing 4 ... Nc6, which was quite obscure when I first started analysing it in depth. Now it’s a thoroughly mainstream system with top players using it and theoreticians recommending it. Furthermore, none of the important subvariations seem to be threatened with refutation; so if you’ve already got 4 ... Nc6 as a weapon I’d recommend that you keep it. Not surprisingly, however, the move is being analysed in extraordinary detail, and quite a few variations hang upon tactical points, which means that it takes quite a bit of study to keep up. I’m presenting two other systems, the old and established 4 ... dxe4 5 a3 Bxc3+ lines, and the fresh 4 ... Nf6, which has remarkably little theory attached to it and yet has already established itself as a sound system at every level of play. It’s an ideal line in the sense of being relatively easy to learn and replete with characteristic French themes. 7.41
4
...
dxe4
5
a3
Bxc3+
6
Nxc3
Nc6
7 Bb5 7 d5 is the main alternative here, but a couple of interesting older moves have their virtues and also deserve to be mentioned: (a) 7 d5 exd5 8 Qxd5 seems to appeal to White on the grounds that he has little risk of losing.
On the other hand, the chances of a draw rise, which may not be a desirable situation for White, and in the main line he pays for his bishop pair by falling behind in development. With plenty of pieces left on the board, there are enough imbalances to keep the play interesting: 8 ... Be6 (8 ... Nge7 also equalizes) 9 Qxe4 (9 Qxd8+ Rxd8 10 Nxe4 Nd4 11 Bd3 Bc4! is strong: I haven’t seen 12 Bxc4!? Nxc2+ 13 Ke2 Nxa1 analysed before, but it looks like Black gets out after the tricky 14 Bf4 Nc2 15 Bxc7 Nd4+ 16 Ke3 Rd7 17 Be5 Nh6!-/+; in this line, Bronstein once played 10 Bf4, but I think that 10 ... f5 favours Black; e.g. 11 Nb5 Nge7 12 Nxc7+ Kf7 13 Nxe6 Kxe6 14 Bc4+ Kf6) 9 ... Nf6 10 Qh4 (10 Qf4 Nd5!? 11 Nxd5 Qxd5 could favour Black’s development, so White might want to bail out by 12 Bd3 Qxg2 13 Be4 Qh3 14 Qxc7 0-0 15 Be3 Rad8 16 Qg3 Qh5 17 Qg5, while after 13 ... Qg4 14 Qxg4 Bxg4 15 Bxc6+ the extra pawn won’t mean anything; 10 Qa4 0-0 11 Bd3 Ne5 12 Be2 Qe8!, or here 12 ... Re8 13 0-0 Nd5=) 10 ... Qe7 (or 10 ... Bf5) 11 Be2 (11 Be3 0-0-0) 11 ... 0-0-0 12 Be3 Nd5 13 Qxe7 Ndxe7= intending ... Nf5, GaponenkoKononenko, Antalya 2002. These positions, by the way, lead to decisive results more often than you might imagine, in large part due to the minor piece imbalance. (b) 7 Bf4!? is an old gambit try which has fallen off the radar. Now 7 ... Qxd4? 8 Nb5 and 7 ... Nxd4? 8 Be5 are bad. InPTF2I analysed 7 ... Nge7, which leads to equality with best play. Black’s most ambitious move is 7 ... Nf6 8 Bb5 Nd5 (8 ... Bd7 9 d5 Nxd5 10 Nxd5 exd5 11 Qxd5 0-0 12 0-0-0 Qf6!=; 8 ... 0-0 9 Bxc6 bxc6 is also unclear) 9 Bd2 (9 Bxc6+ bxc6 10 Qd2!? 0-0 11 0-0-0 Nxf4 12 Qxf4 f5 with complex play) 9 ... Nxc3 10 Bxc3 0-0 11 0-0 (11 Bxc6 bxc6 12 0-0 f5 13 Qd2 Rf6) 11 ... Qd5
12 Ba4 (12 a4 Rd8) 12 ... Rd8 13 Bxc6 Qxc6 14 Qe2 b6 15 Rad1?! (15 Rfe1 Bb7 16 Qg4 Qd5 17 Re3 f5=/+) 15 ... Bb7 16 Qe3 Rd7 17 Qg3 f5 18 Rfe1 Rf8-/+ 19 Rd2 f4 20 Qh3 Ba6 21 Qg4 Qd5 (21 ... e3!) 22 b3 Rf5 23 h4 h5 24 Qd1 e3 0-1 LenksasRomsdal, ICCF 1990; 25 fxe3 fxe3 26 Rxe3 Rdf7 can follow.
(c) 7 Be3 isn’t a bad gambit; White counts upon his two bishops to give him chances: 7 ... Nf6 8 Qd2 (8 Bb5?! Bd7 9 Qd2 h6 10 0-0-0 Ne7! 11 Bc4 Bc6=/+ Westerinen-Agdestein, Gjovik 1985) and several moves are playable, including 8 ... b6 and 8 ... 0-0, but a simple solution is 8 ... Ne7 9 Bg5 Ned5 10 0-0-0 0-0 (10 ... Bd7 11 f3 Nxc3 12 Qxc3 Nd5 13 Qd2 e3= Kotronias-Foisor, Moscow 1987) 11 Bc4 Nxc3 12 Qxc3 Nd5!? 13 Qg3 Qd6 14 Qh4 Rb8?! (14 ... b5! 15 Bxd5 Qxd5=/+) 15 Qxe4 b5 16 Bd3 f5 17 Qe5 Qxe5= Van Mil-Jolles, Dieren 1990. 7 ... Ne7
8 Bg5 White wants to provoke a weakness. (a) 8 0-0?! is too slow; e.g. 8 ... 0-0 9 Bxc6 Nxc6 10 d5 exd5 11 Qxd5 Nd4! 12 Bg5 Qxd5 13 Nxd5 Ne6=/+ Em.Lasker-Kan, Moscow 1936. (b) 8 Nxe4 0-0 (8 ... Qd5 9 Qd3!? results in unclear complications following 9 ... f5 10 Ng3 Qxg2 11 Bg5) 9 c3 (9 Bxc6 Nxc6 10 c3 e5 11 d5 Ne7 12 c4 f5 13 Nc3 f4!=/+; 9 Be3?! f5 10 Ng5 f4 11 Bd2 Qd5 12 Bxc6 Nxc6-/+ Thorhallsen-Moskalenko, Copenhagen 1995) 9 ... e5 (9 ... Qd5!? 10 Bd3 e5 11 Qh5! Bf5 12 Nf6+ gxf6 13 Bxf5 Nxf5 14 Qg4+! Kh8 15 Qxf5 Qe6=) 10 Bxc6 (10 dxe5 Qxd1+ 11 Kxd1 Nxe5= Bronstein-Bondarevsky, USSR Ch., Leningrad 1977) 10 ... Nxc6 11 d5 Ne7 12 c4 (12 d6 Nf5 13 dxc7 Qxc7=/+) 12 ... f5 (12 ... b5!? 13 b3 bxc4 14 bxc4 Nf5=; 12 ... Bf5 13 Ng3 Qd6) 13 Nc3 f4 14 f3 Nf5 with a dynamically balanced position, Horvath-Knaak, Trnava 1981. (c) 8 Be3 is similar to the main line: 8 ... 0-0 9 Qd2 e5 (9 ... f5 is equivalent to 8 Bg5 f6 9 Be3 0-0 10 Qd2 f5) 10 d5 (10 dxe5 Qxd2+ 11 Bxd2 Nxe5 12 Nxe4 Bf5 13 f3 Bxe4 14 fxe4 a6 15 Be2 N7c6 16 0-0-0 Nd4= Mokry-Knaak, Bratislava 1983) 10 ... Nd4! 11 Bc4 (11 Bxd4 exd4 12 Qxd4 c6! 13 Bc4 Nf5! 14 Qc5 b6! 15 Qxc6 Bd7 16 Qb7 Nd6-/+ – Uhlmann) 11 ... Nef5 12 0-0-0 (12 Nxe4? Nxe3 13 fxe3 Qh4+ 14 Nf2 Nf5=/+ – Psakhis) 12 ... Nd6 13 Ba2 N4f5 14 h3 Nxe3 15 Qxe3 f5 16 f4, Garcia MartinezUhlmann, Leipzig 1983; and either 16 ... Qf6 or 16 ... exf4 17 Qxf4 Qf6 favours Black’s superior minor pieces. 8 ... f6 9 Be3 0-0 10 Qd2
10 ... e5 This move has the advantage of clarifying the central situation and maintaining a definite imbalance. On the other hand it is rather complex, so if you simply want an unambitious equality one of these might suffice: (a) 10 ... a6 forces a satisfactory simplification: 11 Bxc6 Nxc6 12 0-0-0 b6 (12 ... b5!? 13 Nxe4 aims at the newly-weakened c5-square; then 13 ... Qd5 14 Nc3!? Qxg2 15 Rhg1 is interesting and dynamically equal) 13 Nxe4 Bb7 14 f3 Qd7 15 Rhe1 Rad8= Hector-Casper, German League 2002. Several later games followed this course and confirmed the assessment. (b) I gave 10 ... f5 in earlier editions.
It is quite as good and worth playing, but gives White some attack if Black isn’t careful. For reasons of space, I’ll merely give an analytical update which should cover the essentials: (b1) 11 f3 exf3 (11 ... Nd5 12 Bxc6 bxc6 13 fxe4 fxe4 14 0-0-0 Nxc3 15 Qxc3 Qd5=) 12 gxf3 (12 0-0-0 fxg2 and 13 Qxg2 f4 14 Bf2 f3 or 13 Rhg1 Nd5 14 Qxg2 g6 15 Bg5 Qd6) 12 ... e5 13 d5?! (13 Bc4+ Kh8 14 d5 f4 15 Bc5 b6!? 16 0-0-0 Na5 17 Bxe7 Qxe7 18 Bd3 Bf5 19 Rhe1 Nb7) 13 ... f4! 14 Bc5 Nd4! 15 0-0-0 (15 Bxd4 exd4 16 Qxd4 Nf5 and White won’t get castled) 15 ... Nef5 16 Qf2 (16 Bxf8 Qxf8 17 Qf2 Ne3=/+), Rogulj-Lindgren, Pula 2011, and the strongest course was 16 ... Rf7! 17 Bc4 Ne3-/+. (b2) 11 0-0-0 a6 12 Bxc6 Nxc6 13 Bg5 Qd7 14 d5 (14 Bf4 Ne7 15 f3 exf3 16 gxf3 b5 17 Rhg1 Bb7=/+ 18 Rg3 Ng6 19 Rdg1 Rf7=/+ V.Gurevich-Dimitrov, Werfen 1990; 14 f3 exf3 15 gxf3 e5 16 dxe5! Qxd2+ 17 Bxd2 Be6=) 14 ... Ne5 (14 ... Ne7 15 dxe6 Qxe6 16 f3 h6 17 Bxe7!? Qxe7 18 fxe4 fxe4 19 Rhe1 Be6= Westerinen-I.Thompson, Gausdal 2006) 15 f3 Nc4 16 Qe2 b5 17 fxe4 Bb7 18 exf5 exd5 ‘with double-edged play’ (Vitiugov); Black has somewhat the better of it. 11 d5
Others: (a) 11 Bc4+ Kh8 12 d5 Nd4 transposes to 11 d5 Nd4 12 Bc4 Kh8 without allowing Black his option 12 ... Nef5. Instead, 12 ... Na5 13 Ba2 b6 is possible, intending the idea ... Nb7-d6. (b) 11 dxe5 is consistent, but harmless at best: 11 ... Qxd2+ 12 Bxd2 Nxe5 13 0-0-0 (13 Nxe4 Bf5 14 f3 Bxe4 15 fxe4 Nc8! 16 0-0-0 Nd6 17 Bd3 Rfe8=/+ Gipslis-Knaak, Berlin 1988) 13 ... f5!
(13 ... Bf5=) 14 Bg5 N7g6 15 Nd5?! (15 h4! c6 16 Be2 Nf7 17 Be3 Be6=/+) 15 ... c6 16 Nc7 Rb8 17 Ba4, Kovalev-Ulibin, Simferopol 1988; and Black can best secure his advantage by 17 ... h6 (17 ... Ng4 18 Bb3+ Kh8 19 Rd2 f4 20 h4) 18 Be3 (18 Bd2 b6 19 Bb4 c5 20 Bc3 Rb7 21 Nb5 Be6 22 Nd6 Re7=/+) 18 ... b6 19 Rd6 (19 g3 Rb7 20 Bb3+ Kh7 21 Ne6 Bxe6 22 Bxe6 f4 23 gxf4 Nxf4-/+) 19 ... f4 20 Bd4 c5 21 Bb3+ Kh7 22 Bxe5 Nxe5 23 Bd5 Bf5=/+ with ideas of ... f3 and/or ... Nf7. 11 ... Nd4
A 7.411 7.412
crucial
juncture.
White 12 12
has
two
main
moves: Bc4 Bxd4
7.411 12 Bc4Kh8 Alternatively, 12 ... Nef5 leads to equality after 13 Nxe4 (13 0-0-0 Nd6=/+) 13 ... Nxe3 14 fxe3 Nf5 15 d6+ Kh8 16 0-0-0
Nxd6 17 Nxd6 cxd6 18 Qxd6 Bg4 19 Rd2 Qxd6 20 Rxd6 Rfd8. 13 0-0-0!? 13 0-0 is probably a safer choice, after which 13 ... c5!? reinforces Black’s central control; e.g. 14 dxc6 (14 Nxe4 Nxd5 15 c3 Nxe3 16 fxe3 Be6!) 14 ... Nexc6 15 Nxe4 f5!? 16 Ng5 f4 17 Bxd4 Qxg5 (17 ... Nxd4 18 Qb4! Bf5!=) 18 Bc3 Bf5 19 f3 Rad8=. 13 ... Bg4! 14 Rde1 c5 15 dxc6 Nexc6 16 Nxe4 White regains his material but is exposed to a queenside attack.
16 ... Rc8 Alternatively: (a) 16 ... Bf5 17 f3 b5! 18 Ba2 (18 Bd3 a5 and ... b4) 18 ... Qc7 (18 ... a5? 19 c3) 19 c3 b4! 20 cxd4 Nxd4+ 21 Kb1 Bxe4+ 22 fxe4 b3 23 Bxd4 bxa2+ 24 Ka1 (24 Kxa2?! Rad8 with the idea 25 Qc3 Qf7+) 24 ... exd4 (24 ... Rad8?? 25 Qc3) 25 Qxd4 Rad8 26 Qe3 Rfe8 27 Rc1 Qe5 28 Kxa2 Rd4 29 Rhe1 a5!?; this should be drawn fairly easily. (b) 16 ... Qc7 17 Bd3 f5 (17 ... Rad8 18 f3 Be6) 18 Nc3 f4 19 Bxd4 Nxd4=/+. 17 Kb1 Nxc2!? 17 ... Na5 was better; e.g. 18 Bd3 Nc4 19 Bxc4 Rxc4 20 f3 Bf5 21 Rc1 Qd5! 22 Nc3 Qc6 with the ideas of ... Nb3 and ... Rc8, so play might go 23 Bxd4 exd4 24 Ne2 Rd8 25 Nf4 Rc8 26 Nd3 a5 with ongoing pressure. 18 Kxc2 Na5 19 Qxd8? It was easy to miss 19 b3!! Nxc4 20 bxc4 Rxc4+ 21 Kb2 Rxe4 22 Qxd8 Rxd8 23 f3 and Black has a nice structure following 23 ... Rxe3 24 Rxe3 Bf5, with two pawns for the exchange, but White’s rook will be active and this should be drawn. 19 ... Rxc4+ 20 Kb1 Rxd8 21 Bxa7 Wallner-Weinzettl, Vienna 2011; and here Black should play 21 ... Bf5! 22 f3 Nc6 23 Rd1 (23 Bb6 Rd2) 23 ... Rxd1+ 24 Rxd1 Kg8 25 Bb6 Bxe4+ 26 fxe4 Rxe4 27 Rd7 Re1+ 28 Ka2 e4! with good winning chances. 7.412 12 Bxd4 exd4 13 Qxd4 Black settles in on the d6 blockade square. If White grabs a pawn he will have to dance around with his king.
Nf5
14 Qxe4 Others: (a) 14 Qb4 Nd6 15 0-0-0 Bf5!? (Finkel suggests 15 ... f5, which secures the centre in preparation for a queenside attack, when Black has equal prospects; 15 ... a6 16 Be2 Qe7 is also sound) 16 h4 a6 17 Be2 Qd7!? 18 h5 (18 Rhg1!? Qe8 19 g4 Bd7 20 Kb1) 18 ... h6 19 Rd2 Rfe8 20 Nd1 Bg4 21 Bxg4 Qxg4 22 Ne3 Qg5 with an equal game, Hertneck-Uhlmann, Austrian Team Ch. 2000. (b) 14 Qd2 Nd6 15 0-0-0 Bd7 (15 ... f5 16 Rhe1 a6 17 Bf1 Qh4 18 h3 Bd7 19 g3 Qf6= Lebredo Zarragoi-Vilela de Acuna, Bayamo 1984) 16 Bxd7 Qxd7 17 Qf4 f5 18 h4 b5! 19 h5 a5 20 h6 g6 21 Rh3 b4 22 Nb1 Rab8 with good attacking chances, Nordsieck-B.Schmidt, German League 2007. 14 ... Nd6 15 Qb4 15 Qa4 Nxb5 transposes. Not 15 Qf3?! Nxb5 16 Nxb5 Re8+ 17 Kf1 Bd7! 18 Nd4 Qe7=/+ intending ... Qc5/ ... Re5. 15 ... Nxb5 16 Qxb5 Re8+ 16 ... c6! may be more accurate, when 18 Qc5 Re8+ 19 Kd2 transposes to the main line, and 17 Qb3 Qe7+ 18 Kd2 Qe5! gives compensation after 19 g3 Qd4+ 20 Kc1 Re8 or 19 Rae1 Qf4+ 20 Kd1 Rd8. 17 Kd2 17 Kf1 has been played from time to time, but it should be met by 17 ... c6! with the idea 18 dxc6 bxc6 19 Qxc6 Rb8-/+. 17 ... c618 Qb3 18 Qc5 has also been played: 18 ... Be6 (after 18 ... cxd5 19 Rad1 Be6 20 Kc1 Qd7, 21 Rd2 has been given ‘!’, but perhaps simply 21 ... Qc6! is okay; e.g. 22 Qd4 Rad8, and 23 Qxa7 d4 or 23 Rhd1 b5 24 Kb1 a5= with complex play) 19 d6 b6 20 Qd4 c5 21 Qf4 Rb8!?. This was used twice by Naumkin, planning a direct assault by ... b5-b4. Zlochevskij-Naumkin, Moscow 2002, continued 22 Rad1 b5 23 b4, and here I think 23 ... cxb4 24 axb4 Rc8! was the most accurate sequence, because then 25 Nxb5?! Rc4! 26 Nd4 Qb6 27 c3 Rxb4! breaks through, in view of 28 cxb4 Qxb4+ 29 Ke2! Bc8+ 30 Kf3 Qa3+ 31 Qe3 Bb7+ 32 Kg3 Rxe3+ 33 fxe3 Qxd6+.
18 ... Qd6 This is untried; to begin with, it threatens ... Qf4+. Instead, 18 ... Be6 is unambitious but adequate. Sarhisov-Tscharotschkin, Neuhausen 2007, continued 19 Rae1!?, when 19 ... Qd7 20 Kc1 cxd5 is equal, and Finkel gives the simpler 19 ... cxd5 20 Kc1 Qd6!. White should prefer 19 Rad1; nevertheless, 19 ... Bxd5 20 Nxd5 Qxd5+ 21 Qxd5+ cxd5 22 Rhe1 Kf7 will draw without much effort. 19 Qb4! c5 A promising pawn sacrifice; for example, 20 Qh4 (20 Qc4 Bd7!? intending ... b5; e.g. 21 Rae1 b5 22 Rxe8+ Rxe8 23 Nxb5 Bxb5 24 Qxb5 Qf4+ 25 Kd1 Re5! 26 Qxc5! Qg5 27 Qc4 Rxd5+ 28 Ke2 Qe5+ 29 Kf1 Qe6 30 Qe2 Qd7 with an unclear attack) 20 ... b5! 21 Rae1! Bb7! and either the pawn falls with complete equality, or something like 22 Nxb5 Qxd5+ 23 Kc1 Qa2 24 Nd6! Red8 25 Qc4+ Qxc4 26 Nxc4 Bxg2 27 Rhg1 Bd5 levels things. 7.42
4
...
Nf6
Black simply attacks the centre and renews his threat on e4. This solution has only recently gotten much attention but is holding its own and is a little more flexible than 4 ... dxe4 or 4 ... Nc6. I’ll just give an overview with some exemplary variations. 5 e5 The critical move. Alternatively, 5 Bg5 is a transposition to the MacCutcheon variation 3 Nc3 Nf6 4 Bg5 Bb4 5 Ne2 in Chapter 14 (14.2), which is perfectly satisfactory for Black. White doesn’t obtain much for a pawn after 5 a3?! Bxc3+ 6 Nxc3
Nxe4 7 Nxe4 dxe4 8 Qg4 (trying to achieve activity; otherwise 8 Be3 0-0=/+ – Moskalenko, or 8 Bc4 Nc6 9 c3 e5 10 d5 Ne7 11 0-0 0-0 12 Qe2 Bf5=/+) 8 ... Qxd4 9 c3 Qf6 10 Qxe4 Nc6 11 Be3 e5 12 Bd3 Be6=/+ Holzhaeuer-Renner, German League 2001. 5 ... Nfd7
The MacCutcheon-like 5 ... Ne4 is also played, but the text is more consistent with our normal ... Nfd7/ ... c5 set-up in the Tarrasch. 6 Qd3 Others: (a) 6 a3 Be7 prepares the standard counterattack with ... c5: (a1) 7 f4 0-0 8 Be3 (8 b4 a5 9 b5 c5 10 Be3 cxd4 11 Nxd4 Nc5=; 8 g3 c5 9 Be3 Nc6 10 Bg2 b6 11 0-0 Ba6 with central pressure, Asis-Moskalenko, Montcada 2006) 8 ... c5 (8 ... a6!? 9 g3 c5 10 Bg2 Nc6 11 0-0 Qc7 12 dxc5 Bxc5 13 Bxc5 Nxc5 14 Nd4 Bd7=) 9 dxc5 Bxc5 10 Qd2 Nc6 11 0-0-0 a6=. (a2) 7 Nf4 c5 8 Nh5 g6 (an alternative is to allow 8 ... 0-0 9 Bh6!, when one safe solution is the standard exchange sacrifice 9 ... g6!?; e.g. 10 Bxf8 Bxf8 11 Nf4 cxd4 12 Qxd4 Nc6 13 Qd2 Ndxe5 with more than enough for the exchange, or 10 dxc5 Nxe5 11 Bxf8 Bxf8 12 f4 Ned7 13 Ng3 Nxc5 with the same verdict) 9 Ng7+? Kf8 10 Bh6 Kg8 11 h4, Fontaine-Lputian, playchess.com 2004; and the simplest solution is 11 ... cxd4!. (b) 6 f4 0-0 7 g3 (7 a3 Be7 transposes to ‘a1’) 7 ... c5 8 a3 Bxc3+ 9 Nxc3 cxd4 10 Qxd4 Nc6 11 Qf2 f6 12 exf6 Nxf6 13 Bg2 d4 14 Ne4 Nxe4 15 Bxe4 e5 16 Qg2 Bf5=. 6 ... 0-0
7 a3 7 Qg3 Kh8 8 Nf4 Nc6!? (8 ... c5 9 a3 Ba5! 10 dxc5 Nc6=/+) 9 a3? (9 Qd3) 9 ... Nxd4 10 axb4 Nxc2+ 11 Kd1 Nxa1 12 Nh5 Rg8 and White’s attack was insufficient in Volokitin-Gasanov, Kharkov 2001. 7 ... Be7 8 Qg3 Kh8 Moskalenko gives the crazy line 8 ... Bh4 9 Qg4 f5 10 Qh5 c5 11 Nf4 cxd4 12 Nxe6 g6 ‘with an extremely unbalanced position’. I guess 13 Qd1 Qe7 14 Nxf8 Nxe5 15 Ne2 Nbc6 would follow, when 16 Bf4 likely favours White, but there are many possibilities. 9 Be3 c5 10 0-0-0 Nc6 11 f4 b5!
12 Nxb5 12 dxc5 b4 13 axb4 Nxb4 (intending ... Qa5, ... Rb8) 14 f5! Bxc5 (14 ... Qa5! 15 Kb1 exf5 16 c6 Nc5! – Moskalenko) 15 fxe6 fxe6 16 Nd4 Bxd4 17 Bxd4?! (17 Rxd4 Qa5 18 Kb1 Rb8 19 Rh4!?) 17 ... Qa5 18 Be2? (18 Bb5! Rb8 19 Rhf1) 18 ... Ba6 (18 ... Qa1+ 19 Kd2 Qxb2=/+) 19 Bg4?? (19 Bxa6 Qxa6 20 Kb1 Rac8) 19 ... Qa1+ 20 Nb1 Rac8 21 c3 Na2+ 22 Kc2 Rb8 23 b4 Nxb4+ 24 cxb4 Rfc8+ 25 Bc5 Qa2+ 0-1 De la Riva Aguado-Moskalenko, Sitges 2006. 12 ... Rb8 13 Nec3 13 Nbc3 c4!. 13 ... c4! R.Perez-Rustemov, Villa de Albox 2002. The game continued 14 Be2?!, when 14 ... Qb6!-/+ followed by ... a6 is easy enough. Rustemov analyses the better 14 f5 Ba6! 15 Nd6 Bxd6 16 exd6 Qb6 17 Na4 Qa5 18 Nc5 (18 Nc3 Rxb2! 19 Kxb2 Rb8+) 18 ... c3. This should favour Black too; e.g. 19 bxc3 Qxc3 20 Nxa6 Rb2 21 Rd2 Qxa3 22 Rd3 (22 Rf2 Ra2+ 23 Kd2 Ra1–+) 22 ... Qa1+ 23 Kd2 Qxa6-/+. 7.5 4 Bd3 This lacks punch and is seldom seen at the top levels any more; nevertheless, strong grandmasters have used it and the move has generated various theoretical articles. I haven’t revised much here, but you still need to know some lines. 4 ... dxe4 4 ... c5 also gives Black equality:
(a) 5 a3 Bxc3+ 6 bxc3 c4!? (or 6 ... Qa5 with the idea 7 exd5 c4! 8 Be4 exd5 9 Bf3 Nf6) 7 Be2 Nf6 8 e5 Ne4 9 Bb2, Akhmedov-Fatullaev, Baku 2000, 9 ... f6 (9 ... 0-0 10 Nh3 f6!? with the idea 11 f3 fxe5 12 fxe4 Qh4+ 13 Kd2 Qh6+=) 10 f3 Ng5 11 f4 Ne4 12 Nf3 0-0 with a dynamically balanced game. (b) 5 dxc5 dxe4 6 Bxe4 Qxd1+ 7 Kxd1 Nf6 8 Bf3 Bxc5 looks harmless, but it’s not so easy for White to equalize completely; e.g. 9 Be3 (this seems best; 9 Ke2 Nc6 10 Be3 Nd4+ 11 Kd2 0-0 12 Rd1 Nxf3+ 13 Nxf3 Bb4=/+) 9 ... Nbd7 10 Bxc5 Nxc5 11 Nge2 Bd7 12 Kc1 0-0-0 and the position is equal but a tad easier for Black to play. (c) 5 exd5exd5 (Simon Williams recommends 5 ... Qxd5 6 Bd2 Bxc3 7 Bxc3 cxd4 8 Bxd4 e5=) 6 dxc5 (6 Nf3 c4! 7 Be2 Ne7=/+; 6 Bb5+ Nc6 7 Ne2 can be answered by 7 ... c4, 7 ... Nf6, or 7 ... Ne7 with the idea 8 a3 Bxc3+ 9 Nxc3 0-0 10 dxc5 d4 and, if the knight moves, 11 ... Qa5) 6 ... Nc6 (6 ... Nf6 7 Nf3 0-0 8 0-0 Bxc5 is approximately equal) 7 Nf3 (7 Bd2 Bxc5 8 Nf3 Nge7 9 00 0-0= Bilek-Osvath, Budapest 1954) 7 ... Bxc5 8 0-0 Nge7 (8 ... Nf6 9 Re1+ Be6 10 Ng5!? 0-0!) 9 h3 0-0 10 a3, Sariego-Paneque, Pinar del Rio 1990. Formally, White has a slight theoretical edge in this IQP position, but Black has an active game and good practical prospects after, say, 10 ... a6. 5 Bxe4 Nf6
The immediate 5 ... c5 is a good alternative (compare the main game); e.g. 6 a3 Bxc3+ 7 bxc3 Nf6 8 Bf3 (8 Bd3 Qc7! 9 dxc5 Nbd7 10 Ne2 Nxc5 11 f4 Bd7 12 0-0 Na4!=/+ intending ... Qc5+, ... 0-0, Bungo-Pederson, corr 1986; or here 8 ... Nbd7 9 Nf3 Qc7! 10 0-0 c4 11 Be2 Nd5=/+ Barlov-Sahovic, Vrnjacka Banja 1984) 8 ... Nc6 (8 ... Qc7 is a useful way to avoid simplification; e.g. 9 dxc5?! Qxc5 10 Qd4, Petkovic-Gavric, Yugoslavia 1991, and 10 ... Qa5 is the best way to play for dynamic prospects) 9 Ne2 e5 10 Bg5 (10 Qd3 cxd4 11 cxd4 exd4 12 Bxc6+ bxc6 13 Nxd4 0-0 14 0-0 c5 15 Nb3 Qb6=)
10 ... exd4 11 cxd4 (11 Bxc6+ bxc6 12 cxd4 Ba6 13 0-0 0-0=/+) 11 ... h6 12 Bxf6 Qxf6 13 c3 0-0 14 0-0 Rd8 15 Qa4 Bd7!?=/+ Ortega-Uhlmann, Polanica Zdroj 1967; or 15 ... cxd4 intending 16 Bxc6 bxc6 17 cxd4 c5! 18 dxc5 Ba6. 6 Bf3 Instead: (a) 6 Bd3 c5 7 dxc5 (7 a3 Bxc3+ 8 bxc3 Qc7= with the idea ... Nbd7 and ... c4) 7 ... Nbd7 8 Bd2 Nxc5 9 Bb5+, AverbakhBotvinnik, USSR Ch., Moscow 1955; 9 ... Bd7! and Black even has a small lead in development. And 7 Nf3 is safe but harmless: 7 ... cxd4! 8 Nxd4 e5 9 Nde2 0-0 10 0-0 Bg4 11 f3, Djuric-Drasko, Niksic 1996; 11 ... Be6 12 a3 Be7 ‘looks very solid for Black’ (Pedersen); I agree, and the kingside 4-3 majority may come in handy later. (b) 6 Bg5 is an innocuous line from the MacCutcheon.
Black can play, for example, 6 ... c5! (6 ... Nbd7 is also equal) 7 Nf3 (7 dxc5 Qxd1+ 8 Rxd1 Nbd7 9 Bf3 Bxc3+ 10 bxc3 Nxc5= intending 11 Be3 Nfd7; here 7 a3? Bxc3+ 8 bxc3 Qa5-/+ was Tal-Karner, Parnu 1971) 7 ... cxd4 8 Nxd4 (8 Bxf6 Qxf6 9 Qxd4 Qxd4 10 Nxd4 Bxc3+ 11 bxc3 Nd7=/+) 8 ... Bxc3+ 9 bxc3 Qa5 10 Bxf6 Qxc3+ 11 Qd2 Qxd2+! 12 Kxd2 gxf6=/+ Honfi-Lengyel, Budapest 1957. 6 ... c5 6 ... 0-0 7 Ne2 e5!? also equalizes: 8 0-0 (8 dxe5 Qxd1+ 9 Kxd1 Ng4 10 Bxg4 Bxg4 11 f3 Bf5 with at least equality – Moskalenko) 8 ... Bxc3 9 Nxc3 (9 bxc3?? e4) 9 ... exd4 10 Nb5 c5 11 Bf4 Ne8 (11 ... a6 12 Nd6 Nc6) 12 c3 a6 13 Na3 Nc6=. 7 Ne2 7 a3 Bxc3+ 8 bxc3 is 5 ... c5 above, and 7 Be3 is easily met by 7 ... cxd4 8 Qxd4 Qxd4 9 Bxd4 Nc6=, or 7 ... Nd5 8 Bd2 Bxc3
9 bxc3 Nc6. 7 ... Nc6 Liquidating the centre immediately by 7 ... cxd4 should also be satisfactory; e.g. 8 Qxd4 (8 Nxd4?! e5 9 Ndb5 a6 10 Qxd8+ Kxd8) 8 ... Qxd4 9 Nxd4 0-0 10 0-0 Nbd7=. 8 Be3 8 a3 Bxc3+ 9 bxc3 e5=. 8 ... cxd4 9 Nxd4 9 Bxc6+!? bxc6 10 Qxd4 may be best, when Black should probably opt for 10 ... Qxd4 11 Bxd4 Ba6+=. Instead, there’s a much-cited game Urban-Uhlmann, Würzburg 1993: 10 ... Qe7 11 0-0 Ba6 with dynamic play; unfortunately, this allows 11 a3!, when White takes control: 11 ... Ba5 (11 ... Bxc3+ 12 Qxc3!±) 12 Qa4 Qc7 13 0-0-0, which is awkward for Black. 9 ... Ne5 Or 9 ... Bxc3+ 10 bxc3 Ne5=.
10 Be2 10 0-0 Nxf3+ 11 Qxf3 Bxc3 12 bxc3 0-0 13 c4 Qc7 14 Bg5 Nd7 15 Qe2 a6 16 Rad1, Kallio-Kristjansson, Gausdal 2003; here 16 ... Nc5 was fine, but 16 ... Nb6=/+ is even stronger. 10 ... Nd5 11 Bd2 Nxc3 12 bxc3 Be7! 13 0-0 0-0 14 f4 14 Re1 Qd5!?=/+; 14 Rb1 Bf6. 14 ... Nd7 15 Bd3 Nc5 16 Qf3 Bd7 17 Rae1 Nxd3 17 ... Qc7! is a good alternative, with a comfortable advantage. 18 cxd3 Qc7=/+ Fichtl-Uhlmann, Zinnowitz 1966. 7.6 4 exd5 This is another version of the Exchange Variation. White frees Black’s bishop on c8 and can’t expect much as a result, but wants to use the advantage of the first move. Most of what we say about the Exchange in Chapter 12 applies here as well. White can’t play an early c4, of course, but he can try to take advantage of Black’s bishop on b4 by playing a3, and if ... Bxc3 follows then Black’s dark squares may become vulnerable. In return, Black gains c4 and quick development. Alternately, Black can play solidly and retreat his bishop. In general the 4 exd5 lines are considered safe but not pointed enough to cause Black many worries. 4 ... exd5
Now 7.61 7.62
White
has
two 5 5
main
moves: Bd3 Qf3
Instead: (a) 5 a3 tries to commit Black to ... Bxc3+ before he can arrange a retreat to d6 (that being the best square in many Exchange lines). Then Black can play simply 5 ... Be7, which is solid, or with more ambition: (a1) Simon Williams suggests 5 ... Qe7+!?, ‘since White does not have his bishop on d3 and so 6 Ne2 is not a very effective way of blocking the check’. This is sensible; e.g. 6 Qe2 (6 Ne2 Bxc3+ 7 bxc3 Nf6; 6 Be2 Bxc3+ 7 bxc3 Nf6 8 Nf3 Ne4!? and Black is at least equal – Williams) 6 ... Bxc3+ 7 bxc3 Nc6 8 Bg5!? Qxe2+ 9 Bxe2 Bf5 10 0-0-0 h6 11 Bh4 Nge7=. (a2) 5 ... Bxc3+ 6 bxc3 Ne7 (6 ... Nf6 7 Bd3 transposes to 6 a3 in the notes to 7.611) 7 Nf3 (7 Bd3 Bf5=) 7 ... 0-0 (or (7 ... Nbc6 8 Bd3 Bg4!) 8 Bd3 Bf5 9 0-0 Nbc6 10 Bf4, Treybal-Nimzowitsch, Carlsbad 1923; and 10 ... Na5=/+ was suggested by Nimzowitsch, while 10 ... Qd7 is also good. (b) 5 Nf3 Ne7 (a snippet from history is 5 ... Nc6 6 Bd3?! Bg4=/+ Em.Lasker-Botvinnik, Moscow 1936; 5 ... Bg4 6 h3 Bh5 is also good; and 5 ... Nf6 is examined by transposition in 17.21) 6 Bd3 (6 Bg5 Bf5 7 Be2 f6 8 Be3 Nbc6 9 Qd2 Qd7 10 a3 Ba5! 11 b4 Bb6= P.Chandler-Brecht, German League 2000) 6 ... Nbc6 (6 ... Bg4 7 h3 Bh5) 7 h3 Bf5 8 Bxf5 Nxf5 9 0-0 Bxc3 10 bxc3 0-0 11 Qd3 Nd6=/+ Mannheimer-Nimzowitsch, Frankfurt 1930. 7.615
Bd3
The main move; White develops logically. Now, as is typical with slow variations, Black has a selection of at least five valid continuations. I’ll concentrate upon a selection of subvariations which don’t require extraordinary precision in their handling. The analysis of last two is abbreviated, but should contain enough information to work with. 7.611 7.612 7.613
5 5 5
... ... ...
Nf6 c6 Nc6
Others: (a) 5 ... c5 6 dxc5 (6 a3 Bxc3+ 7 bxc3 c4) 6 ... Nc6 7 Nf3 (7 a3 Bxc5 with active piece play) 7 ... Nge7 8 0-0 Bxc5 may favour White slightly because of Black’s isolated d-pawn, but that’s not clear, and in case it wouldn’t be much to brag about. Black has the standard IQP deployments such as ... Bg4, ... Re8, ... a6, ... Qd6, etc. Psakhis assigns ‘?!’ to 5 ... c5 and cites LutikovNogueiras Santiago, Jurmala 1978: 9 a3 0-0 10 b4 Bd6 11 Nb5 Bb8 12 Re1 f6 13 Bb2, which is logical to this point. Psakhis calls this ‘+/=’, but the game continued rather normally by 13 ... a6 14 Nbd4 Ne5 15 h3 N7g6?! (15 ... Qc7= is more precise) 16 Bf5?! (16 Nxe5! Nxe5 17 Nf3+/=) 16 ... Qd6 17 Bxc8 Rxc8 and with the outpost on c4 and active pieces, Black stood better. (b) 5 ... Ne7 6 a3 Bxc3+ 7 bxc3 Nc6 transposes to 5 ... Nbc6 6 a3 Bxc3 7 bxc3 Nge7 in the notes to 7.613, but Black can opt for the modest 7 ... Bf5 with full equality. Here’s an example of the play with some typical themes: 8 Qf3 Qd7 9 Ne2 0-0 10 0-0 Nbc6 11 Bxf5?! Nxf5 (or 11 ... Qxf5!; White’s weaknesses don’t go away in the endgame) 12 Ng3 Nxg3 13 Qxg3 Ne7!? (13 ... Na5 is natural and also good; White’s bishop is the inferior minor piece) 14 Bf4 c6 15 Rfe1 Rfe8=/+ 16 Re3? Nf5 17 Qg4 Rxe3! 18 fxe3 Re8-/+ Moshina-Maksimovic, Ikaros 2002. 7.6115 This simple developing move is reliable and well regarded.
...
Nf6
6 Bg5 A pointed continuation. Instead: (a) 6 Nf3 0-0 (6 ... Qe7+ 7 Be3 Ne4 8 0-0! Nxc3 9 bxc3 Bxc3 10 Rb1 with good attacking chances) 7 0-0 Bg4 (‘the simplest solution’ – Psakhis) 8 Bg5 (8 h3 Bh5 9 Re1 Nc6 10 g4 Bg6=) 8 ... c6 9 h3 Bh5 10 Bf5!? (10 Re1 Nbd7) 10 ... Qd6 (White’s last didn’t really prevent 10 ... Nbd7!: 11 Bxd7 Qxd7 12 Bxf6 gxf6 with bishops and kingside prospects; upon 13 Ne5, 13 ... Qf5 14 Ng4 Kh8! is strong) 11 Qd3 Bxc3 12 bxc3 Nbd7 13 Bxf6 Nxf6 14 Ne5 Bg6= Van der Doel-Vaganian, Groningen 1999. (b) 6 a3 Bxc3+ 7 bxc3 0-0 8 Ne2 and: (b1) 8 ... c5!? 9 dxc5 Nbd7 10 Be3 (10 0-0 is 10 dxc5 in note ‘b3’) 10 ... Qc7 is double-edged; e.g. 11 Bd4 Ne4 12 0-0 Ndxc5 13 f3 Ng5 14 Re1 Nge6=. (b2) 8 ... Bg4 9 f3 Bh5 10 0-0 Bg6 is a standard manoeuvre; e.g. 11 c4 Bxd3 12 Qxd3 dxc4 13 Qxc4 Re8 14 Bg5?! Qd5 15 Qxd5 Nxd5 16 Kf2 b5! 17 Rfb1 Nc6 18 Bd2 (18 Rxb5 Rxe2+) 18 ... a6=/+ (or 18 ... Nb6), Preissmann-Hug, Swiss Team Ch. 2010. (b3) 8 ... Nbd7 9 0-0 (9 Be3 Nb6) 9 ... c5 10 Ng3 (10 dxc5 Nxc5 11 Bg5 h6 12 Bh4 Re8= was the actual route of DvoirysEingorn in the note to 6 Bg5 h6 below) 10 ... c4!? 11 Be2 Re8 12 Bf3 Nb6!? (12 ... h6 13 Bf4 Nf8) 13 a4 Bd7 14 a5 Na4 15 Bg5 h6 16 Bxf6 Qxf6 17 Qd2 Bc6 18 Rfb1 Qd8!? 19 Rb4 Qxa5 20 Qf4 with mutual and unclear chances; or here the simpler 18 ... b6. (c) 6 Ne2 0-0 7 Bg5 transposes to the main line. Here 7 0-0 has several valid answers: 7 ... Re8 (7 ... c6 8 a3 Bd6=; 7 ... Bg4 8 f3 Bh5 9 Nf4 Bg6 10 Nxg6 hxg6 11 Bg5, Ivanchuk-W.So, Khanty-Mansiysk 2009; and So recommends 11 ... c5! 12 dxc5 Bxc3 13 bxc3 Nbd7 ‘when Black has good counterplay’; 12 ... Qa5! also levels the game) 8 Bg5 c6 9 Qc1!? (9 Qd2 h6 10 Bxf6 Qxf6 11 Rae1 Nd7 12 Ng3 Rxe1 13 Rxe1 Qd8!? 14 Nf5 Nf6= Fluvia Poyatos-Moskalenko, Illes Medes 2006/07) 9 ... Nbd7 10 Ng3 Nf8! 11 Nce2 h6 12 Bf4 Ng6 (12 ... Ne6 was also good) 13 Bd2, Annaberdiev-Al Modiahki, Doha 2006; and now 13 ... Bd6! is at least equal. 6 ... 0-0 The immediate 6 ... h6 is also effective (often transposing): 7 Bh4 (7 Qe2+ Be6 8 Bh4 g5! 9 Bg3 c5!) 7 ... 0-0 8 a3 (8 Ne2 transposes to 6 ... 0-0; not 8 Qf3?? g5 9 Bg3 Bg4 10 Qe3 c5–+) 8 ... Re8+ 9 Ne2 Bxc3+ (or 9 ... Be7=) 10 bxc3 c5! 11 dxc5 Nbd7 12 0-0 Nxc5= Dvoirys-Eingorn, Lvov 1990.
7 Ne2 Consistent with White’s set-up. Others: a) 7 a3 Bxc3+ 8 bxc3 (this sometimes arises from an earlier a3 by White) 8 ... Qe8+!? (clever, although 8 ... c5 9 dxc5 Nbd7 10 Ne2 Nxc5= is simpler and quite as good) 9 Ne2 (9 Qe2 Qc6! 10 Qd2 Re8+ 11 Ne2 Ne4 12 Bxe4 Rxe4 13 0-0 f6=/+) 9 ... Ne4 10 Bf4?! (10 Bd2 Nd7 11 0-0 Nb6 12 Ng3 Qc6 13 Qh5?! f5 14 Bxe4 dxe4 15 f3? e3!–+ Druckenthaner-Kindermann, Austrian Team Ch. 2002) 10 ... Nxc3 11 Qd2 Nxe2 12 Bxe2 Nc6-/+ Stromer-Degraeve, Cappelle la Grande 1995. (b) 7 Qd2? c5! is already better for Black: 8 a3 (8 dxc5 d4 9 0-0-0 dxc3 10 Bxh7+ Kxh7 11 Qxd8 Rxd8 12 Rxd8 cxb2+ 13 Kxb2 Nbd7-/+; 8 Nf3 Qe8+! 9 Qe2 c4-/+) 8 ... Ba5 9 Nf3 Qe8+ 10 Be3 c4 11 b4 cxd3 12 bxa5 dxc2 13 Qxc2 Nc6 14 0-0 Bg4! 15 Nd2 Nxa5-/+. 7 ... h6 8 Bh4 8 Bf4 c5 9 dxc5 Nc6 10 0-0 Bxc5 11 Na4 Bd6 12 Qd2 Be6=. 8 ... c6 8 ... c5!? 9 dxc5 Nc6 10 0-0 Bxc5 11 Nxd5!? g5 12 Nxf6+ Qxf6 13 Bg3 Qxb2 14 Rb1 Qf6= ½-½ Skripchenko-Barua, Groningen 1997. 9 0-0 Re8
Moskalenko comments: ‘Zugzwang!’, the point being that most natural moves will only help Black. Alternatively, 9 ... Nbd7 10 a3 Bd6 11 Bg3 Nb6 12 Bxd6 Qxd6= Rechel-Eingorn, Metz 1998. 10 f4?!
10 h3 Nbd7 11 a3 Bd6 12 Qd2 Nf8 13 Rae1 Ne6 is comfortable for Black, who can meet 14 f4 by 14 ... Bd7 15 f5 Ng5! 16 Bxg5 hxg5 17 Qxg5 c5! 18 dxc5 Bxc5+ 19 Kh1 Be3 20 Qg3 d4 21 Nb5 Bxb5 22 Bxb5 Ne4 with more than enough for a pawn. 10 ... Bg4 Or 10 ... Nbd7=/+. 11 h3 Bxe2 12 Nxe2 Nbd7 Perpinya Rofes-Moskalenko, Catalonia 2002. ‘Black has at least an equal game’ (Moskalenko). 7.612 5 ... c6 This is recommended in the recent books by both Williams and Vitiugov. What’s more, Uhlmann employed 5 ... c6, and theoreticians generally approve of the move, so it’s probably worth a quick look! 6 Qf3 Oddly enough, this is the main reply. White has several ideas in mind, including pinning a knight on f6 by Bg5, exchanging a bishop on d6 with Bf4, and discouraging Black’s normal simplification via ... Bf5. Alternatively: (a) 6 Ne2 Ne7 and ... Bf5 should equalize without serious problems; e.g. 7 0-0 Bf5 8 Ng3 Bxd3 9 Qxd3 0-0 10 Nce2 Na6 11 c3 Bd6 12 Bf4 Nc7 13 Rae1 Ne6 ½-½ Oll-Short, Parnu 1996. (b) 6 Nf3 Qe7+ 7 Be3 Nf6 8 0-0 0-0 9 Re1 Qc7 10 h3 Bd6 11 Bg5 Nbd7 12 Qd2 b5!? 13 a3 a5 with double-edged play. 6 ... Nf6 6 ... Qf6 has been used twice by Ivanchuk and is recommended by Williams and Vitiugov. It is adequate, although what chances there are seem to be on White’s side; e.g. 7 Bf4 Ne7 8 Ne2 Ng6 9 Qe3+ Qe6 10 Bg3+/=. 7 Bg5 Nbd7 8 Ne2 0-0
9 0-0 Williams cites Fressinet-Pelletier, Istanbul Olympiad 2000, with 9 0-0-0, which looks risky to me. A simple response is 9 ... Re8 10 Rhe1 h6; e.g. 11 Bf4 Nf8 12 h3 Ne6 13 Be5 Nd7 14 Bg3 Bxc3 15 bxc3 Qa5. 9 ... h6! 10 Bh4 Re8 11 h3 Be7 12 Rfe1 Nf8 13 Bg3 Ne6 And ... Bd6 follows with equality. 7.613
5
...
Nc6
In recent years Black has experienced some new challenges here, but they seem to be manageable if you know some theory. I’ve recommended this in previous editions and it has become Black’s most popular solution to 4 exd5, although recently there has also been a turn towards 5 ... Nf6 and assorted solid lines. Here’s the main line, which you can supplement with research: 6 a3 A single recent example with 6 Ne2: 6 ... Bg4 (6 ... Nge7 7 0-0 Bf5=) 7 a3 Ba5 8 h3 Bh5 9 Be3 Nge7 10 Qd2 Bg6 11 Ng3 Qd7 12 0-0-0 0-0-0 13 Bxg6?! hxg6 14 Qd3 Bxc3! 15 Qxc3 Rde8, Doluhanova-Pähtz, Dresden 2011. Suddenly Black’s game looks good. White has a bad bishop and his knight has no forward squares. 6 ... Bxc3+ 6 ... Ba5 has been used by Apicella and Uhlmann and is undoubtedly playable. The capture is thematic. 7 bxc3
7 ... Be6!? This is my previously recommended move, which maintains flexibility and allows Qh5 to be met by ... Nf6, while Black can also play a quicker ... 0-0-0 in several lines. After 7 ... Nge7, Black needs to play a bit more carefully than with his alternatives because of 8 Qh5, preventing ... Bf5, and causing a little awkwardness in Black’s camp. With caution, however, Black can take his share of the board after 8 ... Be6.
Here’s a brief look at White two main choices: (a) 9 Nf3 Qd7 10 Ng5 0-0-0 (with the idea 11 Nxf7? Qe8) has been tested many times. A frequently-played line is 11 Nxe6 Qxe6+ 12 Be3 g6 13 Qf3 Nf5 14 0-0 Nxe3 15 Rfe1 Ne5! 16 dxe5 (16 Qxe3 Nxd3 17 Qxe6+ fxe6 18 cxd3 Rd6=, with even a touch of queenside pressure) 16 ... Nc4 17 Bxc4 dxc4 18 Qe3 b6 19 a4 a5 and Black stands better if anybody. (b) 9 Rb1 Qc8!? (an alternative to the well-investigated 9 ... b6) 10 h3 Bf5 (getting the key move in) 11 Ne2 0-0 12 0-0, Berset-Huss, Lenk 2011; and 12 ... b6 looks most logical, with good chances to equalize. 8 Rb1 This is the most common move by quite a margin. The other important choice is 8 Qf3 Nge7 (8 ... Qd7 is also played) 9 Ne2 (9 Nh3 Qd7 10 Ng5, Kramnik-Carriere, Bonn 2004; 10 ... Bf5!= with the idea 11 g4?! Bg6) 9 ... Qd7 10 Ng3!? (10 h3 is less ambitious and allows 10 ... Bf5!?; e.g. 11 0-0 0-0 12 Ng3 Bxd3 13 cxd3 Ng6 14 Bg5 Rae8= Vishnu-Adrian, Figueres 2011; 10 Nf4 Bf5 11 0-0 can be met by 11 ... 0-0-0!? or the safer 11 ... 0-0) 10 ... Bg4!? (10 ... 0-0 11 h3 Na5 12 0-0 Ng6 13 Nf5 Bxf5 14 Bxf5 Qc6=) 11 Qe3 0-0 12 0-0 Rfe8=. 8 ... b6
9 Ne2 Or 9 Qf3 Nge7 10 Ne2 Qd7 11 Ng3 0-0 12 h3 Na5 13 0-0 Rae8 14 Nh5, Malloni-H.Andersen, Kaunas 2011, and now 14 ... f6=; e.g. 15 Qg3 g6 16 Nf4 Nf5. After 9 Ne2, Nataf-Rustemov, Stokholm 2002, went 9 ... Qd7 10 0-0 Nge7 11 Nf4 Bf5 12 Nh5 0-0 13 h3 (13 Bh6? Bg4!-/+), when the most interesting move is 13 ... Bg6!?; e.g. 14 Qf3 Na5! 15 Re1 Rae8=.
7.62 5 Larsen’s old weapon, with which he won some nice games, but it fell into relative disuse because of Black’s reply. 5 ... Qe7+ This has been the key move for decades. A good solid alternative is 5 ... Nc6, which leads to interesting play after 6 Bb5 Ne7:
Qf3
(a) 7 Bf4 0-0 (7 ... a6 8 Bxc6+ bxc6 9 Ne2 0-0 10 0-0 Rb8) 8 0-0-0 (8 Ne2 Ng6! 9 0-0-0 a6 10 Ba4 Be6 and I prefer Black’s attacking chances), B.Larsen-L.Portisch, Amsterdam Interzonal 1964; and here 8 ... a6! favours Black after either 9 Bxc6 bxc6 or 9 Ba4 Na5! intending ... b5. (b) 7 Bg5!? f6 8 Bf4 0-0 9 0-0-0 a6 (9 ... Be6 may be more accurate) 10 Bd3?! (10 Bf1 Be6=/+; 10 Ba4 b5 11 Bb3 Be6=/+; 10 Bxc6! bxc6 11 Nge2 looks best, with equal and mutual chances) 10 ... Nxd4 11 Bxh7+ Kxh7 12 Rxd4 c5 13 Rd1 Bxc3 14 Qxc3 Bf5 15 Ne2 d4=/+ Sukhanov-Krasnov, Moscow 1971. (c) 7 a3 Bd6!? (7 ... Bxc3+ 8 Qxc3 a6 9 Be2 0-0 10 Nf3 Bg4=) 8 Be3 (8 Nge2 0-0 9 Bd3) 8 ... a6 (8 ... Be6 9 0-0-0 a6 10 Bd3 b5 with mutual chances) 9 Bd3 Be6= intending ... Qd7, Karttunen-N.Pert, Porto Carras 2011; e.g. 10 Nge2 Qd7 11 h3 0-0 12 0-0 Ng6. (d) 7 Ne2 Bf5 8 Bd3 Qd7= 9 a3 (9 0-0 Bxd3 10 Qxd3=) 9 ... Bxd3 10 Qxd3 (10 cxd3 Bd6=/+) 10 ... Bxc3+ 11 Nxc3 Qf5! 12 Qxf5 Nxf5 13 Be3 0-0-0 14 0-0-0 Rhe8 with a slight edge for Black, due to White’s bad bishop.
6 Ne2
The best move. (a) 6 Be3?! runs into trouble after 6 ... Nf6 (or 6 ... Nc6!, when 7 Ne2 transposes to 6 Ne2 Nc6 7 Be3 Nf6 below) 7 Bd3 (7 a3 Bg4 8 Qg3 Bd6 9 Qh4 Bf4 10 Nxd5 Nxd5 11 Qxg4 0-0-/+; 7 Ne2 Nc6 is the same transposition as after 6 ... Nc6) 7 ... c5! 8 Kf1 (8 a3 Ba5) 8 ... Bxc3 9 bxc3 c4 10 Bf5 Bxf5 11 Qxf5 0-0-/+ Mestrovic-Maric, Kraljevo 1967. (b) White may not even quite equalize after the passive 6 Qe3; e.g. 6 ... Nc6 7 Nf3 (7 Bb5 Bf5 8 Qxe7+ Nxe7 9 Bf4 Bxc2 10 Bxc7 Rc8 11 Bf4 a6=/+ Saharov-Antoshin, Sochi 1966) 7 ... Bg4 (7 ... Bf5 8 Bd3 Bxd3 9 cxd3 0-0-0 10 0-0, Ajala-I.Farago, Harrachov 1967, and 10 ... Qxe3 11 fxe3! Nf6 12 Bd2 is about equal) 8 a3 (8 Ne5 Nxe5 9 dxe5 Rd8!=/+; 8 Be2? Bxf3 9 gxf3 0-00=/+) 8 ... Bxc3+ 9 bxc3 Bxf3 10 gxf3 0-0-0. This is a hard-to-assess position, but I think one of the rare times that the knight pair is worth a bit more than the bishops, only because White has two sets of doubled pawns and four isolated pawns. 6 ... Nc6
7 Qd3 I recommended this move as an experimental try inDWF. Given the alternatives, it is doubtless the best try: (a) 7 Be3?! Nf6 is already difficult for White, who has to deal with the moves ... Ne4, ... Bg4, and ... Bxc3+ (with ... Qa3 in response to bxc3); for example, 8 a3 Bxc3+ 9 bxc3 Ne4 (or 9 ... Bg4=/+) 10 Qh5 Be6 11 f3, Biyiasis-J.Watson, Vancouver 1977; and among other good moves, 11 ... Nf6! stands out: 12 Qh4 Bf5 13 Bg5 0-0-0 14 Bxf6 gxf6 and White has an unsafe king with too many weaknesses. (b) 7 Qxd5 is obviously risky after 7 ... Nf6, when Black has a huge lead in development. White is forced along a narrow path: 8 Qb5! (8 Qc4 Be6 9 Qd3 0-0-0 10 Be3 Rhe8 will end in a Black advantage) 8 ... Bd7!? (8 ... 0-0! is more flexible; Black needn’t be in a hurry) 9 Qd3! 0-0-0 10 Be3 Bxc3+!? (10 ... Rhe8 11 0-0-0 Ng4 offers more than enough compensation; here 11 a3 is better, when 11 ... Ba5 is possible, or 11 ... Bxc3+ 12 Nxc3 Bg4!) 11 Nxc3 (11 Qxc3 Nd5; 11 bxc3 Rhe8 or 11 ... Nd5 with good prospects) 11 ... Bg4 (11 ... Nb4 12 Qd2 Bf5 13 Rc1 Ne4 14 Nxe4 Bxe4 is unclear) 12 Be2 Bxe2 13 Qxe2 Qb4 14 0-0-0 Nxd4 15 Bxd4 Rxd4 16 Qe5 Rhd8=. 7 ... Nf6! Williams recommends this good natural move, which I neglected in my book (thinking that Black shouldn’t allow Bg5). It equalizes.
8 f3 This keeps Black’s knight out of e4. Otherwise: (a) Williams analyses 8 a3 Bxc3+ 9 Qxc3 Ne4 (9 ... 0-0 10 f3) 10 Qe3 Bf5 11 f3 Nd6 12 Qxe7+, when I think 12 ... Kxe7! is best and moderately advantageous, because 13 c3 Na5 targets White’s queenside light squares. His 12 ... Nxe7 13 Bf4 0-0-0 14 0-0-0 Ng6 15 Bg3 h5 is less convincing after 16 h4 with the idea 16 ... Nc4 17 Bf2, but in general Black is doing well after 8 a3. (b) 8 Bg5 h6! 9 Bh4 (9 Bxf6 Qxf6 10 0-0-0 Be6 accomplishes little; 9 Be3 0-0 10 0-0-0 Na5! 11 a3 Bd6 and Black’s development is freer than White’s) 9 ... g5 10 Bg3 Bxc3+ gives Black the better game after either 11 bxc3 Ne4 or 11 Qxc3 Ne4=/+. After 8 f3, Black can equalize and retain an unbalanced contest with two moves: (a) 8 ... 0-0 9 a3 (9 Bg5 h6 10 Bh4 Na5!=/+) 9 ... Bd6 (9 ... Bxc3+ 10 Qxc3 Ne8) 10 Bg5 Be6 11 Nb5 Na5=. (b) 8 ... Ba5 (with the idea of ... Nb4 and ... Bf5) 9 a3 Be6 10 Bd2 (10 Bg5 0-0-0 11 0-0-0 h6 12 Bh4 g5 13 Bf2 Ne8!? intending ... Nd6) 10 ... 0-0-0 11 0-0-0 Kb8 12 Kb1 a6. 7.7
4
Qd3
White defends e4. He would like to transfer the queen to pressure the kingside. In addition, he can prepare 0-0-0 by, say, Bd2. 4 ... Ne7
White’s main idea is 4 ... dxe4 5 Qxe4 Nf6 6 Qh4, which may be fine for Black but isn’t worth contesting. The text recognizes that the queen is not well placed on d3, and thus doesn’t need to be forced to a better spot. Another solution is 4 ... Nc6 and then: (a) 5 Nf3 dxe4!? (5 ... Nf6 6 e5 Ne4 7 a3?! Bxc3+ 8 bxc3 f6! 9 exf6, Jerez Perez-Romero Holmes, Cala Galdana 2001; and now 9 ... Qxf6 10 Be2 0-0 gives Black good development and a variety of plans such as ... b6, ... Bb7 and ... e5, or ... a5/ ... Ba6, or ... Bd7-e8-g6/h5) 6 Qxe4 Nf6 7 Qh4 Ne7! 8 Bd3 Nf5 9 Bxf5 exf5 10 Bg5 Qe7+ 11 Ne5 Qe6 12 0-0 Bxc3 13 bxc3 Ne4=. (b) 5 e5 f6 6 a3 Ba5 7 b4 Bb6 8 Nf3 fxe5 9 dxe5 Nge7 10 Bb2, Münch-Reimer, corr 1998; and simplest was 10 ... a6! 11 Na4 Ba7=/+. 5 Bd2 Probably best, thematically preparing 0-0-0. Others: (a) 5 Ne2
Now Black can play a variety of moves, including 5 ... c5. These two are more subtle: (a1) 5 ... Nd7 6 a3 Ba5 (or 6 ... dxe4 7 Qxe4 Bd6) 7 b4 Bb6 (White has stopped ... c5, but now has to face ... e5) 8 Ng3 a5 9 Rb1 axb4 10 axb4 0-0 11 Be2 (11 e5 f6 12 f4 fxe5 13 fxe5 Nc6 14 Nce2 Qh4 intending ... Ndxe5) 11 ... e5! 12 Nxd5 Nxd5 13 exd5 exd4 14 0-0 Ne5-/+ Chevallier-Luce, Torcy 1991. (a2) 5 ... 0-0 6 a3 Ba5 (6 ... dxe4 7 Qxe4 Bd6=) 7 Bd2 Nbc6 8 0-0-0 a6!? 9 h4 b5= Kir.Georgiev-M.Kuijf, Wijk aan Zee 1984. (b) 5 Bg5 0-0 (5 ... f6 is also fine) 6 Nf3 Bxc3+ (6 ... f6 7 Bd2 c5!; or 6 ... b6 intending ... Ba6) 7 bxc3 f6 8 Bd2 b6 (8 ... c5!) 9 Qe3 Bb7 10 Bd3 (10 e5!? Nf5 11 Qf4 Nc6=/+ – Vaganian) 10 ... dxe4 11 Bxe4 Nf5! 12 Qe2 Bxe4 13 Qxe4 Qd5=/+ SmirinVaganian, Naberezhnye Chelny 1988.
5 ... c5 5 ... b6 has been played in several games, but a more common order for this idea is 5 ... 0-0 6 a3 (6 0-0-0 c5 7 dxc5 Nbc6 transposes to the main line; here 6 ... b6 is a sound alternative; e.g. 7 Nh3 Ba6 ½-½ Meszaros-Lahlum, Kecskemet 2003, having in mind 8 Qf3 Bxf1 9 Rhxf1 Bxc3 10 Bxc3 dxe4 11 Qxe4 Qd5=) 6 ... Bxc3 7 Bxc3 b6! 8 0-0-0 (8 Nf3 Ba6 9 Qe3 Bxf1 10 Rxf1 Nd7=/+) 8 ... Ba6 9 Qf3 (9 Qe3 Bxf1 10 Rxf1 a5! 11 Nf3 Nbc6 12 Bd2 b5!=/+ Jerez Perez-Moskalenko, Barbera 1999) 9 ... Bxf1 10 Rxf1 Nbc6 11 Ne2 a5 12 a4?! (but ... b5-b4 is a problem anyway) 12 ... Qd7! 13 Nf4 dxe4 14 Qxe4 Nb4!-/+ 0-1 M.LevittNogueiras Santiago, Bled Olympiad 2002. 6 dxc5 Nbc6 7 0-0-0 0-0 7 ... Bxc5 is probably more accurate, just to avoid the possibilities in the next note. 8 Nf3 Others: (a) 8 exd5 exd5 9 Na4 Bf5 10 Qb3, when 10 ... Qa5! 11 Bxb4 Nxb4 12 c3 Nbc6 offers good queenside attacking chances. (b) 8 Qg3 d4 9 Nb5 Bxc5 10 h4?! (slow) 10 ... a6 11 Na3 f5! 12 exf5 Nxf5 13 Qb3 b5 14 Nf3 Qd6 15 Bd3 Bb7=/+ SuerKanko, Tel Aviv Olympiad 1964. 8 ... Bxc5 9 Be3
9 ... Qb6 9 ... Bxe3+ 10 Qxe3 Qc7 is a calm approach. 10 Bxc5 Qxc5 11 exd5 exd5 12 Nd4
and
After 12 Nxd5 Nxd5 13 Qxd5 Qxf2, Black intends ... Be6 or ... Bg4. 12 Nd4 is Ribeiro-Russek Libni, Maringa 1991, when a simple solution is 12 ... Be6! with the idea 13 Nxe6 fxe6 14 f3 Nf5 a nice advantage.
Chapter
Eight
Winawer
Variation:
Fifth
Moves
1 e4 e6 2 d4 d5 3 Nc3 Bb4 4 e5 c5
The thrust 4 ... c5 attacks the centre in traditional French Defence style. The natural response is 5 a3, so that after 5 ... Bxc3+ 6 bxc3, White’s threatened d4 point is fortified and he has the two bishops. Nevertheless, White has several other fifth moves and it’s essential to know how to answer them. The solutions presented here are not radically different from those in previous editions. 8.1 8.2 8.3 8.4
5 5 5 5
Nf3 Bd2 dxc5 Qg4
8.15 Nf3
This simple development isn’t used much, mainly because White’s centre isn’t well supported. It is nevertheless perfectly
sound. 5 ... cxd4 This is more complicated than I realized at the time of the previous edition, and I’d also recommend the following two options which might make life easier: (a) After 5 ... Ne7, 6 a3 Bxc3+ 7 bxc3 transposes to the Positional Winawer (Chapter 10), and 6 dxc5 will transpose to 5 dxc5 below. I personally like this choice. (b) 5 ... Nc6 is a natural move, open to transposition. An independent line is 6 Bb5 (6 a3 Bxc3+ 7 bxc3 Ne7 and 7 ... Qa5 are Positional Winawers; 6 dxc5 is a direct transposition to 5 dxc5; and 6 Bd2 could transpose to something related to 5 Bd2 after 6 ... Ne7 7 Nb5 0-0 or 6 ... Nh6, though White also has to deal with 6 ... cxd4 7 Nb5 Bc5) 6 ... Ne7 (6 ... Qb6; 6 ... Bd7) 7 0-0 0-0 8 dxc5 Bxc5=; e.g. 9 Na4!? Bb4 10 a3 Ba5 11 b4 Bc7 12 Qe2 a6 13 Bd3 Ng6=.
6 Qxd4 6 Bb5+ Bd7 is unimpressive, while 6 Nxd4 has several answers, the most popular being the double attack on c3 and e5 by 6 ... Qc7: 7 Bb5+ Nc6 (or 7 ... Nd7, or 7 ... Bd7 8 0-0 Ne7=) 8 0-0 Ne7 9 Qg4 0-0 10 Bxc6 Bxc3 11 bxc3 Nxc6; e.g. 12 Ba3 (12 Nxc6 bxc6 13 Re1 f5 14 exf6 Rxf6=) 12 ... Nxe5 13 Qg3 Re8 14 Nb5 Qd7 15 Nd6 Nc4 16 Nxe8 Qxe8 17 Bc1 f6=/+; Black has a powerful centre and his knight is a monster. 6 ... Nc6
7 Bb5 7 Qg4!? isn’t bad: 7 ... Nge7 8 Qxg7 Rg8 9 Qh6 (9 Qxh7 d4 10 a3 Qa5 11 Rb1 dxc3 12 Be3! Bd7! 13 axb4 Qa2 14 Rd1 cxb2
15 c3 Qb3! is about equal; compare 5 Qg4 Ne7 6 dxc5 Nbc6 7 Qxg7 etc in the notes to 8.41) 9 ... d4 (9 ... Bxc3+ 10 bxc3 Qc7 11 Bb5!? Rxg2 12 Kf1 Rg8, and both sides have dynamic chances; simply 9 ... Qc7 is also playable and equal after 10 Bd2 Nxe5 11 00-0! Nxf3 12 gxf3 Bd7 13 Kb1 0-0-0; e.g. 14 Qxh7 d4 15 Ne4 Bxd2 16 Rxd2 Nf5!=/+) 10 a3 dxc3 (or 10 ... Rg6 11 Qxh7 dxc3 12 axb4 Nxb4 13 Qh8+ Rg8 14 Qh7=) 11 axb4 Nxb4 12 Qxh7 Rg6 13 Bb5+ Bd7 14 Bxd7+ Qxd7 15 0-0 Nxc2 16 Rb1 Nd4 17 Nxd4 Qxd4 18 bxc3, and if nothing else Black can bail out by 18 ... Rxg2+ 19 Kxg2 Qg4+=. 7 ... Qa5 8 Bd2 8 0-0 Ne7 9 Qg4 Bxc3 10 Bxc6+ bxc6 11 bxc3 Ba6 12 Re1 Qxc3= Arpi-Harikrishna, Bikaner 1999, with the idea 13 Qxg7? Rg8 14 Qf6 Nf5-/+. 8 ... Ne7
9 Qg4 There are several alternatives, none of them dangerous: (a) 9 a3 0-0 (9 ... Bxc3=) 10 Bxc6 Nxc6 11 Qg4 Kh8!? 12 0-0 Be7!=. (b) 9 Bxc6+ bxc6 10 0-0 (10 Ne4 Bxd2+ 11 Qxd2 Qxd2+ 12 Nexd2 a5 13 0-0 Ba6 14 Rfe1 0-0=/+) 10 ... c5! (10 ... Nf5 11 Qf4 Ba6 12 Rfd1 0-0 13 a3 Bxc3 14 Bxc3 Qc7=) 11 Qg4 Ng6 12 a3 Bxc3 13 Bxc3 Qb5 intending ... d4 and ... Bb7. (c) 9 0-0 Bxc3!? 10 Bxc6+ Nxc6 11 Qxc3 Qxc3 12 Bxc3 0-0=. 9 ... 0-0 10 Bd3!? After 10 Bxc6 bxc6 11 a3 c5! 12 0-0 Bxc3 13 Bxc3 Qc7 Black has a comfortable position with a modest central advantage. 10 Bd3 is Hodgson-Ravi, London 1987. Here I’d recommend simply 10 ... Ng6 11 a3 (11 Bxg6 fxg6!=/+ 12 a3 Be7; 11 0-0 Ncxe5 12 Nxe5 Nxe5 13 Bxh7+ Kxh7 14 Qh5+ Kg8 15 Qxe5 f6=/+ with two bishops and a mobile centre) 11 ... d4!? 12 Bxg6 (12 Nxd4 Ncxe5 13 Qe4 Nxd3+ 14 cxd3 Be7=/+) 12 ... hxg6 13 Nxd4 Qxe5+ 14 Nde2 Be7 15 0-0 b6=/+. The bishop pair suffices for a limited advantage. 8.2
5
Bd2
Credited to Bogoljubow, this developing move prevents the doubling of White’s c-pawns and prepares Nb5, from where the knight can infiltrate to d6 or simply cover d4. Of the fifth move alternatives, this has been the most popular, although it hasn’t caught on among the leading players. I’ll show two main replies: 8.21 8.22
5 5
... ...
Ne7 Nh6
The difference between this and the last edition is that, since there’s really not enough room to cover these systems in full detail, I’m picking 5 ... Ne7 to be the main line. That’s because its theory has stabilized and is unlikely to change fundamentally. Unfortunately, the theory of 5 ... Nh6 (which was a very minor, almost unknown, move when I first recommended it) has simply exploded, and it would take considerable commitment of space to fill in all the details. So I’ll undertake a less expansive investigation, one which will especially concentrate upon lines which have become critical since the previous edition. 8.21
5
This
is
the
traditional
...
line.
In
Ne7
response,
White
plays:
8.211 8.212 8.213
6 6 6
Nb5 a3 f4
There are two other logical moves: (a) 6 dxc5 Nbc6 (Black can also play 6 ... 0-0) 7 f4 (7 Qg4 transposes to 5 Qg4 – see 8.41; 7 Nf3 is note ‘b’) 7 ... 0-0 8 Nf3 f6 (8 ... Nf5!) 9 Bd3, Santo Roman-Kindermann, Uzes 1990, when 9 ... fxe5 10 fxe5 Nf5=/+ is not what White wants. (b) 6 Nf3 is highly transpositional; for example, 6 ... cxd4 7 Nb5 Bxd2+ 8 Qxd2 0-0 transposes to 6 Nb5 Bxd2+ 7 Qxd2 0-0 8 Nf3 cxd4. Black also has the generally more complex 6 ... Nbc6, which contains some instructive themes: (b1) 7 dxc5 Ng6!? (or 7 ... 0-0 and 8 Bd3 Ng6 or 8 a3 Bxc5 9 Na4 Bd4!? 10 Nxd4 Nxd4 11 Bd3 Qc7!) 8 Bb5 (8 Nb5 Bxc5 9 Bc3 0-0=/+ 10 h4 a6 11 Nbd4 Qb6 12 Qd2 Nxd4 13 Nxd4 Nxe5-/+ Macia-Ludevid, email 1993) 8 ... Bxc5 9 0-0 (9 h4?! Bd7 10 Bxc6 Bxc6 11 h5 Ne7 12 h6 gxh6=/+ with the idea 13 Bxh6? Qb6! or 13 Qe2 Rg8 14 g3 Qb6) 9 ... Bd7 10 Re1 a6 (10 ... 0-0!) 11 Bd3?! Qb6! 12 Qe2 Nd4! 13 Nxd4 Bxd4=/+ 14 Be3? Bxe3 15 Bxg6, Holms-Tosti, corr 1994, when 15 ... hxg6 16 fxe3 Qxb2 17 Qd3 Rc8–+ is one of several winning lines. (b2) 7 Nb5 Bxd2+ 8 Qxd2 0-0 9 dxc5 (9 c3 a6! transposes to 9 Nf3 a6 in section 8.2111) 9 ... b6! 10 cxb6 Qxb6 11 0-0-0 Rb8 12 b3, An.Martin Gonzalez-Sanchez Jimenez, Spanish Ch. 1993. Black’s 12 ... Qc5 was at least equal, while 12 ... Bd7! 13 Nbd4 a5 is better still. 8.211
6
Nb5
Bxd2+
7
Qxd2
0-0
The main line. White’s immediate concern is to resolve the central situation, and there is a wide choice of means: 8.2111 8.2112 8.2113
8 8 8
c3 f4 dxc5
8.21118 c3 Nbc6!? This is a good move but not the only one. Here are some other thematic lines; note especially the efficient and effective development in ‘c’: (a) 8 ... f6 9 exf6 Rxf6 10 dxc5 Nd7 11 b4 a5 12 f4!? (12 Nf3 b6 13 cxb6 Qxb6 14 Be2 axb4 15 cxb4 Ba6=) 12 ... b6! 13 cxb6 Qxb6 with dynamic play, Martinez-Arencibia, Cuba 1988. (b) 8 ... a6 9 Na3 (9 Nd6 cxd4 10 cxd4 Nbc6 11 0-0-0 f6 12 f4 Qb6 13 Nf3 Bd7 intending ... Nc8) 9 ... Nbc6 10 f4 (10 Nf3 f6=) 10 ... cxd4 11 cxd4 Nf5 12 Nf3 f6 is complex and equal. (c) 8 ... Bd7! 9 Nd6 (9 dxc5 Bxb5 10 Bxb5 Nd7) 9 ... cxd4 10 cxd4 Qb6=/+ 11 f4?! Nbc6 (or 11 ... Nc8 12 Nxc8 Rxc8 13 Nf3 Bb5!? 14 Qb4?! Bxf1 15 Qxb6 axb6 16 Rxf1 Nc6 17 Kd2 Ra4!=/+ Kristjansson-B.Sorensen, Reykjavik 2011) 12 Nf3 Nc8! 13 Nxc8 Raxc8 with a healthy advantage, Yildiz-Zhukova, Tbilisi 2011. 9 f4 Or 9 Nf3 a6 and then:
(a) 10 Nd6!? cxd4 11 cxd4 f6 12 Nxc8 Rxc8 13 exf6 Rxf6 14 Bd3 g6!? (eliminating Bxh7+ and thus preparing ... Qb6) (14 ... Ng6 15 Bxg6 Rxg6 16 0-0 Qf6 17 Kh1 Rg4 18 Rad1 Rf4 19 Qe3 Re4 with a little pressure) 15 Be2 (15 0-0? Rxf3! 16 gxf3 Nxd4 17 Qe3 Nec6 18 Rad1 Qf6 19 f4 Rf8=/+) 15 ... Qb6 16 0-0 Nf5 17 Rfd1 Rff8 18 b3 Qb4! with the idea 19 Qxb4 Nxb4 20 a3 Rc2=/+. (b) 10 Na3 cxd4 11 cxd4 f6 (11 ... Nf5 12 Nc2 f6 13 exf6 Qxf6 14 Bd3 e5 15 dxe5 Nxe5 with an edge, Turzo-MusiqueWand, ICC 2010) 12 exf6 Rxf6 13 Nc2 Qd6 14 0-0-0 (14 Be2 Bd7! 15 0-0 Be8 16 Rac1 Bg6=/+) 14 ... Bd7 15 Bd3 Raf8 (15 ... Be8!) 16 Rde1?! (16 Kb1) 16 ... Be8 17 Rhf1 Ng6!-/+ Strikovic-Ivkov, Cetinje 1991. 9 ... cxd4 Or 9 ... a6!? 10 Nd6 cxd4 11 cxd4 f6 intending ... Qb6, applying irritating pressure on White’s position. 10 cxd4 10 Nxd4 f6 11 exf6 Rxf6 12 Nxc6 Nxc6 13 0-0-0 Bd7 14 Nf3 Be8= intending ... Bg6, Manolov-Spasov, Elenite 1992.
On
10 ... Nf5 11 Nf3 f6!? 11 ... a6 12 Nc3 Bd7 13 Be2 Rc8 14 0-0 Qb6=/+. 12 Bd3 12 Be2 a6 13 Nc3 fxe5 14 dxe5 (14 fxe5? Nfxd4! 15 Nxd4 Qh4+) 14 ... d4 15 Nd1 Qb6=/+. 12 ... a6!? 13 Bxf5 axb5 14 Bd3 Qa5! Lilienthal-Mikenas, USSR Ch., Moscow 1940. This is usually assessed as favourable for Black, but it’s probably just equal. the other hand, Black had nice improvements earlier.
8.2112
8
f4
As usual, f4 is rather loosening, but this seems the best moment for it. Quite a lot of theory has accumulated on this rather innocuous variation, so I’ll try to focus on the most principled solutions. 8 ... a6
This has been played the most and has a lot of developed theory. On the other hand, Black has easy equality and good chances of an advantage with other moves; for example, you can take your pick of the following three options. The first one is an eye-opener for those who think 8 f4 has any chance of producing an advantage for White: (a) The little-known 8 ... Bd7(!) seems to solve Black’s problems immediately, and hopes for more: 9 Nd6 (9 dxc5 Bxb5 10 Bxb5 Qc7 with the idea 11 b4 a5 12 c3 b6!=/+; or here or 10 ... b6, since 11 cxb6? Qxb6 favours Black; such variations are good as winning tries because they are thoroughly unbalanced) 9 ... Qb6 (or 9 ... cxd4 10 Nf3 f6) 10 0-0-0 Nbc6 11 Nf3 f6? (a mistake; 11 ... cxd4! 12 Bd3 f6 or perhaps 12 ... h6 puts the burden of proof on White; I’d be surprised if this didn’t give Black the advantage already) 12 exf6! Rxf6 (12 ... gxf6 is more dynamic) 13 Ne5!, Tregubov-Vysochin, Lvov 2011; and instead of the risky 13 ... Nb8?, Black should play 13 ... Nxe5 14 fxe5 Rff8+/=. (b) There’s nothing wrong with 8 ... Nbc6 either:
9 Nf3 (9 dxc5 transposes to 8 dxc5 Nbc6 9 f4 in the notes to 8.2113, a nice line for Black) 9 ... a6 (9 ... Qb6 10 dxc5 Qxc5 transposes to ‘c’) 10 Nd6 Nxd4 11 Nxd4 cxd4 12 Bd3 (12 Qxd4 is a decent alternative, with about the same assessment; 12 ... Nc6 13 Qe3 Qa5+ 14 Qd2 Qxd2+ 15 Kxd2 f6 16 exf6 Rxf6= is just one of many lines) 12 ... f6!? 13 0-0 Bd7 14 Nxb7 Qb6 15 Na5 fxe5 16 fxe5 Ng6!? or 16 ... Qxb2 with equal chances. (c) Finally, 8 ... Qb6 9 dxc5 Qxc5 10 Nf3 Nbc6 (this can also arise from 8 ... Nbc6) 11 0-0-0 Bd7 is logical play from both sides.
Then we have: (c1) 12 Nfd4 a6 13 Nb3 Qb6 14 Nc3 a5=/+. (c2) 12 Nd6 and: (c21) 12 ... Nc8!? 13 Nxc8 (13 Nc4? dxc4 14 Qxd7 c3-/+ Comp Deep Thought-McCambridge, Pittsburgh 1988) 13 ... Raxc8 14 Kb1 a5 15 h4 b5 16 h5 a4 17 h6 (17 a3? b4) 17 ... g6 18 g4!? a3 19 b3 b4 and Black stands no worse at any rate. (c22) 12 ... f6! is perhaps even better; e.g. 13 Nxb7 (13 Qe1?! Ng6-/+ 14 Bd3?! Nxf4 15 Bxh7+? Kxh7 16 Qh4+ Kg8 17 Qxf4 fxe5 18 Qh4 Rxf3 19 gxf3 Qxd6–+ Gooris-Garret, Hammelburg 2009) 13 ... Qb6 14 Nd6 Rab8 15 b3 fxe5 16 fxe5 Nxe5 17 Nxe5 Qxd6=/+. (c3) 12 Nbd4 (best) 12 ... Rfc8 (12 ... Nxd4 13 Nxd4 Rac8 14 Kb1 Ba4!? intending ... Nc6 and ... Bb5) 13 Kb1 a5 14 h4 b5! 15 h5 a4 16 Bd3 a3 17 b3 Nxd4 18 Nxd4 b4=. 9 Nd6 cxd4 10 Nf3 10 Nxc8 Qxc8 11 Nf3 Nbc6 12 0-0-0 Nf5=/+.
10 ... f6
11 Bd3 11 exf6 Rxf6 12 Nxc8 Nxc8!? (or 12 ... Qxc8 13 0-0-0 Nbc6 14 Nxd4 Nxd4 15 Qxd4 Nf5 16 Qe5 Qc5 with perhaps a slight edge) 13 0-0-0 Nc6 (13 ... Nd6 14 Bd3 Nc6 15 Rhe1 Qb6 16 a3 Re8) 14 Bd3!? (14 g3 Nd6 15 Nxd4 Qb6=/+) 14 ... Nb6! 15 Kb1 Na4!. 11 ... Nbc6 12 0-0 Bd7
13 Qe1 This is the move advocated by Dzindzihashvili and Perelshteyn; otherwise White’s central position is falling apart. 13 ... Nc8 13 ... Qb6 14 Qh4 (14 a3 h6 intending ... Nc8 is similar, leaving open the option of ... f5) 14 ... h6! 15 g4 fxe5 (15 ... Nb4) 16 Nxe5? (16 fxe5 Rf4!? and ... Raf8) 16 ... Qc7-/+ 17 g5 Qxd6 18 gxh6 g6 19 Qg5 Be8 20 Bxg6 Nxg6 21 Nxg6 Kh7 22 Nxf8+ Qxf8, Ryan-Fluvia Poyatos, Barcelona 2011, and the attack has fallen short. 14 Nxc8 Rxc8 15 a3 This is played in order to prevent the awkward ... Nb4.
15 ... h6! This looks like the most straightforward solution; Black remains a pawn ahead, neutralizes White’s kingside attack and challenges him to show compensation. Alternatively: (a) 15 ... fxe5 16 fxe5 seems risky at first in view of 16 ... Rxf3? 17 Rxf3 Qg5 18 Rf3! with an attack; however, Black can still play 16 ... h6 17 Qg3 Ne7! (intending ... Bb5) 18 Nh4! (18 a4 Nf5 19 Qg6 Rc6) 18 ... Qc7!? 19 Ng6 Rxf1+ 20 Rxf1 Bb5 21 Qg4 Nf5, and White can get within a fraction of equality by 22 Bxf5 exf5 23 Qxf5! Bxf1 24 Qxc8+ Qxc8 25 Ne7+ Kf7 26 Nxc8 d3 27 cxd3 Bxd3 28 Kf2. (b) 15 ... Qb6 16 Qh4 h6 17 Rab1?! (17 exf6! Rxf6 18 Ne5 Nxe5 19 fxe5 Rxf1+ 20 Rxf1 Rf8=) 17 ... Ne7 18 g4 Bb5 19 exf6 Rxf6 20 Ne5, Pedram-Crouse, corr 2004; and 20 ... Qc7 produces a comfortable advantage.
16 Qg3 Now the try 16 Nh4?! isn’t effective after the reply 16 ... fxe5 17 fxe5 Rxf1+ 18 Bxf1 Be8!-/+. 16 ... Qb6 17 exf6 17 Qg6 f5 18 b4 Ne7 19 Qg3 Bb5. 17 ... Rxf6 18 Ne5 Nxe5 19 fxe5 Rxf1+ 20 Rxf1 Bb5 21 h3 This is all analysis that I did several years ago. Andersson’s move 21 h3 is better than my line 21 Qg6 Bxd3 22 cxd3 Qc6 23 h3 Qe8=/+. 21 ... Bxd3 22 cxd3 Qxb2
Taking chances. 22 ... Rf8 is safe and at least equal. 23 Qg6 Rc1 24 Qxe6+ Kh8 25 Qe8+ Kh7 26 Qf7 Qxa3 27 e6 Rxf1+ 28 Qxf1 b5 29 Qf5+ Kh8 30 Qf7 Qxd3 31 e7 Qe3+ 32 Kh1 Qe1+ ½-½ M.Andersson-H.Graf, ICCF server 2008. This whole line is fairly easy for Black; apart from the established variations, I would have you mark the new move 8 ... Bd7 which, if it works, might be an easy way to limit your study. 8.2113
8
dxc5
White’s idea is to overprotect e5 and clear d4 for a knight. This is the probably White’s soundest line; it doesn’t promise a great deal but minimizes his chances for serious disadvantage. 8 ... Nd7 A good alternative is 8 ... Nbc6 9 f4 (9 Nf3 b6! transposes to 6 Nf3 Nbc6 7 Nb5 Bxd2+ 8 Qxd2 0-0 9 dxc5 b6) 9 ... f6 (although I like 9 ... b6! here too), when Otten analyses 10 exf6 Rxf6 11 Nf3 b6! 12 cxb6 Qxb6 13 c3 (13 0-0-0! Rb8 14 b3 is better, though Black has attacking chances) 13 ... Rb8 14 b3 Ng6 15 g3 e5! 16 Nxe5? (but Black stands better anyway) 16 ... Ngxe5 17 fxe5 Bg4. He says this is winning for Black, and it looks strong after the forced line 18 Nd4 Re8 19 Nxc6 Qxc6 20 Qd4 Rxe5+ 21 Qxe5 Re6 22 Qxe6+ Qxe6+ 23 Kd2 Bf3 24 Bd3 Bxh1 25 Rxh1 Qh3.
9 f4?! Too loosening. Instead:
(a) 9 Nf3?! doesn’t hold up the centre well: 9 ... Nxc5 10 Bd3 (10 0-0-0 Ne4 11 Qe1 Bd7 12 Nbd4 Nc6 13 h4 f5 14 g3 Qc7=/+ Ciolac-Zysk, Vienna 1990) 10 ... Nc6 (10 ... Qb6!=/+) 11 0-0, V.Knox-J.Levitt, British Ch., Dundee 1993; and here 11 ... Bd7 12 Nbd4 Rc8 is level. (b) 9 Qc3 is generally considered best, although Black has a number of good replies:
(b1) 9 ... Nc6 10 Nf3 (10 f4 f6 11 exf6 Nxf6 12 Nd6 d4 13 Qd2 b6! or 13 Qa3 Ng4=) 10 ... b6!? (this is a standard sacrificial motif going back to Bronstein; objectively, however, 10 ... a6 11 Nd6 f6 12 exf6 Qxf6! is better and at least equalizes) 11 cxb6 Qxb6 12 a4 a5 with just about enough counterplay; e.g. 13 Be2 (13 Bd3 Ba6 14 0-0 Rfc8) 13 ... Ba6 14 0-0 (14 b3 Rac8 15 Qb2 Qb8=) 14 ... Rac8 and Black has more chance of getting somewhere than White. (b2) 9 ... f6 looks like a good solution: 10 exf6 (10 Nf3 fxe5 11 Nxe5 b6! 12 Nxd7 Bxd7 13 cxb6 Qxb6 14 Qd4 Qb8=/+) 10 ... Nxf6 11 Bd3 Bd7 12 Nd6? (12 a4 e5!) 12 ... Qc7=/+ with the idea 13 b4 b6 14 Rb1 a5 15 a3 axb4 16 axb4 Rfb8=/+. (b3) 9 ... a6 is dull, but equalizes easily:10 Nd6 Qc7 (10 ... b6 11 c6 d4!) 11 Nf3 (11 b4? a5; 11 f4 Qxc5 12 Qxc5 Nxc5=) 11 ... Qxc5 12 Qxc5 Nxc5 13 c4!? Rd8 (or 13 ... Nc6 14 Rc1=, rather than 14 cxd5?! exd5 15 Rd1 Ne4! 16 Rxd5 Be6 17 Rd1 Nxd6=/+) 14 Rc1 (14 Ng5 f6!) 14 ... Nd7! 15 c5 Nc6=/+. 9 ... Nxc5
10 Nd4 Qb6 Also good are 10 ... Bd7 and 10 ... Nc6 11 Ngf3 Ne4 12 Qe3 Qb6 13 0-0-0 Bd7. 11 0-0-0 Sanz Alonzo-Sion Castro, Salamanca 1990; and apart from 11 ... Bd7, Black had 11 ... Ne4 12 Qe3 Nf5! 13 Nxf5 exf5 with a
small 8.2126
advantage a3
and
plenty Bxc3
of 7
play. Bxc3
This variation is equal, but more promising for White than the alternatives. Black can choose between a somewhat drawish line (this might discourage players of White) or one with mutual chances: 7 ... b6 This is often essayed by stronger players and yields a deep strategic game with chances for both sides. Instead, 7 ... Nbc6 8 Nf3 cxd4 can lead to quick liquidation, though there are still ways to keep things alive: (a) 9 Bxd4 Nf5 (9 ... Nxd4 10 Qxd4 Nc6 11 Qg4 0-0 12 Bd3 f5 was approximately equal in Balashov-Svistunov, Pinsk 1993) 10 c3 (10 Bc3 d4 11 Bd2 is double-edged, and more fun) 10 ... Bd7 11 Be2 Rc8!? 12 0-0 0-0 13 Qd2!? Nfxd4 14 cxd4 Na5!= Savon-Hort, Skopje 1968. (b) 9 Nxd4 Nxe5! (otherwise Black is slightly worse) 10 Nxe6 Bxe6 11 Bxe5 0-0 12 Bd3 (12 Be2 Nc6 13 Bg3 Qf6 14 c3?! d4!=/+; but 14 0-0 is better, and equal) 12 ... Nc6 13 Bg3 (13 Bc3 d4 14 Bd2 Ne5! 15 Bxh7+? Kxh7 16 Qh5+ Kg8 17 Qxe5 Re8 18 Qg3 Bc4+ 19 Kd1 Rc8 with a huge attack, Ljubojevic-Nogueiras Santiago, Wijk aan Zee 1987) 13 ... Qf6
(b1) 14 Rb1 Bf5 offers little for either player (also possible is 14 ... g6!?, e.g. 15 0-0 d4 16 Be4 Rac8 and ... Rfd8; whereas Spassky-E.Vladimirov, Solingen 1991, went 14 ... Rac8 15 0-0 Ne5?! 16 Bxh7+ Kxh7 17 Qh5+ Kg8 18 Bxe5 Qf5 19 Qxf5 Bxf5 20 Rbd1 with a few minor chances for White): 15 0-0 Rfe8 (15 ... Bxd3 16 Qxd3 Rad8= Guedon-Naumkin, Cappelle la Grande 1993)
16 Bxf5 Qxf5 17 Qd2 h6 18 Rfe1 d4 ½-½ Roese-H.Franke, German League 1991. (b2) 14 0-0! Qxb2!? (14 ... Rfe8 is boringly safe, and likely superior) 15 Rb1 Qxa3 16 Rxb7 offers reasonable and potentially exciting play for the pawn: 16 ... Rfe8!? (16 ... Qc5 17 Qh5 h6!?, Hector-Sorensen, Berlin 1993; and White should stay aggressive with 18 Bf4, hoping for 18 ... Nb4 19 Bg6!) 17 Qh5 (17 Re1 Re7!, Tringov-Uhlmann, Skopje Olympiad 1972) 17 ... h6?! (17 ... g6 18 Qg5 Qc3) 18 f4! Re7, Obukhovski-Khasin, USSR 1973; and White could demonstrate how this variation has some punch by 19 f5! Bc8 (19 ... Rxb7 20 fxe6+–) 20 Rxe7 Qxe7 21 Re1 Qd8 22 f6!. 8 Bb5+ The most ambitious move. (a) 8 b4 Qc7 9 Nf3 isn’t timed as well: 9 ... cxb4 10 Bxb4 a5 11 Bd2 0-0 12 Bd3 Ba6 and Black is a little more comfortable, J.Friedman-Gulko, Philadelphia 1993. (b) 8 Nf3 is solid; e.g. 8 ... Ba6 (8 ... 0-0 9 Bd3 Ng6!? has ideas of both ... Nf4 and ... Ba6, so White should play 10 b4) 9 Bxa6 Nxa6 10 0-0 0-0 11 b4 Nb8!? (11 ... cxb4! 12 axb4 Nc7=/+ intending ... Qd7; the queenside light squares, especially c4, are effective as outposts) 12 Qd3 cxd4 (12 ... Nd7=) 13 Nxd4 Nbc6 14 Ne2!? (14 Nxc6 Nxc6 is not so easy because Black has the cfile) 14 ... Rc8 15 f4 h5!? (15 ... Qc7!, with the idea 16 b5 Nd8 17 Bb4 Nb7 and ... Nc5) 16 Be1 Nf5 17 Bf2 Nce7 18 c3 Rc4 with active pieces, Cabrilo-Lputian, Cetinje 1991. (c) 8 Qg4 0-0 9 Nf3 Ba6 10 Bxa6 Nxa6 11 Bd2 cxd4 (11 ... Qc7!?) 12 Bg5!? Qc7 13 Bxe7 Qxe7 14 Nxd4 Rac8= De la Villa Garcia-Bareev, Leon 1995.
8 ... Bd7 9 Bd3 Now at least he doesn’t have to deal with ... Ba6. 9 ... Nbc6 10 Nf3 10 f4 cxd4 (10 ... c4 11 Be2 Nf5 12 g3 b5 intending ... a5 is unclear) 11 Bd2 Ng6 (11 ... 0-0 12 Nf3 f6, Spassky-Mohr, German League 1988; 13 exf6! Rxf6 14 0-0 Qc7 15 Re1 and Black is fine, but he’s defending) 12 Nf3 0-0 (12 ... f6!?) 13 Ng5?! (13 0-0=) 13 ... h6 14 Nf3 Nh4!? 15 0-0 Nxf3+ 16 Qxf3 a5= Makka-Ikonomopoulou, Athens 2011.
10 ... Ng6! I called 10 ... cxd4 11 Bxd4 Nxd4 12 Nxd4 Ng6 equal inPTF3, but 13 Qg4 keeps a little plus, even after 13 ... Qh4! 14 Qxh4 Nxh4 15 0-0 Ng6 16 f4, albeit nothing special. 11 0-0 11 Bxg6 fxg6 12 0-0 0-0 13 dxc5 bxc5 14 b4 Qb6!= Leko-Bareev, Wijk aan Zee 1995. 11 ... 0-0 12 Re1 12 dxc5 bxc5 13 b4! is possible, with a dynamic balance after, for example, 13 ... d4 14 Bb2 Ncxe5 15 Nxe5 Nxe5 16 Bxh7+ Kxh7 17 Qh5+ Kg8 18 Qxe5 Qb6 19 c3!?. 12 ... Qc7?! Black should play 12 ... a5!=. 13 Qd2 Better 13 dxc5 bxc5 14 b4!, with the idea 14 ... d4 15 Bd2 Ncxe5 16 Nxe5 Nxe5 17 Bxh7+ Kxh7 18 Qh5+ Kg8 19 Rxe5+/= intending 19 ... g6?! 20 Rxc5. 13 ... a5! 14 Qe3 Nce7= Mastrovasilis-Shirov, Porto Carras 2011. 8.213
6
f4
This doesn’t seem to be any kind of threat to Black; in fact, it’s risky. 6 ... Nf5!? Somewhat ambitious. Others are easier to play: (a) 6 ... 0-0 7 dxc5 f6! 8 Nf3 Nc6 and 7 Nf3 Nbc6 8 dxc5 f6 transposes to 6 dxc5 Nbc6 7 f4 (see note ‘a’ at the start of 8.21). 7 Nf3 can also be answered by 7 ... c4!? (or 7 ... cxd4 8 Nb5 Bc5 9 b4 Bb6) 8 Ne2 Bxd2+ 9 Qxd2 Nbc6 10 Ng3 b5 with a pawn storm, probably about equal. (b) 6 ... cxd4 7 Nb5 Bxd2+ (or 7 ... Bc5! 8 b4 a6 9 bxc5 axb5) 8 Qxd2 0-0 has been played in a number of contests and White has gotten nowhere; e.g. 9 Nxd4 Nbc6 10 Ngf3 Nxd4 (or 10 ... f6 11 exf6 Rxf6 12 Bd3 Nxd4 13 Nxd4 Qb6 14 c3 Bd7 15 0-0 Nc6=) 11 Nxd4 Bd7 12 Bd3 Qb6 13 Nf3?! (13 c3 Nc6=) 13 ... Bb5 (13 ... Qxb2!? 14 0-0 Qb6+ 15 Kh1 Qc5=/+) 14 Bxb5 Qxb5 15 Kf2!? Rac8 16 Rhc1 h6!? ½-½ Gullaksen-Ogaard, Oslo 1998. Black has most of the potential improvements here. 7 Nf3 7 Nb5? a6! 8 Bxb4 cxb4 9 Nd6+ Nxd6 10 exd6 Qxd6-/+ Santo Roman-Raicevic, Athens 1992. 7 ... cxd4! 7 ... Nc6 8 dxc5 Bxc5 9 Bd3, Hebden-N.McDonald, British Ch., Plymouth 1989; and here simply 9 ... 0-0 is slightly better for Black, since White can’t castle and ... f6 will open new lines. 8 Nb5 Ne3! 9 Qc1 Bc5 10 b4 Be7=/+ 8.22
5
...
Nh6
As explained above, I’m going to give this a limited look, with the emphasis on updating critical lines. If you have my previous edition (PTF3), that gives more detail; and fortunately, Simon Williams’ book,Attacking Chess: The French, has an excellent overview with original thoughts (Williams has played 5 ... Nh6 for a decade). What’s the point of the move? At the potential cost of allowing White to weaken his pawn structure, Black brings flexibility into his range of strategies; for instance, if White plays Nb5-d6, Black can challenge that piece via ... f6 and ... Nf7. By not playing ... Ne7, Black’s queen is free to move to the kingside, and can recapture on ... f6 in any line where White plays exf6. Even the move ... Ng4 sometimes enter the picture (e.g. after dxc5), and one should note that White’s standard sacrifice Bxh7+ with Ng5+ and Qh5 is blocked by the knight. The most obvious negative is that White can at some point play Bxh6, shattering Black’s kingside pawns. As in other French variations with ... Nh6, the disadvantage to so doing is that White forfeits his dark-squared bishop and can easily become weak on the dark squares as a result (White’s centre is constructed on dark squares for example). This normally wouldn’t quite balance out Black’s structural loss, except that here it will have cost White two moves to make this capture: Bd2 and then Bxh6. We’ll look at: 8.221 8.222 8.223
6 6 6
Nb5 Bxh6 a3
Other moves, rather briefly: (a) 6 f4 is even less effective than after 5 ... Ne7; e.g. 6 ... cxd4 (or 6 ... 0-0 7 Nf3 cxd4 8 Nb5 Bc5 9 b4 Bb6; 6 ... Nf5!? transposes to 8.213 above) 7 Nb5 Bc5 8 b4 Be7 9 Nf3 Nc6 10 Rb1 Nf5 11 Bd3 a6! 12 Bxf5 axb5 13 Bd3 Rxa2.
(b) 6 Nf3 and then: (b1) 6 ... cxd4 7 Nxd4 (7 Nb5 Bxd2+ 8 Qxd2 0-0=; e.g. 9 Nbxd4 f6!? 10 exf6 Qxf6 11 Be2 Nc6 12 0-0 e5=) 7 ... 0-0 8 Bxh6 gxh6 9 Bd3 (9 Qd2 Qc7 10 Qe3 f6 11 exf6 Rxf6=) 9 ... f6 10 Qg4+ Kh8 11 Qh5 fxe5 12 Qxe5+ Qf6 13 Qxf6+ Rxf6 14 0-0 Nc6=. (b2) 6 ... Nc6 7 Bxh6 (7 Bd3 is note ‘d’, while 7 a3 Bxc3!? 8 Bxc3 is seen via 6 a3 Bxc3 7 Bxc3 Nc6 8 Nf3 in the notes to 8.223) 7 ... gxh6 transposes to 6 Bxh6 gxh6 7 Nf3 Nc6 (see 8.222); but 7 ... cxd4 deserves a look, one fun line being 8 Bg5 (8 Nxd4 gxh6 9 Bb5 Bd7 10 Bxc6 bxc6 11 0-0 Rb8!?) 8 ... dxc3!? 9 Bxd8 cxb2+ 10 Nd2 bxa1Q 11 Qxa1 Nxd8 12 Bb5+ Nc6 13 Qd4 Be7 with mutual chances. (c) 6 dxc5 can also transpose into other lines, but there are several unique moves: 6 ... Nc6 (a modest solution is 6 ... Nd7 7 Bb5 a6 8 Bxd7+ Bxd7 9 Bxh6 gxh6 10 Nf3 Qc7 and ... 0-0-0; or 6 ... Bxc5!? and 7 Bd3 Nc6 8 Qh5 Nf5 or 7 Bxh6 gxh6 8 Qd2 Nc6 with the idea 9 Qxh6 Bxf2+! 10 Kxf2 Qb6+) 7 Bb5 (7 Nf3 Ng4!?) 7 ... Nf5 8 Nf3 Bxc5=. (d) 6 Bd3 Nc6 (6 ... cxd4 7 Nb5 Bxd2+ 8 Qxd2 0-0 9 Nxd4 f6 10 Ngf3 Nf7= intends 11 exf6?! Qxf6 and ... e5; here 10 ... Ng4 is recommended by Williams) 7 Nf3 Nxd4 (or 7 ... cxd4 8 Nb5 Bxd2+ 9 Qxd2 0-0 10 0-0 f6 11 exf6 Rxf6=/+) 8 Nxd4 cxd4 9 Nb5 (9 Ne2 Bxd2+ 10 Qxd2 0-0 11 f4 f6=) 9 ... Bxd2+ 10 Qxd2 0-0= intending ... f6. 8.221 6 Nb5 Bxd2+ This is similar to 5 ... Ne7 6 Nb5, but the knight on h6 has its advantages.
7
Qxd2
0-0
8 dxc5 Alternatively: (a) 8 f4 has several answers; for example, 8 ... Nc6 (8 ... a6 9 Nd6 cxd4 10 Nf3 f6!=/+ is also good) 9 Nf3 a6 10 Nd6 cxd4 11 Bd3!? f6 12 0-0 Ng4!-/+. (b) 8 Nf3 a6 9 Nd6 cxd4 10 Bd3? (10 Nxd4 f6=/+) 10 ... Nc6 11 Nxc8 Rxc8 12 0-0 f6 13 Rae1 Nf7 14 exf6 Qxf6-/+ LeiberShaked, Berlin 1997. 8 ... Nc6 9 Nf3 b6!? A very common theme, sacrificing a pawn for activity. 10 Qc3 Others: (a) 10 cxb6 Qxb6 11 0-0-0?! Ng4 12 Nbd4 Ngxe5 13 Nxc6 Nxc6=/+ A.Hunt-S.Williams, Witley 1999. (b) 10 Nbd4 Nxd4! 11 Qxd4 Nf5 12 Qb4 (or 12 Qc3 d4) 12 ... Qe7 13 Qg4 bxc5 14 Bd3 c4 15 Bxf5 Qb4+ 16 c3 Qxb2 17 0-0 exf5 18 Qd4 Be6 19 Rfb1 Qc2=/+. 10 ... Bd7 11 Bd3 11 Nd6!? bxc5 12 Nb7! Qb6 13 Nxc5 Ng4! 14 Nxd7 Qxf2+ 15 Kd1 Rfc8! (Kholmov). 11 ... bxc5 12 Qxc5 Ng4!=/+ Kholmov-Nikitin, Moscow 1995. 8.222 6 Bxh6 An obvious and important move. 6 ... gxh6 7 Nf3 On 7 a3, Black has: (a) 7 ... Bxc3+ 8 bxc3 Qa5 9 Qd2 Nc6 10 Nf3 (10 Ne2!?, Frittchle-J.Watson, Los Angeles 1999, and most pointed was 10 ... b6!?; e.g. 11 f4 Ba6 12 Nc1 cxd4 13 Bxa6 Qxa6 14 cxd4 Rg8 15 Kf2 0-0-0=) 10 ... Bd7!? (10 ... cxd4 11 cxd4 Qxd2+ 12 Kxd2 looks
adequate; e.g. 12 ... Bd7 13 Rb1 Na5 14 Bd3 Rc8=) 11 dxc5! Rg8 12 g3 Qxc5 13 Bd3 f6! 14 exf6 (14 Bxh7 Rg7 15 Bd3 fxe5) 14 ... Rf8 15 0-0 Rxf6, and a sample line is 16 Nd4 e5 17 Qe3 Kf7 18 Nb5 d4 19 Qe2 dxc3 20 Rab1 Rd8, which is messy but balanced. (b) 7 ... Ba5
8 Nf3 (8 dxc5 Bxc3+ 9 bxc3 Qa5 10 Qd2 Nd7 11 Nf3 Nxc5 12 c4 Qxd2+ 13 Nxd2 dxc4 14 Nxc4 Bd7= with weaknesses on both sides; here Williams likes 8 ... d4 9 b4 dxc3 10 bxa5 Qxa5, which is complex but looks a good option) 8 ... Qb6 9 Rb1 Nc6 10 dxc5 (10 b4 cxb4 11 Nb5 bxa3+ 12 c3 a2!) 10 ... Qxc5 11 Qd2 d4 12 b4 Qxc3 13 bxa5 Rg8 14 Qxc3 dxc3 15 a6 bxa6 16 Rb3 Rg4 17 Rxc3 Bb7!=. 7 ... Nc6 This isn’t forced – which is important in move order terms, because after 6 Nf3, Black could transpose to 6 ... Nc6 7 Bxh6 gxh6 here, but doesn’t have to because of the options of 6 ... cxd4 or 7 ... cxd4 (see that note above). Therefore, if Black doesn’t need to play 7 ... Nc6 here, he isn’t forced to go into the critical line that follows at all. Instead, he can try 7 ... Qa5, leading to 8 Qd2 Rg8 9 a3 (or 9 g3 Rg4!?, and a likely continuation is 10 a3 cxd4 11 Nxd4 Rxd4!? 12 Qxd4 Nc6 13 Qd2! d4 14 axb4 Qxa1+ 15 Nd1 Qa4! 16 Bg2, and either 16 ... Qb5 for complications, or 16 ... Bd7 17 0-0 0-0-0 18 c3 Be8! 19 Qxh6 dxc3 20 Nxc3 Qxb4=) 9 ... cxd4 (9 ... Nd7!?) 10 Nxd4 Nc6 11 Bb5 Bxc3 12 Qxc3 Qxc3+ 13 bxc3 Bd7 14 Bxc6 (14 Nxc6!? bxc6 15 Bf1 Rb8 16 g3 Ke7) 14 ... bxc6 15 g3 c5 16 Nb3 Rc8=. 8 Bb5 Even if, as indicated above, you needn’t enter this line as Black, I think the play is fascinating and satisfactory.
8 ... Qa5
This is logical. Black could also try: (a) 8 ... cxd4 9 Qxd4 Qa5 10 0-0 Qxb5 11 Nxb5 Nxd4 12 Nfxd4 Kd8 or 12 ... Kd7 is plausible, although Black would only be going for a draw. (b) 8 ... 0-0 9 0-0 cxd4 10 Ne2 Qb6 11 a4! f6 12 Bxc6 bxc6 13 Nexd4 Be7 is playable, but should favour White. (c) 8 ... Bxc3+ 9 bxc3 Qa5 is the best alternative to 8 ... Qa5, one line going 10 Bxc6+ bxc6 11 0-0! (or 11 Qd2 Ba6 with plenty of play) 11 ... Qxc3 12 Rb1 Ba6 13 Re1 cxd4 14 Nxd4 Rg8 15 Re3 Qc4 intending ... c5. 9 Qd2 cxd4 Better than 9 ... Rg8 10 Bxc6+ bxc6 11 0-0 cxd4 12 Nxd4 c5 13 Nb3 Qc7+/=. After 9 ... cxd4 play can continue 10 Nxd4 (10 Qxd4?? Qxb5) 10 ... Bd7 11 Bxc6 (11 Nxc6 bxc6 12 Be2 Qc7=, or 12 ... Rb8 13 a3 Be7) 11 ... bxc6 12 0-0 Bf8!? (Black’s idea is to play ... c5 and ... Bc6, probably with ... Rg8 and ... 0-0-0) 13 f4 c5 14 Nf3 (14 Nb3 Qb6 15 Kh1 0-0-0= intending ... Rg8 and/or ... Bc6) 14 ... 0-0-0 15 f5 Rg8 16 a3 Bc6 17 Qf4 Qc7 18 Ne2 (18 Kh1 h5) 18 ... d4 19 Ng3 Kb8 20 Kh1 exf5 (intending ... f6) 21 Qxf5 Qd7=. 8.2236
a3
6 ... Bxc3 7 Bxc3 Or 7 bxc3 Nc6 (7 ... Nf5 8 Bd3 c4!? – Williams) 8 Nf3 Nf5 9 Bd3 Qc7 10 0-0 c4 11 Be2 (11 Bxf5 exf5 12 Qe2 Be6=) 11 ... Bd7 12 Ne1!? (12 a4 Qa5 13 Qe1 0-0 14 Bc1 f6 15 Ba3 Rf7) 12 ... 0-0-0 (12 ... 0-0) 13 g3 f6 14 f4 Rdf8 15 Ng2 h6= Lobron-Hug, Beersheba 1985. 7 ... b6 7 ... Nc6 might lead to the oddball line 8 dxc5 (8 Nf3 cxd4 9 Nxd4 0-0) 8 ... b6!? 9 cxb6 d4! 10 Bd2 Qxb6 11 Nf3 Ng4 12 Qe2 Qxb2 13 Rc1 Bb7 14 h3 d3! 15 cxd3 Ngxe5 16 Nxe5 Nxe5 17 Qe3 Ng6 with a double-edged game. 8 b4 This is similar to the 5 ... Ne7 6 a3 lines. The alternative 8 Bb5+ results in equality after 8 ... Bd7 9 Bd3 cxd4 (9 ... Nc6 10 Nf3 Qc7!?) 10 Bxd4 Nc6 11 Nf3 Nxd4 12 Nxd4 0-0 13 0-0 f6=. 8 ... Qc7 9 dxc5 9 Nf3 cxb4 10 Bxb4 Ba6. 9 ... bxc5 10 Nf3 cxb4 11 Bxb4 a5! 12 Bd2 Ng4! 13 Bb5+ Bd7 14 Qe2?! 14 a4 Nxe5 (14 ... 0-0) 15 0-0 Nc4 16 Bxc4 Qxc4 17 Ne5 Qd4 18 Re1 0-0 and White has compensation, but he’s not better. 14 ... Bxb5 15 Qxb5+ Nd7 16 0-0 0-0 Attacking the e-pawn and using the open files, A.Martin-Shaked, Schwarzach 1997. 8.35
dxc5
With this move, White emphasizes piece play, planning to develop quickly as compensation for losing his foothold in the centre. I will follow a main, main line (with options on moves 7 and 8), because Black can get the positions almost by force and I think that they are at least okay for him. 5 ... Nc6 This is the most aggressive continuation, introducing the threat of ... d4 at the earliest moment. According to theory, 5 ... Ne7 and 5 ... Nd7 are fine, while 5 ... Qc7 6 Nf3 Nc6 transposes to 6 ... Qc7 below. 6 Nf3
6 Qg4 Nge7 transposes to 5 Qg4 (see 8.41), and 6 Bf4? d4 7 a3 Ba5 8 b4 Nxb4! 9 axb4 Bxb4 10 Bb5+ Bd7 11 Qxd4 Bxc3+ 12 Qxc3 Bxb5 is good for Black. 6 ... Nge7 There are two alternatives of interest, both requiring less study: (a) 6 ... Qc7 is complex and apparently a sound choice. In fact, considering the body of theory on 6 ... Nge7, this might be a good way for practical players to go.
(a1) 7 Bd2 Bxc5 8 Bd3 (8 Nb5?! Qb6; 8 Bb5 Bd7 9 Qe2 Nge7 and 10 0-0 a6 or 10 Na4 Bd4=/+) 8 ... a6 (8 ... Nxe5!? 9 0-0 Nxd3 10 cxd3 Nf6 11 Rc1 Qe7 12 d4 Bb6 13 Na4 Ne4=) 9 Qe2 Nge7 10 0-0 Nd4 11 Nxd4 Bxd4 12 Rae1 0-0=/+. (a2) 7 Bb5 Ne7 (7 ... Bxc5 8 0-0 Bd7 9 Bf4 a6 10 Bxc6 Bxc6 11 Qd3 Ne7=) 8 0-0 0-0 9 Na4!? (9 Re1 Bd7=) 9 ... Nxe5 10 Nxe5 Qxe5 11 c3 Ba5 12 f4 Qf6 13 Be3 Nf5 14 Bd4, Gallagher-Carton, London 1986; and 14 ... Nxd4, as played, or 14 ... Qe7!=/+. (a3) 7 Bf4 (the overwhelming choice of tournament players) 7 ... Nge7 and then: (a31) 8 a3!? Bxc3+ 9 bxc3 Ng6 (9 ... Qa5 10 Qd2 Ng6=) 10 Bg3 Qa5 11 Qd2 Qxc5=. (a32) 8 Qd3 Ng6!? 9 Bg3 a6=/+. (a33) 8 Be2 f6! (8 ... a6=/+) 9 exf6 Qxf4 10 fxe7 Bxc5 and Black’s centre and two bishops more than compensate for the time it takes to recapture on e7. (a34) 8 Bd3 is the only ambitious try: 8 ... d4 9 a3 dxc3 (9 ... Ba5 10 b4 Nxb4 11 axb4?! has been played a lot, but in fact 11 0-0! is a better move – discovered by the engine, of course – when Black in turn might play the dramatic 11 ... Nxd3!? 12 Nb5 Nxf4 13 Nxc7+ Bxc7 with three pieces for the queen and a dynamically balanced game) 10 axb4 cxb2 11 Rb1 Nd5 12 Bd2 Ndxb4 13 Bxb4 (13 Rxb2 Nxd3+ 14 cxd3 Nxe5=/+) 13 ... Nxb4 14 Bb5+ Nc6 (14 ... Bd7 15 Bxd7+ Qxd7 16 Qxd7+ Kxd7 17 Rxb2 a5 18 Ke2 Rhd8 19 c3 Nd5 20 Rxb7+ Kc6 21 Rxf7 Nxc3+ 22 Ke3 Nd1+=) 15 0-0 0-0 16 Qd6 Qa5 17 Bxc6 bxc6 18 Rxb2 Ba6 19 Rfb1 Rad8 20 Qxc6 h6 and perhaps Black is slightly for preference; to begin with, ... Bc4-d5 in combination with ... Rc8 is in the works. (b) 6 ... d4 simplifies a bit much for my taste, but seems just adequate to equalize: 7 a3 Ba5 8 b4 dxc3 9 bxa5 Qxd1+ 10 Kxd1 Nge7 11 Bb5 (11 Bd3 Nxa5 12 Ng5 Nd5 13 Rb1 Bd7 14 Ne4 0-0 15 Nd6 b6 16 cxb6 axb6=; 11 a6 bxa6 and 12 Rb1 Bb7! or 12 Bd3 Bb7 13 Rb1 0-0-0=) 11 ... 0-0 12 a6 (12 Bg5 Nd5 13 Bxc6 bxc6; 12 Ke2 Nxa5 13 Rd1 Nd5=) 12 ... bxa6 13 Bd3 Rb8=. 7 Bd3 Others: (a) 7 Bd2 transposes to the irregular line 5 Bd2 Ne7 6 Nf3 Nbc6 7 dxc5, which gives Black no trouble whatsoever (see note ‘b’ at the start of 8.21). (b) 7 a3 Bxc3+ 8 bxc3 and Black has various equalizers: (b1) 8 ... Qa5 9 Qd2 0-0=. (b2) 8 ... Qc7 9 Bb5 Bd7 10 Qe2 0-0 11 a4 (11 0-0?! Nxe5!) 11 ... Ng6 12 Bxc6 Qxc6 13 0-0 Rfc8 14 Be3 b6! 15 cxb6 axb6, at least equal, although the winning chances are small. (b3) 8 ... Ng6 9 Bb5 (9 Be3 0-0 10 Bd4 Qa5 11 Qd2 b6! 12 cxb6 axb6 13 Be2 Ba6!=/+) 9 ... Qa5 10 Bxc6+ bxc6 11 0-0 (11 Qd4 Ba6=/+) 11 ... Qxc3 (11 ... Ba6 12 Re1 0-0=) 12 Bd2 Qc4=. 7 ... d4 This begins a long forced sequence which ends with mutual chances. The main alternatives are 7 ... Qc7, which involves accepting a rather dangerous pawn sacrifice after 8 0-0; and 7 ... Bxc5, which has been played by some strong grandmasters and has done well in practice. Black usually plays a combination of ... a6, ... Ng6, and at some point ... Qc7; e.g. 8 0-0 (8 Na4 Bb4+!? 9 c3 Ba5 10 Qe2 Bc7 11 0-0 0-0 12 b3 Ng6 13 Bxg6 fxg6 with dynamic equality; 8 Bf4 a6 9 a3 Ng6 10 Bg3 Qc7 11 0-0, Fiore-Rausis, Castellaneta 1998; and the most straightforward approach is 11 ... 0-0 12 Re1 Ba7 13 Qd2 h6=) 8 ... Ng6 9 Re1 a6 10 a3 0-0 11 b4!? Ba7 (or 11 ... Be7) 12 b5 Na5, Lagumina-Naumkin, Porto San Giorgio 1995; and instead of 13 Ng5? Qc7, here 13 Bd2 Bd7 14 Rb1 Qc7 would have been equal or better for Black. 8 a3 Ba5 9 b4 Nxb4 10 axb4 Bxb4
A well-known position to which considerable analysis has been devoted. I’ll just put what seems to me best play so as to streamline things, although a few variations are necessarily deep. 11 0-0 Almost automatic, but occasionally other moves are tried: (a) 11 Bb5+ Bd7 12 Qxd4 Bxc3+ 13 Qxc3 Bxb5=/+ 14 Nd4 (14 Bg5 Qd5=/+) 14 ... Ba6 15 Bg5! (15 c6?! Qd5! 16 cxb7 Bxb7! 17 f3 Qxe5+ 18 Qe3 Qxe3+ 19 Bxe3 0-0=/+) 15 ... Rc8!? (15 ... Qc7) 16 f3 (16 Bxe7!? Qxe7 17 c6 0-0 18 f4 f6!=/+) 16 ... Nd5 17 Qd2 Qc7 18 c6! Qxe5+ 19 Kf2 0-0 20 Rhb1 h6 21 Bh4 Qh5 22 Bg3 e5=/+ Vadljan-Venus, bdf-fernschachbund.de 2009. (b) 11 Nd2!? Bxc3 12 Rb1 Qd5 13 0-0 Bxd2 14 Bxd2 resembles the main line, but Black hasn’t had to spend a tempo on ... h6: after 14 ... Bd7 15 Re1 Bc6 16 Be4 Qd7 17 Rb3, I.Upton-Harley, British League 1996, simply 17 ... Bxe4 18 Rxe4 Qd5 and ... 00 leaves White with awful pawns. 11 ... Bxc3 12 Rb1 h6
The point of this move is to stop Ng5-e4. For reasons of space, I won’t go into the alternatives, except to say that 12 ... Bd7 is playable, but I no longer trust 12 ... Qc7 due to 13 Ng5 Qxe5 14 Qf3! (rather than theory’s 14 Qh5). 13 Nd2 Others: (a) 13 Nh4 Qd5! 14 Qg4! (14 f4?! Bd7 15 f5 exf5 16 Bxf5, Zinser-Ackerman, corr 1964; 16 ... Bxf5 17 Nxf5 Nxf5 18 Rxf5 00=/+) 14 ... Bd7 is equal, and clearer than 14 ... Qxe5. (b) 13 Bb2 Bxb2 14 Rxb2 0-0 15 Be4 Qc7 16 Qxd4 Rd8= Zelcic-Sa.Martinovic, Zadar 2007.
13 ... Bxd2 Black prevents Ne4-d6. 14 Bxd2 Bd7 15 Rxb7 15 Qg4 Bc6 16 Rb4 h5! (16 ... Qa5!? 17 Qxd4 Rd8 18 Qg4?! Qxc5=/+ D.Lima-N.McDonald, Hastings 1988/89) 17 Qxd4 a5=/+. 15 ... Bc6 16 Rb4 Instead: (a) 16 Bb5?! 0-0=/+ (or 16 ... Bxb5 17 Rxb5 Qd5). (b) 16 Rb3 Qd5 17 Qg4 a5!? 18 Rfb1 (18 Bxa5! h5 19 Qg5 f6 20 Qg3 h4 21 Qh3 f5=, or here 21 ... g5!?) 18 ... h5 19 Qg3 (19 Qg5 a4 20 Rb8+ Kd7=) 19 ... h4 20 Qg4 h3=.
16 ... a5 16 ... Qd5 is also sufficient: 17 Qg4 h5! 18 Qg3 (18 Qg5 0-0 19 Rfb1 a5 20 Rb6 Nf5) 18 ... a5 (or 18 ... h4 19 Qg4 h3=) 19 Rb6 h4=. 17 Rb6 Qd5 18 Qg4 Qxe5 Again, 18 ... h5 19 Qg3 (19 Qg5 0-0) 19 ... h4 and ... h3 is fine. 19 Re1 19 Rfb1!? is a sound move, leading to equality after 19 ... Rc8 20 Bxa5 h5; or 19 ... 0-0 20 Bf4 Qd5! with the idea 21 Bxh6 Qxg2+ 22 Qxg2 Bxg2 23 Bg5 f6; or, finally, 19 ... g5!? 20 f4! Qxc5 21 fxg5 hxg5 22 Bxg5 Rc8 23 Qf4 f5 24 R1b5 Bxb5 25 Rxe6 Bxd3 26 Bxe7 Qxe7 27 Rxe7+ Kxe7 28 Qe5+ Kf7 29 cxd3 Kg6, which should be drawn. 19 ... h5! 19 ... Qxc5 20 Rxc6! Nxc6 21 Qxg7 Ke7 (or 21 ... Rf8 22 Bxh6 Ke7) 22 Bg6 Rhf8 23 Bxh6, Gallagher-Steenbekkers, Cappelle la Grande 2002; and Steenbekkers analysed 23 ... a4 24 h4 a3 25 Bxf7 Rxf7 26 Bg5+ Ke8 27 Rxe6+ Ne7 28 Qg8+ Rf8 29 Qg6+ Kd7 30 Rxe7+ Qxe7 31 Bxe7 Kxe7 to a draw. 20 Qh3 20 Rxe5 hxg4 21 Rg5!? a4 22 Rxg4 a3 23 Rb1 e5!-/+. 20 ... Qxc5 21 Reb1 f5 Black apparently has a significant advantage. The whole line with 5 ... Nc6 should serve Black well and there are reasonable options throughout; for example, as early as Black’s 6th and 7th moves. 8.45 Qg4 This aggressive queen sortie mobilizes White’s forces quickly at the inevitable cost of letting his central structure be compromised. 5 ... Ne7
Now 8.41 8.42
White
has
two 6 6
sound
moves: dxc5 Nf3
And two moves that can get him into trouble: (a) 6 a3?! Qa5!? 7 Bd2!? (7 axb4 Qxa1 8 Kd1 cxd4 9 Nb5 0-0!-/+ Jansa-Korchnoi, Luhacovice 1969, 10 Nc7 Na6! 11 Nxa8 Bd7; or 7 Ne2 cxd4 8 axb4 Qxa1 9 Nb5 0-0! 10 Nc7 Na6 11 Nxa8 Bd7 12 Kd1 d3! 13 cxd3 Qa4+ 14 Kd2 Nxb4 0-1 Engholm-Freyer, corr 1968) 7 ... cxd4 8 axb4 Qxa1+ 9 Nd1 Nbc6 (9 ... Nf5! – Moles) 10 Nf3 Bd7 11 Qxg7 Rg8 12 Qxh7 a6=/+ (Uhlmann). (b) 6 Qxg7 Rg8 7 Qh6 (7 Qxh7? cxd4 8 a3 Qa5–+; e.g. 9 Rb1 dxc3 10 axb4 Qa2) 7 ... cxd4 8 a3 Bxc3+ 9 bxc3 Qc7 (this can be compared to the main line Winawer with 7 Qg4 Qc7 8 Qxg7 Rg8 9 Qxh7 cxd4 in the next chapter; except that in the position before us White, crucially, hasn’t picked up the pawn on h7) 10 Ne2 (after 10 Kd1, Tisdall suggests 10 ... Nd7! and I think that’s the way to go) 10 ... dxc3 11 f4 Bd7 12 Rb1 Nbc6 13 Ng3 (13 Qxh7 d4 14 g3 0-0-0 15 Qd3, J.Edwards-Mayo, corr 1986, and 15 ... Na5 ‘preserves all the chances’ – Edwards) 13 ... 0-0-0 14 Bd3 Nd4 15 Qh5 (15 Be3? Rxg3! 16 hxg3 Ndf5-/+; 15 0-0 Ndf5! 16 Nxf5 Nxf5 17 Bxf5 Qc5+! 18 Kh1 exf5-/+ – Uhlmann) 15 ... Qc5!-/+ Sigurjonsson-Uhlmann, Hastings 1975/76. 8.416 dxc5 This can transpose from and is related to 5 dxc5. White attacks the bishop on b4. 6 ... Nbc6 Williams suggests 6 ... Bxc3+ 7 bxc3 Qa5 8 Bd2 Ng6 9 Nf3 Nd7, which is a less ambitious but quite reasonable solution. 7 Bd2 Others: (a) InDWF, I recommended 7 Qxg7, with the forced sequence 7 ... Rg8 8 Qxh7 (8 Qh6 Rg6) 8 ... d4 9 a3 Qa5 10 Rb1 (10 axb4 Qxa1 11 Nce2 is weak due to 11 ... Nxb4 12 Nf3 d3! 13 cxd3 Qb1; e.g. 14 Qe4 Ned5 15 Nfd4 f5! 16 exf6 Nxf6 17 Qf3 Nxd3+ 18 Kd2 Nxc1 19 Nxc1 Qxb2+ 20 Nc2 Nd5-/+, or in this line even 18 ... Nxf2!?) 10 ... dxc3, and now the rather amazing 11 Be3!? (the best try, at any rate).
Now it seems as though everyone in the world found what they thought was a refutation to this line involving 11 ... cxb2+, 11 ... Bxc5, or 11 ... Nxe5, but none of them have been convincing to me (or to my computer). Nevertheless, it now looks to me as though Black can gain the advantage by using this sequence: 11 ... Bd7! (in addition, 11 ... Nxe5 equalizes; e.g. 12 Ne2 cxb2+ 13 axb4 Qxb4+ 14 c3 Qb3 15 Ng3 Ng4; while 11 ... cxb2+ 12 axb4 Nxb4 13 c3 Rg6! 14 Qh8+ Rg8 15 Qh7 Rg6 is a perpetual) 12 Nf3 (12 Ne2 Bxc5 13 b4 Bxb4 14 axb4 Nxb4=/+; 12 Qe4 Nd5 13 Ne2 0-0-0 is winning; e.g. 14 axb4 Qa2 15 Rd1 Ncxb4 16 Nc1 Qxb2 17 Bd3 Bc6 18 Bd4 Nxc2+ 0-1 Z.Ivekovic-Pourramezanali, Vung Tau 2008) 12 ... 0-00 13 Bd3 Bxc5! (13 ... Rxg2) 14 b4 Bxb4 15 axb4 Nxb4 16 0-0?! (16 Nd4 Nxd3+ 17 Qxd3 Rxg2 18 Qe4 Qd5 19 Qxd5 Nxd5 20 h4 is a better try, although still not equalizing after 20 ... Nxe3 21 fxe3 Ba4 22 h5 Rh8) 16 ... Nxd3 17 Qxd3 (17 cxd3 Bc6 18 Qh3 Qd5–+; there’s nothing to be done about ... Qxf3, because 19 Kh1 Rxg2 wins, and 19 Ne1 Nf5! intending ... Rh8 is killing) 17 ... Bc6 18 Qc4 Qc7! (18 ... Qd5-/+) 19 Ne1 Qxe5 with a near-decisive advantage. I don’t have much excuse here, perhaps only that the lines in theDangerous Weaponsbooks are supposed to include the speculative variety, and White has won a clear majority of master games in this line since my book appeared! (b) 7 Nf3 isn’t as disadvantageous as theory has generally held:
7 ... d4 8 Bb5 Qa5 9 Bxc6+ bxc6 10 Qxd4 (10 Qxg7 Rg8 11 Qxh7 Bxc3+ 12 Kd1 Rb8 13 Ng5 Bxb2–+) 10 ... Nf5 11 Qc4 (11 Qe4 Bxc3+ 12 bxc3 Qxc3+ 13 Nd2 Rb8!–+) 11 ... Ba6 12 Qb3 Qb5 13 Kd1 Qxc5 14 Bd2, Banas-Prandstetter, Marianske Lazne 1978; 14 ... Nd4 15 Nxd4 Qxd4 with perhaps a very slight edge. (c) 7 Bb5 Qa5 (or 7 ... 0-0 8 Bxc6 Bxc3+ 9 bxc3 bxc6 intending ... Ba6) 8 Bd2 (8 Bxc6+ bxc6 9 Bd2 Nf5 10 Nge2 h5 11 Qf4 Qxc5 12 0-0 Ba6=/+ Yudasin-Lputian, Simferopol 1988) 8 ... 0-0 9 Bxc6 Nxc6 10 Nf3, Hebden-Raicevic, Vrnjacka Banja 1989; 10 ...
f5 (10 ... Bxc3 11 Bxc3 Qxc5=) 11 exf6 Rxf6 12 0-0 Qxc5 13 a3 Bxc3 14 Bxc3 e5 15 Qh5 g6 16 Qh6 (16 Qh4 Rf4) 16 ... Rxf3! 17 gxf3 Bf5 intending ... d4/ ... Qd5; Black has excellent compensation for the exchange. 7 ... 0-0 7 ... Ng6 is also played; it scores well and should be looked into if Black desires an alternative. After 7 ... 0-0, there are two main moves: 8.411 8.412
8 8
Nf3 Bd3
8 0-0-0 is ineffective due to 8 ... f5! (or 8 ... b6!? and 9 cxb6 Qxb6 or 9 Nf3 bxc5 10 Bd3 Ng6 11 Qh5, when Pedersen suggests 11 ... Qe8!, one idea being 12 Nb5 c4! 13 Nc7 Bxd2+ 14 Rxd2 Qe7 15 Nxa8 cxd3-/+) 9 exf6 Rxf6 10 Bd3 Bxc5 11 Nf3 h6=/+ and Black’s 2:0 central majority ensures an advantage. 8.4118
Nf3
f5
9 Qh4 Others: (a) 9 Qh5 Bd7 (or 9 ... d4) 10 a3 Bxc5 11 Bd3 Nd4!=/+. (b) 9 exf6 Rxf6 10 0-0-0 (10 Qh5 e5) 10 ... e5 11 Qh5 Rf5! 12 Qh4 Bxc3! 13 bxc3 (13 Bxc3 Rf4) 13 ... Qa5-/+ PietzschUhlmann, Aschersleben 1963. (c) 9 Qg3 Ng6 10 Bd3 d4 11 Ne2 Bxd2+ 12 Nxd2 Ngxe5=/+. 9 ... d4 Or 9 ... Ng6! 10 Qxd8 (10 Qh3? Ngxe5 11 Nxe5 Nxe5 12 Qe3 Nc6-/+ Popovych-Mednis, US Ch., New York 1972) 10 ... Rxd8 11 Na4 Bd7=/+. 10 Ne2 10 Nb5 improves, but the safe 10 ... Bxc5 or 10 ... Bxd2+ 11 Nxd2 a6! 12 Nd6 Qa5 with the idea 13 f4 Nd5 seems fine. 10 ... Bxd2+ 11 Nxd2 Qd5!=/+ Estrin-Rittner, corr 1966. There might follow 12 Nb3 Ng6 13 Qg3?! (13 Qh3 Qxe5) 13 ... f4 14 Qf3 Qxe5 15 0-0-0 Qg5!-/+. 8.412
8
Bd3
8 ... Bxc3!? A good option is 8 ... Bxc5 9 Nf3 Ng6 10 Qh5 Nb4=/+, Van Perlo-Stigar, corr 1993, getting rid of the bishop pair. This has been confirmed by games played since the previous edition. 9 Bxc3 9 bxc3 Nxe5 10 Bxh7+ Kxh7 11 Qh5+ Kg8 12 Qxe5 f6=/+ with a clear central superiority. 9 ... d4 10 0-0-0 White’s most obvious attempt to keep the initiative. Others: (a) 10 Bd2 Nxe5 11 Bxh7+ Kxh7 12 Qh5+ Kg8 13 Qxe5 f6 (a common theme) 14 Qd6! e5 15 c3 Qe8!=/+. (b) 10 Rd1 dxc3 11 Bxh7+ Kxh7 12 Rxd8 Rxd8 13 Qh3+ (if 13 bxc3 then 13 ... Rd5 14 Nf3 Kg8 favours Black) 13 ... Kg8 14 Qxc3 Nd5=/+.
10 ... dxc3! 11 Bxh7+ Kxh7 12 Rxd8 cxb2+ 13 Kxb2 Rxd8 Black has the ideas of ... Nf5-h6 ... Ng6, and/or ... Bd7-e8 for defence, if necessary. He will also attack via ... Rd5. Here are some possible lines: 14 Nf3 Kg8 15 Qh5 Ng6 16 Ng5! (16 Re1 Rd5=/+; e.g. 17 Ng5 Ncxe5 18 Qh7+ Kf8 19 h4 b5!?) 16 ... Ncxe5 (16 ... Kf8) 17 Qh7+ Kf8 18 h4 Ng4! 19 h5 (19 Qh5 Rd4! 20 f3 Nf2 21 Rh2 Nd1+ 22 Kc1 Nc3-/+) 19 ... Nf6 20 hxg6 Nxh7 21 gxh7 Ke7 22 h8Q Rxh8 23 Rxh8 b6!=. The 5 Qg4 Ne7 6 dxc5 line is clearly a serious try, but in the end White has to be as careful as Black.
8.42
6
Nf3
Here White tries to hold the d4-square and develop quickly, but Black no longer has to worry about blitzkrieg attacks. 6 ... cxd4 This is the simplest path, although 6 ... Nbc6 has been equally successful. Then 7 dxc5 transposes to the previous section (note ‘b’ to White’s 7th move), and 7 a3 Bxc3+ (7 ... Nxd4!? is a little crazy but apparently adequate; 7 ... Ba5 is also playable) 8 bxc3 0-0 9 Bd3 (what else?) transposes into the 7 Qg4 0-0 8 Bd3 Nbc6 Winawer, but with White playing the inferior move 9 Nf3 (seePTF3) So at a first glance, 6 ... Nbc6 is a promising option, though it would take some work to prove that. 7 Nxd4 Ng6 7 ... Qc7 is also adequate, with a double threat to e5 and c3. White gets approximate equality from 8 Bb5+ (8 Ndb5 may be objectively best: 8 ... Qxe5+ 9 Be2! Nbc6 10 Bf4 Bxc3+ 11 bxc3 Qf6 12 Bg5! Qe5 13 Bf4=, and not 12 Nc7+?! Kd8 13 Qf3 Ng6=/+) 8 ... Nd7! (8 ... Nbc6 9 Bxc6+ bxc6 10 0-0 0-0 11 Nxe6 fxe6 12 Qxb4 c5 13 Qa3 Nf5 14 Bd2 Bb7=) 9 0-0 Bxc3 (9 ... 0-0 10 Nxe6 fxe6 11 Qxb4 Nxe5= Minev) 10 bxc3 0-0 and Black has a minimal edge due to the pawn structure; e.g. 11 f4 (11 Ba3 Nxe5 12 Qg3 f6=/+) 11 ... Qxc3!? 12 Nb3 Qc7.
8 Bd3!? This is the main move. Others don’t seem to equalize: (a) 8 Nf3?! Nc6 (or 8 ... Bxc3+ 9 bxc3 Qc7=/+) 9 Bd2 d4 10 Ne4 Bxd2+ 11 Nexd2 Qa5=/+ Kuzmin-Dolmatov, Minsk 1982. (b) 8 Bb5+?! Bd7 9 0-0 0-0 10 Bxd7 (10 Nxe6?? Bxc3!–+; 10 Bd3 Nxe5 11 Bxh7+ Kxh7 12 Qh5+ Kg8 13 Qxe5 Nc6=/+
Murey-Sigurjonsson, Brighton 1983) 10 ... Qxd7 11 Nf3, Fernando Garcia-Matamoros, Seville 1992; and 11 ... Nc6 was advantageous, but 11 ... Bxc3 12 bxc3 Rc8 still more so. (c) 8 Bd2 0-0 9 Nf3 Nc6 10 0-0-0 f5!? (10 ... Qc7=/+; 10 ... Bc5=/+) 11 exf6 Qxf6 12 Kb1 e5 13 Qg3 Nf4=/+ MovsesianShaked, Zagan 1997. 8 ... 0-0 9 Bxg6 If White tries to maintain e5 by 9 Nf3, Black has 9 ... Nc6, when 10 Bd2 Qc7 forces 11 Bxg6 fxg6 anyway. 9 ... fxg6! By this recapture, Black activates his rook, while also preventing problems based upon Nf3-g5. 10 0-0 10 Bd2 Be7 (10 ... Bc5!?) 11 0-0 Nc6 12 Nf3 Qb6 (12 ... Rf5!?=/+ – Kinsman) 13 Nd1 Bd7 14 Ne3 Rad8 15 b3 Bc8 16 Rad1 Qc7 17 Qg3 b5! 18 Bc1? d4! 19 Ng4 h5-/+ Murey-Kinsman, Paris 1996. 10 ... Be7 11 Be3! 11 Nf3?! Nc6 12 Re1 Bd7 (12 ... Qb6! and 12 ... Rf5 both favour Black) 13 Ne2 Qb6 (threatening ... Nxe5; 13 ... Bc5! is also good) 14 Be3 (14 Nf4 Bc5) 14 ... Qxb2 15 Rab1, Hebden-Matamoros, London 1987, when 15 ... Qxa2! 16 Rxb7 Bc8 was easiest. 11 ... Nc6 12 Nxc6 bxc6 13 Na4 13 b4?! of Murey-Marciniak, Condom 2004, could be answered by 13 ... Rf5! (or 13 ... a5!) 14 Qd4 a5 15 a3 (15 g4 Rf3 16 Kg2 axb4 17 Kxf3 c5-/+) 15 ... axb4 16 axb4 Rb8 17 Rfb1 Rxb4! 18 Rxb4 c5=/+. 13 ... Qc7!? 13 ... Rb8! 14 a3 Qa5 15 b3, Murey-I.Farago, Seefeld 2002; now 15 ... Rb5 intending ... c5 secures a slight plus. 14 Qd4 Rf5!? 14 ... Rb8 15 b3 (15 a3 Rb5!? 16 b4 a5=) 15 ... a5=. 15 f4 Friedel-Bhat, Lindsborg 2004; and instead of 15 ... g5, Black should still be able to maintain the balance with 15 ... Rb8 16 a3 (16 b3 a5=) 16 ... c5!? 17 Nxc5 Bxc5 18 Qxc5 Qxc5 19 Bxc5 Rxb2 20 Bxa7 Rf7 21 Be3 Rxc2 22 Rfc1 Rfc7=.
Chapter
Nine
Winawer
Variation:
Poisoned
Pawn
Variation
1 e4 e6 2 d4 d5 3 Nc3 Bb4 4 e5 c5 5 a3 Finally, we reach the main line of the Winawer. 5 ... Bxc3+ 6 bxc3 Ne7 This move is the traditional one and still clear favourite among players on all levels. Nevertheless, Black has options; e.g. 6 ... Qc7 (examined inPTF3), 6 ... Nc6 – and 6 ... Qa5, which we look at in Chapter 11 (the Portisch-Hook Variation). 7 Qg4
This is the quintessential Winawer variation, the subject of ceaseless testing since Alexander’s win over Botvinnik in 1946. With 7 Qg4, White is trying to exploit Black’s lack of a dark-square defender and provoke weaknesses on the kingside. Given his long-term structural disadvantage (the doubled c-pawns), the first player needs to utilize his bishop pair, and his cramping e5-pawn makes the kingside a natural area in which to proceed. Black will often have a lead in development (particularly when White takes time for Qxg7) and will succeed in eliminating at least part of White’s centre by ... cxd4, with a subsequent attack on e5. As in many modern openings, however, both sides (and especially Black) will frequently play on both sides of the board. What’s more, whether and how White or Black should open the game varies from case to case. 7 ... Qc7 In this edition I’m returning to the old Poisoned Pawn Variation which was featured in the first two volumes. InPTF3I recommended 7 ... 0-0. After countless books, articles, arguments, and correspondences, I still think that’s a valid move. It has become top-heavy with theory, however, and demands extremely accurate play, so it will be refreshing to look into some of the relatively newer lines below. Now I’m going to do an odd thing for the beginning of a chapter, which is to present material at great length which may be entirely irrelevant to you! Specifically, I want to introduce you to some critical variations you should know if you choose 7 ... cxd4!?.
This capture has become a popular choice of contemporary grandmasters. With both 7th moves Black heads for the main line, which can arise by 7 ... Qc7 8 Qxg7 Rg8 9 Qxh7 cxd4 or via 7 ... cxd4 8 Qxg7 Rg8 9 Qxh7 Qc7. Essentially Black is choosing between allowing 8 Bd3 after 7 ... Qc7, or avoiding that line by playing 7 ... cxd4. After 7 ... cxd4, 8 Bd3 can be answered by 8 ... Qa5 rather than 8 ... Qc7. Although I had originally chosen 7 ... cxd4 as my main line, I finally decided to let the reader choose between it and 7 ... Qc7, even though they both seem to be satisfactory with correct play. For one thing, there are a few lines after 7 ... cxd4 in which both sides may have to accept a perpetual check; those lines are difficult (but not impossible) to avoid. On the other hand, a lot of players trying to get the Poisoned Pawn Variation want to avoid 7 ... Qc7 8 Bd3 (apparently including some top players, considering their preference for 7 ... cxd4). Here’s a repertoire you should be able to use after 7 ... cxd4, and which I leave to you expand upon: (a) 8 cxd4 has been played a great deal but doesn’t give Black too much trouble after 8 ... Qc7: (a1) 9 Bd2 Qxc2 10 Qxg7 (10 Ne2 Qg6 11 Qf3 Nbc6 12 Qc3 f6=/+; 10 Nf3 0-0 11 Be2 Nbc6 gives White almost no compensation; 10 Bb5+ Bd7 11 Bxd7+ Nxd7 12 Nf3 Qg6 13 Qf4 Qd3!-/+ Kresoja-Milijanovic, Belgrade 1993) 10 ... Rg8 11 Qh6 Nbc6! 12 Ne2 Nxd4! 13 Nxd4 Qb2 14 Bb5+ Bd7 15 Bxd7+ Kxd7=/+ Zufic-Grigoryan, Rijeka 2010. (a2) 9 Ra2 Nf5 10 Nf3 Nc6 11 Bb2 (11 Bd2? Qb6!; 11 Bd3! h5 and 12 Qf4 Ncxd4, or 12 Qg5 Ncxd4 13 Nxd4 Qc3+ 14 Bd2 Qxd4 15 0-0 Qg4 16 Qxg4 hxg4, and White may have barely enough compensation, Vogt-Kosten, Austrian Team Ch. 2004) 11 ... Qa5+ 12 c3 b6 13 Bd3 0-0 14 0-0 Ba6=/+. (a3) 9 Ne2 0-0 10 c3 b6=/+. (a4) 9 Kd1 h5 (this seems pretty easy; the old solution is 9 ... 0-0 10 Nf3 f6 from Spassky-Korchnoi, Belgrade 1978; if 9 ... h5 falls short, check that out) 10 Qf4 (10 Qxg7 Rg8 11 Qh6 Qc3=/+) 10 ... b6 11 Bb5+ Bd7 (or 11 ... Nbc6 12 Nf3 a5 13 Rb1 Ba6 14 Bxa6 Rxa6 15 Re1 a4 16 h3 Ra8, Savchenko-Shulman, Khanty-Mansiysk 2009; Black stands somewhat better in this position, but at some point he probably needed to play ... 0-0 to make progress; as it was, White equalized) 12 Bd3 Ng6! 13 Qe3! Nc6 14 Nh3 Nce7 15 Ng5 Nh4 (15 ... Rc8=/+) 16 Qh3 Nhf5 17 Bb2 f6 18 Nf3 (18 g4 fxg5 19 gxf5 g4=/+) 18 ... 0-0-0 19 Re1? g5! 20 g4?! hxg4 21 Qxg4 Be8! and ... Bh5 with a virtually winning position, Savchenko-Kamsky, Baku 2009. (b) I can find only one game with 8 Nf3 in my database and no mention of it elsewhere, but as I see it, Black has no obvious way to advantage, and double-edged play results.
Here are a few ideas: (b1) 8 ... Qc7 9 Kd1 (9 Ra2 0-0 10 cxd4 b6 11 Bd3 Nf5 12 Bd2 Ba6) and: (b11) 9 ... 0-0 10 Bd3 (10 cxd4 b6 11 Bg5 Ba6 12 Bxa6 Nxa6 13 h4 Nf5-/+) 10 ... f5 (10 ... Nf5 11 Qh3 f6 12 g4 fxe5 13 gxf5 e4 14 Rg1 e5 15 Bh6 Rf7 is unclear) 11 exf6 Rxf6 12 Qh5 h6 13 Bg5 hxg5 14 Qe8+ Rf8 15 Bh7+ Kxh7 16 Qh5+ Kg8 17 Nxg5 Rf5 18 Qe8+=. (b12) 9 ... Qxc3 10 Rb1 0-0 11 Bd3 Ng6 12 h4!? Nc6 13 Qg3 f5 14 exf6 Rxf6 15 h5 Nge7 16 Bd2 Qc5 17 Bb4 Nxb4 18 axb4 with attacking chances. (b2) 8 ... Qa5 9 Rb1 Qxc3+ 10 Kd1 Ng6 (10 ... 0-0 11 Bd3 Ng6 12 h4 Qc7) 11 h4 h5 12 Qg5 Qc7 13 Bd3 Nc6 14 Bxg6 fxg6 15 Qxg6+ Qf7 16 Qxf7+ Kxf7 17 Bb2 Rf8 18 Bxd4 Kg8 19 Bc5 Rf4 20 Bd6 b6=/+. I can recommend 8 ... Qa5, but this line is hardly forced. (c) 8 Bd3 is the real challenge. Then 8 ... Qc7 transposes into 7 ... Qc7 8 Bd3 cxd4, something Black is trying to avoid by playing 7 ... cxd4. So 8 ... Qa5 is the key move, and the key juncture for the 7 ... cxd4 plan.
Now: (c1) 9 Rb1 Qxc3+!? (9 ... Ng6 10 Nf3 Qxc3+ 11 Bd2 Qc7 with the idea 12 Bb4?! Nc6 13 Bd6 Ncxe5! 14 Bxe5 Nxe5 15 Qxg7 Nxf3+ 16 gxf3 Rf8=/+) 10 Kd1 (10 Bd2 Qc7 11 Nf3 Ng6 12 0-0 Nc6 13 Rfe1 0-0=/+) 10 ... Kf8!? 11 Nf3 b6!? 12 Qh5 h6 13 Ng5 g6 14 Qh4, Vuckovic-Grischuk, Khanty-Mansiysk Olympiad 2010; and 14 ... Nd7! seems to neutralize the attack. (c2) White has almost always played 9 Ne2, and Black in response 9 ... 0-0 (in general, Black should stay away from positions like 9 ... Nbc6 10 Qxg7 Rg8 11 Qf6!).
This is actually a line of the Rustemov Variation which goes 7 ... 0-0 8 Bd3 Qa5 9 Ne2 cxd4. Fortunately, this is not the most critical position in that variation, and Black can hold his own versus a wide variety of possibilities:
(c21) 10 Bxh7+?! Kxh7 11 Qh4+ Kg8 12 Qxe7 Nc6 13 Qg5 dxc3=/+ Rocha-De la Villa Garcia Linares 2002. (c22) 10 Bd2 is logical: 10 ... dxc3 11 Bxc3, and now Macieja-Vysochin, Warsaw 2010, went 11 ... Qc7 12 0-0 (12 h4 Nd7 13 Qg3 Nc5 14 h5 h6 15 Rh4 Nxd3+ 16 Qxd3 b6=/+; 12 f4 Nd7 13 0-0 Nc5), and here I like 12 ... Nd7! 13 Qh5 (as usual, the minicombination 13 Bxh7+ Kxh7 14 Qh4+ Kg8 15 Qxe7 recovers the pawn only at the cost of White’s best piece and attacking chances; a good solution is 15 ... b6 16 Bb4 Ba6 17 Rfe1 Rfe8 18 Qd6 Qxd6 19 Bxd6 Bxe2 20 Rxe2 Rac8 21 g3 Rc4 with pressure on White’s weak pawns) 13 ... Ng6 14 f4 Nc5 15 Rf3 Nxd3 16 cxd3 Bd7 17 Rh3 h6 18 Rf1 (18 f5 Nf4! 19 Nxf4 Qxc3 20 Rf1 exf5–+) 18 ... Qc5+ (or 18 ... d4!? 19 Bxd4 Rfd8) 19 Bd4 (19 Kh1 Qxa3 20 f5 exf5 21 e6 d4 22 Bxd4 Bxe6 23 Bxg7 Kxg7 24 Nf4 Kf6 and the attack falls short) 19 ... Qc2 20 f5 exf5 21 e6 Bxe6 22 Rg3 (22 Bxg7 Kxg7–+) 22 ... Qa4!, with the idea 23 Rxg6 fxg6 24 Qxg6 Rf7 25 Bxg7 Qg4 26 Qxh6 Qxg7 27 Qxe6 Qf6 28 Qxd5 Re8=/+. (c23) 10 0-0 dxc3 11 Ng3 Nbc6 (or 11 ... Nd7) 12 Nh5 (12 Bh6? Ng6) 12 ... Ng6 13 Re1 d4-/+. (c24) 10 Bg5 is the move most top players have tried. This leads to a final split: (c241) After 10 ... Nbc6 11 f4, ‘Black has really disastrous results in practice’, according to Golubev. But he could try 11 ... Ng6 (11 ... dxc3 12 0-0 Ng6 13 Qh5 intending Rf3-h3 may be simply winning!) 12 h4 f5 with the idea 13 exf6 e5! 14 Qg3 exf4 15 Bxf4 Rxf6. (c242) 10 ... Ng6 11 f4 (11 Qh5 Nd7 12 f4, and 12 ... Nc5 13 0-0! Nxd3 14 cxd4 dxc3 15 Rf3! is deceptively powerful, so Black should play 12 ... f5!; e.g. 13 exf6 Nxf6 14 Bxf6 Rxf6 15 0-0 dxc3 16 Rf3 Nf8 17 Nxc3 Bd7=/+) 11 ... Nd7!
12 Bxg6 (12 h4 f5! 13 Qg3 Rf7! 14 h5 Ngf8=/+ Short-Shulman, Ohrid 2001; Black’s advantage was confirmed by later games and analysis) 12 ... fxg6! 13 Qxe6+ Rf7 14 0-0 (14 Qe8+ Rf8 15 Qe6+ Rf7=) 14 ... dxc3 (14 ... Nb6?! 15 Qe8+ Rf8 16 Qe7 dxc3, Svidler-Grischuk, Nalchik 2009; 17 e6! Qa4!? 18 f5 Qe8 19 g4+/=) 15 Qd6 (here 15 Qe8+ may lead to a repetition after 15 ... Rf8 16 Qe6+=; 15 Nd4?! Nf6 16 Qd6 Ne4) 15 ... Nf8 (15 ... Nc5!? 16 Qxd5 Be6! 17 Qd4!, Friedel-Shulman, US Ch., St Louis 2009; 17 ... Bf5! 18 Qxc3 Qxc3 19 Nxc3 Ne6=) 16 Nd4 (16 Be7 Bg4 17 Nd4 Qb6 18 Qxb6 axb6 19 Bd6 Ra4!=) 16 ... Qb6 17 Rad1 Bg4 18 Rd3 Ne6 19 Qxb6 axb6 20 Nxe6 Bxe6 21 Rxc3 d4 and Black stands slightly better, although this will be tough to convert. Just be aware that you are allowing perpetual checks in this line. (c25) 10 Bh6 Ng6 11 h4 is quite dangerous if you play too slowly.
I don’t think there’s any straightforward way to advantage, but there are a couple of equal lines: 11 ... Nc6 (11 ... gxh6!? 12 h5 Nc6 13 f4!+/=) 12 Bxg7 (12 f4 f5! 13 exf6 Rxf6 14 h5 e5=/+) 12 ... Kxg7 13 h5 Ncxe5 14 Qxd4 Kg8 (for the very bold: 14 ... f6 15 hxg6 hxg6 16 Qh4 Rg8 17 Qh7+ Kf8 is apparently about equal!; or 14 ... Qb6 15 hxg6 Qxd4 16 Rxh7+ Kf6, Cabrera-E.Berg, Petra 2007, and White should probably continue 17 Nxd4 Nxd3+ 18 cxd3 Kxg6 19 Rh4 with balanced play) 15 hxg6 Nxd3+ 16 Kd2! fxg6 17 Rxh7 Kxh7 18 Rh1+ Kg8 19 Qh8+ Kf7 20 Rh7+ Ke8 21 Qg7 Qc5 22 Qxg6+ Kd8 23 Qg5+=. c26) 10 Rb1!? dxc3 (10 ... Nbc6 is probably playable, but I find 11 Rb5! particularly irritating, with 11 ... Qa4 12 Qh5 hard to assess after any of 12 ... Nf5, 12 ... h6, or 12 ... Ng6) 11 Qh5 (11 Rb5 Qc7; 11 h4 Nd7 12 f4 Nc5! with two clean extra pawns; 11 Qb4 Nbc6 12 Qxa5 Nxa5 13 Nxc3 b6=/+) 11 ... Ng6 12 Rb4 Nc6 13 Rh4 h6 14 Bxh6 (14 Bxg6 fxg6 15 Qxg6 Ne7 16 Qg4 Nf5 17 Bxh6 Qc7=/+) 14 ... Ncxe5 15 Bxg6 Nxg6 16 Rh3 (16 Nf4 Qb5!) 16 ... e5 17 g4 gxh6 18 Qxh6 Re8 19 Qh7+ Kf8 20 Rf3 Re7 21 Qxg6 Qb6 and Black is more than holding.
Finally returning to 7 ... Qc7,
we’ll
be
examining
9.1 9.2
these
lines:
8 8
Qxg7 Bd3
There are a couple of minor options: (a) 8 Kd1!? 0-0 9 Nf3 f5!? 10 exf6 Rxf6 11 Bd3 (11 Bg5 Rf8 12 Bd3 Nbc6=/+ intending ... c4 and ... e5 – Uhlmann) 11 ... e5! 12 Qg3 c4 13 Be2 (13 dxe5?! Rxf3 14 Bxh7+ Kxh7 15 Qxf3 Kg8 16 Qh5 g6 17 Qg5 Bf5=/+) 13 ... Nbc6 14 h4 (14 Nxe5? Nf5 15 Qf4 Nfxd4; 14 dxe5 Rg6-/+ – Uhlmann) 14 ... Rg6 15 Qh2 Qa5! 16 h5 Rg4-/+ Enders-Uhlmann, Erfurt 1985. (b) 8 Ra2 Nbc6 (8 ... cxd4 9 Qxg7 Rg8 10 Qxh7 Nbc6 11 f4 dxc3 12 Nf3 Bd7=) 9 Ne2 (9 Nf3 cxd4 10 cxd4 Nf5 11 Bd3 h5 12 Qg5) 9 ... 0-0!=/+ intending ... f6 (ECO). 9.1
8
Black
threatens
Qxg7
...
Qxc3+
Rg8
and
has
an
9
eye
on
Qxh7
e5.
In
response,
cxd4
White
has:
9.11 9.12
10 10
Ne2 Kd1
Instead: (a) 10 Qd3 dxc3 tends to transpose, but 10 ... Qxe5+ is independent: 11 Ne2 dxc3 12 Bf4 Qf6 13 Qxc3 Qxc3 14 Nxc3 Nbc6 with the idea 15 Nb5 e5!. (b) 10 Rb1!? is extremely rare, but looks like a legitimate option; for example, 10 ... Qxe5+ (10 ... Nbc6 11 Bf4 Bd7) 11 Ne2 dxc3 12 Bf4 (12 h4!? d4 13 Bf4 Qf5=; 12 Qd3 Nbc6 13 Qxc3 Qe4!?) 12 ... Qf6 13 h4 Nd7!? 14 Bg5 Qe5 with the idea 15 Bxe7 Rh8 16 Bd6 Rxh7 17 Bxe5 Nxe5 18 Nxc3 Ke7= or 18 ... f5=. It’s surprising that we have almost no tests of this idea. 9.11 10 Ne2 This introduces one of the most exciting and controversial of Winawer lines. White has two bishops, a passed h-pawn and, often, an extra pawn. In return, Black gets superior development and activity, play along the g-, h-, and c-files, and ongoing play against White’s weak pawns and king. 10 ... Nbc6 10 ... Qxe5? 11 cxd4 leaves White a pawn up with a healthy centre.
11 f4 11 cxd4? Nxd4! is weak, and 11 h4 dxc3 12 f4 transposes to 11 f4 dxc3 12 h4. Others: (a) 11 Bd2 dxc3 12 Bxc3 Nxe5 wins a centre pawn while tying White’s knight down to defence of c3. (b) 11 Bf4 and: (b1) 11 ... dxc3!? 12 Qd3 Ng6 (12 ... Nxe5 13 Qxc3 Qxc3+ 14 Nxc3 f6!=) 13 Bg3 Ngxe5 14 Qxc3 f6=. (b2) 11 ... Bd7 is more complex: 12 Qd3 (12 Rb1 dxc3 13 h4 0-0-0 14 Qd3 Ng6 15 Bg5 Ngxe5!=/+ Bakali-Uhlmann, Lugano Olympiad 1968) 12 ... dxc3 13 Qxc3 Rc8!? (13 ... 0-0-0, perhaps objectively best, is equal, with ideas of ... d4 and ... Nf5 or ... Nd5 or ... Ng6; then 14 Nd4 Nxd4 15 Qxd4 Nf5 ‘with counterplay’ is given by Psakhis) 14 h4 d4 15 Qd2 Nd5 (or 15 ... Ng6; e.g. 16 Bg3 Ngxe5 with the idea 17 Nxd4 Rxg3! 18 fxg3 Nxd4 19 Qxd4 Ng4!, threatening ... Qxg3+ or ... Qa5+; White is really forced into 20 Qxg4 Qc3+ 21 Kf2 Qxa1 22 Qg8+ Ke7 23 Qg5+ and 23 ... Ke8= or 23 ... Kf8 with a complicated game in store) 16 Bg3 (16 Bh2 Rg4?) 16 ... Rxg3! 17 fxg3 (17 Nxg3 Qxe5+ 18 Qe2 Ne3!) 17 ... Ne3=/+. 11 ... dxc3
Over the past few years this has become a true competitor with the traditional 11 ... Bd7 (which was featured inPTF1andPTF2). I should say that the system introduced by the latter move, which has also undergone a revival over the last decade, strikes me as still playable, and fully satisfactory or better in many of the lines which have been proposed to obtain the advantage against it. This is after 50 years of heavy use at the top levels of the game. Nevertheless, there are always a few lines (out of dozens) in which White is claiming an edge, usually strictly limited; and a group of lines (including those) which continue to be debated at ridiculous depth. So, in my decision to use 11 ... dxc3, I am not only going by the fact that it seems to hold its own very well against all of White’s efforts, but also that such an enormous investment of time is required to learn the 11 ... Bd7 lines. It wouldn’t take too thorough a research project to write a full-length book on this traditional variation, and hence it would be irresponsible, given the level of detail in this book, to use it as part of our repertoire. I have to say that this is sad, considering my lifetime attachment to the traditional Poisoned Pawn; and this in no way means that you can’t still find secrets and wonderful new moves within its complexities. Having said that, you will notice that the theory I present on 11 ... dxc3 is also detailed. The difference is that, in the old Poisoned Pawn variation, White can play perhaps 8-10 fundamentally different set-ups, and several of them are ultra-critical, requiring ingenious tactics for Black to stay in the game. With 11 ... dxc3, Black’s ideas are repeated over and again; for all but a few lines, you really can rely on ‘ideas’ to carry you through new positions. Black has his normal lead in development and activity, but his kingside is shattered, while White has the bishop pair and will pick up either the c- or d-pawn. White tends to work on the dark squares and Black on the light ones; however, that description applies to static lines and frequently breaks down when the action gets going. One thing that I should mention is that, with 11 ... dxc3, Black preserves the option of putting his bishop on b7 in many lines, from where it exerts more influence than on d7, especially after ... d4 is played. On the flip side, 11 ... Bd7 takes only one move to develop instead of two, and in the ... dxc3 lines you will see that Black takes yet another tempo to play ... d4, something he often delays or foregoes completely in the traditional variations. So in our lines, Black’s development isn’t quite as speedy, and he depends more upon positional factors. 12 Qd3 The overwhelming choice, both bringing the queen back for defence and to win the c-pawn before Black is able to defend it. Still, there are a few alternatives of note: (a) 12 h4 is seldom seen without a quick transposition, but it deserves notice:
(a1) 12 ... Bd7!? 13 h5 0-0-0 14 Qd3 (14 h6?! Nf5 gives White difficulties in extracting his queen) 14 ... Nf5 transposes into a traditional Poisoned Pawn position, usually considered equal, but risky: 15 Rb1! (a classic example went 15 h6 Rg6 16 h7 Rh8 17 Rh3 d4! 18 Rb1 Be8! 19 Qf3 Qd8! and Black was definitely better in Bronstein-Uhlmann, Tallinn 1977) 15 ... Kb8!? (15 ... d4 16 Rg1!, intending g2-g4, is an extremely dangerous line called the Tait Variation; it may be playable for Black, but if you want to enter this line you’ll have to do some serious independent study and research) 16 Qxc3 Rg4!? (16 ... Rg7 and 16 ... Rc8 are options), hoping for 17 Qf3? (but 17 Bd2! favours White; e.g. 17 ... Rh8 18 Qf3 Rgg8 19 Rh3+/=) 17 ... Ncd4! 18 Qxg4 (18 Nxd4 Nxd4 19 Qxg4 Qxc2=) 18 ... Nxc2+ 19 Kd2 Rc8! with a very strong attack. Still, Black could use an improvement here, perhaps one of the alternatives on move 15 or 16. The whole line is obscure, but it’s characteristic of some major Poisoned Pawn themes that I haven’t mentioned elsewhere. Rather than abandon it entirely, the above gives the reader some incentive to investigate the Tait Variation, and to look around elsewhere. In general, I don’t mind giving a second option in which Black’s chances look inferior, as long as it’s still unclear and the reader is forewarned. Funny that 12 h4 is played so seldom – I count 9 non-transpositional games out of 4300 or so PPs inMegabase(1600 games after move 11); and as far as I can see at a quick glance none going into the Tait (that seems wrong, but I’ll bet the number is tiny). (a2) 12 ... b6 is an easier way to establish dynamic chances: 13 Nxc3! (13 Qd3 d4 transposes to 9.112 below; 13 Rh3 d4 or 13 ... Ba6 14 Rxc3 Rc8!?; e.g. 15 Nd4 Bxf1 16 Kxf1 Nxd4 17 Rxc7 Rxc7; here 14 ... Bc4 is also interesting) 13 ... Nd4 14 Rh3! Bb7 (14 ... Qc5!?) 15 Nb5 Qxc2!; for example, 16 Qd3 Qxd3 17 Nd6+ Kd7 (17 ... Kf8!?) 18 Bxd3 Bc6 19 Be3 Nef5 20 Nxf5 Nxf5 21 Bxf5 exf5 with a drawn position. (b) 12 Rb1 encourages 12 ... b6, blocking off the b-file; e.g. 13 Qd3 d4 14 h4 Bb7 15 h5 0-0-0 16 h6 Rg6 (remember this standard rook shift versus an onrushing h-pawn) 17 h7 Rh8 18 Rh5 Rg7=/+ planning ... Ng6 and ... Rgxh7, Hindicky-Mego, CCASR email 2005. (c) 12 Nxc3 has only recently been taken seriously, and by some of the strongest players in the game!
Since 12 ... Nxe5?! 13 Nb5 hasn’t worked out well for Black, he has two important moves: (c1) 12 ... Nd4 is the aggressive way to hit c3 and attack along the c-file: (c11) 13 Nb5?! Qxc2 (13 ... Nxb5 14 Bxb5+ Bd7 15 Bxd7+ Kxd7=/+ is also good) 14 Nd6+ Kd7 15 Qd3 Qxd3 16 Bxd3 Rxg2 17 Nxf7 Nef5 and Black’s control of the light squares gives him some advantage. (c12) 13 Qd3 Nef5
14 Kd1?! (14 Bd2 Bd7! 15 0-0-0 Qb6=/+; 14 Bb2 is best: 14 ... Ne3! 15 Nb5! Nexc2+ 16 Kd1 Nxb5 17 Qxc2 Qb6 is equal and double-edged; 14 Rg1 has also been played, but should be met by 14 ... Qc5!, when White has to either retreat the rook to h1 or try 15 g4 b6! 16 Rg2 Ba6! 17 Qh3 Rxg4!! 18 Rxg4 Nxc2+ 19 Kd2 Qf2+ 20 Be2 Nxa1-/+ with the idea 21 Rg8+ Kd7 22 Rxa8 Bc4!, aiming for moves such as d4 or ... Nd4 and Nab3+; Black’s attack is way too strong) 14 ... Bd7 15 a4 Rc8 16 Nb5? (16 Bd2 Nxc2! 17 Kxc2 d4=/+) 16 ... Bxb5 17 axb5 Nxc2 18 Rxa7 Qb6 19 Ra4 Qf2 20 Be2 Rxg2 21 Qf3 Nfe3+ 0-1 Khedkar-Shome, Winnipeg 2003. (c13) 13 Bd2 Bd7 14 0-0-0 0-0-0 with great pressure and more than enough for a pawn; e.g. 15 Qd3 Nef5 16 g3 Kb8 17 Bh3 Rc8=/+. (c14) 13 Bb2 has been the most popular move and may well be best:
(c141) 13 ... Qb6!? 14 0-0-0 Bd7 15 Qd3 Ndf5 16 Nb5 (16 Ne4!? dxe4 17 Qxd7+ Kf8 and Black owns key files and central squares), Nijboer-Wirschell, Antwerp 1996; and here either 16 ... Rc8 with compensation or 16 ... a6 is equal, an example of the latter being 17 Nd4 Rc8 18 Kb1 Ba4 19 Rd2 Rc4 20 Nxf5 Nxf5. (c142) 13 ... Bd7! (developing a new piece) 14 0-0-0 (14 Nxd5 Nxc2+ 15 Kd1 exd5 16 Qxc2 Bg4+ 17 Kc1 Qb6 18 Qa4+ Bd7=/+; White’s king position is airy) 14 ... Ndf5! (14 ... Qb6!? 15 Qd3 transposes to 13 ... Qb6!?) 15 Nb5 (15 Qh3 can be met by the immediate 15 ... 0-0-0 with at least equality, or by 15 ... Qb6 16 Rd2 0-0-0 17 Nd1 Bc6 18 Qb3 Qxb3 19 cxb3 d4= T.LutherRuppert, Wiesbaden 2011) 15 ... Bxb5! (15 ... Qb6!? 16 Nd4 0-0-0=; but here 16 Nd6+! Nxd6 17 exd6 Qxd6 18 g3 gives the bishops some real scope in a position where White will soon catch up in development) 16 Bxb5+ Kf8 17 Bd3 (17 Rhg1 Rc8 18 Bd3 Ne3 19 g3 Qc5=/+) 17 ... Ne3 18 Rdg1! Nc4!?= Bartsch-Hirr, corr BdF 2006; or of course 18 ... Rc8=. (c2) 12 ... a6 is completely sound. It can transpose into the old Poisoned Pawn lines which aren’t bad (see the explanation in ‘c21’): (c21) 13 Qd3 Bd7 (13 ... Na5!? is interesting, in view 14 h4 Nf5 15 Rh3 Bd7 16 h5 0-0-0) 14 Bd2 (here both 14 Rb1 and 14 Ne2 transpose to traditional main lines; however, Black can then claim that he’s bypassed everything that doesn’t involve Nxc3, a category that includes difficult variations with Qxc3, h4-h5, and h3; so anyone who has studied or wants to study the old 11 ... Bd7 12 Qd3 dxc3 13 Nxc3 a6 might be happy to play 12 ... a6) 14 ... 0-0-0 (14 ... Na5=) 15 g3 Nf5 16 Bg2 Na5 (or 16 ... Ncd4!, with the idea of ... Nxc2 and ... d4 in some lines) 17 Bh3 Bc6 18 Bxf5 exf5 19 Qxf5+?! (19 Qd4! Nc4 20 Bc1) 19 ... Kb8 20 Qd3? (20 Ne2 d4 21 0-0 Nc4=/+) 20 ... d4 21 Ne4 f5! 22 Nf2 Bxh1 23 Nxh1 Rh8 24 h4 Nc4 25 Kf1 Qc6 and Black was winning in Kosintseva-Hou Yifan, Hangzhou 2011. (c22) 13 Bd2 Bd7 14 h4!? 0-0-0 15 Qxf7? (15 Rh3! with the idea 15 ... Nf5 16 Qxf5! exf5 17 Nxd5 Qb8 18 Nb6+ Kc7 19 Nd5+, drawing), Battesti-Pieri, Porticcio 2008, and here Black can gain a clear advantage with 15 ... Rdf8! 16 Qh7 Nf5. (c23) 13 Bb2!? Bd7 14 0-0-0 0-0-0 15 Qd3 (this gets in the way of White’s control of c4; 15 Qh4 Kb8 16 Kb1, ThavandiranDvirnyy, Chotowa 2010, is an option, although 16 ... Nf5 17 Qf2 Na5! gives more than enough play for a pawn) 15 ... Na5!
16 g3?! (better 16 Qd4!? Nf5 17 Qa7 Bc6 18 Bd3, or 16 Qf3 Bc6 17 Qf2, although Black has a small edge in both cases) 16 ... Kb8 (16 ... Bc6 17 Rg1 Nf5=/+ is difficult for White) 17 Ne2 (to blockade d4) 17 ... Bb5?! (17 ... Rc8!) 18 Qc3 Qb6 19 Qd4 Qc7 20 Qc3 Qb6 21 Nd4?! Rc8 22 Qb4 Nec6 23 Nxc6+ (23 Qd6+? Ka8 threatens ... Rgd8, trapping the queen, as well as ... Bxf1 followed by ... Nc4) 23 ... Nxc6 24 Qd2, Shirov-Dvirny, Arvier 2010; and 24 ... Na5 retained an advantage, but 24 ... Bxf1! 25 Rhxf1 Na5 intending ... Nc4 would have very strong. Black’s powerful knight contrasts with White’s very bad bishop; for example, 26 Qb4 (26 Qd3 Rc4! 27 Rf3 Rgc8) 26 ... Qe3+ 27 Kb1 (27 Qd2?? Nb3+) 27 ... Nc4 (with the idea ... Rc6-b6, among others) 28 Rd3! Qe2 29 Qe1 Qxh2 and Black has a clear advantage. 12 ... d4
Here the capture 13 Nxd4 is the ‘main’ move; but the others are important and deep enough to deserve their own sections. Thus: 9.111 9.112 9.113 9.114
13 13 13 13
Rb1 h4 Ng3 Nxd4
13 Rg1?! Qb6! (both guarding d4 and preventing Rb1) 14 Ng3 Bd7 15 Ne4 0-0-0 16 Nd6+ Kb8 is a good version for Black of 9.113 (13 Ng3), with which it should be compared. D.Ledger-E.Berg, London 2011, continued 17 Nc4 Qc5 (or 17 ... Qc7=/+) 18 Rb1 and here 18 ... Na5 19 Nd6 was only slightly in Black’s favour, but much better is 18 ... Nc8! 19 Be2 a6 20 Kf2 N6e7!-/+ with the idea ... Ba4 and ... Bxc2, among several others. 9.111 13 Rb1 This is arguably the best move in the traditional Poisoned Pawn variation, but here it allows a couple of interesting answers not available there: 13 ... b6 This is my preference, exploiting the fact that the bishop hasn’t committed to d7. Others are legitimate, if less clear: (a) 13 ... Nf5 defends the d-pawn and eyes White’s weakness at e3. White can try: (a1) 14 Ng3 Bd7!? 15 Nxf5 (15 Ne4 0-0-0 16 Nf6 Rg7=) 15 ... exf5 16 g3 0-0-0 17 Bg2 Be6 18 0-0 Ba2 19 Ra1 (19 Qxf5+ Kb8 20 Ra1 Bc4!) 19 ... Be6=. (a2) 14 Rg1 prepares g4:
14 ... Qd8! (14 ... Qe7!? is objectively fine, but can be drawish: 15 g3 b6 16 Bg2 Bb7 17 Bxc6+ Bxc6 18 Nxd4 0-0-0, and 19 Nxf5 exf5 20 Qxc3 Kb7=, or 19 Nxc6 Qc5 20 Qa6+! Kc7 21 Rf1! Kxc6=) 15 g3 b6 16 Bg2 Bb7 with balanced play; e.g. 17 Kf2 (17 Bxc6+ Bxc6 18 Rb4 a5! 19 Rxd4 Nxd4 20 Nxd4 Qd5 21 Qxc3 Rc8=/+) 17 ... Rh8 18 h4 (18 h3 Na5) 18 ... Na5 (18 ... Qd7!=) 19 Qb5+ Kf8 20 Bh3 Rxh4! (20 ... Kg7!?) 21 gxh4 Qxh4+ 22 Ng3 Qxh3 23 Nxf5 Qh2+!= 24 Kf1 Rc8 25 Nxd4 Qh3+ 26 Kf2 Qh2+=. (a3) 14 h4 b6!? (14 ... Bd7 15 h5 0-0-0 16 Rg1 is the Tait Variation mentioned in the note with 12 h4 above; 14 ... Qa5! is an original approach, with the idea 15 h5?! b6! and ... Ba6, or 15 Qc4 Qd5!? 16 Qxd5 exd5 17 h5 Nh6 intending ... Bf5) 15 h5 a5!? (intending ... Ba6 and ... d3; 15 ... Bb7 16 h6 0-0-0 17 h7 Rh8 is unclear) 16 Qb5! Qd7! 17 h6 Ba6 18 Qxb6 Qd5 is a dreadfully complicated and apparently equal position. (b) 13 ... Bd7 14 g3 (14 Nxd4 Nxd4 15 Qxd4 Nf5 transposes to 9.1142; 14 Rg1 0-0-0 15 g4 Kb8 16 h4 Bc8 17 h5 b6 intending ... Bb7 gives Black decent chances, which says something about the value of playing ... b6 and ... Bb7 straight away, and here 15 ... Be8 16 Rg3 Na5!? would also be entertaining; after 14 h4 0-0-0 15 h5, Tait suggests 15 ... Kb8!? 16 h6 Rg6 17 h7 Rh8 intending 18 Nxd4 Nxd4 19 Qxd4 Nf5, as in the main lines with 13 Nxd4) 14 ... 0-0-0 15 Bg2 Nf5 (15 ... Na5!? 16 Qxd4 Bc6 17 Qf2 Bxg2 18 Qxg2 Nf5 gives Black plenty of play; then 19 Qf3 Qc5! 20 Qxc3 Nc4 21 Rb4 b5 22 Qf3 Nce3 is good enough to level the game) 16 0-0 Kb8
17 Be4 (White lacks an obvious plan; on 17 a4, 17 ... Na5 18 Ba3 Bxa4!? 19 Be4!? Ne3! initiates the lengthy forcing line 20 Bd6 Rxd6 21 exd6 Qxd6 22 Nxd4 Nxf1 23 Nc6+ Kc7 24 Nxa5 Qxd3 25 Bxd3 Ne3 26 Rxb7+ Kd6=) 17 ... b6!? 18 a4 Bc8 (18 ... Na5!?) 19 Ba3 Bb7, Smirnov-D.Petrosian, Moscow 2011; and instead of 20 Rfd1 Na5!, here 20 Bxf5 exf5 holds the balance; for example, 21 Bd6!? (21 Qxf5? Na5-/+) 21 ... Rxd6 22 exd6 Qxd6, which is hard to assess due to White’s exposed king; e.g. 23 Rbd1 Re8! 24 Nxd4 Nxd4 25 Qxd4 Qc6 26 Qd6+ Qxd6 27 Rxd6 Re2=. 14 Nxd4 14 Rg1 Bb7 15 g4?! 0-0-0 16 Qc4, Barbosa-Grover, Chennai 2011; 16 ... Na5! 17 Qxc7+ Kxc7-/+ with all the normal advantages plus domination of the long diagonal and the threat of ... Be4. 14 ... Nxd4 15 Qxd4 Bb7
16 Rb3 Black has classic Poisoned Pawn activity to compensate for the pawn after 16 Bb5+ Kf8=/+, and 16 Rg1 Nf5 17 Bb5+ Kf8 18 Qf2 Rd8 19 Rb4 Qc5=. 16 ... Bxg2!? Slightly speculative. Simply 16 ... Rc8! keeps the cramping pawn alive; if anyone has the advantage, it’s Black. 17 Rxc3 Qxc3+! 18 Qxc3 Bxh1 This strikes me as totally unclear; e.g. 19 Qc7 Nf5 20 Bd2 Kf8 with a dynamic balance.
9.112 13 h4 Another nice thing about 11 ... dxc3 is that White’s plans involving pushing his passed h-pawn tend to be less immediately critical, although Black naturally has to keep an eye on its potential to reach the last rank. 13 ... b6!
Once again this move. Essentially, Black will profit from putting the bishop on this active square when White doesn’t do something specific to stop it (see 13 Ng3 and 13 Nxd4, for example, where Black lacks the time). 14 Nxd4 Others: (a) 14 Ng3 Bb7 (14 ... Ba6! 15 Qxa6 Rxg3=/+) 15 Ne4 0-0-0 16 Nd6+ Kb8 17 Nxf7 Rd5! with counterplay (Moskalenko). (b) 14 Rg1 Bb7 15 g4?! is loosening: 15 ... 0-0-0 16 h5 Kb8 17 Rg3 Qc8!?=/+ intending ... Ba6. (c) 14 h5 Bb7 15 h6 0-0-0 16 h7 Rh8 is dynamically equal, I think, but also very critical and subject to reassessment, so let’s look in depth: 17 Rh3 (17 Rb1 Nd5 18 Rh6 Qe7! 19 g4 f6 20 exf6 Nxf6 21 g5 Nd7=/+ 22 Qh3?! Nc5 23 Ng1 Kb8 24 g6 Qg7 25 Qg3 Ka8 26 Rh5 Ne7!, Pallett-Bell, ICCF email 2009; White is fatally underdeveloped)
17 ... Ng6!? (the only move I’ve seen; perhaps this is the time to clarify matters with 17 ... Nd5!, threatening ... f5, when 18 Nxd4 Nde7! 19 Nb5 Rxd3 20 Rxd3 Qb8 21 Nd6+ Kd7! isn’t attractive, nor is 18 g4?! Ndb4! – fortunately for White, 18 Qc4! f6 19 exf6 Rxh7 20 Nxd4 Nxd4 21 Qxd4 Rxh3 22 gxh3 Rf8 23 Bc4! Qc6 24 Bxd5 exd5 25 Be3 Qxf6= saves his bacon) 18 Kf2
From a mass of interesting analyses I’ll pick these: (c1) 18 ... Nf8 19 Ng3 Kb8 20 Rb1 f6 (this simplifies; there are some double-edged alternatives such as 20 ... Nd7 21 Ne4 Ne7 intending ... Bxe4 and ... Nc5) 21 exf6 Rxh7 (21 ... e5!?) 22 Rxh7 Nxh7 23 Qg6! (23 Ne4? e5! 24 g3 Rf8 has the idea 25 Ng5 Nxf6! 26 Ne6 e4 27 Qc4 Qh7!, winning) 23 ... Rf8 24 Nh5 Qf7 25 Qxf7 Rxf7 26 f5! Nxf6 27 fxe6 (White’s other try is to get a passed pawn by 27 Nxf6 Rxf6 28 g4! exf5 29 g5; Black has no winning chances, but his extra pawns help him keep the draw in hand; e.g. 29 ... Re6 30 Bd3 Ne7 31 Rb4 Be4 32 Rxd4 Bxd3 33 Rxd3 Kb7 34 Rxc3 Rc6=) 27 ... Rf8 28 Bf4+ Ka8 29 Nxf6 Rxf6 30 g3 Rxe6 31 Re1!, Bernadskij-Vysochin, Lvov 2010; and 31 ... Bc8 was sufficient for a draw, but 31 ... Rxe1 32 Kxe1 Bc8 33 Bg2 Bb7= intending ... Nd8 is more straightforward. (c2) 18 ... Nb8 19 a4 (19 Kg1 Na6 20 Rb1 Nc5) 19 ... Na6 (or 19 ... Ba6 20 Qe4 Bb7 21 Qd3 Ba6 etc; rather than 22 Qd1 Nc6, which looks easier for Black to play than White) 20 a5 Nb4!? 21 Qb5 (21 Qd1? d3 22 cxd3 Rxh7!) 21 ... Qc5 (21 ... Nxc2 22 Ra2 Be4 23 axb6 axb6=) 22 Qxc5+ bxc5 23 a6 Be4 24 Ra5 Rd5=. 14 ... Nxd4 15 Qxd4 Nf5 15 ... Bb7 is a good and possibly even better alternative: 16 Rh3! (16 Bb5+ Kf8=/+, because White has to waste a precious tempo defending his g-pawn) 16 ... Bxg2!? (16 ... Rc8 17 Qf2+/=) 17 Rxc3 Qd7 18 Qxd7+ Kxd7 19 Be2 Nd5 20 Bb5+ Ke7 with equal chances due to Black’s superior structure. 16 Bb5+ Bd7 17 Qxd7+ Qxd7 18 Bxd7+ Kxd7 ‘with a strategically better ending for Black’ (Moskalenko), although a pawn is a pawn and White has as many chances as Black:
19 Rh2 19 Kf2 Rg4 20 Be3 Rag8!? (20 ... Nxe3 21 Kxe3 Rxg2 22 Rac1 Rh8 23 h5=) 21 Rag1 Nxh4 22 g3 Nf5 23 Rh3 Ke7 24 Kf3 Nxe3 25 Kxe3 Rd8 26 Rh2= Bologan-Kamsky, Reggio Emilia 2009/10. 19 ... Rg4 20 h5 Rag8 21 h6 Rh8 This is sufficient, but 21 ... Rh4! 22 h7 Rh8 23 Rxh4 Nxh4 24 Kf2 Rxh7 intending ... Rg7 appears to get to a drawn position more easily. 22 Kf2 Rxh6 Black drew in Svetushkin-Giri, Spanish Team Ch. 2011, after the similar 22 ... Rg6 23 h7 Rg7 24 Be3 Rgxh7 25 Rd1+ Kc6 26 Rxh7 Rxh7 27 Rd3 Rg7 28 Rxc3+ Kd5 intending ... Ke4, which only slightly favours White with his extra pawn. After 22 ... Rxh6, Borosova-E.Berg, Gibraltar 2012, continued 23 Rxh6 Nxh6 24 Kf3 Rg8 25 Be3 Kc6, and Black even went on to win following 26 a4?!. Instead, 26 Rd1 Nf5 27 Bf2 Ne7 28 Rd6+ Kc7 29 Be1 at first appears to give White a few chances, but Black’s strong knight outpost and superior pawn structure will hold the draw after 29 ... Rh8 30 Rd3 (30 Bxc3 Rh1) 30 ... Nd5 31 Bxc3 Kc6 32 Bd2 b5. 9.11313 Ng3 Bd7 Because White’s knight is coming so quickly to e4 and d6 or f6, it seems as though Black has no time for the desirable 13 ... b6. Actually, it’s a tribute to the essential soundness of Black’s position that even after 14 Ne4 Rh8 15 Nd6+ Kf8 with the idea of ... Bb7, or ... Nf5, or ... a5 and ... Ba6, Black’s disadvantage is limited. 14 Ne4 Many strong players have waited to develop before rushing the knight forward: 14 Be2 (with the knight on g3, the gpawn is protected)
14 ... 0-0-0 (14 ... Qb6! is a move-order option, preventing Rb1; then 15 0-0 0-0-0 transposes to 14 ... 0-0-0; 14 ... Nf5 is also playable, especially in view of 15 Ne4?! Rxg2 16 Nf6+ Ke7 17 Bf3 Rg6=/+) 15 0-0 (15 Rb1 Na5 16 Qxd4 Bb5 17 Qf2 Bxe2 18 Qxe2 Nc4 19 Ne4 Nf5 20 0-0 Qc6 is unclear, and here 15 ... Nf5 keeps the game level) 15 ... Qb6 16 Ne4 Nd5 (16 ... Kb8! 17 Nd6 Be8=, since White never won his pawn and Black’s centre cramps him; even 16 ... Nf5 is alright, e.g. 17 Nf6 Rg6 18 g4 Nh4 19 Qh3 Rh8=) 17 Nd6+ (17 Qb5 Qxb5 18 Bxb5 Kc7 19 Rb1 a6= Khachiyan-Shulman, US Ch., St Louis 2009) 17 ... Kb8 18 Nxf7 Rdf8 19 Nd6 Nce7 20 Bf3 Bc6 21 g3 (21 a4? Nb4 22 a5 Qc5 23 Qh7, Karjakin-Kamsky, Nalchik 2009; and 23 ... Nf5 24 Nxf5 exf5 25 Ba3 d3+ 26 Kh1 Bxf3 27 Rxf3 dxc2 ends in Black’s favour) 21 ... Nf5! (21 ... Nxf4!? 22 Bxf4 Rxf4 23 Bxc6 Rxf1+ 24 Rxf1 Nxc6 eventually comes out equal) 22 Nxf5 Rxf5=/+ intending ... Bb5.
14 ... 0-0-0 15 Nd6+ Or 15 Rb1 Nxe5 (15 ... b6!?) 16 fxe5 Qxe5 17 Qe2 Bc6 and now:
(a) 18 Ng3 Qxe2+, which is recommended as best by Moskalenko, looks equal. Here 18 ... Qd5 is more exciting; e.g. 19 Qf2 (19 Rb5!? of Oll-Eingorn, Odessa 1989, was answered by 19 ... Qa2, which was fine, but it’s hard not to like Black after 19 ... Bxb5 20 Qxb5 f5!) 19 ... f5 20 Rb4?! (20 Bd3 Qxg2!? 21 Qxg2 Bxg2 22 Rg1 Be4=) 20 ... a5! 21 Rc4 e5! 22 Be2 (22 Nxf5 Qe4+) 22 ... f4 23 Bxf4 (23 Nf1 Rxg2 24 Qh4 Rxe2+ 25 Kxe2 Qxc4+) 23 ... Rdf8 24 Rxc6+ Qxc6 25 Bf3 Rxf4 26 Bxc6 Rxf2 27 Bxb7+ Kxb7 28 Kxf2, Calistri-Bunzmann, French Team Ch. 2007; and most moves are good, but 28 ... Kc6! 29 Re1 Rf8+ 30 Kg1 Nd5! with the idea 31 Rxe5 Ne3 is virtually decisive. (b) 18 Nf2!? hasn’t been used, as far as I know: 18 ... Qd5 (or 18 ... Qxe2+ 19 Kxe2, when Black can play 19 ... Bxg2, or the more aggressive 19 ... e5 20 Rg1 f5 21 Kd1 f4 22 g3 Nd5! with lots of play; finally, 18 ... Qf5 19 g4!? Qd5 20 Rb4 Qa2 21 Rg1 e5 is simply hard to assess) 19 Nd3 Qa2! 20 Rb3 e5! with dynamic equality. 15 ... Kb8
16 Rb1 16 Nxf7 Rdf8 17 Nd6 Nf5 18 Nxf5 Rxf5=/+ is very similar to the main line, and once again there are sacrificial themes such as 19 g3? Nxe5! 20 fxe5 Rxe5+ 21 Kd1 (21 Kf2 Rf8+ 22 Bf4 Rg5!–+) 21 ... Bc6 22 Bf4 (22 Rg1 Be4–+) 22 ... Bxh1 23 Qe2 Rgg5! 24 Rb1 Rgf5 25 Qg4 Rxf4 26 Qxf4 Re3 27 Be2 Qxf4 28 gxf4 d3!, winning a piece. 16 ... b6 16 ... Bc8 has been played in many games, with mixed results; the text is simpler. But if it’s essential to win, 16 ... Bc8 avoids the possible draw by repetition in the following note. 17 Nxf7 Instead: (a) 17 Nb5 Qb7 18 Nd6 repeats moves, unless Black wants to try 17 ... Nxe5!?. (b) 17 g3 Nf5 18 Nxf5 exf5 19 Bg2 Be6=/+ with light-square control, Reinhart-Carpentier, Le Port Marly 2009; 20 0-0 Na5 21 a4 Qc4! might follow. 17 ... Rdf8 18 Nd6 18 Ng5 Rxg5!? (18 ... Nxe5 19 fxe5 Qxe5+ 20 Be2 Rxg5 21 Bxg5 Qxg5 22 Qxd4 Qxg2 23 Rf1 Rxf1+ 24 Bxf1 Qxh2! 25 Qxd7 Qh4+ with perpetual check) 19 fxg5 Be8!, threatening ... Bg6 and the e-pawn; White has to neutralize the attack immediately with 20 Qg3! Bg6 21 Bd3 Nxe5 22 Bxg6 N7xg6, but Black is developed and controls a lot of squares (as well as threatening ... Qc4), so I’d say that it’s equal in theory and difficult for White in practice. 18 ... Nf5 19 Nxf5 Rxf5= 20 g3 20 Qc4 prevents a sacrifice on e5, at least temporarily: 20 ... Qb7! (this stops g3, so White still can’t develop his f1bishop), and a fun line is 21 h3!? Nxe5! 22 Qxd4! Rxg2! 23 fxe5 Re2+ 24 Kxe2 Qf3+ 25 Ke1 Qg3+ 26 Kd1 Qf3+=. 20 ... Nxe5! 21 fxe5 Bc6 A standard Winawer sacrificial sequence; sometimes White has g4 in instead of g3, but the principle is the same. 22 Rg1 Be4! 23 Qxe4! Black stands better after either 23 Qc4 Qxc4 24 Bxc4 Bxc2, or 23 Qe2? d3! 24 Qxe4 d2+ 25 Bxd2 Rxe5!–+. 23 ... Rxe5 24 Qe2 24 Qxe5?! Qxe5+ 25 Be2 Qe4 26 Kd1 Rf8 with pressure. 24 ... Rxe2+ Better 24 ... Rf8!=/+. 25 Bxe2 e5= With chances for both sides, Kovacevic-Bukal, Zadar 2010. 9.114
13
Nxd4
Nxd4
14
Qxd4
Bd7
This 9.1141 9.1142
is
the
all-important
variation.
I’ll
split 15 15
the
material
once
more,
for
clarity’s
sake: Rg1 Rb1
Only two other moves have much value: (a) 15 Qf2 Qc6!? (15 ... Bc6) 16 Qf3 (16 Rg1 Qe4+ 17 Qe2 Qh7!? 18 g3 Bc6 19 Bg2 0-0-0 20 Bxc6 Nxc6 21 Be3 Nd4!=/+ Milliet-Wirig, Nancy 2010, intending 22 Qc4+ Kb8 23 Bxd4 Qe4+) 16 ... Qa4!? (16 ... 0-0-0) 17 Qxc3 Rc8 18 Qd3 Rxc2 19 Bd2 a6 20 Rc1 ½-½ Velimirovic-B.Lalic, Yugoslav Ch. 1988; but 20 ... Rxc1+ 21 Bxc1 Bb5! 22 Qf3! Bc6! definitely favours Black, who will recover his pawn with the better game. (b) 15 Be3 has grown some in popularity:
(b1) 15 ... Nf5 16 Qc5 Qc6! (16 ... 0-0-0) 17 0-0-0 (17 Qxc6 Bxc6 18 Kf2 Nxe3 19 Kxe3 Rd8 20 Rg1 Rd2 21 Rc1 Kf8=, according to Lukacs & Hazai; I still prefer Black slightly). 17 0-0-0 is Moreno Carnero-Matamoros Franco, Suances 1997; and here 17 ... Qb6! 18 Qxb6 axb6 19 Bxb6 Rxa3 20 Kb1 Ra4 21 Bc5 Bc6 gives Black the initiative (Lukacs & Hazai); then 22 Rd3 b6 (22 ... Rc4!?) 23 Bf2! (23 Be3?? Nxe3 24 Rxe3 Ra3–+) 23 ... Rxf4 24 Rxc3 Bxg2 25 Bxg2 Rxg2 26 Bxb6 will be drawn. (b2) 15 ... b6 has a whole range of answers, none particularly threatening. Just a few lines: 16 g3 (16 Rd1 0-0-0 17 Qd6
Nd5! 18 Qxc7+ Kxc7 19 Bc1 Ba4=/+ Fier-Jatoba de Oliveira Reis, Americana 2010; 16 Rg1 Nf5 17 Qe4, Lopez Gomez-Rangel Garcia, corr Cuba 1998; and here 17 ... Bc6! was interesting, with the idea 18 Bb5 0-0-0! 19 Bxc6 Nxe3=) 16 ... Nf5 17 Qe4!? Bc6 18 Bb5 Bxb5! 19 Qxa8+ Ke7 20 Qxg8 Nxe3 21 Qg5+! Ke8 22 Qg8+ Ke7 23 Qg5+=. (c) 15 a4?! Nf5 16 Qf2 Qc6!=/+ and ... Qe4+ is awkward to meet. 9.1141 15 Rg1 Nf5 16 Qf2 Qc6 There have been many games and loads of analysis about 16 ... Bc6, but the queen move gives Black more control over the direction of the play. 17 Bd3 Direct and solid. Instead: (a) 17 Rb1 Qd5 18 Rb4 transposes to the main line of 9.1142 (15 Rb1). (b) 17 g4 has turned out to be one of the few realistic tries to muddy the waters. White tries to drive Black back and consolidate: 17 ... Qe4+ 18 Qe2 is crucial;
Black has at least two good replies: (b1) 18 ... Qd4 19 Rg2 Bc6!? 20 gxf5! (20 Rf2!? is too passive to play for the advantage after 20 ... Nh4) 20 ... Bxg2 21 Bxg2 (at first sight this looks practically decisive, since Be3 is threatened; but ... ) 21 ... Rxg2! 22 Qxg2 0-0-0 23 Qg4! (the only try for advantage) 23 ... exf5! 24 Qxf5+ Kb8 and although it’s not self-evident, thorough analysis shows that this is dead drawn. (b2) Black can avoid that and stay more on the offensive with 18 ... Qd5! and then:
(b21) 19 Qf2 Qe4+ repeats, while 19 ... Nd4 20 Bg2 Qc5 transposes to ‘b24’. (b22) 19 Be3?! Nd4 20 Bxd4 Qxd4 21 Rg3 (21 Qf2 Qe4+ 22 Be2 Qxc2=/+) 21 ... Qxf4=/+. (b23) 19 Qc4 Nd4 20 Qxd5 exd5 21 Kf2 Rxg4!? (21 ... Nxc2!?) 22 Rxg4 Bxg4 23 Bd3 Bf5 24 Be3 Nxc2 25 Bxc2 Bxc2 26 Rc1 Be4 27 Rxc3=. (b24) 19 Bg2! Qc5 20 Qf2 Nd4! 21 Be4 (forced) is complicated. One final split, but I’ll keep the analysis brief: (b241) 21 ... 0-0-0 22 Be3 (22 Rg3? Bc6) 22 ... Bc6! 23 Rd1! Bxe4 24 Bxd4 Qxa3 25 Rg3 Kb8 26 Rxc3 Qa5 and in this fairly standard type of position the chances are equal because Black’s pieces can generate more threats. A strange, not-at-all-forced draw results from 27 Qe3 Bc6 28 Kf2! Rxg4 29 Bxa7+ Kc8 30 Rxc6+! bxc6 31 Bb6 Qxb6! 32 Qxb6 Rxd1 33 Qxc6+ Kd8 34 Kf3 Rgg1 and it turns out to be impossible for White to avoid perpetual or loss of another pawn. (b242) 21 ... Bc6 22 Bxc6+ (22 Be3 Bxe4 23 Bxd4 Qc4 24 Rd1 Rd8 25 h3 Bxc2 26 Qxc2 Rxd4 27 Rxd4 Qxd4 28 Rg3 Qxf4 29 Qxc3 Kf8 results in a balanced game) 22 ... bxc6 (threatening ... Nf3+) 23 Rg3 Qc4 24 Be3 0-0-0 25 h4 Qa4 26 Rc1 Qxa3 27 h5 Qb2=. (c) 18 ... Qa4 tries to create more winning chances but is riskier. Alekseev-Gupta, Moscow 2012, saw 19 Rb1 Nd4 20 Qd3 Bc6 21 Kf2 Nf3 22 Rg3 Rd8 23 Qxc3 Nxh2 24 Rb4 Nxf1! 25 Rxa4 Nxg3 26 Kxg3 Bxa4 27 Qc7 Bc6 28 Be3 Rd7 and Black had sufficient counterplay to draw.
17 ... Qd5 This has become a standard way to reorganize; after the coming ... Bc6, Black will have an effective battery on the long diagonal.
18 Rb1 Others: (a) 18 a4 is similar to the main lines if White plays Ra3/Rxb3, but the bishop can also go to a3 at some point: (a1) 18 ... Bc6 19 Ra3 (19 Kf1 0-0-0=/+ 20 g4 Nd4, Talla-Cech, Pardubice 1991) can legitimately be answered by 19 ... 0-00 (19 ... Rh8 and 19 ... Qa5 are playable but seem to leave White ahead by just a shade) 20 Rxc3 Kb8 21 Ba3, Milman-Tuhrim, Philadelphia 2011, when 21 ... b6 is best, and White can’t seem to establish himself; e.g. 22 g4 Qa5 23 Kd2 Nd4 24 Rg3 Rd5! 25 Bb2 Rgd8 26 Kc1 Bxa4 27 Kb1 (or 27 Rc4 Nxc2 28 Bc3 Qb5 29 Rxa4 Qxa4 30 Qxc2 Qxf4+ 31 Kb2 b5 is equal) 27 ... Nc6! 28 Ra3 Nb4=. (a2) I think 18 ... 0-0-0! is a more efficient solution:
(a21) 19 Ra3 Kb8 20 Rxc3 Bxa4 is comfortable for Black; e.g. 21 Ra3 (21 Qc5!? Qd7!) 21 ... Qc6 22 Bb2 Ne7! (or 22 ... Rh8 23 g4 Nh4 24 Rg3 b6, having ... Bb5 in mind) 23 h4 Nd5 24 h5 b6 25 h6 Nb4 26 Rc3 Qd5=/+ threatening ... Qa2 and intending ... Na2 in several lines. (a22) 19 Ba3 Bxa4 (19 ... Bc6!? 20 Bb4 Nd4! 21 Bxc3 Qc5! threatens ... Nf3+, and then 22 Bxd4 Rxd4 intends ... Qc3+; Black seems to have just enough play for the two pawns) 20 Qc5+ Bc6 21 Qxd5 Rxd5 22 Kf2 Rh8 23 h3 b5. The idea of ... a5 and advancing Black’s queenside pawns holds the balance. (b) A simplifying line is 18 Be3 Nxe3 19 Qxe3 Rxg2 20 Rxg2 Qxg2.
Then R.Robson-Shankland, Milwaukee 2009, went 21 Be4 (21 0-0-0 Qd5 22 Rg1 0-0-0 23 Qxa7 Bc6 24 Rd1 Rd7!= intending ... Kc7 and in some cases ... Qa2) 21 ... Qxh2 22 0-0-0 (22 Bxb7 Rb8 23 Be4 Qh4+ 24 Qf2 Qxf2+ 25 Kxf2 Rb5= intending
... Ra5; e.g. 26 Ke3 Ra5 27 Kd3 Ra4 28 Rb1 Rxa3 29 Rb8+ Ke7 30 Rb3 Ra4 31 Rxc3 f5! 32 exf6+ Kxf6=), and now 22 ... Bc6! was suggested immediately by commentators all over the world, most of whom found 23 Kb1 (or 23 Bxc6+ bxc6 24 Qd4 – here 24 Qc5? Qxf4+ 25 Kb1 Qa4 favoured Black in Barrientos-Pozo Vera, Panama City 2011 – and now the computer move 24 ... a5!, preventing Qb4 in many lines and thus helping when Black’s queen checks on the b-file; e.g. 25 Kb1 Qe2! 26 Qd7+ Kf8 27 Qd6+ Kg7 28 Rg1+ Kh8=) 23 ... Bxe4 24 Qxe4 Qf2 25 Qxb7 Qb6+ 26 Qxb6 axb6 27 Ka2 Ra4 with a drawn ending. It takes some analysis to show how easy this is, but Black has few problems. (c) 18 Bxf5 exf5 19 Be3 0-0-0 20 Rd1 (20 Bxa7?! Bc6 21 Be3 Qb5!? and ... Qb2 gives Black the advantage), Spitz-Debray, Evry 2005; 20 ... Qa2!=; e.g. 21 Bc5 Bc6 22 Bd6 Rg6 23 Qxa7 Rgxd6 24 Rxd6 Qb1+ 25 Kf2 Qxc2+ 26 Kg3=. (d) 18 g3 Nd4 (18 ... Bb5! 19 Bxf5 exf5 20 Qe3 0-0-0 21 Kf2 f6 22 Qxc3+ Bc6=) 19 Rf1 Bc6 20 Be3 Nf3+ 21 Ke2 Qa5 with compensation, Semina-Ludwig, German League 2000. (e) 18 g4?! Nd4 19 Rg3 Qh1+ 20 Qf1 Qxh2 21 Rh3 Nxc2+ 22 Kd1 Qxh3 23 Qxh3 Nxa1 24 Qh4 Rc8 25 a4 c2+ 26 Ke2, R.Jones-D.Mason, High Wycombe 2012; and here 26 ... Rc5! threatening ... Nb3 is extremely strong. 18 ... Bc6 19 Rb3 19 Rb4, remarkably, would transpose into the main line of 9.1142 (15 Rb1) again. 19 ... 0-0-0 20 Rxc3 Kb8 This position has been tested and analysed a lot, and it seems that Black’s activity and light-square control is enough for two pawns, especially since White’s king is exposed.
21 Rc5 There are many complex alternatives here, and this one is probably no better nor worse than the others. Although correspondence games in this variation almost inevitably turn into draws, the complications illustrate that both sides will have winning chances over-the-board: (a) 21 Qc5 Rxg2 22 Qxd5 Rxg1+ 23 Kf2
23 ... Rdg8!? (brilliant, but perhaps not as precise as 23 ... Bxd5 24 Kxg1 Rg8+ 25 Kf1, when 25 ... Rh8 26 Bxf5 exf5 will undoubtedly be drawn; while 25 ... Ne7 intending ... Rg2 should work as well) 24 Qd8+ Rxd8 25 Kxg1 Rg8+ 26 Kf1 (almost the same position, but the bishop was better on d5) 26 ... Nh4!? (or 26 ... Rh8 27 Bxf5 exf5 28 h3 Kc8 29 Bd2 Kd7 30 Be1 Ke6 with practical equality) 27 Be2, Muzychuk-Lahno, Moscow 2010; and 27 ... Nf5 28 Bd2 Bg2+ 29 Ke1 Bd5 makes it hard for White even to mount a winning attempt. (b) 21 Rc4 b6 22 g4!? Bb5 23 Rc3 Nd4 24 Bxb5?! (24 Rg3! Qh1+ 25 Qf1 Qxh2 26 Rh3 Bxd3 27 Rxh2 Bxf1 28 Kxf1 Rxg4 29 Rd3! and White manages to equalize) 24 ... Qxb5 25 Qf1 Qa5 26 Bd2, S.Kuipers-Stellwagen, Dutch Team Ch. 2011, and now 26 ... Qd5 27 Qg2 Qa2 28 Qe4 Rd5! gives Black excellent chances. (c) 21 g4 Nd4 22 Rg3 Qh1+ (22 ... Qa5!?) 23 Bf1 b6 (23 ... Ka8 24 Bb2 Be4!?) 24 a4 Qe4+ 25 Rge3 Qd5 26 Red3 Qa5 27 Ba3 Nb5! 28 Qd2 Rxd3 29 Rxd3 Nxa3 30 Rxa3 Qxd2+ 31 Kxd2 Rxg4 32 Ke3 Rh4 33 h3 a5 34 Ra1 ½-½ Soldano-Girard, ICCF 2010; according to the computer, this was a perfectly played game!. (d) 21 Bd2 Qa2 22 Qc5 (22 Rxc6 Nd4!! 23 Ba5! Qxa3 24 Bc7+ Ka8 25 Kd1 Nxc6 26 Bxd8 Rxd8=/+) 22 ... Qa1+ 23 Kf2 (23 Ke2 Nd4+ 24 Kf2 Rxg2+! 25 Rxg2 Qd1! 26 Bf1! Qxd2+ 27 Kg1 Qxf4 and Black stands better) 23 ... Rxg2+! 24 Rxg2 Qd1! 25 Be2 Qxd2 26 Rd3?! (26 Rg4! may still equalize) 26 ... Rxd3 27 cxd3 Qxf4+ 28 Kg1 Bxg2 29 Qf8+ Kc7, Gallagher-Nemet, Swiss Team Ch. 1995; and now 30 Kxg2!= was correct. (e) 21 g3 Nd4 22 Be3 Nf3+ 23 Kd1 Nxg1!? (23 ... Rh8! equalizes; e.g. 24 Bxa7+ Ka8 25 Rg2 Qb5! 26 Kc1 Qa5 27 Rxc6 Qxa3+ 28 Kd1 Rxd3+ 29 cxd3 Qxd3+ 30 Kc1 bxc6 31 Bc5 Qc3+ 32 Qc2 Qa1+=) 24 Bxa7+ Ka8 25 Qxg1 Bb5 26 Bb6 Bxd3! 27 cxd3 (27 Rxd3 Qb5= Pacher-Petrik, Stiavnica 2011) 27 ... Qb5 28 Kc2 Rc8 29 Bc7 Qa4+ 30 Kb2 Qxf4 31 Rc4 Qd2+ 32 Ka1 Rxc7 33 Rxc7 Qxd3; this should be drawn. 21 ... Qa2
22 Rxc6!? A safe decision; with the bishop pair, White can make enough threats on the queenside to draw. There wasn’t much else to do in terms of winning attempts; e.g. 22 Qd2 Nh4 23 Rxc6 Nf3+ 24 gxf3 Rxg1+ 25 Kf2 Rxc1 26 Qa5 Rd7 27 Rc8+! Kxc8 28 Qc5+ with perpetual check! After 22 Rxc6, Ganguly-Shulman, Ningbo 2011, continued 22 ... bxc6 23 Qc5 (23 Qd2 Rd4 24 g4 Qb1 25 Kf2 Ne7=) 23 ... Rxd3 (heading for the draw; 23 ... Rd5 might do the same) 24 cxd3 Rxg2 25 Rxg2 (25 Qf8+ Kc7 26 Qxf7+ Kb6 transposes) 25 ... Qxg2 26 Be3 (26 Qf8+ Kb7 27 Qxf7+ Kb6, and not 28 Qxe6?? Nd4, but 28 Qf8 Qg1+ 29 Kd2 Qe3+ 30 Kd1 Qxd3+=) 26 ... Nxe3 27 Qxe3 Qh1+ 28 Ke2 Qxh2+, and the game was drawn shortly through perpetual check. 9.1142
15
Rb1
This is a transposition from the traditional line 11 ... Bd7 12 Qd3 dxc3 13 Rb1 d4 14 Nxd4 Nxd4 15 Qxd4. These days we get the position more often by our 11 ... dxc3 move order, but in any case a good deal of theory has built up on it. 15 ... Nf5 16 Qf2 Qc6 I’m not generally recommending ... Bc6 systems, which is another wide and unresolved subject, but I find this particular one promising:
16 ... Bc6 17 Rg1 Rd8! (17 ... 0-0-0 and 17 ... Qd7 are the more normal moves, currently looking a tad better for White), and now:
(a) 18 Bd3? (natural, but ... ) 18 ... Rxd3! 19 cxd3 Qd8! and: (a1) 20 Qxa7 Qxd3 21 Qb8+ (21 Rb4 Qc2=/+ 22 Be3?? Kf8 was simply winning in Palkovi-F.Portisch, Eger 1987) 21 ... Ke7 22 Rxb7+ Bxb7 23 Qxb7+ Kf8=/+. (a2) 20 g4 Qxd3 (20 ... Nd4!? 21 Rg3 Qd5 seems to yield a small advantage) 21 gxf5 Qe4+ 22 Kf1 Qd3+ 23 Ke1 Qe4+ 24 Kf1 ½-½ Dolzhikova-Sedina, St Petersburg 2009. (b) 18 Qxa7 Nd4 19 Bd3 Qe7! 20 g3 Rh8 21 h4 Rxh4! 22 gxh4 Qxh4+ leads to at least a perpetual check. (c) 18 Rb4 has several answers:
18 ... Qa5 (or 18 ... Rd5!? 19 Be2!? Qd8! 20 Rc4 Nd4 21 Rxc3, Rytshagov-Guanciale, Saint Vincent 2001; and 21 ... Nxe2 22 Qxe2 Bb5 is equal after 23 Qf3 Bc6 24 Qe2 Bb5 or 23 Qh5 Qb6 24 Be3 Qa5 25 Bd2 Qb6 26 Be3=; 18 ... Rh8 has 19 Bd3 Rxd3! in mind, whereas 19 h3 Rg8! 20 g4 Qd7 21 Be2 Nd4 is complex) 19 Bd3 Rh8 (19 ... Qd5; 19 ... Rxd3!? 20 cxd3 Qd5 is also interesting) 20 h3 Rxd3!? 21 cxd3 Qd5= 22 g4 Nh4 23 Ke2 Ng2 (23 ... b5!?) 24 Rxg2, Adelseck-Vicoli, corr 2000; and Black didn’t go in for the wild 24 ... Rxh3!? 25 Rg3 Rh2!! 26 Qxh2 Qa2+ 27 Bb2 cxb2 28 Ke3 b1Q 29 Qh8+ Kd7 30 Rxb1 Qxb1, when White cannot escape a perpetual check. 17 Rb4!? This dynamic move stops ... Qe4+ and covers key squares along the fourth rank; ironically, however, it can sometimes be a waste of time. The immediate 17 Rg1 allows a draw: 17 ... Qe4+ (17 ... Qd5!? 18 Bd3 returns to the main line of 9.1141 above) 18 Qe2 Qd4=. After 19 Rh1?! Bc6 20 Rb4, as in Zunker-Müller Ludwig, German League 2009, 20 ... Qd7! is particularly strong, because ... 0-0-0 is prepared, and 21 Qd3 can be answered by 21 ... Qe7!, when ... Qh4+ and ... Rd8 are problems for White. 17 ... Qd5
18 Rg1 Or 18 Qf3 Qxf3 19 gxf3 Bc6, Singer-Liepold, Kaufbeuren 1998, with obvious compensation in view of White’s pawn structure. Then 20 Kf2 Nh4 21 Rc4 Bb5! 22 Rxc3 Bxf1 23 Kxf1 Rd8 24 Be3 Rg2 gives Black enough for the two pawns, shattered as they are: 25 Bf2 Rd1+ 26 Be1 Ke7 27 f5! (27 Rc7+ Kd8 28 Rxf7 Rgd2 29 Rg1 Rxe1+ 30 Kxe1 Nxf3+ 31 Kf1 Nxh2+=) 27 ... Rgd2=. 18 ... Bc6 19 Bd3
Instead: (a) 19 Rc4 0-0-0 20 Bd3 Kb8 transposes. (b) White loses time with 19 Be2?! Qa2! (19 ... 0-0-0 20 Bf3 Qd7 21 Qc5 Nh4 or 21 ... a5) 20 Bd3 0-0-0=/+. (c) 19 g4 Rd8!? (19 ... 0-0-0) 20 Bd3 Nd4 21 Rxd4 Qxd4 22 Qxd4 Rxd4 23 h4 Kf8! 24 Kf2 Kg7 justifies playing ... Rd8, although this is still very complex. 19 ... 0-0-0!
Remarkably, few players have chosen this move over the years, the great majority preferring 19 ... Rd8, which leaves White with much the better game and has scored very poorly. Presumably they were afraid of 20 Qxa7, which is basically a blunder. Incidentally, this position can arise from the strange order 16 ... Bc6 17 Rg1 0-0-0 18 Bd3 Qa5 19 Rb4 Qd5. 20 Rc4 After 20 Qxa7? Rxg2 21 Rxg2 Qxg2-/+, the white king is very hard to defend; Black’s rook can come into play via h8. 20 ... Kb8
21 Rxc3 A position that occurred way back in V.Liberzon-Pietzsch, Leipzig 1965! I keep coming across old references to the 11 ... dxc3 lines, which has shattered my illusion that the ideas were recently conceived. More importantly, White is now a tempo behind on the same lines in which he plays Rb1-b3 and Rxc3 (as in 9.1141), rather than Rb1-b4-c4 and Rxc3. 21 ... Qa2!? The same move Black uses with a tempo less. I’m not sure that anything is good enough to force an advantage, but a
cute line is 21 ... b6 22 Rb3 Rh8 23 h3 Rhg8 24 Bd2 Qxb3! 25 cxb3 Rxd3 and Black stands better in spite of his material disadvantage. 22 Be3 Others: (a) 22 Bxf5 exf5 23 Qc5 Qb1 24 Ke2 Rxg2+ 25 Rxg2 Qxc1 26 Rd3 Rc8 27 Rg1 Qxf4 28 Qd4 Be4= Hyldkrog-Pasko, ICCF 1999. (b) 22 Qc5? Qa1 23 Kd2 Nd4 24 a4, Strobel-Cech, German League 1997; and among other moves, 24 ... Qb1 and 24 ... Bxg2 are advantageous. (c) Curiously, 22 Rxc6!? bxc6 transposes to the main line of 9.1141. 22 ... Nxe3 23 Qxe3 Rxg2 24 Rxg2 Bxg2 Mamedyarov-Alekseev, Ohrid 2009. White’s weaknesses cancel out his extra pawn; the game continued 25 Kf2 Bc6 26 Be4 (White’s king is a little exposed, so he chooses the safest defence) 26 ... Bxe4 27 Qxe4 Qb1 28 Re3?! (28 Qe1 Qb6+ 29 Kf3=), and here 28 ... Qd1! has the twin ideas of ... Rc8 and ... Rh8, when 29 Re1 Qh5 30 Rh1 Qh3 keeps the pressure on. 9.12 10 Kd1!? Euwe’s move, both avoiding the piece congestion that 10 Ne2 brings and planning Nf3-g5. White’s king is stuck in the centre, of course, but it often stays there in the 10 Ne2 lines anyway. This is a complicated, underestimated line, since several of the solutions which have been recommended for Black in dozens of books aren’t very good at all. It turns out that Black can’t avoid entering into some very dense theory. 10 ... Nbc6
A famous old line goes 10 ... Nd7 11 Nf3 Nxe5 12 Bf4 Qxc3 13 Nxe5 Qxa1+ 14 Bc1, and the analysis of both 14 ... Rf8 and 14 ... d3 goes on in depressing length; I believe there are draws there for Black if you desperately need one. 11 Nf3 11 f4 shores up the centre but is slow: 11 ... Bd7 12 Nf3 (Williams suggests 12 Qd3 ‘!’, but Black has an excellent game following 12 ... dxc3 13 Qxc3 d4 followed by ... Nf5 and ... 0-0-0) 12 ... 0-0-0! 13 Qd3 (13 cxd4? Nf5; e.g. 14 c3 Rh8 15 Qxf7 Ncxd4–+ intending ... Ba4+) 13 ... dxc3 14 Qxc3 d4 15 Qc5 (15 Qc4 Nd5 16 Bd2, Vasiukov-Krasnov, Cheboksary 1960; and among other things, 16 ... Nb6 17 Qd3 Ne7! is strong, with the idea 18 Nxd4 Ba4=/+) 15 ... Nd5 16 Bd2, Kosenkov-Novotelnov, USSR 1961; and the break 16 ... f6! 17 exf6 Nxf6, intending ... Ne4 or ... e5, gives Black a distinct advantage. 11 ... dxc3
A position with remarkably many tactical possibilities. 12 Ng5 The most forcing line. A couple of the alternatives are important: (a) 12 Bh6? Bd7 (or 12 ... Qb6) 13 Ng5 Nxe5 14 f4 f6 15 fxe5 fxg5 with an extra pawn and far superior position. (b) 12 Bg5 Bd7 (12 ... Qb6 13 Ke1 Bd7) 13 Bf6 0-0-0 (13 ... Qb6!, hoping for 14 Ng5? Rxg5 15 Bxg5 Qd4+ 16 Kc1 Qxe5–+) 14 h4 (14 Rb1=), Lukin-Cherepkov, Leningrad 1983; and 14 ... Nf5! (14 ... Qb6=/+) 15 Bxd8 Rxd8, with the ideas ... Nxe5 and ... Qb6, gives Black a substantial advantage. (c) 12 h4 Bd7 13 Ng5 (13 Bf4 Qb6!=/+) 13 ... Rf8 (13 ... Nxe5!? 14 Bf4 f6 15 Qh5+ N7g6 16 Bd3 0-0-0, or 14 f4? Ng4! 15 Qxf7+ Kd8 intending ... Qc5 and ... Nf5) 14 f4 Nf5! 15 g4 (15 Bd3 Qb6=/+) 15 ... Ng3 16 Rg1 Nxf1 17 Rxf1 Nd4=/+ DaviesJ.Watson, Manchester 1978, or here 17 ... 0-0-0!? 18 Nxf7? Rxf7! 19 Qxf7 Nd4 with a decisive advantage. (d) 12 Rb1 Bd7 13 Bg5 0-0-0 14 Qd3 Rxg5!? (14 ... d4=) 15 Nxg5 Nxe5 16 Qd4 Kb8 (16 ... Rg8 17 f4 Ng4) 17 Ba6, MinicIvkov, Yugoslav Ch. 1965, and 17 ... Bc8 is equal. (e) 12 Bf4 Qb6! 13 Be3 (13 Ke1 Bd7 14 Ng5 Nxe5! 15 Bxe5 Rxg5 16 Bf6 Rg6 17 Bxc3 Rc8 18 h4, Wiedenkeller-J.Watson, Oslo 1980; and 18 ... Rg8 19 Qd3 Qc7 20 Bb2 Rg4 was the easiest way to a large advantage) 13 Bg3 Qb2 14 Rc1 Qxa3 15 Bd3 Nb4-/+) 13 ... Qb2 14 Rc1, B.Stein-I.Farago, Budapest 1986; 14 ... Qxa3! 15 Rb1 b6 16 Bb5 Bb7=/+. 12 ... Nxe5 After the frequently-recommended 12 ... Qxe5?!, I think that 13 Qxf7+ Kd7 14 h4 favours White; e.g. 14 ... Qd4+ 15 Bd3 e5 16 Be3 Qg4+ 17 Kc1.
13 f4 13 Bf4 Qb6! is fascinating material: 14 Bxe5 ( ... Qxf2 and ... Qd4+ were threatened) 14 ... Rxg5 15 Qh4 (best; there’s are scads of choices, none convincing: 15 f4 Qe3 16 Qd3 Rxe5! 17 fxe5 Qxe5=/+; 15 h4 Qxf2 16 Bxc3 Rg4 is good, or 15 ... Rg8; 15 Bxc3 Bd7! 16 h4 Rg8=/+; 15 Be2 Rxg2! 16 Bg3 Nf5 17 Bh5 Qd4+ 18 Ke1 Qe4+ 19 Be2 Bd7!-/+ Schutt-Balogh, corr 1971) 15 ... Rg8 16 Qd4 Qxd4+ (16 ... Qd8!?) 17 Bxd4 Nf5 18 Bxc3 d4 19 Bd2 e5 20 g3 Nd6 with a better central position and perhaps a very slight advantage. 13 ... f6!
Originally Uhlmann’s suggestion and probably the only acceptable move. The old sacrifice 13 ... Rxg5?! seems to give White too many pluses after 14 fxg5 N5g6 (14 ... N7g6 15 Qg8+ Kd7 16 Be3 or 15 ... Ke7 16 a4 is unpleasant) 15 h4!? (in fact, 15 Rb1 with the idea 15 ... e5 16 Bb5+, and Williams’ suggestion of 15 Be2 e5 16 h4 are both a little better for White) and, for the record: (a) 15 ... Qg3 (15 Be2 would not have allowed this, as Williams points out) 16 Rh3 Qg4+ 17 Ke1 Nf5 18 h5 Nf8 19 Qh8! is apparently winning for White (Williams gives 19 Bb5+ Bd7 with good counterplay); e.g. 19 ... Bd7 20 h6 d4 21 h7 0-0-0 22 Qe5. (b) 15 ... e5 16 h5 Nf8 17 Qg7 (or 17 Bb5+, which Williams likes for White after 17 ... Nc6 18 Qd3) 17 ... Bg4+ 18 Ke1 0-0-0 19 Qxf7 Bf5, B.Stein-I.Farago, Lugano 1985; and Williams gives 20 g4! (20 Be2 Ne6 21 Rf1 Rf8 22 Qxe6+ Bxe6 23 Rxf8+ Kd7 24 Bb5+ Nc6 25 h6 also looks good) 20 ... Bxc2 21 g6 with a substantial advantage. Then Black can hope for the best with 21 ... Bxg6 22 hxg6 Nfxg6, but this is unlikely to play out well after 23 Rh3. In any case, 13 ... Rxg5, which is still widely recommended, looks extremely shaky. 14 fxe5 The alternatives are complex: (a) 14 Bb5+ Bd7 (14 ... Kd8!? achieves only a small plus) 15 Nxe6 Qb6 with these continuations: (a1) 16 Ng7+? Rxg7! 17 Qxg7 Bxb5 18 Qh8+ Kf7! 19 Qh5+ (19 Qh7+ Ke6 20 Qh3+ Nf5! 21 Qxc3 Ng4 22 Re1+ Kd7-/+) 19 ... N5g6 20 Qh7+ Kf8 21 Qh6+ Kg8–+ Greene-Vaughan, corr 1991. (a2) 16 Bxd7+ Kxd7 17 fxe5 Kxe6! 18 Qh3+ (18 exf6 Nf5 19 Rf1? Qd4+ 20 Ke1 Qe4+ 21 Kd1 Ne3+–+) 18 ... Nf5 19 Rf1 Qd4+ 20 Ke1, Garmendez-Moran, Havana 1993; and 20 ... Rg4! wins. (a3) 16 Qh5+ can be met by several moves; e.g. 16 ... Rg6! (or 16 ... N5g6 17 Bxd7+ Kxd7 18 Re1! Nf8=/+; or 16 ... Nf7 17 Bxd7+ Kxd7 18 Qxf7 Qxe6=/+) 17 fxe5 (17 Bxd7+ Kxd7 18 fxe5 Kxe6-/+) 17 ... Qxe6 18 Bxd7+ Kxd7 with a significant advantage. (b) 14 Qh5+! improves the order; there might follow 14 ... N5g6 15 Nh7?! (15 Bb5+! Bd7 16 Nxe6 Qb6=/+ transposes to ‘a3’) 15 ... Kf7 16 g4!? (16 Bd3 Rh8-/+) 16 ... Rh8 (16 ... Kg7; or 16 ... Qc5 with the idea 17 g5 Qd4+ 18 Bd3 e5 19 gxf6 e4 20 fxe7 exd3-/+) 17 Bd3 (17 f5 exf5 18 Bd3 fxg4) 17 ... Qb6 18 f5 exf5 19 Re1 fxg4-/+. 14 ... fxg5 15 Qh5+ After 15 h4 gxh4 16 Qh5+ Kd8 17 Bg5 Qc5 18 Rh3 Bd7 19 Qxh4, Busquets-Remlinger, Los Angeles 1993, it wasn’t too early for the standard exchange sacrifice 19 ... Rxg5! 20 Qxg5 Kc7 with multiple threats; this is actually better for Black than the main line. 15 ... Kd8 16 Bxg5 Qc5!
Black intends ... Bd7 and ... Kc7 to untangle, after which White’s dark squares and king are vulnerable. White’s two bishops and passed h-pawn do not really compensate him enough, although by accurate play and simplification he can hold. 17 h4 Others: (a) 17 Qh4 Rxg5!? (17 ... Bd7 18 Bd3 Bc6 19 Rf1 Kd7=/+ and 20 Rf7 Qg1+ intending 21 Bf1 Raf8!-/+, or 20 g4 Raf8 of B.Stein-Bukal, Balatonbereny 1985, when best was 21 Rxf8! Rxf8 22 Ke2 Rg8!? or 22 ... Qd4, in both cases with equality) 18 Qxg5 Bd7 19 Bd3 Kc7 20 Qh4! (20 g4 Rf8 21 Re1 Rf3!, winning; for example, 22 Kc1 Rxd3! 23 cxd3 Qf2–+ Popovych-J.Watson, New York 1980, or 22 a4 Qd4! 23 Qxe7 Rxd3+ 24 Kc1 Rd1+ 25 Rxd1 Qe3+, mating) 20 ... Nc6 (20 ... a5 prevents Qb4; e.g. 21 Rf1 Rg8 22 Rf2!? Rh8! 23 Qf4 Rxh2! 24 Ke2 with unclear prospects) 21 Rf1 Rg8 22 Qf2 d4 23 Re1, and Black has at least three moves that yield an equal game, but none a forced advantage; e.g. 23 ... Kc8 (23 ... Ne7!? 24 Rb1! Nd5 25 Qf7 Rxg2 26 Rxb7+! Kxb7 27 Qxd7+ etc; 23 ... Rf8 24 Qg3 Ne7 25 Rb1 Kb8 26 Be4 b6=) 24 Re4!= (24 Bh7 Rf8 25 Qg3 Na5=/+; 24 Qe2 Ne7! 25 Be4 Bc6 26 Bxc6 Nxc6=/+) 24 ... Nxe5 25 Rxd4 (25 Qxd4 Qxd4 26 Rxd4 Rxg2 27 Bf1 Rf2=/+) 25 ... Ng4 26 Qf4 Ne5 27 Qf2=. (b) 17 Bd3 Kc7 18 Qh4 (18 h4 Bd7 is line ‘b’ in the next note) 18 ... Nc6! 19 Re1 Bd7 20 g4 (to free the g5-bishop) 20 ... Raf8 21 Bf6 Qd4 (21 ... Qb6=/+ intending ... Nd4, ... Qb2 is also promising) 22 h3 Qf4, Sisniega-Agdestein, Taxco 1985; Black won by bringing his queen to f3 with check and then sacrificing the exchange on f6. 17 ... Bd7 18 Rh3?! Instead: (a) 18 Qf7?? Rxg5 19 hxg5 Qe3–+. (b) 18 Bd3!? occurred in Schvartsman-J.Watson, Midway City 1993: 18 ... Kc7 19 Bxe7 Qxe7 20 Qe2 Qc5 (20 ... Qg7!? 21 h5 Qh6!) 21 Rf1 (21 h5 Raf8! 22 h6 Rf2 23 h7 Rxe2–+) 21 ... Raf8 22 Rf6 Be8! 23 Ke1 (a cute line is 23 g4 Rxf6! 24 exf6 Rxg4! 25 Qxg4 Qe3 26 Qe2 Bh5–+) 23 ... Qd4 24 Rb1 Rxg2! 25 Qxg2 Qxe5+–+. (c) 18 Qg4 a5 (18 ... d4 19 Bd3 Kc7 20 Bxe7 Qxe7 21 Qxd4 Rad8 22 Qxc3+ Kb8=) 19 Rb1 Rxg5 20 hxg5 Nf5=; e.g. 21 Qf4 Kc7 22 Rh3 Rf8 23 Rf3 Ne3+ 24 Qxe3 Qxe3 25 Rxe3 Rxf1+ 26 Re1 Rf2 27 Re2 Rf1+ 28 Re1=. 18 ... Kc7! 19 Bxe7 Qxe7 20 Rxc3+ Bc6 For his pawn, Black has secured his king and threatens ... Rh8, ... Raf8, and ... d4. 21 Rf3 Raf8 22 Rb1 22 Kc1 Rxf3! (or 22 ... d4 23 Rxf8 Qxf8) 23 gxf3 Rg1 24 Kb2 Qc5 25 Qh7+ Bd7 26 Qd3 Rh1 27 Be2 Rxh4=/+ (or 27 ... Qb6+). 22 ... d4 23 Rxf8 Popovych-J.Watson, New York 1981; and now 23 ... Rxf8 led to a win, but 23 ... Qxf8! with the idea 24 Qe2? Qf4! or 24 Ke1 Bxg2 was stronger still. White can probably draw with very accurate play against 13 ... f6, so it doesn’t refute 10 Kd1; but you probably wouldn’t want to be White in a real game. 9.2
8
Bd3
As explained in the very first part of this chapter, quite a few players are trying to avoid this rather dangerous move. They prefer to play 7 ... cxd4, which I analysed in a lengthy note to 7 ... Qc7, and transpose to the Poisoned Pawn in the case of 8 Qxg7 Rg8 9 Qxh7 Qc7. I can sympathize with this point of view but, as you can see from that analysis there, 7 ... cxd4 takes a lot of preparation and involves some very tactical lines as well. I wouldn’t discourage you from playing either or both ways. The point of 8 Bd3 is to develop quickly and get castled, both rare occurrences in the Qxg7 lines. It also makes room for the development of White’s knight to e2 in many variations, whereas in the main lines that move blocks off the bishop on f1. While Black has plenty of resources, he should have a very specific line worked out here; otherwise it’s easy to suffer a quick defeat. 8 ... cxd4 It’s time to break up the centre. 8 ... Nbc6 is a more solid alternative, but I think it leaves White with a meaningful advantage after 9 Qxg7 Rg8 10 Qf6! cxd4 11 Nf3 Ng6 (Moskalenko) 12 Bxg6! fxg6 (12 ... hxg6 13 0-0 dxc3 14 Bg5 b6 15 Rfd1 Bb7 16 Rd3 Qe7 17 Qf4 Qc5 18 Nd4+/=) 13 0-0 dxc3 14 Bh6 Qe7 15 Qf4+/=. Maybe Black can improve upon this. 9 Ne2 dxc3 10 Qxg7 10 f4?! 0-0 11 0-0 Nd7 12 Bxh7+ Kxh7 13 Qh4+ Kg8 14 Qxe7 Qc5+!? (14 ... b6; 14 ... Qc4) 15 Qxc5 Nxc5, Los-Blees, Dutch Team Ch. 1994. In spite of the opposite-coloured bishops, endings such as this are torture for White due to Black’s c-file pressure (following Nxc3) and superior outposts (e4, a4, e4). 10 ... Rg8 11 Qxh7 11 Qh6!? is important.
It was given ‘!’ by French Poisoned Pawn legend Uhlmann; White threatens Bxh7 with tempo. It would be interesting to try 11 ... b6 here, intending ... Ba6; but I’ll stick to more conventional moves: (a) 11 ... Nbc6 and: (a1) 12 Bf4 Bd7!? 13 Bxh7 (13 Bg3 Ng6 14 0-0 0-0-0 15 f4 Kb8 should give reasonable play; e.g. 16 Qxh7 Bc8 17 Rab1 d4 18 Qh5 b6 19 Qf3 Bb7 20 Qf2 Nge7=) 13 ... Rxg2! 14 Bg3 (14 Ng3 Qb6) 14 ... 0-0-0 with good counterplay. The exchange sacrifice 15 Kf1 (15 Qf6 Be8) 15 ... Rxg3 16 hxg3 Nxe5 is clearly satisfactory, with two pawns and an exposed white king. (a2) 12 Bg5 Qxe5 13 Bf6 Qd6 14 Nxc3 Ne5 (14 ... Rxg2) 15 0-0 Nxd3 16 cxd3 Rg6 (16 ... e5!) 17 Qxh7 Qf4 18 Qh8+?! (18 Bxe7 Kxe7 19 Rae1 Rxg2+=) 18 ... Rg8 ½-½ Pedzich-W.Schmidt, Czestochowa 1992. (a3) 12 Bxh7
12 ... Rxg2 (12 ... Rh8 13 Bf4 Nf5 14 Qg5! Nce7 15 Bxf5, Gazik-Uhlmann, Budapest 1987; 15 ... Rg8! 16 Qf6 Nxf5 17 Bg5 b6 18 h4 Ba6 19 Ng3 Qc4 with dynamic equality; Moles’ 12 ... Rf8!? also deserves consideration) 13 Ng3! (13 Qh3 Rg7 14 Qh6 Rg4 15 f4 Qb6=/+) 13 ... Nxe5 14 Qh3 (14 Kf1?? Rxf2+) 14 ... Nf3+ 15 Kf1 and the most forcing sequence is 15 ... e5! 16 Qh6! (16 Qxg2? Nh4 17 Qg1 Bh3+ 18 Ke2 Qc4+!–+; 16 Qh5 Bg4! 17 Qh6 Rxg3 18 hxg3 0-0-0) 16 ... Ng8 (16 ... Qc4+ 17 Bd3 Qg4 18 Kxg2 Nh4+=) 17 Bxg8 Qc4+ 18 Kxg2 Nh4+ 19 Kg1 Nf3+=. (b) 11 ... Qxe5 and: (b1) 12 Bf4 Qh8 13 Bxh7 and now:
(b11) 13 ... Rg4!? 14 h3 Rxf4 (14 ... Rxg2!?) 15 Nxf4 Ng8 16 Qh4 Nf6 17 0-0!, Sevillano-J.Watson, San Francisco 1995; and easiest was 17 ... Nc6! 18 Rfe1 (18 g3? Nd4 19 Kg2 Nxh7) 18 ... Bd7 19 g3 0-0-0 (or 19 ... Nxh7!? 20 Nxd5 Kf8=/+) 20 Bd3 Qxh4 21 gxh4 Rh8 22 Ng2 e5=/+. (b12) 13 ... Rg7 14 Bd3 Qxh6 15 Bxh6 Rxg2! (15 ... Rg8 is simply unclear; Black has the powerful mobile centre but White has the bishop pair and potentially strong h-pawn) 16 Ng3 (16 h4 Nd7 17 h5 Ne5=/+) 16 ... e5 17 Bf1 (17 h4 f6 18 Kf1 Rxg3 19 fxg3 Nf5-/+) 17 ... Bh3 18 Bxg2 Bxg2 19 Rg1 Bf3=/+. (b2) 12 Bxh7 Rxg2 13 Bf4 Qh8 (13 ... Qg7! 14 Ng3 Qxh6 15 Bxh6 e5=/+) 14 Bg3 (14 Ng3 Nd7 15 Kf1 Rxf2+! 16 Kxf2 Nf6 17 h3 Nxh7) 14 ... e5 15 Kf1 Bf5!? 16 Kxg2 Bxh7=/+ (hoping for 17 ... Be4+); e.g. 17 f3 Nbc6 18 Nxc3 0-0-0 19 Rhe1 Nf5 with a comfortable advantage. 11 ... Qxe5 This is the most forcing move, clearing out White’s centre and threatening ... Rh8. The cost is in development.
12 Bf4 Instead: (a) 12 f4 Qf6 renews the threat of ... Rh8. (b) 12 0-0 Nd7 (rejecting the draw by 12 ... Rh8 13 Bf4 Qf6 14 Bg5 Qe5=) 13 Qh6 Nf5 (13 ... Rh8 14 f4 Rxh6 15 fxe5 Rh8 16 Nxc3 a6 gives Black a positional advantage but White is better developed; about equal) 14 Qf4 b6 (14 ... Ne7 15 Qh6 Rh8 16 f4 Rxh6 17 fxe5 Rh5=; 14 ... Nd6 15 Re1) 15 Re1 Ne7 16 Qa4 Bb7 17 Bf4 Qg7 18 Bg3 a6!? (versus Bb5) is equal. 12 ... Qf6
13 h4 Not 13 Qh5?! (13 Bxb8? Rxb8 14 Bb5+ Bd7 15 Bxd7+ Kxd7-/+ Ashley-Morris, New York 1981) 13 ... Nbc6 14 h4 Bd7 (14 ... Rxg2!) 15 Rb1 0-0-0 16 Rb3? (16 Bg5=) 16 ... e5-/+ Madl-N.McDonald, Oakham 1988. 13 ... Nd7 This all-purpose move covers e5 and f6, while keeping open the possibility of ... Nc5 or ... Nf8. 13 ... Rxg2 is also good:
(a) 14 Bg3?! e5 (14 ... Nbc6 with the idea 15 Kf1 Qf3) 15 Kf1 Rxg3 16 Nxg3 Nbc6=/+ with more than enough compensation, and intending 17 Nh5 Qf3. (b) 14 Kf1 Rg8! (14 ... e5 15 Kxg2 exf4 has been analysed to death; White has the tiniest of advantages) 15 Bg5 Qh8! 16 Qxh8 Rxh8 17 Bf6 Rg8 is a curious version of the trade-off in the main line, discussed below. The difference is that Black has an extra pawn, but is a couple of tempi behind. This makes the h-pawn more dangerous; still, Black has time to use his centre: 18 Bxc3 (18 h5 Nd7 19 Bb5 Nc6 20 Bxc3 e5! 21 f4 a6! with the idea 22 Bxc6 bxc6 23 fxe5 c5) 18 ... Nd7 19 Re1!? b6 (19 ... e5 20 h5 f6 21 h6 Kf7) 20 h5 f6 21 h6 e5 22 f4! e4 23 Nd4 Kf7 24 Be2 Nc5 25 h7 Rh8= Wurschner-Strnad, ICCF server 2007. Black might get more winning chances from 19 ... e5 etc. 14 Bg5 Qh8! 14 ... Qg7 is playable, but somewhat more difficult because the rook is relatively misplaced on g7 after 15 Qxg7 Rxg7. 15 Qxh8 Rxh8 16 Nxc3
Here’s a position illustrating the basic issues. White has a passed pawn supported by a rook from behind, as well as two bishops in the vicinity. Black’s pieces are cramped and his queenside undeveloped. But in the French, we often see how a strong central majority (2:0 in this case) can outweigh a passed h-pawn (sometimes even with a passed g-pawn thrown in!). A central advance not only facilitates the quick transfer of pieces to the defence, it also cuts off the bishops from supporting the passed pawns. This situation with rook pawns versus centre pawns can also arise in the Grünfeld Defence and Noteboom Variation, for example. In the case before us the strengths of the centre almost precisely balance Black’s several disadvantages. 16 ... a6 Stopping Nb5 and preparing to expand in the centre. 16 ... f6 has also been played in several games: 17 Bd2 (nobody plays 17 Nb5, but that should be checked carefully; there might follow something like 17 ... fxg5 18 Nc7+ Kf7 19 Nxa8 Nc5 20 Nc7 Bd7 21 Kd2 gxh4 22 Nb5 e5 23 Rab1 Bxb5 24 Bxb5 Nf5, which may favour White’s slight material lead – or not – but Black’s position is harmonious and poised for conflict) 17 ... Nc5 18 Be2 Bd7 19 Rb1 (19 Nb5?! Bxb5 20 Bxb5+ Kf7=/+ Pulkkinen-Kokkila, Finnish Team Ch. 2005) 19 ... Bc6 (19 ... Rc8 or 19 ... Kf7 leaves Black in good shape) 20 Nb5, Van den Doel-Hellenberg, Leiden 2008; and 20 ... Bxb5! 21 Rxb5 (21 Bxb5+ Kf7=/+) 21 ... b6 22 h5 Kf7 is at least equal. 17 0-0-0 Here 17 h5 f6 18 Bd2 Nc5 19 f3 Kf7 20 Kf2 e5 gives equal chances. Szarvas-Molnar, Pecs 1996, went 17 Bxe7 Kxe7 18 f4 b5 19 Rb1 Nb6 (19 ... Nf6!) 20 Kf2 Bd7 21 Ne2?! f6 22 Bg6 e5=/+. 17 ... f6 17 ... b5 18 Rde1 Bb7 19 Bf4 Rc8 20 Kb2 Nc5= Jakovenko-Voekler, Pardubice 2001; here 20 ... d4!? was also interesting. 18 Be3 18 Bd2 Nc5 wasn’t an improvement in Kadziolka-Toth, Heraklio 2004, which continued 19 h5 Nxd3+ 20 cxd3 Bd7=. 18 ... Ne5! Recommended by A.Rodriguez after his blunder 18 ... e5? 19 Nxd5! in Rodriguez Cespedes-Matamoros Franco, Bayamo 1989. 18 ... b5 19 f4 Bb7= with ... Kf7 next has also been played. 19 Bb6?! The following variation illustrates the balance-of-forces in detail: 19 h5 Nxd3+ (19 ... Bd7 20 f4 Nxd3+ 21 Rxd3 Bc6 22 Bd4 Kf7=) 20 Rxd3 (20 cxd3 Bd7 21 Kb2 Rc8 22 f3 Kf7=) 20 ... Kf7
21 f4 Bd7 22 g4 Bc6 23 Bd4 Rag8 24 Rh4 (24 g5 e5! 25 fxe5 Rxg5 26 exf6 Nf5=/+) 24 ... Rh7 25 Bc5 Rhg7 26 Bxe7 Kxe7 27 Rg3 Kf7! 28 h6 Rh7= with the idea 29 g5? e5!. 19 ... Bd7 Black can claim the better position now that his forces are out. Lechowicz-Grzesiewicz, corr 1992, continued 20 f4 Nxd3+ 21 Rxd3 (21 cxd3 Kf7 22 Bf2 Rag8 23 g3 Nf5 24 Ne2 Ba4 25 Rd2 Bildat-Negele, corr 1998; and at this point 25 ... d4!=/+ was good) 21 ... Rc8 (or 21 ... Kf7!) 22 Kb2 Rc4 23 g3, and the game’s 23 ... Kf7 is fine, while 23 ... Bc6! intending ... e5 and ... Nf5 was also recommended.
Chapter
Ten
Winawer
Variation:
Positional
Main
Lines
1 e4 e6 2 d4 d5 3 Nc3 Bb4 4 e5 c5 5 a3 Bxc3+ 6 bxc3 Ne7
This chapter deals with ‘positional’ approaches to the Winawer. In contrast to our experience last chapter with 7 Qg4, White tends to keeps his queen close to the centre and doesn’t immediately try to get at Black’s king (even the apparently attacking 7 h4 introduces what is essentially a positional attack via h5-h6, with Black’s king often going queenside). White then has numerous ways to attack on both wings; e.g. a4 with Ba3 and a5, h4-h5-h6, or even dxc5 with Nd4, hitting both sides of the board. For his part, Black can apply pressure down the c-file, attack d4, blockade the queenside (often by ... Qa5-a4, with or without ... c4), castle on either wing, and will very often attack White’s centre by ... f6 at some point. The extreme flexibility of both sides’ strategies has produced fascinating strategic battles over many years. Arguably, the Positional Winawer has been one of the most intellectually fruitful openings not just of the French Defence but of modern chess. To begin with, White chooses one of several continuations: 10.1 10.2 10.3
7 7 7
Nf3 a4 h4
White has a few other rare tries: (a) 7 f4 cuts off the bishop on c1, but is playable. One answer is 7 ... Qa5 (or 7 ... Nbc6 8 Nf3 Nf5) 8 Bd2 (8 Qd2 can be met by 8 ... b6 intending ... Ba6, or by 8 ... Bd7 9 a4 Nbc6) 8 ... Nbc6 (8 ... Qa4 is another normal move) 9 Nf3 cxd4 10 cxd4 Qa4 11 Bc3 b6 (at least equal), intending ... Ba6 (Korchnoi). (b) 7 Bd3!? commits the bishop rather early, but deserves some attention:
(b1) 7 ... Nbc6 8 Qg4 (8 Nf3 transposes to 10.11) 8 ... Qa5 (8 ... c4 9 Be2 0-0 10 h4 f6! – Dvoretsky; Black stands well;ECOgives 10 Nf3 f5!? 11 Qg3=) 9 Bd2 c4 10 Be2 0-0 (10 ... Nf5 11 Nf3 Bd7 12 Ng5 h6 13 Qh5 Rf8 14 Nh7 and 14 ... Rh8=, or 14 ... Rg8 15 g4 Nfe7 unclear, Velimirovic-Timman, Sarajevo 1984) 11 h4 f6 12 f4 (otherwise Black breaks down the centre) 12 ... Qa4 13 Bd1 Qb5 (threatening ... Qb2) 14 Bc1 Qa5 15 Bd2 and in Short-Shulman, Dhaka 1999, Black repeated by 15 ... Qb5 16 Bc1 Qa5 etc; but 15 ... Qb6 would have been very interesting, since ... Qb2 is still a threat, and now 16 Bc1? fails to 16 ... fxe5 17 fxe5 Nxe5!. (b2) 7 ... Qa5 8 Bd2 Nbc6 (8 ... c4 9 Bf1 Qa4) 9 Qg4!? c4 10 Be2 0-0 11 Nf3 Qa4 12 Ra2 f6 13 exf6 Rxf6=. (b3) 7 ... Qc7 8 Rb1 Nbc6 9 Qg4, Short-Poulton, British League 2002, and as pointed out inChessBase Magazine, Black should play simply 9 ... 0-0! 10 Qh5 (10 Nf3 f5 11 exf6 Rxf6 12 Bg5 e5!=/+) 10 ... Ng6 11 Nf3 c4 12 Bxg6 fxg6, which is practically a tempo up on the main line of 7 Qg4 0-0 (which was featured inPTF3), since Rb1 looks largely wasted. (c) 7 Ne2!? Qc7! (stopping Nf4; 7 ... Nbc6 is also natural) 8 Rb1 (8 g3 b6! 9 Bg2 Ba6 10 0-0 Nbc6=/+) 8 ... b6 (8 ... Nd7 and ... Nb6 is promising, or simply 8 ... 0-0) 9 Ng3?! cxd4 10 cxd4 Qc3+ 11 Bd2?! (a strange pawn sacrifice) 11 ... Qxd4 12 f4, SpasskyBruch, French Team Ch. 1994; and 12 ... 0-0 is natural and strong, intending ... Ba6. 10.1 7 Nf3 The older move, still the most popular at most levels. 7
...
Nbc6
I use this move for convenience and consistency with the rest of the repertoire and with previous editions. You may want to file away the idea that 7 ... Qa5, frequently transposing to our main line after 8 Bd2 Nbc6, avoids some minor issues below. Otherwise, there’s a maze of transpositions and move order ideas here, and all sorts of moves have proven legitimate; for example, 7 ... Bd7, 7 ... Qc7, 7 ... b6, and even 7 ... h6 have all been played in top-level chess. Over the last decade or more, 7 ... b6 has become popular, I think with good reason. The immediate exchange of lightsquared bishops cuts down on White’s attacking possibilities, and ... Ba6 tends to be better-motivated after White plays Nf3, a generalization which holds in both the Positional Winawer of this chapter and the Portisch-Hook Variation of the next. In large part that’s because the knight on f3 blocks the white queen’s path to the kingside, so Black has the luxury of being able to take an extra two moves; also, the bishop isn’t needed as much to protect e6 when White is slow getting to the kingside. To the extent that you don’t feel comfortable with 7 ... Nbc6 (or related 7 ... Qa5 systems), 7 ... b6 is a way to go. After 7 ... Nbc6, we look at: 10.11 10.12
8 8
Bd3 Be2
What are arguably the two most important moves, 8 a4 and 8 h4 are dealt with in the next two sections by transposition, but there are odds and ends. Specifically: (a) 8 a4 Qa5 transposes to 7 a4 lines, after 9 Qd2 (10.21) or 9 Bd2 (10.22). (b) One of the very most important moves is 8 h4, which I will discuss at the beginning of section 10.3 on 7 h4. I should say that h4 ideas pop up in almost every basic move order, so you need to get used to when they are dangerous and when not. (c) 8 g3 is rather slow and has several answers. One is the manoeuvre 8 ... Qa5 (or 8 ... 0-0 9 Bg2 f6=) 9 Bd2 Qa4 (or 9 ... cxd4 10 cxd4 Qa4); e.g. 10 Qb1 (10 dxc5!? Ng6 wins the e-pawn safely, whereas 10 ... Nxe5?!, which I have given previously, isn’t satisfactory because of 11 Nxe5 Qe4+ 12 Qe2 Qxh1 13 0-0-0! Qe4 14 Qb5+ Kf8 15 Re1 Qf5 16 Qa5!) 10 ... c4!? (10 ... 0-0=) 11 Bg2 Bd7 12 0-0 (12 Qxb7?? Rb8 13 Qc7 Qa6!) 12 ... 0-0-0 13 Ng5 Rdf8 14 f4 f6 15 Nf3 fxe5 16 fxe5 h6 with chances of a kingside initiative. Black’s king is secure. Compare the next chapter, which features this ... Qa5-a4 manoeuvre. (d) 8 Rb1 Qa5 (8 ... Qc7 is also natural): (d1) 9 Rb5 has had a good reputation, but after 9 ... Qc7! (the right square for the queen anyway) 10 Rxc5? (10 Bd3 c4 11 Be2 f6), 10 ... a6 and ... b6 follows. (d2) Socko-Bartel, Warsaw 2001, went 9 Qd2 c4 (9 ... 0-0 would keep more options open; 9 ... b6) 10 h4, and Black should just play the natural 10 ... 0-0 11 h5 h6 12 Be2 f6 13 exf6 Rxf6=. (e) 8 dxc5 leaves the valuable e-pawn up for grabs: 8 ... Qa5 (8 ... Ng6!? has the idea 9 Bb5 Qa5!; while 9 Be3 0-0 10 Bd4 Qa5! 11 Qd2 b6! 12 cxb6 axb6 opens lines and prepares to take over the light squares; this is a standard pawn sacrifice and effective here) 9 Bd2 (9 Qd2 Ng6=) 9 ... Ng6 10 c4 Qxc5 11 cxd5 exd5 12 Be3 Qa5+ 13 Bd2 Qc7 and the e-pawn will soon fall. 10.11
8
Bd3
This is relatively uncommon these days, but perfectly playable. Compare 7 Bd3 above. 8 ... Qa5 Or 8 ... c4 9 Bf1 (sometimes White reroutes the bishop in this way rather than play Be2/0-0/Re1/Bf1 and now g3) 9 ... 0-0 (9 ... Qa5 10 Qd2 Qa4!?) 10 g3 f6 (10 ... h6 11 a4 Qa5 12 Bd2 Bd7 13 Bg2 f6=) 11 exf6 Rxf6 12 Bg2 Qa5 13 Qd2 Bd7 14 0-0 Be8 15 Ne5 Nxe5 16 dxe5 Rf7=.
9 Qd2 Instead: (a) 9 0-0 is generally not played because it allows the draw after 9 ... c4 10 Be2 Qxc3 11 Bd2 Qb2 12 Rb1 Qxa3 13 Ra1 Qb2 etc. Apart from that, Black can also play for the win:
(a1) 10 ... Qa4!? 11 Ng5 h6 12 Nh3 Bd7 13 f4!? 0-0-0!? (13 ... Nf5!) 14 Bf3 f6 15 Qe2, N.Rogers-Edelman, USA 1987, and now 15 ... Nf5 16 Nf2 fxe5 17 fxe5 Rhf8=. (a2) 10 ... Bd7 11 a4 Nc8 12 Qd2 Nb6 13 Qg5 Rg8!? 14 Ra3 h6 15 Qh5 Nxa4 16 g3 b5 17 Bd2 Qd8 18 Rb1 Qe7 19 Raa1 a6 20 Nh4= Van der Sterren-Korchnoi, Wijk aan Zee 1978, assessed as equal by Short. (b) 9 Bd2 c4 and: (b1) 10 Be2 Bd7 (10 ... 0-0 intending ... f6 is also played) 11 0-0 (11 Ng5 is a little slow: 11 ... h6 12 Nh3 0-0 13 Nf4 f6 14 exf6 Rxf6) 11 ... f6 (11 ... 0-0-0 12 Ng5 Rdf8 is complex but also satisfactory; e.g. 13 f4 f5 – 13 ... f6! – 14 exf6?! gxf6 15 Nh3 Nf5 16 Bg4 Nce7 followed by taking over the g-file, Ljubisavljevic-Ugrinovic, Italy 1970; and here 16 ... Nd6 17 Nf2 f5! intending ... Ne4 was also good) 12 Re1 (12 exf6 gxf6 13 Nh4 0-0-0=) 12 ... fxe5 13 dxe5 (13 Nxe5 Nxe5 14 dxe5 0-0=) 13 ... 0-0 14 Bf1 Rf5! 15 g3 Raf8 16 Re3 R5f7 17 Bg2 Qc7=/+ De Firmian-Gulko, Malmö 2001. (b2) 10 Bf1 and then:
(b21) 10 ... Bd7 11 g3 f6 12 Bh3 0-0-0 (the conventional 12 ... fxe5 13 dxe5 0-0 was also fully playable) 13 0-0 Ng6 14 exf6 (14 Re1 Rde8!?) 14 ... gxf6 15 Qe2, G.Garcia-J.Watson, Philadelphia 1998; 15 ... e5=. (b22) 10 ... 0-0 11 g3 f6 12 Bh3 (12 exf6 Rxf6 13 Bg2 Bd7 14 0-0 Raf8=) 12 ... fxe5 13 Nxe5 Nxe5 14 dxe5, Ricardi-
J.Watson, Ubeda 1999, and now 14 ... Nf5!? (14 ... Qc7 15 Qe2 Nf5 16 0-0 Bd7=) 15 0-0 Qc5 (15 ... Bd7) 16 Bxf5 Rxf5 17 Be3 Qa5 18 Bd4 Bd7 is somewhat better for Black. 9 ... c4 9 ... b6! is a very good solution, as is often the case after White answers ... Qa5 with Qd2. But what follows is extremely instructive: 10 Be2 Qa4
11 h4 Posohov-Kurochkin, Kharkov 2002, saw a manoeuvre that is occasionally available to Black: 11 0-0 Bd7 12 Ne1 0-0-0 13 f4 f6 14 Bg4!? f5 15 Bd1 Rdg8 16 Kh1 h6 17 Nf3 Nd8! 18 Rb1 Nf7 19 Ng1 g6 20 Be2 Rg7, followed by ... Rhg8 and ... g5 with strong pressure. 11 ... Bd7 12 h5 h6 13 Nh4 Nf5!? 14 Nxf5 exf5 15 Rg1?! Better 15 0-0, when 15 ... Be6 is equal. 15 ... Ne7 16 g4 fxg4 17 Bxg4 Bxg4!? Botvinnik called this a mistake; he suggests 17 ... Bf5 18 Bxf5 Nxf5 19 Rg4 Qa6 intending ... Qe6/ ... g5, achieving ‘a won position’. Needless to say, White has options, but Black should obtain some advantage. 18 Rxg4 Nf5 19 Qe2 Qd7 Here it might be better to play 19 ... 0-0-0! 20 Rf4 g6 21 Kd1 Rhe8 22 Qf3 Re6=/+. On the other hand, I’m not sure what Black’s plan is; perhaps ... Kb8-a8/a7 and then opening a kingside file? 20 Rf4 Qe6 21 Qf3 g6 22 a4 0-0-0 23 Ba3 Smyslov-Botvinnik, USSR Ch., Leningrad 1941. Here Botvinnik says, “Black cannot begin anything active since after Kd2 and Rg2 he is constantly forced to reckon with the threat (of an exchange sac on g6).” Indeed, 23 ... Kb8 (23 ... b6 24 Kd2 Rhg8 25 Rb1 Kb7 26 a5+/= was the game) 24 Kd2 Ka8 25 Rg1 Rhg8 26 hxg6 fxg6 27 Rxg6 Qxg6 28 Rxf5 is probably drawn but can only favour White (thanks to John Tomas for his help here). 10.12
8
Be2
This modest move, less dynamic and threatening than, for example, systems with h4, is still an important one that reflects the desire to prop up the centre quickly via 0-0 and moves like Re1. 8 ... Qa5 Black attempts to tie White to defence of c3, and prepares the blockading move ... Qa4 in some situations. This move order might seem a little odd because White doesn’t have to defend the pawn on c3 immediately and thus Black apparently gives White more leeway to develop. One can also argue that Black’s deployment is less effective when White hasn’t yet committed to a4. But Black gains new possibilities, and in some cases White forfeits his idea of Ba3, so White often plays a4 at some point anyway. It’s a little surprising that so few players have made the simple developing move 8 ... 0-0; after all, White has played the relatively passive Be2 and not the attacking Bd3. White may be able transpose to the text with 9 0-0!, assuming that Black doesn’t have much better than 9 ... Qa5; but the option of 9 Bd3 threatening Bxh7+ is instructive and relevant to the next note. Black should be fine due to White’s waste of a move; e.g. 9 ... f6 10 exf6 (10 Bf4? fxe5 11 Bxe5 Nxe5 12 Nxe5? Qa5-/+ with the idea 13 Qd2 cxd4 14 cxd4 Qxd2+ 15 Kxd2 Rxf2+) 10 ... Rxf6 11 Bg5 Rf7 12 0-0 (12 dxc5 e5) 12 ... c4 13 Be2 h6! (13 ... Qd6!?) 14 Bh4 (14 Be3 Qc7 15 Qd2 Bd7) 14 ... Qa5 15 Qd2 Bd7 16 Rfe1 Nf5 17 Bg3 Nxg3 18 hxg3 Raf8 19 Nh4! Rf6! 20 f4 Be8=.
9 0-0! This to some extent argues for 7 ... Qa5 and 8 ... Nbc6 rather than 7 ... Nbc6 and 8 ... Qa5. Otherwise, White has two important options:
(a) 9 Qd2 is usually played to prepare a4 and Ba3, but here Black hasn’t committed to ... Bd7 (compare 10.21), so a good answer is 9 ... b6! intending ... Ba6; e.g. 10 dxc5 (10 a4 Ba6 11 Bxa6 Qxa6 12 dxc5 Ng6!? 13 cxb6 axb6 14 Qe3 0-0 with great pressure) 10 ... bxc5 11 c4 (11 0-0 Ba6 12 c4 0-0) 11 ... Qxd2+ 12 Bxd2 d4 (12 ... Rb8!=/+) 13 c3 Ng6 14 cxd4 cxd4 15 Bb4! Ngxe5 16 Nxe5 Nxe5 17 Rd1 d3 18 Bxd3 Nxd3+ 19 Rxd3 Ba6=. (b) 9 Bd2
9 ... 0-0 (a straightforward solution to this position is 9 ... Qa4, which meets both 10 Rb1 and 10 Qb1 with 10 ... c4, while contesting the light squares after 10 0-0 b6; Moskalenko gives the line 10 dxc5 Ng6 11 Be3 0-0! 12 Bd4 b6! 13 cxb6 axb6 14 0-0 Ba6 and calls it ‘good for Black’, while here 14 Bxb6 Nf4! is already worse for White – this deserves a look!) 10 Bd3 (this tends to be one of the problems with an early ... 0-0; instead, 10 0-0 c4!? 11 Ng5 Qa4 12 Nh3 Ng6 is fine; while 10 c4 Qc7 weakens White’s centre, e.g. 11 cxd5 exd5 12 c3?! Bg4 13 Be3 Qa5 14 Qd2 Nf5=/+) 10 ... Ng6 (White threatened Bxh7, but 10 ... h6 intending 11 0-0 c4 12 Be2 f6 is also satisfactory) 11 Ng5 (11 c4? Qc7 and White’s centre collapses; 11 0-0 can be met by 11 ... Qa4 or 11 ... c4 12 Be2 Qa4) 11 ... h6 12 Nh3 Qa4! 13 dxc5 Ncxe5=/+. 9 ... b6
A rather obscure move that I’ve previously recommended, intending ... Ba6, which tries to exploit the fact that White has inserted neither Bd2 nor a4. Normally 9 ... Bd7 or 9 ... c4 are played. Of course, Black cannot grab the pawn in view of 9 ... Qxc3?? 10 Bd2 Qb2 11 Rb1 Qxa3 12 Rb3 Qa2 13 Qc1 and Ra3. 10 a4 White tries to deny the black queen access to a4 and thus give c4 added effect, in contrast to:
(a) 10 Bd2 Qa4 11 dxc5 bxc5 12 c4 0-0 13 cxd5 exd5=. (b) 10 c4?! dxc4! (10 ... cxd4 11 Bd2 Qa4 12 cxd5 Nxd5=) 11 dxc5 Ba6! 12 cxb6 axb6=/+. 10 ... Ba6 11 Bxa6 11 Bd2 Bxe2 (11 ... Bc4!?, Van der Weide-Wemmers, Triesen 2006; 12 Bxc4! dxc4 13 dxc5 Qxc5 14 Qe2 0-0=) 12 Qxe2 0-0 13 c4?! Qa6=/+. 11 ... Qxa6 12 dxc5 bxc5 13 Ba3 Babula-Trichkov, Czech League 1992. Black keeps the advantage after 13 ... Qa5 14 c4 0-0 15 cxd5 Nxd5=/+; e.g. 16 Qd3 h6 17 Qb5 Rfc8 18 Bxc5 a6! 19 Qxa5 Nxa5 20 Bd6 Rxc2 with a superior position. 10.2
7
a4
White not only prepares Ba3 (exploiting the absence of Black’s dark-squared bishop), but also prevents ... Qa5-a4. 7 ... Qa5 Black attacks c3, which can be logically defended in two ways: 10.21 10.22
8 8
Qd2 Bd2
10.21 8 Qd2 Nbc6 9 Nf3 Bd7 The 8 Qd2 variation isn’t seen that much in modern play, but it may be due a revival. White allows an endgame after ... cxd4/cxd4 and hopes that possession of the two bishops will secure him an advantage, although practice to date has not confirmed this. What’s more, in many of the most important lines, Black can bypass the ending and achieve dynamic play. It should be said, however, that when Black does so, White can open lines for his bishop pair and fight on an equal basis with Black’s superior development. In general, there’s no variation that stands out against 8 Qd2 and you can expect a double-edged game.
I should mention that 9 ... b6, intending ... Ba6, has scored rather well and would serve as an acceptable alternative. 10 Be2 The most important move, but there are many other legitimate choices: (a) 10 Bb5 Nxe5! 11 Nxe5 Bxb5 12 Ba3! Ba6 13 Bxc5 f6 14 Nd3 Nc6 15 0-0 Kf7= Mnatskanian-Korchnoi, Yerevan 1965. (b) 10 dxc5 Ng6 11 Qe3 Qc7 12 c4?! dxc4 13 Bb2 Nb4 ‘-/+’ (Kmoch); 14 Qd2 a5 15 Bxc4 Bc6! might follow. (c) 10 g3 Rc8 (or (10 ... f6 11 Bg2?! fxe5 12 Nxe5 Nxe5 13 dxe5 0-0 14 0-0 Rad8= intending ... Bc6) 11 Bg2 cxd4 12 cxd4 Qxd2+ 13 Bxd2 Nf5=. (d) 10 Ba3 cxd4 11 cxd4 Qxd2+ 12 Kxd2 is a well-analysed ending. Black has excellent practical results and no theoretical problems; e.g. 12 ... Nf5 13 c3 (13 Bb2 Na5 14 Bc3 Nc4+ 15 Bxc4 dxc4=; 13 Rb1 b6 14 c3 Na5 15 Bb4 Nc4+ 16 Bxc4 dxc4 17 a5 Bc6=/+ Kan-Botvinnik, Sverdlovsk 1943) 13 ... Na5 14 Ra2
14 ... Rc8 (14 ... f6 15 Bb5 Bxb5 16 axb5 Nc4+ 17 Kd3= Brunner-Kindermann, Bern 1990) 15 Bb4?! (15 Bb5 a6 16 Bxd7+ Kxd7 17 Rb1 b6!=/+; 15 Bb2 f6 16 Bd3 Nc4+ 17 Bxc4 dxc4 18 Ba3 Bc6=/+) 15 ... Nb3+ 16 Kd1 a5 17 Ba3 Bxa4 (or 17 ... Rxc3) 18 Bb2 Bd7 19 Bd3 a4-/+ Curi-Filguth, San Jose 1989. (e) 10 Bd3 offers Black several good choices, including 10 ... c4 and even 10 ... cxd4. The most interesting move is 10 ... f6 and then:
(e1) 11 exf6 gxf6 12 dxc5 e5 13 Be2 (13 c4 dxc4 14 Bxc4 Qxc5 15 Bb3 Na5! 16 Ba2 0-0-0=) 13 ... 0-0-0 14 0-0 Rhg8 15 Kh1 Ng6! 16 c4 (16 Qxd5? Nf4! 17 Bxf4 Bh3-/+) 16 ... Qxd2 17 Bxd2 e4! 18 Ng1 d4, Tringov-Vaganian, Vrnjacka Banja 1971. (e2) 11 0-0 fxe5 12 dxe5 (12 Nxe5 Nxe5 13 dxe5 0-0 14 c4 Qxd2 15 Bxd2 d4=, or here 14 ... Qc7!? to leave more play on) 12 ... 0-0 13 Re1 (if 13 Ba3 – Korchnoi; then 13 ... Rxf3! 14 gxf3 Nxe5 15 Qe3 Nxd3 16 cxd3 d4!) 13 ... h6 14 Nh4 (14 Ba3, Smyslov-Uhlmann, Havana 1964; 14 ... b6=) 14 ... Be8!? 15 c4?! (15 Qe3 c4 16 Be2 d4 17 Qh3 Qxc3 18 Qxc3 dxc3 19 Bxc4 Nd5= intending ... g5 and ... Bg6) 15 ... Qxd2 16 Bxd2 Rd8 17 cxd5 Nxd5-/+ Mariano-Wang Hao, Kuala Lumpur 2005. 10 ... Rc8 Uhlmann’s idea, to improve Black’s ending by waiting a move before ... cxd4 and putting a piece on the potentially open file. 10 ... f6 is also fine, since the bishop on e2 tends to be passive. 11 dxc5! Before Black can play ... cxd4, White creates tripled pawns! The first point is to obtain d4 for a knight. Here 11 Bd3?! cxd4 loses a tempo, and 11 Ba3 cxd4 12 cxd4 Qxd2+ 13 Kxd2 is again harmless: 13 ... Na5! 14 Rhb1 b6 (or 14 ... f6 and ... Kf7) 15 Bb4 (15 Ba6 Nc4+ 16 Kd1 Rd8=) 15 ... Nc4+ 16 Kd1 Nc6 17 Be1 N6a5 18 Nd2, D.Myers-Van der Kleij, corr 1990; and 18 ... f6 gives Black comfortable equality. 11 ... Ng6 12 0-0 12 Qe3 has assorted answers, such as 12 ... a6 and the simplest move 12 ... 0-0. Note also that in response to 12 ... Nf4, White has nothing better than 13 Bf1, and after 13 ... Ng6, it’s hard to find a better move than 14 Be2 !
12 ... Ncxe5!? 13 Nd4
Or 13 Nxe5!? Nxe5 14 Qg5! (14 Qe3?! Ng6 15 Rb1=; 14 Qd4 f6 15 Rb1= or 15 Be3=) 14 ... Ng6 15 c4 0-0 (15 ... dxc4?? 16 c6+–; 15 ... h6 16 Qg3 dxc4=) 16 cxd5 exd5!? (or 16 ... h6 17 Qg4 Rfd8=) 17 Qxd5 Bc6 18 Qd2 (not 18 Qf5? Bxg2!; or 18 Bd2?, D.Myers-West, corr 1981, as 18 ... Nf4! favours Black, while 18 ... Bxd5 gives a smaller edge) 18 ... Qxc5! 19 Ba3 Qf5 20 Bxf8 Nf4 with threats of both ... Qg5 and ... Qh3!, hence 21 Rfd1! Rxf8 22 Ra3! Bxg2 23 Rg3 Nxe2+ 24 Qxe2 Bc6= with a pawn for the exchange and a sound pawn structure, in contrast to White’s four isolated pawns.
13 ... 0-0 I criticized this in previous editions but, given White’s pawn structure, Black has some leeway in developing. Instead, 13 ... a6 is sound but unnecessary, whereas 13 ... Qxc5 is surprisingly playable, perhaps even best; e.g. 14 Nb5 (or 14 Ba3 Qxc3 15 Qxc3 Rxc3 16 f4 Nc6 17 Nb5 Rxa3! 18 Rxa3 Ke7 and Black has even material with clearly more promising piece play) 14 ... Bxb5 15 Bxb5+ Kd8! 16 Ra3?! (16 Re1! a6 17 Bf1 Nc4 18 Qg5+! Kc7! 19 Qg3+ Qd6=) 16 ... Nc4 17 Bxc4 Qxc4=/+. 14 Ba3 Bc6! 14 ... Rfd8?! 15 Bb4 Qc7 16 Nb5+/= Kaplan-Tukmakov, Puerto Rico 1971. 15 f4 15 Nb5 Nc4 16 Bxc4 dxc4 17 Bb4 Qa6! 18 Nd6 Rcd8 19 f3 Bd5=. 15 ... Nc4 16 Bxc4 dxc4 17 f5 exf5 18 Nxf5 Rcd8 19 Nd6 f6 20 Qe3 Ne5= 10.228 Bd2 The main move. White abandons the idea of Ba3 (although Bc1-a3 can occur later), but keeps the kingside dark squares covered and gives solid support to the weak c3-square. 8 ... Nbc6 9 Nf3 Instead: (a) 9 Qg4 compares poorly with the main 7 Qg4 lines because a4 is then practically wasted; play might go 9 ... 0-0 (a cute game went 9 ... Bd7 10 Qxg7 0-0-0 11 Qxf7? Qb6! 12 Bd3 Rdf8 13 Qh5 c4 14 Bf1 Nxd4! 15 0-0-0 Nxc2! 0-1 Rebej-Spitzenberger, corr 1965, due to 16 Kxc2 Qb3+ 17 Kc1 Bxa4) 10 Nf3 f6 (10 ... f5 11 Qh5 Qb6!? 12 dxc5 Qc7=/+) 11 exf6 Rxf6 12 Qh5 (12 Be2? e5) 12 ... Bd7 13 Bd3 h6 14 0-0 c4 15 Be2 Be8 16 Qh4 Bg6=/+. (b) 9 Ne2 Qc7. Now that the bishop can’t come to a3, this is an ideal spot for the queen. It also ties White to defence of d4. Myreng-Glück, St Paul 1982, continued 10 Bf4 Ng6 11 Bg3 Qa5 12 Qd2 f6= (or 12 ... 0-0! and ... f6). 9 ... Bd7
Now 10.221 10.222
White
normally 10 10
plays: Be2 Bb5
Of the many other 10th moves, some are instructive, but none are terribly critical: (a) 10 Ng5?! intends Qh5, but abandons the centre: 10 ... h6 11 Qh5 (11 Nh3? Qc7! hits d4: 12 Bc1!? cxd4 13 cxd4 Nb4! 14 c4 dxc4 15 Be2 Nd3+! 16 Bxd3 cxd3 17 Qxd3 Bc6-/+ Milton-J.Watson, San Diego 1986; or here 12 Qg4 Nf5 13 Bd3 cxd4 14 cxd4 Nxe5!-/+ – Suetin) 11 ... g6 12 Qh3 cxd4 13 cxd4 Qb6! 14 c3 Qb2 15 Rd1 Na5=/+. (b) 10 Qb1 Qc7 11 Bd3 c4 (or 11 ... 0-0 12 0-0 h6=) 12 Be2 f6 13 Bf4?! (13 exf6 gxf6 14 0-0 0-0-0 is double-edged) 13 ... Ng6 14 Bg3 fxe5 (14 ... f5 – Korchnoi) 15 dxe5 Qa5 16 0-0!? Qxc3 17 Qxb7 Rb8 18 Qc7, Schoneberg-Tischbierek, Leipzig 1986; 18 ... Rc8 (18 ... Qxc2!) 19 Qd6 Qxc2 20 Rfe1 Nge7=/+ (Tischbierek). (c) 10 c4 Qc7 11 cxd5 exd5 (or 11 ... Nxd5=) 12 dxc5?! (12 c3 0-0, planning ... cxd4 and ... Bg4) 12 ... 0-0 13 Be2 Nxe5 14 00 Nxf3+ 15 Bxf3 Qxc5=/+ Dubinin-Boleslavsky, USSR Ch., Leningrad 1947. (d) 10 h4 is a little slow after a4, although it means White no longer has to worry about ... Qa4:
10 ... f6!? (or 10 ... 0-0-0!? 11 h5 h6; 10 ... 0-0! 11 h5 h6= is the safest, intending ... f6) 11 h5 (11 exf6 gxf6 12 c4 Qc7 13 cxd5 Nxd5; compare 10 Be2) 11 ... fxe5 12 Nxe5 (12 h6 gxh6 13 Nxe5 Nxe5 14 dxe5 0-0-0 15 Rxh6 Qc7 and ... Nf5) 12 ... Nxe5 13
dxe5 Qc7 (13 ... h6=) 14 f4 0-0-0 15 Qg4 h6= Pirttimaki-Farago, Helsinki 1983. (e) 10 Rb1 Qc7 11 Be2 (11 Bb5 is well met by 11 ... Na5 or 11 ... a6, with the idea 12 Be2 f6 or 12 Bxc6 Nxc6, and 11 ... c4 is also good; compare 10 Bb5) 11 ... f6!? (11 ... Na5 is appropriate, since it hits a4 as well as planning to occupy c4; and of course 11 ... 0-0 can’t be bad) 12 Bf4!? (12 exf6 gxf6 13 dxc5 e5 produces the normal trade-offs that you’ll see after 10 Be2) 12 ... Ng6 13 Bg3 fxe5 14 Bb5?! cxd4 15 cxd4 Qa5+ 16 Nd2 exd4 17 Bd6 Qd8 18 h4? Qf6-/+ intending ... e5, Hirsig-Jurek, Bad Ragaz 1993. (f) 10 g3 is interesting:
10 ... 0-0-0 (10 ... 0-0 looks safest, intending 11 Bg2 f6= or 11 Bh3 f6 12 exf6 Rxf6=) 11 Bh3 (11 Bg2 f6 12 c4 Qa6!? 13 cxd5 Nxd5= with many options for both sides) 11 ... f5! 12 Bg2 (12 exf6 gxf6 13 0-0 e5 14 Bxd7+ Rxd7 15 c4!?, Doda-Raicevic, Banovici 1979; and 15 ... Qa6!= controls the key c4-square) 12 ... h6 13 h4 Qc7 14 Bc1 (14 0-0 Na5 15 dxc5 Nc4 16 Re1=/+) 14 ... cxd4 15 cxd4 Nb4 16 Kd2 Be8!? 17 Ba3 Nbc6 18 Bd6 Rxd6! 19 exd6 Qxd6 20 Kc1 (20 c3 f4 21 g4 h5 22 g5 Bg6) 20 ... f4! with a strong attack, Kavalek-Hort, Montreal 1979. (g) 10 Bd3 c4 (10 ... f6 11 0-0 c4) 11 Be2 f6 12 0-0 fxe5 13 Nxe5 (13 dxe5 0-0! 14 Re1 Ng6 15 Bf1 Rf5) 13 ... Nxe5 14 dxe5 0-0 15 f4 Be8 16 Bg4 Qb6+ (16 ... Nf5) 17 Kh1 Bg6 18 Qb1 Qc6! 19 Qb5 Qc8! 20 Ra2 Be4=/+ S.Nikolic-Ivkov, Sarajevo 1967. 10.221 10 Be2 The old main line, and in a sense the main reason to play 7 a4, which stops variations with ... Qa4 and discourages those with ... b6. 10 ... f6
This is the most dynamic move. It is well established after many years, although the computer seems consistently to underestimate Black’s central and kingside play. In the previous edition I recommended 10 ... 0-0-0 in order to avoid certain move orders by White, but now I don’t think that’s an issue. Nevertheless, I have been investigating a way for Black to play more safely and solidly in these lines; to that end, I’m going to give the rudiments of a set-up which goes back to some old moves: (a) 10 ... c4 11 Ng5! (this has long been the theoretical drawback to 10 ... c4; previous games and theory have shown that otherwise Black can play on the kingside with ... f6, regardless of which way he castles, with no problems; sometimes ... 0-0-0 can even go with ... f5/ ... h6 and playing for ... g5) 11 ... h6 12 Nh3 (‘threatening’ Nf4-h5) 12 ... Ng6 (here’s the move that I think deserves a closer look; the first idea is to neutralize Nf4; however, 12 ... 0-0 is also reasonable, when the game is level following 13 0-0 Ng6 or 13 Nf4 f6!? 14 exf6 Rxf6 15 0-0 Raf8) 13 0-0 (13 Bh5 Nce7 14 Qg4 0-0) 13 ... 0-0. This is the point! 10 ... c4 has mostly been associated with 13 ... 0-0-0, especially in the very few old games with 12 ... Ng6. But I think that the king is happy enough on the kingside and this opens up possibilities on the other side of the board. For example (out of numerous possibilities), 14 Qc1 (contemplating both Qa3 and sacrifices on h6; I should note that Black doesn’t usually need to defend against the latter, but if needed, ... Kh7 is available) 14 ... Rfb8!? (this is only one of a number of moves, but I like the idea of blasting away on the queenside right away):
(a1) 15 Bh5 Nce7 16 Qa3 b5!? 17 Qd6 (17 axb5 Qxa3 18 Rxa3 Rxb5 19 Rfa1 a5=, a basic position in which Black gets his pawn to a4 and then has liberty to manoeuvre) 17 ... Qd8 18 axb5 Rb6 19 Qa3 Rxb5=. (a2) 15 Ra3 (this protects c3 so that the d2-bishop can move, and makes ... Bxa4 somewhat less attractive because then
Qa1 forces ... b5, though even that sort of position is notoriously drawish) 15 ... Nce7 16 Bh5 Kh7 (16 ... b5!?) 17 Nf4 Nxf4 (17 ... Bxa4 18 Qa1 b5 can’t be too bad either) 18 Bxf4 Kg8!? with the idea 19 Bxh6?! gxh6 20 Qxh6 Nf5 21 Qg5+ Ng7 22 f4 Qd8 23 Qg4 f5!-/+. (a3) 15 g3 Nce7 (or 15 ... b5 16 axb5 Qxb5=) 16 Qa3 b5 17 axb5 Qxb5 (17 ... Qxa3 18 Rxa3 Rxb5 19 Rfa1 Nc6=) 18 Qa2 Qb2=. (b) I’m not sure if 10 ... 0-0, intending ... c4, is a better or worse order: 11 0-0 (11 Ng5 is just asking for Black to keep the pressure on d4 by, for example, 11 ... Qc7; the only possible justification is 12 Bd3, but then follows 12 ... Nf5! with the idea 13 g4 Nh4! and 14 Bxh7+ Kh8, when the threat of ... f6 means that White’s d-pawn falls, or 14 Nxh7 cxd4! 15 Nxf8 Rxf8 and White has to work a bit to stay out of real trouble; his best defence is 16 0-0 Nxe5 17 Be2! dxc3, and not 18 Be3 f5!, but 18 Bg5! Nhg6 19 Qd4, though Black still has an edge after 19 ... Qa5, intending ... Nc6 at some point) 11 ... c4 (Moskalenko analyses the old line 11 ... Qc7 12 Re1 h6 13 Bc1 b6, very heavily tested and written about in the 1970s and 1980s; this may turn out to be a decent line for Black), when 12 Ng5 h6 13 Nh3 transposes to 10 ... c4 (with 12 ... 0-0), and even 12 ... f6!? 13 exf6 gxf6 14 Nf3 Nf5 is quite interesting. 11 c4 Aggressive but loosening. Here are typical examples of other moves: (a) 11 0-0 c4 illustrates what I mean above about ... c4 usually working out in the absence of Ng5 (11 ... fxe5 is also equal here; e.g. 12 Nxe5 Nxe5 13 dxe5 0-0 14 c4 Qc7 15 f4 Bc6=); e.g. 12 Re1 fxe5 13 dxe5!? (13 Nxe5 Nxe5 14 dxe5 0-0=) 13 ... 0-0 14 Bf1?! Rf5! 15 g3 Raf8-/+ Pein-Plaskett, British Ch., Swansea 1987. (b) 11 exf6 gxf6 12 0-0 (12 c4 Qc7 transposes to our main line; 12 Nh4!? 0-0-0 13 Bh5 Qc7! 14 Be3?! Ng6! 15 Bxg6 hxg6 16 Nxg6 Rxh2 17 Rg1 Rh7-/+ Gold-J.Watson, New York 1978)
and: (b1) 12 ... c4 13 Nh4 0-0-0 (a typical pawn sacrifice is 13 ... Ng6!? 14 Bh5 0-0-0 15 Bxg6 hxg6 16 Nxg6 Rhg8 17 Nf4 e5 18 Ne2 Rg7 19 Ng3 Rh8 with an attack) 14 g3 (14 Bh5 Ng6!) 14 ... e5 (14 ... Ng6!=/+) 15 f4 exd4 (or 15 ... Rhg8=/+) 16 cxd4 Qc7 17 c3 Nf5 18 Ng2 Na5 19 Bf3 Nb3 20 Ra2 Qa5= Spassky-Shaked, Hoogeveen 1998. (b2) 12 ... 0-0-0 13 c4 Qc7 14 cxd5 Nxd5 15 c4 Nde7 (15 ... Nf4)
16 a5 (16 dxc5 e5 17 Qc2 Rhg8 18 Rfe1 Bf5 19 Qb2 Be4 20 g3 Nf5 21 Bc3 Ncd4 22 Bxd4 exd4= 23 Nd2?? 0-1 DeikoBelozerov, Minsk 1996; 23 ... Nxg3 follows) 16 ... Nxd4! (16 ... cxd4!? 17 a6 e5? 18 axb7+ Kb8 19 Qa4± Kleijn-Janse, Fermo 2009; 17 ... b6 was necessary) 17 Nxd4 cxd4 18 a6 b6! 19 c5 bxc5 20 Ba5 Qd6 21 Bxd8 Kxd8 with a double-edged and balanced game in which Black’s two pawns and central mass compensate for the exchange and a somewhat exposed king. (c) 11 Rb1!? is Wedberg’s active move, important to know but no longer feared: 11 ... Qc7! 12 Bf4 (12 exf6 gxf6 13 dxc5 e5 14 c4 Be6 15 cxd5 Bxd5 16 0-0 0-0-0 17 Qc1 Rhg8 18 g3 Nf5 19 c3, Poretti-Weber, email 2000; and after 19 ... Be4!? 20 Rb5 Bd3! 21 Bxd3 Rxd3, the light squares and White’s shattered pawns give Black at least a satisfactory game) 12 ... Ng6 13 Bg3 fxe5
(c1) 14 Bb5 cxd4 15 cxd4 Qa5+ 16 Kf1 0-0 (or 16 ... e4) 17 Bxc6 Bxc6 18 Ng5? (18 Nxe5 Nxe5 19 Bxe5 Bxa4! 20 Kg1 Rf7-/+) 18 ... h6 19 Nxe6 Rf6 20 Nc5 exd4 21 Qxd4 b6–+ and ... Qxa4, Weinzettl-Dückstein, Nendeln 1986. (c2) 14 0-0 cxd4 (or 14 ... Qc8!? 15 dxe5 0-0 16 c4 Nce7! 17 Qd2 Bc6= Maki-Ogaard, Oslo 1983) 15 cxd4 0-0 16 Bb5 Nf4 (or 16 ... Qc8 17 dxe5 Nce7) 17 Bxc6 (17 Nxe5?! Nxe5 18 dxe5 Bxb5 19 Rxb5, Karell-D.Myers, corr 1990; 19 ... Rf7!=/+ – Myers) 17 ... Bxc6 18 Nxe5 Bxa4 19 Rb4 (a much-tested position: 19 Qg4 h5! wins material; 19 Qd2 Nh5 is good; and 19 Rb2 b6 is equal, with a messier line being 19 ... Rac8!? 20 Nd3! Bxc2 21 Bxf4! Qc3 22 Rxc2 Qxc2=) 19 ... Qxc2! (19 ... b5 20 Qd2 Nh5=/+) 20 Rxa4 (20 Qg4 h5; 20 Qxc2 Bxc2 21 Bxf4 Rxf4 22 g3 Rff8 23 Rxb7 a5=/+) 20 ... Ne2+ 21 Kh1 Qxd1 22 Rxd1 Nc3-/+ 23 Rda1 Nxa4 24 Rxa4 Rfc8 25 h3 a5 26 Nd7 b5 27 Ra1 Rc2 28 Nc5 a4–+ Bickel-Hund, corr 1983. 11 ... Qc7 12 exf6 12 cxd5 Nxd5 13 c4 is more accurate, preventing ... Nf4, when 13 ... Nde7 14 exf6 gxf6 transposes below. Here 13 c3 cxd4 14 cxd4 fxe5 15 dxe5 (15 Nxe5 Nxd4) 15 ... Nxe5 16 0-0 0-0! 17 Re1 Rad8 gives equal chances.
12 ... gxf6 13 cxd5 13 Bc3 0-0-0! 14 dxc5 d4! and 15 Bxd4 Be8 or 15 Nxd4 Be8-/+. 13 ... Nxd5
14 c4 Instead: (a) 14 dxc5 is tricky: 14 ... 0-0-0 (14 ... e5!? 15 a5 a6 16 c4 Nde7 17 Qb1!? 0-0-0 18 0-0 Bg4 19 Rd1 Bxf3!? 20 Bxf3 Nd4 and ... Qxc5 is about equal, but 19 ... Rhg8 gives more positive chances) 15 a5! (15 0-0 e5 16 c4 Ndb4 17 Bxb4 Nxb4 18 Qd6 Rhf8 19 Qxc7+ Kxc7 20 Rfd1 Na6-/+ McCrory-J.Watson, Denver 1976; this shows why White should play c4 early on) 15 ... a6! 16 c4 Nde7 17 Bc3 e5 18 0-0
At this point Black should target the king with 18 ... Rhg8!; for example, 19 Qc2 Bh3 20 Ne1 Bf5 21 Qb2 Nd4 22 Bxd4 Rxd4 23 Nf3 Rd7 24 Rfd1 Rxd1+ 25 Rxd1 Ng6=, and White has to be a little careful of ... Nf4; e.g. 26 g3 (26 Kf1 Nf4 27 Nh4 Be6=) 26 ... Nf4! 27 Nh4 Nxe2+ 28 Qxe2 Bg4 29 f3 Be6=/+. (b) 14 c3 0-0-0 15 0-0 Rhg8 16 Re1 e5!? 17 c4 (17 dxe5 Bh3=/+) 17 ... Bh3! 18 Bf1 Nb6 (18 ... Nf4!? 19 Bxf4 exf4 20 d5 Bg4! has been tested in practice and equalizes) 19 d5 Nxc4! 20 dxc6 Qxc6 21 g3 Bxf1 22 Rxf1 e4= 23 Qc2 (23 Qb3?! Qd5 24 Rac1, Spassky-Korchnoi, Belgrade 1977; 24 ... Ne5! 25 Rxc5+ Qxc5 26 Rc1 Qxc1+ 27 Bxc1 Nxf3+ should favour Black) 23 ... Qd5 24 Bf4 exf3 25 Rfc1 (25 Rac1 Ne5=) 25 ... Ne5?! (25 ... Rge8! 26 Qxh7 Ne5=) 26 Qf5+!? (26 Qxc5+=) 26 ... Rd7 27 Qxf6? Nd3-/+ Mazi-Farago, Bled 1992. 14 ... Nde7
Here Black can opt for 14 ... Nf4! with excellent play, but that would have been avoided by 12 cxd5 etc.
15 Bc3 Others: (a) 15 d5!? exd5 16 cxd5 Nxd5 17 Qc2?!, Ratsch-Raderer, corr 1982; 17 ... Ncb4! 18 Qe4+ Kd8! 19 Rd1 Re8 20 Qc4 Kc8 and White is a pawn down and tied up (21 0-0?? Nb6). (b) 15 dxc5 0-0-0 16 a5 a6 (or 16 ... Rhg8!? with the idea 17 0-0 Ne5 18 a6 Bc6) 17 0-0! (17 Bc3 e5 18 Qb3 Rhg8 19 g3 Bg4 20 Qb2 Nf5!? with good play and the idea 21 h3? Nxg3!!–+ 0-1 Mauro-Pieretti, corr 1999; 22 fxg3 e4!–+) 17 ... e5 18 Qb1! Rhg8 19 g3 (19 Qxh7? Rh8 20 Qb1 Bg4=/+) 19 ... Bg4 20 Rd1 Ng6!? 21 Be3 f5 (or 21 ... Rxd1+ 22 Bxd1 Kb8) 22 Rxd8+ Rxd8 23 Bg5 (23 h3 Bxh3 24 Ng5 f4=) 23 ... Rd7 24 Kg2 Nxa5=. 15 ... 0-0-0 16 d5 16 0-0 Be8! 17 d5 Rf8 18 Qc2 exd5 19 Qxh7 Bg6-/+ – Korchnoi. 16 ... exd5 17 Bxf6 17 cxd5 Be6 18 dxe6?! (18 Bxf6 Bxd5 19 Qc1 Rhf8=/+) 18 ... Rxd1+ 19 Rxd1 Qf4!? (19 ... Rd8-/+) 20 0-0 Rg8-/+ EkstromBeliavsky, Stockholm 1986. 17 ... Rhg8!? 17 ... Rhf8!=/+ improves; one idea is 18 Bh4 Qa5+!. 18 cxd5 Be6 19 0-0 19 Qc2! Rxd5=. 19 ... Bxd5! 20 Qc1? Not 20 Qc2?? Qf4!; 20 Qd3 Rd6 21 Bxe7 Qxe7 22 Rfe1 Qf7-/+ with ... Rf6 and in some cases ... Nd4 are in the air. 20 ... Rdf8! 21 Qc3 There’s no defence; e.g. 21 Bxe7 Nd4!. 21 ... Qf4-/+ A.Rodriguez-Roselli Mailhe, Uruguayan Ch. 1989, intending 22 Bxe7 Rxg2+! 23 Kxg2 Qg4+ 24 Kh1 Rxf3 25 Qxc5 Be4–+. 10.222 10 White tries to provoke ... a6, hoping that it will be a weakness.
Bb5
10 ... a6 Anyway! I’m not really sure that any logical move isn’t good here, but a particularly sound alternative is 10 ... Qc7 11 0-0 0-0, which I analysed inPTF3; or else the somewhat dull 10 ... c4 11 0-0 f6, when one main line goes 12 Re1 0-0-0 13 Bc1 fxe5 14 dxe5 Nf5 15 Bg5 Rdf8 16 Bxc6 Bxc6 17 Nd4 h6 18 Be3 Nxd4 19 Bxd4 g5= Djurhuus-Ulibin, Tunja 1989. 11 Bxc6 Others are no better: (a) 11 c4!? Qc7 12 cxd5 Nxd5 (12 ... axb5!? 13 d6!) 13 Bxc6, Casella-Shulman, Connecticut 2002; and 13 ... Qxc6 is unclear, but 13 ... Bxc6! gives a healthy advantage due to the long diagonal; e.g. 14 0-0 0-0 15 dxc5 Ne7 16 Qe2 Ng6 17 Rfd1 Rad8=/+. (b) 11 Bd3 c4 12 Be2 (12 Bf1 0-0 13 g3 f6!; e.g. 14 exf6 Rxf6 15 Bg2 Raf8 16 0-0 h6 17 Re1 Be8!) 12 ... f6 13 0-0?! (13 exf6 gxf6 14 0-0 e5=), Dineley-J.Levitt, British Ch., Swansea 1995, and instead of 13 ... Qc7!?, 13 ... fxe5! would have the standard effect: 14 Nxe5 Nxe5 15 dxe5 0-0 16 Qb1! Qc7 17 f4 Nf5=/+. (c) 11 Be2 is the most popular move (on the theory that ... a6 would be a weakness by comparison with 10 Be2, but it hasn’t proven to be the case) 11 ... f6 12 c4 Qc7 13 cxd5 (13 exf6 gxf6 14 c3, Morris-Klein, New York 1991; 14 ... dxc4 15 Bxc4 cxd4 16 cxd4 Nxd4!?) 13 ... Nxd5 14 c4 Nde7 (14 ... Ndb4!?) 15 exf6 gxf6
16 dxc5 (16 d5!? exd5 17 cxd5 Nxd5 18 Qb3 Be6=/+ 19 0-0 Qg7?! 20 Bc4 0-0-0 21 Rfe1 Rhg8 22 g3 Qf7 unclear, M.Chandler-Kummerow, German League 2000; Black might have played better earlier) 16 ... 0-0-0 17 Bc3 e5 18 Qd6 Bf5!? (18 ... Nf5 looks more accurate: 19 Qxc7+ Kxc7 with full compensation, Tischbierek-Poldauf, German Cup 1991 – compare 10 Be2;
then 20 0-0-0, as in Zhang Zhong-Berkes, Calvia Olympiad 2004, could be met by 20 ... Ng7!; e.g. 21 Rd5 Ne6 22 g3 Kc8 23 Rhd1 Ne7 24 Rd6 Nf5 25 Rb6 Bc6 26 Bb4 Nfd4=) 19 Qxc7+ (19 Qxf6!? Rhf8 20 Qh6 Ng6 – Pelletier) 19 ... Kxc7 20 Rc1 (20 0-0 Ng6 21 g3 Rhg8 intending ... Nf4 is at least equal) 20 ... Ng6!? 21 g3 Nf8= intending ... Ne6xc5, Short-Pelletier, Leon 2001. 11 ... Nxc6 11 ... Bxc6 keeps the e7-knight defending the kingside. Briefly: (a) 12 Ng5 h6 (12 ... 0-0 13 Qh5 h6) 13 Qh5 g6 14 Qh3 Qc7 15 0-0? cxd4 16 cxd4 Bxa4-/+ Berzinsh-Kahn, Berlin 1994. (b) 12 c4?! Qc7 13 cxd5 Bxd5 14 0-0 Rd8=/+ C.Lopez-Celis, Buenos Aires 1996. (c) 12 0-0 h6 (12 ... Qc7!? 13 dxc5 Qa5! is a cute idea that seems to equalize at once, especially with opposite-coloured bishops)
13 dxc5 (13 Qc1 c4 14 Qa3 Qc7 15 Rfe1, Suetin-Misiano, Biel 1997, and 15 ... 0-0 was equal; 13 Qe1 Qc7 14 dxc5 Qd7! 15 Nd4 Bxa4 16 f4 Bc6!? 17 Qg3 Nf5! 18 Qh3 g6 19 g4 ½-½ Gashimov-Guseinov, Baku 2000) 13 ... Qxc5 14 Nd4 0-0 (14 ... Nf5 15 Nxf5 exf5 16 Qh5 d4 17 Qxf5 dxc3 18 Be3 Qc4 ½-½ Velicka-Jurek, Lazne Bohdanec 1999) 15 Qg4 Kh7 16 Rfe1 Bd7 17 Rab1 Qc7 18 Re3, Kovalov-Tischbierek, German League 1991; and apart from 18 ... Nf5 19 Nxf5 exf5=, Black had 18 ... Ng6 19 Nf3 f6 20 exf6 Rxf6=/+. 12 0-0 12 Ng5!? h6 13 Qh5 0-0!? (13 ... Rf8 14 Nf3 0-0-0) 14 Nf3! (14 Nh3 Be8 or 14 ... cxd4 15 cxd4 Qb6=/+) 14 ... Be8 15 Qg4 f5 16 exf6 Rxf6 17 c4 Qb6 18 cxd5 exd5 19 dxc5 Qxc5=/+. 12 ... 0-0 12 ... Qc7 13 Re1 Na5 14 dxc5 Nc4 (14 ... Qxc5!) equalized in El Taher-Shulman, Pardubice 1999. 13 Re1 Qc7 14 dxc5 Black answers just about any move with ... Na5. 14 ... Na5 15 Be3 15 Nd4 Nc4 16 Bg5!?, Wedberg-Spielmann, Stockholm 2002; and easiest is 16 ... Qxc5 17 Qg4 Kh8!?=. After 15 Be3, Cabrilo-Shulman, Belgrade 1998, continued 15 ... Nc4 16 Qb1!? Rab8!? (16 ... Nxe3! 17 Rxe3 Rab8 would have won back the c5-pawn and exposed White’s queenside, without dark-square worries on the the other side of the board) 17 Bd4 b6! 18 cxb6 Nxb6 19 Qd1?! (19 Qc1 Nxa4 20 Re3=) 19 ... Nxa4 20 Re3 Qb7? (20 ... Nb2!=/+ with ... Nc4 next) 21 Nd2 h6 ½-½. Here with 22 Rg3, White would have had a very real attack to more than compensate for his weaknesses. Nevertheless, Black stood well coming out of the opening. 10.3 7 h4 An ultra-modern way of playing it: White will grab space on the kingside and worry about his development later. He also preserves some leeway as to where to put his pieces, depending upon how Black replies. Before turning to 7 h4 proper, I want to address the similar idea 7 Nf3 Nbc6 8 h4 (equivalent to the less common 7 h4 Nbc6 8 Nf3). In both move orders, White waits to see how Black responds; for example, after 7 Nf3, White may plan to answer 7 ... Bd7 or 7 ... Qc7 with 8 a4, but prefer to answer 7 ... Nbc6 with 8 h4, which is the position in the diagram.
Then 8 ... Qa5 9 Bd2 Qa4 is usually answered by 10 h5, when 10 ... cxd4 11 cxd4 transposes to our main line below. Relatively strong players have also used 10 Qb1 c4 11 h5 h6 to transpose to Chapter 11: specifically, the Portisch-Hook Variation with 7 ... Qa4 8 Qb1 c4 9 h4 Nc6 10 h5 h6 11 Nf3 Ne7 (see 11.111). That line is full of complexities and ultimately balanced. In the past I’ve presented two unusual alternatives to 8 ... Qa5, but let me just stick with one, 8 ... f6!?, which as far as I know is not suggested in the major sources and is almost untested, yet it appears to be quite good: 9 h5 (9 exf6 gxf6 cedes the centre and ought to be fine for Black; e.g. 10 Be2 Qc7 11 0-0 Bd7 12 Rb1 0-0-0 13 dxc5 e5 with a more-or-less standard and balanced position; 9 Bb5 fxe5 10 Nxe5 0-0=/+ and 9 Be2 fxe5 10 dxe5 Qc7 are unattractive; or 9 Bd3 fxe5 10 dxe5 c4 11 Be2 00=/+) 9 ... fxe5
Now play might go: (a) 10 h6, when I analysed 10 ... exd4 inPTF3, but 10 ... g6 is a much simpler solution and equal; e.g. 11 Nxe5?! (11 dxe5 Qa5 12 Bd2 0-0=) 11 ... Nxe5 12 dxe5 Qc7!=/+ 13 Bb5+ Bd7!? (13 ... Nc6=/+) 14 Bxd7+ Qxd7 (14 ... Kxd7!? 15 Qe2 Nc6 16 Bg5 Qxe5 17 Qxe5 Nxe5 18 Bf6 Nc4 19 Bxh8 Rxh8 and only Black can win) 15 Bg5! 0-0 16 0-0 Nc6 17 Bf6 Na5 and I prefer Black’s chances, though it’s an open contest. (b) 10 Nxe5 Nxe5 11 dxe5 Qc7 12 f4 Bd7 (12 ... Nf5 13 Rh3!?) 13 Be2 0-0-0 14 0-0 g6! 15 Be3 (15 h6? Nf5), FontaineBergez, Clichy 1998; 15 ... gxh5! 16 Bxh5 Nf5 17 Bf2 Bb5 18 Re1 Rhg8 19 Qf3 Rdf8. (c) 10 dxe5 h6 11 Bf4 0-0 12 Qd2 Qa5 13 g3 Bd7 14 Nh4 Be8 15 Be2 Nf5! 16 Kf1 (16 Nxf5 Rxf5) 16 ... Nxh4 17 Rxh4 Ne7 18 Bd3 c4 19 Be2 Kh7=/+.
7 ... Nbc6 Considering how delicate the balance becomes in our main line, this is a good moment to see that back-ups are available: (a) 7 ... Qc7 is a relatively solid choice. Briefly: (a1) After 8 Nf3, Black can play 8 ... b6 and switch into a good version of the ... b6 plan (Nf3 in this or the Portisch-Hook Variation is often well met by ... b6). I’m not going to analyse it in depth, but 9 Bb5+ Bd7 10 a4 can be answered by 10 ... Bxb5 11 axb5 a5 (among others), and here 10 Bd3 h6 11 h5 Nbc6 (11 ... c4) 12 0-0 0-0 is a traditional Black set-up; e.g. 13 Bf4 f5 14 Be2! cxd4 15 cxd4 Na5 (15 ... Be8=) 16 a4?! Be8 17 Nh4 Qc3=/+ Mainka-Kindermann, Lippstadt 1993. (a2) 8 h5 cxd4 9 cxd4 Qc3+ 10 Bd2 Qxd4 11 Nf3 Qe4+ 12 Be2 is a position with which we will become familiar with Black’s queen’s knight on c6 and White’s pawn on h6 (see note ‘c3’ to White’s 11th move below). The difference here is that Black can play 12 ... Nf5
13 Kf1!? (13 0-0 Nc6 14 Re1 0-0 15 h6 Nxh6 16 Bxh6 gxh6 17 Rb1 Kh8=/+ Almasi-Vallejo Pons, Beijing 2011; 13 h6 gxh6 14 Kf1, and I like 14 ... b6! in these positions, though 14 ... Rg8 15 Bd3 Qg4 was equal in Kuczynski-W.Schmidt, Slupsk 1988) 13 ... b6 14 Bb5+ Bd7!? (14 ... Nd7!=/+) 15 Bd3 Qg4 16 h6! gxh6 17 Bxh6?! (17 Bxf5? Qxf5 18 Rxh6 Rg8 19 Bb4 Bb5+ 20 Kg1 Qe4!-/+ Najer-Rustemov, Moscow 1995; 17 Ne1! Rg8 18 Qxg4 Rxg4 19 Bxf5 exf5 20 Rxh6 Bb5+ 21 Nd3 Kf8=) 17 ... Nc6 18 Bg5 Rc8 (18 ... Rg8!) 19 Bf6 Rg8= Aseev-Rustemov, St Petersburg 1994. (b) 7 ... Qa5 8 Bd2 Qa4 is complex.
Lines which include ... Ne7 tend to come via this move order, while other variations with ... Nbc6 are mostly in Chapter 11: (b1) 9 Qb1 Nbc6 10 Bb5 (10 Nf3 c4 11 h5 h6 is a Portisch-Hook variation – see 11.111; if Black wants something unique he can try 10 ... 0-0 11 Bb5 Qa5) 10 ... Qa5=, intending 11 c4?! Qc7. On 11 Be2 0-0 12 Nf3, 12 ... c4 intending ... f6 yields good play; or 12 ... Qc7, planning either an early ... f6 or perhaps the old-fashioned ... b6. The reason this seems to work out is that White’s queen on b1 is misplaced and will have to spend time becoming useful, since none of the queenside squares are particularly impressive in that respect. (b2) 9 Qg4 0-0 10 Nf3 Qxc2 11 Rc1 Qg6 12 Ng5?! f6 13 h5 Qf5=/+. (b3) 9 h5 h6 10 Nf3 Nbc6 (or 10 ... cxd4 11 Rh4 Nbc6) 11 Rh4 cxd4 transposes into the main line of section 11.3 in the next chapter. 8 h5 8 Nf3 is analysed at the beginning of this section, in the note about 7 Nf3 Nbc6 8 h4. 8 ... Qa5 9 Bd2 9 h6? Qxc3+ 10 Bd2 Qxd4 11 Nf3 Qe4+ 12 Be2 reaches the complex position arising from 11 h6 below, except that Black has captured an extra pawn on c3 and retained his very important one on c5. 9 ... cxd4 This move order sidesteps 9 ... h6 10 Qg4, which in my opinion favours White. In the previous edition I talked extensively about 9 ... Bd7, which has undergone much further and deeper analysis. New discoveries for White have, at the very least, compelled Black to walk a tightrope to stay competitive after 9 ... Bd7, so I’m going to forego it here in favour of the text. 10 cxd4 Qa4
With the explosion of popularity of h4 on moves 7 or 8, this position has become of the main lines of Winawer, all the more so because it can arise from either the Positional Winawer of this chapter (6 ... Ne7) or the Portisch/Hook of the next (6 ... Qa5 7 Bd2 Qa4). The good thing is that it is straightforward and clear in comparison to the ultra-flexible ones in which Black has pawns on c5 and f6, where you never know whether Black will advance them, exchange them, or keep the tension. The bad news is that it’s difficult (for me, anyway) to keep track of the move orders by which you can reach one or another of the resulting positions, especially if you play both systems. I’ve tried to indicate transpositions, but the reality is that you should try to pick a particular set of move orders and master those before you try to absorb the whole complex. At the risk of being repetitive, let me urge you to study the h4-h5 lines in Chapter 11 simultaneously with these. 11 Nf3 A pawn sacrifice designed to bring White’s pieces out as quickly as possible and open lines for the bishop pair. It’s important to understand the alternative ideas: (a) Upon 11 c3 Black activates quickly on the queenside and has no real trouble after 11 ... Qxd1+ 12 Rxd1 h6
13 g4!? (or 13 Bd3 Bd7 14 Rb1 b6 15 Ne2 Na5 16 g4 Nc4= Dowden-Wang, Wanganui 2003) 13 ... Bd7 14 f4 Rc8 15 Bd3 Na5 16 Rb1 b6= Shaked-Remlinger, Los Angeles 1993. If White tries to enforce f5 by 17 Ne2 Nc4 18 Bc1 and Rf1 or Ng3, Black beats him to the punch with 18 ... f5!, when 19 g5 (to deny Black an outpost on f5) results in a position where White can’t make progress on the kingside. (b) 11 Bc3 b6 is easy for Black; e.g. 12 h6 gxh6 13 Qd3 (or 13 Nf3 Ba6 14 Bxa6 Qxa6 15 Qe2 Qxe2+ 16 Kxe2 Rc8 17 Kd3 Rg8 18 Rag1 Rg6= Pritchett-I.Rogers, British Ch., Blackpool 1988) 13 ... a5 14 Qd2 Nf5 15 Rb1!? Ncxd4 16 Kd1, Short-Korchnoi,
Wijk aan Zee 1987; and now 16 ... Ba6!=/+ was best. (c) 11 h6 is rarely played with this move order, but leads to forcing and very unclear play, so it’s important. Some possible solutions:
(c1) 11 ... gxh6 12 c3! Qxd1+ 13 Rxd1 looks awkward for Black. One interesting plan would be to play 13 ... Ng8, discouraging capture on h6, then make the normal ... Na5/ ... Bd7/ ... Rc8 moves on the queenside, and finally, when the time is right, break with ... f6. If White allows his bishop to be traded off after Bxh6, Black should be able to defend the kingside and continue with his queenside plans. Having said that, I don’t quite trust 13 ... Ng8 to solve all of Black’s problems. (c2) 11 ... Nxd4 is complex and I think best: 12 hxg7 Rg8 13 Bc3 (13 Bd3 Rxg7 14 Kf1 b6 15 Bh6, and 15 ... Rg8? 16 Rh4! is strong, but 15 ... Rg6! 16 Rh4! Ba6 17 Be3 Nef5 18 Rxd4 Nxd4 19 Bxd4 Qxd4 20 Bxa6 Qxe5 21 Nf3 Qf6 is up for grabs) 13 ... Ndc6 (13 ... Ndf5) 14 Qd3 (14 Rxh7?? Qe4+) 14 ... Rxg7 15 Rxh7 Rxh7 (15 ... Rg4 16 Ne2 Rh4 is only marginally in White’s favour, if that) 16 Qxh7 b6 17 Nf3 Kd7! 18 Qxf7 Qxc2 19 Bd2 Qb2!= 20 Rc1 Bb7 21 Bg5! (intending 22 Rxc6) 21 ... Qxa3=, or 21 ... Rc8 22 Rxc6 Qa1+=. (c3) 11 ... Qxd4 12 Nf3 Qe4+ 13 Be2 is another option.
Remarkably, this same position could arise via 7 ... Qc7 8 h5 cxd4 9 cxd4 Qc3+ 10 Bd2 Qxd4 11 Nf3 Qe4+ 12 Be2, except that White’s pawn would be on h5, and Black’s c6-knight back home on b8. That move order allows Rustemov’s 12 ... Nf5!, which from the evidence seems quite decent for him. But Black has resources in the diagrammed position too: 13 ... Nxe5 14 Bc3 (14 0-0Nxf3+ 15 Bxf3 Qd4 16 c3 Qc4!? 17 hxg7 Rg8 18 Re1 Rxg7 19 Rb1 a5! gives White a lot of play, but Black has two pawns; 14 hxg7 Rg8 15 Bc3 f6 transposes, and here 15 Kf1 Nxf3 16 Bxf3 Qd4 doesn’t establish any advantage for White either)
14 ... f6 15 hxg7 (15 Nxe5 fxe5 16 0-0 0-0 17 hxg7 Rf6=) 15 ... Rg8 16 Nxe5 fxe5 17 Qd3 Qxd3 18 Bxd3 d4 19 Bb4 Rxg7 20 Rxh7 Rxh7 21 Bxh7 Nc6, Miton-Shabalov, Stratton Mountain 1999. I’d assess this as ‘unbalanced but equal’. White has the bishop pair, Black’s extra pawn is doubled but potentially useful, and White’s a- and c-pawns are isolated. Black should try to exchange light-squared bishops, probably on the long diagonal. 11 ... Nxd4 12 Bd3
This is the ‘Kasparov Gambit’ I refer to elsewhere, which can also arise from various lines of Chapter 11 via 6 ... Qa5 7 Bd2 Qa4. It tends to provide both sides with dynamic chances. Now the variation splits into two main lines: 10.31 10.32
12 12
... ...
Nec6 h6
After first suggesting it 27 years ago, let me just briefly point out that 12 ... Ndf5 is similar to 12 ... Nec6, and makes 13 h6 relatively harmless after 13 ... Nxh6 14 Bxh6 gxh6 15 Kf1 Bd7. If you compare some of the lines below, you can see that this move has pluses and minuses, though the overall assessment is about the same. It’s also interesting that the simple 12 ... 0-0, intending 13 h6 g6 followed ... f6 or ... f5, has no apparent drawbacks that are more serious than those faced by 12 ... Nec6. But of course the latter move is battle-tested. 10.31 12 ... Nec6 This threatens the e5-pawn in some cases, and also keeps an eye on b4, which can be used by either a white bishop or rook to cause Black problems. I’m currently leaning in favour of 12 ... h6 (10.32), so out of a large mass of material about 12 ... Nec6, I’ll select some major and critical points in the interests of (relative) brevity.
13 Kf1 This is the original idea, to threaten Rh4 because ... Nxf3 won’t come with check. Here’s a survey of other possibilities: (a) 13 Rh4?? Nxf3+. (b) 13 0-0 Nxf3+ 14 Qxf3 Qh4! (Kasparov); although I’m not sure that this is so much better than other typical positions after 15 h6 g6 16 Rfe1 with unclear compensation. (c) 13 h6 is the major and, I think, underrated alternative, putting into question whether Black can completely equalize: 13 ... Nxf3+ (here 13 ... g6!? should be taken seriously) 14 Qxf3 Qd4 15 0-0 gxh6 (15 ... Nxe5!? 16 Qg3 Nxd3 17 cxd3 0-0 with two pawns but serious weaknesses) 16 Rfe1 Rg8 17 Qf6 (17 Bf4!+/=) 17 ... Qg4 18 g3 Nd4 19 Kg2 Rg6!? 20 Bxg6 Qxg6, Fedorowicz-Yermolinsky, US Ch., Durango 1992; and White can keep an edge with 21 Rh1 or 21 Rac1. 13 ... Nf5 13 ... Nxf3 14 Qxf3 has been under a cloud ever since the original game Kasparov-Anand, Linares 1992, where 14 ... b6? 15 h6!± was played.
14 h6 The most important alternative is 14 Bxf5 exf5 15 h6!. This isn’t held in particularly high regard, but it gives White a good chance for a modest advantage: (a) 15 ... gxh6!? 16 Rxh6!? (16 Bxh6 Be6? 17 Bg5 h6 18 Kg1 d4?! 19 Bf6 Rg8 20 Rxh6± Vocaturo-Caruana, Cremona 2006; Black needed to play 16 ... Rg8 or even 16 ... f4!?, but may not feel satisfied anyway) 16 ... Be6 (16 ... Nd4 17 Ng5! Be6 18 Bb4 Qa6+, Hossain-Natarajan, Mumbai 2003; and White had 19 Kg1! Ne2+ 20 Kh1 Qc4 21 Qd2±) 17 Bg5 Rc8! 18 Rh4 Qa6+ 19 Kg1
Rg8 20 Rxh7 Kd7! ½-½ Fedorowicz-Yermolinsky, St Martin 1993. Black’s position is solid after, for example, 21 Rb1 b6 22 c4!? Qxc4 23 Rc1 Qa6 24 Bf6 Nd8. (b) 15 ... Rg8 16 Bg5! (16 Ng5 Be6! 17 Nxh7 0-0-0 gives Black enough play; 16 hxg7 Rxg7 17 Bg5! Be6 18 Bf6 Rg4! intending ... Kd7, ... Rag8 is roughly equal) 16 ... Be6 17 Rh4 Qa6+ 18 Kg1 gxh6, Hellers-Gulko, Biel 1993; and here White should play 19 Rxh6! with the idea 19 ... Nxe5?! (19 ... Rc8 20 Rxh7 Kd7! is deceptively solid, but a c4 sacrifice from White will be critical and force Black to defend) 20 Nxe5 Rxg5 21 Qd4+/= with command of the dark squares and pressure on Black’s vulnerable king.
14 ... Rg8 14 ... g6 is worth considering, whereas 14 ... Nxh6 15 Rh4! Ng4 16 Rb1 h5 17 Qc1! is very difficult for Black. Likewise, 14 ... gxh6 15 Bxf5 exf5 allows 16 Rxh6! (16 Bxh6, as in Vocaturo-Caruana, Cremona 2006, could have been met by 16 ... f4! and ... Bf5) 16 ... Be6 17 Bg5, when Black stands worse, albeit in a defensible position. 15 hxg7 Again 15 Bxf5! is best – see 14 Bxf5 above (note ‘b’). 15 ... Rxg7 16 Ng5?! Bd7 17 Qh5?! Or 17 g4 Nfe7 18 Nxh7 0-0-0 19 Bg5 Rxg5! 20 Nxg5 Nxe5 with full compensation. 17 ... 0-0-0 18 g4 Pfreundtner-Rosenberger, Wiesbaden 1994; and here 18 ... Nfd4! gives Black a clear advantage. 10.32
12
...
h6
Stopping h6 once and for all. The main game I’ll follow is Nakamura-Shulman, US Ch., St Louis 2010, which typically arose from the entirely different move order 6 ... Qa5 7 Bd2 Qa4 8 Nf3 Nc6 9 h4 cxd4 10 cxd4 Nge7 11 h5 Nxd4 12 Bd3 h6. If you’re aiming to play ... h6, be careful to choose your move order carefully or you may not get time to play it. 13 Kf1
As we’ve already seen in 10.31, this odd-looking move is actually routine; it prevents ... Nxf3 with check (as in 13 Rh4?? Nxf3+). Alternatives: (a) 13 Bb4 Ndc6! (I think this improves upon 13 ... Nxf3+?! 14 Qxf3 Nc6 of Berzinsh-Ptrik, Slovenian Ch. 1998, which tempts White into 15 Bd6 or 15 Rb1! Nxe5 16 Qg3 Nxd3+ 17 cxd3 Rg8 18 0-0 planning Rfc1-c7, which is at any rate difficult to defend) 14 Bd6 0-0. This appears to be evenly balanced; e.g. 15 Rh4 Qa5+ 16 Kf1 f6 17 Rg4 Rf7 or here 17 ... fxe5 18 Nxe5 Rf6!=. (b) 13 Rb1 threatens 14 Rb4. Thus 13 ... Nec6 14 Bb4!? Nf5 with a very complicated trade-off of forces; for example, 15 00 (15 Bxf5 Nxb4! 16 axb4 exf5 17 Nd4 0-0 18 0-0 Be6=/+) 15 ... Nxb4 16 axb4 (16 Rxb4 Qxa3 17 Bb5+ Kf8 with balanced chances) 16 ... 0-0
17 Qd2 (or 17 g4 Ne7 18 Qd2 Qe8!? intending ... Bd7; e.g. 19 g5?! f6, hitting h5) 17 ... Ne7 18 Nd4 (18 g4 f5!) 18 ... Qe8!? intending ... f6, exemplified by 19 Nb5 Qc6 20 Nd6 f6 or 20 Rb3 Bd7 21 Nd6 f6. 13 ... Nxf3 An interesting and possibly very useful option is 13 ... Nef5 (13 ... Nec6?? 14 Rh4! will win a piece shortly), when Black has won several games after 14 Rb1 Nc6 15 Bb5; e.g. 15 ... Qg4 (15 ... Qe4 16 Rb4 d4=/+ Hnydiuk-Zschaebitz, Police 2011) 16 Rb4 d4 17 Rxd4?! Nfxd4 18 Rh4 Qxf3! 19 gxf3 Nxb5-/+ Semenova-Mkrtchian, Rijeka 2010. Of course it’s not that easy. 14 Qxf3 b6 The point. Black gets rid of White’s good bishop and gains access to the light squares. He is willing to give back his pawn to do so. Watch out for 14 ... 0-0?? 15 Bxh6!! gxh6 16 Qf6 and White wins.
15 Qg3 Ba6! Sacrificing pawns in order to seize the initiative, in particular via the light squares. 15 ... Rg8 16 Rh4 is no fun, to say the least. 16 Qxg7 Bxd3+ 17 cxd3 Rg8
18 Qxh6 18 Qh7 Qg4 19 Rh2 (19 g3 Rc8=/+) 19 ... Rc8=/+. 18 ... Qd4 Winning back the pawn and heading towards White’s king. 19 Re1!? White protects his e-pawn. Instead, 19 Rc1 Qxd3+ 20 Kg1 is more active:
(a) 20 ... Nf5 21 Qf4 b5 22 Rc3 (22 Rc7!? Kd8 23 Rc5 Kd7 24 h6 Qe2! 25 Rh3! Rg4 26 Qf3 Qxf3 27 Rxf3 Rb8+/=) 22 ... Qe2 23 Rhh3!? (23 h6 Rg4 24 Qf3 and 24 ... Nd4 25 Qf6 Qe4 26 f3 Ne2+ is unclear, while a draw results from 24 ... Qxd2 25 Qxg4 Qxc3 26 Qg8+ Ke7 27 Qxa8 Qc1+ 28 Kh2 Qxh6+ 29 Kg1 Qc1+=) 23 ... Rg4 24 Qf3 Nd4! 25 Qf6 Qd1+ 26 Kh2 Ne2 27 Rhg3 Rxg3 28 fxg3 Nxc3, Manhardt-Petrik, Guarapuava 1995; and instead of 29 Qh8+??, here 29 Bg5! would have led to perpetual check. (b) The ambitious move 20 ... d4!? has the idea of 21 Bg5 Nd5. I suspect that Black is doing well enough but has no advantage; e.g. 22 Rh4! Kd7 23 Bf6 Rac8 24 Rxc8 Kxc8 25 Qc1+! Kb7 26 h6 Nxf6 27 exf6 Qf3 28 Qf1 Qxf6 29 g3 d3 30 Qxd3 Qxh4 31 h7! Qd8 32 hxg8Q Qxg8 33 Qd7+=. (c) 20 ... Qe4 21 g3 Qxe5 22 Qf4 Qxf4 23 Bxf4 Rc8 (or 23 ... f6 24 Rc7 e5 25 Bd2 Kf7), Vlcek-Petrik, Slovakian Team Ch. 1998; and White should play 24 Rxc8+ Nxc8 25 Kg2. 19 ... Qxd3+ 20 Kg1 Rc8!? This has gone unquestioned in the notes that I’ve seen, but 20 ... d4 appears to be a legitimate option, preventing Qe3. And the obvious 20 ... Nf5 doesn’t seem clear either; e.g. 21 Qf4 (21 Qh7?? Rxg2+ 22 Kxg2 Ne3+) 21 ... Nd4 (or 21 ... 0-0-0) 22
Kh2 Rc8 23 Rc1 Rc4! is double-edged, though I prefer Black. 21 Bg5!? White should try 21 Qe3!; e.g. 21 ... Qh7 22 Rh3 Rc4 23 h6 Nf5 24 Qe2 Qg6 25 Qf1 Rh4 with equal chances. 21 ... Qf5 22 f4 22 Bxe7 Kxe7 leaves Black with an obvious advantage, since his rooks are on open files and White’s are tied to defence, while the white king is subject to attack. Then 23 Qe3 is almost forced, when 23 ... Qg5 24 Qxg5+ Rxg5 25 h6 Kf8! favours Black. 22 ... Rc2 23 Rh2?? Losing. He had to try 23 Qf6! Qxf6 24 exf6 Nf5 25 Rh3 (25 g4 Nd6) 25 ... Nd6! (25 ... d4 26 Rd3 Rh8 27 h6 Rc4!? 28 g4 Ne3 29 f5 is a dynamically balanced mess; 25 ... Kd7 26 h6 also looks pretty complicated) 26 h6 Kd7, when Black’s superior knight offsets the fact that the h-pawn ties him down; probably he stands a bit better. After 23 Rh2, Nakamura-Shulman, US Ch., St Louis 2010, continued 23 ... Qd3 (threatening ... Nf5 and ... Qd4+, with the mating idea ... Ng3) 24 Qf6 Rxg5! (pretty; although 24 ... Nc6!, threatening ... Qd2, will win just as easily) 25 Qxg5 (or 25 fxg5 Nf5 26 Qh8+ Kd7 27 Rh3 Qd2) 25 ... Qd4+ 26 Kh1 Qe3! 0-1. There could follow 26 ... Qe3 27 Rh3 (27 Rg1 Rc1; 27 Qh4 Rc1) 27 ... Qxe1+ 28 Kh2 Rc1 and White will be mated.
Chapter
Eleven
Winawer
Variation:
Portisch-Hook
Variation
1 e4 e6 2 d4 d5 3 Nc3 Bb4 4 e5 c5 5 a3 Bxc3+ 6 bxc3 Qa5
I’ll call this the ‘Portisch-Hook Variation’, since, at least in conjunction with the next move ... Qa4, Lajos Portisch was the strongest advocate of 6 ... Qa5 in its early days, and William Hook was its dedicated practitioner from about the same time on. Today, as you will see, Viktor Moskalenko is its foremost advocate and theoretical backer. This variation complements the Winawer Poisoned Pawn solution of Chapter 9 extremely well. As we will see, the Portisch-Hook involves strategies in non-tactical settings, often closed, and the first direct contact between forces sometimes doesn’t take place until past the 20th move. The tactically sensitive Poisoned Pawn lines, by contrast, require specific and accurate play. While they offer a greater degree of control over the direction of such play, they have the drawback that very concrete continuations are predictable and easily prepared for. Double-edged theoretical variations being what they are, there is also the possibility that some part of the Poisoned Pawn repertoire will fall into disrepute. By contrast, the Portisch-Hook is so positionally-oriented and flexible that it will always be playable whether or not a particular branch is thrown into question. The first thing we notice about 6 ... Qa5 is that Black attacks c3, limiting his opponent’s response. Importantly, this means that he can play this way without letting White have the choice of complicated alternatives he gets after, for example, 6 ... Ne7 or 6 ... Qc7. In fact, White’s next move is chosen almost exclusively in practice: 7 Bd2 Black probably can’t force an advantage versus the other logical moves, but they give White no prospects for advantage other than by transposition: (a) 7 Nf3 Qxc3+?? loses the queen after 8 Bd2 Qb2 9 Rb1 Qxa3 (or 9 ... Qa2 10 Qc1) 10 Rb3 Qa2 11 Qc1 and Ra3, but Black has more effective answers of course: (a1) 7 ... Ne7 8 Bd2 Nbc6 transposes into a Positional Winawer, or here 8 ... Qa4 to 7 Bd2 Qa4 8 Nf3 Ne7 (note ‘d1’ after 7 ... Qa4 below). Notice that 7 ... Nc6!? 8 Bd2 Qa4= also transposes, this time to 7 Bd2 Qa4 8 Nf3 Nc6 (which is note ‘d2’ similarly). (a2) 7 ... cxd4 is considered the real drawback to 7 Nf3:
(a21) 8 Bd3 gives up another pawn for development, but Black stands solidly; e.g. 8 ... dxc3 9 0-0 Nc6 (or 9 ... Ne7 10 a4 h6 11 Ba3 Nbc6) 10 a4 (10 Rb1 Nge7 ‘with two extra pawns’ – Moskalenko) 10 ... h6 (versus Ng5, or a later Bxh7+) 11 Ba3 Nge7 12 Bd6 0-0 13 Nd4 Re8 14 Nb5 Bd7 15 Nc7 Nc8! 16 Nxa8 Nxd6 17 exd6 Rxa8=/+ intending ... Qc5/ ... Qxd6. (a22) 8 Qxd4 Nc6 9 Qe3 Nge7 gains a lot of time for development; notice that ... d4 followed by capture on e5 is becoming an idea, so play can go 10 Bd2 Qa4 (or 10 ... Qc7; or 10 ... b6! with the idea 11 c4 Qa4 12 cxd5 Nxd5 13 Qd3 Bb7=/+ and Black has an open long diagonal and superior pawn structure in return for the bishops) 11 Bd3 b6 (11 ... Nf5 12 Qf4 Qxf4 13 Bxf4 Na5 14 Nd4 is unclear) 12 Qg5! (12 0-0 Ba6; 12 Qf4?! Qxf4 13 Bxf4 Na5 14 Nd4 Bd7) 12 ... Kf8 13 0-0 h6 14 Qh5 Ba6, and a sample line is 15 c4 Bxc4 16 Ng5 Nd8 17 Bb4 Bxd3 18 cxd3 Qd7 19 Rfc1 Kg8 20 Nf3 Ndc6=/+. (b) 7 Qd2 diverts the queen from its potentially damaging presence on g4, while forcing White to always be ready for an exchange of queens. After 7 ... Nc6 8 Nf3, we have: (b1) 8 ... Nge7!? 9 a4 Bd7 10 Ba3 (10 dxc5 Qc7 11 Qg5 Ng6 12 Rb1 0-0=/+ Massoni-Moskalenko, Barbera del Valles 2009) 10 ... cxd4 11 cxd4 Qxd2+ 12 Kxd2 Nf5 13 Bb2 Na5 14 Bd3 Nc4+
15 Kc1 (15 Bxc4 dxc4 16 Kc3 Rc8! 17 Ba3, and 17 ... Bc6 18 a5 Ne7 19 Bxe7 Kxe7 20 Nd2 f6, or 17 ... b6 18 g4 Ne7 19 Bxe7 Kxe7 20 Nd2 Bc6 21 f3 Rhf8 intending ... f6) 15 ... Rc8 16 Re1 Ne7 17 Nd2 Nxb2 18 Kxb2 Nc6 19 Nb3 Nb4 20 a5 Ke7 21 Nc5 Rc7= Westerinen-Farago, Balaguer 2005. (b2) 8 ... b6!? intending ... Ba6 is a little more ambitious: 9 dxc5 bxc5 10 a4 Ba6 (Moskalenko mentions 10 ... Nge7!?, which I think is equal after, for example, 11 Ba3 0-0 12 Bb5 a6 13 Be2 Rd8=) 11 Bxa6 Qxa6 12 Ba3 Rc8 (12 ... Qxa4!? 13 0-0 Qc4 – Moskalenko; but this isn’t as easy) 13 Qd3, Poulton-N.Pert, British League 2002; and 13 ... Qa5 yields a good solid game with a
pawn structure familiar from several Winawer variations. 7 ... Qa4
This is what really distinguishes the Portisch-Hook. The queen move comes at the cost of development but has several advantages: 1) it attacks d4; 2) it aims at c2, discouraging White’s queen from wandering off to the kingside for fear of ... Qxc2, which in addition to winning a pawn provides a nice defensive/offensive function on the b1-h7 diagonal; 3) it blockades the pawn on a3. This last may seem odd (isn’t the queen a poor blockader?) but it prevents White’s valuable move a4, and thus a transfer of White’s queen or queen’s bishop to the a3-f8 diagonal with attack on Black’s abandoned dark squares. The fascinating thing about this variation is the remarkable variety of plans and set-ups both sides have. Let’s consider Black first. He can: 1) play for a direct attack on the d-pawn with ... Nc6 and perhaps ... Ne7-f5; 2) exchange light-squared bishops by ... b6 and ... Ba6; getting rid of his ‘bad’ bishop is not that important for Black, since it otherwise serves a valuable role; but eliminating White’s light-squared bishop is a major achievement, since it is often essential for his long-term attack on the kingside; 3) play to break up the centre with ... f6, or opt for ... f5 with a mind towards queenside play or even a later ... g5; 4) play ... c4, preventing White’s central/queenside action stemming from c4 or dxc5, while stabilizing the queenside situation so as to turn his attention elsewhere; 5) attack with the pawn storm ... b5/ ... a5/ ... b4, usually in conjunction with ... c4. White for his part can: 1) bring his queen to b1, which forces ... c4, then concentrate upon a kingside attack; 2) develop his queen to g4 with the traditional Winawer attack on the g-pawn; 3) bring his knight to e2, foreshadowing Nf4 or the advance f4; 4) bring his knight to f3, supporting the centre and perhaps heading to h4 or g5 later; 5) bring his knight to h3 with an eye on f4 and g5; 6) play in the centre with c4 or dxc5; 7) play on the flank with h4-h5. Because of this multitude of set-ups (and others), it’s impossible to map out a repertoire versus every combination of White’s moves. So, while attempting to be as systematic as is practical, this chapter illustrates ideas and leaves it to you to fill in the continuations when White’s moves haven’t been specified in a precise move order. All that said, let’s divide up the material: 11.1 11.2 11.3
8 8 8
Qb1 Qg4 h4
There are other ways to react to the pressure on d4: (a) 8 dxc5 shows White’s cards too early. A simple solution is 8 ... Ne7 9 c4 (9 Bd3 Nd7!) 9 ... dxc4 10 Qg4 Qxc2 11 Bxc4 (11 Qxg7 Qe4+ 12 Ne2 Rg8 13 Qh6 Nd7=/+) 11 ... Qg6=. (b) 8 c4 resembles other c4 lines below; for example, 8 ... Ne7 (8 ... dxc4!? 9 Qg4 Ne7 10 Qxg7 Rg8 11 Qxh7 Nbc6 with counterplay) 9 c3 Qxd1+ 10 Rxd1 dxc4 11 Bxc4 cxd4 12 cxd4 b6=. (c) 8 Rb1 is rare; it threatens Bb5+.
Unlike the 8 Qb1 lines, however, it has the drawback that the a3-pawn hangs, so White may have to waste resources looking after it: (c1) 8 ... a6 is better than against 8 Qb1, since in that case 9 Qb3 is a problem. Here White’s best move is probably 9 dxc5, when a sample line would be 9 ... Ne7 (9 ... Nd7) 10 Nf3 Qxa3 11 Bd3 Qxc5 12 0-0 Nd7 13 Re1 b5!? (versus c4) 14 Ng5!? h6 15 Qh5 g6 16 Qh4 Qc7, and 17 f4?! Nc5 or 17 Nf3 Nc5 18 Nd4 Nf5 19 Bxf5 gxf5 20 Qf6 Rg8 21 Qxh6 Bb7 22 f3 0-0-0= intending ... Rg6 and ... Rdg8. (c2) 8 ... c4 is normally played: 9 Qc1 (to protect a3; 9 f4!? Nh6!) 9 ... Nc6 10 Be2 was Ivanchuk-Socko, Heraklio 2007; here one idea is to get castled quickly while leaving a knight flexibly placed on g8; e.g. 10 ... b6 11 h4 (11 Nf3 Bd7 12 0-0 h6 13 Nh4 Nge7 14 f4 0-0-0=) 11 ... Bd7 12 h5 h6 13 Rh4 0-0-0 14 Rf4 f6 with mutual chances. (d) 8 Nf3 looks natural and is an important move. But it’s also a bit inflexible because the knight gets in the way of the path of White’s queen to the kingside, while the options of Nh3-f4 or Ne2-f4 are eliminated.
White will frequently play h4 next, which, as explained in the following notes, transposes to either 8 h4 in this chapter, or 7 h4 or 8 h4 in the previous one (see section 10.3). But Black also has some independent replies: (d1) 8 ... Ne7 is committal, because Black doesn’t wait to see if ... f6 might be a good move. Nevertheless, 9 Be2 will lose time after 9 ... b6 and ... Ba6, and 9 h4 transposes to 8 h4 in most cases; e.g. 9 ... Nbc6 10 h5 cxd4 (or 10 ... h6 11 Bd3!? cxd4 12 cxd4 Nxd4) 11 cxd4 Nxd4 12 Bd3 is the Kasparov pawn sacrifice we will refer to often in this chapter (see ‘d21’). (d2) 8 ... Nc6 hits d4 again. Then 9 dxc5 wins a pawn for White, but after 9 ... f6!? 10 exf6 Nxf6, Black gains a 2:0 central majority, opens lines, and develops quickly. Therefore White more frequently White plays 9 h4 and then:
(d21) 9 ... cxd4 10 cxd4 Nge7 11 h5 Nxd4 (11 ... h6 will come to the same thing after 12 Bd3 Nxd4; whereas if White plays 12 Rh4, we transpose into the notes to section 11.3 which uses the order 8 h4 Nc6 9 h5 h6 10 Nf3 Nge7 11 Rh4 cxd4 and now 12 cxd4; playing 11 ... h6 also allows options like 12 c3 and 12 c4, neither considered serious problems because 12 ... Qxd1+ equalizes; nevertheless, 11 ... Nxd4 bypasses these possibilities) 12 Bd3. Now we’ve transposes into the Kasparov Gambit of the previous chapter (see section 10.3), where it arises via 6 ... Ne7 7 h4 Nbc6 8 h5 Qa5 9 Bd2 cxd4 10 cxd4 Qa4 11 Nf3 Nxd4 12 Bd3. In fact, there are several alternate orders after both 6 ... Ne7 and 6 ... Qa5 which arrive at this position, and I’ll be pointing them out at various junctures. Since its theoretical status is fine for Black, I’ll be using it as part of a repertoire choice when it fits. (d22) 9 ... Nge7 10 h5 is purely transpositional, that is, 10 ... cxd4 11 Rh4 h6 transposes to the main line of section 11.3, as does 10 ... h6 11 Rh4, while 11 cxd4 Nxd4 12 Bd3 is again the Kasparov Gambit of Chapter 10, section 10.3. (d3) 8 ... b6
This is a good reaction to the move Nf3 in many different lines of the Winawer, and it is considered a bit of a disincentive to 8 Nf3. Please pay attention to the following ideas, which recur throughout Winawer practice: (d31) 9 Qb1 Ba6 10 Bxa6 Qxa6 (or 10 ... Nxa6 11 Qb3 Qd7!? 12 c4 Ne7 13 cxd5 exd5 14 c3 0-0 with a comfortable game) 11 a4 Qc8 12 Qb5+ Nd7 13 0-0 Ne7 14 Qd3 h6! 15 Nh4 cxd4 16 cxd4 Qc4= Macieja-Jacimovic, Kallithea 2002. (d32) 9 h4 is better suited to lines in which Black has played ... Nge7 or at least ... Nc6. The light-squared bishops come off quickly here: 9 ... Ba6 10 h5 (10 Bxa6 Qxa6 or 10 ... Nxa6; we’ll discuss this common exchange elsewhere in this chapter) 10 ... Bxf1 11 Kxf1 cxd4 (11 ... Nc6 12 h6 g6!?) 12 cxd4 h6 13 Bb4!? Nc6 14 c3 Qxd1+ 15 Rxd1 Nge7=/+. (d33) 9 Bd3 Ba6 (9 ... c4 leads to what several strong players have as a favourite position, with the idea ... Nc6/ ... Bd7/ ... 0-0-0 and ideally ... f6, or ... Nge7 if appropriate; the extra ... b6 is useful because Black’s king goes to b7 in this line) 10 0-0. Here Moskalenko thinks that 10 ... Nd7 may be the most accurate move, because it prepares to answer dxc5 with ... Nxc5, and fits with ... Ne7-c6, attacking d4 and in some cases e5 (after either the dissolution of the d4-pawn or after ... f6). At any rate, White gets no attack with his queen cut off from the kingside and his light-squared bishop being traded off. (d34) 9 c4!?
9 ... dxc4! (this tries to win the light squares; e.g. the long diagonal for ... Bb7 and d5 for a knight; 9 ... Ba6!? cxd5 Bxf1 11 Kxf1 exd5 is also satisfactory) 10 Ng5!? (10 dxc5 Ne7!, ready to occupy d5) 10 ... Nc6 11 Qf3!? Nh6 12 dxc5 Nf5 13 Bxc4 Ncd4!? 14 Qxa8 Nxc2+ 15 Kf1 Qxc4+ 16 Kg1 0-0-/+ with a terrific game for a mere exchange. Here 11 Ne4 Nxd4 12 Nd6+ Ke7 is also good, and 11 dxc5 should be answered with 11 ... Nge7! intending 12 Qf3 Nf5. 11.1 8 Qb1 White takes the time to threaten Bb5+ and forces Black to close the queenside, thus releasing pressure on d4. On the other hand, this costs a tempo and moves the queen to a square from where it can’t join in the kingside fray.
8 ... c4 8 ... a6 9 Qb3! offers an ending in which White’s pawns will be straightened out and the bishop pair will become strong. In response, 9 ... Qc6 10 a4 Ne7 11 Nf3 or 11 a5, intending h4-h5, is not particularly attractive either. After 8 ... c4, White has: 11.11 11.12
9 9
h4 Nh3
Alternatives: (a) 9 Nf3 obviously isn’t bad but, as explained in note ‘d’ above about 8 Nf3, it does block White’s queen off from returning to g4 or h5, and you can argue that a knight can do more from f4 (after Ne2-f4 or Nh3-f4) than from h4 or g5. Black plays 9 ... Nc6 and: (a1) 10 Be2 Nge7 (or 10 ... Bd7 intending ... 0-0-0 and ... f6) 11 0-0 Bd7 12 Re1 h6 13 Bf1 0-0-0 14 g3 Rdf8!? 15 Bh3 g5 with counterplay, Byrne-Jamieson, Haifa Olympiad 1976. (a2) 10 g3 Bd7 11 Bg2 Nge7 (11 ... 0-0-0!?) 12 0-0 b6 13 Nh4 0-0=. (b) 9 Ne2 Nc6 10 g3 f6!? (10 ... Bd7 11 Bg2 0-0-0 12 0-0 Nge7) is a direct attempt to get some space; e.g. 11 exf6 (11 f4 Nge7) 11 ... Nxf6 12 Bg2 (12 Bf4 0-0 13 Qb2 Bd7 14 Bg2 Be8!? 15 Bd6 Rf7 16 Nf4 Rd7 17 Be5 Ne4 18 f3 Ng5=) 12 ... 0-0 13 0-0 Bd7 14 Qb2 b6 15 Nf4 Qa5 16 Rad1 Ne4=. 11.11
9
h4
This has become the main line by quite a margin. White wants to advance h5-h6, while in the typical case where Black plays ... 0-0-0, the pawn on h4 or h5 can also clamp down on counterplay by ... g5. That is true in general, and particularly so in the popular position where White has developed by g3 and Bh3. The h-pawn advance to h5 also frees White’s rook to come to h4 and then to f4 or g4. Of course, such a set-up is time-consuming, and Black has his choice of trying to fight back in the centre right away, or develop first and then try to exploit the presence of White’s advanced pawn to open lines. His king will be ultrasafe on the queenside. If the position stabilizes (an important ‘if’), then typically neither side can break through without elaborate preparation. 9 ... Nc6 10 h5 h6 10 ... Bd7!? is sometimes played, and could actually be an easier move order if you aren’t scared of h6:
(a) 11 h6 Nxh6 12 Bxh6 gxh6 (Black has a lead in development and the open g-file to compensate for his kingside weaknesses) 13 Nf3 (13 Rxh6? Qa5 14 Qb2 Nxd4) 13 ... 0-0-0 14 Be2 (14 Rxh6 Rdg8 15 g3 Rg6 16 Rxg6 hxg6 17 Bg2 Qa5 18 Qb2 Qd8= intending ... f5) 14 ... Rdg8 15 Kf1 Rg6 16 Qc1 f6! 17 Nh4 Rg7 18 Qxh6 (18 exf6 Rf7=/+ in view of 19 Qxh6 Qxc2; or 18 f4? fxe5 19 fxe5 Rhg8 20 Rb1 Nd8 and White can’t do anything useful, whereas Black can shift his pieces to the kingside; e.g. ... Nf7g5-e4, sometimes with ... Rg3, and consolidating with ... b6/ ... Kb7 if needed) 18 ... Rf7 19 Bh5 Rff8=/+. (b) 11 Ne2 f6!? 12 h6 g6 13 exf6!? (13 f4 0-0-0 14 Ng3 Rf8) 13 ... Nxf6 and ... 0-0. (c) 11 Nf3 0-0-0!? (11 ... h6! is equivalent to having gone into the 10 ... h6 line, but in this case having only to deal with 11 Nf3 instead of the array of other options) 12 Be2 Rf8 13 Bf4 f6 14 Qb2 Nge7 15 Rb1 b6, Flyingfatman-Rajlich, playchess.com 2007. ‘=’ (Moskalenko). That might be so. Black can play ... Rhg8 and ... Be8 among other useful moves, so perhaps 16 h6 is a decent try: 16 ... g5 17 Bh2 (17 Bg3 Nf5) 17 ... Rhg8 18 exf6 Rxf6 19 Ne5! Rxh6 (19 ... Nf5=) 20 Nf7 Rf6 21 Nd6+ Kd8 with an unclear situation. After 10 ... h6,
White 11.111 11.112
has
two 11 11
main
tries: Nf3 Ne2
Alternatives: (a) 11 Rh4 is a standard manoeuvre, with ideas of going to f4 and/or g4 to attack Black’s kingside pawns; for example, 11 ... Nge7! 12 Rg4 and then: (a1) 12 ... Rg8 13 Rf4 Bd7 14 Nf3 Rf8 (14 ... 0-0-0!, and 15 Rxf7?! Be8 wins the h-pawn while activating the bishop, so 15 g3 Rdf8 16 Bh3 Kb8 17 Qd1 Bc8=, Modiakhi-Toufighi, Subic Bay 2009, might be preferred) 15 g3 0-0-0 16 Bh3 Kb8 17 Kf1, Hamdouchi-Sethuraman, Guingamp 2010; and 17 ... Bc8 introduced a protracted manoeuvring game with approximate equality. A more active sequence would have been 17 ... Ka8 18 Kg2 f5! threatening 19 ... g5. There might follow 19 exf6 (19 Rh4!? Nc8 20 Qb2 Nb6 intending ... Qa5 and ... Na4 is unclear) 19 ... gxf6 with an eventual ... e5 in store. At any rate, Black’s main plan in these positions is to open a few lines on the kingside. (a2) 12 ... Nf5 is natural; there might follow 13 Nf3 Bd7 (13 ... 0-0!?) 14 Rf4 g5! 15 hxg6 (15 Rg4? 0-0-0 16 Qb2 Be8-/+) 15 ... fxg6 16 g4 (16 Nh4 g5 17 Nxf5 exf5 18 Rf3 0-0-0=/+; 16 Nh2 g5 17 Rf3 0-0-0 18 Ng4 Be8=) 16 ... g5 17 gxf5! exf5! 18 Nh4 gxf4 19 Ng6 Rg8 20 Nxf4 0-0-0 21 Nxd5 (21 Bg2 Ne7 22 Kf1 Bc6=/+) 21 ... Qa5 22 Nf4 Nxe5! 23 dxe5 Qxe5+ 24 Ne2 Rde8 25 Be3 f4 26 Bd4 Qd5 and Black has an effective attack with good winning chances. b) 11 g3 Bd7 12 Bh3 Nge7 13 Ne2 will usually transpose to 11 Ne2 Nge7 12 g3 Bd7 13 Bh3 below; 13 Nf3 is the 11 Nf3 Nge7 12 g3 Bd7 13 Bh3 main line. A rare option for White is 13 f4 (13 Qb2 0-0!? or 13 ... 0-0-0 14 f4 f6) 13 ... 0-0-0 14 Nf3 Rdf8 15 0-0 g6!? (or 15 ... b6 with the idea 16 Nh4 g6!) 16 Kh2?! (16 hxg6 Nxg6=), Becerra-Xzibit, ICC 2010; 16 ... gxh5! 17 Nh4 Ng6 18 Ng2 f5!=/+. (c) 11 Be2 has been played a few times, keeping options open:
(c1) 11 ... Nge7 12 Nf3 (the equivalent of 11 Nf3 Nge7 12 Be2) 12 ... Bd7 and: (c11) 13 g3 0-0-0 (13 ... b6 14 0-0 0-0) 14 Kf1 Kb8 15 Kg2 Ka8 16 Qb2 Rb8 with complications; Black might play ... Nc8-b6 next. (c12) 13 Bf4 0-0-0 (when White commits to Bf4, perhaps it’s time for 13 ... Qa5; e.g. 14 Qb2 Nc8 15 0-0 Nb6 16 a4 Nd8 17 Qc1 Nxa4 18 Ra3 b5) 14 Qc1 Kb8 15 0-0 Ka8 16 Nh2, Jakubowski-Sieciechowicz, Ostroda 2011; and one idea was 16 ... Nc8 17 Ng4 Nb6 18 Qd2 Qa5 19 a4 Nxa4 20 Ra3 Ne7 21 Ne3 Qc7 22 Rfa1 b5 with an odd position in which both sides may have trouble making progress; eventually I suppose Black might organize to advance on the queenside, and White on the kingside (where their respective kings are!). (c13) 13 0-0 0-0-0 (13 ... b5 has the idea ... 0-0 and ... Kh7, perhaps followed by a combination of ... Qa5-c7 and ... Nc8b6-a4) 14 Nh4 g5 15 Nf3 Ng8! (this wouldn’t work if White could play f4) 16 Nh2?! (16 Be3 f5 17 exf6 Nxf6 18 Ne5 Be8=) 16 ... f5 17 exf6 Nxf6=/+ (Black has ideas of ... Ne4 along with ... e5), Shashikant-T.Luther, Mumbai 2010. (c2) Moskalenko says that he might prefer 11 ... Bd7, keeping open the possibility of ... f6, and after exf6, ... Nxf6. Here’s some random analysis with that idea: 12 Nf3 0-0-0 13 0-0 f6 (13 ... Be8 could prepare this by covering g6 and h5) 14 Nh4!? fxe5 15 Ng6 Rh7 16 dxe5 Nge7 17 Nxe7+ Nxe7 18 Qb4 (18 Bg4 g5 19 hxg6 Nxg6 20 Qb4 Qc6 is dynamically balanced) 18 ... g5 19 hxg6 Nxg6 20 Qc5+ (20 Qxa4?! Bxa4 and Black stands well) 20 ... Kb8 21 Rfb1 Qc6 22 Qd4 Ne7 23 Be3 Qa6 24 Qf4 Rg8 25 Bc5 h5 26 a4 Nf5 27 a5 Ka8 with mutual chances, although I prefer Black’s. 11.111 11 Nf3 Nge7 11 ... Bd7 again keeps open the possibility of omitting ... Nge7 and playing for ... f6 or ... f5 (see the note ‘c2’ on 11 Be2). This would also apply to g3 systems; e.g. 12 g3 0-0-0 (or 12 ... Nge7 13 Bh3 b6 14 Qb2 0-0) 13 Bh3 b6 14 0-0 f5!? 15 Nh4 Nge7, and both sides will be stuck for a while; as always, there are options along the way. 12 g3 12 Be2 returns to 11 Be2 above (see note ‘c1’).
12 ... Bd7 13 Bh3 0-0-0 14 0-0
This position has arisen quite a few times. A couple of examples: 14 ... Rdf8 15 Nh4 15 Re1 is important, discouraging an f-pawn move by indirectly hitting e6: (a) 15 ... g5!? 16 Nh2! f5!? (Black doesn’t want to be squeezed to death after Ng4; but 16 ... Ng8 17 Ng4 f5 18 exf6 Nxf6 19 Nxf6 Rxf6 is quite reasonable; here 17 f4 meets with 17 ... f5!) 17 exf6 Ng8!? (17 ... Rxf6 18 f4 gxf4 19 Bxf4 Rg8 affords counterplay, and here 18 Ng4 Rf7 intending ... Nf5 looks playable; still, Black’s 16th-move option is more straightforward) 18 Bxe6 Bxe6 19 Rxe6 Nxf6 20 Qd1 Kb8 21 g4 Ne8 22 Nf1 Nc7 23 Re1 Rh7 24 Ng3 Ne7 25 Re3 Rhf7 26 f3 and Black lacked sufficient compensation in Svidler-D.Lima, Khanty-Mansiysk 2011. (b) 15 ... b6 16 Nh2 Kb7 17 Ng4 Nf5 is one of those positions where both sides will be searching for a plan that doesn’t boomerang; this may only be equal. (c) 15 ... g6 is more active; then 16 Qc1 Ng8!? might be alright, because 17 hxg6 fxg6 18 Nh4 Nce7 prepares ... g5 followed by ... Ng6 or ... Nc6, when the other knight slides into e7. 15 ... g5! 16 Ng2 f5 17 exf6 Rxf6 18 Ne3 Rf7 19 Re1 Ng8! 20 Ng4 Nf6 21 Nxf6 Rxf6 22 Re2 Rhf8 23 Bg4 Nd8= Sadvakasov-Pogorelov, Moscow 2002. 11.112
11
Ne2
11 ... Bd7 Without a knight on f3, this could be a time to venture 11 ... f6, assuming that 12 Nf4 fxe5 13 Ng6 Rh7 isn’t a problem. With White’s bishops unable to get active, I think Black should be fine; e.g. 14 dxe5 (14 Nxe5 Nxe5 15 dxe5 Ne7 16 Be2 Nc6, or 16 ... b6 intending ... Bd7, ... 0-0-0, and perhaps ... Rf8 with ... Be8; it’s not clear what White can do to hurt Black) 14 ... Nge7 15 Nxe7 Kxe7 16 f4 Rh8=. 12 Nf4 0-0-0 13 g3 Nge7 13 ... f6!? 14 Bh3 Re8, Willemze-Heemskerk, Groningen 2010; and White’s best bet was to try for a bind with 15 Ng6 Rh7; e.g. 16 exf6 (16 Qb2 fxe5 17 Rb1 b6 18 Nxe5 Nxe5 19 dxe5 Ne7 is satisfactory for Black at any rate) 16 ... Nxf6 (hitting h5) 17 Qd1 e5!? 18 Bxd7+ Nxd7 19 dxe5 Ncxe5 20 0-0 Nxg6 21 hxg6 Rhh8 22 Bf4 Nf6 with chances for both sides. 14 Bh3 Kb8 15 0-0 This is a basic position which has arisen many times and gone in many different directions. I like the following plan: 15 ... Nc8
This is Black’s alternative to challenging the e5-pawn or attacking on the kingside. The strategy of ... Nb6 and ... Na4 depends upon the idea that White will have difficulty making progress on the kingside. 16 Ra2 After this, White can still play Qd1-g4. Instead: (a) 16 Kh2 Qa5 17 Qd1 Nb6 18 Ng2 Rdg8 19 f4 g6 (a standard idea) 20 Rf2 Ne7 21 Rf3 Bc8 (21 ... Nf5 22 g4 Ne7) 22 Ne3 Nf5 23 g4 (23 Ng4! Nd7=/+) 23 ... Nxe3 24 Rxe3 gxh5 25 gxh5 Nd7 26 Qe1 Qa4 and again Black stood better in Najer-Zhang Pengxiang, Ergun 2006. My impression is that these positions are generally equal, but perhaps White has a slightly harder position to play. On the other hand, he didn’t have to commit to positive action. (b) 16 Qb2 Nb6 17 Rfb1 Bc8 18 Ra2 Rde8 19 Kg2 Rhg8 20 Ne2 Re7 21 Raa1 (I’m not sure what White’s plan is) 21 ... f6!? 22 Re1 Nd7 23 Nf4!? (23 exf6 Nxf6 24 Nf4 Ne4+/=) 23 ... fxe5 24 Ng6 Ree8 25 dxe5 Nf8 26 Nxf8 Rexf8= Ohtake-Soza de la Carrera, ICCF server 2006.
16 ... Nb6 17 Qd1 Qa5 18 Qg4 Rhg8 18 ... Nxd4 19 cxd4 Qxd2 20 Rd1 isn’t worth it. 19 Ne2 Stopping ... Nxd4 and preparing f4-f5. It’s not a good sign that White might need such a slow plan, which seems unlikely to succeed, but it’s hard to see what he can do. Of course, Black also has a difficult time finding an effective plan. 19 ... Na4 20 f4?! This backfires positionally. 20 ... f5! 21 exf6 21 Qf3 Be8 followed by ... Qc7 and ... g6 gives Black a kingside initiative. 21 ... gxf6 22 Qh4 f5!
Securing e4; Black is definitely better now. 22 ... e5 23 Bxd7 Rxd7 is also possible. After 22 ... f5, Kosteniuk-Hoang Thanh Trang, Tbilisi 2011, continued 23 Kh1 Rdf8 24 g4 (White doesn’t want to wait for Black to bring a knight to d6 via c8 and double on the g-file) 24 ... fxg4 25 Bxg4, and here 25 ... Rf7! with the idea of ... Ne7-f5 achieves a nice edge for Black. White’s rook on a2 and bishop on d2 aren’t doing enough. 11.12
9
Nh3
In this manner White retains the option of f4 and leaves open a path for his queen on the kingside. 9 ... Nc6 9 ... f6!? 10 f4 led to an interesting struggle in Konguvel-Moskalenko, Andorra 2006, where Black, when he couldn’t make any progress on the kingside, turned to a characteristic if somewhat rare plan which is worth mentioning: 10 ... Nc6 11 Nf2 Nge7 12 g3 0-0 13 Bh3 Rb8 (13 ... fxe5 14 fxe5 Rb8) 14 0-0 Bd7 15 Qd1 f5 16 Bg2 b5 17 h3 Qa5 18 Qe2 Qc7 19 Nd1 a5
Here it is: the attack by ... b5/ ... a5/ ... b4. Often this is used as a last resort, but several strong players have employed it, and in this position it’s perfectly good: 20 Ne3 b4 21 axb4 axb4 22 g4 Kh8 23 Be1 Be8! (this enables ... Bg6 if desired, and is a well-known idea in this race between pawn attacks) 24 Qd1?! (24 cxb4 Nxb4 25 Bc3=) 24 ... b3 25 cxb3 and here 25 ... cxb3 would have kept the advantage. 10 Nf4 10 Be2 Bd7 11 0-0 0-0-0 12 Nf4 is an alternate way to reach 10 Nf4 Bd7 etc in the following note. 10 ... Nge7 Or 10 ... Bd7 11 Be2 (11 Nh5 g6 12 Nf4 Nge7 13 h4 h6) 11 ... 0-0-0 (11 ... Nge7 12 0-0 0-0!? 13 Nh5 Ng6 with the idea 14 Qxb7?! Qxc2) 12 0-0 f6 13 exf6 (13 Nh5!? fxe5 14 Nxg7 exd4 15 Bg5!, Fedorov-Rustemov, Russia 2002; 15 ... dxc3!? 16 Bxd8 Nxd8 with compensation – Fedorov) 13 ... Nxf6 14 Qb2 Ne4 15 Rab1 b6 16 Be1 (16 Be3 Qa5) 16 ... Rhf8 17 Nh5 g6 18 f3! unclear (Rustemov). 11 g3 b6 12 Bg2 Bd7 13 0-0 0-0-0
13 ... g6 is better, according to Fedorov; then ... h6 and possibly ... g5 could follow. 14 Nh5 This move can sometimes be embarrassing, because Black usually can’t play ... g6 without a serious amount of preparation. But there are ways of dealing with the problem. 14 ... Rhg8 15 Qb2 Kb7 16 Rae1 Rdf8 17 f4
Fedorov-Asrian, Bled Olympiad 2002; and here simply 17 ... Nf5 looks fine, with various ideas; one is ... f6, either next or after preparatory moves such as ... Be8 and ... Nd8. In the game, 17 ... Ka6 18 g4 h6 19 Re2 might have been met by 19 ... f5!? 20 exf6 gxf6 21 h3 f5!? 22 g5 Rh8 with complications. 11.2
8
Qg4
White’s most aggressive move, forcing Black to attend to g7. We’ll look at two ways of doing this: 11.21 11.22 11.21 8 ... Kf8
8 8
... ...
Kf8 g6
This is an old move but with a modern flavour, refusing to create weaknesses at the cost of losing the castling privilege. Black also threatens 9 ... Qxc2. I’m going to recommend 8 ... Kf8 as my first choice, with a less-detailed examination of the ambitious 8 ... g6 as well, if only to present the rudiments of an alternative should 8 ... Kf8 fall into disrepute. 9 Qd1 This normal move protects c2; White is happy to have the king looking forlorn on f8. This is easily the main continuation, but by no means the only one: (a) 9 Nf3 hasn’t been played here as much as 8 ... g6 9 Nf3, but is beginning to appear more often. White has the same ideas (Qf4, for example), and can get either an attack or favourable ending if Black lets him. One thing to be aware of is that if Black plays ... Qxc2, g6 is available as a retreat square for Black’s queen, so White will not have the Qg3 strategy that he uses after 8 ... g6 9 Nf3 or 9 Qf4.
The play is complex: (a1) 9 ... Qxc2 can be answered in two main ways: (a11) 10 Be2 b6!? (10 ... Ne7) 11 dxc5!? bxc5 12 Rc1 Qe4 13 Qxe4 dxe4 14 Be3! Nd7 15 Ng5 Bb7 16 Rd1 Nxe5 17 Bxc5+ Ne7 18 Rd4 f6 19 Nxe4! (19 Nxe6+ Kf7 20 Nxg7 N7c6 21 Nf5 Nxd4 22 cxd4 Nd3+ 23 Bxd3 exd3 24 Kd2 Bxg2 25 Rg1 Ke6 26 Ng7+ Kf7 27 Nf5=) 19 ... Bxe4!? 20 Rxe4 Rc8 21 Bxa7 Rxc3 22 a4 Rc1+ 23 Bd1 Kf7 24 0-0 Rc7 25 Bb6 Rc6 26 Bd4 Nc4. Two knights versus two bishops is generally pretty bad in chess, but Black is well within bounds here, and you’ll see several similar situations in this variation with the knights at least holding their own, usually due to a relatively better pawn structure and active pieces. (a12) 10 Rc1 Qg6 11 Qxg6 hxg6 12 c4 Ne7
13 dxc5 (13 cxd5 Nxd5! 14 Rxc5 Ke7 15 Bd3 b6 16 Rc2 Bb7 17 Be4 Nd7 18 0-0 Rhc8=) 13 ... Nbc6!? (13 ... Bd7! brings the bishop to an active long diagonal; play might go 14 cxd5 Nxd5 15 Be2 Bc6 16 Nd4 Nd7 17 f4 g5! 18 fxg5 Rh4 19 Nxc6 bxc6 20 0-0 Nxe5 21 Rb1 Ra4 22 Rb3 Rd8=, when Black’s active knights and White’s inferior pawn structure compensate for the bishop pair) 14 Bd3 Rh5 15 Bf4?! (15 0-0! dxc4 16 Rxc4 Nxe5 17 Nxe5 Rxe5 18 Bf4 with compensation and probably a bit more) 15 ... Na5 16 cxd5 Nxd5=/+ 17 Bg3 Bd7 (17 ... Nb3 18 Rc2 Bd7 19 0-0 Rc8=/+) 18 Nd2 Ba4?! (18 ... Bc6=) 19 0-0 Rd8 20 Be2! Rh8 21 Ne4+/= Cabrera-Scavo, Burguillos 2009. A very instructive example showing both sides’ ideas. (a2) 9 ... Nc6 (it’s often useful to delay a capture on c2 until White protects his queen) 10 Qf4 (now dxc5 is a kind of positional threat; 10 Ra2 can be answered with 10 ... Nge7, but this is also a point where 10 ... b6 makes sense: 11 Qf4 Ba6 12 Bd3 Bxd3 13 cxd3 c4 14 Ng5 Nd8=; this defensive retreat is an advantage of playing ... Nc6 early on) 10 ... Qxc2
11 dxc5 (11 Be2 Nge7 12 dxc5 h6 keeps a knight out of g5 and takes advantage of the fact that White’s queen is somewhat restricted in its movements; e.g. 13 0-0 g5! 14 Qe3 Nf5 15 Qd3 Qxd3 16 Bxd3 Nh4! 17 Nxh4 Nxe5 18 Bb5! gxh4 19 Bf4 f6 20 Bxe5 fxe5 21 Rae1 Rh7 22 Rxe5 Rc7=) 11 ... Nge7 12 Be2 h6 13 0-0 (13 Nd4 Qe4=) 13 ... Qe4 (13 ... g5 is odd but fully playable in view of 14 Qf6? Rh7, so 14 Qe3 Qe4 15 Nd4 Qxe3 16 Bxe3 Nxe5 17 f4 gxf4 18 Bxf4 N7g6=) 14 Rfe1 Qxf4 15 Bxf4, Khachiyan-Shulman, ICC 2011; and several moves suffice, but 15 ... g5 is direct: 16 Bd2 (16 Be3?! Nf5 17 Bc1 g4) 16 ... Ng6 17 Rab1!? Kg7 18 Bd3 Rd8 19 h3 Rb8 and ... Bd7=. b) 9 c4 is rare but interesting: 9 ... Nc6!? 10 dxc5 Qxc2 (or 10 ... Nxe5 11 Qf4 Nxc4 with mutual chances following 12 Nf3! f6!? 13 Bxc4 Qxc4 14 Qc7 a5! 15 Rb1 Qa4=) 11 Nf3 h6 12 Qf4 Qe4+ 13 Qe3 Nge7 14 cxd5 exd5 15 Bd3 Qxe3+ 16 Bxe3 Bg4=. 9 ... b6
9 ... Nc6 is not my first recommendation, but is also used a lot and deserves a brief look. After 10 Nf3, here are a few samples of play from the two main responses (excluding ... b6 systems): (a) 10 ... Nge7 can lead in many directions, including these two:
(a1) 11 Qb1 c4 12 h4 Ke8 13 h5 h6 14 g4 Kd8!? 15 g5 (active, but creating structural problems; 15 Bg2 Kc7 16 Nh4 Bd7 17 f4 Raf8 or 17 ... f6 ‘and I prefer Black in the long run – at least his king is more safe’ – Moskalenko) 15 ... hxg5 16 Nxg5 Ke8 (typically running back-and-forth; Moskalenko has pointed out this strange tendency in the ... Kf8 lines) 17 Be2 b6 18 Bg4 Bd7 19 Qd1 Ng8! 20 Be3 Nh6 21 Kd2, Karjakin-P.Nikolic, Wijk aan Zee 2005; Black kept a small edge after the amusing reversal 21 ... Kd8, but 21 ... Ne7 is more appealing in the sense that it completely stops any play on the kingside and takes control of the outpost on f5. (a2) 11 dxc5 Qe4+ (Moskalenko mentions 11 ... Ng6!?, when 12 Be3! and Bd4 is somewhat advantageous for White, because 12 ... Ngxe5 13 Nxe5 Nxe5 14 Qd4! can go into an ending in which the bishop pair is the main factor; then instead of 14 ... Qxd4 15 cxd4 Nc4 16 Bf4, Black should play 14 ... Nc4, but it’s likely that White retains an edge) 12 Be2 Nxe5 13 Nxe5 (13 0-0 Nxf3+ 14 Bxf3 Qc4 15 Be3 Qxc3 16 Bd4 Qc4 17 c3 intending Be2 is dangerous for Black) 13 ... Qxe5 14 0-0 Bd7 and ... Bc6, which may still be better for White because of Black’s king position, but Black has a superior pawn structure that keeps prospects fairly balanced. (b) 10 ... c4 11 h4 Bd7 12 h5 h6 13 Rh4 (13 g4 Re8! intends ... Ke7-d8, according to Moskalenko; then play might go 14 Nh4 Ke7 15 Qb1 b6 16 Rh3 Rf8 17 Rf3 Ke8 18 Be2 Nge7 19 Kf1 Kd8 20 Kg1 Kc7=) 13 ... Nge7 14 Rf4 Kg8! ‘and then ... Rf8 is more than sufficient’ – Moskalenko. I’d be more cautious and assess this as equal.
10 h4 Instead: (a) 10 c4 used to be considered critical, but a few simple solutions have been found:
(a1) 10 ... Ba6 11 dxc5 bxc5 12 c3!? Qxd1+ 13 Rxd1 Ne7 14 cxd5 (or 14 Bg5, Marques-Pereira, Brazil 2011; and here simply 14 ... Nbc6 looks satisfactory) 14 ... Bxf1 15 Kxf1 Nxd5 and the knights are a match for White’s mediocre bishop, especially as Black will get his pieces out to better positions. (a2) 10 ... Ne7 11 cxd5 (11 dxc5 dxc4!? 12 Ne2 Bb7 13 Nc3 Qc6 14 cxb6 axb6= 15 Bf4 Nd7 16 Qd6 Qxd6 17 exd6 Nf5! – Wei Ming) 11 ... Qxd4 12 dxe6 Qxe5+ 13 Be2 and in this position, Black has two recaptures of interest and a developing move: (a21) 13 ... Qxe6!? 14 Nf3 Ba6 (14 ... Nbc6 15 0-0 Bb7 16 Re1 Qd5 17 c4 Qf5 18 Bc3 Rd8 19 Qb3 f6 20 a4 Ng6!= intending ... Nf4) 15 Be3 Nd5 16 Ng5 Qd7 17 Bxa6 Nxa6 18 c4 Nf6 19 Qc2 h6 20 Nf3 Ng4=. (a22) 13 ... Bxe6 14 Nf3 Qd5 is double-edged; Black has better central control, but needs to get his king tucked away. (a23) 13 ... Bb7 14 Bf4 (14 Be3 Nbc6 15 Nf3 Qxe6=/+) 14 ... Qc3+ 15 Bd2 Qe5 16 Bf4 Qc3+ 17 Bd2 ½-½ Grischuk-Shulman, Khanty-Mansiysk 2005. (b) 10 Nh3 isn’t as good as in positions where ... c4 has been played. Here Black can counterattack on d4; for example, 10 ... Nc6 (or 10 ... Ne7; e.g. 11 dxc5 bxc5 12 Bd3 Nd7! 13 f4 Rb8 14 0-0 Ba6=/+, planning to answer 15 Ng5 with 15 ... h6 16 Nf3 g6 and ... Kg7) 11 dxc5 (11 Rb1 Ba6; 11 Bd3 cxd4 12 Qe2 Nge7 13 0-0 dxc3 14 Bxc3 Bd7 15 Ng5 h6 16 Nf3 Kg8 and White has some compensation but no way to get to the king) 11 ... bxc5 (11 ... Ba6!?) 12 f4 Nge7 13 Bd3 Rb8 14 0-0 Kg8 15 Be3 c4 16 Be2, and Black stands better after either 16 ... h6, securing the usual safe square for his king on h7, or 16 ... Nf5 17 Bc1 (versus ... Rb2) 17 ... h6 18 Bg4 Nce7=/+. (c) 10 Nf3 has several answers. I’ll recommend 10 ... Ba6 with these ideas:
(c1) 11 Bxa6 gives Black a choice of recaptures: (c11) 11 ... Qxa6 12 dxc5! (12 a4 Ne7 13 Qe2 cxd4 14 Qxa6 Nxa6 15 cxd4 Rc8=) 12 ... Nd7!? 13 cxb6 axb6 14 Qb1!+/= intending Qb4, though this is quite playable for Black. (c12) 11 ... Nxa6! 12 Qb1 (12 0-0 Ne7=/+) 12 ... Ne7 (12 ... c4!?) 13 Qb3 Qc4 14 Qxc4 dxc4 15 0-0 h6= with ... Nd5 next. (c2) 11 dxc5 Bxf1 12 Kxf1 bxc5 (or 12 ... Nd7! 13 cxb6 axb6 with excellent play, an idea which pops up in a lot of variations) 13 g3 Nd7 14 Kg2 Ne7 15 Rb1 h6 16 h4 (16 Qe2 g5! 17 Rhe1 Nb6) 16 ... Nc6 17 Qe2 Qe4 (or 17 ... Ke7). (c3) 11 Bd3 leads to a typical ending following 11 ... Bxd3 12 cxd3 Qxd1+ 13 Kxd1 cxd4 14 cxd4 (14 Nxd4 Ne7 15 Ke2 Nbc6 16 Nxc6 Nxc6 17 f4 h5= intending ... Rc8, ... f6) 14 ... Nc6 15 Ke2 Nge7 16 Rhc1 (16 Rac1 Ke8 17 Bb4 Kd7=/+) 16 ... Ke8 17 Bb4 Kd7, at least equal, planning to expand on the kingside by ... g5 and, if g4, ... Ng6-f4.
10 ... Ba6 Not the only move; in fact, both of the following look good: (a) 10 ... Ne7 11 h5 h6 12 Rh4 Ba6 13 Bxa6 Nxa6 (not bad, but 13 ... Qxa6! intending ... Nbc6 and ... Nf5 if convenient, combined with ... Qc4, should give Black excellent play) 14 Rf4 Qd7!? (14 ... Kg8; 14 ... Qc4) 15 Qf3! Nc6 16 Nh3 with mutual chances, R.J.Fischer-Hook, Siegen Olympiad 1970. (b) 10 ... Nc6 forces White to defend d4 and thus interferes with him bringing his queen to the kingside. Then the exchange of bishops should lead to equality; e.g. 11 Nf3 Ba6 12 h5 h6 13 Rh4 Bxf1 14 Kxf1 Nge7?! (14 ... c4! with an easy game) 15 Rf4!? (15 dxc5! Qa6+ 16 Qe2 Qxe2+ 17 Kxe2 bxc5 18 Be3! c4 19 Rb1 Nd8 20 Bc5 Ke8 21 Rhh1 Kd7 22 Rb2+/=, although this is playable for Black) 15 ... Rc8 16 Kg1 (16 dxc5 Qa6+ 17 Kg1 bxc5=; then 18 c4 d4 is a normal position, or you can mix it up a little
with 18 ... g5!? 19 hxg6 Nxg6) 16 ... cxd4 17 cxd4 Ke8 (a typical swing to the queenside) 18 g4?? (18 Rg4! g5! 19 hxg6 Nxg6=, threatening ... Ncxe5) 18 ... g5! 19 hxg6 (19 Rf6 Ng8) 19 ... Nxg6 20 Kg2 Nxf4+ 21 Bxf4 Ne7 winning, Renteria-Ruiz Sanchez, Totana 2011. 11 Bxa6 Others: (a) 11 Nf3 Bxf1 12 Kxf1 h6 and ... Ne7/ ... Kg8-h7 is normal, but in this case 12 ... h5 should be considered, since it both limits White’s expansion and secures f5 for a knight. (b) 11 h5 Bxf1 12 Kxf1 h6 13 Rh4 Ne7 (13 ... Qa6+ 14 Ne2 Nc6 15 Kg1 Nge7 16 Rf4 Kg8 17 Ng3, Ethan Hunt-Captain Nemo, playchess.com 2007; then 17 ... Rh7!? – versus any Rg4-Rxg7+ ideas – 18 Qf3 Rf8 19 Rg4! Kh8 20 Rf4 Kg8 21 Rg4 repeats) 14 Rf4 Nbc6 15 Nf3 (15 Qf3?! Qxc2 16 Rxf7+ Ke8) 15 ... Na5 16 dxc5 Qb5+ 17 Kg1 bxc5 18 Rb1 Qc6=.
11 ... Qxa6 11 ... Nxa6 is also fine; e.g. 12 h5 h6 13 Rh4 Rc8 14 Rf4 cxd4 15 cxd4 Ne7! 16 Qf3 Ke8 17 c3 Rf8 18 Nh3 Nb8 (18 ... Nc7!) 19 Qg4 Rg8 20 Qf3 Rf8 21 Qg3 Rg8 22 Qf3 ½-½ Cheparinov-P.Nikolic, Wijk aan Zee 2005. 12 h5 12 Nf3 can be met variously; for example, by 12 ... cxd4 13 cxd4 Ne7 14 h5 h6 15 g4!? (15 Qe2 Qxe2+ 16 Kxe2 Nbc6 17 Kd3 Ke8 18 a4 Kd7 19 a5 b5! 20 a6 Rab8 21 Rhb1 Rb6=) 15 ... Nbc6 16 Bb4!? (16 Qe2 Qxe2+ 17 Kxe2 Ke8 18 Kd3=) 16 ... Rc8 17 Bd6 Na5=/+. 12 ... h6 13 Rh4 Ne7
14 Rf4 14 Nf3 Nd7 prepares ... Nbc6 and may well favour Black slightly; instead, S.M.Larsen-Slawinski, ICCF 2010, continued 14 ... Nbc6 15 Qe2 Qxe2+ 16 Kxe2 c4!? (16 ... Na5) 17 a4 Ke8 18 Rf4 Rf8 19 Bc1 f6 20 exf6 gxf6=. 14 ... Nbc6 15 a4 Rc8 16 a5 cxd4 Or 16 ... b5!? 17 Qf3 Nf5 18 Ne2 b4, which is complex and dynamically equal. 17 cxd4 Castellanos-Vilela de Acuna, Balaguer 2006. Here Moskalenko suggests that after 17 ... bxa5!, at least ‘Black cannot lose!’. 11.22
8
...
g6
A lot of theory has accumulated on this move which, by contrast with 8 ... Kf8, keeps open the possibility of castling (probably queenside). It comes at the cost of weakening the kingside dark squares when there is no dark-squared bishop left and, because of that, the defence can be difficult, especially in practical terms. Still, 8 ... g6 is extremely instructive and, if you know the concrete lines well, you might find that it is sufficiently safe and offers more dynamic counterattacking chances than 8 ... Kf8. 9 Nf3 I consider this quite dangerous, but White has several other challenging continuations: (a) 9 Rc1 Nc6 (9 ... b6) 10 Nf3 h6!? (10 ... b6 11 Qf4 h6) 11 Bd3 c4 12 Be2 Bd7 13 Qf4 Rh7 14 h4 0-0-0, Parma-L.Portisch, Bled 1961. (b) 9 Qd1 is the traditional move, returning to protect the c2-pawn. In general, I don’t think Black has many problems here, so I’ll limit myself to one reply and not give many details: 9 ... b6 10 h4 Ba6 (10 ... h5 is a little strange, but stops h5; as long as a knight is covering f6 Black seems alright; e.g. 11 Nf3 Ba6 12 Bd3 Nd7! 13 0-0 Bxd3 14 cxd3 Qxd1 15 Rfxd1 Ne7 16 Bg5 Nc6= Skorchenko-Lysyj, Vladimir 2002) 11 Bxa6 Nxa6 (11 ... Qxa6 12 h5 Nc6! 13 hxg6 fxg6 14 Qg4 Kd7! 15 Nf3 h5! 16 Qxg6 Nge7 17 Qd3 Qxd3 18 cxd3 cxd4 19 cxd4 Nf5 with compensation, Apicella-Jacimovic, Fügen 2006) 12 h5 (12 Ne2 h5 13 Bg5 Rc8 14 0-0 Ne7 15 Bf6 Rf8= M.Chandler-N.Pert, London [m/6] 2000) 12 ... cxd4 13 hxg6 fxg6 14 Qg4 (14 Rh4 Ne7 15 Nf3, and Wei Ming suggests 15 ... Nc5! intending ... Ne4, or 16 Rxd4 Qd7! 17 c4 Nc6 18 Rh4 0-0=/+) 14 ... Ne7! 15 cxd4 (15 Qxe6? Nc5! 16 Qg4 Qxc2 17 Qxd4 0-0–+; after 15 Nf3, Black can insist by 15 ... h5! 16 Qxe6 and play a series of forcing moves to advantage: 16 ... Nc5 17 Qf6 Rf8 18 Qg7 Ne6! 19 Qh7 dxc3 20 Bh6 Rxf3! 21 gxf3 Rc8!-/+ intending ... Nd4) 15 ... h5! 16 Qh4 (16 Qxe6? Qxd4 17 Rd1 Qe4+! 18 Ne2 Nc5 19 Qh3 0-0) 16 ... Qxc2 17 Nf3 Rf8 18 Rc1 Qd3 19 Rc3 Qb1+ 20 Rc1 Qd3= (Wei Ming). (c) 9 Ra2 b6 (Black’s main strategy against slower moves) 10 Nf3 Ba6 11 Bd3! (probably the best way to get developed; 11 Bxa6 Qxa6 12 Ng5!? Nc6! 13 Qf4 Nd8 14 h4 cxd4 15 cxd4 h6 16 Nf3 Rc8=/+) 11 ... Bxd3 (11 ... Nd7) 12 cxd3 c4!? (thematic, going for the light squares, although it does take time from development; 12 ... Nd7 13 0-0 Ne7 is double-edged, e.g. 14 Bg5 Nf5 15 Qf4 h6 16 Bf6 g5 17 Qg4 Rg8=) 13 0-0 cxd3 14 Ng5 Qd7 15 a4 Nc6 16 Bc1 Na5 17 Rd1 h6 18 Qh3 Kf8 19 Rxd3 Kg7 20 Nf3 Ne7 with balanced play. (d) 9 Qf4 resembles 9 Nf3 and is very important.
White has the idea of dxc5, opening the position, after which the extra tempi spent on ... Qa5-a4 and ... g6 can look suspicious. Play can go: (d1) 9 ... Qxc2!? is playable, although dangerous after 10 Nf3! and then:
(d11) 10 ... c4!? 11 Rc1 Qf5 12 Qg3!. These lines look quite scary. White can play Be2 and either make threats on the kingside or play the surprisingly strong Bd1-c2. (d12) 10 ... Qe4+ isn’t attractive after 11 Qxe4 dxe4 12 Bb5+ Nc6 13 Ng5+/=. (d13) 10 ... h6! 11 dxc5 threatens 12 Nd4, so Black probably plays 11 ... Qe4+ 12 Qe3!? Ne7 13 Bd3 Qxe3+ 14 Bxe3 Nf5 15 Bd4 Bd7, when 16 g4 Nxd4 17 cxd4 Nc6 seems adequate, with the ideas of ... b6 and/or ... Na5 in mind. (d2) 9 ... c4!? 10 Nf3 (if White plays too slowly, ... Qxc2 will start to make sense) 10 ... h6!? (10 ... Nc6 11 Be2 h6 12 Ra2! seems to misplace the rook, but squelches all ideas of ... Qxc2; then according to Moskalenko, Black should play 12 ... Bd7 intending ... Rh7 and ... 0-0-0, but I prefer White) 11 Be2 Qxc2. Moskalenko gives this an ‘!’, but maybe he’s a little optimistic; for example, I like 12 Rc1! and: (d21) 12 ... Qa4 13 h4 Nc6 14 0-0 and most of Black’s pieces are a little stuck, so he might try a raid with 14 ... Na5 (14 ... Qxa3 15 Ra1 Qf8 16 Bc1 Qg7 17 Ba3) 15 Rb1 Nb3 16 Be3 Qxa3 17 Nd2 Nxd2 18 Bxd2 Qf8 19 Bc1 Qg7 20 Ba3 g5 21 Qf3 and I still think White is doing pretty well; watch out for Bxc4 and Rxb7 sacrifices. (d22) 12 ... Qe4 13 Qg3 g5 14 h4 g4 15 Ng1 h5 16 f3 gxf3 17 Qxf3! Qxf3 18 Bxf3 with more than enough for the pawn in terms of development, Black’s weak squares, and the potential to win the h-pawn. This is the sort of variation that makes me think that 8 ... Kf8 is more practical than 8 ... g6.
(d3) 9 ... Nd7 is a careful way to play it: 10 Nf3 (10 h4 h6 11 Nf3 b6!=) 10 ... h6 11 dxc5 Qxf4 12 Bxf4 Nxc5 gives White’s bishops scope, but because of the pawn structure Black may stand acceptably; e.g. 13 Be3 (13 c4 Ne7 14 Nd2 Bd7, about equal) 13 ... b6 14 a4 (14 c4 Ne7) 14 ... Ne7 15 a5 Bb7 16 Nd4 Kf8 17 axb6 axb6 18 Rxa8+ Bxa8 19 Bd3 with a modest advantage for White. 9 ... Nc6 One of the ideas behind this move is that Ng5 can be met by ... Nd8 in some lines, followed by driving the knight back. 10 Qf4
White intends dxc5. 10 ... c4!? Black should probably play 10 ... Qxc2 at this point if he wishes to show that monitoring the c-pawn is a benefit of ... Qa4. Then play can continue 11 dxc5! Qe4+ 12 Qxe4 (or 12 Be3!? Qxf4 13 Bxf4 h6, when 14 Nd4 Nge7 15 Nb5 Kf8 16 Nd6 g5 17 Bg3 Ng6 is about equal) 12 ... dxe4 13 Ng5 (13 Nd4 Nxe5 14 Nb5 Ke7 15 Bf4 f6 and White has only a slight edge at best) 13 ... f5 14 exf6 Nxf6 15 Bc4 Ne5! (15 ... Ke7 16 f3 Ne5 17 Be2 exf3 18 Nxf3+/=) 16 Bb5+ Bd7 17 Bf4 Nd3+ 18 Bxd3 exd3 19 Rd1 h6 20 Nf3 Ne4=. 11 Be2 After 11 Ra2, Moskalenko’s plan is 11 ... h6 (11 ... Qb5 12 Bc1 h6!? 13 Be2 simply asks for a4 later on) 12 Be2 Bd7 13 0-0 Rh7 14 Rb1 0-0-0. Black, by contrast with White, has no serious pawn weaknesses, but has to live with the general situation of his own dark-square weaknesses and bad bishop on d7. Following 15 Qh4 (15 h4!?, intending Nh2-g4, is the idea played in the main game) 15 ... Kb8 16 Bg5 Re8, White has no obvious way to make progress. Black has the same problem; a move such as 15 ... Nge7? is punished by 16 Ng5 (or 16 Bg5) 16 ... Nf5 17 Qh3 Rg7 18 Nf3 Rh8 19 Bg5 and Bf6. But 15 ... Rf8 is possible, looking to advance the f-pawn, perhaps in conjunction with ... Rff7 and ... g5. This all deserves investigation. 11 ... h6
12 h4 12 Ra2 resembles the previous note; and 12 0-0 can be safely played because 12 ... Qxc2?! (12 ... g5 13 Qg3 Qxc2 14 Bd1 Qg6 15 h4 Nge7 16 hxg5 hxg5 17 Nxg5 Na5 18 Ra2 Nb3 19 Bf4+/=; 12 ... Bd7 13 Ra2 g5 14 Qg4 0-0-0 15 Qh5 f6!?) 13 Bd1 gives White good attacking chances regardless of what Black does; e.g. 13 ... Qb2 (13 ... Qf5 14 Qg3 g5 15 Rc1 Qg6 16 Bc2 Qg7 17 h4+/=) 14 Ba4 Qb6 15 h4+/= Cabrera-Pomes Marcet, La Roda 2009. 12 ... Bd7 Not bad, although 12 ... Qxc2 is still consistent; e.g. 13 0-0 (13 Rc1 Qb3) 13 ... Qe4 (13 ... Nge7?! 14 Bd1 with the idea 14 ... Qf5? 15 Qh2 or 14 ... Qe4 15 Qg3+/=) 14 Qxe4 dxe4 15 Nh2 Na5 intending ... b6 and ... Bb7. 13 0-0 Rh7 14 Nh2 0-0-0 15 Ra2 Rf8 Or 15 ... g5!? 16 Qg3 f6. 16 Ng4 Qa5!?
Relieving the blockade is a double-edged decision; Black wants to cover the dark squares on the h4-d8 diagonal. 17 Re1 The same position arose in Lacasa Diaz-Moskalenko, Sitges 2011: 17 a4! (making way for Bc1-a3) 17 ... Qd8 18 Qg3 Nce7 (18 ... f5 19 exf6 Nxf6+/=) 19 Ne3 and here 19 ... f5 was best, but Black has few positive prospects. 17 ... Qd8 18 Nf6 18 Qg3 g5 19 Qh3 gxh4. 18 ... g5 19 hxg5 hxg5 20 Qxg5 Nxf6 21 Qxf6
Muzychuk-I.Rajlich, Ohrid 2009; and here I think 21 ... Qa5 is best (threatening ... Nxd4), when White has to be careful: 22 Rd1! Rfh8 23 f3 Rh5! 24 Qf4?! (keeping an eye on d4 for the following reason: 24 Qxf7 Rh1+ 25 Kf2 Rxd1 26 Bxd1 Qb6 intending ... Qb2; or if 27 Ra1, then 27 ... Nxd4! 28 Be3 Qb2 is strong; but 24 Kf2!= is better) 24 ... Qb6!? (or 24 ... Be8 25 Raa1 Ne7=/+) 25 Kf2 Nxe5 26 a4 Nc6 with the idea 27 Qd6 e5! 28 f4? Qd8! 29 Bxh5 Qh4+ and Black has too many threats. 11.3 8 h4 This has become a popular idea in the 6 ... Qa5 systems (for example, 9 h4 is a main line after 8 Qb1 c4). There are numerous transpositions here, including to the Positional Winawer systems in Chapter 10. To make it a little easier to navigate this mess, the general rule is that if White plays h4-h5 and secures a pawn on d4 by means of the move Rh4 (which means that Black is usually playing the move ... h6), it will be treated in this chapter. Instead White can sacrifice the d-pawn, and this very often means using the Kasparov Gambit from Chapter 10 (section 10.3), which I’ll be referring to below. To help you recognize that transposition without having to flip back-and-forth between chapters, here’s a diagram of the initial position of the Kasparov Gambit:
8 ... Nc6 8 ... Ne7 will often transpose; e.g. 9 h5 h6 10 Rh4 cxd4 11 Nf3 Nc6 is the main line below. A few variations which involve ... Ne7 without an early ... Nbc6 are considered in Chapter 10, section 10.3, in the note on 7 ... Qa5. 8 ... b6 is also playable:
(a) 9 h5 Ba6 10 Qg4 Kf8 11 Bxa6 Nxa6 12 Rc1!? h6 13 Ne2 Ne7 14 Rh3 Nc7 (14 ... Kg8! 15 Rf3 Rf8) 15 Rf3 Kg8 16 Qh4 (16 Qf4 Rf8 17 dxc5 Qxf4 18 Rxf4) 16 ... Qd7 17 g4, Iordachescu-Agdestein, Isle of Man 2003; 17 ... Rc8!=/+, with the idea 18 g5?! hxg5 19 Bxg5 Nf5 20 Qg4 Nb5 21 Ng3 Nh6. (b) 9 Qg4 Kf8 10 Rh3 h5!? 11 Qf4 Ba6 12 Nf3 (12 Rf3 Nh6) 12 ... Bxf1 13 Ng5 Nh6 14 Kxf1 (14 Nxe6+ Kg8 15 Nxg7 Qc4 16 Qf3 Kxg7 17 Qf6+ Kg8 18 Qg5+ Kf8 19 Qd8+=) 14 ... Ke7! 15 Rf3 Nf5 16 Kg1 g6=/+. 9 h5 9 Qg4 Kf8! is a position from 7 Qg4 Kf8 theory where h4 doesn’t fit in very well; Black should have easy equality. 9 ... h6 Black can also ignore h6 to develop quickly; e.g. 9 ... Nge7 and then: (a) 10 Nf3 cxd4 (10 ... h6 is the main line) 11 cxd4 Nxd4 12 Bd3 would be – surprise – the Kasparov Gambit (in section 10.3), and here 11 Rh4 b6 12 cxd4 h6 would be the note on 12 cxd4 below; but here White has the option of 12 h6 gxh6 13 Nxd4 Nxd4 14 Rxd4!? intending Qh5, which is yet another can of worms I’d simply stay away from as Black. (b) 10 h6 gxh6!? 11 Nf3 cxd4 12 cxd4 Nxd4 13 Bd3
This is very similar to the Kasparov pawn sacrifice, but with the moves h6 and ... gxh6 in, which matters a lot and may even help Black. The exact position in the diagram won’t arise naturally from normal orders in Chapter 10, however, without allowing other problems; e.g. 6 ... Ne7 7 h4 Nbc6 8 h5 Qa5 9 Bd2 cxd4 10 cxd4 Qa4 11 h6 gxh6 12 c3! (instead of 12 Nf3 Nxd4 13 Bd3, transposing) 12 ... Qxd1+ 13 Rxd1 causes positional damage to Black (perhaps reparable, as explained there in note ‘c1’ to White’s 11th move, but it’s a chore). Therefore, since this Portisch/Hook sequence is the ‘correct’ one, I’m going to examine it here: 13 ... Bd7 (13 ... b6!? 14 Bxh6 Ndf5) 14 Bb4 (14 Kf1?! Bb5!=/+) 14 ... Ndc6 15 Rh4?! (15 Rb1 Nxb4 16 axb4 Bb5=; 15 Bd6!? Qg4 16 g3 Nf5 17 Bxf5 Qxf5 18 Rxh6 – McDonald; and here 18 ... Rc8 19 Kf1 Na5 20 Nd4 Qg5 21 Rh4 Nc4 yields mutual chances) 15 ... Ng6! 16 Bxg6 hxg6 (with two extra pawns in return for some worry on the dark squares; Moskalenko thinks that Black is better and I agree) 17 Bd2! Qa6 (17 ... d4 isn’t bad, with the idea 18 Bg5 Nxe5! 19 Nxe5?? Qa5+) 18 Bg5 Rc8 19 Bf6, Volokitin-Zhang Pengxiang, Fügen 2006; and Moskalenko’s 19 ... Rf8 20 Rxh6 Ne7 with ideas of ... Ba4 and ... Rc4 would preserve the advantage.
10 Nf3 10 Qg4 Kf8 (10 ... Rh7!?, with ideas of ... Nxe5, ... cxd4, and ... Qc2, might run into 11 Qg3!? or simply 11 Nf3 Qxc2 12 dxc5+/=) 11 Nf3 (11 Rc1? Nxe5) 11 ... Nge7 (11 ... f5 12 Qg3 b6!?) 12 Qf4 Qxc2 (Moskalenko prefers 12 ... c4 in such positions, but sometimes greed is justifiable) 13 dxc5 Qe4+ 14 Qxe4 (14 Be3 Nf5) 14 ... dxe4 15 Nd4 Nxe5, and the strange position after 16 Bf4 f6 17 Bxe5 fxe5 18 Nb5 Nd5 is level. 10 ... Nge7 10 ... cxd4 11 Rh4 Nge7 transposes below, while 11 cxd4 Nxd4 12 Bd3 is the Kasparov Gambit from Chapter 10 again. 11 Rh4 cxd4 The most ambitious move, but not the only one: (a) With this move order, 11 ... Nf5 12 Rf4 c4 or 12 Rg4 b6! 13 Rf4! c4 is equal, but note that sometimes Black is already committed to ... cxd4 by the time he reaches this point. (b) 11 ... c4!? 12 Rg4 Nf5 13 Qc1 Bd7 14 Be2 0-0-0 15 Nh4 Nce7 16 Nxf5 Nxf5 is roughly equal; e.g. 17 Kf1 Rhg8 18 Rf4 g5 19 hxg6 fxg6 20 Bg4 h5! 21 Bxf5!? gxf5 22 Rh4 Be8= Schüppen-Valli, ICCF Champions League 2010. 12 Bd3!? 12 cxd4 b6 prepares to trade off the light-squared bishops.
Since White has no meaningful attack on the kingside, this usually leads to an ending; e.g. 13 Rg4 Nf5 (13 ... Kf8!? keeps the knight on e7, so 14 c4 Qxd1+ 15 Rxd1 dxc4 16 Bxc4 Bb7 is typical: Black’s pawn structure, long diagonal, and d5 outpost compensate for White’s space and the bishop pair) 14 c4 (14 c3 Qxd1+ 15 Rxd1 Na5!=) 14 ... Qxd1+ 15 Rxd1 dxc4 16 Bxc4 Nce7
17 Bb4 Nd5 18 Bxd5!? exd5 19 Rf4 Be6 20 g4 Ne7 21 Rc1 Rc8 with equality. 12 ... Bd7 13 Kf1 0-0-0 14 Rf4 g5! Black has saved a tempo over the lines in which the f-pawn has to be protected before castling queenside is possible.
15 Rxf7!? White invests an exchange. Following 15 Rg4, 15 ... Ng8! intends ... f5 (or ... f6); e.g. 16 cxd4? (16 Rb1 f5!? 17 Bb5! Qxa3 18 Bxc6 Bxc6 19 Rxd4 is difficult to assess) 16 ... f5 17 Rg3 f4 18 Rg4 Be8. 15 ... Be8 16 Rxe7 Nxe7 17 Nxd4 White controls the key squares g6, f5 and b5 and attacks e6. But Black is safe on the queenside and has his usual compact pawn structure. 17 ... Bd7 18 Rb1 Kb8 19 Nb3?! 19 Qc1 intending Qb2-b4 is more accurate, although Black has a good position. 19 ... Bb5 19 ... b6! is strong, since 20 Nc5 Qc6 21 Na6+ Ka8 achieves little. 20 Nc5 Bxd3+ 21 cxd3 Qxd1+ 22 Rxd1 Nc6=/+ R.Robson-Shulman, Lubbock 2011.
Chapter
Twelve
Exchange
Variation
1 e4 e6 2 d4 d5 3 exd5 exd5
This is the Exchange Variation, which tends to produce a simpler game, or at least a less tactical one, than most other variations in the French Defence. Some players dismiss it as ‘drawish’; with only one open file, they reason, both sides will mindlessly exchange their heavy pieces and a sterile position will ensue. But there is a hole in this reasoning: in most cases where one side is fighting for a win, at least one rook doesn’t belong on the e-file; and in numerous lines, neither side’s rooks should be there. Why? Because there will be no points of penetration along the file; for example, squares like e2, e3 and e4 for White and e7, e6, and e5 for Black will customarily be covered two or three times by that side’s pawns, knights, and bishops. Moreover, only one such defender is really necessary. So if, for example, White doubles or triples along the e-file while Black is blithely mounting a kingside pawn storm (backed by his rooks and queen), the second player has every chance of winning. Furthermore, Black can always make the struggle unbalanced should he so choose. This has been pointed out and demonstrated by great players for at least 70 years. I might add that it’s rather silly to call any position in which every piece and all but two pawns remain on the board ‘drawish’. In fact, even very strong players hoping for a draw with White have had little success in trying to force one. At any rate, White’s main moves are: 12.1 12.2 12.3 12.4
4 4 4 4
Nf3 Bd3 c4 Bf4
White has other choices, but Black can always achieve an imbalance if he wants: (a) 4 Nc3 Bb4 is dealt with in Chapter 7, under the order 3 Nc3 Bb4 4 exd5 exd5 (see section 7.6). Here 4 ... Nf6 is also equal; e.g. 5 Bg5 Be7 6 Bd3 0-0 7 Nge2 h6 8 Bh4 Nb4. (b) 4 c3 is slow: 4 ... Bd6 (or 4 ... Nc6; then 5 Bb5 might look appealing, but c3 doesn’t go well with this move – compare the reversed positions in Chapter 7), and 5 Bd3 (5 Nf3 Nf6 6 Bd3 Qe7+ transposes, and here 6 Be2 is slow) 5 ... Nf6 (5 ... Nc6 transposes to 4 Bd3 Nc6 5 c3 Bd6 in section 12.2) 6 Nf3 (6 Ne2 0-0 7 0-0 Re8) 6 ... Qe7+! 7 Be3 (7 Qe2 Qxe2+=/+) 7 ... Ng4 wins the bishop pair. (c) 4 Be3 is passive. Then 4 ... Nf6 is normal, and 4 ... Bd6 intending ... Ne7 and perhaps later ... Nf5 is also fine. (d) 4 Qf3 is rather eccentric, yet keeps an eye on d5 and has in mind castling queenside. Black seems to get equality with almost any natural move, including 4 ... c6 and 4 ... Bd6. Alternatively: (d1) A nice example was 4 ... Nc6 5 Bb5 Nf6 6 Ne2 Bg4 7 Qc3 Bd6! 8 Bxc6+ bxc6 9 Qxc6+ Bd7 with more than enough compensation: 10 Qa6 Ng4!? (10 ... 0-0!=/+) 11 h3? Qf6! 12 Nbc3? (to stop ... Bb4+; White should try 12 Qd3 Qxf2+ 13 Kd2! Kf8! 14 hxg4 Bxg4-/+) 12 ... Qxf2+ 13 Kd1 Qxg2–+ Bykhovsky-Ulibin, Stockholm 1997. (d2) 4 ... Nf6 5 h3 (versus ... Bg4, which is often followed by ... Bh5-g6) 5 ... c5 6 Bb5+!? Nc6 7 Ne2 Qb6! 8 Qd3!? c4 9 Qe3+ Be6 10 Ba4 Bd6=/+ with a nice lead in development, M.Hermann-Vaganian, German League 1992.
12.1 4 Nf3
White develops normally and is thinking about 5 c4. Kasparov played this move and it is now more popular than any other. I’ll recommended one of two moves, both breaking the symmetry:
12.11 12.12
4 4
... ...
Nc6 Bd6
Instead: (a) In previous editions I have recommended 4 ... Bg4, but I confess to changing my mind about its practical value. In the line made famous by Kasparov, 5 h3 Bh5 6 Qe2+ Qe7 (lest 7 Qb5+) 7 Be3, Black can reach effective equality in the main line 7 ... Nc6 8 Nc3 0-0-0 9 0-0-0, but only with very accurate play, and he still may be faced with the more difficult over-the-board decisions thereafter. So I’m using two lines with excellent reputations, the first a fashionable unbalancing variation, and the second an older solution with a few optional twists that have been discovered in the past several years. (b) Of course, 4 ... Nf6 is fully playable, and I should point out that 5 c4 (a key White plan) allows 5 ... Bb4+ 6 Nc3, or here 5 ... Be7 6 Nc3, both transposing to 4 c4 lines (see 12.312 and 12.32 respectively). 12.11
4
...
Nc6
I didn’t use this move in previous editions, both because there are good alternatives and because White seems to have an ideal response. Nevertheless, this has become a favourite of grandmasters who are seeking to retain an imbalance and thus greater winning chances. Over the last decade, it’s become clear that Black’s chances are every bit as good as White’s, so I’m going to give it the most thorough analysis in this chapter. 5 Bb5 The most important move. Interestingly, we have now reached a Winawer Exchange Variation 3 Nc3 Bb4 4 exd5 exd5 5 Bd3 Nf6 (see 7.611; note that Black has other options here), except that in our line, White is a tempo ahead! However, it turns out that this variation (like so many symmetrical ones) has the paradoxical property that, because White has had to commit his pieces first (Nf3, in this case), Black can adjust his play accordingly. In other words, his extra information about what White has done is enough compensation for the tempo that he is able to maintain equality. Clearly, it doesn’t hurt that the variation under consideration is generally slow and non-tactical (except for our main line, as you’ll see). Less aggressive moves than 5 Bb5 don’t achieve much; for instance, 5 c3 Bd6 6 Bd3 and 5 Bd3 Bg4 6 c3 Bd6 both transpose to 4 Bd3 lines (see 12.2). But 5 c4 can be independent following 5 ... Bg4 (5 ... Bb4+ 6 Nc3 Bg4 transposes; and of course 6 ... Nge7 and 6 ... Nf6 are also playable and will probably transpose to 4 c4 Bb4+ lines in 12.31) 6 Nc3 Bb4!? (or 6 ... Nf6 7 Be2 dxc4 8 0-0 and 8 ... Be7 or 8 ... Bd6!? 9 Bxc4 0-0) 7 a3!? (7 cxd5 Qxd5 8 Be2 transposes to the Danish/ Göring Gambit! – that is, 1 e4 e5 2 d4 exd4 3 c3 d5 4 exd5 Qxd5 5 cxd4 Nc6 6 Nf3 Bg4 7 Be2 Bb4+ 8 Nc3; then a standard recipe is Capablanca’s 8 ... Bxf3 9 Bxf3 Qc4, considered equal) 7 ... Bxf3 8 gxf3 (8 Qxf3 Nxd4 9 Qe3+ Qe7) 8 ... Bxc3+ 9 bxc3 Nge7
10 Qc2!? (10 Bd3 0-0 11 0-0 dxc4 12 Bxc4 Nd5 13 Qc2 Qd6=) 10 ... 0-0 11 Bd3 dxc4 12 Bxh7+? Kh8 13 Be4 f5! 14 Bxc6 Nxc6 15 Bf4? (15 0-0 Qh4 with a nice advantage) 15 ... Nxd4!? (strong, but 15 ... Qh4! leaves no viable defence; e.g. 16 Be5 Nxe5 17 dxe5 Rae8 18 0-0-0 Rxe5 and White has a pawn less, with five isolated ones remaining) 16 cxd4 Qxd4 17 Qc1 Qd3 18 Be5 f4! 19 Ra2?, Hayrapetian-D.Petrosian, Kajaran 2011; and Black should have put the game to rest by 19 ... Rfe8 20 Re2 Rad8 21 Rg1 Re7 threatening ... Red7, ... c3-c2 and other winning moves. 5 ... Bd6 6 c4 This is the most forcing and tactical move; notice that Black lacks time for it in the reversed, Winawer position. White attacks d5, and also has c5 in mind. From the time 4 ... Nc6 5 Bb5 Bd6 became popular, 6 c4 has been considered the only way to cast the viability of Black’s position into doubt. Nevertheless, White can try slower moves and hope for an edge. Thus far they haven’t had much luck: (a) 6 0-0 Ne7 and:
(a1) There’s still time for 7 c4, but 7 ... dxc4 should not be answered by 8 d5?! (8 Bxc4) 8 ... Nxd5! having in mind 9 Qxd5?? Bxh2+. (a2) 7 Bg5 f6 is typical: (a21) 8 Bh4 isn’t as bad as it might seem, even though it tempts Black’s pawns forward; for example, 8 ... 0-0 (of course, Black can also reserve the idea of ... 0-0-0; e.g. 8 ... Bg4 9 Bg3 Bxg3 10 hxg3 Qd6; or 8 ... Bf5!?) 9 c4 dxc4 10 Bxc4+ Kh8 11 Nc3 Nf5 creates threats on d4 which are hard to meet, and 12 d5 (12 Bd3! is a better try, when Black can improve his position with 12 ... Re8) 12 ... Ne5 13 Nxe5 Bxe5 14 Bg3 Nxg3 15 hxg3 Qd6 or 15 ... a6 leaves White nothing that matches Black’s bishop pair. (a22) 8 Be3 Bg4 (8 ... a6 9 Bd3 Bf5 is an option, probably equal) 9 c3 Qd7 10 Nbd2 (10 c4 dxc4 11 d5? Nxd5) 10 ... 0-0-0!? 11 a4!? Nf5 (going after the bishop; other normal moves are 11 ... Rde8, freeing d8 in some lines for a knight retreat, and 11 ... g5, initiating a pawn storm) 12 Bd3 (12 h3 Bh5!) 12 ... Rde8 (12 ... Nxe3 13 fxe3 Rhe8 14 Re1 Re7 will result in an edge even after White plays e4, but it’s not much) 13 Re1
13 ... g5! (enterprising; of course capturing on e3 and doubling rooks isn’t bad either) 14 Qc2 Nce7 15 b4 h5?! (15 ... Bh5! threatens ... g4, winning a piece, when 16 h3 exposes the kingside to direct attack by 16 ... Bxf3! 17 Nxf3 h5!) 16 Bb5!? (setting up a target, but it’s slow; 16 b5 looks better) 16 ... c6 17 Bd3 Kb8 18 b5, Fataliyeva-Sengupta, Mumbai 2011; and a good way to increase the pressure is 18 ... Qc7 (18 ... Rc8=/+ was played) 19 bxc6 Nxe3! 20 Rxe3 Nxc6 21 Rxe8+ Rxe8. (a3) 7 Re1 0-0 8 h3 (versus ... Bg4) 8 ... Bf5 9 c3 (solid, but in conjunction with 8 h3, this is rather slow) 9 ... Qd7 (9 ... a6 10 Ba4 Qd7) 10 Ne5?! (10 Nbd2 a6 11 Ba4 h6 12 Bc2, about equal) 10 ... Bxe5! 11 Rxe5 a6 12 Bxc6 Nxc6 13 Re1 Rae8 14 Rxe8 Rxe8 (White’s pieces are on their initial squares) 15 Bf4, Safarli-Mamedyarov, Baku 2011; and here I like I like 15 ... Nd8!
followed by ... Ne6 ( ... Qb5 becomes an idea as well). Black would have a small but irritating pull. (b) 6 h3 stops ... Bg4, but is again slow;
for example, 6 ... Ne7 7 0-0 0-0 8 c3 a6?! (driving the bishop where it wants to go; 8 ... Re8 would secure a slight advantage) 9 Bd3 Bf5 10 Re1 Qd7 11 Qc2 Rae8 12 Nbd2 Bxd3 13 Qxd3, Wang Chen-Ganguly, Qinhuangdao 2011; and instead of 13 ... Qf5 with a miniscule advantage, 13 ... Ng6! keeps a little more play on. (c) 6 c3 Ne7 7 0-0 0-0 8 Bg5!? f6 (as so often, playing this move with gain of tempo makes up for the generally inaccessible weakness on e6; but 8 ... h6 is also fine, since 9 Bxc6? hxg5 10 Ba4 Ng6 takes the initiative on the kingside) 9 Bh4 Nf5 10 Bg3 Nxg3 11 hxg3 Ne7!? (11 ... Re8 is also logical; White doesn’t want to exchange on c6 and lose all the light squares) 12 Nbd2 (12 Bd3! Bg4=) 12 ... Be6 (12 ... Bf5!) 13 Bd3 c6 14 Re1 Qd7 15 a4 (neither side has a dramatic plan) 15 ... a5 16 Qc2 g6 17 Re2 Bf7 18 Rae1 Rae8= Dzhakaev-Psakhis, St Petersburg 2002. 6 ... dxc4 Not 6 ... Ne7?? 7 c5!. 7 d5 a6
8 Ba4 The best move. Otherwise: (a) 8 Bxc4 Qe7+! 9 Be3 Ne5 give Black superior activity. (b) 8 Qa4? fails to 8 ... axb5! 9 Qxa8 Nb4, with both 10 ... Nc2+ and 10 ... Nd3+ threatened. White’s best is 10 0-0! Nc2 (even 10 ... Ne7 11 Nc3 0-0! 12 Nxb5 Nexd5 gives more than enough compensation) 11 Nc3 Nxa1 12 Nxb5 Ne7! 13 Qa4 0-0 14
Qxc4 Nxd5! 15 Bg5! Qd7 16 Rxa1 Re8 17 Nxd6 Qxd6 and although it isn’t much, Black’s pieces are better placed. (c) 8 Bxc6+ bxc6 9 dxc6 Ne7 10 Nbd2 Be6 (10 ... c3 11 bxc3 0-0 12 0-0 Nxc6 13 Ne4=) 11 0-0 0-0 12 Ng5 (12 Qc2 Bf5! intending 13 Qxc4? Bd3!) 12 ... Bd5 13 Qc2 Ng6 14 Nxc4 h6 15 Nxd6 Qxd6 16 Ne4 Qxc6 17 Qxc6 Bxc6 18 Nc5 Rfe8=/+ with the idea 19 Be3? Nh4-/+. 8 ... b5 9 dxc6 bxa4 10 0-0 10 Qxa4 allows 10 ... Bg4 11 Nbd2 Qe7+ and the king has to move; for instance, 12 Kf1 Nf6 13 Nxc4 0-0 14 Bg5 Qe4! 15 Bxf6 Bxf3 16 gxf3 Qxf3 17 Rg1 Qxf6, Ozgibcev-Volkov, Sochi 2004; Black has recovered the pawn and can target White’s weaknesses. 10 ... Ne7 11 Qxa4
11 Nbd2 0-0 (11 ... Be6! is more accurate, I think, and forces White to move his queen if he is to recover the pawn; e.g. 12 Qxa4 c3!?= or 12 Ng5 Bd5! 13 Qxa4 c3 14 bxc3 0-0=/+) 12 Nxc4 Nxc6 13 Qxa4 Bb7 14 Bg5 (14 Nxd6 cxd6!? 15 Bg5!? Qe8= L.Webb-N.Pert, British League 2011) 14 ... Qb8?! (14 ... Qe8!?; 14 ... Ne7 targets f3 and equalizes) 15 Rac1 Rc8?! 16 Nxd6 cxd6 17 Qg4+/= A.Gonzalez Perez-Moskalenko, Sitges 2010. 11 ... Rb8 With a dynamic counterattack; though it’s not the only reasonable move: (a) 11 ... Be6 12 Nbd2 Bd5 (12 ... c3! 13 bxc3 0-0=) 13 Nxc4 Bxc6 14 Qd1+/= 0-0 15 Nxd6 cxd6 16 Nd4 Qd7 (16 ... Bb7) 17 b3 Bb7 18 Bb2 Ng6 ½-½ J.Iglesias-Moskalenko, Spain 1999. (b) 11 ... 0-0 is similar: 12 Nbd2 c3!? (12 ... Rb8! is clearest, with the idea 13 Nxc4 Rb4 14 Qc2 Bf5; upon 13 a3, 13 ... Rb5 14 Nxc4= transposes to 11 ... Rb8, and here 13 ... Be6 14 Nxc4 Bd5! also yields active play) 13 bxc3 Be6 14 Nd4 Bd5 15 Nc4 Nxc6 16 Nxc6 (16 Nf5!?) 16 ... Bxh2+ 17 Kxh2 Qh4+ 18 Kg1 Bxg2! (the old Lasker-Bauer sacrifice! – but only to draw) 19 Kxg2 Qg4+ 20 Kh2 Qh4+ 21 Kg2 Qg4+ ½-½ Hayrapetian-Ara.Minasian, Kajaran 2011.
12 a3 Preventing ... Rb4 at the cost of time. White hasn’t found a way to advantage here and, while the game is equal, he has to watch out for direct attack on Black’s part: (a) 12 Nbd2 gets the pawn back with more activity, but loses a number of tempi: 12 ... Rb4 (12 ... 0-0 13 Nxc4 Rb4 14 Qc2 Bf5 15 Qc3 Nd5 16 Qd4 transposes) 13 Qc2 Bf5! 14 Qc3 0-0 15 Nxc4 Nd5 16 Qd4 Nb6 17 b3 Nxc4 18 bxc4 Be6 19 Bb2 f6. After all that, Black has definite compensation for the pawn. Vajda-Glud, Obro 2006, continued 20 Rfc1 Ra4 21 a3 (21 Qd1! Rxc4 22 Rxc4 Bxc4 23 Qa4 Bf7! 24 Qxa6 Qa8=) 21 ... Qe8 (21 ... Qe7) 22 Qe4? Bxc4 with material equality and two strong bishops. (b) 12 Qxc4 is less purposeful than Nbd2/Nxc4: 12 ... 0-0 13 Nc3 Rb4 14 Qe2 Nxc6 15 Bg5? (15 Be3 Qf6 is mildly better for Black) 15 ... f6 (15 ... Nd4! produces a meaningful advantage following 16 Bxd8! Nxe2+ 17 Nxe2 Rxb2!? 18 Bxc7 Bxc7 with two powerful bishops; here 16 Nxd4 Qxg5 makes White’s defence difficult) 16 a3 Nd4 (16 ... Rg4!? 17 Be3 Qd7! 18 g3 Bb7=/+) 17 Nxd4 Rxd4 18 Be3, Djingarova-Bhat, Andorra 2006; 18 ... Re8 19 Rad1 Rxd1 20 Rxd1 is about equal. 12 ... 0-0 13 Nbd2 Rb5!? An interesting rook lift with tactics in mind, although the positionally-based 13 ... c3 14 bxc3 Be6 was perfectly fine. 14 Nxc4 Nxc6 15 Bg5 Qd7! 16 Rad1 Qg4!?
Preparing a trick. 16 ... Bb7 is the next best alternative, when things could peter out in a line such as 17 Nxd6 cxd6 18 Bf4 d5 19 Rd2 d4! 20 Nxd4 Nxd4 21 Rxd4 Qc6 22 f3 Rxb2 23 Qxc6 Bxc6=. 17 h3 ‘Falling for it’, but this is probably best.
17 ... Qxf3! 18 gxf3 Rxg5+ 19 Kh1 Bxh3 20 Nxd6? A blunder under pressure. White could simplify and keep things level by 20 Rxd6 cxd6 21 Rd1 Bg2+ 22 Kh2 Bxf3 23 Rxd6 Ne7! 24 Ne3 Ng6=; McDonald gives 25 Rxa6 Bb7 26 Ra7 Ne5 (26 ... Rh5+ 27 Kg3 Rg5+=) 27 Rxb7 Nf3+ 28 Kh3 Ng1+ 29 Kh2=. 20 ... Bg2+ 21 Kg1 It seems that brutal materialism is going to triumph after 21 ... cxd6 22 Qf4 f6 (or 22 ... Rg6 23 Rxd6) 23 Qxg5! fxg5 24 Kxg2 and that all Black’s ingenuity has only led to a lost endgame. But instead comes a shocker: 21 ... Ne5!! The point; f3 is indefensible by normal means. 21 ... cxd6? can be answered by 22 Qf4! f6 23 Qxg5 fxg5 24 Kxg2. 22 Rd3 Bh3+ 23 Kh1 Bxf1 24 f4? But White will end up far behind in material after 24 Rc3 cxd6. 24 ... Rh5+ 25 Kg1 Bxd3 26 Qd1 Bg6 It’s all over now. Black will end up with a decisive material advantage and an ongoing attack. 27 fxe5 cxd6 And Black went on to win, Slavina-I.Rajlich, Budapest 2005. 12.12
4
...
Bd6
5 c4 Instead: (a) 5 Nc3 is rare, because 5 ... c6 restricts White’s knight (probably he would prefer to set up with Bd3 and Nge2 in this case), and 5 ... Ne7 is also fine, because 6 Nb5 and capturing on d6 gives Black a nice lead in development in a position where knights are about as good as bishops. b) 5 Bd3 Ne7 (5 ... Nc6 might transpose to 4 Bd3 lines; i.e. if White plays c3) 6 0-0 Bg4 (or again 6 ... Nbc6) 7 Bg5 (7 h3 Bh5 8 c4 dxc4 9 Bxc4 0-0 10 g4 Bg6 11 Ne5 c5!? 12 Nxg6 Nxg6 13 dxc5 Bxc5 14 Qxd8 Rxd8 15 Bg5 Rd7= Thesing-Knaak, German League 1992; here 11 ... Nbc6 would have been more confrontational; instead, 7 Re1 Nbc6 8 c3 transposes to the main line of 4 Bd3 in 12.2) 7 ... Nbc6 8 c3 Qd7 9 Nbd2 f6 10 Bh4 h5 11 b4 Nd8 (or 11 ... 0-0-0), Mokcsay-L.Portisch, Zalakaros 1991; Black’s coming kingside pawn storm in this type of position will be seen again under 4 Bd3.
After
5
c4,
we
12.121 12.122 But
look
5 5 not
5
at
two
... ... ...
Ne7??
moves: c6 Nf6
6
c5!.
12.121 5 ... c6 This is a solid if relatively dull line. 5 ... c6 has the advantage of reserving a later ... Ne7, when in the event of cxd5, Black’s knight on e7 will arguably be better placed than White’s on f3. 6 Nc3 6 cxd5 cxd5 7 Nc3 (7 Bd3 Ne7 8 Bg5 h6 9 Bh4 Qb6!= Holubcik-Vavrak, Tatranske Zruby 2000) 7 ... Ne7 transposes below (see 7 cxd5 cxd5 in the main line), and here 7 Bg5 f6! only helps Black. 6 ... Ne7 6 ... Nf6 isn’t bad, and often arises from 5 ... Nf6 6 Nc3 c6. In Kramnik-J.Polgar, Novgorod 1996, Black played a fascinating gambit: 7 Bg5 (trying to get Black to play ... dxc4 without losing a tempo by Bd3 or Be2; 7 c5 can be answered by the aggressive 7 ... Bc7, maintaining an eye on the kingside, which is probably best; or by the patient 7 ... Be7, after which Black will almost always play 8 ... b6; e.g. 8 Bd3 b6 9 cxb6 axb6 10 0-0 0-0 11 Bf4 Ba6 12 Ne5 Qc8 13 Re1 Bxd3 14 Qxd3 Re8 15 Rac1 Nbd7=) 7 ... 0-0! 8 cxd5 h6!
Now White faces a choice: (a) Kramnik suggests 9 Bxf6 Qxf6 10 Be2 Rd8 11 dxc6 (11 0-0 cxd5= 12 Nxd5?! Bxh2+ 13 Kxh2 Rxd5 14 Rc1 Nc6=/+ with the ideas ... Bg4 and ... Qd6+) 11 ... Nxc6 12 0-0. But then Black has 12 ... Nxd4! 13 Nxd4 (13 Qxd4 Bxh2+) 13 ... Be5 with an edge; e.g. 14 Qb3 Rxd4 (or 14 ... Bxd4 15 Bc4 Be6 16 Bxe6 Qxe6 17 Qxe6 fxe6 18 Rfe1 Bxc3 19 bxc3 Kf7) 15 Bc4 Bd7 16 Rad1 Bc6 17 Nd5 Qd6 18 Rxd4! Bxd4 19 Nb4 Qf6 20 Nxc6 bxc6=/+, which will undoubtedly be drawn with good play, but Black has what chances there are. (b) The game went 9 Bh4 Qb6!? (9 ... Re8+ 10 Be2 g5 11 Bg3 cxd5 12 0-0 Nc6) 10 Qd2?! (10 Bxf6 Qxb2 11 Rc1 Bb4 12 Qd2 Qxd2+ 13 Kxd2 gxf6 14 Bc4 Rd8 is still in Black’s favour) 10 ... Nxd5 (10 ... Re8+!? 11 Be2 Ne4! 12 Nxe4 Rxe4 13 0-0 Bg4 threatens the bishop on h4, and 14 Bg3 Bxg3 15 hxg3 cxd5 is somewhat to Black’s advantage as well) 11 Bg3 Bxg3 12 hxg3 Bg4 13 Bd3 Re8+ 14 Kf1 Nd7 15 Nxd5 cxd5 16 Rh4, and here 16 ... h5! maintains some pressure (16 ... Be6 ‘!’ 17 Re1 Nf8 with equality – Kramnik); e.g. 17 Nh2 (17 a4 g6) 17 ... Nf6 18 Nxg4 Nxg4 19 Rxh5 g6 20 Rh4 Qxd4=/+.
7 Bd3 Others: (a) 7 Be2 might transpose after 7 ... dxc4 8 Bxc4, though Black might also simply develop by ... 0-0 and ... Bf5, in view of the passivity of White’s bishop. (b) 7 cxd5 cxd5 yields a static central pawn structure where knights are at least the equal of bishops. Thus White is illmotivated to waste two tempi to win the d6-bishop by 8 Nb5?! 0-0 (8 ... Nbc6) 9 Nxd6 Qxd6 10 Be2 Bg4 (or 10 ... Nbc6 11 0-0 Re8=) 11 0-0 Nbc6 (aiming for ... Nf5) 12 h3 Bh5 13 b3!? (13 Be3 Nf5 14 Qd2 Rae8=/+; 13 g4 Bg6 14 Nh4 Rae8=/+ intending ...
f5) 13 ... Rfe8 14 Bb2? (14 Re1) 14 ... Ng6! (a knight on f4 spells trouble, and ... Rxe2 and/or ... Nh4 are potential threats) 15 g4 (15 g3 Qe6!; 15 Bc1 Rxe2!) 15 ... Nf4 16 Ne5 Bg6 17 Bf3 Nxh3+ 18 Kh1 Be4–+ M.Johansson-Renman, Swedish Ch. 1980. (c) 7 c5 isn’t bad, but lacks impact because it releases the pressure on the centre;
e.g. 7 ... Bc7 8 Bd3 (8 Bg5 f6 9 Be3 0-0 10 Bd3 Bf5=) 8 ... 0-0 (8 ... Bf5 is the more common move order, to force the kind of position that arises next) 9 0-0 h6 (9 ... Bf5 is normal, but this isn’t bad) 10 h3 (10 Qc2 Ng6!= 11 Bxg6 fxg6 12 Qxg6 Bf5 13 Qh5 Nd7 with obvious compensation) 10 ... Bf5 11 Re1 Bxd3 12 Qxd3 Nd7 13 Nh4!? (versus ... Ng6) 13 ... Re8 14 Bd2 Nf8! 15 b4 Qd7 16 a4 Nfg6 (16 ... Ne6) 17 Nxg6 Nxg6= George-Hutchings, England 1986. 7 ... dxc4 8 Bxc4 0-0 9 0-0 Nd7
After Black gets ... Nb6 in, the game can fairly be assessed as equal. 10 Re1 Naturally there are alternatives to this routine move: (a) 10 Bg5 Nb6 11 Bb3 Kh8 (11 ... Bg4!? has a similar idea; e.g. 12 h3 Bh5 13 g4 Bg6 14 Ne5 Kh8! intending ... f6) 12 Bxe7 (12 Qd2 f6 13 Bf4 Bxf4 14 Qxf4 Nbd5=) 12 ... Qxe7 13 Re1 Qc7 14 Qd3 f6 15 d5 (15 Bc2 g6; 15 Ne4 Bf5 16 Qc3 Rad8) 15 ... Bg4 16 h3 Bxf3 17 Qxf3 cxd5 18 Nxd5 Nxd5 19 Bxd5 Rae8=. (b) 10 Ne4!? Nb6 11 Bb3 Bf5 (11 ... Bg4 12 h3 Bh5) 12 Ng3 (White gets nothing out of 12 Nxd6 Qxd6 13 Re1 Rae8!? 14 Bg5 Ned5) 12 ... Bg6 13 Bg5 Qc7= Roitzsch-A.Graf, Dresden (rapid) 2010. (c) 10 h3 (to prevent ... Bg4 after ... Nb6) 10 ... Nb6 11 Bb3 Ned5 12 Re1 Re8 13 Bg5 Rxe1+ 14 Qxe1 Qf8=, aiming for ... Be6 or ... Bf5 and ... Re8.
10 ... Nb6 11 Bb3 Nbd5 11 ... Ned5 12 Bg5 f6 13 Bh4 Re8 14 Bg3 Bxg3 15 hxg3 Be6 16 Qd2 ½-½ Neubauer-Borges Mateos, Santa Clara 2002. 12 h3 Bf5 12 ... Qb6!? 13 Bc2 f6 keeps things complicated. 13 Ne4 Bxe4 14 Rxe4 ½-½ Leko-Varga, Budapest 1993. 12.1225 A currently popular approach.
...
Nf6
6 Nc3 Others: (a) 6 c5 Be7 7 Bd3 b6! 8 cxb6 (8 b4? a5 9 cxb6 axb4 10 bxc7 Qxc7=/+) 8 ... axb6 9 0-0 0-0 10 Bg5 (10 Nc3 c5=) 10 ... c5!?, Zatonskih-Savchenko, Helsingor 2011; and White should play 11 Nc3 (instead of 11 b3? h6 12 Bh4 g5 13 Bg3 Ne4!=/+) 11 ... Nc6!? (11 ... Bb7=) 12 dxc5 (12 Bb5 Bb7 13 dxc5?! bxc5, having in mind 14 Bxf6 Bxf6 15 Qxd5 Qb6!) 12 ... bxc5 13 Bxf6 Bxf6 14 Nxd5 Bxb2 15 Rb1 Bd4, when Black stands better due to the d4 outpost. (b) 6 Bg5 0-0! 7 cxd5 (7 c5?! Be7 followed by ... b6 already favours Black) 7 ... h6 8 Bxf6 Qxf6 is at least equal; compare what follows, as well as the Kramnik-Polgar game in the previous section (with 5 ... c6 6 Nc3 Nf6 7 Bg5 0-0!). 6 ... 0-0! 6 ... c6 transposes to 5 ... c6 6 Nc3 Nf6 (in the notes to 12.121); and 6 ... dxc4 7 Bxc4 0-0 8 0-0, while not conventional (White’s bishop came out in one jump), is still perfectly playable because of the aggressive bishop on d6 (compare 4 Nf3 dxc4 5 Bxc4 Bd6, a fairly common line among strong players); e.g. 8 ... Nc6 (planning ... Bg4) 9 h3 h6!? (9 ... Bf5)
10 a3 (10 Qc2!? Nb4!? 11 Qb1 c6 12 a3 Nbd5 intending ... Be6; several grandmasters have employed the move 10 ... Na5 as well; or 10 Re1 Re8 11 Rxe8+ Qxe8 12 Nb5 Qe4!? 13 Qb3 Na5 14 Bxf7+ Kf8 15 Qc3 Kxf7 16 Nxd6+?! cxd6 17 Qxa5 Bxh3! 18 Qc7+ Kg8 19 Qxd6 Bg4=/+ Moskalenko-Navara, Catalonian Team Ch. 2011) 10 ... Bf5= 11 b4 a6 12 Bb2 Re8 13 Ne2?! Be4!? (13 ... Qd7) 14 Ng3 (14 Nc3! Bf5) 14 ... Bd5 15 Bxd5 Nxd5 16 Nf5, Andrijevic-Ulibin, Paracin 2011; and Black stands better in any case due to his powerful outpost and White’s horrendous bishop, but 16 ... Qd7! 17 Nxd6 cxd6! would keep White’s knight out of e5. 7 cxd5 Not 7 Nxd5 Nxd5 8 cxd5 Bb4+ 9 Bd2 Bxd2+ 10 Qxd2 Qxd5; in general, White doesn’t want to exchange pieces, especially his c1-bishop, in an isolated pawn position. As usual, 7 c5 Be7 releases the pressure on d5, and White still doesn’t have a convenient way of preparing an answer to ... b6 that doesn’t compel the reply cxb6 and thus allow easy equality; e.g. 8 Be2 b6 9 cxb6 axb6 10 0-0 c5. 7 ... Nbd7 7 ... Re8+ is not unplayable, but it’s nice to get moves like ... Nbd7-b6 and/or ... h6 in quickly so as to recover the pawn. 8 Be2 Alternatively: (a) 8 Bg5 h6 9 Bh4 Nb6 (or 9 ... Re8+) 10 Be2 Be7 11 0-0 Nfxd5=. (b) 8 Bc4 Re8+ 9 Be3 Ng4!? (9 ... Nb6 10 Bb3 Ng4) 10 0-0 Nxe3 (10 ... Rxe3!?) 11 fxe3 Rxe3 12 Qd2 Re8 is unclear; naturally, both sides have options.
8 ... h6 Having omitted ... Re8, Black gets to the d-pawn in time and can afford this tempo to prevent Bg5. Nevertheless, allowing Bg5 works, too: 8 ... Nb6 9 Bg5 (9 Qb3?! Nbxd5 10 Nxd5 Nxd5) 9 ... h6 10 Bxf6 (10 Bh4 transposes to the previous note, or Black can try 10 ... g5 11 Bg3 Nbxd5 12 0-0 Be6=) 10 ... Qxf6 11 0-0. White stays a pawn ahead, but it’s an isolated doubled d-pawn and Black has the bishops, so probably this is about equal. There might follow 11 ... Bf5 12 Re1 Rae8 13 Qd2 Bg6!? (13 ... Bg4) 14 Bd3 Bf4 (or 14 ... Bh5 with the idea 15 Ne5 Bxe5 16 dxe5 Rxe5 17 Rxe5 Qxe5 18 Re1 Qd6) 15 Qc2 Bh5 16 Be4 (16 Ne5 Bxe5 17 dxe5 Rxe5=) 16 ... Bxf3 17 Bxf3 Qxd4 with no serious problems. 9 0-0 Nb6 10 Ne5 Nfxd5 11 Nxd5 11 Bd3 has ideas of Bb1 and Qd3. Then 11 ... c5! counters in the centre; e.g. 12 Bc2 cxd4 13 Qd3 Nf6 14 Qxd4 Bc7=. 11 ... Nxd5 12 Bc4 Nb6!? This can’t be too bad, but it does decentralize, so I prefer 12 ... c6! with complete equality. After ... Be6, White will be hard-pressed to create any pressure. 13 Bb3 Qf6 14 Re1 Be6!? 14 ... a5 with the idea 15 a4 Be6 probably helps Black in what follows: 15 Bc2 Bf5 16 Ng4!
16 ... Bxg4!? White gets a little something after this. Instead, 16 ... Qg6! 17 Bxf5 Qxf5 18 Bxh6! Rae8 gives sufficient compensation for the pawn; e.g. 19 h3 Rxe1+ 20 Qxe1 Qd5 21 Qe3! Nc4 22 Qc3 f5 23 Ne3 Nxe3 24 Bxe3 f4 25 Bd2 f3 26 Qb3! Qxb3 27 axb3 a6 28 g3 Rf5! intending ... Rb5 and/or bringing to king into the centre. 17 Qxg4 Rfe8 18 Bd2 Bologan-Socko, Khanty-Mansiysk 2011. Here Black could play with only a minimal disadvantage by taking the outpost with 18 ... Nd5; e.g. 19 Re4 Bb4 20 Rae1 Rf8. 12.2
4
Bd3
Through most of the last century, this was the main move in the Exchange Variation, in some periods played more than seven times as much as 4 Nf3; but today the latter move is somewhat more common. An old semi-arranged draw used to continue 4 ... Bd6 5 Ne2 Ne7 6 0-0 0-0 7 Bf4 Bf5 or something similar, with the handshake in a few moves. As explained in previous editions, it’s easy to break the symmetry at an early stage, and grandmasters use a variety of ideas to do so. My selection of lines in this 4 Bd3 section is mostly unchanged from earlier editions, and their theory has changed around the edges rather than in a substantial way. Still, I have revised my view of one of my favourite subvariations; sometimes the computers give you no leeway! 4 ... Nc6 As far as I can tell, the dynamic move 4 ... c5 is still holding its own, and of course 4 ... Nf6 and 4 ... Bd6 can’t be bad; but for reasons of space I’ll limit myself to one system. 5 c3 Here 5 Nf3 Bg4 6 c3 Bd6 transposes to the main line, which can also arise via 5 ... Bd6 6 0-0 Bg4 7 c3. Others: (a) 5 Ne2 is natural and can’t be bad, but Black can play aggressively by 5 ... Bg4 (or 5 ... Bd6, of course), when White’s two most common moves are: (a1) 6 c3 Bd6!? 7 Qb3 (7 f3 Bh5 8 Bf4 Nge7=, when ... Bg6 follows, swapping off White’s good bishop) 7 ... Nge7 and:
(a11) 8 Qxb7 Rb8 9 Qa6 0-0 10 0-0 Rb6 with a kingside attack is fine; or Black can play the clever 10 ... Bd7!, threatening 11 ... Rb6 12 Qa4 Ne5, which White has serious trouble meeting. Then 11 Bc2 may be best, but 11 ... Nb4! 12 cxb4 Bb5 13 Qa3
Bxe2 14 Qh3! h5! 15 Re1 Bg4 still leaves Black with more active pieces. (a12) If White plays 8 0-0, 8 ... 0-0 is simplest, but a more aggressive approach was 8 ... Qd7!? 9 Re1!? 0-0-0 10 Nd2 h5 11 Nf1 h4 12 Qc2 h3 13 g3, J.Eriksson-B.Lindberg, Stockholm 2008; here 13 ... Rde8 14 b4 Bf5 is the easiest way to equality and perhaps a bit more. (a13) 8 Na3!? a6 9 f3 Bf5 10 Bxf5 Nxf5 11 0-0 0-0 12 Nf4 (12 Qc2 Qf6-/+ Lau-Watson, Philadelphia 1978) 12 ... Re8!? 13 Qxb7 Ncxd4! 14 cxd4 Nxd4, when ... c6 as well as ... Bxf4 is threatened, while 15 Qxd5! Bxf4 16 Qxd8 Ne2+ 17 Kh1 Raxd8 18 Bxf4 Nxf4 still leaves White with the tougher task. (a2) 6 f3 Bh5
7 c3 (7 Nf4 Bg6; 7 Ng3 Qh4 8 Qe2+ Nge7 9 c3 Bg6=) 7 ... Bd6 8 Bf4 (8 0-0 Qf6!? or 8 ... Bg6 9 f4!? Bxd3 10 Qxd3 Qd7 11 Qb5 Nf6 12 Qxb7 Rb8 13 Qa6 0-0 14 Nd2 Ne7! with compensation) 8 ... Nge7 (8 ... Bg6) 9 0-0 (9 Qc2 Bg6 10 0-0 Qd7 11 Nd2 0-00!? with dynamic equality) 9 ... Bg6 10 Bxd6 Qxd6 11 f4!? (11 Ng3 h5! 12 Bxg6 Qxg6=/+) 11 ... Qe6! (11 ... Bxd3 12 Qxd3 0-0-0 13 Nd2 g6 14 b4 Rde8, Ljubisavljevic-Lenz, Celle Ligure 1997; and instead of 15 a4? Nf5, 15 Rae1 Nf5 16 Ng3 Nce7 17 Nb3 would be about equal) 12 Rf3 Nf5 (12 ... Be4!=/+) 13 Kf2?! (13 Ng3 0-0-0) 13 ... Nd6 14 Re3 Be4 15 Ng3 f5=/+ Sollid-J.Watson, Gausdal 1981. (b) A particularly mischievous move is 5 Bb5!?, which transposes into a Winawer with colours reversed (White has wasted a tempo)! I don’t know how many players this has occurred to, but my entire database has only one actual example. As we show in the section on 3 Nc3 Bb4 4 exd5 exd5, the resulting position is objectively equal, with chances for both sides (see 7.6). (c) 5 Nf3 Bd6 (5 ... Bg4 tries to force things more; then 6 c3 Bd6 transposes to the main line, and here 6 ... Qf6 has been played; finally, an accelerated queenside castling plan is 6 ... Qd7 7 0-0 0-0-0, when White knows where his target is sooner, but Black has leeway where to put his pieces) 6 0-0 Nge7 7 c4 (7 Re1 Bg4 8 c3 transposes to the main line) 7 ... dxc4 8 Bxc4 0-0 9 h3, J.Polgar-V.Kramnik, Dos Hermanas 1997; and Kramnik’s 9 ... Be6!? 10 Bxe6 fxe6 was roughly equal, while an active alternative is 9 ... Nf5! (threatening captures on d4 and ... Bh2+) 10 d5 (10 Be3 Qf6 11 Nc3 Bd7= intending ... Rae8) 10 ... Ne5 11 Nxe5 Bxe5 12 Nd2 Re8 13 Re1 (13 Nf3 Bf6) 13 ... h6 14 Nf3 Bf6 15 Rxe8+ Qxe8 with the ideas ... Qe4 and ... b6 (or ... b5) and ... Bb7 and pressure on the weak d5-pawn. 5 ... Bd6 6 Nf3 Instead: (a) 6 Ne2 – the slightly asymmetric placement of pieces here creates all sorts of interesting play. Then 6 ... Bg4 transposes to 5 Ne2 Bg4 6 c3 Bd6!? above.
Others: (a1) 6 ... Qh4!? is an ingenious move played by Alekhine and repeated in several hundred games since with a small performance rating edge for Black. I strongly recommended it in previous editions. Predictably, the computers are good at finding ways to neutralize the more dynamic variations, but the play is still complex and this is a good practical line:
(a11) 7 g3 Qh5 8 Bf4 (8 Nf4 Qxd1+ 9 Kxd1 Nf6 10 Re1+ Ne7= Apsenieks-Alekhine, Buenos Aires Olympiad 1939; 8 0-0 Bg4 9 Qb3 0-0-0 10 Nf4 Bxf4 11 Bxf4 g5 12 Be3 f5! having in mind 13 ... f4!, or otherwise some combination of ... Bh3/ ... Qmoves/ ... h5-h4) 8 ... Bg4 (8 ... Bxf4 9 Nxf4 Qxd1+ 10 Kxd1 Bg4+= McDonald) 9 Bxd6 cxd6 10 Qb3!? (10 Nd2 intending 0-0 and f3 equalizes) 10 ... 0-0-0 11 Nf4 Re8+ 12 Kd2 Qh6! (intending ... Nf6-h5) 13 Kc2! (13 Qxd5 Nf6 14 Qb5 a6 15 Qc4 Bf3-/+; 13 Na3 Nf6! and 14 h4? Nh5 or 14 Kc2 g5 15 Ng2 Be2!) 13 ... Kb8 14 Nd2 Rc8 15 Qa4! Nf6 16 Rae1 g5 17 Ng2 Bd7, with slightly the better of it in an interesting position. (a12) 7 Ng3 Nge7 (7 ... Nf6; or 7 ... Nh6!? 8 Bxh6?! Qxh6 9 0-0 Ne7=/+ Westerinen-Kavalek, Solingen 1986) 8 Nd2 Bg4!? (8 ... Qf6 9 0-0 h5) 9 Nf3 Qf6 10 Be2 Qg6 11 0-0 h5 12 h3 Bxh3 13 gxh3 h4 14 Nxh4 Rxh4 15 Bg4 Rh8=. (a13) 7 Nd2 Bg4 8 Qb3 (8 Qc2 is well answered by 8 ... 0-0-0) 8 ... 0-0-0?! (in Rathnakaran-Ganguly, Chennai 2008, Black chose the more promising 8 ... Nge7! 9 Nf1 0-0-0 10 Ne3 Be6 11 Bd2 Rhe8 12 0-0-0 Kb8=/+) 9 Qxd5 Nf6 (Alekhine). He calls this clearly better for Black, but again, the computer spoils things by 10 Nf3! Qxf2+ 11 Kxf2 Nxd5 12 Ng3 with an unclear position. (a2) Alekhine also suggested 6 ... Qf6, when 7 0-0 is most easily met by 7 ... Bf5=, and more ambitiously by 7 ... Nge7!? 8 Ng3!? (versus ... Bf5) 8 ... h5 9 h3! g6 10 Re1 Bd7= and ... 0-0-0. (a3) 6 ... Nf6 is simplest and retains plenty of positive chances; for instance, 7 Bf4 (7 0-0 0-0 8 Nd2 Ne7!? 9 Qc2 Re8 10
Nf3 Bf5!=; 7 Bg5 h6 8 Bh4 0-0 9 Nd2 Re8 10 Qc2, Kirnos-Savchenko, Alushta 1998; and 10 ... Qe7! wins the bishop pair if White plays 11 Bxf6 Qxf6, avoiding the risky 11 0-0-0 Rb8! and ... b5, or 11 0-0? Bxh2+! 12 Kxh2 Ng4+) 7 ... 0-0 8 Bxd6 (8 0-0 Re8 9 Bg3 Bg4!) 8 ... Qxd6 9 0-0 Ng4!? 10 Ng3 f5 11 h3 f4! 12 Nf5 (12 Ne2! f3 13 hxg4 Bxg4! 14 Ng3 fxg2 15 Qxg4 gxf1Q+ 16 Nxf1 Rf4!), Mann-Arkhipov, Kecskemet 1990; 12 ... Nxf2! 13 Nxd6 Nxd1 14 Rxd1 cxd6=/+. (b) 6 Qf3 has been used quite a bit.
It attacks d5 and covers f5, creating a further imbalance without risking much. Black has several reasonable replies: (b1) 6 ... Nce7, as played by a number of grandmasters, is a subtle move, when 7 Ne2 is normally met by 7 ... c6; but 7 ... h6, intending ... Nf6, and even 7 ... Nf6!? with the idea 8 Bg5 Bg4 are fine. Instead, 7 Bf4 Nf6 8 Ne2?! Bg4 9 Qg3 Bxe2! 10 Bxe2 Bxf4 11 Qxf4 0-0 leaves White’s development lagging: 12 Bd3?! (12 0-0 Ng6 13 Qf5 Re8) 12 ... Ng6 13 Bxg6 Re8+ 14 Kd1 hxg6=/+ Lazic-Kosten, Varallo 1991. (b2) 6 ... Be6 7 Ne2 Qd7 (7 ... Qf6=) 8 Bf4 0-0-0 9 Bxd6 Qxd6 10 Qg3 Nge7 11 Qxd6 Rxd6 12 Nd2 Bf5 13 Bxf5+ Nxf5 14 Nf3 Re8 15 Kd2= Saidov-Volkov, Tashkent 2009. (b3) 6 ... Nf6!? 7 Bg5 (7 h3 Ne4! 8 Ne2 0-0! 9 0-0 f5=; here 9 Bxe4?! dxe4 10 Qxe4 Re8 11 Qf3 Be6! gives Black too much play) 7 ... Bg4 8 Qxg4! (8 Qe3+ Kd7! 9 Ne2 h6 10 Bh4 Re8 11 Qd2 Ne4!) 8 ... Nxg4 9 Bxd8 Nxd8 (or 9 ... Rxd8 10 Nf3 0-0=) 10 Nf3 (10 h3 Nf6 11 Nd2 Ne6 12 Ne2 0-0 13 0-0 Rfe8 14 Rfe1 Re7 15 g3 Rae8 is obviously playable for White but not terribly attractive, Bukavshin-Meier, Neustadt 2008) 10 ... g6 11 0-0 Ne6 12 h3 Nf6 13 Re1 0-0 14 Nbd2 Rfe8= A.Larsson-Hillarp Persson, Swedish Team Ch. 2001. Returning to 6 Nf3: 6 ... Bg4 6 ... Nge7 tends to transpose: after 7 0-0, 7 ... Bg4 followed in 27 of 30 top-rated games. 7 0-0 7 h3?! Bh5 has scored miserably for White; all he’s done is weaken his kingside and bring Black one move closer to the desirable move ... Bg6. 7 ... Nge7
One of the positions most frequently reached in the Exchange French. White can develop smoothly by Re1 and perhaps Bg5 with Nbd2, whereas if Black brings his king to the queenside, he has the ready-made attack b4-b5 and/or Nb3-c5. For his part, Black has an easy and safe equality by continuing ... 0-0 and arranging an exchange of White’s good light-squared bishop. Alternatively, he can play ... 0-0-0 and hurl his kingside pawns at White’s king, although that naturally involves some risk. 7 ... Qd7 is also popular, and again may well transpose after ... Nge7. One point of the move is that Black prepares the option of ... 0-0-0, but he also needs to protect the bishop on g4 if he is to play ... Nge7 and ... 0-0, for fear of White’s Bxh7+. Finally, the queen supports the move ... Bf5 if it proves desirable later. 8 Re1 I think that this is best. Instead: (a) 8 Bg5 is extremely common in practice but arguably inferior because ... f6 helps Black’s kingside attack: 8 ... Qd7 (or 8 ... f6, almost invariably transposing) 9 Nbd2 f6 (or 9 ... 0-0-0 with ... f6 next; or 10 Bh4 Rde8 11 Bg3 Ng6 12 Bxd6 Qxd6, as in Zozulia-Antic, Corinth 2002)
And now: (a1) White plays 10 Bh4?! most often (by a wide margin), but after 10 ... 0-0-0, the idea ... Nf5 and ... g5, ... h5-h4 is very dangerous; for example, 11 b4 Nf5 12 Bg3 Nxg3 (Black has scored overwhelmingly after 10 ... 0-0-0, and 90%+ from this position) 13 hxg3 (13 fxg3 Rde8) 13 ... Ne7 14 Re1 h5 15 Qb3 h4 16 gxh4 Nf5 17 g3 Nxg3! 18 fxg3 Bxg3 19 Re2 Bxh4 20 Rg2 Qe7 21 Rxg4 Qe3+ 22 Kg2 Qf2+ 0-1 Sotnikov-Stepanov, Tula 2010. (a2) 10 Be3 0-0-0!? (10 ... Bf5 and 10 ... 0-0 11 Qc2 Bf5 are equal) 11 b4 (11 Qa4 Kb8 12 b4? Bxf3 13 Nxf3 Nxd4-/+
Trzaska-Köpf, German League 1988; 11 c4 dxc4! 12 Nxc4 Kb8 13 Nxd6 Qxd6=/+), and Black can develop smoothly by 11 ... Nf5, or by 11 ... Rde8 12 b5 Nd8 (12 ... Na5! makes it difficult for White to attack on the queenside) 13 c4!? dxc4 14 Nxc4 Nd5 15 Rc1 Kb8 16 Be2 h5 17 Kh1 g5! (Black has achieved his ideal position) 18 a4 Bf4 19 Qd2 Re7!? 20 Bxf4 Nxf4 21 Bd1?, E.SheinHaimovich, Israeli Team Ch. 2000; and here 21 ... Nxg2! was easiest, intending 22 Kxg2 Bh3+. (b) 8 h3!? (“inaccurate” – Psakhis) 8 ... Bh5 9 Re1 Qd7 10 Nbd2 0-0-0 11 Nf1 (11 Qc2!? Rde8 12 b4 Bg6 13 Nb3 Bxd3 14 Qxd3 Qf5=; 11 b4 Ng6) 11 ... f6 12 Ng3 Bg6 13 Qc2 (13 Bf1 Rhe8 14 b4 Bxg3 15 fxg3 Nf5 16 Bf4 Nd6) 13 ... Bxd3 14 Qxd3 h5! 15 Nh4 g5! (15 ... Bxg3 16 fxg3 Rhe8=) 16 Nhf5 (16 Ng6 Rh6 17 Nxe7+ Nxe7 18 h4 Rg8 with a promising attack) 16 ... Bxg3 17 Nxe7+ Nxe7 18 fxg3 Nf5 19 Rf1 Nd6 20 Rxf6 Ne4=/+. 8 ... Qd7
9 Nbd2 0-0-0 An all-out move, speculative but fun for the attacker. When 8 Bg5 f6 is interpolated, Black’s attack is very fast, but after 8 Re1 it’s not as clear. Objectively, then, 9 ... 0-0 is the safer and surer solution, used by many strong players.
For example: (a) 10 Nf1 Bf5 (10 ... h6 11 Ng3 f5! 12 h3 f4 13 Nf1 Bf5= Thesing-D.King, German League 1998) 11 Ng3 Bxd3 12 Qxd3 Bxg3!? (12 ... Rae8=) 13 hxg3 Rae8 14 b3 Nc8 15 Bf4 f6 16 c4 N6e7 (16 ... g5 17 Bd2 dxc4 18 Qxc4+ Qf7) 17 Nd2 c6 18 Nf1 Ng6 19 Bd2 Rxe1 20 Rxe1 Rd8 21 c5 ½-½ Slobodjan-T.Luther, Schoeneck 1996. (b) 10 Qc2 Ng6 (10 ... Bf5 11 Nf1 Bxd3 12 Qxd3 Nf5 13 Bg5 f6 14 Bh4 Nce7 15 Bg3 Rae8=) 11 h3?! Bxh3! 12 gxh3 Qxh3 13 Bf5 (13 Bf1 Qg4+ 14 Kh1 Rae8-/+) 13 ... Qh5 14 Re3, Stoma-Bugajski, Warsaw 2005; and 14 ... Nce7 (or 14 ... Bf4) 15 Bd3 Bf4 will
win the exchange. (c) 10 b4 Ng6 (10 ... f6 11 Qc2 Bf5 12 Nb3 Rae8=) 11 Nb3 b6 12 Be2 Rfe8 13 b5 Nd8 14 a4 Ne6=/+ 15 Ne5? Bxe2 16 Qxe2 Nxe5 17 dxe5 Bxe5-/+ K.Lee-Murshed, Guangzhou 2010. 10 b4 The logical 10 Nf1 frees the bishop on c1, protects White’s kingside, and prepares Ne3 or Ng3: 10 ... f6 11 h3 Bf5 (or 11 ... Bh5) 12 Ne3 Bxd3 13 Qxd3 g5!? (13 ... Kb8 14 b4 Rde8 is level, with the idea 15 b5 Na5!) 14 Ng4 (perhaps a pawn race with 14 b4!? will serve White better; a sample line would be 14 ... Kb8 15 a4 h5 16 a5 Rdg8 17 b5 Nd8 with complications, or Black could venture upon 14 ... h5 right away) 14 ... Ng8 15 Ne3 Nge7 16 Ng4 Ng8 17 Ne3 Nce7! (avoiding the draw) 18 c4 dxc4 19 Qxc4. Now in M.R.Savic-Ulibin, Paracin 2011, Black played 19 ... Nh6, but 19 ... h5 was a bit more direct; e.g. 20 d5! (20 b4?! g4 21 Nh4 gxh3; 20 h4 gxh4!? 21 Nxh4 Nh6 and Black’s attack is further along) 20 ... g4 21 hxg4 hxg4 22 Qxg4 Qxg4 23 Nxg4 Nxd5=, with chances to play for a win.
10 ... Rde8 This is less to win the e-file than to free d8 for a knight retreat; from that square, the centralizing move ... Ne6 is possible. I previously recommended 10 ... Ng6, which has been much more common and extremely successful, but I find White establishing a small advantage after the simple but hardly ever played 11 h3! with the idea 11 ... Bh5?! (11 ... Bf5 12 Nb3; 11 ... Nf4 12 Bf1) 12 b5 Nce7 13 Qa4 Kb8 14 Ne5!. 11 b5 11 Nb3 is the main alternative:
11 ... Ng6 (or 11 ... Nf5; e.g. 12 Be2 Qe7!? 13 Bd2 Nh4 14 h3 Nxg2!? 15 Kxg2 Bxh3+ 16 Kxh3?? Qe6+ 17 Kg2 Qg4+ 18 Kh1 Qh3+ 19 Kg1 Re6 0-1 H.Hoffmann-Hofstetter, corr 1998) 12 Be3 (12 Rxe8+ Rxe8 13 Nc5?? Bxc5 14 bxc5 Nh4 and Black was winning in Tipu-Psakhis, Toronto 1997) 12 ... Nh4 13 Be2 Nf5 14 Bd2 Rxe2! 15 Qxe2 Nh4 16 b5 Nd8 17 Ne5 Bxe2 18 Nxd7 Bxb5 19 Ne5 f6 20 Ng4 Nf5 21 Ne3 ½-½ Carrabeo Garcia-Matamoros Franco, Coria del Rio 2005. 11 ... Nd8 12 Qa4 12 c4 dxc4 13 Nxc4 Nd5 14 Bd2 Rxe1+ 15 Bxe1 f6 is equal; e.g. 16 a4 Nf4 17 a5 Kb8 18 b6 (18 Bf1 Nde6 19 a6 Ng5!) 18 ... cxb6 19 axb6 a6 20 Bf1 (20 h3 Bh5 21 Bf1 Nde6) 20 ... Nc6 21 Bc3 Rd8. 12 ... Kb8 13 Ba3 13 c4 dxc4 14 Bxc4 Ng6 15 Ba3 Bxa3 16 Qxa3, Asztalos-Peredy, Budapest 1996; and 16 ... Nf4 was good enough, but I think 16 ... f6= is best, protecting e5 and preparing ... Nf7-d6 in some lines. 13 ... f6 13 ... Bxa3 14 Qxa3 f6 is fully equal; e.g. 15 Qb3 Ng6 16 a4 Nf4 17 Bf1 h5, Melnikova-Praveen Kumar, Kuala Lumpur 2010. After 13 ... f6, Kholmov-Psakhis, USSR 1984, continued 14 Bxd6 Qxd6 15 Re3 Nc8 16 Rae1 Rxe3 17 Rxe3 Qd7 18 Qd1 g6 19 a4 Nf7 20 Nb3 b6 21 h3 Bxf3 22 Rxf3 f5 23 c4 dxc4 24 Bxc4 Re8=. 12.3 4 c4 This move has become one of White’s main options. It can transpose to 4 Nf3 and 5 c4, but also has independent significance. I’ll concentrate upon two solutions: 12.31 12.32
4 4
... ...
Bb4+ Nf6
12.31 4 ... Bb4+
This is a very sound move which at the same time limits White’s options. 5 Nc3 5 Bd2 Bxd2+ tends to be weak (or at best harmless) in such positions because, in an isolated pawn position following ... Bxd2 and ... dxc4, the exchange of bishops favours Black; e.g. 6 Qxd2 Nf6 7 Nc3 0-0 8 cxd5 (8 Bd3 Re8+ 9 Nge2 dxc4 10 Bxc4 Nc6) 8 ... Re8+ 9 Be2 Bg4!. After 5 Nc3 there is another fork in the road: 12.311 12.312
5 5
... ...
Ne7 Nf6
I’m not thrilled with 5 ... Qe7+!? 6 Be2 Bg4, as in Okhotnik-Caruana, Arvier 2007, because the surprising 7 Kf1! Bxe2+ 8 Ngxe2 dxc4 (8 ... Nf6 9 Bg5!) 9 Qa4+ c6 10 Nd5! b5 11 Nxe7 bxa4 12 Nf5, while playable for Black, is more comfortable for the first player. 12.311
5
...
Ne7
Black avoids getting pinned by Bg5 and maintains great flexibility.
6 Nf3 Others: (a) 6 a3 Bxc3+ 7 bxc3 0-0 8 Nf3 Nbc6= intending ... Na5 and/or ... Bg4; e.g. 9 Bd3!? dxc4 10 Bxc4 Nd5 11 Bxd5?! Re8+ 12 Be3 Qxd5 13 h3?(13 0-0 Bg4)13 ... Qc4! 14 Qc2 b6-/+ H.Speck-T.Luther, Liechtenstein 1993. (b) 6 Bd3 prepares Nge2 (to avoid an awkward pin by ... Bg4), but a knight is passively placed on e2 and Black may even get slightly the better of it by 6 ... Nbc6 7 cxd5 (7 Ne2 dxc4 8 Bxc4 0-0) 7 ... Nxd5 8 Ne2 Bg4 9 f3 Be6. 6 ... 0-0 6 ... Bg4! will usually transpose and pretty much amounts to the same thing. I actually prefer this less commonly chosen order because it puts immediate pressure on the centre. These useful early pins are characteristic of the Exchange French.
(a) 7 Be2 dxc4 8 Bxc4 0-0 transposes to 7 Be2 etc in the main line; but 7 ... Nbc6 8 a3 Bxc3+ 9 bxc3 dxc4 10 Bxc4 Qd7!? 11 0-0 0-0-0 is an interesting alternative. (b) 7 Qa4+ Nbc6 8 Ne5 Be6=. (c) 7 a3 Bxc3+ 8 bxc3 Nbc6 9 Bg5 (9 a4 Na5 10 cxd5! 0-0!= with the idea 11 c4? Re8; 9 cxd5 Nxd5 10 Bd2 Qe7+ 11 Qe2 Bxf3 12 Qxe7+ Ncxe7 13 gxf3 Ng6=/+; 9 Bd3 dxc4 10 Bxc4 0-0 11 0-0 Nd5 12 Bd2 Nb6 13 Be2 Na5 14 Re1 Nac4 15 Bf4 Re8) 9 ... 0-0! 10 Be2 (10 cxd5 Qxd5 threatens capture on f3, hitting the g5-bishop) 10 ... f6 (10 ... Qd7 may be more accurate: 11 0-0 dxc4 12 Bxc4 Nd5 intending ... Na5, ... b5 etc) 11 Bd2 Na5=. Black’s control of the light squares compensates for the bishop pair:
if 12 c5, 12 ... b6!; and 12 cxd5 Qxd5 13 Rb1 (13 0-0 Nc4 14 Re1 Rfe8 15 Rb1 b5 16 a4 a6) 13 ... Nc4! plans ... b5. 7 a3!? 7 Be2 dxc4 8 Bxc4 Bg4 9 0-0 Nbc6 is a typical position in which Black has pressure on the d-pawn and light squares. Several games have gone 10 Be3 (10 a3 Ba5 11 Be3 Nf5 12 Re1 Qd6=) 10 ... Nf5!? (10 ... Rb8!? 11 a3 Bxc3 12 bxc3 Nd5 13 Bd2 Na5 14 Ba2 b5 and the light squares compensate for White’s bishops) 11 Qd3 Nd6 12 Bd5 (12 Bg5!? Qc8 13 Bd5 Bf5!? 14 Qd1 Re8=) 12 ... Bf5 13 Qd1 (13 Qe2 Na5 14 Bg5 Qc8 15 Rfc1 Re8 16 Qf1 c6 17 a3 Bxc3 18 Rxc3 Be6 19 Bxe6 Qxe6 20 b4 Nac4=/+ Ashley-Remlinger, New York 1994) 13 ... Ne7 14 Bb3 c6 15 Ne5 Kh8 16 Rc1 f6 17 Nd3 Bxc3 18 bxc3 Nd5 19 Nf4 Nxe3 20 fxe3 Qe8=/+ Waitzkin-Shaked, Bermuda 1997. 7 ... Bxc3+ 8 bxc3 Nbc6
9 Be2 9 c5 b6!=/+, since 10 cxb6 axb6 gives Black the a-file and all the light squares, perhaps even with ... Ba6 next. 9 ... dxc4 10 Bxc4 Nd5 11 Bd2?! A better course is 11 0-0! Nxc3 12 Qd3 Nd5 13 Re1 Qd6 14 Rb1 a6 (versus Rb5) 15 Ba2 Nce7=. After 11 Bd2?!, Santo Roman-Shaked, Cannes 1997, continued 11 ... Re8+ 12 Be2 Qe7-/+ 13 c4 Nb6 (or 13 ... Bf5! with the idea 14 cxd5? Bc2) 14 Be3 Bg4 15 0-0 Rad8 16 d5 Bxf3 17 gxf3 Ne5 18 Qb3 Nbxc4! 19 Bxc4 Nxf3+ 20 Kg2 Qe4! with an overwhelming attack. 12.312
5
...
Nf6
6 Nf3 Others: (a) 6 Bd3 followed by Ne2 tries to avoid the pin by ... Bg4, but a knight is passively placed on e2 and Black may even get slightly the better of it after 6 ... Nc6 7 cxd5 (or 7 Ne2 dxc4 8 Bxc4 0-0) 7 ... Nxd5 8 Ne2 Bg4 9 f3 Be6. (b) 6 a3 Bxc3+ 7 bxc3 0-0 disturbs White’s development because of checks on the e-file; e.g. 8 Bd3 (8 Nf3 Re8+ 9 Be2 Ne4!) 8 ... dxc4 9 Bxc4 Qe8+! and White has difficulties in view of 10 Ne2 (10 Be2 Qc6 11 Nf3 Ne4!; 10 Qe2? Qc6 11 Nf3 b5; 10 Kf1 Be6!) 10 ... Qc6!, hitting c4 and g2. (c) 6 cxd5 Nxd5 7 Bd2 0-0 8 Bd3 c5! 9 dxc5 Nxc3 10 Bxc3 Bxc3+ 11 bxc3 Nd7 12 c6 Nc5! 13 Bc2 Re8+ 14 Ne2 Qe7-/+ M.Kuijf-Korchnoi, Tilburg 1992. 6 ... 0-0 6 ... Nc6 is another move that may transpose after 7 Bd3 0-0 8 0-0 dxc4 etc, or play can develop uniquely by 7 Be2 (7 Bg5 Bg4 is natural; but also interesting is 7 ... 0-0 8 cxd5 Qe8+!? 9 Be2 Nxd5 10 Rc1 Be6 with a pleasant game) 7 ... 0-0 (7 ... dxc4 8 Bxc4 0-0 is our main line) 8 0-0 h6 9 a3 Bxc3 10 bxc3 Re8. 7 Be2 7 Bd3 is frequently played, although it gives Black the option of 7 ... Re8+ (otherwise 7 ... Nc6 8 0-0 dxc4 9 Bxc4 and 7 ... Bg4 8 0-0 dxc4 9 Bxc4 Nc6 transpose to our main line again, while 9 ... Bxf3 is seen in the note with 8 ... Bg4 9 0-0) 8 Be3 Bg4 (8 ... Ng4?! 9 0-0!) 9 0-0 Nc6
10 h3 Bh5 (10 ... dxc4) 11 Rc1 (11 c5 Bxc3 12 bxc3 Ne4, Gaponenko-Moskalenko, Tenerife rapid 2004; and White should play 13 Rc1 Qf6 14 Re1 when 14 ... Rad8= might follow) 11 ... dxc4 12 Bxc4 Qd7 13 a3, T.Taylor-Stopa, Los Angeles 2011; and here instead of 13 ... Ba5 14 Na4, I like 13 ... Bxc3!? 14 bxc3 Na5! 15 Ba2 b5 16 Bg5 Nd5 (16 ... Ne4 17 Be3 Nc4=) with lightsquare control. White will get the worst of it after 17 g4? (17 Ne5 Rxe5 18 Qxh5 g6) due to 17 ... Bxg4 18 hxg4 Qxg4+ 19 Kh2 (19 Kh1 Qh5+ 20 Kg1 Re4) 19 ... Qh5+ 20 Bh4 (20 Kg1? Re4) 20 ... Nf4! (20 ... Re4 21 Bxd5) 21 Rg1 Re2 (hitting f2) 22 Kg3 Re4! intending ... Ne2+. 7 ... dxc4 7 ... Nc6 8 0-0 dxc4 9 Bxc4 Bg4 again transposes. 8 Bxc4
8 0-0 might get messy after 8 ... Be6!? 9 Ng5 Qd7!? (or 9 ... Bf5 10 Bxc4 Nc6 11 Be3 Qd6!) 10 Nxe6 Qxe6 11 Bf3 Qa6!. 8 ... Nc6 Instead: (a) 8 ... Re8+ 9 Be3 Be6 10 Bxe6 Rxe6 11 0-0 of Spassky-Korchnoi, Zürich 2001, is slightly better for White. (b) 8 ... Bg4 should equalize in view of some tactical ideas: 9 Be3 (9 0-0 Nc6 transposes to the main line, but gives Black the option of 9 ... Bxf3 10 Qxf3 Qxd4 11 Qxb7 Qxc4 12 Qxa8 Bd6!, threatening 13 ... Nc6 and 13 ... Bxh2+, and 13 Qxa7 Bxh2+ 14 Kxh2 Qxf1 15 f3 Qe1 will likely end in perpetual) 9 ... Nd5 10 0-0!? Bxc3! 11 bxc3 Nxc3 12 Qb3 Bxf3 13 gxf3 b5 14 Bd3 (14 Qxc3 bxc4 15 Qxc4 Nd7-/+) 14 ... Nd5 15 Qxb5 Nxe3! 16 fxe3 Nd7 17 Rac1 c5!= Gulko-Psakhis, USSR Ch., Riga 1985. 9 0-0 Bg4 10 Be3
A key position that has been reached repeatedly over the years. The play is intricate and, in my opinion, perfectly satisfactory for Black. 10 ... Rb8 There have actually been nine moves here(!). I like two of them in particular. 10 ... Rb8 has the virtue of simplicity: it defends the b-pawn and thus threatens ... Bxf3, (the immediate 10 ... Bxf3 11 Qxf3 Nxd4 12 Qxb7 favours White). The rook move also prepares ... b5, contesting the light squares, which tend to be Black’s strength in these positions. The other effective move which produces equality is 10 ... Qd7, connecting rooks and again focusing on the light squares. Without going into too much detail, here are some excerpts: (a) 11 a3 Ba5 (11 ... Bxc3 12 bxc3 Rfe8 has the idea of ... Nd5) 12 Rc1 Rad8 13 Na4 Bb6 14 Nxb6 axb6 15 b4 (15 h3 Bh5), Comp Deep Fritz-Hübner, Dortmund 2001; and simply 15 ... Rfe8 16 h3 Bh5 leaves White under some pressure and wondering what to do. (b) 11 h3 Bh5 12 Be2 Rfe8 transposes to 11 Be2 Rfe8 12 h3 Bh5. (c) 11 Be2 and now:
(c1) 11 ... Rfe8 has scored well for Black. White may have nothing better than 12 h3!? (12 Qb3 Rad8 13 Rad1 Bxc3 14 bxc3 Nd5 15 c4 Nf4 16 Bxf4 Rxe2 is at least equal) 12 ... Bxh3!? (12 ... Bh5 13 a3 Bxc3!? 14 bxc3 Nd5 15 Bd2 Qd6!?=) 13 gxh3 (13 d5? Bxg2 14 Kxg2 Qg4+ 15 Kh1 Qh3+ 16 Kg1 Rxe3! 17 fxe3 Qg3+ 18 Kh1 Ng4) 13 ... Qxh3 14 Bd3 Qg4+ 15 Kh1 Qh3+ 16 Kg1, repeating. (c2) 11 ... Rad8 is quite as good; for example, 12 a3 Ba5 13 Na4 (13 h3 Bxf3 14 Bxf3 and 14 ... Nxd4, or 14 ... Bb6 with the idea 15 d5 Ne5) 13 ... Bb6 14 Nxb6, Miezis-Masserey, Geneva 1999, when Black got creative by 14 ... cxb6!?, but simply 14 ... axb6 15 Rc1 (15 h3 Bh5) 15 ... Rfe8=/+ is a bit awkward for White. 11 Be2 11 a3 Bxc3 12 bxc3 Nd5 13 Bd2 b5! fixes the light squares, which should guarantee equality; ... Na5-c4 is one idea, as well as ... Re8 and ... Be6 in some lines. 11 ... Nd5 11 ... Re8 seems more accurate, since ... Nd5 can wait. 12 Qc2 12 Nxd5 Qxd5 13 Rc1 has been played twice; if nothing else, 13 ... Qxa2 will draw after 14 Bc4! (14 h3 Bf5!=/+) 14 ... Qxb2 15 Rc2 Bxf3 16 gxf3 Qa3 17 Ra2 Qc3 18 Rc2=.
12 ... Be7!? 12 ... Bxc3 13 bxc3 Qe7 is a decent option, since 14 Ng5 (14 Rae1 Nxe3 15 fxe3 Qxe3+ 16 Kh1 Qh6 17 Bc4 is more enterprising, and unclear) 14 ... g6 15 Bxg4 Nxe3 16 fxe3 Qxe3+ 17 Qf2 Qxg5 18 Bf3 Ne7 gives White dubious compensation for the pawn. After 12 ... Be7!?, Wehmeier-Claesen, Lippstadt 1998, continued 13 Rad1 Ncb4 14 Qb1 Bh5!? (14 ... Re8 15 a3 Nxc3 16 bxc3 Nd5 with ideas of ... Bd6 and ... b5) 15 Rd2?! (15 a3 Nxc3 16 bxc3 Nd5 17 Bd2 b5=) 15 ... Bf6 16 a3 (16 Nxd5 Nxd5 and ... Re8) 16 ... Nxc3 17 bxc3 Nd5 with excellent play for Black. 12.32 4 ... Nf6 5 Nc3 Be7 A traditional placement and the starting point of a host of games through the years. 5 ... Bb4 is examined by transposition in 12.312. I should mention that 5 ... Nc6 is also playable and will generally transpose, but there are unique lines that look satisfactory including 6 Nf3 Bg4 (6 ... Bb4 is also covered in 12.312; see the note there with 6 ... Nc6) 7 cxd5 Nxd5 8 Be2 (8 Bc4 Qe7+! 9 Kf1 Nb6) 8 ... Bb4 9 Qb3!? 0-0 10 0-0 Be6=. 6 Nf3 Sometimes White plays 6 Bd3 first, either to avoid the ... Bg4 pin or to play Nge2. An independent line is 6 ... Nc6 (6 ... Be6!? 7 cxd5 Nxd5) 7 cxd5 Nxd5 8 Nge2 Be6 9 0-0 0-0 10 Be3 g6 (10 ... Ncb4 11 Bb1 g6 12 a3 Nxe3 13 fxe3 Nd5=) 11 Qd2 Re8 12 Rac1 Qd7 13 Rfd1 Rad8= Ashley-Granda Zuniga, New York 1997. 6 ... 0-0 7 Bd3 7 Be2 sidesteps some lines with ... Nb4 and is placed to meet ... Bg4; but this also means that after cxd5 and ... Nxd5, his bishop won’t be on its ‘proper’ d3-square aimed at the kingside. A few ideas: (a) 7 ... Be6 has become a popular move in this and related positions:
8 0-0!? (8 Ng5?! Bf5; 8 c5 b6 9 cxb6 axb6 10 0-0 c5 11 Ne5 Ne4!=; 8 Qb3 Nc6! 9 Qxb7 Nb4 10 0-0 dxc4 is unclear, or 10 ... Rb8 11 Qxa7 Ra8=) 8 ... dxc4 9 Ng5 Bd5 (an extra pawn is as important as the bishop pair) 10 Nxd5 (10 Qc2 c6 retains the pawn) 10 ... Qxd5 (10 ... Nxd5 11 Nf3 Nb6!?) 11 Bf3 Qb5 12 Qe2 Re8 13 b3 Nc6 14 bxc4 Qa4 with no problems. (b) 7 ... dxc4 8 Bxc4 Bg4 (Black can always head for a standard isolated queen’s position; e.g. 8 ... Nbd7 9 0-0 Nb6 10 Bb3 Nbd5 11 Ne5 Be6 12 Qf3 c6 13 Re1 Nc7 14 Be3 Bxb3 15 axb3 Ne6= Kharlov-Maslak, chessassistantclub.com 2004; these choices are a matter of taste) 9 0-0 Nc6 (an absolutely typical position which can arise from several 4 c4 move orders) 10 Be3 Qd6!? (10 ... Rb8 has the idea ... Bxf3 and if Qxf3, ... Nxd4, since b7 is protected; then 11 Be2 Be6 is solid) 11 h3 Bxf3 12 Qxf3 Nxd4 13 Qxb7 ½-½ Vlassov-Nikolenko, Moscow 2002; there might follow 13 ... Nf5 14 Rad1 Qe5 15 Rfe1 Bd6 16 g3 Nxe3 17 Rxe3 Qh5 and White has at most a nominal edge. (c) 7 ... Nc6 8 0-0 Be6!? 9 cxd5 (9 c5!? Ne4) 9 ... Nxd5 10 Bb5 (10 Re1 a6 11 Bd3 Qd7) 10 ... Ncb4 11 Re1 (11 a3 a6 12 Ba4 b5=) 11 ... a6 12 Bf1 (12 Ba4 c5!?) 12 ... Re8= Miezis-Short, Leon 2001. 7 ... Nc6 7 ... dxc4 8 Bxc4 is ‘b’ in the previous note.
8 cxd5 Trying to make use of the bishop on d3. Instead: (a) 8 0-0 Bg4 9 Be3 Nb4!? (9 ... dxc4 10 Bxc4 is 7 Be2, note ‘b’ again) 10 Be2 dxc4 11 Bxc4 c6 12 h3 Bh5!? (12 ... Bf5) 13 Re1? (13 g4 Bg6 14 Ne5 Nbd5 15 Bg5) 13 ... Bxf3 14 gxf3 (14 Qxf3 Nc2) 14 ... Nbd5=/+ Voitsekhovsky-Zolnierowicz, Ceske Budejovice 1995.
(b) 8 h3?! gives Black an extra tempo for 8 ... Nb4!, aiming for 9 Be2 Bf5 10 0-0, and here 10 ... Bc2! is good: 11 Qd2 Ne4! 12 Nxe4 Bxe4 13 a3 Nc2 14 Ra2 dxc4 15 Bxc4 Qd6=/+. 8 ... Nxd5 Or 8 ... Nb4 9 Bb1 Nbxd5 10 0-0 Re8 11 Re1 Be6 with the idea 12 Ng5 Nxc3 13 bxc3 Bd5=. 9 0-0 Be6 10 a3 Another typical set-up is 10 Re1 Re8 11 Qc2 h6 12 a3 Qd7 13 Be3 Rad8= Afromeev-Yagupov, Tula 2006.
10 ... Bf6 Perhaps not best. Two others equalize: (a) 10 ... Re8 11 Re1 (11 Bc2 Qd7 12 Qd3 achieves little after 12 ... g6 intending ... Bf5, ... Rad8, and/or ... Bf6) 11 ... Qd7=. (b) 10 ... Qd7! is a little more complicated: 11 Bb5 (11 Qc2 g6) 11 ... Bd6! 12 Ng5 Nxc3 13 bxc3 Bf5 14 d5 Qe7! 15 Re1 (15 dxc6 Qe5 16 Nf3 Qxb5) 15 ... Ne5 16 f4 a6! 17 Bc4 b5 18 fxe5 Bc5+ 19 Kh1 bxc4=. After 10 ... Bf6, Gjuran-Korchnoi, Rogaska Slatina 2011, continued 11 Be4 g6!? (11 ... Qd7!) 12 Bh6 Re8 13 Re1 Qd7 14 Qd2 Bh8!? 15 Rad1 Rad8 16 h3 Nce7 17 Qc2 c6 with approximate equality. 12.4 4 Bf4 This has been touted as the best way for White to play for a draw, but as mentioned above, it’s childish to talk about a draw with so many pieces on.
In fact, White may not only be trying to neutralize ... Bd6, but also intending to play ambitiously, for example, by castling queenside. This strategy has received attention of late, so I’ll devote some space to it. For his part, Black isn’t too worried: in the final analysis, every logical move should equalize. For the sake of an active repertoire, I’ll present a few of ways to play for an imbalance; naturally, there are others: 4 ... Bd6 4 ... Nf6 is obviously playable, whereas 4 ... Nc6 creates a desirable tension (assuming you are playing for a win). The first idea of 4 ... Nc6 is to prevent 5 Bd3 by attacking d4. Play might go 5 Nc3 (pursuing his queenside strategy; 5 Nf3 Bd6) 5 ... Nf6 (5 ... Bb4!? intending 6 a3 Qe7+!) 6 Nf3 (6 Qd2 Bd6 7 0-0-0 0-0=/+) 6 ... Bd6 7 Qd2 0-0 8 0-0-0 Re8 9 Bd3 (9 Bg5 Bb4) 9 ... Bxf4 10 Qxf4 Bg4 with the idea ... Bxf3. None of this says much for 4 Bf4. 5 Bxd6 5 Qd2 plans Nc3 and 0-0-0.
Then: (a) 5 ... Nc6 6 Nc3 Nf6 transposes to 5 ... Nf6 below. (b) 5 ... Qe7+ 6 Be2 Nf6 (the sane move, and already at least equal; the exciting alternative is 6 ... Bxf4 7 Qxf4 Qb4+ 8 Nd2 Qxb2 9 Rb1 Qxc2 10 Ngf3 Ne7 11 0-0 – A.Marshall; an adventurous sort might dig in with 11 ... Nbc6 12 Rfc1 Qg6 here; but 8 Nc3 Qxb2 9 Qe3+ Ne7 needs to be analysed too) 7 Nc3 0-0 (7 ... Bb4 is another problem, after which 8 f3 0-0 isn’t comfortable, and 8 Qe3 Ne4!? 9 a3 Bxc3+ 10 bxc3 g5! 11 Bg3 Nc6 intends 12 Nf3 f5! 13 Be5 0-0 and ... f4) 8 Nf3 (8 0-0-0 Nc6 9 Nf3 Ne4!=/+, or here 8 ... Bf5) 8 ... Nc6 9 0-0 (9 Bxd6 Qxd6) 9 ... Ne4 10 Bxd6! Qxd6 11 Qe3 Nxc3!? 12 Qxc3 Bg4 13 Rae1! Rae8=, but still with ideas for both sides. (c) 5 ... Nf6 6 Nc3 (6 Bd3 Nc6 7 Ne2 Nb4) 6 ... Nc6 (6 ... 0-0 allows 7 Bd3, although 7 ... Nc6 8 Nge2 Nb4 gets rid of the bishop on d3, which guarantees at least equality) 7 Bb5 0-0 8 Nge2!? (8 Bxc6 bxc6 9 Nge2 is slightly passive; e.g. 9 ... Rb8 10 b3 Re8 11 0-0 Bf5) 8 ... Bxf4 9 Nxf4 (9 Qxf4 Nb4 10 Qd2 c6 11 Ba4 Na6 12 0-0 Nc7 and White’s bishop is out of play) 9 ... Re8+ 10 Nfe2 Qd6 11 0-0 Bd7 12 h3 Re7 13 Rae1 Rae8 14 Bxc6 Bxc6 15 Ng3 Bd7 with an insignificant advantage but considerable play left. 5 ... Qxd6 6 Nc3 Instead: (a) 6 Bd3 Nf6 7 c3 0-0 8 Ne2 b6!? (8 ... Bg4) 9 0-0 c5! 10 Nd2 Bg4 11 f3 Bd7 12 Qe1 Re8 13 Qf2 Nc6 gave Black the initiative and better game in Stefanova-Kindermann, Vienna 1996. (b) 6 c3 is safe and prevents ... Qb4+, but 6 ... Nf6 (6 ... Bf5 7 Bd3 Ne7) 7 Nf3 (7 Bd3 transposes to ‘a’) 7 ... 0-0 maintains a lead in development: 8 Be2 Qb6 (or 8 ... b6 9 0-0 c5) 9 b3 (9 Qb3 Qe6) 9 ... Re8 10 0-0 c5! 11 dxc5 Qxc5=/+ and it’s not so easy for White to develop; for example, 12 Qd4 Qc7 or 12 Nd4 Nc6 13 Nxc6 bxc6 14 Bd3 Bg4. 6 ... Nc6 Played at once in order to prevent Bd3. 7 Nf3 7 Bb5 Nf6 8 Nge2 0-0 9 0-0 Ne7! and it’s nothing much, but I’d prefer to be Black. 7 ... Nf6 Or 7 ... Bg4 8 Be2 Nf6. 8 Be2 8 Qd2 0-0 9 0-0-0 Re8 10 h3 (10 Bd3 Bg4) 10 ... Ne4=/+. 8 ... 0-0 9 0-0 Re8 Black’s position is without problems and, in my opinion, easier to play.
Chapter
Thirteen
Classical
Variation
with
4
e5
1 e4 e6 2 d4 d5 3 Nc3 Nf6 The Classical Variation of the French Defence. 4 e5 Now the pawn structure becomes similar to the Advance Variation and to those variations of the Tarrasch that include ... Nf6 and e5. In the next two chapters we consider 4 Bg5. Instead: (a) 4 Bd3 is seldom played due to 4 ... c5 and then: (a1) 5 exd5 cxd4 (or 5 ... exd5=; the bishop isn’t optimally placed on d3 if this turns into an isolated queen’s pawn position) 6 Bb5+ (6 Nb5?! Bc5! 7 Nf3 e5! 8 Nxe5 a6 9 Na3 0-0 10 0-0 Qxd5 11 Re1 b5 12 Bg5 Bb7 13 f3 Qd6=/+; I think this line improves upon the traditional 6 ... Nxd5, which is equal) 6 ... Bd7 7 Bxd7+ (7 Qxd4?! Bxb5 8 Nxb5 Nxd5 9 Ne2 Nc6=/+) 7 ... Qxd7 (7 ... Nbxd7!? with the idea 8 dxe6 dxc3 9 exd7+ Qxd7=/+) 8 dxe6 Qxe6+ (8 ... fxe6! 9 Nce2 Nc6 10 Nf3 Bc5=/+) 9 Nce2 Nc6 10 Nf3 Bb4+ 11 Bd2, Em.Lasker-Marshall, New York (m/2) 1907; and 11 ... Bc5 12 0-0 0-0 13 Re1 Qc4= was sounder than 11 ... 0-00?!. (a2) 5 Nf3 cxd4 (or 5 ... Nc6=) 6 Nxd4 e5 (6 ... Nc6 7 Bb5 Bd7 8 exd5 exd5 9 0-0 Be7 10 Be3 0-0= Em.Lasker-Bogoljubow, Zürich 1934) 7 Nf3 d4 8 Ne2 Bg4 9 0-0 (9 Nxe5?? Qa5+) 9 ... Nc6=. (b) 4 exd5 is harmless after 4 ... exd5, a version of the Exchange Variation with no bite (see Chapter 12), or 4 ... Nxd5=. 4 ... Nfd7
Now 13.1 13.2 13.3
White 5 5 5
has: Nf3 Nce2 f4
There are two other interesting moves: (a) 5 Qg4!? c5 and the centre is breaking up so White must try to get Nb5-d6 in, or perhaps 6 dxc5 at the right moment: 6 Nf3 (6 Nb5 cxd4 7 Bf4 Qa5+ 8 Bd2 Qb6=/+) 6 ... cxd4 7 Nb5 (7 Qxd4 Nc6 8 Qf4 a6! intending ... Qc7) 7 ... Nc6 8 Nd6+?! Bxd6 9 Qxg7? (9 exd6 0-0 10 Nxd4 Nde5=/+) 9 ... Bb4+! (9 ... Bxe5 10 Nxe5 Qf6 11 Qxf6 Nxf6=) 10 c3 dxc3 11 Qxh8+ Nf8 12 Kd1 cxb2 13 Bxb2 Qb6-/+. (b) 5 Qh5!? is Haldane’s move:
5 ... c5 (the naive 5 ... g6 also looks fine; e.g. 6 Qg5 Qxg5 7 Bxg5 a6, intending ... c5) 6 Nf3 (6 dxc5 Nc6 7 Nf3 Bxc5) 6 ... cxd4 (or 6 ... g6) 7 Nxd4 g6! (or 7 ... Qb6=/+ with the idea 8 Ndb5 Nc6) 8 Qg4?! (8 Qg5 Qxg5 9 Bxg5 a6=/+ with ideas of both ... h6 and ... Nc6, but this isn’t as bad) 8 ... Nxe5 9 Qg3 Nbc6 (9 ... Bg7 10 Bf4 Nbc6 11 Nxc6 transposes) 10 Nxc6 Nxc6 11 Bf4 Bg7 12 Bc7 (McDonald notes that 12 Bd6 is inaccurate due to 12 ... Qa5!) 12 ... Qf6 13 Bd6. White is a pawn down versus Black’s unopposed centre with no lead in development. I think there are a few good answers, including Neil McDonald’s suggestion 13 ... Ne7 (or 13 ... a6 14 0-0-0 Bd7) 14 Bd3 (or 14 Bb5+ Bd7 15 Bxd7+ Kxd7 and ‘Black king’s is safe behind his pawns and the threat is 16 ... Nf5, winning the white bishop’ – McDonald) 14 ... 0-0 15 f4 Nf5 16 Bxf5 Rd8! 17 Bc7 Rd7=/+, at least. 13.1 5 Nf3 Aiming for piece play instead of pawn expansion; this is considered fairly harmless, but there are some challenging lines. 5
...
c5
6 Ne2 Now we have reached a position which could come from 5 Nce2 c5 6 Nf3. Others: (a) I could place 6 dxc5 here, but since it famously arises via 1 e4 e6 2 Nf3 d5 3 Nc3 Nf6 4 e5 Nfd7 5 d4 c5 6 dxc5, I’ve put it with the second moves of the ‘Uncommon Early Moves’ chapter, section 17.21, under that order. (b) 6 Bb5 has more than one answer, but the most ambitious is 6 ... Nc6!; e.g. 7 0-0 cxd4 8 Nxd4 Ndxe5! (8 ... Qc7 9 Bf4!) 9 f4 (9 Bf4 a6 10 Bxe5 axb5 11 Bg3 Bc5! 12 Ndxb5 0-0 with the centre and bishop pair; 9 Re1 Bd6! 10 f4! Nd7 11 Nxd5 Nxd4 12
Qxd4 0-0 13 Ne3 Qc7=/+; 9 Qh5 Bd6 10 Bg5 Qd7=/+) 9 ... a6! 10 Ba4 Nd7! 11 f5 (11 Nxd5?? Nxd4 12 Qxd4 b5–+) 11 ... Nxd4 12 Bxd7+ Bxd7 13 Qxd4 Rc8-/+ D.Schneider-J.Watson, Chicago 1992. (c) 6 Bg5!? is a quick-developing and fascinating move called ‘The Jackal’.
It has a mediocre reputation, but theoreticians haven’t looked at it carefully and I see no refutation for Black. Both sides can keep things interesting with all kinds of exciting action. Here’s an overview: 6 ... Qb6 7 dxc5! (7 Bb5 is unproductive due to 7 ... a6 8 Bxd7+ Bxd7, preventing Na4)
and now: (c1) 7 ... Qxb2!? 8 Nb5 Nxe5 (the only move; 8 ... Na6? 9 Rb1 Qxa2 10 Ra1 Qb2 11 Rxa6; 8 ... Qb4+? 9 c3 Qa5?? 10 Qa4! wins due to 10 ... Qxa4 11 Nc7 mate, while 9 ... Qxc5 10 Be3 is awful) 9 Nd2!? (9 Nc7+ Kd7 10 Nxa8 Nxf3+ 11 gxf3 Qe5+ 12 Be3 Bxc5 is unclear) 9 ... Na6 10 Rb1 Qxa2 11 Ne4! h6 12 Ra1 (12 Ned6+ Bxd6 13 Nxd6+ Kf8) 12 ... Qb2 13 Bd2 Nc4 14 Nbd6+ Nxd6 15 Bc3 Qxa1 16 Nxd6+ Bxd6 17 Qxa1 Nxc5 18 Bxg7 Rg8 19 Bxh6 Bd7 and the position is unresolved. (c2) 7 ... Nxc5 8 Rb1 (8 Be3 a6!; 8 Be2 Nc6 9 0-0 a6 10 Rb1 transposes to 8 Rb1) 8 ... a6 9 Be2 Nc6 10 0-0 Qc7 11 Bf4 b5 12 b4 Nd7 13 a4=. (c3) 7 ... Bxc5 is utterly chaotic:
(c31) 8 Na4 (a touch speculative) 8 ... Bxf2+ (8 ... Qa5+!? 9 c3 Bf8 10 b4 Qc7 11 Bf4 Nc6 12 Bd3 Ncxe5 13 Nxe5 Nxe5 14 Bb5+ Bd7 15 Bxd7+ Kxd7 16 0-0 with some attack) 9 Ke2 Qa6+ (9 ... Qb5+? 10 c4!) 10 Kxf2 Qxa4 11 b3 (11 Bd3 Nxe5!; 11 c4!? Qa5 12 cxd5 exd5 13 Bd3 Nc6 14 Re1=/+; though this is still complex) 11 ... Qa5 12 Bd3 Nc6 13 Re1, probably favouring Black but full of life. (c32) 8 Qd2!, with one more division: (c321) 8 ... Qxb2? 9 Rb1 Qa3 10 Nb5±. (c322) 8 ... Bxf2+ initiates a lengthy and romantic-style forcing sequence: 9 Qxf2 Qxb2 10 Kd2! Qxa1 11 Bb5! (11 Nb5? 0-0 12 Nc7 Na6! 13 Nxa8 Nac5–+) 11 ... Qxh1 12 Qc5 Qxg2+ 13 Kc1 (13 Ne2 Nc6 14 Bxc6 Qxg5+ 15 Nxg5 bxc6 16 Qxc6 Rb8 17 Nxe6!=) 13 ... Nc6 (13 ... Qh1+?? 14 Kb2; 13 ... f6?? 14 exf6) 14 Bxc6 Qxg5+ 15 Nxg5 bxc6 16 Qxc6 Rb8 17 Nxd5! is a variant on this theme: 17 ... exd5 18 e6 f6 19 Qd6 Rb6 20 exd7+ Bxd7 21 Qxd5! Rf8 22 Ne4 and White has all the winning chances. Amazing! (c323) 8 ... h6 9 Bh4 (9 Be3!?) 9 ... g5 10 Bg3 Qxb2 11 Rb1 Qa3 12 Nb5!? (12 Rb3 Qa5 13 Nb5 Qxd2+ 14 Nxd2 Bb6!=/+) 12 ... Qxa2 13 Rc1 is unclear. (c324) 8 ... Nc6 9 0-0-0 (9 Na4 Bxf2+ 10 Qxf2 Qb4+ 11 c3 Qxa4 12 Be2 Qc2=/+) 9 ... Bxf2 10 Na4 Qb4 11 Qxf2 Qxa4 12 a3 (12 Kb1 0-0 13 b3!?) 12 ... 0-0 13 Qg3 f6! (13 ... Kh8 14 Bd3 with a serious attack) 14 exf6 (14 Bh6 Rf7) 14 ... Nxf6 15 Bd3 Bd7 (a position played several times!; 15 ... Ne4!? 16 Bxe4 Qxe4) 16 Kb1 (or 16 Rhe1=), Roy Chowdhury-Illingworth, Sydney 2009; and 16 ... Ne4 was best and about equal.
6 ... Nc6
6 ... cxd4 7 Nexd4 Nc6 is also reasonable; e.g. 8 Nxc6 bxc6 9 Bd3 Nc5 (9 ... Qc7 10 Qe2 Be7=) 10 Be2 Be7 11 0-0 0-0 12 Be3 Rb8= Jimenez Chacon-Hamdouchi, Dos Hermanas 1998. 7 c3
This position also arises from 3 Nd2 Nf6 4 e5 Nfd7 5 Ndf3 c5 6 c3 Nc6 7 Ne2. 7 ... cxd4 7 ... Qb6 and 7 ... b5 have also been played successfully. The simplest alternative is 7 ... Be7 and:
(a) 8 Ng3?! Qb6 9 Be2 cxd4 10 cxd4 Bb4+ 11 Kf1 f6=/+. (b) 8 Nf4 cxd4 9 cxd4 Qb6 10 Be2?! (10 a3 g5! 11 Ne2 g4 12 Nfg1 f6!), Yudkovsky-Coello, Cannes 1997; 10 ... g5! 11 Nh5 g4 12 Ng5 Bb4+ 13 Kf1 Qxd4-/+. (c) 8 g3 0-0 9 Bh3 b5! (9 ... b6 intending ... Ba6 is also logical, when 10 Qa4 Qc7 11 0-0 a5=/+ doesn’t prevent Black’s plan) 10 0-0 b4 11 Nf4, Ponomariov-Ni Hua, Khanty-Mansiysk 2011; and 11 ... bxc3! 12 bxc3 Ba6 13 Re1 cxd4 14 cxd4 Nb6 begins an effective queenside attack while White struggles to find play. (d) 8 a3 a5 (or 8 ... 0-0 with the idea 9 b4 cxd4 10 cxd4 b5!) 9 Nf4 Qb6 (preventing Bd3) 10 h4 a4=. 8 cxd4 f6
9 Nf4 9 exf6 Nxf6 10 Nc3 Bd6 is balanced and comfortable for Black (compare the 3 Nd2 Nf6 Tarrasch; here White is committed to an early Nc3, which precludes Bf4 and Nf4 plans): (a) 11 Bb5 (to control e5) 11 ... 0-0 12 0-0 Qc7 (12 ... Qb6=) 13 h3 Bd7 14 Be3 a6 15 Bxc6 (15 Bd3 Be8!? 16 Rc1 Bh5 with the idea 17 g4 Nxg4 18 hxg4 Bxg4 19 Be2 Qe7 20 Kg2! Rf6! with an attack sufficient to draw) 15 ... bxc6 (15 ... Bxc6= KuzminLempert, Alushta 1994) 16 Rc1 Be8 17 Ng5 Qe7=/+. (b) Typical tactical themes arise after 11 Bd3 0-0 12 Bg5:
12 ... Qe8!? (12 ... Qb6 13 0-0, and now 13 ... Bd7= or, if desired, 13 ... Qxb2 14 Na4 Qa3 15 Bc1 Qb4 16 Bd2 Qa3 17 Bc1=) 13 Qd2? Nh5!? (13 ... e5!=/+) 14 Ne2?! (14 Qc2! Nf6=; 14 0-0 Nf4) 14 ... e5 15 Nxe5 (15 dxe5? Bb4 16 Nc3 Bg4) 15 ... Nxd4 (15 ... Bb4!) 16 Nxd4 Qxe5+ 17 Be3 Nf4 with a pleasant edge, Nijboer-Glek, Wijk aan Zee 1999. 9 ... Bb4+ For some reason, no one seems to have taken 9 ... fxe5, a thematic central break-up, very seriously: 10 Nxe6 (10 dxe5 Ndxe5 11 Nxe5 Bb4+ 12 Bd2 Bxd2+ 13 Qxd2 Nxe5) 10 ... Qb6
and now: (a) 11 Qb3!? Bb4+ 12 Kd1? (12 Bd2! Ke7 13 Bxb4+ Qxb4+ 14 Qxb4+ Nxb4 15 Nc7 Nc2+ 16 Kd2 Nxa1 17 dxe5 Rb8 18 Nxd5+ Ke6 19 Nc7+ Ke7=) 12 ... Nf6-/+ 13 Nxg7+? Kf7 14 dxe5 Ng4–+ 15 Nh5 Nd4 16 Nxd4 Qxd4+ 17 Bd3 Nxf2+ 18 Kc2, Motwani-Rrhioua, Gibraltar 2004; and here 18 ... Bd7 would win immediately. (b) 11 Be2 was suggested as an improvement, when after 11 ... e4! 12 Nfg5 Nf6 13 0-0 (13 Bf4 Bxe6 14 Nxe6 Kf7 isn’t clear), Black might try the extremely messy line 13 ... h6!? (13 ... Bd6 14 Nxg7+ Ke7) 14 Bh5+ g6! 15 Bxg6+ Ke7, which can go in several directions; e.g. 16 Nxf8 (16 Nxe4 dxe4 17 Nf4 Bg4) 16 ... hxg5 17 Bxg5 Rxf8 18 Qh5 Be6 19 Qh7+ Bf7=. 10 Bd2 Qe7 10 ... Bxd2+ 11 Qxd2 Qe7 transposes. 11 Bxb4 11 exf6 Nxf6 12 Bd3 Ne4! 13 Bxe4 dxe4 14 Ne5 0-0 15 0-0 Bxd2 16 Qxd2 Nxe5 17 dxe5 Qg5 18 g3 Qxe5 and ... b6 with a very slight advantage. 11 ... Qxb4+ 12 Qd2
12 ... Qe7 The most reliable solution. 12 ... Ke7 13 exf6+ gxf6 is playable as well; e.g. 14 Qxb4+ Nxb4 15 Kd2, Shirov-Ivanchuk, Tilburg 1993; and 15 ... Kd6!, intending an early ... e5, establishes equality. 13 exf6 13 Bb5?! fxe5 14 dxe5 0-0 15 Bxc6 bxc6 16 0-0 Rb8 17 b3 c5=/+.
13 ... Nxf6 14 Bd3 Ne4! 15 Qe3! Qb4+ 16 Kf1 0-0 Velimirovic-Moskalenko, Belgrade 1988. There followed 17 Re1 (17 Bxe4 dxe4 18 Qxe4 Qxb2=), and Moskalenko suggested 17 ... Qd6 18 Ne2 e5 as advantageous, although 19 Bxe4 dxe4 20 Qxe4 exd4 (20 ... Bf5 21 dxe5 Qc5 22 Qe3! Qxe3 23 fxe3 Rae8 24 Ned4 Nxe5 25 Nxe5 Rxe5 26 Kg1 and Black has compensation but White should be fine) 21 Nexd4 Nxd4 22 Qxd4 Qa6+ 23 Kg1 Qxa2 24 h4 with the idea h5-h6 looks about equal. 13.2
5
Nce2
This section is largely concerned with move orders, transpositions and, most importantly, ways in which Black can bypass the main lines entirely should he so choose. When White plays 5 Nce2, he is usually trying to get to the formation with Ne2/f4/Nf3/c3. This is examined in Chapter 6 in the line 3 Nd2 Nf6 4 e5 Nfd7 5 f4 c5 6 c3 Nc6 7 Nf3. But there’s a serious question whether White can get to that main line without running into attractive options for Black, regardless of whether he plays 5 Nce2 or 5 f4. So, although 5 f4 is the order used in the next section (13.3), I’m going to look at that move right now, but in the context of trying to achieve that Nce2 reorganization. The move order issues discussed here and in the next few notes are potentially of great importance for both players and repay study. Black can bypass all of the Classical Variation lines with the structure e5/f4/c3 by denying White the chance to get that far. Thus: 5 f4 c5 6 Nf3 (6 dxc5 and other alternate moves not leading to Nce2 systems are given in 13.3) 6 ... Nc6 7 Ne2.
This is a very important line that can also come from 5 Nce2 c5 6 f4 Nc6 7 Nf3, as well as from 3 Nd2 Nf6 4 e5 Nfd7 5 f4 c5 6 Ndf3 Nc6 7 Ne2. As I pointed out in the last edition, books tend to skimp on this position or even ignore it. Black has several ways to respond: (a) After 7 ... Qb6 White can return the game to normal channels with 8 c3 (transposing to section 6.1). His most important alternative is 8 g3, after which Black has two good lines: (a1) 8 ... cxd4 9 Nexd4 Nxd4 10 Nxd4 Bc5 (10 ... Nb8!? 11 c3 Bc5) 11 c3
11 ... 0-0 (11 ... a5!? 12 Rb1!? 0-0 13 Bh3 Kh8 14 Kf1? f6! 15 Kg2 Nxe5! 16 fxe5 fxe5 led to a Black win in RuhrigReefschläger, German League 1986; as a whole, 11 ... a5 seems equal) 12 Bh3 Nb8 13 Nb3 Nc6 (13 ... Na6!) 14 Nxc5 Qxc5 15 Qd3 d4!? (15 ... b6! 16 Bg2 Rd8 17 Be3 Ba6! 18 Bxc5 Bxd3 19 Bd6 Na5 20 0-0-0 Bf5= with a double-edged game) 16 0-0 dxc3+ 17 Be3 Qa5 18 Qxc3 Qxc3 19 bxc3 b6 20 Bg2 Bb7, Zysk-H.Franke, German League 1988; White may have a little something, but his space advantage is at least partially cancelled by the weak c-pawn on an open file. (a2) 8 ... f6 9 Bh3 (again 9 c3 transposes to a main line – see 6.11; after 9 exf6?! Nxf6 10 Bg2, 10 ... Bd6 11 c3 0-0 12 dxc5 Bxc5 13 Ned4 Ne4 14 0-0 e5! 15 fxe5 Bg4 could follow) 9 ... Nxd4 10 Nexd4 cxd4 11 exf6 (11 Nxd4 fxe5+/= intending 12 Nxe6? Nf6!) 11 ... Bb4+ (or 11 ... Nxf6 with the idea 12 0-0 d3+ 13 Kg2 dxc2 14 Qxc2 Bc5) 12 Kf1 Nxf6 with equality or better. (b) 7 ... b5 looks sound (even if it seems even more logical once White has played g3):
8 a3!? (8 g3 b4 and perhaps ... Ba6 next; or 8 c3 b4 with ... Qa5 and ... Ba6 as typical follow-up moves; White needs to watch out for something like 8 dxc5 Bxc5 9 Ned4? Nxd4 10 Nxd4 Nxe5! 11 Bxb5+ Bd7 12 Bxd7+ Nxd7 or 11 fxe5 Qh4+) 8 ... Rb8
Black achieves productive queenside play in all cases: (b1) 9 dxc5 Nxc5 10 Ned4 Nxd4 11 Qxd4 Ne4! 12 Qxa7?! (12 b4!? a5 13 Bd3 Be7=) 12 ... Rb7 13 Qe3 Bc5 14 Nd4 f6. (b2) 9 Be3 b4 (9 ... Qc7 10 c3 Na5) 10 axb4 Rxb4 11 b3 a5 12 c3 Rb8 13 dxc5 was Yudasin-Glek, Tilburg 1994, and Glek suggests 13 ... Nxc5 14 Ned4 Nxd4 15 Nxd4 Bd7 16 Bd3 Nxd3+ (16 ... a4! 17 bxa4 Qa5) 17 Qxd3 Be7=. (b3) 9 c3 Be7 (9 ... c4) 10 b4 (10 g3 0-0 11 Bg2 a5 12 0-0 a4 with ideas of ... Nb6-c4 and ... Na5-b3, advancing by ... b4 at some point; 10 dxc5 Nxc5 11 Be3 Bd7) 10 ... cxb4 11 axb4 g6 (11 ... Nb6!?) 12 Ng3 (12 g3?! a5 13 bxa5 Nxa5) 12 ... a5 13 bxa5 Nxa5 14 Bd3 Qc7 with unclear prospects. (c) 7 ... Be7 is a productive waiting move, not committing to ... Qb6, ... cxd4, or ... b5.
The first thing I should say is that playing ... Be7 before ... Qb6 is not necessarily a universal solution to all f4 lines. For example, in our Tarrasch Defence order 3 Nd2 Nf6 4 e5 Nfd7 5 f4 c5 6 c3 Nc6 7 Ndf3, 7 ... Be7 is sometimes played, but it does allow 8 Bd3 under rather favourable circumstances, since the d4-pawn doesn’t hang as it would after 7 ... Qb6. Be that as it may, in our position with Nce2 already played, ... Be7 is more appropriate and White has to show his cards: (c1) 8 c3 allows Black to transpose into a main line of the Tarrasch French by 8 ... Qb6 (see 8 ... Be7 in the notes to 6.1) or he can go his own way by 8 ... 0-0; e.g. 9 a3 (9 g3 allows another known position after 9 ... Qb6 10 Bh3 cxd4 11 cxd4 f6!? – see 6.211; or Black can expand on the queenside by 9 ... a5!? 10 Bg2 b5!? with ideas of ... Ba6 and ... b4) 9 ... a5 (9 ... a6 10 b4 cxd4 11 cxd4!? b5! and ... Nb6) 10 g3
10 ... cxd4 (10 ... b6!?; 10 ... a4!? 11 Bh3 b5) 11 cxd4 (11 Nexd4 Nc5= intends moves like ... Bd7 and ... a4) 11 ... f6!? (11 ... f5 12 Bg2?! a4 13 0-0 Nb6 wins key queenside squares) 12 Bh3 fxe5 13 dxe5 (13 Bxe6+ Kh8 is messy, but the first point is 14 dxe5?! Ndxe5!) 13 ... Qb6 (or 13 ... Nc5 intending ... a4) 14 Nc3? (for better or worse, 14 Bxe6+ Kh8 15 Bxd5 was correct, when 15 ... Ndxe5 16 fxe5 Bg4! transposes to Chapter 6, specifically note ‘a1’ after 9 g3 in section 6.42; that is unclear, but here Black has the option of 15 ... Nc5!, which gives him more than enough play for the two pawns) 14 ... Ndxe5!? (14 ... Nc5! ties White down completely, since 15 Nd4 Nxd4 16 Qxd4 Qa6 threatens ... Nb3 and ... Nd3+) 15 fxe5 Nxe5 16 Bg2 Bd7 17 Rf1 Nc4 with a strong attack for the piece; Bologan-M.Gurevich, Cap d’Agde (rapid) 2002, went 18 Rb1 Bf6 19 Qd3 Rac8 20 Ng5 Bxg5 21 Bxg5 Rxf1+ 22 Bxf1 Rf8 23 Qe2 Rf5! 24 Bf4 g5 25 Nxd5 exd5 26 Bb8 Rf8 0-1. (c2) 8 Be3 Qb6 forces White to lose a valuable tempo defending his b-pawn. (c3) 8 g3 Qb6 gives White a similar option of a Tarrasch Variation after 9 c3 (see 8 ... Be7 9 g3 in the notes to 6.1), but with g3 already in, which isn’t his plan in most of these 5 Nce2 lines. Another possible plan for Black might be 8 ... 0-0 9 Bg2 b6!? intending 10 0-0 Ba6. In fact, even the immediate 8 ... b6 is worth thinking about. (d) 7 ... cxd4 8 Nexd4 transposes to the Tarrasch variation 3 Nd2 Nf6 4 e5 Nfd7 5 f4 c5 6 Ngf3 Nc6 7 Nb3 cxd4 8 Nbxd4, which is dealt with in Chapter 6 in note ‘b31’ to 6 c3. e) 7 ... Qa5+!? 8 Bd2 Qb6 9 Bc3 transposes to 6 f4 Qa5+ 7 Bd2 in the next note. 5 ... c5 6 c3 The most obvious move. Instead, 6 Nf3 transposes to 5 Nf3 c5 6 Ne2 in section 13.1 above. But 6 f4 is an important option because of the issues raised in the next note.
Then 6 ... Nc6 7 Nf3 returns to the previous note, but 7 c3! will transpose to our main line without having to deal with the 7 ... f6 variation of the next note. However, Black has the opportunity to deviate from the main lines here as well. For example, these experimental ideas are sound enough: (a) 6 ... Qb6 7 Nf3 (7 c3 Be7 8 Nf3 0-0 transposes) 7 ... Be7 is a sort of waiting game; e.g. 8 c3 0-0 9 dxc5 (9 Qb3 Qxb3 10 axb3 Nc6) 9 ... Nxc5 10 Ned4 Nc6 11 Be2 Bd7 intending ... Rac8 and/or ... Ne4. (b) 6 ... Be7 is a similar idea, when White might try 7 Nf3 0-0 8 c3 (8 Be3 Qb6) 8 ... Nc6 (8 ... Qb6; 8 ... f6!? 9 Be3 fxe5 10 fxe5 Qb6 11 Qd2 Nc6 12 0-0-0! Qa5=) 9 Be3 b6 with mixed prospects. (c) 6 ... Qa5+!? tries to take advantage of White’s slow development and contemplates a queenside advance:
(c1) 7 Bd2 Qb6 8 Bc3 Nc6 9 Nf3 f6!? (or 9 ... cxd4!? 10 Nexd4 Nc5=/+) 10 exf6 (10 f5 fxe5! 11 fxe6 cxd4 12 exd7+ Bxd7 13 Bd2 e4, or 13 Nexd4 exd4 14 Nxd4 0-0-0 intending ... Bc5) 10 ... Nxf6 11 g3 Ne4 with pressure. (c2) 7 c3! b5!? (double-edged; 7 ... Nc6 is normal) 8 Nf3 (8 dxc5 b4 9 a3 Bxc5 with the idea 10 Bd2 Qb6 or 10 cxb4 Bxb4+ 11 Bd2 Bxd2+ 12 Qxd2 Qb6, about equal) 8 ... b4 9 Bd2 Nc6 10 cxb4 Nxb4 11 Nc3 cxd4 12 Nxd4 Ba6=. 6 ... Nc6 Assuming that Black has achieved this exact position after all the move order jockeying implicit in the last two notes, yet another deviation 6 ... cxd4 7 cxd4 f6! creates new problems for White and seems adequate to maintain a balance:
(a) Black’s development is fast after 8 exf6 Nxf6 9 Nf3 Bd6 (9 ... Nc6 10 Nc3 transposes to 9 exf6 Nxf6 10 Nc3 in the notes to 13.1) 10 g3 Nc6 11 Bg2 0-0 12 0-0 Qb6 13 Rb1 Bd7 14 Bf4 Ne4 gave Black plenty of activity in Bologan-Glek, German League
1992. (b) 8 f4 fxe5 9 dxe5 (9 fxe5? Bb4+ 10 Kf2 0-0+ 11 Nf3 Nc6 with the initiative and ideas of sacrificing on e5) 9 ... Nc6 10 Nf3 Bb4+ 11 Nc3 (11 Bd2 Nc5) 11 ... Nc5 12 Be3, Shirov-Ivanchuk, Tilburg 1993, and a straightforward path was 12 ... Ne4 13 Qc2 (13 Rc1 Qa5 14 Bd2 0-0) 13 ... Qa5 14 Rc1 Qxa2 15 Bd3! Nxc3 16 0-0 Qa4!=/+. (c) The main line is 8 Nf4 Bb4+ 9 Bd2 Qb6 10 Bxb4 (10 Rc1 0-0 11 Bxb4 Qxb4+ 12 Qd2 Qe7= Atalik; 10 Qh5+ g6 11 Nxg6 Bxd2+ 12 Kxd2 Qxb2+ 13 Ke3 Qc3+ 14 Bd3 Qxa1 15 Nxh8+ Kd8-/+) 10 ... Qxb4+ 11 Qd2 Qxd2+ 12 Kxd2 Ke7
In this well-known theoretical position Black has equal chances: (c1) 13 exf6+ (the most commonly played) 13 ... gxf6 14 Re1 (14 Nf3?! Kd6; 14 Nge2 Nc6 15 Nd3 e5) 14 ... Nb6 and now two important moves:
(c11) 15 Bd3 allows Black to force things in the centre; e.g. 15 ... Nc6 16 Nf3 Kd6! 17 Nh5 Rf8! (17 ... e5 18 Nxf6 e4 19 Ng5 Kc7 unclear) 18 Bxh7 (18 Re2 e5=/+) 18 ... e5 19 dxe5+ (19 Ng3 e4 20 Nh4 Nxd4 21 f3 Rh8! 22 Nhf5+ Bxf5 23 Nxf5+ Ke5 and Black’s king was a monster in Bezgodov-Sakaev, Russian Ch. 1999) 19 ... fxe5 20 Kc1 (20 Rhf1 Bg4-/+) 20 ... Bg4 21 Ng3 Bxf3 22 gxf3 Nd4-/+ Arakhamia Grant-Gleizerov, Isle of Man 2001. (c12) 15 Nf3 Kd6!, when White must find something better than transposing into ‘c11’ by 16 Nh5 Rf8 17 Bd3 Nc6, and 16 Bb5 a6 17 Bd3 Nc6 is no improvement. (c2) 13 Nf3 fxe5 14 Nxe5 Nxe5 15 dxe5 Nc6 16 Nd3 Bd7 17 f4 Be8 18 Rc1 Rf8 19 g3 (‘+/=’ according to theory) 19 ... Bg6! 20 Nc5 Rfc8!? (or 20 ... Rab8) 21 Nxb7 (21 Bb5 Nb4!) 21 ... Nb4!=; White loses the c-file. 7 f4
At this point 7 ... Qb6 8 Nf3 transposes into the main lines of Chapter 6. But the most important revelation of this section is that Black needn’t enter the main lines of the 4 e5 Classical lines if he doesn’t want to, assuming that some of the deviations above are sufficiently interesting to him. 13.35 f4 c5 6 Nf3 Here, too, 6 Nce2 can be played; whereas 6 dxc5 surrenders the centre prematurely: 6 ... Bxc5 (6 ... Nc6 7 a3 Bxc5 8 Qg4 0-0 9 Nf3 Nd4!? 10 Bd3 f5 11 Qh3 Nxf3+ 12 Qxf3 Bb6 13 Ne2 Nc5=/+ Tarrasch-Spielmann, Nuremberg 1906; 9 ... Qb6 is also effective) 7 Qg4 0-0 8 Bd3 f5 9 Qh3 Bxg1!? (an original idea, playable but not inspiring – nevertheless, it was enough to confuse Fischer himself; the alternative 9 ... Nc6 is fine, and better still is 9 ... Bb6 10 Nf3 Nc5! 11 Bd2 Ne4=/+ or 9 ... Qb6=/+) 10 Rxg1 Nc5 11 Bd2 Nc6!? 12 Nb5?! (12 0-0-0) 12 ... Qb6 13 0-0-0 Bd7 14 Nd6 Na4! 15 Bb5 Nd4 16 Be3 Ne2+! 17 Bxe2 Qxb2+ 18 Kd2 Qb4+ 19 Kc1 Nc3-/+ R.J.Fischer-Benko, Curacao 1962. 6 ... Nc6
7 Be3 This is arguably the main line of the entire Classical Variation. White decides to hold the centre with pieces rather than pawns. Instead, 7 Ne2 is covered in the note to 5 Nce2 in 13.2 above. Other moves are interesting but less effective: (a) 7 Bd3?! cxd4 8 Ne2 Qb6 9 0-0 f6! 10 exf6 Nxf6 11 Kh1 Bc5 12 a3 a5 13 b3 0-0 14 Bb2 g6? (14 ... Ng4! 15 Qe1 e5! 16 Nxe5 Ncxe5 17 fxe5 Rxf1+ 18 Qxf1 Nxe5 and Black has the better of it) 15 Qe1 Bd7!? 16 Qh4 is unclear, Abdel Aziem-Peng Zhaoqin, Cairo 2002. (b) 7 a3? meets the fate of most slow moves: 7 ... cxd4 8 Nxd4 Ndxe5! with the idea 9 fxe5 Qh4+. (c) 7 Bb5?! can be answered by various moves, among them: 7 ... cxd4 (7 ... a6 8 Bxc6 bxc6 9 0-0 cxd4 10 Nxd4 c5 11 Nf3 Be7=/+ with the bishop pair and mobile centre, Megibow-Cotton, Concord 1995) 8 Nxd4 Nxd4 (8 ... Qb6 9 Be3 a6 10 Bxc6 bxc6=/+) 9 Qxd4 a6 10 Bxd7+ Bxd7 11 0-0 Rc8 12 Be3 Bc5= 13 Qd2 Qb6 14 Bxc5 Rxc5 15 Qd4 Rc6 16 Qxb6 (16 Ne2 Qb5=/+) 16 ... Rxb6 17 Rab1 Ke7=/+. (d) 7 dxc5 Bxc5 8 Bd3 f6 (or 8 ... a6 9 Qe2 Qc7 10 Bd2 b5 11 a3 Rb8= Boleslavsky-Pachman, Saltsjöbaden Interzonal 1948) 9 exf6 (9 Na4=) 9 ... Nxf6 10 Qe2 0-0 11 Ne5?! (11 Bd2 Bd7 12 0-0-0 Nd4=/+) 11 ... Qc7=/+ (11 ... Nd4 12 Qd1 Bd7 is also attractive), with the idea ... Nd4 or ... Nxe5, as well as ... Bd4 or ... Nd7. 7 ... Be7
Something new for this edition. InPTF3, we examined 7 ... cxd4, which is still going strong. The modest move 7 ... Be7 has become popular only in the past few years, and is now established as one of the main lines. Unlike 7 ... cxd4, it leaves the pawn structure as is, waiting for White to commit, and has a rather subtle point described in the next note. Since the purpose of 7 ... Be7 is clear – that is, development and preparation for castling – I won’t outline Black’s strategy in depth. Nonetheless, it should be noted that he often keeps the centre unresolved and plays with ... a6/ ... b5 or ... b6 on the queenside, and/or ... f6 on the kingside, without exchanging any pawns. White tries to develop harmoniously, and is left with the question of how to make progress against Black’s solid structure. One really attractive feature of 7 ... Be7 followed by ... 0-0 is that Black has a much easier time playing ... f6, a move which not only pressures White’s centre but discourages the advance f5, since White’s pawn on e5 will hang. As it happens, the move exf6 isn’t an effective solution in most cases because it tends to free Black’s game and expose weak squares in White’s camp. 8 Qd2 The most popular reply. White keeps his position flexible, and if Black plays ... Qb6, or ... cxd4 and ... Bc5, White’s bishop on e3 is protected. In the meantime, the knight on c3 has a useful retreat square on d1 should Black manage to get ... b5-b4 in. Finally, f2 is sometimes a powerful square for the queen. The alternatives are also critical: (a) 8 dxc5 is a frequently-played and extremely significant option, often connected with castling queenside. It’s particularly important to see that, with the main line move order 8 Qd2 0-0, 9 dxc5 will transpose to lines analysed within this note.
(a1) 8 ... Nxc5 9 Be2 0-0 10 0-0 has been considered better for White, although 10 ... b6 11 Qe1 (11 a3 Bb7 12 Qe1 Ne4) 11 ... f5 looks quite reasonable. (a2) 8 ... 0-0 might be valuable as a transpositional finesse: (a21) Notice that 9 Qd2 is the same as the common move order 8 Qd2 0-0 9 dxc5. Now 9 ... Bxc5 10 Be2 transposes to line ‘a22’, or 10 Bxc5 Nxc5 11 0-0-0 to line ‘a3’. But with White making the queen move now, Black has the alternative 9 ... Nxc5!?, when 10 ... b6 can follow, or in some cases ... a6 and ... b5. For example, 10 0-0-0
10 ... Qc7!? (10 ... Qa5?! 11 Kb1 Bd7? falls into 12 Nxd5! of Zhigalko-Li Shilong, Cappelle la Grande 2006; 10 ... b6 11 Kb1 Bb7 can go in various directions, with Black shifting between using the c-file, advancing with ... f6 or ... f5, or playing ... Ne4; e.g. 12 h4 Ne4 13 Qe1 Bb4!? 14 Bd2 Nxd2+ 15 Qxd2 d4!? 16 Nxd4 Nxd4 17 Qxd4 Bxc3 18 bxc3 Qc7, when 19 Qd7 Qxc3 20 Qxb7 Qb4+ draws) 11 Kb1 b6 12 Bb5 Bb7 13 h4 Na5!? (risky; Black seems to stand quite well in this type of position; e.g. 13 ... Ne4! 14 Nxe4 dxe4 15 Ng5 Rfd8 16 Qe2 h6 17 Nxe4 Nxe5=/+) 14 Bd4 (14 Bxc5!? bxc5 15 f5 f6! 16 exf6 Rxf6 is hardly clear, but Black seems to be holding his own) 14 ... Rac8!? 15 Ng5 Nc4 (15 ... h6? 16 Bxc5 bxc5 17 Qd3±) 16 Qe2! h6 17 Bxc4 dxc4 18 Bxc5, Nakamura-Vitiugov, Reggio Emilia 2011/12; and for the last time Black can still keep a balance by 18 ... bxc5 19 Nge4 Rb8, aiming for 20 Qxc4 Bxe4! 21 Qxe4 Rb4 22 Qf3 Rfb8 23 b3 c4 with more than enough compensation. (a22) The other unique line is 9 Be2 Bxc5 10 Qd2 Bxe3 11 Qxe3 f6 12 exf6 Nxf6 (White’s queen position allows Black a key gain of tempo because it’s vulnerable to both ... d4 and ... Ng4) 13 Nd4! (13 0-0-0 Ng4 and ... Rxf4) 13 ... e5! 14 fxe5 Ng4 15 Qd2! (15 Bxg4? Bxg4 16 Nxc6 bxc6 17 Nd1 d4 18 Qg3 Qa5+ 19 c3 Rae8–+ Shomoev-Riazantsev, Moscow 2007) 15 ... Ncxe5 (15 ... Qe7!?) 16 Rf1! Rxf1+ 17 Bxf1 Nxh2 18 0-0-0 Nxf1 19 Rxf1 Ng4 20 Nf5 Bxf5 (or 20 ... Be6 21 Nxd5! Kh8!=) 21 Rxf5 Nf6 22 Nxd5
Nxd5 23 Rxd5 Qf6=. (a3) 8 ... Bxc5 is the most common and produces complex, interesting positions.
If White now plays 9 Nd4?!, we have precisely the same position as last edition’s 7 ... cxd4 8 Nxd4 Bc5, except that it’s Black’s turn! That’s because White played extra move dxc5, and while Black had to move his bishop twice (to e7 and c5), he saved a tempo by not playing ... cxd4. Since 7 ... cxd4 8 Nxd4 Bc5 is still such a popular variation for Black, it shouldn’t be surprising that White doesn’t often play 9 Nd4 with a tempo less. So instead we see White answer 8 ... Bxc5 with 9 Bxc5 (9 Qd2 0-0 10 0-0-0 Qa5 11 Bxc5 Nxc5 transposes, or 11 Nd4 Bxd4 12 Bxd4 Rb8! 13 Kb1 b5 14 Ne2 b4 15 Be3 Ba6 16 Nd4 Rfc8= with a slight initiative, Shomoev-Potkin, Dresden 2007) 9 ... Nxc5 10 Qd2 0-0 11 0-0-0 (notice that 8 ... 0-0 9 Qd2 Nxc5 sidesteps this position if you’re so inclined – see line ‘a21’ above) 11 ... Qa5!? (the attacking move; 11 ... Qb6 intending ... Ne4 is also fine; but since White now launches a kingside attack, you could also choose to keep the queen closer to that flank: some possibilities include 11 ... Qe7, 11 ... Bd7 12 h4 Rc8, and 11 ... b6)
Nevertheless, 11 ... Qa5 seems satisfactory, as shown by: (a31) 12 Kb1 Bd7 13 Bd3?! (13 Be2=) 13 ... Nb4 14 a3 (14 Nd4 Rfc8 15 g4? Ba4!-/+ Saric-M.Gurevich, Cannes 2007) 14 ... Nbxd3 15 cxd3 Nb3 16 Qc2 Qb6 17 Ka2 Na5! intending ... Rfc8 and ... d4, Vila Gazquez-Berczes, Szeged 2008, which continued 18 Na4 (18 d4 Rac8 19 g4 Bb5!) 18 ... Qe3!=/+. (a32) 12 h4 (the most dangerous move) 12 ... Bd7 (12 ... Rb8!? 13 Qe3 Bd7 14 h5 Rfc8 is unclear, Kurnosov-Kotsur, Moscow 2011)
After 12 ... Bd7, play diverges: (a321) 13 Kb1 could be met conservatively by a combination of ... Rfd8 and ... Rac8; but the continuation of ShadrinaGrabuzova, Sochi 2007, was sharper and probably better: 13 ... Rfc8 14 Rh3 Rab8 15 Rg3, and the attack by 15 ... Nb4! 16 a3 Nxc2! 17 Qxc2 Ne4 would have been dynamically equal after 18 Ng5! Ba4 19 Qd3! Rxc3! 20 bxc3 Nxg3 (or 20 ... Rc8, ultimately winning a bunch of pawns for a piece) 21 Qxh7+ Kf8 22 Qh8+ Ke7, and whether White takes the rook or g-pawn, a drawn position will result. (a322) 13 h5 b5!? (13 ... h6 is safe) 14 h6 (14 Bxb5 Nb4 15 Bxd7 Nxa2+=/+), and now 14 ... g6 of Li Chao-Ni Hua, Chinese Team Ch. 2008, would have been equal had White played 15 Kb1. McDonald suggests that Black might get more with 14 ... b4!? e.g. 15 hxg7 bxc3 16 gxf8Q+ (16 Qxc3 Rfb8!) 16 ... Kxf8! 17 Qxc3 Nb4 18 Qe1 Nxa2+ 19 Kb1 Nb4 with attack. Probably White holds after 20 Nd4 Qa2+ (20 ... Rc8) 21 Kc1 Qa1+ 22 Kd2 Qxb2 23 Ke3 Nxc2+ 24 Nxc2 Qxc2 25 Rc1! Qe4+ 26 Kf2 Qxf4+ 27 Kg1 Ne4 with equal chances; but none of this really justifies 13 h5. (a323) 13 a3 Rfd8 14 Rh3 Rac8 15 Rg3 Kh8 16 Ng5?! (16 Be2 h6) 16 ... Be8! 17 f5 exf5 18 Nxh7? Ne4! 19 Nxe4 dxe4 20 Qxa5 Nxa5=/+. (a324) 13 Rh3
13 ... Rac8 (13 ... Rab8!? 14 h5, Naer-Riazantsev, Sochi 2007; and instead of 14 ... b5?! 15 f5! exf5 16 Rg3! with an attack, 14 ... h6 is unclear; Moskalenko also analyses the pawn sacrifice 13 ... b5 14 Bxb5 Nb4; e.g. 15 Bxd7 Nxa2+! 16 Nxa2 Qxa2 17 Qb4 Nxd7 18 Nd4 a5=) 14 Kb1 Rfd8! 15 h5 h6 16 g4 Be8! 17 Nd4 Nxd4 18 Qxd4 Ne4 19 Be2 Nxc3+ 20 Rxc3 Rxc3 21 Qxc3 Qxc3 22 bxc3 Rc8 and after a fairly forced set of moves, Black stood better in Kurnosov-Vitiugov, Taganrog 2011.
(b) 8 Be2 0-0 (or the immediate 8 ... a6 9 0-0 b5, intending ... b4, when 10 a3 0-0 11 Qd2 transposes into the main line 13.31 below; here 11 dxc5!? Bxc5 12 Bxc5 Nxc5 13 Nd4 Bb7 poses Black no problems, or he can try to keep more action with 11 ... Nxc5 12 Qe1 Bb7 13 Rd1 Qc7 14 Qg3 Ne4!?, as in Kogan-Bruno, Sort 2008) 9 0-0
This will very often transpose to the main line 13.31 after 9 ... a6 10 Qd2; but Black has some other options: (b1) 9 ... b6 is solid, and should be compared with 8 Qd2 0-0 9 Be2 b6 of 13.32 below; indeed, 10 Qd2 transposes to that line. Instead, 10 Qe1!? f6!? (10 ... cxd4 11 Nxd4 Nxd4 12 Bxd4 Bc5 13 Rd1 a6 14 Qf2! Qc7 with equal play; 10 ... f5!? plans queenside action) 11 Qg3 cxd4 (11 ... f5!?; White’s queen is not ideally placed) 12 Nxd4 Nxd4 13 Bxd4 Bc5 14 Bxc5 (14 Nb5 Ba6! 15 c3 Qe7 is comfortable) 14 ... bxc5 15 Nb5 Ba6= Tomczak-Piorun, Wroclaw 2009. (b2) 9 ... a6
and now: (b21) As mentioned, 10 Qd2 is the main line 13.31. (b22) 10 a3 b5 11 dxc5 Bxc5 12 Bxc5 Nxc5 (with the idea ... d4) 13 Bd3 (13 Nd4 Bb7 14 Nxc6 Bxc6 15 Qd4 Qe7= Wei Ming) 13 ... f5 (or 13 ... Qb6!? 14 Kh1 h6) 14 exf6 Qxf6 15 Ne5 Bd7 (15 ... Bb7!) 16 Qe2 Nxe5 17 Qxe5, Bologan-Yemelin, Budva 2009; 17 ... Qxe5 18 fxe5 Rxf1+ 19 Rxf1 Rc8=. (b23) 10 Qe1!? f6! 11 exf6 (without the queen on d2, ... fxe5 is an especially strong threat) 11 ... Nxf6 12 h3!? ( ... Ng4 was threatened) 12 ... cxd4 (12 ... Nh5 13 g3 cxd4 14 Nxd4 Bc5 =) 13 Nxd4 Bd6 14 Rd1 Qc7 15 Qh4 Bd7 with at least equality; ... Ne7 and ... Rae8 can follow, contemplating ... e5. (b3) 9 ... f6!? threatens ... fxe5: 10 exf6 (10 dxc5 Bxc5 11 Bxc5 Nxc5=; 10 Qd2? fxe5 with the idea 11 Nxe5 Nxd4! or 11
fxe5? Rxf3! 12 Bxf3 cxd4, winning a second piece for the rook) 10 ... Nxf6 11 Kh1! (versus ... Ng4) 11 ... Bd6! 12 dxc5 Ng4 13 Ng5! Nxe3 14 Qd3 Rf5 15 Qxe3 Bxf4 16 Rxf4 Qxg5 17 Raf1 Bd7=. (c) 8 a3 is a clever order, recently used by elite players and suddenly accumulating a mass of theory and practice.
One point is to reply to 8 ... Qb6 with 9 Na4 Qa5+ 10 Bd2! hoping for 10 ... Qxa4?? 11 b3. Instead, the correct 10 ... Qc7 11 dxc5 Nxc5 has equalized in two high-level encounters, but I’d like to look in another direction: (c1) Black can opt for 8 ... b6with play resembling 13.32 (8 Qd2 0-0 9 Be2 b6) below; e.g. 9 Bd3!? 0-0 10 0-0 (10 Qd2 transposes to ‘c2’) 10 ... a6!? (10 ... f5 11 exf6 Nxf6!? 12 Ne5 Bb7!? 13 Qf3!+/= Morozevich-Vitiugov, Saratov 2011; here 11 ... Bxf6 is worth a try) 11 Ne2?! (a crazy move, or perhaps an oversight; White actually gets something for his piece, but really 11 Qe1 was a safe way to threaten a kingside attack or prepare the handy move Qf2, with a double-edged game ahead) 11 ... c4 12 Bxh7+ Kxh7 13 Ng5+ with a lunatic’s attack that is harder to defend against than it looks, Shirov-Morozevich, Saratov 2011; still 13 ... Bxg5 14 fxg5 Ne7 should defend. (c2) 8 ... 0-0 9 Qd2 b6 is another version of this idea:
10 g3 (10 Bd3 Bb7 11 0-0 Rc8 12 Kh1 cxd4 f6!?, or here 13 Nxd4 Nc5, should give enough counterplay) 10 ... Kh8?! (10 ... a6 11 h4 b5 is double-edged, losing a tempo on the grounds that g3 isn’t useful; 10 ... Bb7 may be best; e.g. 11 h4 Rc8 12 0-0-0 a6 13 Bd3 cxd4 14 Nxd4 b5 with a complex battle ahead) 11 h4! f6 12 exf6 Nxf6 13 Bh3!? (13 0-0-0! is better, and perhaps a reason for Black to try an earlier improvement) 13 ... Bd7 (13 ... Nh5!? has the idea 14 Ne2 Nf6, proposing repetition with 15 Nc3 Ne2; the option is 15 0-0-0 Ne4 16 Qe1, about equal) 14 Rd1? (better 14 0-0-0!+/=) 14 ... cxd4 (or 14 ... Qe8!, intending either ... Qh5 or ... Qg6) 15 Nxd4 e5!? 16 Nxc6 Bxc6 17 fxe5 d4!? (17 ... Nh5!) 18 exf6 Bxf6 19 0-0 dxc3 20 Qxd8 Raxd8 21 Rxd8
Rxd8, Ivanchuk-M.Carlsen, Sao Paulo/Bilbao 2011; and here 22 bxc3!=, with the idea 22 ... Bxc3?! 23 Rf7, was better than 22 b3 Re8 with an edge for Black. Returning to 8 Qd2 at last: 8 ... 0-0 8 ... b6 can be played in almost any order, it seems (section 13.32 is devoted to 8 ... 0-0 9 Be2 b6); here 9 Bb5 Ndb8, Benderac-Vojinovic, Pancevo 2011, intends to exchange pieces on d4 followed soon by ... Nc6. 9 Be2 We’ve reached a main line. Note that 9 dxc5 transposes to 8 dxc5 0-0 9 Qd2 (see line ‘a21’ in the notes to 8 Qd2); while 9 a3 b6 is note ‘c2’ just above. However, other moves have independent significance: (a) 9 0-0-0? is an error that has occurred many times at the master and grandmaster level. Black answers 9 ... c4! and nothing can be done about his furious queenside attack, so White is reduced to flailing away on the other wing. Probably the position isn’t lost, as some commentators have suggested, but Black will very likely come out better:
For example, 10 f5 (almost always played; White suddenly realizes that he’d better throw everything into the mix before he is overrun. 10 g4? is too slow after 10 ... b5! 11 f5 b4 12 Ne2 c3 13 Qe1 cxb2+ 14 Kxb2 Nb6 15 Ka1 exf5, winning; and 13 bxc3 Qa5! is no improvement) 10 ... b5!
(a1) 11 Nxb5 Rb8 12 Nd6 Bxd6 13 exd6 Nf6 14 c3 Ne4 15 Qc2 exf5-/+ E.Pinter-Lipka, Slovakian Team Ch. 2007. (a2) 11 fxe6fxe6 12 Nxb5 (12 h4? is too slow after 12 ... Qa5) 12 ... Rb8 13 Nd6 Bxd6 14 exd6 Nb4 (or 14 ... Nf6) 15 Kb1 Nf6 16 a3 Ne4 17 Qe1 c3! 18 b3 (18 axb4? Rxb4 19 b3 Qb6 intending ... Rxb3+) 18 ... Qa5 (18 ... Nxc2! 19 Kxc2 Rxb3! is even
more decisive) 19 Bc1 Nxc2! 20 Kxc2 Bd7! and wins, Pruijssers-Li Shilong, Dieren 2006. (a3) 11 f6! gxf6 is critical: (a31) 12 exf6 Bxf6 13 Bh6 b4 14 Ne2, and Black could play 14 ... Bg7, but can also try 14 ... c3 15 bxc3 Qa5!, when 16 Bxf8 wins the exchange, but subjects White to attack (minus the bishop pair): 16 ... Kxf8 17 Qe3 (17 Kb1?? loses to 17 ... Nc5!) 17 ... Qxa2 18 cxb4 Nxb4 19 Qc3 Be7!-/+ prepares ... Nf6 as well as the obvious queenside attack. (a32) 12 Bh6 fxe5! (12 ... b4 13 Ne2? fxe5 14 dxe5, Cabrera-P.Carlsson, Petra 2007; 14 ... Ncxe5! 15 Bxf8 Bxf8 16 Ned4 c3! 17 Qe1 Bg7 with too much pressure; but 13 Qf4! Kh8 14 exf6 keeps White in the game) 13 dxe5 Ndxe5 14 Bxf8 Qxf8!? (14 ... Bxf8!-/+; Black’s bishops and centre are extremely strong) 15 Nxb5 (15 Nxe5 Nxe5 16 Nxb5 Bd7 17 Nd4 Qg7=/+) 15 ... Qg7 (or 15 ... Rb8 – McDonald) 16 c3 Rb8 17 Nbd4 Bd7 with more than enough for the exchange; but at least White is still playing. (b) 9 g3 is seen less often these days, but still appears.
It can be met by various moves; for example, 9 ... Rb8 (9 ... b6!? is a clear solution, with the obvious idea 10 Bg2?! Ba6, but also preparing the ideal ... bxc5 if White tries dxc5; then 10 Be2 Bb7 11 0-0 Rc8 is solid; other moves are 9 ... Qa5 and 9 ... a6 10 Bg2 b5 11 0-0 Qc7) 10 a3 (10 Bh3 b5) 10 ... b6 11 Be2 f6!? 12 exf6 Nxf6, and in view of 13 0-0-0 c4 or 13 0-0 Ng4, White tried 13 Rd1 Ng4 14 Bg1 Nh6!? 15 Bd3 Bd7=/+ Bologan-T.Luther, Moscow 2007. (c) 9 Bd3 was played in the early days of 7 ... Be7, but Black has found good answers: (c1) 9 ... c4!? initiates queenside play at the risk of allowing White the f5 advance: 10 Be2 Rb8 (10 ... b5! 11 a3! Re8!? with the idea 12 f5 Nf8!) 11 f5!? b5 12 a3 Nb6! 13 fxe6 (13 f6!?) 13 ... Bxe6= Van Delft-M.Gurevich, Hoogeveen 2006. (c2) 9 ... f6! 10 exf6 Bxf6
11 Be2 (11 Bb5 contests the dark squares, but White may not want to exchange this bishop; e.g. 11 ... cxd4! 12 Nxd4 Bxd4 13 Bxd4 a6 14 Bxc6 bxc6=/+ intending ... c5 or in some cases ... e5) 11 ... cxd4 12 Nxd4 Bxd4 13 Bxd4 e5! 14 Bxe5 (14 fxe5?? Qh4+) 14 ... Ndxe5 15 fxe5 Be6 (15 ... d4! 16 Nb5 Be6 cuts down White’s options, because 16 Ne4?! Qh4+ favours Black) 16 0-0-0 d4 17 Nb5 (17 Ne4 Qd5 18 Ng5 Qxa2 leads to a draw: 19 Qd3 Qa1+ 20 Kd2 Qa5+ 21 Kc1 Bf5 22 Qb3+ Kh8 23 Nf7+ Rxf7 24 Qxf7 Qa1+ 25 Kd2 Qa5+ 26 Kc1 Qa1+ 27 Kd2 Qa5+ ½-½ Eliseev-Vysochin, Peterhof 2007; but 15 ... d4! avoids this) 17 ... Qd5 18 b3 (18 c4! Qxe5 19 Rhe1 Qc5 20 Nxd4 ½-½ Efimenko-Goloshchapov, German League 2007; there might follow 20 ... Bxc4 21 Qc3 Qg5+ 22 Qd2 Qxd2+ 23 Rxd2 Bxa2 24 b3 Nxd4 25 Bc4+ Kh8 26 Rxd4 b5 with unclear complications) 18 ... Qxe5 19 Bd3 Qc5 (19 ... a6! and 20 Rde1?! Qc5 or 20 Na3 b5 looks good, though it hasn’t been played for some reason) 20 Kb1! (20 Kb2 Rac8 21 Rhe1 Bg4 22 Rc1 a6 23 Na3 b5=/+ Shaposhnikov-Kirusha, St Petersburg 2008) 20 ... Rad8, Shyam-Sethuraman, New Delhi 2010; and here 21 Rhf1!= was best. After 9 Be2,
we 13.31 13.32 13.31
arrive
at
the
main
lines
of
this
9 9 9
Classical
Steinitz
... ... ...
a6
10
variation: a6 b6 0-0
10 a3 b5 11 Nd1?! doesn’t gain a tempo for reorganization because of 11 ... f6!; e.g. 12 c3 fxe5 13 dxe5 (13 fxe5 Nb6! with the idea 14 dxc5 Nc4 15 Bxc4 bxc4=/+) 13 ... Na5 14 Qc2 c4=/+. 10 ... b5
Readers may recognize this as similar to 7 ... a6 and other early ... a6 lines. The difference is that Black has gotten castled and not committed his other pieces. This gives him a lot of flexibility with the placement of his queen and queen’s bishop. This variation is very largely a matter of positional decisions, and there are a huge number of possible lines based upon the choices and timing of these moves among others: (a) ... Qc7 or ... Qb6; (b) ... f6, or ... f5; (c) ... Bb7, ... Bd7, or ... a5 with ... Ba6; (d) ... cxd4 or ... c4; and (e) every conceivable combination of rook placements along the bank rank. To make matters much worse, White is playing his own unique set-ups; I have paid the most attention to lines with a3 and/or Nd1, Kh1, f5, and dxc5, but obviously I can only give a small selection of reasonable possibilities. Therefore, in this section of the book more than any other, I want to emphasize that, if doubt is cast upon one or another line I mention, you can adjust your repertoire without much effort and be confident in the basic ideas. To find out which lines are best and/or most convenient, examining recent games of the leading players will be extremely useful. There is no shortage of such games in current practice and, if dissatisfied, I suspect that the respectable variations with an early ... a6 (for example), will always be a decent alternative and involve similar ideas. 11 a3
This is currently considered White’s best approach from this position, and it has been played more than any other move by leading grandmasters. The main idea is to hold up ... b4; then White would like either to capture on c5 to occupy d4, or attack on the kingside, usually by the move f5. Black has various options, sometimes in combination: (a) develop his queenside with ... Qb6/ ... Qc7 and ... Bb7, connecting rooks; (b) expand on the queenside with ... b4, ... a5-a4 and ... Ba6; (c) play ... f6 and a timely ... fxe5, or play ... f5
(which seems to me an underrated strategy); (d) liquidate with ... cxd4 and, if White recaptures with a piece, follow with moves such as ... Bc5 and/or ... Nc5-e4. Not surprisingly, White has several logical alternatives to 11 a3. One drawback is that they tend to allow ... b4: (a) 11 dxc5 Bxc5 12 Bxc5 Nxc5 13 Kh1 (for 13 a3 Qb6 or 13 ... Bb7, compare 11 a3) 13 ... b4 14 Nd1 a5 15 Nf2 Qb6 and ... Ba6, with more space on the queenside and good prospects for Black. (b) 11 Rad1 b4 12 Na4 cxd4 13 Nxd4 Nxd4 14 Bxd4 Qa5 15 b3 Nc5 16 Nxc5 Bxc5 17 a4 Qb6 (17 ... Bd7=) 18 Qe3 Bxd4 19 Qxd4 Qxd4+ 20 Rxd4 a5 21 Rf3 Ba6 22 Bd1 Rac8 23 Kf2 Rc5 24 Ke3 f6 25 Rd2, Miladinovic-Sedak, Mataruska Banja 2007, and here Anka suggests 25 ... d4+ (25 ... f5!?) 26 Kxd4 Rfc8 27 Ke3 Bb7 28 Rff2 fxe5=/+. (c) 11 Kh1 is a waiting move designed to make Black commit his queen or bishop:
(c1) 11 ... Qb6 gives a good impression: 12 f5 (for 12 a3, see 11 a3 Qb6 12 Kh1 in 13.312) 12 ... cxd4 13 Nxd4 Bg5! (Anka), intending 14 f6 (14 Rad1 Bxe3 15 Qxe3 Bb7! with the idea 16 f6 Ncxe5 17 Qg5 Ng6 18 fxg7 Kxg7=/+) 14 ... Qxd4! and, if anything, Black stands better. (c2) 11 ... Qc7 introduced a famous game: 12 a3 Bb7 13 Rad1 Rac8 14 Qe1 (14 dxc5 Bxc5 15 Nd4 Nxd4 16 Bxd4 Bxd4 17 Qxd4 Qc5=) 14 ... cxd4 15 Nxd4 Nxd4 16 Bxd4 Bc5 17 Qh4 Bxd4 18 Rxd4 f6!
19 Bd3?! (but 19 exf6 Rxf6 is fine for Black) 19 ... h6! 20 exf6 Rxf6 21 f5 (21 Qg4 Rcf8 22 Rf3 Nb8! and 21 Qg3 Rcf8 22 Rf3 Nb8! both have the idea ... Nc6 and clearly favour Black) 21 ... Rcf8 22 Rg1 Nc5! 23 fxe6 Nxe6 24 Rg4 (24 Nxd5 Bxd5 25 Rxd5 Nf4 26 Rd4 Qc5, and a cute line is 27 c3?? Nxd3 28 Rxd3 Qxg1+ 29 Kxg1 Rf1 mate) 24 ... Nf4 25 Qg3 Qe7 26 Rxf4 ( ... d4 was threatened) 26 ... Rxf4 with a winning game, Karjakin-M.Carlsen, Wijk aan Zee 2010.
(d) 11 Nd1 used to be very popular, but it makes it easier for Black to progress on the queenside: 11 ... b4 (11 ... cxd4 12 Nxd4 Nxd4 13 Bxd4 Nc5 is another respectable line) 12 Nf2 (12 c3 a5 13 dxc5, Zelcic-Riazantsev, Dresden 2007, and here 13 ... Bxc5! equalizes easily after 14 Bxc5 Nxc5 15 Qe3 Qb6) 12 ... Qb6 (12 ... a5!? 13 dxc5 Bxc5 14 Bxc5 Nxc5 15 Bd3 Ba6 16 Qe3 Bxd3 17 cxd3 Na4 18 Rab1 Ne7 19 Rfc1 Qd7= Kamsky-Morozevich, Amsterdam 1996; here 14 c4 might be worth a try; I like 12 ... c4 best, intending ... a5-a4) 13 c3 a5 14 g4?! (less ambitious but safer for White was 14 Rad1 Ba6 15 f5 exf5 16 Bxa6 Qxa6, ½-½ Hermansson-P.Carlsson, Malmö 2007) 14 ... Ba6 15 Nd3 (15 Bxa6 Qxa6 16 f5 bxc3 17 bxc3 Nb6 and ... Nc4) 15 ... bxc3 16 bxc3 cxd4 17 cxd4 Nb4!? (17 ... Rab8=/+; 17 ... Bb4) 18 f5 Rac8 19 Nfe1 Bg5! 20 Nxb4 Bxe3+ 21 Qxe3 Bxe2 22 Rf4 (22 Qxe2 Qxd4+ 23 Kh1 Qxa1) 22 ... Qxb4 23 Qxe2 g5! 24 fxg6 fxg6 25 Rxf8+ Rxf8 26 Qe3 Nb6-/+ Timofeev-Wang Hao, Taiyuan 2006. After 11 a3, Black has a varied choice of set-ups. I’ll concentrate upon: 13.311 13.312
11 11
... ...
Bb7 Qb6
Naturally, 11 ... Qc7 is also fully playable, with the idea 12 Nd1 f6!; in this variation move orders are important, as always, but less so than strategic ideas. 13.31111
...
Bb7
Black foregoes the more aggressive placement resulting from ... b4/ ... a5/ ... Ba6. On the other hand, the latter is timeconsuming, and with 11 ... Bb7 Black can develop quickly and connect rooks, while preserving the choice between ... Qb6 and ... Qc7. 12 Nd1 The standard move. Nevertheless, others deserve consideration: (a) 12 dxc5 Nxc5 (I like 12 ... Bxc5 better) 13 Rad1 Qc7 14 Qe1!? (heading for the kingside) 14 ... Rac8 15 Qg3 Rfd8 (15 ... b4 16 axb4 Ne4! is also satisfactory) 16 f5! exf5 (16 ... Ne4 17 Nxe4 dxe4 is also equal) 17 Nxd5 Rxd5! 18 Rxd5 Ne4 19 Qf4 Nb4! 20 axb4 Bxd5 21 Qxf5, Corrales Jimenez-Granda Zuniga, Sao Paulo 2009; and here 21 ... Bxb4! would have ensured equal chances. (b) 12 Rad1 Rc8 (or 12 ... Qb6) 13 Kh1 cxd4 (13 ... Qb6 14 f5!? cxd4 15 Nxd4 Nxd4 16 Bxd4 Bc5 is roughly equal) 14 Nxd4 Nxd4 15 Bxd4 Bc5 16 Qe3, Vitiugov-Maslak, Novokuznetsk 2008; and Vitiugov suggests 16 ... Bxd4 17 Rxd4 Qb6 18 Bd3 f6!. (c) 12 f5 Qb6 13 f6 gxf6 14 exf6 Nxf6 15 Bh6 (Browne) 15 ... Kh8! 16 Bxf8 Rxf8 17 dxc5 Qxc5+ 18 Kh1 d4 19 Nd1 Ne4 20 Qd3 f5 (or 20 ... Qd5 21 Nf2 f5) 21 Nf2 e5! 22 Nxe4 fxe4 23 Qxe4 Rf4 24 Qd3 e4 25 Qd2 Bd6 and if the knight moves, 26 ... Rh4. (d) 12 Bd3 Rc8 (or 12 ... f6=) 13 Qf2 cxd4!? (13 ... b4 14 axb4 Nxb4) 14 Nxd4 Nxd4 (14 ... Na5!?) 15 Bxd4 Nc5 16 f5 exf5 17 Bxf5 Ne6 18 Bb6 Qxb6 19 Qxb6 Bc5+ 20 Qxc5 Rxc5 is level; e.g. 21 Ne2 d4!? 22 Bxe6 fxe6 23 Nxd4 Rxe5 24 Rfe1 Rxe1+ 25 Rxe1 Bd5. 12 ... cxd4 Alternatively: (a) 12 ... Rc8 is also good for equality: 13 c3 Na5 14 Qe1 Nc4 15 Bd3 (15 Bf2 cxd4 16 cxd4 Qb6 17 Bd3 f5! slows any potential kingside attack by White, and 18 exf6 Nxf6 19 Bh4 Rc7 prepares ... Ne4) 15 ... f5 (15 ... cxd4 16 cxd4 f5 is more accurate, with the idea 17 exf6 Nxf6 18 Ng5 Qd6 19 f5 Ng4 20 Bf4 e5 21 Qg3! exf4 22 Qxg4 Qf6 23 Ne6 Bd6!=) 16 Bf2 (16 exf6 Nxf6 17 a4! initiates some queenside play) 16 ... cxd4 17 cxd4, Zherebukh-Bajarani, Kirishi 2010; and 17 ... Nb8! 18 a4 Nc6 would establish at least equality due to Black’s queenside pressure.
(b) 12 ... Qb6 resembles 11 ... Qb6 of the following section (13.312).
Then one interesting example went 13 Kh1 a5 14 c3 b4 15 a4 (15 Nf2 Ba6 16 Nd3!? bxc3 17 bxc3 cxd4 18 Nxd4 Nxd4 19 cxd4 Rab8 20 f5 exf5 21 Rxf5 Bc4 22 a4 Qa6= intending ... Rb3) 15 ... Ba6 16 Bg1?! (16 Nf2 Bxe2 17 Qxe2 Rab8 18 Rfc1 Rfc8 19 Nd3 Qa6=/+) 16 ... bxc3 17 bxc3 cxd4 18 cxd4 Nb4 (or 18 ... Bb4 19 Qa2 Bxe2 20 Qxe2 Ne7!, versus f5) 19 Ne3 f5!=/+ KurnosovSwiercz, Lublin 2010. 13 Nxd4 Qc7 14 Nf2 14 Nxc6 Qxc6 15 Bd4 Rac8 is dynamically balanced; e.g. 16 Rc1 (16 Ne3 Nc5) 16 ... Nb6 17 f5?! (17 Ne3 Bc5=) 17 ... Nc4 18 Bxc4 dxc4 19 f6 Bc5! 20 c3 Qe4! 21 Re1 Qg6 22 Re3 Qg5=/+. 14 ... Rac8 15 c3 Nc5 Or 15 ... Na5=. 16 Qc2 16 Qe1 was played in Hou Yifan-Ding Liren, Chinese Ch. 2009. After 16 ... Nxd4 17 cxd4, 17 ... Nb3 is perhaps more interesting (than 17 ... Ne4); e.g. 18 Rb1 Qb6 19 Qd1 Na5 20 Nd3 Nc4 21 Bf2 Rc7 22 Bg4 a5 23 f5 Bc8! 24 fxe6 Bxe6=/+. 16 ... Nb8!? After 16 ... h6, neither side has a compelling threat. Black can try ... Na4 and/or ... Nd7-b6, but may end up settling for simply ... Nxd4. 17 Rac1 Nbd7= Aiming for ... Nb6 and either ... Ne4 or ... Na4. 13.312
11
...
Qb6
12 Nd1 White makes this temporary retreat, also the most popular move, with two principle ideas: to reroute the knight to f2 or e3 in support of a kingside attack, and to shore up the centre if needed by c3. Naturally any backwards move opens up possibilities for the opponent, and Black will try to implement his ideas forthwith. There are a few other important set-ups: (a) 12 Kh1 Bb7 (or 12 ... cxd4 13 Nxd4 Bc5 14 Rad1 Bb7 15 Bf3 Nxd4 16 Bxd4 Bxd4 17 Qxd4 Rac8 and if White retreats the queen, Black can play ... Qc5 and ... Nb6; if not, ... Rc7 and ... Rfc8 is a good plan) 13 Rad1 (13 Bg1 cxd4 14 Nxd4 Nxd4 15 Bxd4 Bc5 16 Bxc5 Nxc5= 17 Rad1 Ne4 18 Nxe4 dxe4 19 Qd4 Qc7! 20 c3 Rfd8= Sadvakasov-Ni Hua, Moscow 2007) 13 ... cxd4 14 Nxd4 Nxd4 15 Bxd4 Bc5 16 Qe3 Bxd4 17 Qxd4 Qxd4 18 Rxd4 f6!? (18 ... Rfc8=) 19 exf6 gxf6 20 Bg4 f5 21 Bf3 Rac8 22 Rc1 Nc5= Fedorov-Kayumov, Tashkent 2008. (b) 12 Rad1!? lends support to d4, although the rook doesn’t have much effect on the central and queenside struggle after 12 ... Bb7
(b1) 13 dxc5 Bxc5 14 Bxc5 Nxc5 (or 14 ... Qxc5=) 15 Qe3 Rac8 16 Bd3 (16 f5!? Na4 17 Qxb6 Nxb6 18 Bd3 Nd7 19 fxe6 fxe6 20 Ng5 Rxf1+ 21 Rxf1 Nf8=) 16 ... Na4 (16 ... b4=) 17 Qxb6 Nxb6 18 Ne2, Mekhitarian-Rodriguez Vila, Campinas 2010, and here 18 ... Na4! 19 b3 Nc5 20 Nfd4 g6 equalizes. (b2) 13 Rfe1 cxd4 14 Nxd4 Bc5 15 Bg4!? Bxd4 16 Bxd4 Nxd4 17 Qxd4 g6!? 18 Ne2 a5 (expanding and threatening to free his bishop after ... b4 and ... Ba6) 19 c3 Rfc8 20 Qxb6 Nxb6 21 Nd4 Bc6 22 Ra1 Bd7 23 Kf2 Na4 24 Ra2 Rab8 25 Bd1 Nc5=
K.Szabo-Manca, Budapest 2009. (c) 12 dxc5 Bxc5 13 Bxc5 Nxc5 gives Black easy equality; e.g. 14 Kh1 (14 Qe3 Bb7 15 b4 Na4=) 14 ... d4! 15 Nd1 (15 Nxd4? Rd8 16 b4 Nb7-/+) 15 ... Rd8 16 Nf2 Ne7! 17 b4 Na4 18 Bd3 Nf5!=/+ Wang Li-Gleizerov, Dubai 2008.
12 ... a5 It’s important to note that 12 ... Bb7 transposes to 11 ... Bb7 12 Nd1 Qb6, which comes out satisfactorily for Black. 13 c3 White prepares Bd3 at the cost of weakening b3. 13 ... a4! This sets up ... Na5-c4 and preserves the possibility of ... b4. When Black plays ... Ba6 without inserting this move, it can often happen that White will play b4 and stifle the bishop. Svidler-Hou Yifan, Amsterdam 2009, continued cautiously with 13 ... Bb7 14 Bf2!? (14 Bd3 is more promising, intending f5), and instead of 14 ... f6?! 15 exf6! Nxf6 16 dxc5 Bxc5 17 Bxc5 Qxc5+ 18 Qe3! with an edge, Black can hold his own with 14 ... b4 or 14 ... a4 15 Ne3 Na5 16 Rad1 Nc4. 14 Bd3 Natural, finally aiming for the attack with f5; this is White’s best plan in general, although the timing is critical. The alternative 14 Nf2 can be met by 14 ... Na5;
e.g. 15 dxc5 Bxc5 16 Bxc5 Nxc5 17 Qe3 Nc4 18 Bxc4 bxc4 19 Rae1 Rb8= Kosintseva-Xu Yuhua, Ningpo 2008. 14 ... Na5 14 ... cxd4 15 Nxd4!? Nc5 16 Nf2 Na5! was at least equal in Motylev-Yemelin, Ulan Ude 2009, with the idea 17 Bxb5?!
Ba6! 18 Bxa6 Qxa6, winning all the light squares. White should play 15 cxd4, when 15 ... Na5!? creates an equal, tension-filled position.
15 Qe2 15 Qf2 f5!? 16 exf6 Nxf6 17 dxc5 Qc7 was unclear in Xu Yuhua-Xiao Zhang, Jiangsu Wuxi 2008. 15 ... Nc4 A natural move, although 15 ... Nb3! 16 Rb1 Ba6 intending ... b4 is a smoother solution; e.g. 17 dxc5! Ndxc5 18 Nf2 Qc6=. 16 Bf2! f6 This holds up White’s intended f5 and creates kingside counterplay. 17 exf6 Nxf6 18 dxc5 Bxc5 19 Bxc5 19 Bd4 is more testing, after which 19 ... Bd7 20 Nf2 Bxd4 21 Nxd4 Rae8 can follow. 19 ... Qxc5+ Djukic-Kosic, Cetinje 2009; here, instead of 20 Nd4?, Wei Ming suggests 20 Nf2! Bd7= followed by ... Rae8. 13.329
...
b6
This is a logical and recently popular way of treating the position; Moskalenko, as so often, is a chief agitator for it. The most basic ideas behind the move are these: (a) if, as in previous lines, White plays dxc5(?), then ... bxc5 takes over the centre and even threatens ... d4; (b) when Black plays ... f6, he threatens ... fxe5 followed by ... cxd4, winning a pawn. In earlier lines,
we saw White respond to ... f6 with exf6 and dxc5, which isn’t normally effective here since Black replies ... bxc5. It turns out that Black has other ideas as well, such as playing ... f5, ... cxd4, and ... Nc5-e4; or even ... a5 and ... Ba6. This flexibility serves him well. For the same idea in less-tested contexts, see the notes on 8 a3 b6 and 8 ... 0-0 9 a3 b6 above. 10 0-0 Instead: (a) As was the case after 9 0-0-0, here 10 0-0-0 might again be met by 10 ... c4 (or the calmer 10 ... Bb7= intending ... Rc8) 11 f5 b5! 12 Nxb5 (12 f6 gxf6 13 Bh6 fxe5) 12 ... Rb8 with a powerful attack. (b) 10 Nd1 prepares c3 or c4. 10 ... a5 and ... Ba6 is one approach, in addition to:
(b1) 10 ... cxd4 11 Nxd4 Nxd4!? (11 ... Bb7 12 0-0 Nc5= transposes to 10 ... Bb7 in note ‘b’ to the main line with 10 ... f6!? below) 12 Bxd4 Nb8!? (12 ... Nc5; 12 ... Bb7 13 0-0 Rc8=) 13 0-0 Nc6 14 c3, Dembo-Sedina, Rogaska Slatina 2011; White has a small positional edge following 14 ... Bb7 15 Ne3 Nxd4 16 cxd4 Rc8, although it’s hard to do anything with it. (b2) 10 ... f5 is an important interpretation of the ... b6 lines; one point is that after ... cxd4 and a white piece recapture, Black can play ... Nc5-e4; e.g. 11 c4!? (after 11 c3, Black occupies the c-file anticipating queenside play: 11 ... Bb7 12 0-0 Rc8 13 Ng5 Bxg5 14 fxg5 Qe7=; 11 0-0 transposes to 10 0-0 f5 11 Nd1 below) 11 ... cxd4! (11 ... dxc4? 12 d5) 12 Nxd4 Nxd4 13 Bxd4 dxc4! 14 Bxc4 Nc5
15 Nc3 (15 b4? Ne4 16 Qb2 a5!-/+ – Kritz) 15 ... Qc7, Motylev-Vitiugov, Poikovsky 2010; and instead of 16 0-0-0?!, White should play 16 0-0; e.g. 16 ... Rd8 (16 ... Ba6) 17 Rac1 Ba6! 18 Bxa6 Nxa6 19 Nd5 Qd7 20 Nxe7+ Qxe7 21 Rfd1 (Kritz) 21 ... Nc7! and Black transfers to the excellent outpost on d5.
10 ... f6!? Thematic and most often played, though Black has other good ideas here: (a) 10 ... f5 is a solid move, discouraging White’s kingside aspirations: 11 Nd1 (11 dxc5 Nxc5 12 Nb5 Ne4 13 Qd1 Bc5= Paikidze-Vorobiov, St Petersburg 2010; 11 exf6 transposes into the main line) 11 ... cxd4 12 Nxd4 Nxd4 13 Bxd4 Nc5 14 Ne3 (14 Nf2 Qd7 15 Kh1 Bb7=) 14 ... Ne4 15 Qe1 Bd7 (15 ... Bb7!) 16 c4 dxc4 17 Bxc4 Kh8 18 Rd1 Qe8 19 Qe2 Rd8 20 a3 Qg6 21 Rd3 Bc6= Bobras-Naer, German League 2010. (b) 10 ... Bb7 (‘also playable and attractive’ – Moskalenko) develops and awaits White’s reply; by protecting c6, Black makes ... f6 sounder (that is, White’s Nxd5 won’t work; see the next note on 11 dxc5); e.g. 11 Nd1 (11 Kh1 f6) 11 ... cxd4 12 Nxd4 Nc5 (12 ... Nxd4 13 Bxd4 Nb8 14 c3 Nc6 15 Bf2 f6= Nijboer-Narciso Dublan, Barcelona 2007) 13 Nxc6 Bxc6 14 Bd4 Ne4 15 Qe3 Bc5 (15 ... f5!? is more dynamic – McDonald) 16 Nf2 f5 (16 ... Bxd4! 17 Qxd4 Qe7=) 17 Nxe4 Bxd4 18 Nf6+! (18 Qxd4 fxe4) 18 ... Qxf6 19 Qxd4 Qe7 20 Rad1 Rac8, Kokarev-Volkov, Plovdiv 2008. This appears roughly equal. 11 exf6 11 dxc5 Nxc5 (11 ... bxc5? 12 Nxd5!) 12 Bb5 Bb7 13 exf6 Bxf6 14 Bxc6 Bxc6 15 Bd4 Rc8 16 Rad1 Be8!? 17 Bxf6 Rxf6 18 Ne5 b5 19 a3 Qb6=. 11 ... Nxf6 12 Bb5 Trying to strengthen his hold over the e5-square. Alternatively, 12 Kh1 Bb7 (12 ... Qc7 13 Ne5 a6! 14 Nxc6 Qxc6 15 Qe1 Bb7 16 Bf3 Rae8 17 Rd1 Bd6 18 Qh4 Qc7 19 Ne2, Bazeev-Iskusnyh, St Petersburg 2010, and 19 ... Ne4 was fully equal)
(a) 13 Rad1 Ne4 (13 ... Qc7=) 14 Nxe4 dxe4 15 Ne5 Nxd4 16 Bxd4 cxd4 17 Bc4 Bf6! 18 Bxe6+ Kh8 19 Nf7+ Rxf7 20 Bxf7 Qc7 with obvious compensation in the form of the bishops and centre pawns, especially since 21 Bb3?! (21 Bh5 Rc8 22 Rc1 Qc4) 21 ... Ba6 will win the exchange back. (b) 13 Bd3 is more aggressive: 13 ... a6 (‘Black intends the central set-up ... Qc7/ ... Qd6, ... Rae8, followed by ... Ng4, so White cannot afford to play casually’ – Kosintseva) 14 a3?! (14 Rae1 c4!? 15 Be2 Bb4; 14 Bg1 Qd6=) 14 ... Qc7 15 Qe1 (15 Rae1 Bd6 16 g3 Rae8=/+), Kosintseva-I.Rajlich, Novi Sad 2009; and instead of releasing the central tension by 15 ... c4, Kosintseva suggests 15 ... Rae8, when there might follow 16 Qh4 cxd4 (16 ... Bd6; 16 ... g6!?) 17 Nxd4 Ne4 18 Qh3 Nxd4 19 Bxd4 Bf6=/+. 12 ... Qc7 White can take over e5 after 12 ... Bb7 13 Bxc6 Bxc6 14 Ne5, but it can be challenged; e.g. 14 ... Be8 (or 14 ... Qc7 15 Rad1 Nd7 16 Nxc6 Qxc6= with ... Nf6 to come) 15 Rae1 Rc8 16 a3 (16 Qe2 b5!) 16 ... Bd6 17 Bf2 Bb8 18 Bh4 Qd6=. 13 Kh1 Bd7!? 13 ... a6 14 Bxc6 Qxc6 is straightforward. 14 Rad1 a6 15 Bxc6 Bxc6 16 Qe2 b5 Or 16 ... Bd6 17 Bg1 Rae8 with the idea 18 Qxa6? c4!. 17 Bg1 Qc8?! Here 17 ... Bd7 is fine, since 18 Ng5 Qc6 19 Rde1 Rae8 secures the centre and can be followed by a queenside advance. After 17 ... Qc8?!, White could get a limited edge with 18 Rfe1. Instead, in Baramidze-Berczes, Biel 2008, he blundered by 18 dxc5? b4 19 Rfe1 (in view of 19 Nb1 Bb5-/+) 19 ... bxc3 20 Qxe6+ Rf7 21 Ng5 Qxe6 22 Rxe6 Bd7 and Black went on to win.
Chapter
Fourteen
MacCutcheon:
Introduction
1 e4 e6 2 d4 d5 3 Nc3 Nf6 4 Bg5 Bb4
This is the MacCutcheon Variation. In the last edition, the solution I gave to 4 Bg5 was based upon ... dxe4 (i.e. 4 ... dxe4 5 Nxe4 Be7), which is fine but not in the style of the other systems I have recommended. That is, in the main lines of the main systems of this book, Black has always held the d5 strongpoint until White decided upon e5, setting up a traditional pawn chain. That is also the case with the MacCutcheon. It is considered quite dynamic, and resembles the Winawer Variation in several respects, not the least of which are the doubled pawns White usually incurs when Black captures on c3 and surrenders his bishop pair. Moreover, in White’s main line with Qg4, Black has to defend g7 and choose between active and passive play in so doing. The resulting positions are rich and varied, with the bonus that Black can choose between wide-open play (e.g. lines based upon ... cxd4) and a complex positional struggle (lines with ... c4). Sometimes White can make this decision for him by opening the position with dxc5, a move used much more often here than in the Winawer. In the main line, there are early piece exchanges, but the imbalance of forces and structure keeps the positions extremely complex in spite of such formal simplification. As usual, all this is best shown by examining the variations themselves: 14.1 14.2 14.3
5 5 5
exd5 Ne2 e5
Others: (a) Black has the bishop pair and no problems after 5 Bxf6 Qxf6 6 e5 Qd8 intending ... c5, or here 6 ... Qe7. (b) 5 Bd3 is well answered by 5 ... dxe4 (or 5 ... c5) 6 Bxe4 Nbd7 (or 6 ... h6 7 Bxf6 Qxf6 8 Nf3 0-0 9 0-0 Nd7=) 7 Ne2 h6 8 Bxf6 Nxf6 9 Bf3 c5= Gipslis-Bronstein, USSR Ch., Baku 1961. 14.1 Here 14.11 14.12 14.11
5 Black
exd5 has 6 6 6
Qxd5
6 satisfactory
two ... ... ...
Bxf6 moves: gxf6 Bxc3+ gxf6
This goes with moves like ... Rg8, ... e5, and/or ... b6 to achieve rapid development. 7 Ne2 7 Qd2 Qa5 (7 ... Bxc3 8 Qxc3 Nc6 9 Nf3 Rg8 is also satisfactory) 8 Ne2 Nd7 9 a3 Nb6 has been played often; e.g. 10 Rd1 Be7 11 b4!? (11 Nf4 Bd7 12 Be2 0-0-0) 11 ... Qg5 12 f4 Qh6 (or 12 ... Qh4+ 13 g3 Qh5=) 13 Nc1 (13 Nb5 Kd8! intending ... c6/ ... Kc7) 13 ... Bd7 14 Nb3 Rg8 and Black had an edge in Ramsauer-Behrends, corr Germany 1994. 7 ... Nc6
8 a3 8 Qd2 Bxc3 9 Nxc3 (9 Qxc3 Bd7 10 Qb3 Qa5+ 11 c3 0-0-0 is more than adequate for Black) 9 ... Qxd4 10 Qxd4 Nxd4 11 00-0 c5 12 Ne4 b6 13 c3 Nf5 14 g4 Nh4 15 Nxf6+ Ke7 16 g5 h6 17 h4 Rb8=; compare the main line. After 8 a3, we enter a lengthy forcing sequence: 8 ... Bxc3+ 9 Nxc3 Qxd4 10 Qxd4 Nxd4 11 0-0-0 c5! 11 ... Nc6 12 Nb5 Ke7 13 Nxc7 Rb8 14 Bb5!+/= with the idea 14 ... a6? 15 Nxa6. 12 Ne4 b6 13 c3
13 Nxf6+ Ke7 and ... Bb7 can only favour Black. 13 ... Nf5 Black can also play 13 ... Nb3+ (note that this isn’t available after 8 Qd2 Bxc3 9 Nxc3 Qxd4 etc) 14 Kc2 Na5 15 b4 Nb7! 16 Nxf6+ (16 g4 Ke7) 16 ... Ke7 17 Ng4 (or 17 Ne4 Bd7) 17 ... Bd7, which has equalized in several games. 14 g4 Nh4 15 Nxf6+ Ke7 16 g5 h6 17 f4 Rb8! With the idea ... Ng6 and in some cases ... hxg5. 18 Bb5 a6! 19 Bc6 Ng6
This position is balanced and offers equal chances. 20 f5!? 20 Rhf1 is objectively better: 20 ... hxg5 21 fxg5 Ne5 22 Bh1 (22 Be4 Rxh2 23 Ng8+ Ke8=) 22 ... c4 (22 ... Rxh2 23 Ng8+ Ke8 24 Rfe1) 23 Rd2 Bb7 24 Bxb7 Rxb7=. 20 ... exf5 21 Rhe1+ Be6 22 Rd7+ Kf8 Black has a slight edge, with a more serious advantage if White tries 23 Ra7? (23 gxh6?! Ne7; 23 Bd5! Bxd7 24 Nxd7+ Kg7 25 gxh6+ Kxh6 26 Nxb8 Rxb8 27 Bxf7 Nf4=/+) 23 ... hxg5! 24 Rxe6 fxe6 25 Nd7+ Kf7 26 Nxb8+ Ne7 27 Nxa6 g4!. 14.12 6 ... Bxc3+ In this line the opponents have weaknesses and open files of about equal worth.
7
bxc3
gxf6
8 Qd2 Instead: (a) After 8 Nf3, Black has some leeway, but he usually fianchettoes; e.g. 8 ... b6 9 Be2 Bb7 10 0-0 Rg8 has worked out well in numerous games. Alternatively, 8 ... Bd7 is interesting, with the idea 9 Be2 Bb5. (b) 8 Qg4 Qa5 9 Ne2 Ke7!? (9 ... Bd7! is preferable, with the idea 10 Qg7 Rf8 11 Qxf6 Bc6 followed by ... Nd7, after which White will be well behind in development and has difficulties getting his pieces untangled) 10 Qe4 (or 10 g3) 10 ... Bd7 11 Qxb7 (11 Qe3! Bc6 12 Nf4 intending Bc4 is better) 11 ... Nc6 12 Qb3 Rab8 13 Qc4 Rb2 with a nice initiative, Brendel-Glek, Swiss Team Ch. 2004.
8 ... Nc6 Again Black has ideas of ... Rg8, ... e5, and/or ... b6 with rapid development. Alternatively, the natural 8 ... c5 is often played here, as is 8 ... Nd7 (9 c4 Qc6! 10 Nf3 Rg8). Finally, 8 ... Qa5 is a safe choice; e.g. 9 Bd3 (9 g3 Qd5!? 10 f3 Nc6; 9 Ne2 b6 10 Qf4?! Nd7!, Kovalevskaya-Volkov, Togliatti 2003, with the idea 11 Qxc7 Ba6) 9 ... Bd7 10 Ne2 Bc6 11 Nf4 (11 f3 Nd7 12 0-0 00-0 13 Be4, D.King-N.McDonald, British League 2009; and here Wei Ming suggests 13 ... Bxe4 14 fxe4 Rdg8 intending to double rooks and attack) 11 ... e5 12 Nh5 (12 dxe5 fxe5 13 Nh5 Nd7 14 Bf5 0-0-0 15 Nf6 Qc5 16 0-0 Qe7 17 Nd5 Bxd5 18 Qxd5 Kb8 19 Rfe1 f6 intending for the knight to take up permanent residence on c5) 12 ... Nd7 13 Bf5 0-0-0 14 c4 (14 Nxf6? exd4, hitting f5) 14 ... Qxd2+ 15 Kxd2 exd4 16 Nxf6 h5= intending ... Rh6.
9 Qf4 9 Nf3 (after 9 Ne2 e5!? 10 Qh6 Ke7 11 dxe5 Qxe5=, Black’s free development compensates for his structure) 9 ... Rg8 10
Qd3 Bd7 11 Rb1 0-0-0 12 Qa6!? Na5!? 13 Qxa7 Bc6=/+ intending ... Qxa2 or ... b6. 9 ... Qa5 10 Ne2 10 Kd2 f5 11 Nf3 Bd7=/+ and ... 0-0-0. 10 ... Ke7!? Or 10 ... f5= intending ... b6. 11 Qd2 Here 11 Qd2 (11 Rd1 Rd8 12 Qe4 Qxa2 13 Qxh7 e5=/+ Feldmann-Canovas Pordomingo, IECG 2003) 11 ... b6! 12 Ng3 Bb7 is equal, since White will have to go through contortions to get castled after ... Rad8. 14.2 5 Ne2
This position, which can also arise from the Winawer line 3 Nc3 Bb4 4 Ne2 Nf6 (see 7.42) 5 Bg5, tries for a small positional edge. Black has two good ways to tackle it: 14.21 14.22
5 5
... ...
dxe4 h6
14.21 5 ... dxe4 6 a3 Be7 7 Bxf6Bxf6 Black can also play 7 ... gxf6 8 Nxe4 f5 (8 ... b6 is possible and theoretically okay, but somewhat passive after 9 g3 Bb7 10 Bg2 c6) 9 N4c3 Nd7 10 Qd2 Nf6 11 0-0-0 Qd6=. 8 Nxe4
8 ... 0-0 Here Black can play solidly with 8 ... b6, when 9 g3 Bb7 10 Bg2 Nc6 or 10 ... c6 is playable; or else 8 ... Nd7, e.g. 9 Qd2 0-0 10 0-0-0 Be7. But 8 ... Nc6 is a more active and direct solution: 9 c3 (9 Nxf6+ Qxf6 10 Qd3 Bd7 11 g3 e5) 9 ... 0-0 10 N2g3 (10 g3 e5 11 d5 Na5 with balanced prospects; e.g. 12 Bg2 Nc4 13 b3 Nd6 14 Nxf6+ Qxf6=) 10 ... Be7 11 Bc4 Kh8 12 0-0 f5!? (or 12 ... e5 13 d5 Na5 14 Ba2 b5!? with the idea 15 d6 cxd6 16 Bxf7 Nc4) 13 Nc5 (13 Nd2 e5) 13 ... e5 14 Ne6 Bxe6 15 Bxe6 exd4 16 Bxf5 dxc3! 17 Qh5 Rxf5 18 Qxf5 cxb2 19 Rad1 Qg8 and the b-pawn is a real threat. 9 Qd2 9 Nxf6+?! is well met by 9 ... Qxf6 10 Qd2 c5=/+. 9 ... Be7 9 ... Bd7 10 0-0-0 Bc6 is a reasonable alternate set-up.
10 0-0-0?! This is rather risky because, in spite of his unimpressive development, Black can target White’s king. Instead: (a) 10 N2c3?! Nc6 (or 10 ... b6 11 Be2 Bb7; e.g. 12 Bf3? Nc6 13 Rd1 e5! 14 dxe5 Nxe5 15 Nf6+ Bxf6 16 Bxb7 Qxd2+ 17 Rxd2 Rfe8!) 11 0-0-0 f5! 12 Ng3 e5! 13 Bc4+ Kh8=/+ has the idea 14 dxe5? Bg5. (b) 10 Qc3 Nc6 11 0-0-0 f5! 12 Nd2 (12 Nc5 Qd5) 12 ... Bf6 and it’s nearly equal, but Black is for choice. (c) Probably White should prefer the simple 10 g3!; e.g. 10 ... Nd7 11 Bg2 c5= or 11 ... c6=. 10 ... Nd7 11 g3
Or 11 Qc3 b5! and an attack: 12 Nf4 Bb7 13 Nc5 Bxc5 14 dxc5 Bc6 with ideas of ... Qg5 or ... Qe7 and ... a5. 11 ... b5! 12 Bg2 Rb8 12 ... Bb7! intending ... a5 is also strong here; e.g. 13 Nf6+ Nxf6 14 Bxb7 Rb8 15 Bg2 b4. 13 d5 Black also stands better after 13 Nf4 b4 14 axb4 Bxb4 15 c3 Be7. 13 ... e5 14 d6 cxd6 15 Nxd6 Nc5 16 Qd5 Rb6=/+ N.Friedrich-Lputian, Berlin 14.22 Black trades space for the bishop pair. 6 Bxf6
5
... Qxf6
7
1989. h6
a3
Ba5!?
This makes it a gambit, but one for which Black gets very obvious compensation. 8 exd5 If White decides not to get greedy, 8 b4 Bb6 9 e5 is probably the most interesting way to proceed: 9 ... Qe7 (9 ... Qf5!? 10 Ng3 Qf4 and 9 ... Qg5!? are good options) 10 Na4 Bd7 11 c3 0-0 12 Nf4 Bxa4 (12 ... Be8 13 g3 f6 14 exf6!? Qxf6 15 Ra2?!, Hector-Glek, Copenhagen 1995; and 15 ... Bxa4 16 Qxa4 g5 17 Nd3 Qf3 18 Rg1 e5! gives Black the advantage; here 14 Bh3! fxe5 15 dxe5 Bf7 16 0-0 Nd7 is highly unbalanced and hard to assess) 13 Qxa4 Nd7 14 Bd3, R.Jacobs-C.Jones, corr 2000; and the most dynamic course would be 14 ... a5!? 15 b5 g5 16 Nh5 f6 17 Nxf6+ Nxf6 18 exf6 Qxf6 19 0-0 e5 with complex equality.
8 ... 0-0! This is the gambit. Whether or not White accepts, Black scores well from this position. 9 Qd3 9 Qd2 Rd8 10 dxe6!? (10 Qf4 Bxc3+ 11 bxc3 Qe7!, and White still does well to avoid 12 dxe6 Qxe6! by playing, say, 12 Qe3 Rxd5 13 Nf4, when 13 ... Ra5!? 14 Bd3 Nd7 15 0-0 Nf6 could follow) 10 ... Bxe6
11 Qe3?! (11 0-0-0! Nc6 12 Qf4 Qxf4+ 13 Nxf4 Nxd4 is only modestly in Black’s favour) 11 ... Nc6 (11 ... c5! 12 dxc5 Nc6 lets Black take a strong initiative, with the idea 13 b4 Nxb4! 14 axb4 Bxb4) 12 0-0-0 Bxc3 13 Nxc3 Nxd4 14 Bd3 c5 15 Rd2?! (15 Rhe1=) 15 ... b6!?=/+ ½-½ Mamedov-Antic, Kavala 2010. 9 ... Rd8 10 dxe6 Bxe6 Black has obvious compensation. 11 0-0-0!? Qxf2 12 Ne4 Qf5 Or 12 ... Qh4 with a minor edge. 13 Nc5 Qxd3 14 Rxd3 Bc8
This is called ‘=‘ inECO. Unzicker-Piskov, German League 1991, continued 15 g3 Nd7 16 b4 Bb6 17 Bg2 c6 18 Nf4, when an ambitious continuation is 18 ... g5!? (18 ... Nxc5 19 bxc5!? Ba5 20 d5 Bf5 21 Rb3 g5 22 Ne2 cxd5 23 Rxb7 is dynamically balanced) 19 Nh5 a5 20 Nxd7! (otherwise ... Nxc5 is coming and Black retains the bishop pair) 20 ... Bxd7 21 Nf6+ Kg7 22 Nxd7 Rxd7 23 c3 axb4 24 axb4 Re7=. 14.3
5
5 ... h6 White 14.31 14.32
has
three
main
moves
here, 6 6
e5
two
covered
in
this
chapter: Bc1 Be3
The traditional main line 6 Bd2 is investigated in Chapter 15. One of the things that I noticed in researching the MacCutcheon is that, at several junctures, White’s ‘irregular’ moves were not only satisfactory for him, but sometimes difficult to counter. The following alternatives are better than their reputation (one author for example, gives both 6 Bxf6 and 6 Bh4 ‘?’s, which are total misassessments):
(a) 6 exf6 hxg5 7 fxg7 Rg8 isn’t dangerous if Black takes a bit of care, and his bishop pair and central majority are potential advantages: (a1) 8 Qh5 Qf6 9 0-0-0 (9 Ne2 Qxg7 10 0-0-0 Nc6 11 h4 Rh8 12 Qg4 f5=/+) 9 ... Qxg7 10 Qf3 a6 (10 ... Nc6 11 Qe3 Ne7!=) 11 Qe3 Be7! 12 Kb1 Nc6=/+. Black has the bishop pair and queenside prospects. (a2) 8 h4! is the most promising move: 8 ... gxh4 (8 ... Nc6 9 h5 Rxg7 has been played, but I like 9 Qd3! for White)
and: (a21) 9 Qg4 Qf6 (9 ... Nc6 10 0-0-0 Qf6 is also equal) 10 Nf3 (10 Rxh4 Rxg7 11 Rh8+ Kd7 and ... Qxd4, when White has compensation but also the burden of proof, since Black stands solidly) 10 ... Rxg7 11 Qxh4 Qxh4 12 Rxh4 Nd7 13 g3 Ke7 14 Bg2 c5=. (a22) 9 Nf3 Qf6 10 Rxh4 Nc6 11 a3!? (11 Qd3 Bd7=), Stambulian-Skorchenko, Krasnodar 2002; and 11 ... Bxc3+! 12 bxc3 Bd7, e.g. 13 Qd2 Qxg7 14 0-0-0 f6 15 Qh6 Qxh6+ 16 Rxh6 Kf7=. (a23) 9 Qh5Qf6 10 Qxh4 (10 Nf3 Qxg7 11 Qxh4 Nc6 12 0-0-0 Bd7=; 10 Rxh4 Nc6 11 Bb5 Bd7 12 Nf3 0-0-0 13 0-0-0 Rxg7= Timman-Ree, Wijk aan Zee 1983) 10 ... Qxg7 11 0-0-0 (11 Nf3 Nc6 12 0-0-0 Bd7 13 Nb5 Ba5 14 Bd3 f6!=/+) 11 ... Bd7 (11 ... Be7!? – Moskalenko – might lead to 12 Qh2 Bd6 13 Qh3 Bd7 14 Qe3 Be7!) 12 Nf3 Nc6 13 a3 Bxc3 14 bxc3 f6=. (b) 6 Bxf6 is fully playable but harmless because White’s centre can come under immediate pressure: 6 ... gxf6 7 a3 (7 Nf3 Nd7 8 Be2 c5 9 exf6 Qxf6 10 0-0 cxd4 11 Nxd4 0-0=) 7 ... Bxc3+ 8 bxc3 fxe5 9 dxe5 c5 (9 ... Qg5!? 10 Nf3 Qg7=) 10 Nf3 Nc6 11 Qd2 Qc7 12 Qf4 (12 Qe3 Qa5 intending ... d4) 12 ... Qa5! 13 Qd2 Qa4 (13 ... Qc7 repeats; 13 ... b6, intending ... Ba6, and 13 ... Bd7 are alternatives) 14 Be2 Rg8 15 0-0! (15 g3 b6 16 0-0 Bb7) 15 ... Qe4!? (15 ... b6 is safe and sound) 16 g3 Nxe5 17 Nxe5 Qxe5 18 Rfe1 Bd7 19 Rab1 b6 20 a4! Qf6 21 a5 Rb8 22 axb6 axb6 with compensation but no more than that for White. (c) 6 Bh4 is considerably better than the books generally indicate: 6 ... g5 7 Bg3 (here 7 exf6 gxh4 8 Qf3 Nd7 will give Black an extra pawn, albeit a weak one, and the bishop pair) 7 ... Ne4 8 Ne2 and now Black has nothing that gains an advantage, but a raft of reasonable moves:
(c1) 8 ... h5 9 h4 (9 f3 Nxg3 10 hxg3 Bd7 with the idea ... Nc6/ ... Qe7/ ... 0-0-0, or if 11 Qd3, then 11 ... Be7 and ... c5 is double-edged) 9 ... c5! 10 a3 Nxg3 11 Nxg3 Bxc3+ 12 bxc3 Qa5 13 Qd2 g4 with sufficient play; e.g. 14 Bd3 Bd7 15 Rb1 Bc6, to meet 16 Qg5 with 16 ... Qxc3+ 17 Kf1 Nd7 18 Nxh5 Qxd4 19 Nf6+ Ke7 20 Nxd5+ Kf8 21 Qe7+=. (c2) 8 ... c5 has been the most common move, with a double-edged game after 9 a3 Nxc3 (or 9 ... Bxc3+ 10 Nxc3 Qa5, when 11 Qd3 Bd7 looks alright, but 11 dxc5 with the idea 11 ... Nxc3 12 Qd2 Qxc5 13 bxc3 might give White a little something) 10 Nxc3 Bxc3+ 11 bxc3 Qa5 12 Qd2 Nc6 13 dxc5 Bd7 14 Be2 Qxc5 15 f4!? gxf4 16 Bxf4 0-0-0 17 Be3 Qa5 18 Bd4 Rhg8=. (c3) 8 ... Nd7 9 a3 (9 Qd3! c5 10 0-0-0+/=) 9 ... Nxg3 10 hxg3 Be7 11 b4? (11 f4+/=) 11 ... a5 12 b5 c5 with good play, Karjakin-Volkov, Panormo 2002. (c4) 8 ... f5 has been a popular line at the top levels:
9 f3! (to me, this casts a few doubts upon 8 ... f5; alternatively: 9 h4 f4 10 Bh2 c5 11 a3 Nxc3 12 Nxc3 Bxc3+ 13 bxc3 Qa5, with the idea 14 Qh5+ Kd8!=; 9 f4?! h5! 10 fxg5 Qxg5 11 Bf4 Bxc3+ 12 bxc3 Qg4 13 h3 Qg7, when White is tied up and faces problems on his queenside light squares; 9 exf6 Qxf6 10 a3 and 10 ... Ba5 is fine, whereas 10 ... Bxc3+ 11 Nxc3 Nxg3 12 hxg3 Nc6 13 Bb5?! appeared in two blitz games by Volkov, in which Black played 13 ... Bd7 and rejected the double attack 13 ... 0-0! 14 Qd2 Qxd4 or 14 ... Nxd4) 9 ... Nxg3 (9 ... Nxc3 10 Nxc3 c5 11 a3 cxd4!? 12 axb4 dxc3 13 bxc3+/=) 10 hxg3 Nc6 11 a3 Bf8 12 f4 Bd7 13 Qd2+/=. 14.31
6
Bc1
This has been played by many strong grandmasters over the years and deserves more respect than it generally receives.
Not only won’t White’s dark-squared bishop be traded off, it may head for a3 at the right moment, and a queen or rook on the third rank will have a clear view of the queenside (as opposed to after 6 Be3). White’s main problem is that c3 and d4 are left undefended. 6 ... Ne47 Qg4 7 Ne2 c5 8 a3 is well met by 8 ... Ba5 (a typical retreat; Black will often go to c7 or b6 with central pressure; the alternative 8 ... Bxc3+ 9 bxc3 Nc6 10 f3 Ng5 intending ... Qa5 is also playable, only because White is undeveloped and almost as disorganized as Black!)
9 dxc5 (9 b4 Nxc3 10 Nxc3 cxd4 11 Nb5 Bc7 12 f4 0-0 13 Bd3 Nc6 14 Nxc7 Qxc7 15 0-0 f5 16 Bd2 a6 proved more than adequate for Black in Heidl-B.Lalic, Biel 1989) 9 ... 0-0 10 b4 Nxc3 11 Nxc3 Bc7 12 Bf4 (12 f4 f6!=/+) 12 ... Nc6 13 Qh5 f5!? (hitting e5; 13 ... b6 is an interesting alternative) 14 Be2 (14 Bxh6? Bxe5 15 Bd2 b6!=/+; 14 Qe2 g5! 15 Qh5 Qe7) 14 ... Bxe5 15 Bxe5 Nxe5 16 0-0 b6! 17 cxb6 axb6 18 f4 Nd7 and ... Nf6, when White’s queenside pawns are vulnerable.
7 Qg4 is a serious try for advantage; Black can equalize but has no easy solution or escape from complications. White’s chances are apparently as good as they are in the main 6 Bd2 lines, which have undergone perhaps fifty times as much analysis. To begin with, Black is faced with the same crucial choice about defence of his g-pawn that arises in many MacCutcheon and Winawer variations: 14.311 14.312
7 7
... ...
Kf8 g6
7 ... g5!? is rather weakening, but adds dynamism to the position, beginning with the idea of ... h5 and ... g4. Given that Black has to tolerate some rather dull positions after 7 ... Kf8 and 7 ... g6, I think this deserves a look: (a) 8 Ne2 c5 (8 ... h5 9 Qf3 is ineffective; e.g. 9 ... c5 10 a3 Ba5 11 b4! cxb4 12 Nxe4 dxe4 13 Qxe4 b3+ 14 c3±) 9 a3 Ba5 and White has two ambitious courses:
(a1) 10 b4!? is a typical attempt to gain the dark squares: 10 ... Nxc3 11 Nxc3 cxb4 (11 ... cxd4 12 Nb5 Bc7 resembles a 4 e5 c5 5 a3 Ba5 Winawer; I suspect that White stands slightly better, as he does after 12 bxa5 dxc3 13 h4 Rg8) 12 Nb5 b3+; this is rather chaotic; e.g. 13 Bd2 (13 c3 Nc6 14 h4 Kd7!? with the ideas ... a6 and ... f6, and answering 15 hxg5?! by 15 ... a6) 13 ... bxc2! 14 h4! (14 Nd6+ Ke7 15 Qf3 Bxd2+ 16 Kxd2 f5 17 Qh5 Qg8 with chances for both sides) 14 ... Qb6!? (14 ... a6!? 15 Nd6+ Kd7 16 hxg5 Qb6 17 Rc1 Qb2 18 Rxc2 Qxc2 19 Bxa5 Nc6 is also wild and apparently balanced) 15 hxg5 Bd7 16 Nd6+ Kf8 17 Rc1
Qb1 18 Ke2 Bb5+ 19 Nxb5 Qxb5+ 20 Ke3 Bxd2+ 21 Kxd2 Qa5+ 22 Kxc2 Rg8 with even chances. (a2) 10 dxc5 Qc7 11 Qf3 (11 b4 can be met by 11 ... Nxc3 12 Nxc3 Qxe5+ 13 Kd2 Bc7; e.g. 14 Nb5 Qf4+ 15 Qxf4 Bxf4+ 16 Kd3 Bd7 17 c4 Be5 18 Rb1 a6 19 Nc3 Nc6) 11 ... Nxc3 (11 ... Nc6!? 12 b4 Nxb4 13 axb4 Bxb4 is perhaps too speculative) 12 Nxc3 Qxe5+ 13 Be2 Nc6 14 0-0 0-0. This looks quite playable because of Black’s good centre and potential pawn expansion, although naturally White has prospects of his own on the kingside. (b) 8 a3 can be answered by 8 ... Ba5 9 Ne2 c5, transposing to 8 Ne2. An independent line is 8 ... h5 9 Qf3 (9 Qh3!? g4 10 Qe3 Bxc3+ 11 bxc3 c5) 9 ... Bxc3+ 10 bxc3 c5 and: (b1) 11 Qe3 Qa5 12 Ne2 Nc6 13 Bb2 Nxd4 14 Nxd4 cxd4 15 Qxd4 b6 16 f3 Nc5 17 h4 Na4 (17 ... Ba6) 18 0-0-0 (18 Bc1 Ba6!) 18 ... Ba6! 19 Bxa6 Qxa6 20 hxg5 Rc8=. (b2) 11 Bd3 Nxc3 12 dxc5 Nc6 13 Qe3 (13 Qf6 Qxf6 14 exf6 Na4 15 Bxg5 Nxc5=) 13 ... Rg8 14 Ne2 d4 15 Qd2 Qd5 and the dissolution of White’s centre establishes equality. 14.311
7
...
Kf8
8 a3 8 Ne2 is the other natural way to develop: 8 ... c5 9 a3 (9 f3 cxd4 10 fxe4 dxc3 11 exd5 Ba5!)
9 ... Bxc3+ (the solid option; 9 ... Ba5 transposes to the main line) 10 bxc3 (10 Nxc3 Nxc3 11 bxc3 Qa5 12 Bd2 Qa4 is unclear but I think quite playable for Black; compare the Portisch-Hook Variation of Chapter 11) 10 ... cxd4 11 cxd4 Nc6
intending ... Qa5 and ... Ne7. 8 ... Ba5 9 Ne2 This the thematic reply. 9 Bd3 is easy to handle: 9 ... c5 (9 ... Nxc3 equalizes: 10 Bd2 Na4!? 11 Bxa5 Nxb2, when 12 Be2 b6 13 Bc3 Na4 14 Bd2 c5=/+ might follow) 10 Bxe4 cxd4 11 b4 dxc3 12 bxa5 (12 Bd3 Bc7 13 f4 a5!=/+) 12 ... dxe4 13 Qxe4 (13 a4 Qd5!) 13 ... Nc6=. 9 ... c5 9 ... f5 is the main recommendation, although I think White retains a limited advantage following 10 Qh3! Nxc3 11 Nxc3 c5 12 dxc5 d4 13 b4 dxc3 14 bxa5 Qxa5 15 Be3. 10 dxc5! 10 b4?! Nxc3 11 Nxc3 (11 bxa5 Nxe2 12 Bxe2 Nc6 13 c3 Qxa5 14 Bd2 h5 15 Qf4 b6=) 11 ... cxb4 12 axb4 (12 Nb5?! b3+ 13 c3?? a6) 12 ... Bxb4=/+. 10 ... Qc7 10 ... h5 11 Qh3 d4 12 b4 dxc3 13 Qe3 (13 bxa5?! Qd2+! 14 Bxd2 cxd2+ 15 Kd1 Nxf2+=/+) 13 ... Nd2!? 14 Qxc3 Nxf1 15 bxa5 Bd7 16 Rxf1 Bb5 with compensation based upon light-square pressure. 11 Qf4! The only other way to shore up the centre is 11 f4, when Black targets the weak queenside pawns by 11 ... Nxc3! 12 Nxc3 Bxc3+ 13 bxc3 Nd7=. 11 ... Nxc3 12 Nxc3 Bxc3+ 13 bxc3 b6!
A standard theme, not only giving the option of ... bxc5, but preparing ... Ba6 before White’s bishops become a force. 14 c4 On 14 Rb1, 14 ... Qxc5! 15 c4 Bb7 is probably best, intending ... dxc4 or ... Nc6, but even 15 ... Nc6 16 Rb5 Qd4 17 cxd5 Qxf4 18 Bxf4 Nd4 19 Rb2 exd5 is balanced. After 14 c4, a sample line would be 14 ... bxc5 15 cxd5 exd5 16 e6!? (16 Rb1 Nc6) 16 ... Qxf4 17 Bxf4 fxe6 18 Bd6+ Kf7 19 Rb1 (19 Bxc5 Ba6 gives Black the better pawn structure) 19 ... Nd7 20 Bb5 a6 21 Bxd7 Bxd7 22 Rb7 Rhd8 and the oppositecoloured bishops ensure equality; e.g. 23 Kd2 Kg6 24 Bxc5 Rab8 25 Rhb1 Rxb7 26 Rxb7 Bb5. Thus 7 ... Kf8 is a completely sound move, whose only drawback involves a lack of winning prospects. 14.312
7
...
g6
8
Ne2
c5
9
a3
9 ... Ba5 A flexible move which keeps many options open. On the other hand, 9 ... Bxc3+ might be the objectively correct move in this case: 10 bxc3 (10 Nxc3 Nxc3 11 bxc3 Qa5 12 Bd2 Qa4 is unbalanced and roughly equal; again, compare the Portisch-Hook Variation of Chapter 11) 10 ... cxd4 11 cxd4 Qa5+ 12 c3 Bd7 13 f3 Bb5! 14 fxe4 Bxe2 15 Qh3 (15 Kxe2 Qxc3 wins the rook!; 15 Qg3 Bxf1 16 Rxf1 dxe4 is similar) 15 ... Bxf1 16 Rxf1 dxe4 17 Kf2 Nd7 18 Kg1, Rytshagov-D.Anderton, Gausdal 2000. Now 18 ... Rh7!? wasn’t bad, but Black can get right to work on the light squares with 18 ... Rc8! 19 Bd2 Rc7 (19 ... Qd5=) 20 Rae1 Nb6 21 Rxe4 (21 Qh4 Nc4 22 Qf6 Rh7 23 Bxh6 Qxc3) 21 ... Qxa3 22 Ref4 Qa2 23 Qd3 a6= intending ... Nd5. 10 b4!? 10 dxc5 appears to be White’s best try: 10 ... Nc6 (perhaps 10 ... Qc7 11 Qf4 Bxc3+ 12 Nxc3 Nxc3 13 bxc3 Nd7 14 Be3 Nxc5 is better, but in that case Black’s dark-square weaknesses seem more important than White’s on the queenside) 11 b4 Nxe5 12 Qh3! Bc7 13 Nxe4 dxe4 14 Nc3 a5 15 Rb1 axb4 16 axb4 Bd7, Negi-Nepomniachtchi, Wijk aan Zee 2007; and here 17 Nxe4! Bc6 18 Nc3 should result in some advantage for White. 10 ... Nxc3 11 Nxc3 cxb4!? 11 ... cxd4 12 Nb5 Bc7 is also equal: 13 Qxd4 (13 f4 Bd7 – an idea from the 5 a3 Ba5 Winawer – 14 Bb2 Bxb5 15 Bxb5+ Nc6=) 13 ... Nc6 14 Nxc7+ Qxc7 15 Bb5 Bd7 16 Bxc6 Qxc6=. 12 Nb5 b3+ 13 c3 Or 13 Bd2 Bxd2+ 14 Kxd2 bxc2 15 Nd6+ Kf8 16 Qf4 f5!. 13 ... a6!? A touch speculative. 13 ... 0-0 is sound enough. 14 Nd6+ Qxd6! 15 exd6 Bxc3+ 16 Bd2 Bxa1 Black has active play in a double-edged position. 14.32
6
Be3
A simple move that can lead to extraordinarily complex and rich play. While not as popular as 6 Bd2, theory on this move has recently expanded dramatically. 6 ... Ne4 7 Qg4 g6 This and 7 ... Kf8 are considered about equally good. I’ve chosen 7 ... g6 mainly because of the main line which forces the play somewhat and therefore simplifies matters. 7 ... Nxc3? is inferior due to 8 Qxg7 Rf8 9 Bd2 (9 a3 Ba5 10 Bd2 is also advantageous, although 10 ... c5 11 bxc3 Qb6 gives Black some play) 9 ... Nxa2 10 c3 Nxc3 11 bxc3 Be7 12 Bxh6±. 8 a3 Bxc3+ 9 bxc3 c5 9 ... Nxc3 10 Bd3 c5 11 dxc5! Nc6 is seen as well. Attacking the pawn chain by 9 ... c5 is not necessarily better, but it’s thematic and relatively easy to play. 10 Bd3 Nothing else makes sense; White must move quickly in the face of ... Qa5 and ... Nc6. 10 ... h5 Moskalenko’s favourite move, which forces White’s queen to commit. We examine:
14.321 11 Qf3 14.322 11 Qf4 The other two queen moves are a bit easier to handle: (a) 11 Qd1 Nxc3 12 Qd2
and now: (a1) Moskalenko analyses 12 ... cxd4 13 Bxd4 Ne4 14 Bxe4 dxe4 15 Ne2 (or 15 Qe3 Nc6 16 Bc3 b6 17 Qxe4 Bb7 and the opposite-coloured bishops give Black the advantage) 15 ... Nc6 16 0-0 b6 17 Rfd1 Bb7 and White’s pawns are vulnerable to ... Qc7, for example; here 17 ... Ba6! looks better still. (a2) 12 ... Ne4 13 Bxe4 dxe4 14 Ne2 (14 dxc5 Qxd2+ 15 Kxd2 Bd7 16 Ne2 h4!? 17 Rab1 Bc6 18 Nd4 Rh5 is unclear) 14 ... Nc6 15 Rd1 cxd4 16 Nxd4 Nxd4 17 Bxd4, Kanarek-Lortkipanidze, Fermo 2009; and Moskalenko suggests the simple 17 ... b6 with equality or better. (b) 11 Qh3 Nxc3 12 dxc5 d4! 13 Bf4 (after 13 Bd2 Nd7, Black will win the e- or c-pawn with an easy game) 13 ... Nd7 (13 ... g5!? 14 Bd2 Qd5!?) 14 Ne2! Nxe2 15 Bxe2 Qa5+!? 16 c3! Qxc5 17 Rc1 d3! 18 Bxd3 Nxe5 with unclear prospects. 14.321
11
Qf3
Nxc3
12
dxc5
A typical structure from this line. Black goes after White’s centre as usual: 12 ... Nc6 13 Qf4 d4 14 Bd2 g5 15 Qxg5 15 Qg3 can be met by 15 ... g4 or 15 ... h4!? 16 Qxg5 Qxg5 17 Bxg5 Nxe5 18 Nf3 Nxf3+ (18 ... Rh5!?) 19 gxf3 Nd5! 20 Be4 f6! with counterplay (McDonald). 15 ... Qxg5 16 Bxg5 Nxe5 17 Nf3 17 Bf6 Nxd3+ 18 cxd3 Rg8 19 Bxd4 Nd5 (Black’s unopposed bishop and open lines for his rooks give him fair chances; 19
... Na4 is another option) 20 g3 Bd7 21 Nf3 Bc6 (21 ... f6!?) 22 Ne5 Nb4 23 Nxc6 Nc2+ 24 Kd2 Nxa1 25 Na5 b6 26 Rxa1 bxa5 27 Rb1 Rg5 with balanced chances, Kosteniuk-Zhukova, Gibraltar 2010. 17 ... Rg8 18 Bf6 Nxd3+ 19 cxd3 Nd5 20 Bxd4 Rxg2=/+ Sutovsky-Wang Hao, Poikovsky 2008. With the disappearance of the g2-pawn, White’s light squares become more vulnerable. 14.322
11
Qf4
g5!
12 Qf3 Nxc3 13 dxc5 13 Ne2 clarifies the centre at the cost of time; for example, 13 ... Nxe2 (13 ... Nc6 14 Nxc3 cxd4 15 Bxd4 Nxd4 16 Qe3 is unclear) 14 Bxe2 cxd4 15 Bxd4 (Moskalenko cites 15 Bd2?! Nc6=/+ of Larrat-Delabaca, French Team Ch. 2009) 15 ... Nc6 16 c3 Nxd4!? (16 ... b6; 16 ... Bd7 17 0-0 Rc8) 17 cxd4 Qa5+ 18 Kf1 h4!? 19 Qf6 Rg8 and Black has his share of the play. 13 ... d4 13 ... Nc6 is a good alternative: (a) 14 Qf6 Qxf6 15 exf6 e5! 16 Bxg5 (16 Ne2 Nxe2 17 Kxe2 Bd7 18 Rhd1 0-0-0=; 16 Bd2 e4 17 Bxc3 exd3 18 cxd3 d4 19 Bb2 Bf5 20 Nf3 g4 21 Nxd4 Nxd4 22 Bxd4 0-0-0=) 16 ... Bd7 (or 16 ... Be6=, or 16 ... h4=) 17 f4 e4 18 Ne2 Nxe2 19 Bxe2 Nd4=. (b) 14 Bd2!? Nxe5 15 Qg3 Nxd3+ 16 cxd3 d4! 17 Bxg5 f6 18 Bd2 Kf7=. (c) 14 Qg3!?
14 ... Qa5! 15 Bd2 Qxc5 16 Nf3! (16 Qxg5? Ne4 17 Bxe4 Qd4!-/+ - Moskalenko) 16 ... g4 17 Ng5 Qd4! (17 ... d4?! 18 0–0 Qxe5 19 Rae1!, Hamdouchi-Rivas Pastor, Salou 2011) 18 Qf4! (18 0–0? Ne2+! 19 Bxe2 Qxd2; 18 Be3 Qxe5 19 Qxe5 Nxe5 20 Bd4 f6! 21 Bxc3 Nxd3+ 22 cxd3 e5=/+; 18 Qe3 Na4! 19 Qxd4 Nxd4 20 h3 Nc5 21 Rc1 Nc6 22 Bf4 gxh3 23 Rxh3 Bd7 with good play) 18 ... Qxf4 19 Bxf4 d4 20 h3 Rg8 21 Nh7 Nd5 22 Bg3 Bd7 23 hxg4 hxg4 24 Be4 0–0–0 25 Bh4 Nc3 26 Bxc6 Bxc6 27 Bxd8 Kxd8 28 Nf6 Rg5 29 Rh5, E.Berg-Tikkanen, Halmstad 2012; and now 29 ... Rf5!; e.g. 30 Rxf5 exf5 31 Kd2! Bxg2 32 Kd3 Bf3 33 Re1! (33 Kxd4 Nd1!) 33 ... Ke7 34 Kxd4 Ne2+ 35 Ke3 f4+ 36 Kd3 g3 37 fxg3 Nxg3, and I think that Black can draw this rather easily after ... Ke6. 14 Bd2
14 ... Bd7 Keeping some play on the board. Here 14 ... Nc6 has been the main line, and is also satisfactory. Then after 15 Bxc3! (15 Qg3 Qd5 16 Qxg5 Qxe5+= Nemcova-Guo Qi, Gaziantep 2008; 15 Qf6 Qxf6 16 exf6 Na4=) 15 ... dxc3, Black has full equality but few positive prospects; e.g. 16 Qe3 (16 Qf6!? Qxf6 17 exf6 Ne5!=) 16 ... Qd4 17 Qxd4 Nxd4 18 Ne2 (18 Rb1 Bd7) 18 ... Nc6! 19 Nxc3 Nxe5 20 h4 g4 21 0-0 Ke7=. 15 Bxc3 Moskalenko calls 15 Qg3 Bc6 unclear; a sample line would be 16 Nf3 (16 Ne2 Nxe2 17 Bxe2 Nd7=) 16 ... h4 17 Qg4 h3 18 Nxd4!? (18 Nxg5 hxg2 19 Rg1 Rg8=) 18 ... Rh4 19 Nxe6 Rxg4 20 Nxd8 Kxd8 21 Bxc3 hxg2 22 Rg1 Nd7=/+. 15 ... Bc6 16 Be4 dxc3 17 Rd1 Qc7 This is also assessed as unclear by Moskalenko. An interesting way for White to proceed is 18 Bxc6+ (18 Qf6 Rg8!=/+) 18 ... Nxc6 19 Rd6!, although 19 ... Nxe5 20 Qf6 Ng6 21 Qxg5 Rd8 is fully equal.
Chapter
Fifteen
MacCutcheon:
Main
Lines
with
6
Bd2
1 e4 e6 2 d4 d5 3 Nc3 Nf6 4 Bg5 Bb4 5 e5 h6 6 Bd2
This is the main line of the MacCutcheon Variation. Since 6 ... Ne4?? would lose a piece to 7 Nxe4, Black needs to eliminate the knight on c3 first. 6 ... Bxc3 7 bxc3 The pawn recapture occurs in the vast majority of games, but 7 Bxc3 is playable and should be taken seriously. White has many legitimate tries after Black’s response 7 ... Ne4:
(a) 8 Bd3 Nxc3 9 bxc3 c5 10 Nf3 Qa5! 11 Qd2 Nc6 prevents White from castling and therefore at least equalizes. Probably 12 dxc5 Qxc5 13 0-0 0-0 14 Rfe1 Bd7 represents best play. (b) 8 Qg4 0-0!=/+ takes advantage of the fact that White has no kingside attack. (c) 8 Ba5 0-0! 9 Bd3 Nc6 10 Bc3 Nxc3 11 bxc3 f6 is already better for Black; e.g. 12 f4 fxe5 13 fxe5 Ne7! 14 Nf3 c5 15 0-0 Qa5 (15 ... c4 16 Be2 Bd7 17 Qd2 Be8=/+) 16 Qe1 Bd7 (16 ... c4! 17 Be2 Bd7) 17 c4 Qxe1 18 Rfxe1 dxc4 19 Be4 (19 Bxc4 b5 20 Bf1 Rac8) 19 ... cxd4 20 Bxb7 Rab8 21 Ba6 Rb4 22 Rad1 d3! 23 cxd3 cxd3 24 Rxd3 Bc6, R.J.Fischer-T.Petrosian, Curacao 1962; with control of d5 and the long diagonal, Black has a limited advantage. (d) 8 Bd2 Nxd2 (8 ... c5 is more ambitious; e.g. 9 Be3 cxd4 10 Bxd4 b6 11 f3 Nc5) 9 Qxd2 c5 10 dxc5 Nd7 11 Bb5 0-0 12
Bxd7 Bxd7=. (e) 8 Bb4 isn’t as bad as it’s made out to be, but Black can equalize:
(e1) 8 ... Qh4? has done well, but in fact White has 9 g3! Qg5 (the obvious 9 ... Nxg3 loses to 10 fxg3 Qe4+ 11 Kf2 Qxh1 12 Nf3 intending Be1 and then Bg2; 9 ... Qd8 10 f3 Ng5 11 h4 Nh7 12 Qd2) 10 Ba3 intending Be2 followed by f3 as well as the ideas of Nh3-f4 or simply Nf3. (e2) 8 ... c5! 9 Bxc5 (9 dxc5? Nxf2 10 Kxf2 Qh4+ 11 g3 Qxb4) 9 ... Nxc5 10 dxc5 Nd7 11 Qd4 Qc7 12 f4?! (12 Nf3! Nxc5= Spielmann-Nimzowitsch, Gothenburg 1920, is a typical French position) 12 ... Nxc5 13 0-0-0 Bd7 14 Nf3 Rc8 15 Kb1 0-0 16 Bd3, Kovchan-Vysochin, Swidnica 1999; and 16 ... Qb6! should be considered. Then Black has ... Bb5 and/or doubling on the c-file in mind, with potentially exciting play in store. (f) 8 Ne2 is arguably the most logical move, followed by 8 ... c5 9 dxc5
and then: (f1) 9 ... Nc6 10 Bd4 Qe7 11 Nc3 Nxc5 12 f4 (12 Be2 0-0 13 0-0 Rd8= intending 14 f4?! Ne4) 12 ... 0-0 13 Qd2, SutovskyComas, Pamplona 1998; and Black has several good moves; a straightforward plan is to prepare ... Ne4 by 13 ... Rd8 14 0-0-0 Nxd4 15 Qxd4 b6 and ... Bb7, or similarly, 13 ... b6 14 Bb5 Nxd4 15 Qxd4 Bb7. (f2) 9 ... Nxc3 10 Nxc3 0-0 11 Qd2, and here relatively simple moves will do; e.g. 11 ... Qe7 (or 11 ... Nd7 12 Qd4 f6 13 exf6 Qxf6 14 Qxf6 Rxf6 15 Na4!? Ne5 16 f3 Bd7 17 Nc3 Rc8 18 b4 b6=/+) 12 Qe3 (12 f4 Nd7) 12 ... f6 13 exf6 Qxf6 14 0-0-0 Qxf2=. 7 ... Ne4 8 Qg4
This is the main line of the MacCutcheon. There are a couple of important alternatives, however, which Black should take note of: (a) 8 Bd3 Nxd2 9 Qxd2 c5 is unsurprisingly equal; e.g. 10 dxc5! (10 Nf3 can be answered by 10 ... c4=, 10 ... Nc6 11 0-0 Qa5, at least equal, or 10 ... Qa5; e.g. 11 dxc5 Nd7 12 0-0 Nxc5 13 c4 Qxd2 14 Nxd2 Ke7=) 10 ... Qc7 (10 ... Nd7 11 Nf3 Nxc5 12 0-0 0-0 is comfortable for Black) 11 Qe3 0-0 12 Nf3 Nd7 13 0-0 Nxc5 14 Nd4 Bd7 15 f4 Rac8 16 g4, Fink-Stoppel, Austria 1966. This has been thought to favour White, but in fact Black stands better after 16 ... Nxd3 (16 ... f6?! was played in the game) 17 cxd3 Qxc3 18 Rac1 Qa3 19 Rxc8 Rxc8 20 f5 (one annotator assessed this as winning for White! – but he is practically lost ... ) 20 ... Rc4! (or 20 ... Rc3 with the idea 21 Qf4 Qc5 22 Kh1 Qb6-/+) 21 fxe6 (21 f6 Qc3 22 Ne2 Rxg4+) 21 ... Bxe6 22 g5 hxg5 23 Nxe6 fxe6 24 Qh3 Qc5+ 25 Kh1, and now the nice 25 ... g4! 26 Qh5 Rf4! 27 Qe8+ Qf8 28 Qxe6+ Kh7–+. (b) 8 Nf3 allows Black’s knight to maintain its post on e4 and cuts the queen off from the dangerous squares g4 and h5. The sequence 8 ... c5 9 Bd3 Nc6 10 0-0 Qa5, threatening ... Nxd2, has served Black well. (c) 8 Bc1!? looks stupid, as in some lines White is a full tempo behind 6 Bc1.
That’s not the whole story, however, because White has ideas of Ne2 and won’t commit to Qg4 until Black grabs the c3pawn. In some cases, White bishop can still go to a3. Finally, if White plays dxc5, Black’s move ... d4 doesn’t come with tempo. The game is fairly balanced, since Black will be a pawn up for a while following 8 ... Nxc3 (after 8 ... c5, 9 Ne2! is hardly ever seen, but I think that the idea of f3 gives White reasonable play, especially since 10 ... Qa5?! 11 f3! Nxc3 12 Qd2 ends well for White; so 9 ... 0-0 10 f3 Ng5 11 h4 Nh7 might follow, with unclear consequences) 9 Qg4 g6 (9 ... Kf8 is also played) 10 Bd3 (10 Qh3!? is another rare move which might be best answered by the odd 10 ... Ne4 11 f3 Nc3!) 10 ... c5 (10 ... Nc6?! 11 a4! with the idea 11 ... Nb4? 12 Bd2), and now:
(c1) 11 Bd2 Ne4! 12 Bxe4 dxe4 13 dxc5! Qd4!? (13 ... Bd7 14 Ne2 Bc6 15 Rd1 Nd7 is also complicated) 14 Rd1 Qc4 (14 ... Qxe5 15 Ne2! Qxc5 puts Black terribly behind in development for his two pawns, although that’s not easy to exploit) 15 Ne2 Bd7 16 0-0! Bc6 (16 ... h5? 17 Qh4 Qxe2 18 Qf6 Rh7 19 Bg5 Nc6 20 Rd6) 17 Rfe1 Nd7 18 Ng3 Nxc5 19 Be3 Bd5 20 Bxc5 Qxc5 21 Nxe4 Bxe4 22 Qxe4 0-0=. (c2) 11 dxc5 has been played in 90% of the games in this line. Then a reliable answer is 11 ... Qc7 (11 ... Qa5 is a bit crazier after 12 Bd2 Qa4 13 Qf3! Ne4 14 Bxe4! dxe4!, and 15 Qb3 Qa6 or 15 Qf6 Rh7! intending ... Nd7 and/or ... Qxc2) 12 Qd4 Ne4 13 Bxe4 Nc6 14 Qc3 dxe4 15 Ne2 Qxe5! 16 Qxe5 Nxe5 17 Bb2 f6 18 Bxe5 fxe5 19 Ng3 b6 20 Nxe4 Ke7. This is unbalanced, but Black is in no danger of standing worse.
8 ... g6 Black is faced with the most fundamental choice in this entire variation, whether to play a position with ... g6 or ... Kf8. The former weakens Black’s kingside (notably the f6-square), but retains the possibility of castling on either side of the board or, importantly, of moving the king to d7, a familiar spot for it in the French. This is by contrast with 8 ... Kf8, which avoids kingside weaknesses and has proven fully satisfactory in practice. The main drawback is that Black’s rook on h8 is difficult to activate (and indeed, sometimes Black plays ... Ke8-d7 later in the game!). Given that there is far too much theory to examine both, I felt that 8 ... g6 was a dynamic option most in tune with this book. According to contemporary theory, the choice is more a matter of taste than objective superiority. After 8 ... g6, White has an array of moves, with these three arguably the most challenging:
15.1 15.2 15.3
9 9 9
Qf4 Be3 Bd3
Others: (a) 9 h4 usually transposes after 9 ... c5 10 Bd3 (10 h5?! g5 11 Bd3 Nxd2 12 Kxd2 Qa5 13 Nf3 Nc6 – contrast the main line, in which White’s queen gets to f6 or at least the break f4 is achieved) 10 ... Nxd2 11 Kxd2 (see 15.31 below). (b) 9 Nf3 c5 also transposes after 10 Bd3 Nxd2 11 Kxd2 (see 15.32). A unique line is 10 dxc5?! Qa5 (or 10 ... Nd7 11 Bd3 Ndxc5) 11 Bd3 Nxd2 12 Kxd2 Nc6 13 Qh4 Qxc5 14 Qf6 Qxf2+ 15 Kc1 Qe3+ 16 Kb2 Qb6+ 17 Kc1 0-0!. (c) Saving the bishop pair with 9 Bc1 is slow but not bad for White. A reasonable line is 9 ... c5 10 Bd3 Nxc3 11 dxc5 (11 Bd2 Ne4) 11 ... Qa5 12 Bd2 Qa4! 13 Qf3! Ne4 14 Bxe4 dxe4 15 Qb3 Qc6= or 15 ... Qa6=. 15.1
9
Qf4
9 ... c5 This is almost always played. 9 ... g5 aims to misplace the queen and seems an acceptable alternative. Of the many possibilities, a typical line is 10 Qe3 c5 (10 ... f5 11 exf6 Qxf6 12 Bd3 Nxd2, Hungaski-Cubas, Sao Paulo 2005; 13 Kxd2! Nc6 14 h4 g4 15 Qg3 is perhaps slightly in White’s favour due to Black’s weakened kingside) 11 Bd3 Nxd2 12 Qxd2 Nc6 13 Bb5 Qa5 14 Bxc6+ bxc6 15 h4 g4 16 Ne2 Ba6 with a dynamic equality, Amonatov-Skorchenko, Voronezh 2005. 10 Bd3 Nxd2 Or 10 ... Bd7 (10 ... g5 11 Qe3 transposes to 9 ... g5 above) 11 Nf3 (as opposed to 11 Bxe4 dxe4 12 Ne2 Bc6 13 0-0 g5!?, when I think that 14 Qg4! is the best move and not so clear, in spite of Black’s light-square superiority; perhaps +/= is a fair verdict; here 13 ... Qd5 is also possible) 11 ... Bc6 12 h4 Nd7 13 dxc5!? (13 0-0!?) 13 ... Qc7 (13 ... Nxd2 14 Qxd2 Qc7) 14 Be3 Nxc3 15 a4 Ne4 16 0-0, Kobalia-Zhukova, playchess.com 2004; and now 16 ... Ndxc5 is at any rate satisfactory for Black. 11 Qxd2
This is the point of 9 Qf4: White’s king isn’t exposed. 11 ... Nc6 Alternatively: (a) 11 ... Qa5 12 Nf3 b6 13 h4 Ba6 (or 13 ... Nc6 14 dxc5 Qxc5 15 0-0 Bd7, contemplating ... Na5) 14 dxc5 bxc5 15 0-0 Nd7=. (b) 11 ... Bd7 preparing ... Bc6 and ... Nd7 is a common idea, especially when there are no Bxg6 sacrifices in the position; e.g. 12 Nf3 Bc6 (12 ... c4 also looks adequate for equality: 13 Be2 Bc6 14 h4! Nd7 15 h5 g5 16 Nh2!, Kobalia-Glek, Ajaccio rapid 2004; and here I think Black should react to the Ng4-f6 idea with 16 ... f5! 17 exf6 Qxf6 18 Ng4 Qf4=, or 18 ... Qf5 and ... 0-0-0) 13 h4 (after 13 dxc5 Nd7 14 0-0, Black has the rare opportunity to stash his king away by 14 ... Kf8 15 Rfe1 Kg7 16 Nd4 Nxc5=) 13 ... Nd7
14 a4!? (14 h5 g5 15 Nh2, Kayumov-Porat, Budapest 2003; and the most direct course was 15 ... Qa5! 16 Ng4 0-0-0; e.g. 17 a4 Rdf8 18 Be2?! f6 19 exf6 Nxf6 20 Nxf6 Rxf6=/+) 14 ... Qa5 15 0-0 (15 dxc5 Nxc5 16 0-0, Grigoriants-Glek, Hoogeveen 2003; and Glek likes 16 ... Ne4!=/+) 15 ... c4!? (or 15 ... Rc8 intending ... cxd4) 16 Be2 Nb6 17 h5 Nxa4!? (17 ... gxh5) 18 hxg6 fxg6 19 Nh4 g5 20 Bh5+ Kd7 21 Ng6 Qxc3! 22 Qxc3 Nxc3 23 Nxh8 Rxh8=. 12 Nf3 c4 13 Be2 Bd7 13 ... Qe7 and 13 ... f5 are legitimate options. 14 h4 White does no better with 14 Rb1 b6 15 h4 f5 16 exf6? Qxf6 17 Qe3 0-0-0 18 Kd2 Rdf8=/+ M.Hoffmann-Istratescu,
Gibraltar 2010.
14 ... f5! Wintzer’s novelty, protecting against Qf4-f6 or Nh2-g4. Black now has long-term prospects of ... g5, and thus White captures, giving up the cramping influence of the pawn on e5. Previously, 14 ... Qe7 had been played. 15 exf6 Qxf6 Now the move h4 actually proves a drawback by weakening White’s kingside. 16 h5 g5 17 Nh2 0-0-0!? 17 ... e5! (Wintzer) would have brought Black at least equality. 18 0-0 And here, again, 18 ... e5 looks to be the clearest solution to the position, since 19 dxe5 Nxe5 (19 ... Qxe5!?) 20 Qxd5 Kb8 and ... Bf5 should be equal. 15.2
9
Be3!?
This resembles 6 Be3 from the previous chapter (section 14.32). Although White has made an extra bishop move, Bd2e3, he has gained a tempo by not playing a3. 9 ... c5 10 Bd3 h5 Or 10 ... Nxc3 11 dxc5 Qa5 12 Bd2 Qa4! – Moskalenko, who points out that the lack of a3 means that Qb4 isn’t available
for White. This looks about equal. 11 Qf4 11 Qf3 Nxc3! 12 dxc5 Nc6, or 11 Qd1 Nxc3 12 Qd2 cxd4 13 Bxd4 Ne4 14 Bxe4 dxe4 ‘with equal chances’ (Moskalenko). 11 ... g5! 12 Qf3 Nxc3
13 dxc5 White’s alternatives don’t impress: (a) 13 Ne2 Nxe2 (13 ... Nc6!? 14 Nxc3 cxd4 is unclear) 14 Bxe2 cxd4 15 Bxd4 Nc6 is at least equal. (b) 13 Qg3? c4 14 Ne2 (Moskalenko gives 14 Bxg5 Qb6 15 Be2 Qb2, 14 Bf1 Ne4 15 Qf3 Qa5+ 16 Kd1 Nc3+–+, and 14 Bd2 Nb5 15 Bxg5 Qb6 16 Be2 Nxd4) 14 ... h4 15 Qg4 (White can bail out with 15 Qxg5 Qxg5 16 Bxg5 Nb5 17 Bf6; e.g. 17 ... Rh6 18 Bg7 Rh5 19 g4 hxg3 20 Nxg3 Rh3, but Black will still emerge with an extra pawn) 15 ... Na4 16 0-0 (16 Bxg5 Rg8 17 f4 cxd3) 16 ... cxd3 with a nearly winning game, Felgaer-Moskalenko, Barcelona 2005. 13 ... Nd7 Two alternatives are 13 ... Nc6 14 Qf6 Qxf6 15 exf6 e5=, and 13 ... d4, when 14 Bd2 Bd7 15 Qg3 Bc6! gives satisfactory play (compare 14.322). 14 Bd4 Qa5 Moskalenko stops here. There might follow 15 Kf1 Na4 16 Qe3 Naxc5=. 15.3 9 Bd3 This is the traditional and most popular move.
9 ... Nxd2 10 Kxd2 c5 Now we
come
to
15.31 15.32
a
critical
choice
11 11
for
White: h4 Nf3
11 Qf4 gives Black a choice of transpositional responses; for example, 11 ... Bd7 12 Nf3 can be found under the order 11 Nf3 Bd7 12 Qf4; and 11 ... Nc6 12 Nf3 transposes to 11 Nf3 Nc6 12 Qf4. The immediate 11 ... Qa5 and 11 ... cxd4 12 cxd4 Qa5+ intending ... b6 are also fine. 15.31
11
h4
This position is the current main line of the 8 ... g6 MacCutcheon; it can also arise from 9 h4 c5 10 Bd3 Nxd2 11 Kxd2. Now I apologize for one more split. Unfortunately, this is the sort of line that will change rapidly with theory, and it’s important for Black to have alternatives; furthermore, the following are stylistically very different: 15.311 15.312 15.313
11 11 11
... ... ...
Nc6 c4 Qa5
Instead, 11 ... Bd7, with the idea of ... Bc6 and ... Nd7, can lead to surprising action: 12 Rb1 (12 h5 g5 13 f4! Qa5! 14 dxc5 d4 15 Ne2 Nc6 16 fxg5 dxc3+ 17 Nxc3 Qxc5 is hard to assess; 12 Nf3 transposes to 11 Nf3 Bd7 12 h4 – see the main line 15.32 below) 12 ... Nc6!? 13 Rxb7 cxd4 14 cxd4 Qa5+ 15 Ke3 Nb4 (15 ... Qe1+ 16 Qe2 Qa1 is unclear) 16 Ne2 0-0-0 17 Rxb4 Qxb4 with doubleedged play, Blomqvist-Skoldebrand, Täby 2007; White has some attack for the exchange. 15.311
11
...
Nc6
12 Nf3 Others: (a) 12 Rh3?! cxd4 (or 12 ... Bd7 13 Qf4 Qc7!? 14 Nf3 cxd4 15 cxd4 f5!?) 13 cxd4 Qb6 with the idea 14 Ne2? (14 Nf3 Bd7=/+) 14 ... Qb4+ 15 Kd1 Nxe5!. (b) This is not the most compelling time for 12 dxc5, due to 12 ... Qa5, when 13 Qf4 (13 h5 g5 14 f4! Qxc5 15 fxg5 Qf2+ 16 Qe2 Qf4+ 17 Qe3 Qxe3+ 18 Kxe3 Nxe5 19 gxh6 Ng4+ 20 Kd2 Rxh6 intending ... e5) 13 ... Qxc5 (13 ... Bd7 14 Qf6 Rg8!?) 14 h5 g5 15 Qf6 Rg8 16 Ne2 (16 Rb1 b6 17 Ne2 Bd7=) 16 ... Qe7 is at least equal. (c) 12 Qf4 is most easily met by 12 ... Qa5 (when 13 Nf3 bring us back to the main line of 15.313 below), or by 12 ... c4. Instead, 12 ... cxd4!? 13 cxd4 Qa5+ can lead to 14 c3 (14 Ke2 b6 15 Nf3 Ba6; 14 Ke3!? b6 15 h5 gxh5 16 Ne2 Ba6 17 Bxa6 Qxa6 18 c3 Rc8=/+ Van den Doel-Glek, Zwolle 2002) 14 ... b5 15 Qf6 Rf8 16 Ne2 b4 17 Rhc1 Rb8 18 Ke1 bxc3 19 Kf1 Ba6=. (d) 12 h5!? g5 13 f4!? doesn’t seem to gain any advantage; e.g. 13 ... cxd4!? (13 ... c4 is safe, when 14 Bf1 f5! 15 exf6 e5 16 Qe2 Qf6 or 16 ... e4!? is comfortably equal, and the speculative 14 fxg5!? cxd3 15 gxh6 probably falls a tad short to 15 ... Ne7 or perhaps 15 ... Qe7) 14 cxd4 Qa5+
(d1) 15 Ke3 f5! 16 Qf3 Bd7 17 Ne2 0-0-0=/+ (17 ... g4!?) 18 g3 Rhg8 19 Rhb1?! gxf4+ 20 gxf4 Rg4 and Black stands distinctly better, Skorchenko-Kobylkin, Krasnodar 2002. (d2) 15 c3! f5! (15 ... Nxd4? 16 fxg5! Nb3+!? 17 axb3 Qxa1 18 g6!) 16 exf6 e5 17 Bg6+ Kd8 with great complications. 12 ... c4
This is playable, simply closing the position in preparation for ... 0-0-0 and a later ... f5. But it’s not the only move: (a) 12 ... Qa5 transposes to 11 ... Qa5 12 Nf3 Nc6 (see 15.313). (b) 12 ... cxd4 13 cxd4 Qa5+ has a good reputation.
For example, 14 c3 (14 Ke2 b6 15 Qf4 Ba6 16 Rhc1 Qa3 17 Qe3 Rc8 puts pressure on White, Moskalenko-Gharamian, Belgian Team Ch. 2007/08) 14 ... b6 15 Qf4 (15 h5 g5 16 Nh2 Bd7! 17 Rhc1 Rc8 18 a4 Rf8=/+ intending ... f5) 15 ... Ba6 16 Qf6 (16 Bxa6 Qxa6 17 Nh2 h5 18 Nf3, Naiditsch-L’Ami, Wijk aan Zee 2010; 18 ... Qa3! 19 Rhc1 Rc8 20 Ng5 Rc7=/+) 16 ... Rg8 17 Bxa6 (17 Bxg6?? Rxg6 18 Qh8+ Kd7 19 Qxa8 Qb5; 17 h5? Bxd3 18 Kxd3 Qb5+ 19 Ke3 Rc8–+; 17 Bc2 Rc8 18 Rh3 Nb4! – Ganeson) 17 ... Qxa6 18 h5 Rc8=/+ 19 Nh4? Nxd4 (or 19 ... g5 – Ganeson) 20 cxd4, Homer-Ganeson, English League 2010; and 20 ... Qa5+ is winning. 13 Be2 Bd7 Black has no real problems in these positions; e.g. 13 ... Qe7 14 Rhb1 b6 15 a4 Bd7 16 Qf4 0-0-0 of Castella-Moskalenko, Ampolla (rapid) 2006, looks perfectly good, and 13 ... Qa5 14 a4 Ne7 15 Qf4 Bd7 16 Qf6 Rh7 17 Rhb1 b6 is equal. 14 Qf4 14 Rab1 b6!? 15 Qf4 Qe7 16 h5 g5 17 Qe3?! 0-0-0=/+ Nakamura-Laznicka, Oropesa del Mar 1999. Here 17 Qf6 was probably equal. Black himself had a good option in 14 ... b5. 14 ... Qe7
“We have reached the critical position of this line” – Moskalenko. But since 12 ... cxd4 and others are good options, I’m going to limit my coverage to a few lines here:
(a) 15 Nh2 h5 (versus Ng4; the alternative is 15 ... 0-0-0 16 Ng4 Rdf8 17 Nf6 Nb8!? 18 a4 Bc6 19 a5 Nd7=, or 17 ... Qa3) 16 Nf3 0-0-0=. This position is so closed that it’s unlikely either side will be able to achieve a successful break. (b) 15 g4 b5!? (or 15 ... 0-0-0) 16 Rhb1 Rb8 17 Qf6 Rh7=/+. (c) 15 a3 0-0-0 16 Nh2, Golubev-Moskalenko, Alushta 1994; and best is 16 ... Rdf8! 17 Ng4 f5 18 exf6 h5! 19 fxe7 Rxf4 20 Ne5 Nxe5 21 dxe5 Re8, at least =. (d) After 15 Qf6, of V.Panov-Korchnoi. USSR Team Ch. 1953, the easiest course is Moskalenko’s 15 ... Qxf6 16 exf6 0-0-0=. 15.312 11 ... This is appropriate for those who like closed positions. 12 Be2 Nd7! Black has a number of moves, of course; the text is unique in that it prepares ... f5 in several lines.
c4
13 Rh3 Another interesting sequence is 13 h5 g5 14 Qg3! (14 f4 f5! 15 Qg3 g4; 14 Nh3 f5 15 Qg3 Nb8 16 f4 g4 17 Nf2 Nc6=/+) 14 ... f5 15 exf6 Nxf6 16 f3 Bd7 17 Nh3, Bergez-Libiszewski, Montpellier 2003; and the game’s 17 ... Qe7 was okay, but Hungaski’s suggestion 17 ... Qb8! 18 Qxb8+ Rxb8 19 Nf2 Ke7=/+ has the idea of ... Kd6 and ... e5, and White has nothing particularly effective against this. 13 ... Qe714 f4 Hungaski and De Dovitiis analyse 14 a3 (preventing ... Qa3) 14 ... Rg8 15 Qf4 g5 16 hxg5 hxg5 17 Qe3 f5 18 exf6 Nxf6 19 f3. Now 19 ... Bd7! 20 Qe5 0-0-0 is effective; White’s pieces aren’t co-ordinating. 14 ... Rg8! 15 Rg3 15 Bf3 f5 16 exf6 Nxf6 17 Qg3 Bd7 18 Qe1 0-0-0=/+ Sedlak-Delchev, Subotica 2003. 15 ... f5 16 Qh3 Nf8 I think this is the safest, preparing to castle queenside. Alternatively: (a) 16 ... Nb6 17 h5 g5 18 fxg5 hxg5 19 h6 Bd7 20 Qh5+ Kd8 21 h7 Rh8 22 Qxg5+/= Rxh7 23 Nh3 Kc7 is very close to equality. (b) 16 ... b5 17 h5 (17 Nf3 Nf8 18 Rb1 Bd7=) 17 ... g5 18 fxg5 hxg5 19 h6, Van der Weide-Tereick, Senden 2006; and here Black should play 19 ... Nf8! 20 Qh5+ Kd8 21 Nh3 Nh7=. 17 h5?! 17 Rb1 b6 18 Nf3 Bd7 is fairly level; White lacks a plan, and Black only has hopes for ... g5. 17 ... g5 18 fxg5 hxg5 19 h6 g4 20 Qh5+ Here 20 ... Kd8! holds out chances for Black to gain a modest edge. 15.31311
...
Qa5
12
Nf3
This variation is a sort of main line in the MacCutcheon, although if Black’s play is as good as it seems, we’ll probably see White players looking in a different direction. Others are harmless: (a) 12 Ne2 Nc6 (not 12 ... c4?, when 13 Bxg6 Rg8 14 Bxf7+ Kxf7 15 Qh5+ is killing – White gets too many pawns and the attack) 13 dxc5 Qxc5 14 Qf4 Bd7 15 Qf6 Rg8 16 h5 g5 yields a complex position in which Black seems to be holding his own; e.g. 17 Rab1 b6 18 Rh3 Qe7!. (b) 12 h5 g5 13 f4 cxd4 14 Ke2! (14 Nf3?? dxc3+ 15 Ke2 Qa4-/+ Seirawan-Makarichev, Reykjavik 1990) 14 ... Qxc3 15 Nf3 f5! 16 exf6 e5 17 Bf5 Qxc2+ 18 Bxc2 Bxg4 19 fxe5 Nc6 20 Kf2 Kf7 with perhaps a minor edge for Black, but it’s complicated. 12 ... Nc6 13 Qf4 White should hurry to occupy f6 and thus prevent Black from castling. 13 dxc5 is slow after 13 ... Qxc5 14 Rhf1 Bd7 15 Qf4 0-0-0!? (or 15 ... Rc8) 16 Qxf7 g5! (16 ... d4!?) 17 hxg5 hxg5 18 Rab1 g4 followed by ... Nxe5. 13 ... cxd4 13 ... b6 is a reasonable option, but has favoured White in practice. 14 h5! Instead: (a) 14 Nxd4 Nxd4 15 Qxd4 Bd7 is calmer.
Then 16 Rhb1 has been called ‘+/=’ (16 h5 g5 17 Rhb1 Bc6), although Black has equal chances following 16 ... Qc7 (or 16 ... Rb8 17 Ke2 h5) 17 a4 (17 Ke2 b6 18 Kf1 Rc8 19 Rb3 0-0!? 20 h5 Kg7 21 Qf4 Qd8=) 17 ... Bc6 (17 ... b6 18 Ke2 0-0 19 Qf4 Qxc3
20 h5 f6!) 18 Rb4, Nyysti-I.Glek, Helsinki 2002; and 18 ... a5! 19 Rb3 Qe7 may have been slightly more accurate than 18 ... Qe7, although they’re both equal. (b) 14 Qf6 gives Black more options than the text; e.g. 14 ... dxc3+ (or 14 ... Qxc3+ 15 Ke2 Rf8) 15 Ke2 Rf8 16 h5 gxh5! (16 ... g5 17 Qxh6 transposes to 16 ... dxc3+ in the main line) 17 Rxh5 b6! 18 Rxh6 Ba6 19 Rh8 Bxd3+ 20 cxd3 Kd7 21 Rh7 Qb4!, threatening ... Qb2+, while 22 Rxf7+ Rxf7 23 Qxf7+ Ne7 is very depressing for White. 14 ... g5!? Black delays 14 ... dxc3+ 15 Ke2 or 14 ... Qxc3+ 15 Ke2 until he sees in what direction the play is heading. On the other hand, if ... dxc3+ is played later, White may gain options other than Ke2, so it might be more accurate to play it now; compare lines below, and especially the note to White’s 16th move.
15 Qf6 15 Nxg5!? is tricky, but ultimately unsatisfactory: 15 ... hxg5 16 Qf6 dxc3+ 17 Ke2 Rf8 18 h6 Qd8 19 Qg7, J.HitselbergerM.Koch, ICCF email 1998, and best is 19 ... Nd4+! 20 Kf1! Kd7 21 h7 Rh8 22 Qxf7+ Kc6, when Black should be able to consolidate; e.g. 23 Rh6 Bd7 24 Rf6 Qf8 25 Qg6 Qa3 26 Qg7 Raf8 27 Re1 Kc7 with the idea 28 Rf7? (28 Qxg5 Qe7) 28 ... Rxf7 29 Qxh8? Nxc2! 30 Bxc2 Qc5. 15 ... Rg8! 15 ... Rf8?! 16 Qxh6 g4!? 17 Ng5 Nxe5 18 Nh7 Rh8? 19 Ke2! led to a big White advantage in Anand-Shirov, Bilbao 2010. Here 18 ... Qxc3+! 19 Ke2 Nxd3! 20 Qxf8+ Kd7 21 Qxf7+ Kd6! improves, but White still has the better of things. 16 Qxh6 16 g4?! stops ... g4 and prepares Qxh6, but 16 ... dxc3+ (or 16 ... b6) 17 Ke1 (17 Ke2 Qa4 18 Rhg1 b6! 19 Qxh6 Ba6 20 Bxa6 Qxa6+ 21 Kd1 Qc4 and Black is penetrating too quickly for White to progress on the kingside) 17 ... Qa4! (17 ... b6 18 Qxh6 Ba6 19 Qf6 Bxd3 20 cxd3 Qb5 21 h6 Rf8 22 h7 Kd7 is also good, though it takes a lot more analysis to prove that) 18 Qxh6 Qf4 19 Qf6 Nxe5! reaps a clear advantage.
16 ... g4? The wrong order. 16 ... dxc3+! 17 Ke2 g4 18 Ng5 transposes below, without allowing the move 17 Nxd4! in the next note. Here 18 Ne1? d4 hits the e5-pawn. 17 Ng5 17 Nxd4! exploits Black’s inaccurate 16th move: 17 ... Nxd4 (17 ... Nxe5 18 Qf4) 18 Qf6 Nf5 19 Bxf5 exf5 20 h6 Be6 21 h7 Kd7 22 hxg8Q Rxg8 23 Rh7 Rf8 is strong for White after 24 Rb1! b6 25 Rb4 Qxa2 26 Rh8 Rxh8 27 Qxh8. 17 ... dxc3+ 18 Ke2 Nxe5
19 Qf6 19 Nh7!? is messy and balanced. Here’s a very abbreviated analysis, mainly of 19 ... Bd7:
(a) 19 ... Nd7 20 Nf6+ Nxf6 21 Qxf6 Qd8 is equal; e.g. 22 Qxc3 (22 Bb5+ Bd7 23 Bxd7+ Kxd7 24 Qxf7+ Qe7=) 22 ... d4! 23 Qc5 b6 24 Qc6+ Bd7 25 Qe4 f5 26 Qe5! Kf7! 27 h6 Qf6 28 Qc7 Qd8 29 Qe5 Qf6=. (b) 19 ... Bd7 is also equal: 20 Nf6+ Ke7 21 Nxg8+ (21 Qf4 Nxd3 22 cxd3 doesn’t improve, e.g. after 22 ... d4!? 23 Nxg8+ Rxg8 24 h6 Bb5! 25 h7 Bxd3+! 26 Kxd3 Qb5+ 27 Kc2 Qb2+ 28 Kd3 Qb5+=) 21 ... Rxg8 22 Qf4 Nxd3 (22 ... f6 23 h6 Rh8 24 h7 Nxd3 25 cxd3 e5 transposes into the repetition below) 23 cxd3
23 ... d4 (or 23 ... f6 24 h6 e5 25 h7 Rh8, which is drawn after 26 Qh6 Kf7 27 Qh5+ Kg7 28 Qh6+, and otherwise Black probably stands better) 24 Qxd4 (24 h6 Bb5 25 h7 Bxd3+! 26 Kxd3 Qb5+ 27 Kc2 Qb2+ 28 Kd3 Qb5+=) 24 ... Bb5 25 Qc5+ Kf6 and White has no better than 26 Qd4+ Ke7 27 Qc5+ repeating, since after 27 a3!? Rd8 28 Qb4+ Qxb4 29 axb4 Rxd3 30 Ke1 Bc6!? (30 ... Rd2 is also drawn) 31 Rh2 (31 h6 allows the beautiful 31 ... Bxg2!! 32 h7 Bf3 33 h8Q c2 and Rd1+–+; 31 Ke2 Bb5=) 31 ... Be4! 32 h6 Bh7 Black will at least hold the balance. 19 ... Qd8 20 Qxe5 This time 20 Nh7 can be met by 20 ... Nd7 21 Qxc3 (21 Bb5 a6=) 21 ... e5 (21 ... f5!?) 22 Kf1 Qe7 23 Re1 Rh8 24 h6 f6 25 Bg6+ Kd8 26 Qa5+ Nb6 27 Rd1 Kc7! 28 Rxd5 Bd7=/+. 20 ... Qxg5 Rather sadly, we are still following a couple of correspondence games. 20 ... Rxg5 is also equal: 21 Qxc3 d4 22 Qb4 (22 Qb2 Kf8 23 h6 Kg8) 22 ... a5 23 Qb2 Kf8 24 h6 Kg8= Kharlamov-Turkov, ICCF server 2007. 21 Qxc3 Bd7!? 21 ... Qh6 22 Rab1 f5!? 23 Qe5 Qg7 24 Qd6 Qe7 25 Qe5 ½-½ Garscha-Schubert, remoteschach.de 2008.
22 h6 e5 23 Rae1 White’s other tries also end in equality; e.g. 23 Qc5 a6!= (versus Bb5) 24 Kf1 (24 Qxd5? Bc6) 24 ... Rc8 (24 ... g3!? 25 h7 Rh8) 25 Qxd5 Bc6 26 Qa5! (26 Qb3?? g3) 26 ... g3 27 f3 Bxf3 28 Re1 f6 29 h7 Rh8 30 gxf3 g2+ 31 Kf2 gxh1Q 32 Rxh1, and either 32 ... Rd8 or 32 ... Qf4 will draw, but Black can’t seem to get more. 23 ... Bc6 Or 23 ... Qe7 24 Kf1 f6. 24 Kd1?! Better 24 Kf1 f6=. 24 ... f6! With a solid game and slight advantage, Frangi-Turkov, ICCF email 2002. 15.32
11
Nf3
11 ... Bd7 Glek’s move, with the idea ... Bc6 and ... Nd7 (made safer by ... Qe7). This spurred renewed interest in the MacCutcheon in the mid-1990s. It is still being debated. Black can choose other instructive set-ups; for example, 11 ... Nc6 (11 ... Qe7 12 h4 Bd7 transposes below – see note ‘b’ to Black’s 12th move) and then: (a) 12 h4 returns to the main line 15.311 above. (b) 12 Qf4 Qc7 (or 12 ... Qa5 13 Rab1!? Nxd4 14 Nxd4 cxd4 15 Qxd4 Qc7=) 13 h4 (13 Qf6 Rg8 14 h4 Qa5 15 h5 gxh5 16 Rxh5 cxd4 17 Rah1 dxc3+ 18 Ke2 Qd8!=) 13 ... f5!? (an idea worth remembering; now if Black can castle queenside, he can prepare ... g5; but to avoid dxc5, 13 ... c4! 14 Be2 f5! was better) 14 g4 (14 dxc5! Na5 15 Nd4 Bd7 16 Nb3+/=) 14 ... cxd4 15 cxd4, R.J.Fischer-Rossolimo, US Ch., New York 1965; and here Rossolimo’s suggestion 15 ... Bd7 was perhaps the most direct solution: 16 gxf5 gxf5 17 Rhg1 0-0-0 18 Rg6 is assessed in various sources as clearly better for White, but Black stands well enough after a number of moves; e.g. 18 ... Qa5+ 19 Ke2 Qb4=, or 18 ... Nb4 19 Rxh6 (19 a3 Nxd3 20 cxd3 Qb6! 21 Rc1+ Kb8 22 Rxh6 Qb2+ 23 Rc2 Qb1= intending ... Ba4; 19 Ke2 Qc3=/+) 19 ... Kb8 20 Rxh8 Rxh8 21 a3? (21 Ke2 Qc3 22 Rb1 Nxa2=/+) 21 ... Rc8! with a clear advantage. (c) 12 dxc5!? Qe7 13 Qf4 Qxc5 14 Nd4 (14 Qf6 Qxf2+ 15 Kc1 Qe3+ 16 Kb2 Qb6+ 17 Kc1 Rg8 looks safe enough) 14 ... Nxd4!? (14 ... a6 15 Rab1 Na5!? has also been tried; or 14 ... Bd7 15 Nb5 0-0 16 Nd6 Rab8 with complications – allowing a knight into d6 is often acceptable in the French Defence) 15 cxd4 Qa5+ 16 Ke2! b6 17 Qf6 Rg8 18 Rhc1 Ba6 19 Kf1 Rc8 should be satisfactory for both sides.
As we’ve seen, the set-up with White’s queen on f6 versus Black’s rook on g8, while optically nice for White, is not necessarily favourable for him, because the queen is as out of touch with the centre and queenside as Black’s rook is. You will see this in several of the MacCutcheon variations with ... g6. Returning to 11 ... Bd7:
12 h4 Others: (a) An interesting order is 12 Qf4 (also reached via 11 Qf4 Bd7 12 Nf3) 12 ... Bc6 13 h4 (13 Rhb1 can be answered by 13 ... c4!?, trying to block the position, or 13 ... Nd7, keeping things flexible; I prefer the latter) 13 ... Nd7 (13 ... Qe7 14 Nh2 Nd7 15 Ng4 0-0-0 was unclear in Amonatov-Skorchenko, Moscow 2008; the idea is 16 Nxh6 Rdf8 and ... f6) 14 Nh2?! (probably not best, but Black plans 14 ... 0-0-0 and ... f6 with good play; instead, after 14 Rh3 Qe7 15 dxc5 of Spraggett-Glek, Cappelle le Grande 1998, 15 ... Nxc5 was the simplest course, considering ... Ne4+; and 14 dxc5 Nxc5 15 Nd4 Ne4+ 16 Bxe4 dxe4 shows one good point behind having the knight on d7) 14 ... Qe7 (or 14 ... g5 15 hxg5 hxg5 16 Qe3 Qe7 17 Nf3 Rg8=; 14 ... Qa5 15 Ng4 0-00! 16 Nxh6 Rxh6! 17 Qxh6 cxd4 was also promising for Black in Bruzon-Glek, Esbjerg 2002) 15 Ng4
15 ... 0-0-0 16 Nxh6 (16 Nf6!? can be answered by 16 ... Nb6 or 16 ... c4 17 Be2 Qa3, among others) 16 ... f5 17 exf6 Nxf6 18 f3 cxd4! 19 cxd4 Ne4+! 20 Ke2 (20 fxe4 dxe4 21 Be2 Rxd4+ 22 Ke1 Qb4+ 23 Kf1 Rf8 24 Nf7 Rd7–+), Tseshkovsky-Skorchenko, Krasnodar 2002; and here 20 ... Rdf8! is strong; e.g. 21 Qe5 Rxh6 22 fxe4 Rh5 23 Qg3 dxe4 24 Bc4 Rhf5 25 Raf1 Qd7 26 c3?? Bb5. (b) 12 Rab1 c4! (12 ... b6!? 13 dxc5! bxc5 14 c4 Nc6 is unclear; and 12 ... Bc6?! 13 Bxg6! is supposed to be a draw, but I think White does better than that, which you can confirm with analysis). After 12 ... c4, Black gets to the standard set-up following 13 Be2 Bc6, and allows White no more than a draw if he sacrifices: 13 Bxg6 Rg8 14 Bxf7+ Kxf7 15 Qh5+ Kg7 16 Rxb7 Nc6 17 Qg4+ Kh8 18 Qh5, and Black can risk playing on with 18 ... Rg7 19 Qxh6+ Kg8, but it’s probably best to accede to 18 ... Kg7 19 Qg4+ Kh8 20 Qh5 Kg7=. (c) 12 dxc5 Qe7! 13 Rab1 (13 Rhb1 Bc6 14 Nd4 Nd7 15 Nb5!? 0-0! 16 f4 Nxc5-/+ Hnydiuk-W.Schmidt, Police 2011) 13 ... Bc6 14 Nd4 (14 h4 Nd7 15 Rhe1 Nxc5 16 Nd4 Rc8= Spassky-Relange, French Team Ch. 1991) 14 ... Nd7 15 Nxc6 bxc6 16 Rb7 Qxc5=. 12 ... Bc6
Here Black has some nice options: (a) 12 ... c4! 13 Be2 (13 Bxg6 Rg8 14 h5!? Qe7-/+) 13 ... Bc6 14 a4 transposes to the main line without allowing White’s possibility of 13 Bxg6 in the next note. (b) 12 ... Qe7 has been played occasionally. It squelches any idea of Bxg6 while keeping the option of ... Nc6 alive (and thinking about ... cxd4 and ... Qb4+ or ... Qa3). There can follow 13 Qf4 Bc6 14 Nh2 Nd7 15 Ng4 0-0-0!? 16 Nxh6 f5! (or 16 ...
Rdf8 and ... f5 or ... f6 with attack) 17 exf6 Nxf6 with good play, Sutovsky-Zifroni, Israeli Team Ch. 2000. 13 a4 Here 13 Qf4 transposes to 12 Qf4 above; but 13 Bxg6 fxg6 (13 ... Rg8?? 14 Qxe6+) 14 Qxe6+ is promising; e.g. 14 ... Qe7 15 Qxg6+ (15 Qc8+ Qd8 16 Qg4!?) 15 ... Qf7 (15 ... Kd8!? 16 e6 is unclear) 16 h5 Rg8 17 Qxh6, Rozkov-Belenov, Nizhnij Novgorod 2007. This is worth avoiding by 12 ... c4.
13 ... c4 14 Be2 Once again, the sequence 14 Bxg6 fxg6 15 Qxe6+ Qe7 16 Qxg6+ is possible, but now a4 is a lost tempo, so Black should be at least equal after 16 ... Kd8 17 e6 Na6 18 Ne5 Be8.
14 ... a5!? To prevent a5, although I think Black should ignore this and develop quickly by 14 ... Nd7 15 a5 0-0!?; or 14 ... Qa5, intending ... Nd7 and perhaps ... Kf8-g7 or ... 0-0-0. 15 Qf4 15 h5 g5 16 Nh2 has the idea of breaking with f4 and/or placing a knight on g4. This leads to unclear play. 15 ... Nd7 Black has equality. 16 Ra2 Apparently afraid of the fall of the a-pawn. After 16 Ra2, Wang Zili-Glek, Hefei 2010, continued 16 ... Qe7 17 Nh2 g5! 18 Qe3 0-0-0 19 h5 f5 20 exf6 Qxf6 (or 20 ... Nxf6) 21 Ng4 Qf5 22 Bf3 Rde8 23 Re1 Kc7 24 Re2 Kd6!, and ... e5 followed with some
advantage.
Chapter
Sixteen
King’s 1 e4 e6 The 16.1 16.2
King’s
Indian Indian
Attack
in
the
French 2 2
Defence
Attack begins
with
two
different
moves: d3 Qe2
16.1 2 d3
This introduces the King’s Indian Attack proper, so called because White sets up the same basic position as Black often does in the King’s Indian Defence to 1 d4. The sequence of White moves tends to be e4/d3/Nd2/Ngf3/g3/Bg2/0-0, with Re1 and/or c3 to follow. Over the years, this 2 d3 system has been used by some of the elite, notably Fischer (although he didn’t stick with it long, preferring the more dynamic 2 d4 and 3 Nc3). Today, however, few leading players employ the King’s Indian Attack (‘KIA’) consistently, or even much at all. Nevertheless, it is still played by some grandmasters, and is very popular at lower levels in spite of its passive reputation. Essentially that’s because the opening is easy to play, and most of the time you can make your first 7 or 8 moves without thought. There are perhaps six or seven fully established sound systems against the King’s Indian Attack, and I’ve proposed a number of them in the previous editions. Because theory never seems to change much except around the edges, and for reasons of overall space in this book, I’ve eliminated a system that I used over the first three editions (with ... d5/ ... c5/ ... Nc6/ ... Bd6/ ... Nge7), which is still completely valid, as is the Closed Sicilian-related variation fromPTF1andPTF2. I’ve kept the other three solutions fromPTF3, which require the KIA player to respond in unstereotyped fashion. Old readers will mark that I’ve also taken the liberty of repeating much of the analysis and much of the prose from the last edition. I suppose that’s what a ‘revised edition’ does once in a while. 2 ... d5 3 Nd2 Move order issues are important in the first few moves of the King’s Indian Attack. Here 3 Qe2 is discussed in 16.2 (about 2 Qe2). The move 3 g3?!, trying to save the move Nd2 and keep open Qg4, has a few potential drawbacks, but the main one is 3 ... dxe4 (or 3 ... Bd6!? 4 Qg4 g6 5 Qe2 Nc6; e.g. 6 Nf3 dxe4 7 dxe4 e5 8 Bg2 Nf6 9 0-0 Bg4 10 c3 h6 11 h3 Be6 and Black is at least equal) 4 dxe4 Qxd1+ 5 Kxd1 Nf6! (5 ... Bc5 also leaves White short of equality; e.g. 6 f3 Nc6 7 Nd2 e5=/+) 6 f3 (6 Nc3 Bb4!? 7 Nb5 Ba5 8 Bd2 Bb6 9 f3 0-0-/+) 6 ... Nc6 (6 ... Bc5=/+) 7 Be3 (7 Bb5 Bd7 8 Be3 0-0-0 9 Nd2 a6 10 Be2 Be7 11 c3 e5 12 Kc2 Be6=/+) 7 ... b6!? 8 Nd2 Bb7 9 Bb5 0-0-0=/+. 3 ... Nf6
4 Ngf3 4 g3 intends to leave White’s options open; for example, 4 ... c5 5 Ngf3 (or 5 Bg2 and 6 f4) takes out certain systems Black might want to play. He could try 4 ... dxe4 5 dxe4 e5 here; then the f1-bishop can still develop along the f1-a6 diagonal in conjunction with attacks on the e5-pawn, but at the cost of a tempo, g3 being a largely wasted move. That’s a good sequence for Black, though he has two other interesting ideas: (a) 4 ... Nc6 5 Bg2 (5 f4?! dxe4 6 dxe4 e5 and White has major weaknesses; 5 Ngf3 dxe4 6 dxe4 Bc5!; 5 c3 and Black has 5 ... e5= or 5 ... a5)5 ... dxe4 6 dxe4 e5
(a1) 7 Ne2?! is rather passive because it takes the pressure off e5. Black can play simply 7 ... Bc5 or try to force weaknesses by 7 ... Bg4 8 0-0 (8 f3 Be6 9 0-0 Bc5+ 10 Kh1 Qe7 11 Nb3 Rd8=/+) 8 ... Nd4 9 f3 Bc5 10 Kh1 Be6 11 c3 Nc6 12 Nb3 Bb6=/+. (a2) 7 Ngf3 is the position that Black has been looking for: White’s bishop is passive on g2 while both of Black’s bishops develop freely and, in the absence of Bb5, his e-pawn is simple to defend. See the instructive example of this position below via the order 4 Ngf3 Nc6 5 g3 dxe4 6 dxe4 e5 7 Bg2 (note ‘b2’ to White’s 5th move in section 16.13). (b) 4 ... b6 intends to occupy the a6-f1 diagonal once White plays Bg2, and just ... Bb7 if White decides to play passively and waste the g3 move: 5 Bg2 (5 Ngf3 dxe4 6 dxe4 Bc5) 5 ... Nc6 6 Ngf3 dxe4 7 dxe4 Ba6! 8 c3 (8 e5 Nd7; 8 c4 Bc5) 8 ... Qd7 9 Qa4 Bd3!=/+. After 4 Ngf3, I will investigate three set-ups introduced by these moves:
16.11 16.12 16.13
4 4 4
... ... ...
Bc5 b6 Nc6
16.11 4 ... Bc5
I have to admit that, many years ago, when I noticed a game with this move I thought it a very strange one one. But now with so many games and extensive experience with 4 ... Bc5, there still doesn’t seem any convincing reason to be wary about it. Black’s idea is a provocative one: the bishop stands extremely well on the a7-g1 diagonal, so White will either have to suffer its presence or advance his central pawns to e5 and d4, allowing Black to undermine them in traditional French Defence fashion. That is precisely what most King’s Indian Attack players are trying to avoid. 5 e5 The most important move. Others: (a) 5 g3? dxe4 6 dxe4 (6 Nxe4 Nxe4 7 dxe4 Bxf2+ 8 Kxf2 Qxd1 9 Bb5+ Qd7-/+) 6 ... Ng4, and White (who has fallen for this more times than one would possibly think) is objectively lost. (b) 5 Be2 dxe4 (5 ... Nc6!? 6 0-0 e5) 6 Nxe4 (6 dxe4!? e5, or Black can try the funny Philidor-like 6 ... Ng4 7 0-0 Bxf2+ 8 Rxf2 Ne3 9 Qe1 Nxc2 10 Qd1 Nxa1 11 b4! with unclear play) 6 ... Nxe4 7 dxe4 Qxd1+ 8 Bxd1 Nc6 9 Bf4 Bb6 (9 ... f6!? 10 Bxc7?! e5) 10 0-0 f6 11 Re1?! (11 e5=) 11 ... e5 12 Be3 Be6=/+ Gushpit-A.Ivanov, Barlinek 2002. (c) 5 d4 is almost never played
and may very well transpose to lines below after 5 ... Bb6 6 e5 Nfd7, while 5 ... Be7 6 e5 Nfd7 is a direct transposition to the main line of the 3 ... Be7 Tarrasch Universal System (3 Nd2 Be7 4 Ngf3 Nf6 5 e5 Nfd7) covered inPTF3, both sides having played an extra move. Apart from whether White thinks that desirable, the other problem is the independent line 5 ... dxe4!? 6 dxc5 exf3 7 Qxf3 Nc6, when Black’s free play makes for a level game, e.g. after 8 Ne4!? (8 Bd3 Ne5; 8 c3 0-0 9 Be2 e5) 8 ... Nxe4 9 Qxe4 0-0 10 Be2 (10 Bf4?! e5 11 Bd3 f5 12 Qc4+ Kh8=/+; 10 Bd3 f5 11 Qe3 e5=) 10 ... e5 11 0-0 Be6=. (d) 5 c3 Bb6!? 6 e5 Nfd7 7 d4 c5 also transposes to variations considered below: 8 Bd3 is note ‘c’ to White’s 7th move, while 8 Nb3 is the note (with 8 c3) to White’s 8th in the main line. Alternatively, 5 ... a5!? may well lead to 6 e5 Nfd7 7 d4 Be7 8 Bd3 c5 9 0-0 Nc6, an exact transposition to one of the main lines of the 3 Nd2 Be7 Tarrasch. Obviously that might not be to your taste, but it’s even less likely to appeal to your opponent, who not only needs to know the line, but feel comfortable with its highly tactical nature, including all the ... g5 ideas which appear in extremely similar form in Chapter 4. Otherwise 6 Be2 0-0 7 00 is passive enough that even 7 ... Be7 is satisfactory. Finally, 5 ... 0-0 6 e5 Nfd7 7 d4 Be7 8 Bd3 b6 intending ... Ba6 is also logical: the sooner Black can get ... Ba6 in, the better. 5 ... Nfd7
6 d4 Here White’s options are more serious: (a) 6 g3!? Nc6 (6 ... Bb6 7 Bg2 c5 is a sound sequence; e.g. 8 0-0 Nc6 9 Re1 Bc7 10 Nb3 Ncxe5 11 Nxe5 Nxe5 12 Nxc5 0-0=) 7 Nb3 Be7 (a fascinating line is 7 ... Bb6!? 8 d4 f6!? 9 c4! dxc4 10 Bxc4 fxe5! 11 Bg5 Nf6 12 Nxe5 Nxd4 unclear) 8 d4 f6!? (8 ... 00) 9 exf6 (9 Bh3 fxe5! 10 Bxe6 e4=/+) 9 ... Nxf6 10 Bb5 Bd7 11 Qe2 (11 0-0 0-0 12 Bxc6 Bxc6 13 Ne5 Qe8 14 Bf4 Bb5 15 Re1
Bd6=) 11 ... 0-0! 12 Bxc6 Bxc6 13 Qxe6+ Kh8 with sufficient compensation. (b) 6 c3 is subtle: where does Black wish to put his pieces?
(b1) 6 ... Bb6 7 g3?! (7 b4!? f6 8 d4 fxe5 9 Nxe5 Nxe5 10 dxe5 Qh4!? 11 g3 Qe7 12 Bd3 0-0= Bruno-Playa, Buenos Aires 1994; 7 d4 c5 transposes to the main line after 8 Bd3 or 8 Nb3) 7 ... f6 8 Bh3?! (8 exf6 Qxf6) 8 ... fxe5 9 Bxe6 Nc5! 10 Bxc8 Nxd3+ 11 Ke2 Nxc1+ 12 Rxc1 Qxc8-/+. (b2) 6 ... 0-0!? 7 d4 Be7 8 Bd3 c5 (you could argue that this can wait until later, and that 8 ... b6 9 0-0 Ba6 is a higher priority) 9 0-0. Black has castled early in this 3 ... Be7 Tarrasch Variation, but that may be alright because, for one thing, he has reserved the possibility of ... b6 and ... Ba6, whereas he would usually have ... Nc6 already in. Alternatively, 7 ... Bb6 is definitely still worth considering; e.g. 8 Bd3 c5 9 Nf1!? f5?! 10 Ng3 Nc6 11 Ne2 cxd4 12 cxd4+/= with a comfortable edge, KasparovRendle, London (simul) 1998. However, Black could gain valuable time here by 9 ... f6! 10 Qc2 f5!?, luring White’s queen onto a bad square, which looks satisfactory for Black (equal); or in this line 10 ... h6!, which seems to force 11 exf6 and therefore might be even better!
6 ... Bb6 That Black’s system is fundamentally sound is shown by the fact that 6 ... Be7 7 Bd3 (or 7 c3 c5) 7 ... c5 8 c3 transposes to a complicated main line of the 3 ... Be7 Tarrasch (covered inPTF3) in which Black has equal chances; for example, after 8 Bd3 Nc6 (8 ... b6!? 9 Nf1 Ba6) 9 0-0, and now 9 ... a5 and 9 ... g5 are two popular moves. Probably 6 ... Be7 is the ideal way to go against the average KIA player. 7 Nb3
A logical move to neutralize ... c5 and get the knight out of the way of its own pieces. Instead, 7 c4 c5 equalizes immediately, while the most popular move 7 Bd3 gives Black sufficient counterplay after 7 ... c5: (a) 8 c4 Nc6 9 0-0 cxd4 10 cxd5 exd5 11 Nb3 Nf8?! (11 ... 0-0 12 Re1 Re8 13 Bf4 Nf8!=/+) 12 Bg5 Qc7 13 a4!? Ne6 14 Re1 Bd7!?= Ree-U.Geller, Netanya 1968. (b) 8 dxc5 and: (b1) 8 ... Nxc5 9 Nb3 Nxd3+ 10 Qxd3 Nc6 11 0-0 f6 (11 ... 0-0 12 Be3 f6 13 Bxb6 axb6 keeps White out of c5 and equalizes; 11 ... h6 12 Re1 0-0=) 12 exf6 Qxf6 13 c3 0-0 14 Be3 (14 Bg5 Qf7 15 Rae1 Qh5!) 14 ... Bc7= 15 Rad1 h6 16 Nbd4 Rd8, planning ... e5. (b2) 8 ... Bc7!? 9 Nb3 Nxe5 (9 ... Nc6? 10 Bg5) 10 Nxe5 Bxe5 11 0-0 Nc6 12 c3 0-0 13 Be3 (13 Qe2 Bf6 14 f4 Re8=; 13 Bc2 g6 14 Be3 Bg7 15 f4 e5!? 16 fxe5 Nxe5 17 Bd4 Re8=) 13 ... Bc7 (13 ... Bf6 14 Qc2 g6 15 Rad1 Bd7=) 14 Qh5 g6 15 Qh6 Be5 16 Bg5 Qc7 17 Qh4 b6 18 Rad1 Bg7=. (c) 8 c3 Nc6 puts pressure on White’s centre and even favours Black. Here 9 dxc5 is a traditional White strategy, but overprotection of e5 is too difficult:
(c1) 9 ... Nxc5 has the idea 10 Bc2?! (but nothing else preserves White’s bishop very long) 10 ... Nd7 11 0-0 (11 Qe2 Qc7 12 Ba4 0-0 13 Bxc6 bxc6=/+) 11 ... Ndxe5 12 Nxe5 Nxe5 13 Re1 Qf6 (13 ... Bc7=/+) 14 Qe2, Andriasian-Goh Wei Ming, playchess.com 2005; and now instead of 14 ... 0-0? 15 Qxe5 Bxf2+ 16 Kh1 Bxe1 17 Qxe1, Black had the natural 14 ... Bc7 and White is a little short of compensation even after the best 15 Ba4+ Bd7 16 Bxd7+ Nxd7 17 Qb5 0-0-0. (c2) 9 ... Bxc5 10 Nb3 Bb6 11 Qe2?! (11 Bf4) 11 ... Qc7 12 Bf4 (12 Bb5 0-0 13 Bxc6 bxc6 14 0-0 a5! 15 Re1 Ba6=/+ Orsagova-Kiss, Rimavska Sobota 1996) 12 ... f6! 13 0-0!? (13 exf6?! Qxf4 14 fxg7 Rg8 15 Qxe6+ Ne7; 13 Nbd4 Ndxe5 14 Nxe5 fxe5 15 Nxc6 exf4=/+) 13 ... Ndxe5=/+ Dulik-Wrba, email 1997. 7 ... c5 8 dxc5 8 c3 Nc6 (alternatively, 8 ... cxd4 9 cxd4 Nc6 10 Bd3 f6 11 exf6 Nxf6 12 0-0 0-0 is a fairly standard Tarrasch position; e.g. 13 Be3 Bd7!? 14 Rc1 Be8 15 Nc5 Bxc5! 16 Rxc5 Bh5=; here 13 ... Qd6 intending ... a5-a4 is also interesting, and finally, the move ... Bc7 can be useful in any of these positions) 9 Be3 cxd4 (9 ... c4!? 10 Nbd2 f6 11 exf6 Nxf6 12 Be2 0-0=) 10 cxd4, KhalifehKristinsdottir, Istanbul Olympiad 2000; and instead of 10 ... Ba5+? 11 Nxa5 Qa5+ 12 Bd2 Qb6 13 Bc3 0-0 14 Bd3, Black had 10 ... 0-0 11 Bd3 f6 12 exf6 Qxf6=. 8 ... Nxc5
9 Be3 Others: (a) 9 Bg5 Qc7!=/+ (Kindermann). (b) 9 Nxc5 isn’t played much but seems to be the most challenging move: 9 ... Bxc5 10 c3!? (10 Bd3 Nc6 11 0-0 with the idea 11 ... Qc7!? 12 Re1 would favour White mildly; Black can try 11 ... f6 12 exf6 Qxf6, e.g. 13 c3 0-0 14 Bg5 Qf7 15 Bh4 Qh5 16 Bg3 Bd7, which is solid although still giving White a limited positional plus) 10 ... Nc6 11 Bd3 Qc7 12 Qe2 (12 0-0!?) 12 ... 0-0 13 0-0 (13 Bxh7+? Kxh7 14 Ng5+ Kg8 15 Qh5 Re8) 13 ... a5 14 Bf4 (14 Be3 b6) 14 ... f6 15 Rae1 (15 Bg3 f5 16 Bf4 Bd7=) 15 ... f5!? intending ... h6; or here 15 ... Bd7. 9 ... Qc7!? 9 ... Ncd7! improves. Again, White will struggle to defend e5 against a direct attack by ... Nc6 and ... Qc7; e.g. 10 Bd4! a5 11 a4 Bxd4 12 Nbxd4 Nc6 13 Nb5 0-0=/+. Thus 9 Nxc5 Bxc5 10 Bd3 is White’s best chance to retain an edge in this line. 10 Nxc5 Bxc5 11 Bxc5 Qxc5 12 c3 Bd7 13 Qb3 To prevent ... Bb5. Instead: (a) Kindermann analyses 13 Bd3 Bb5 14 0-0 Bxd3 15 Qxd3 Nc6=. (b) 13 a4! Nc6 14 Bd3 d4!? 15 0-0! dxc3 16 Rc1 and I’d rather be White here, although it’s not clear. 13 ... Nc6 14 Bd3
14 ... f6!? After 14 ... d4 a critical sequence is 15 Be4 (15 0-0 dxc3 16 bxc3=) 15 ... dxc3 16 Qxc3?! (16 Rc1! Qb4 17 Qxb4 Nxb4 18
Rxc3 Bc6=) 16 ... Qxc3+ 17 bxc3 Na5 (17 ... Rc8 may be better still) 18 Rb1 Rb8 19 Nd4 b6, and Black has at least levelled the play. 15 exf6 gxf6 16 Qxb7 16 Qc2 0-0-0 17 0-0-0 e5 18 Bf5 Ne7 with the advantage (Kindermann). 16 ... Rb8 17 Qc7 Rxb2 Black could apparently have at least drawn following 17 ... e5!?, which cuts the queen off from the kingside after 18 0-0 (18 0-0-0 Rc8 19 Qb7 Rb8 20 Qa6? e4 21 Rhe1 Bc8 22 Qa4 0-0) 18 ... Rc8 (18 ... e4 19 Rae1) 19 Qb7 Rb8 (19 ... e4 is a winning try) 20 Qa6? (20 Qc7 Rc8) 20 ... e4-/+. 18 0-0 Qxc3 19 Rad1 White had significant but unclear compensation in Kindermann-Sarana Hungeling, Bad Wiessee 2002. Here, instead of 19 ... Qa5, a possible continuation would be 19 ... Rg8 20 Bxh7 Rg7 21 Bd3 Qb4 22 Rb1 Qg4 23 Ne1 Rxb1 24 Bxb1 e5 25 Qd6 Qe6 26 Qc5 e4 or 26 ... d4. White has his pawn back, but Black’s centre is a plus. 16.12
4
...
b6
This is now a well-established solution to the KIA. Black’s c8-bishop can go to b7 or a6 depending upon what White does; in particular, the desired g3/Bg2 set-up will encourage an early ... Ba6. 5 c3 This is probably the move that most irritates players of 4 ... b6. Others are hugely instructive: (a) 5 g3 dxe4 6 dxe4 Bc5!? (note that 6 ... Nc6 is more accurate in one important line: 7 Bg2 Ba6! 8 c4 Qd3=/+ with the idea 9 Qb3? Nb4!-/+)
and: (a1) 7 Bd3 e5 8 0-0? Bh3 9 Re1 Ng4 10 Re2 Qf6 11 c3 g5! 12 Qe1 a5 13 b3 h5 14 a4 h4-/+ Rolletschek-Kindermann, Graz 2001. (a2) 7 Bg2 can be answered by 7 ... e5 or 7 ... Nc6, both equal. You should be aware that the attractive move 7 ... Ba6 (good after 6 ... Nc6 7 Bg2) isn’t appropriate here due to 8 c4 Qd3 9 b4!, when 9 ... Bxf2+ 10 Kxf2 Ng4+ falls a little short after 11 Ke1 Ne3 12 Qe2 Nc2+ 13 Kf2 Qxe2+ 14 Kxe2 Nxa1 15 Bb2 Nc2 16 b5 Bb7 17 Rc1. (a3) 7 Bb5+ Bd7 8 Bd3 Nc6 9 c3 (9 Nb3 Bd6 10 c3 0-0 11 h3 e5=/+ Bhat-G.Rey, San Francisco 2000) 9 ... e5 10 b4 Bd6 11 00 0-0 12 Qc2 Bg4 13 Be2 a5 14 b5 Nb8 15 Nc4 Nbd7 16 Rd1 Qe7 17 Nh4 Be6 (17 ... Bc5 18 Nf5 Qe6=) 18 Nf5 Bxf5 19 exf5 h6 20 Nxd6 cxd6 21 c4 Nc5, Morozevich-Dreev, Yalta (rapid) 1995. Black has achieved a powerful knight outpost on the open c-file. He went on to win, though in this position the bishops are an equalizing factor. (b) 5 e5 Nfd7 6 g3 (6 d4 c5 7 c3 transposes to the main line) 6 ... c5 is a standard ... b6 position that features both the attack on e5 and trade of light-squared bishops by ... Ba6. Black has done well here, mainly because the d2-knight has no good squares to go to and, in the meantime, it gets in the way of Bf4. For example: 7 Qe2 Qc7 (7 ... Nc6 8 Nb3 Qc7 9 Bf4 c4!? – 9 ... a5! is better – 10 dxc4 Ba6 11 Nbd4?! Bxc4 12 Qe3 Bxf1 13 Kxf1 Nxd4 14 Nxd4 Bc5 15 c3 0-0=/+ R.Montgomery-A.Shaw, corr 1983) 8 c4 Bb7 9 Bg2 dxc4!? 10 Nxc4 b5 11 Na3 a6 12 0-0 Nc6 13 Re1 Be7 14 Nc2 0-0 15 b3 Nb6=/+ Lushnikov-Grishanovich, St Petersburg 1999. 5 ... c5
Or 5 ... Be7 6 Qa4+!? Bd7 7 Qc2 c5 8 Ne5 Nc6 9 Nxd7 Qxd7=; White’s acquisition of the bishop pair comes at the cost of
space and development. 6 e5 Others: (a) 6 Qa4+ Bd7 (6 ... Qd7 7 Qc2 Qc7! 8 Be2 Nc6 9 0-0 Bd6 10 Re1 0-0=/+ C.Bauer-Bareev, Cannes rapid 2001) 7 Qc2 Qc7 8 Be2 Be7 9 0-0 0-0 10 Re1 Nc6 11 Bf1 Rac8=/+ Van Delft-Hertneck, Tegernsee 2003. (b) 6 g3 Be7! (6 ... Ba6?! 7 exd5+/= Short-Bareev, Sarajevo 1999) 7 Bg2 Ba6 (or 7 ... 0-0 8 0-0 Nc6 9 Re1 Qc7) 8 exd5 exd5=. (c) 6 Be2 Be7 (6 ... Qc7 7 0-0 Nc6 8 Re1 Bd6 is another valid set-up for Black) 7 0-0 0-0 8 Re1 Nc6 9 Qc2 Qc7 10 Bf1 Bb7 11 a3 a5= Kindermann-Klinger, Munich 1987. 6 ... Nfd7 7 d4
7 ... Be7 Alternatively: (a) 7 ... Ba6 poses different problems, since this comes before White can play Bb5 and a4 (as in our main line). For example: 8 Bxa6 Nxa6 9 0-0 Be7 10 Ne1 b5 (A battle between wing attacks begins, but 10 ... 0-0 11 Qg4 f5 is also reasonable) 11 Qg4 g6 (11 ... 0-0 12 Ndf3 Kh8=) 12 Ndf3 h6 13 h4!? Qb6 14 Be3 c4= Van Weersel-De Jong, Wijk aan Zee 2002. (b) I like the course of Chudnovsky-Bhat, Catonsville 2000: 7 ... Nc6 8 Bb5 (8 Bd3 Be7) 8 ... Bb7 9 0-0 a6 10 Ba4 g5!?; White is hard-pressed to defend his centre; there should follow 11 Nb1 g4! (11 ... Be7 12 Be3+/=) 12 Bxc6 Bxc6 13 Ne1 Qh4!?=. 8 Bb5 8 Bd3 is a more obvious development, but Black can get counterplay by exploiting the queenside light squares; for example: (a) 8 ... 0-0 (8 ... Nc6 9 0-0 g5!? would be the speculative way to go about things) 9 Ne4!? (9 Nf1 Ba6 10 Bxa6 Nxa6 11 Ng3 transposes; 9 0-0 Ba6 10 Ne1 Bxd3 11 Nxd3 Nc6 12 Nf3 c4!? 13 Nf4 b5 14 a3 Qc7 or here 12 ... Qc7) 9 ... Ba6 10 Bxa6 Nxa6 11 Ng3 cxd4!? (this should equalize; 11 ... Nab8 12 0-0 Nc6 is a surer way to initiate a queenside attack; e.g. 13 Qe2 a6 14 Bd2 b5 with the idea 15 a3 Na5=/+) 12 cxd4 Qc7 13 0-0 Rac8 14 Bd2 Qc4 15 a3 Nab8 16 Rc1 (16 Re1 h6 17 Nh5 Qb5 18 b4 Qd3) 16 ... Qb5 17 b4 Qd3 18 Qe2 Qxe2 19 Nxe2 b5 20 Nc3 a6 21 a4 Nb6! 22 axb5 Bxb4 23 bxa6 Nxa6 24 Nb5 Nc4=/+. (b) 8 ... Ba6 9 Bxa6 Nxa6 10 Qe2 (after 10 0-0 0-0 Black equalizes without problems; e.g. 11 Qe2 Nc7 12 Re1 Qc8 13 Nf1 Qa6 14 Bg5 Bxg5 15 Nxg5 Qxe2 16 Rxe2 Rfc8= ½-½ Smyslov-Hug, Zürich 1998) 10 ... Qc8 11 0-0 0-0 12 a3 Qb7=. 8 ... Ba6 It’s not at all clear that this is best once White chooses 8 Bb5 over 8 Bd3. Instead: (a) 8 ... 0-0 9 0-0 Bb7 10 Re1 Nc6 takes advantage of the fact that White’s bishop will have to waste time to get back of the b1-h7 diagonal. This looks playable if Black can get something going quickly on the queenside. (b) 8 ... a6 is particularly interesting:
9 Bd3 (9 Ba4 b5 10 Bc2 Nc6 11 Nf1 b4 12 Ng3 bxc3 13 bxc3, Orsag-Jurek, Czech League 1998; and Black can work on the weakened light squares by 13 ... cxd4 14 cxd4 a5 15 0-0 Ba6 16 Re1 0-0 followed by ... Rc8, ... Nb6 etc) 9 ... Nc6 10 0-0 g5!? (10 ... 0-0; or 10 ... Bb7, hoping for 11 Re1?! g5!=/+) 11 h3?! h5 12 g4 hxg4 13 hxg4 b5 (Black should connect rooks quickly: 13 ... Qc7! 14 Kg2 Bb7 15 Rh1 0-0-0=/+) 14 a4 (14 Kg2) 14 ... b4 15 Re1 a5? (15 ... Qb6!) 16 Bb5 Na7 17 Be2 Nc6 18 Bb5 Na7 19 Be2 ½½ S.Berry-N.Pert, British Ch., Liverpool 2008. A lot to think about in that short game!
9 a4 This move has taken over practice. Trading his good bishop on a6 gets White nowhere when it has taken two moves to do so. 9 ... cxd4 By no means forced. An important alternative is 9 ... Bb7!? 10 0-0 (10 Nf1 Nc6 11 Ng3 0-0 12 0-0 Rc8=) 10 ... Nc6 11 Re1
11 ... a6! (the best reply; 11 ... Rc8 with the idea 12 Nf1 c4 13 Ng3 Na5!? is an option) 12 Bd3 (12 Bxc6 Bxc6 13 a5!? 0-0 14 axb6 Qxb6 15 Nb3 a5; 12 Be2 Rc8= intending ... cxd4 and if cxd4, then ... Nb4) 12 ... g5! (the possibility of this move makes a case for delaying castling)
13 h4! (13 h3 h5 14 Nf1?! g4 15 hxg4 hxg4 16 N3h2 cxd4 17 cxd4 Nxd4 18 Nxg4 Bb4!, intending ... Qh4, forces 19 Re3 Nc6=/+) 13 ... gxh4 14 Nf1 Rg8 15 Kh1 h5 16 Bf4 Nf8 and Black has at least equal prospects in this messy position. In these last two notes, the move ... g5 has played a major role. 10 cxd4 0-0!? This is alright, but again, delaying castling has advantages; for example, 10 ... Qc8 11 Qe2 (11 Nf1!?) 11 ... Qb7 12 0-0 Bxb5 13 axb5 a6 14 Ne1?! (14 Nb1! axb5 15 Rxa8 Qxa8 16 Qxb5 0-0 17 Nc3 Rc8 18 Be3 Rc4=) 14 ... axb5 15 Rxa8 Qxa8 16 Qg4 (16 Qxb5 Qc6! 17 Qxc6 Nxc6 18 Nb3 f6 19 f4 Kf7=/+) 16 ... 0-0 17 Ndf3 Kh8 (17 ... Rc8! 18 Bh6 Bf8) 18 Nd3 Nc6 19 Nf4, Andriasian-Hou, Moscow 2010; and Black could keep a plus with 19 ... Rc8 20 Nh5 g6=/+. 11 0-0 Qc8 12 Re1 Nc6 13 Re3
13 ... Qb7 An example of what not to do: 13 ... Bxb5 14 axb5 Nb4 15 Rc3 Qb7 16 Nf1 a6 17 bxa6 Rxa6 18 Rxa6 Nxa6 19 Ng5! (Black has neglected the kingside) 19 ... Bxg5 20 Bxg5+/= f6? 21 exf6 Nxf6 22 Bxf6! Rxf6 23 Ne3 with a comfortable edge, M.AdamsBareev, Frankfurt (rapid) 2000. 14 Nf1 14 Qe2 Nb4= (Psakhis). 14 ... Rfc8 14 ... Na5!? gives Black some counterplay; compare the main line. 15 Bd2 Bxb5 16 axb5 Na5! 17 b3 17 Bxa5 bxa5 18 Rxa5 Bd8=. 17 ... a6 18 bxa6 Rxa6 19 h4!? Nc6 20 Rc1 Raa8 21 Rec3 M.Adams-Bareev, Sarajevo 1999; and it’s a good time to simplify by 21 ... Na7=, or 21 ... Ba3 22 R1c2 Nb4= (Psakhis). 16.13
4
...
Nc6
Here Black simply develops and fights for control of e5. If he were given a free move, 5 ... e5 would be a Philidor Defence with colours reversed, while 5 ... Bc5 would also be strong. 5 c3
This move (common in the Philidor) flexibly supports d4, prepares b4, and gives his queen a path to the queenside. Against moves other than 5 c3, Black will usually play for ... e5, with or without ... dxe5: (a) 5 Be2 Bd6! (a rare move which, however, has scored extremely well, with a 200 point performance rating lead for Black over White; the first idea is to cover e5, which means that the move ... e5 itself will be well supported) 6 0-0 (6 c3 0-0 7 00 transposes) 6 ... 0-0 7 c3 a5 (7 ... Re8 8 Qc2 e5 9 b4, K.Berg-Repkova, Aarhus 2009; and 9 ... a6 is equal, while 9 ... b6!? 10 Bb2 Bb7 is more dynamic, e.g. ... Ne7-g6 might follow)
(a1) 8 a4 b6 9 Qc2 Ba6 10 Re1 Re8 11 Bf1 dxe4 12 dxe4 Bxf1 13 Rxf1 Ng4 14 Nc4 Nge5= A.Lein-L.Portisch, Lone Pine 1978. (a2) 8 Qc2 e5 9 Re1 (9 exd5 Nxd5 10 Nc4 a4!?) 9 ... Re8 10 b3 Bg4 11 Nf1 d4! 12 c4 Bb4=/+ I.Ivanov-Castaneda, Tula 2006. (a3) 8 Re1 Re8 (8 ... h6 9 h3 a4!?) 9 h3 h6 (9 ... a4!) 10 b3 (a basic Philidor set-up, but it’s really designed for defensive purposes) 10 ... b6 11 Bf1 e5 12 a3, Baklan-Gdanski, Magdeburg 2000; and as an alternative to 12 ... d4=, Black could keep the tension by 12 ... Be6 13 Bb2 Qd7, also about equal. (b) 5 g3?! dxe4 6 dxe4 has two good replies: (b1) 6 ... Bc5 (threatening ... Ng4) 7 Bb5 (7 Bg2 e5 transposes to ‘b2’) 7 ... Bd7 8 0-0 Qe7 9 Qe2 e5 10 Nc4 h6 11 c3 a6 12 Bxc6 Bxc6 13 Nfxe5 Bxe4=/+. (b2) 6 ... e5, and here g3 weakens the light squares if White plays 7 Bc4 or 7 Bb5, but 7 Bg2 has drawbacks too, as we’ll see. This is an example of how having a tempo less can mean getting more information about what the opponent is doing. In this case, Black gets to see the move 5 g3 and react accordingly: 7 Bg2 (this invites Black to try to occupy the a6-f1 diagonal by ... b6 and ... Ba6, but putting the bishop elsewhere would leave his weakened kingside light squares undefended; e.g. 7 Bb5 Bc5 and if White plays 8 0-0, 8 ... Bh3 or 8 ... Bg4 is good; whereas 8 Bxc6+ bxc6 9 Qe2 Qe7 10 Nc4 Ba6 11 Nfxe5 0-0 intends ... Bd4, but 12 c3 Rfe8 13 Qf3 Nxe4-/+ is just as bad) 7 ... Bc5 8 0-0 0-0
and now: (b21) 9 c3 a5 10 Qc2 b6 (or 10 ... Be6 11 Ng5 Bd7=/+) 11 Nc4 Ba6=/+ 12 Rd1 Qe7 13 Ne3 Rad8! 14 Nd5 Nxd5 15 exd5 e4! 16 Re1 (16 Ne1 Ne5! 17 Bxe4 Ng4! – Fappas; 16 dxc6? exf3 17 Bxf3 Rxd1+ 18 Qxd1 Rd8 19 Bd2 Qf6-/+) 16 ... Bd3 17 Qb3 Qf6 18 dxc6 exf3 19 Bh1 Be2–+ Fargags-Fappas, Germany 1979. (b22) 9 Qe2 a5 10 Nc4 Qe7 11 Ne3 (11 c3 b6 and ... Ba6, when the queen is not happy on e2) 11 ... Nd4!? (11 ... Rd8=/+) 12 Nxd4 exd4 13 Nc4 Bg4 14 Qd3 Nd7 (14 ... Rfe8 is mentioned by Wintzer; then 15 e5 Nd7 16 Bxb7 Rab8 17 Bc6 Qe6 18 Bxd7 Qxd7 gives Black more than enough for a pawn) 15 a4 f6! 16 Bd2 b6 17 Kh1 (17 Rae1 Ne5=/+) 17 ... Rad8 18 Bf4? (18 f4 Rfe8=/+) 18 ... Bb4! 19 f3 (19 Qxd4?? Ne5) 19 ... Nc5 20 Qe2 Be6 21 Bc1 (21 Rfd1 d3 22 cxd3 Nxd3!-/+) 21 ... d3 22 cxd3 Rxd3–+ Matthaei-Wintzer, German League 1997. (c) After 5 e5 Nd7 6 d4, Black is a tempo up on the Guimard Variation (i.e. 2 d4 d5 3 Nd2 Nc6 4 Ngf3 Nf6 5 e5 Nd7), due to the loss of a time involved with d3-d4. Now Black can play slowly with 6 ... Be7, or directly assault the centre with 6 ... f6.
Then the play is balanced: (c1) 7 Bd3? fxe5 8 dxe5 Ndxe5 9 Nxe5 Nxe5 10 Qh5+ Nf7 11 Bxh7 Qf6 12 Nf3 Bb4+ 13 Kd1 Bd7-/+ J.Rigo-F.Portisch, Kecskemet 1979. (c2) 7 Bb5 fxe5!? (7 ... Be7 8 0-0 fxe5 9 dxe5 0-0 10 Re1 Nc5 11 Nf1, Tomba-Sumets, Milan 2009; and 11 ... Bd7 gives Black more than enough play) 8 dxe5 Be7 9 Bxc6?! (9 0-0 0-0 10 Re1=) 9 ... bxc6 10 Nd4 (10 0-0 0-0=/+ with the bishop pair and ... c5 to come) 10 ... Nxe5 11 Qh5+ Nf7 12 Nxc6 Qd6 13 Nxe7 Qxe7 14 0-0 0-0 15 Nb3 e5!=/+. (c3) 7 exf6 Qxf6! 8 Nb3 Bd6 (or 8 ... e5 9 dxe5 Ndxe5 10 Nxe5 Qxe5+ 11 Be2 Bf5=/+ and ... 0-0-0) 9 Bb5?! (9 Be2! 0-0 10 0-
0 b6! 11 c4 Bb7=) 9 ... e5! 10 dxe5 Ndxe5 11 Qxd5?! Be6 12 Qe4 0-0-/+. 5 ... Bd6!?
This flexible move addresses the dual issues of e5 and b4 while getting developed. Normally Black plays 5 ... a5 to stop b4, which is supposed to favour White, but that’s not clear either and deserves a close look: 6 e5 (6 Qa4 Be7!? 7 e5 Nd7 8 d4 00 9 Bb5!? Na7 10 Bd3 c5= Friedman-Shulman, Connecticut 2003; 6 Be2 Be7 7 0-0 0-0 8 Re1 e5=) 6 ... Nd7 7 d4
Again, Black is a full tempo up on the Guimard Variation that goes 2 d4 d5 3 Nd2 Nc6 4 Ngf3 Nf6 5 e5 Nd7 6 c3 a5, when it’s White’s move. Here it’s Black’s, and he seems to have time for the desirable 7 ... f6 and then: (a) 8 Bb5 fxe5 9 dxe5 (9 Nxe5?! Ncxe5 10 dxe5 c6 11 Bd3 Nxe5=/+) 9 ... Be7 10 0-0 0-0 11 Qe2 (11 Re1 Qe8 12 Nf1 Nc5 13 Be3 Bd7=), Kindermann-Farago, Budapest 1988; and Kindermann suggests 11 ... Qe8!? intending ... Qh5, which is equal, while 11 ... a4 intending ... Nc5 also looks fine. (b) 8 Nh4!? Qe7 (8 ... g6!? 9 exf6 Qxf6 10 g3 e5 11 Bg2 Nb6=) 9 Bd3? (9 Qh5+ Qf7 10 Qxf7+ Kxf7 11 f4 g5!; 9 exf6 Qxf6=) 9 ... fxe5! 10 Qh5+ Qf7 11 Bg6 hxg6 12 Qxh8 e4! 13 Nb1 (13 Nb3?! a4-/+; 13 Qh7 Ne7!) 13 ... Nf6 14 f3 Bd7 15 fxe4 dxe4 16 0-0 00-0 17 g3 e5 18 Be3 g5!–+ Ljubojevic-Vaganian, Belgrade 1974. (c) 8 exf6 Qxf6 9 Bb5 Bd6 10 Nf1 (10 0-0 0-0 intending ... Kh8 and ... e5) 10 ... 0-0 (10 ... e5 11 Ne3 e4 12 Nxd5 Qf5!? 13 Bxc6 bxc6 14 Ne3 Qf4 15 Nc4 Qf5 16 Ne3 Qf4=) 11 Ne3 Nb6!? (to guard d5 and prepare ... e5) 12 0-0 (12 Bd3? e5) 12 ... Bd7 (12 ... e5!? 13 Nxd5 Nxd5 14 Bc4 exd4 15 Bxd5+ Kh8 16 Bxc6 dxc3 is unclear) 13 Bd3 e5 14 Nxd5 Nxd5 15 Bc4 exd4 16 Bxd5+ Kh8 17 cxd4 Bg4=. 6 Qe2
Threatening e5. Black seems to consolidate his position in every case, usually with commonsense moves: (a) 6 Be2 0-0 7 0-0 transposes to 5 Be2 Bd6! 6 0-0 0-0 7 c3 above. (b) 6 g3 0-0 7 Bg2 and: (b1) 7 ... dxe4 8 dxe4 e5 (8 ... a5 9 0-0 e5 10 Qc2, Bologan-O.Danielian, Moscow 1995; 10 ... Bc5=) 9 Qe2 Bg4 (9 ... Be6=) 10 Nc4 h6 11 h3 Be6 12 0-0 Qd7 13 Kh2 Rad8 14 Rd1 Qc8=. (b2) 7 ... e5 8 exd5 (8 0-0 Be6= with the idea 9 Ng5 Bg4 10 Qb3 h6 11 exd5 Na5 12 Qc2 hxg5 13 b4 c6 14 bxa5 cxd5) 8 ... Nxd5 9 0-0, Mazi-Caruana, Trieste 2007; and here 9 ... Bg4 10 Re1 Re8= or 9 ... Bf5 10 Nc4 Nb6=. (c) 6 b4 has been used more often than anything else versus the rare 5 ... Bd6:
(c1) 6 ... a6 7 Be2 (7 Bb2 e5) 7 ... 0-0 8 0-0 e5 (8 ... Re8) 9 Qb3 Ne7!? (9 ... Be7!=) 10 d4 (10 c4!+/=) 10 ... dxe4 11 Nxe5 Be6 12 Bc4 Ned5=/+ Ansell-Kosten, British League 2005. (c2) 6 ... e5 7 exd5 (7 b5 Ne7 8 Bb2 0-0=/+; 7 a3 0-0 8 Be2 Ne7 9 0-0 Ng6, Sadvakasov-Abdennabi, Dubai 1999) 7 ... Nxd5 8 b5 (8 Bb2 0-0 9 Nc4 Bg4=) 8 ... Nce7 and Black has no problems. 6 ... e5 7 exd5!? Surrendering the centre. Since 7 b4 0-0 8 b5 Ne7 achieves less than nothing, White should prefer 7 g3 0-0 8 Bg2 Be6=. 7 ... Nxd5 8 Nc4 0-0 9 Nxd6!? It’s hard to know what to do, since it’s likely that Black already has an edge; e.g. 9 g3 Bf5! 10 Nh4 Be6 11 Bg2 Nf4!=/+. 9 ... Qxd6 10 Bd2 Bf5 11 0-0-0 Rad8-/+ With excellent play against the backward d-pawn. It seems to me that Black has sufficient play after 4 ... Nc6, and in particular both 5 ... a5 and 5 ... Bd6 should hold their own with correct play. 16.2
2
Qe2
An old move used by Chigorin back in the 19th century; 2 Qe2 was revived by Morozevich and others, and today it is as popular among strong players as 2 d3 itself. I don’t think that 2 Qe2 poses many problems for Black, but you’ll undoubtedly run into it, perhaps from the same players who used to play 2 d3 ! The first point of 2 Qe2 is to meet 2 ... d5 with 3 exd5 Qxd5 4 Nc3. Otherwise White tends to enter a King’s Indian Attack set-up where the queen is oddly placed except for lines with fairly early e5 thrusts. One problem for White is that Black can choose systems in which e5 is ineffective. I’ll examine two of these: 16.21 16.22
2 2
... ...
Nf6 c5
Before moving on to those, let me discuss the related order 2 d3 d5 3 Qe2.
In a sense, this ‘spoils’ the obvious idea behind 2 Qe2, which as explained above is to answer 2 ... d5 with 3 exd5 and force Black’s queen to recapture on d5. Presumably White’s idea is to wait for ... d5 before playing Qe2, because otherwise he’s worried about 2 ... c5 without ... d5 (Sicilian-style). After 3 Qe2: (a) I mentioned inPTF3that Black might be able to upset the apple cart by 3 ... Bc5!?, the idea being that White will have to waste another move with the d-pawn to kick the bishop off its excellent diagonal; for example, 4 Nf3 (4 exd5 Nf6! is more
fun than 4 ... Qxd5=) 4 ... Ne7!? (my further suggestion, now used by all of two players in my database!; obviously 4 ... Nf6 is perfectly playable, and has been used at least eight times – compare 3 Nd2 Nf6 4 Ngf3 Bc5 in 16.11 above) 5 d4 (5 g3 0-0 6 Bg2 Nbc6!? 7 0-0 Bb6 has the idea ... f6 and/or ... e5, as appropriate; 5 e5 0-0 6 g3 is well met by 6 ... Bb6 7 Bg2 c5) 5 ... Bb6!; e.g. 6 e5?! (6 c3 c5 7 dxc5 Bxc5=) 6 ... 0-0 7 g3 c5! 8 dxc5 Bxc5 9 Bg2 b6! intending ... Ba6 and Black has a significant advantage. (b) A saner solution is 3 ... Nf6 (why not?) 4 Nf3 Be7 5 g3, and McDonald recommends the clever move order 5 ... b5! 6 e5 (6 Bg2 dxe4 7 dxe4 b4! 8 c4 bxc3 9 bxc3 Ba6 10 c4 Nc6 11 0-0 Nd4! 12 Nxd4 Qxd4=/+ Illner-Kosic, Budapest 2007) 6 ... Nfd7 (as McDonald points out, a useful transposition is 2 Qe2 Nf6 3 Nf3 d5 4 e5 Nfd7 5 d3 Be7 6 g3 b5) 7 Bg2 c5 8 h4 Nc6 9 0-0 a5 10 Bf4 b4 11 Nbd2 Ba6 (McDonald: ‘A typical scenario: White is all dressed up for a kingside attack but his opponent is refusing to play ball by putting his king on g8’) 12 c4 bxc3 13 bxc3 (Black already stands better) 13 ... c4!? (I’m not thrilled with this; 13 ... 00 and 13 ... Nb6 are sound and good) 14 d4 a4! 15 Rac1? (15 Ng5! is very complicated – McDonald) 15 ... h6! 16 Qe3 Rb8 17 Rb1 Bb5 18 Bh3 a3 19 h5 Ba4! 20 Rxb8 Ncxb8 21 Rb1 Bc2 and with the ‘bad’ bishop now very good indeed, Black had a substantial advantage in Schunk-Meier, Altenkirchen 2005. 16.21
2
...
Nf6
I think that this has been underestimated: Black prepares ... d5 and develops. To make sense of 2 Qe2 White has to commit his centre, arguably too early. 3 e5 White has numerous logical options: (a) After 3 d4, Black can of course play 3 ... d5 (and even the gambit 4 exd5 Be7!), but I also like 3 ... c5!
with: (a1) 4 d5 d6 5 c4 exd5 6 cxd5 Be7 (6 ... g6 aims for a Benoni with the queen misplaced on e2; then 7 Nc3 Qe7 8 Nf3 Bg7 9 Qc2 0-0 10 Bd3 a6 11 a4 Nbd7= is normal) 7 Nc3 0-0 8 Nf3 Nbd7!? (8 ... Re8; 8 ... Bg4) 9 a4 Re8=. (a2) 4 e5 Nd5 5 c4 (5 Nf3 cxd4 6 Nxd4 d6 7 Nf3 Nc6=/+) 5 ... Nb4!? 6 dxc5 (6 Nf3 cxd4 7 Nxd4 N8c6=/+) 6 ... Bxc5 7 Nf3 N8c6 8 Nc3 (8 a3?! Qa5 9 Bd2?? Nc2+ 10 Kd1 Qa4 11 Nc3 Qb3–+) 8 ... Qa5! 9 a3 Nd4 10 Nxd4 Bxd4 11 Bf4 Nc6 and e5 is under attack three times, although 12 Qd2! Bxe5 (or 12 ... Bxc3 13 Qxc3 Qxc3+ 14 bxc3 b6!=) 13 Bxe5 Qxe5+ 14 Be2 0-0 15 Rd1 gives just about a pawn’s worth of compensation. (b) 3 Nf3 is the main line, and the following play is instructive: 3 ... d5 4 e5 Nfd7 5 d3 (5 d4?! c5 6 c3 b6 – Rustemov; or 5 g3 b6 6 Bg2 Bc5!? 7 d3 Nc6 8 c3 a5 9 0-0 Ba6 10 Re1 Be7 11 Qd1 Nc5 12 Bf1 d4! 13 c4 Bb7=/+ Troianescu-Cappello, Bari 1971; actually, I think 5 ... c5 is easier and probably better than 5 ... b6, but at any rate the latter is a nice idea) 5 ... c5 (or 5 ... Be7 6 g3 b5; see 2 d3 d5 3 Qe2 Nf6 above) 6 g3 Nc6 7 Bg2 Be7 (7 ... b5 is a good option)
(b1) 8 0-0 g5! has been used repeatedly with success; then ... g4 is threatened, to win the e-pawn, and Bf4 is prevented. Consequently: 9 h3 h5 10 g4 hxg4 11 hxg4 f6!-/+ (11 ... Qc7 12 Re1 f6! 13 exf6 Nxf6=/+); e.g. 12 exf6 Nxf6 and it’s suicidal to grab the g-pawn because of Black’s attack down the h-file. (b2) 8 h4! (preventing ... g5) 8 ... h6 (8 ... f6 9 exf6 Nxf6 10 Bf4 Bd6 11 Bxd6 Qxd6 12 0-0 0-0 13 Re1, Lin-Ding, Los Angeles 2008; and 13 ... Bd7!=/+ was best; here 8 ... b5 transposes to 2 d3 d5 3 Qe2 Nf6 again) 9 c4 (Rustemov assesses 9 Bf4!? b5 10 c4 bxc4 11 dxc4 Ba6 12 Nbd2 as unclear, although the aggressive 12 ... Nb4! looks strong, with the idea 13 0-0 g5! 14 Be3 d4; here 10 h5 Ba6 11 Nbd2 Qb6 is already better for Black, Kernazhitsky-Jasny, Tatranske Zruby 2011) 9 ... dxc4 (9 ... 0-0 is less
committal and at least equal) 10 dxc4 Nd4 11 Nxd4 (Rustemov mentions 11 Qe4 with a slight advantage, but Black stands well after 11 ... 0-0!, since 12 Nxd4 cxd4 13 Qxd4 is met by 13 ... Qa5+!, and 12 0-0?! f5! 13 exf6+ Nxf6 14 Qd3 e5 opens lines for a powerful attack) 11 ... cxd4 12 Bf4 (12 0-0 Qc7 13 Re1 0-0 14 Nd2 Rd8 15 Nb3 d3! 16 Qxd3 Nxe5 17 Qe4 Nd3 18 Rd1 Nxc1 19 Raxc1 Rxd1+ 20 Rxd1 a5=) 12 ... Qb6! 13 Nd2 (13 0-0 g5 14 hxg5 hxg5 15 Bc1 Nc5; 13 b3 Qa5+!? 14 Nd2 Nc5=/+) 13 ... Nc5 (or 13 ... Qxb2 14 Rb1 Qc2) 14 Rb1 a5 15 h5 (15 0-0? g5–+) 15 ... Bd7 16 0-0 0-0 17 b3 (Rustemov offers 17 a3?! a4-/+; 17 Rfd1 Rfd8=/+; 17 Ne4 Bc6 18 Nxc5 Bxc5 19 Be4 Bxe4 20 Qxe4 a4=/+) 17 ... Rfd8 18 a3 Bc6=/+ Fedorov-Rustemov, Vilnius 1997. (c) 3 f4 doesn’t co-ordinate well with the queen on e2; for example, 3 ... d5 4 e5 (4 d3 can be met by 4 ... dxe4 5 dxe4 Nc6 having ... Bc5 in mind, or by 4 ... b6, when the idea of ... dxe4 and ... Ba6 forces White to do something about his centre) 4 ... Ng8!? (4 ... Nfd7 5 Nf3 c5) 5 Nf3 Nh6!? (or 5 ... c5) 6 Nc3 c5= 7 Nd1?! Nc6 8 g3 Nf5 9 c3 Be7 10 Bh3 Rb8 11 d3 b5 with a queenside attack underway, Khouseinov-Villamayor, Bled Olympiad 2002. (d) 3 g3 d5 4 d3 has the drawback of allowing 4 ... dxe4 5 dxe4 e5!. A more conventional position can arise from 4 ... Be7 5 f4 c5 6 Bg2 Nc6 7 Nf3 0-0 8 0-0; for example, 8 ... b5!? 9 e5 (9 Nc3 b4 10 Nd1 dxe4 11 dxe4 a5, planning ... Ba6) 9 ... Nd7 10 c3 (10 c4 bxc4 11 dxc4 Rb8=/+) 10 ... Nb6 11 d4 b4=/+ with a rapidly-developing queenside attack, L.Kovacs-Uhlmann, Sarajevo 1969. 3 ... Nd5 4 Nf3 Instead: (a) 4 d4 c5 (4 ... d6=) transposes to 3 d4 c5 4 e5 Nd5. (b) 4 c4 Nf4!? (4 ... Nb4 5 d4 c5 6 d5!? exd5 7 cxd5 Nxd5 8 Qe4 Nb6 9 Nf3 Be7 10 Nc3 Nc6 11 Be3! 0-0 12 0-0-0 d5 13 exd6 Bf6 14 Bxc5 Re8 15 Qc2 Be6 and now Black is the one a pawn down, but has ... Rc8 coming with clear compensation) 5 Qe4 Ng6 6 Nf3 (6 h4 Nc6 7 d4 d5 8 exd6 f5 9 Qd3 Qxd6 10 Nf3 e5=) 6 ... d5 (or 6 ... c5 7 Nc3 Nc6 8 Be2 b6=) 7 exd6 Bxd6 8 d4, Motwani-Nogueiras Santiago, Istanbul Olympiad 2000; 8 ... 0-0 9 Nc3 e5! 10 Bg5! Qe8 11 0-0-0 exd4 12 Nxd4 Na6=. (c) 4 g3 d6 5 d4 c5 6 c3?! cxd4 7 cxd4 b6! (7 ... Nc6 8 Nf3 dxe5 9 dxe5 Bc5=/+) 8 Bg2 Ba6-/+ Arques Lopez-Barria, Alicante 2000. 4 ... c5!
5 c4 5 g3 Nc6 6 Bg2 d6=, while 5 Nc3 Nc6! 6 Nxd5 exd5 is a very nice version of a Nimzowitsch Sicilian (1 e4 e5 2 Nf3 Nf6). 5 ... Nb4! 6 d3 d5 7 Qd1! 7 a3? Qa5! 8 Nc3 d4 9 Rb1 dxc3 10 axb4 Qa2. 7 ... dxc4 7 ... Be7=/+. 8 dxc4 Qxd1+ 9 Kxd1 Nd7 10 Nc3 Be7 11 a3 Nc6 12 Bf4 0-0= 16.22 2 ... c5 A good reply: Who wants the queen on e2 in a Sicilian? The position can go in many directions and I will look at some of the most common ideas. 3 Nf3 Or 3 f4 Nc6 4 Nf3
when two good solutions are: (a) 4 ... Be7 5 Nc3 (5 g3 d5 6 d3 Nf6 7 Bg2 was looked at briefly in note ‘d’ on 3 g3 in 16.21 above) 5 ... d5 6 d3 (6 e5 Nh6 intending ... Nf5) 6 ... Nf6 7 e5 Nd7 8 g3 b5! 9 Bg2 Qb6 10 0-0 b4 11 Nd1, Lendwai-Uhlmann, Graz 1991; and the straightforward 11 ... 0-0=/+ with the idea ... Ba6/ ... Nd4 is a good course. (b) 4 ... g6 5 g3 Bg7 6 Bg2 Nge7 7 c3 (7 d3 0-0 8 0-0 d6 9 Nc3 is a Closed Sicilian in which Qe2 isn’t usually in at this point; if nothing else, that means that a plan with Be3 and after ... Nd4, Bf2 followed by Nxd4 is not possible, because the queen on e2 will be en prise) 7 ... 0-0 8 0-0 d6 (or 8 ... d5 intending ... b6) 9 Na3 Rb8 10 Kh1 a6 (10 ... Qd7 11 Nc2 b6 12 d4 Bb7=) 11 Nc2 b5 12 d4 cxd4 13 cxd4 b4! 14 b3 a5 15 Bb2 Qb6 16 Rfc1 Ba6 17 Qd1 d5! 18 e5 Rfc8=/+ Onischuk-Shaked, Tilburg 1997. 3 ... Nc6 4 g3 g6 5 Bg2 Bg7 6 0-0 Nge7 7 d3
7 ... e5 7 ... 0-0 is another common continuation, with either ... d5 or ... e5 to follow. If White plays 8 Nc3, Black has 8 ... d6, with a standard position except for the strange and committal Qe2; while 8 c3 can also be met by 8 ... d6, e.g. 9 Nbd2 h6 10 a4 e5! 11 Nc4, transposing to the main line. 8 Nbd2 0-0 9 Nc4 d6 10 a4 h6 11 c3 Be6 Black is well centralized and White has no breaks, whereas the prospects for ... d5 and/or a kingside attack are good. 12 Re1 12 Rd1 f5! threatens ... f4 and 13 exf5 Nxf5 clamps down on d4 while preparing the attack. 12 ... Qd7 13 Nfd2 Rae8 14 Nf1 f5 15 f4?! exf4 16 Bxf4
an
Van der Weide-J.Watson, London 1979; and here the easiest course is 16 ... Bxc4! 17 dxc4 g5 18 Bc1 f4, obtaining e5 as outpost; Black can follow with ... Ng6.
Chapter
Seventeen
Uncommon
Early
Moves
1 e4 e6
In this chapter I’ll examine White’s irregular second and third moves. One advantage of the French Defence is that most of White’s moves other than 2 d4 are harmless. The following four options are all fully playable, but only 2 Nf3 d5 3 Nc3 and 2 Nf3 d5 3 e5 present a serious challenge to Black: 17.1 17.2 17.3 17.4
2 2 2 2
d4 d4
d5 d5
3 3
b3 Nf3 Be3 Bd3
The remainder of the second moves are given here, and third move alternatives in the first note to 17.3. It’s worth mentioning that Black generally has two or more good solutions in addition to the one I’ve presented: (a) 2 e5 c5 (2 ... d6 3 exd6 Bxd6 4 d4 Nf6 5 Nf3 0-0=, and here 3 ... d6 is also played, giving Black a central majority; 2 ... d5 3 Nf3 is given under the order 2 Nf3 d5 3 e5 in 17.22) 3 f4 Nc6 4 Nf3 Nh6 (4 ... d6=) 5 g3, Steinitz-J.Mason, Vienna 1892; and here 5 ... Nf5 6 Bg2 d6=/+ is simple enough. (b) 2 Bb5 c6 3 Ba4 d5 4 Qe2 Qa5!? 5 Bb3 Nf6 6 e5 Nfd7 7 c3 c5 8 d3 Nc6 9 f4 Nd4! 10 Qd1 Nxb3 11 Qxb3 c4! 12 dxc4 Nc5=/+ Wahls-Vaisser, Berlin 1988. (c) 2 f4 d5 (naturally, 2 ... c5 is also fine) 3 e5 (3 Nc3, and 3 ... d4 4 Ne2 c5= or 3 ... dxe4 4 Nxe4 Nf6=) 3 ... c5 4 Nf3 Nc6 5 c3 (5 c4 d4 6 d3 Nh6=/+) 5 ... d4 (or 5 ... Nh6 intending 6 d4 Qb6) 6 d3 Nh6= M.Weiss-Maróczy, Budapest 1895. I think that Black’s position is easier to play, and White doesn’t seem to have a good way of making 2 f4 interesting. (d) 2 c4 d5 (2 ... c5 is a tame variation of the Sicilian Defence, and one reason not to play 2 c4) 3 exd5 (3 cxd5 exd5 4 Qa4+ Bd7 5 Qb3 Nc6!=/+ 6 Qxd5?! Nf6 7 Qc4 Be6 8 Qa4 Bc5-/+ –ECO) 3 ... exd5 4 cxd5 (4 d4 is an Exchange Variation, section 12.3) 4 ... Nf6 5 Nc3 (5 Qa4+ Nbd7 6 Nc3 Bd6! 7 Bc4 0-0 8 d3 a6 9 Qd1 b5 10 Bb3 Nc5=/+; 5 Bb5+ Nbd7 6 Nc3 Be7 7 Nge2!? 0-0 8 0-0 a6 9 Bxd7 Qxd7 10 Nf4 b5 11 Qf3 Bb7 12 a3 a5-/+ Grob-H.Johner, Zürich 1941; here 7 d4 and 8 Nf3 is better and equal, but a move earlier, 6 ... a6 7 Ba4 b5 8 Bb3 Bb7 is a good alternative) 5 ... Nxd5 6 Bc4 Nb6 7 Bb3 Nc6 8 Nge2 (8 Nf3 Qe7+ 9 Qe2 Bg4!=/+) 8 ... Bd6 (8 ... Be6!) 9 d4 0-0 10 0-0 Qh4! 11 g3 Qh5 with a double-edged game, Masserey-Kindermann, Horgen 1995. (e) 2 g3 d5 3 Nc3 (3 Bg2 c5=, or 3 ... dxe4 4 Nc3 Bd7!= and ... Bc6, or 3 ... Nf6 4 e5 Nfd7 5 d4 c5 6 c3 Qb6 7 Nf3 Nc6 8 0-0 Be7= intending ... 0-0, ... f6 –ECO) 3 ... dxe4 (or 3 ... Nf6 4 e5 Nfd7 5 f4 c5 6 Nf3 Nc6 7 Bg2, and 7 ... Be7 or 7 ... a6 8 a4 Be7 9 0-0 0-0 10 d3 Nd4 11 Ne2 Nxf3+ 12 Bxf3 f6= Vavra-Totsky, Pardubice 1995) 4 Nxe4 Nf6 (4 ... Bd7 5 Nf3 Bc6 6 Qe2 Nf6= BartschReefschläger, German League 1981) 5 Nxf6+ Qxf6 6 Bg2 Bc5 7 Qf3 Qe7! 8 Ne2 e5= Suttles-Uhlmann, Palma de Mallorca Interzonal 1970. (f) 2 Ne2 d5 3 exd5 exd5 4 Ng3 Nf6 (4 ... Qh4!? –ECO) 5 d4 Bd6 6 Bd3 0-0 7 0-0 Re8= Tartakower-Fine, Kemeri 1937. (g) 2 Nc3 d5 either transposes to main lines after 3 d4, or to 2 Nf3 below after 3 Nf3 (see 17.21), or to note ‘e’ above after 3 g3. 17.1 2 b3 A safe move, but b3 and e4 do not mix very well.
2 ... d5 This meets the challenge head on. I’ve promoted and analysed 2 ... b6 in all previous editions, and it’s a good solution; for example, 3 Bb2 Bb7 4 Nc3 (the e-pawn is awkward to defend: 4 d3 d5, or 4 Qe2 Nf6 5 Nc3 c5 and ... Nc6) 4 ... c5 5 Nf3 Nc6 6 d4 (the most serious attempt, since otherwise Bb2, Nc3 and e4 don’t go together) 6 ... cxd4 7 Nxd4 Nxd4 8 Qxd4 Nf6 9 f3 (9 Be2 Bc5 10 Qd3 0-0) 9 ... Bc5 10 Qd2 0-0 11 Na4 Be7, Bury-J.Watson, Leominster 1977. Black has the idea 12 0-0-0 Nxe4! and is better developed. 3 Bb2 3 exd5 exd5 simply frees Black’s pieces. 3 ... dxe4 Biting the bullet; in a previous edition I chickened out with 3 ... Nf6. 4 Nc3 Nf6 5 Qe2 5 g4?! is weak after 5 ... Nc6! 6 g5 (6 Bg2 Be7 with an edge, or 6 ... e5! 7 g5 Bg4!) 6 ... Nd5 7 Nxe4 and:
(a) 7 ... h6! 8 g6 (8 gxh6 Rxh6=/+) 8 ... f5! 9 Nc3 (9 Ng3 Qd6 10 Nf3 Bd7 11 d4 0-0-0-/+ Chichirivichi-Pawntobewild, playchess.com 2007) 9 ... e5! 10 Bb5 Nxc3 11 Bxc6+ (11 Bxc3 Qd5) 11 ... bxc6 12 dxc3 Bb7! and ... c5 – Moskalenko. (b) 7 ... e5 8 Nf3 (8 Ne2? Bg4-/+ Arvola-Westerinen, Tromsø 2008) 8 ... Bd6 9 Qe2 0-0=/+. 5 ... Be7 5 ... Nc6!? is a perfectly good option: 6 Nxe4 (6 g4? Nd4 7 Qd1 e5-/+; 6 0-0-0 a5!? and 7 a4 Be7 or 7 Nxe4 a4 8 Nxf6+ gxf6 9 Qf3 axb3 10 axb3 Ba3!) 6 ... Be7 (6 ... Nxe4 7 Qxe4 Qd5!?) 7 Nf3 0-0 8 0-0-0 a5!? 9 d4 (9 a4 Nd5!) 9 ... a4 with a queenside attack, Karasev-Farago, Polanica Zdroj 1974. 6 0-0-0 6 g4 Nc6 7 Nxe4 runs into 7 ... Nb4!=/+. 6 ... 0-0
7 g4 White should settle for 7 Nxe4, but Black has easy equality (at least) after 7 ... a5 (or 7 ... Nc6 8 Nf3 a5 with the idea 9 a4 Nd5) 8 a4 Nbd7 (8 ... Nd5 9 Nf3 b6!?) 9 Nf3 b6! 10 Nxf6+ Bxf6 11 d4 Bb7 with equal prospects. 7 ... a5!? Not bad, but I really like exploiting the weakness of the g4 plan by 7 ... Nc6! 8 Nxe4 (8 g5 Nd5 9 Nxe4 Bxg5) 8 ... Nd4 9 Qd3 e5 10 h3 Nd5!. 8 a4 b6 Black should play 8 ... Nc6! and keep his advantage; e.g. 9 g5 Nd5 10 Qxe4 Bxg5 11 Nf3 f5! 12 Qe2 Bf6=/+. I’ll show a few more moves because they’re fun, but you can see that Black has won the opening stage: 9 g5 Ba6!?
9 ... Nd5 is safer, with the idea 10 h4!? Nc6 and 11 Nxe4 e5=. 10 gxf6! I don’t know if this works, but it deserves an ‘!’. 10 ... Bxe2 11 fxe7 Qxe7 12 Ngxe2 Rzayev-Tran Tuan Minh, Budapest 2011. Now Black should play 12 ... e5! 13 Rg1 Nc6 14 Nxe4 f6 15 f4 Kh8, which is unclear and about balanced. Instead, the game continued 12 ... f5?! 13 Rg1 (13 Nf4! e5 14 Bc4+ Kh8 15 Rhg1! keeps the attack going, and for the first time White might even have an advantage!) 13 ... Nd7? (13 ... Rf7 14 Nf4 Qd6=) 14 Nf4! and White has a nice attack, although the game ended in a draw.
17.2 2 Nf3 d5 Now of course White can play 3 exd5 exd5 4 d4 with an Exchange Variation (see 12.1), but there are two other important variations which are used by loyal adherents: 17.21 17.22
3 3
Nc3 e5
17.21 3 Nc3 Obviously 2 Nc3 d5 3 Nf3 transposes here. 3 ... Nf6
3 ... d4 will also equalize, according to theory; that may be, but in any case I find it less interesting. 4 e5 4 Qe2 achieves nothing after 4 ... Be7, and 4 exd5 exd5 5 d4 is a slightly harmless type of Exchange Variation. Black has various answers such as 5 ... c6 and 5 ... Be7 that are equal, but I’d recommend 5 ... Bb4 with an Exchange Winawer in which Nf3 is not a terribly attractive move; e.g. 6 a3 Bxc3+ 7 bxc3 0-0 8 Be2 (8 Bd3 c5! 9 dxc5 Qe7+ and 10 Qe2 Qxc5 or 10 Be3 Ng4=/+) 8 ... Re8 (or 8 ... Ne4 9 Bb2 Nd7 10 a4 Nb6 11 a5 Nc4 12 Bxc4 dxc4=/+) 9 0-0 Ne4 10 Bb2 Nd6!=/+. 4 ... Nfd7 5 d4 c5 This position is partly analysed in Chapter 13 via the Classical move order 2 d4 d5 3 Nc3 Nf6 4 e5 Nfd7 5 Nf3 c5, but I’ve chosen to analyse 6 dxc5 in this chapter, since it famously arises by this 2 Nf3 d5 3 Nc3 move order and becausePTFveterans will expect to see it here.
Black can also wait for White to commit by 5 ... Be7; for example, 6 Bd3 (Minev suggests 6 Ne2 c5 7 c3 Nc6 8 Nf4, but 8 ... cxd4 9 cxd4 Qb6=/+ threatens ... g5 and ... Bb4+) 6 ... c5 7 0-0 (7 dxc5 Nxc5=) 7 ... Nc6 8 Be3? (8 dxc5 Nxc5=) 8 ... cxd4 9 Bxd4 Nxd4 10 Nxd4 Nxe5 11 Bb5+ Bd7 12 Re1 Nc6-/+ Fleger-Sermier, Biel 1993. 6 dxc5 The other major moves transpose to section 13.1. Those include the main Classical move 6 Ne2, as well as 6 Bb5 and the exciting ‘Jackal’ Attack, 6 Bg5!?. 6 ... Nc6 Attacking e5, which is a key central pawn that must be defended. 7 Bf4 7 Bg5 Be7 8 Bxe7 Qxe7 helps Black develop rapidly: 9 Bb5! (to protect e5) 9 ... Nxc5 (9 ... 0-0! threatens e5, and 10 Qe2 a6 11 Bxc6 bxc6 12 0-0 Qxc5 is not what White wants) 10 0-0 0-0 11 Re1 Bd7= Gelashvili-Radjabov, Greek Team Ch. 2002. 7 ... Bxc5 7 ... Nxc5 is frequently played, with fair success (it can be characterized as ‘equal but dull’); and 7 ... Be7!? with the idea ... g5/ ... h5 is unusual but probably underrated. 8 Bd3 f6
9 exf6 White cannot maintain e5: (a) 9 Bb5? fxe5 10 Nxe5?! (10 Bxe5 0-0! 11 Bg3 d4! 12 Bxc6 dxc3 13 Bxd7 cxb2 14 Rb1 Bb4+ 15 Ke2 Bxd7–+) 10 ... Ndxe5
11 Bxe5 Bxf2+! 12 Kxf2 0-0+ and ... Nxe5. (b) 9 Qe2 fxe5 10 Nxe5 Ndxe5 11 Bxe5 Nxe5 12 Qxe5 0-0 13 0-0 Qf6 (13 ... Qh4!? 14 g3 Qd4) 14 Qxf6 Rxf6-/+ with two bishops and the mobile centre; e.g. 15 g3 Bd7 16 Kg2 g5! 17 Rae1 Raf8 18 f3 Bc6-/+. 9 ... Nxf6 9 ... Qxf6 is fairly popular and recommended in various books and articles; but capture with the knight guarantees dynamic play and is also consistent with other lines in this book such as the Tarrasch with 3 ... Nf6. 10 0-0 10 Qe2 0-0 11 0-0-0 is a major alternative, heavily analysed in the past few years. I’m going to limit myself to what I think is the best solution for Black; i.e. 11 ... a6 and:
(a) 12 Kb1 is unambitious. Black does well with 12 ... b5, and safe development by 12 ... Bd7 13 Ne5 Rc8 is also promising. It’s not much, but I’d rather be Black. (b) 12 Rhe1 Qa5!? (12 ... b5!=/+ and 12 ... Bd7 are natural and somewhat better alternatives) 13 Kb1 b5 14 Ne5, Aabling Thomsen-Gershenov, Helsingor 2011; and 14 ... Qb6! (14 ... Nxe5 15 Bxe5 Ra7!? is double-edged) 15 Nxc6 Qxc6 16 Be5 Ra7! produces equal chances. (c) 12 Ne5 is the original idea, but Black has 12 ... Bd6!, pinning the knight and loosening White’s control over e5; e.g. 13 Bg3 (13 Kb1? Qc7! 14 Nxc6 Bxf4 15 Nd4?! e5 is miserable for White) 13 ... Qc7 14 Ng4! (simplifying; 14 Nf3 Bd7! 15 Rhe1 b5=/+) and then: (c1) 14 ... Bd7 15 Nxf6+ Rxf6 16 Kb1 Raf8 (16 ... b5 17 Bxh7+! Kxh7 18 Ne4 dxe4 19 Rxd6 with an unclear attack; 16 ... Rc8 17 Bxd6 Qxd6 18 g3!=) 17 f3 Ne5!? 18 Rhe1 Nxd3 19 Bxd6 Qxd6 20 Qxd3 Qf4=. (c2) 14 ... Bf4+!? 15 Kb1 Bd7 16 Bxf4 Qxf4 17 h3=. (c3) 14 ... Nxg4 15 Qxg4 Ne5!? 16 Qd4 Bd7 17 Rhe1 Nxd3+ 18 Rxd3 Rac8 19 a3 b5 20 Kb1 Rf7, Alvarez PedrazaMoskalenko, Barcelona 2011; and here 21 Bxd6! Qxd6 22 g3! intends f4. Then 22 ... Rc4! 23 Qe3 b4!? 24 Ne4! Qf8 25 Nd2 Rc7 26 Nf3! Rc4 27 Nd2 repeats, or 26 ... Bb5 (26 ... Rc4 27 Nd2=) 27 Ng5 Rxf2!? 28 Nxe6 Rcxc2 29 Nxf8 Rxb2+ 30 Kc1 Rfc2+ 31 Kd1 Ba4 32 Qe6+ Kh8 33 Ng6+ hxg6 34 Qh3+ Kg8 35 Qe6+=. 10 ... 0-0
The key position for this variation. White’s control of e5 and the e-file are at least compensated for by Black’s central majority and f-file pressure. Objectively, the position is equal; but in practice White has more ways of going wrong and has done rather poorly. We have: 17.211 17.212
11 11
Bg3 Ne5
Others: (a) 11 Qe2 Nh5! (or 11 ... Bd7 12 Rad1 a6= 13 Na4?! Ba7 14 c4?? e5!-/+ 15 cxd5 Nd4! 16 Nxd4 Bxd4–+ J.Rigo-Gulko, Rome 1988) 12 Bg5 (12 Bg3 is 11 Bg3 Nh5!? 12 Qe2?! in 17.211) 12 ... Nf4! (12 ... Be7=) 13 Qd2 Qc7 14 Nb5 Qb8 15 c4? (15 Bxf4 Rxf4 16 Rae1 Bd7=/+) 15 ... Nxd3 16 Qxd3 Ne5 17 Nxe5 Qxe5 18 Bh4 a6! 19 Nc3 Qf4-/+ Larsen-Spassky, Stockholm 1969. (b) 11 Ne2 Qe7 (or 11 ... Nh5= 12 Bg5 Qd6) 12 Ne5 Nxe5 13 Bxe5 Ng4 14 Bg3 Qf6!, hitting b2 and f2. (c) 11 Qd2 a6 (11 ... Qe8!? is an interesting option) 12 Na4?! (12 Rae1 Nh5 13 Bg5 Qc7 14 Bh4 Nf4 15 Bg3 Bd6 16 Bxf4 Bxf4 17 Qd1 Bd7=/+ J.Rigo-H.Meyer, Delmenhorst 1986) 12 ... Ba7 13 Rae1 b5! (13 ... Nh5 14 Be3! Rxf3!? 15 gxf3 and 15 ... d4 16 Bg5 Qc7, or 15 ... Bb8 16 Nb6 Qh4 17 f4 Qg4+ 18 Kh1 Qf3+=) 14 Nc3 Nh5 15 Bg5 Qd6 16 Be4 Kh8!-/+ with the ideas ... Nf4 and ... Bb8. (d) 11 Na4 Bd6 (11 ... Be7 12 c4 Nh5!? 13 Be3 Nf4 14 Bxf4 Rxf4 15 cxd5 exd5 16 Nc3 Bg4 17 Nxd5 Bxf3 18 gxf3 Rh4! is double-edged) 12 Bxd6 Qxd6 13 c4 Bd7 14 Nc3 (14 cxd5 exd5 15 Rc1!? Ng4 16 h3 Rxf3 17 hxg4 Rf4=/+) 14 ... Rad8=. 17.211 11 This is the natural move and probably White’s safest option. But it can get tricky: 11 ... a6
Bg3
Previously I assessed 11 ... Nh5!? as favourable, but it is only equal: (a) 12 Qe2?! Nxg3 13 hxg3 Bd7 (13 ... Qf6!?=/+) 14 Na4 Bd6 15 c4 Nb4! 16 Nc3 Rc8=/+. (b) 12 Be5? Nxe5 13 Nxe5 Nf4 14 Ne2 Qg5 15 Nxf4 Qxe5 16 g3 Qg5 17 Qc1 e5!=/+ intending 18 Nxd5 Qxc1 19 Raxc1 Bh3. (c) 12 Bh4 Qc7 (12 ... Qd7! looks artificial, but probably gives the most positive chances) 13 Nb5 (13 Re1 h6 14 Na4 Bd6; 13 Bxh7+? Kxh7 14 Ng5+ Kh6–+) 13 ... Qb6! (intending ... Nf4; now 13 ... Qd7!? is less appropriate due to 14 Re1! g6!? 15 c3! Qg7 16 Bc2! Nf4 17 Bg3 with an edge) 14 Ng5! Nf6 15 c4 Ne5 16 cxd5 exd5 17 Nc3 h6 18 Nf3 Bg4 19 Be2, and here 19 ... Bxf3 leads to equality after 20 Bxf3 Nxf3+ 21 Qxf3 Ne4 22 Nxd5! Rxf3 23 Nxb6 axb6 24 gxf3 Nd2 25 Bg3 Nxf1 26 Kxf1 Bd4. 12 Qd2 This has been the preferred response, although White has also tried: (a) 12 a3 Nh5 13 Bh4 Qe8 (or 13 ... Qc7=/+) 14 b4 (14 Ng5 h6 15 Nh3 Qf7!? 16 Be2 Nf6=/+ with the idea 17 Nf4 g5 18 Nd3 Bd4, F.Mayer-Suetin, Bad Wildbad 1997) 14 ... Ba7 15 Re1 Nf4 16 Bg3 Qh5 17 b5 axb5 18 Nxb5 Bc5 19 a4?! e5!? (19 ... Bd7=/+) 20 Nc7 Bg4!? (20 ... e4 21 Bxe4 dxe4 22 Rxe4 is unclear) 21 Bxf4? (21 Nxa8! Nxg2!? 22 Be2 Nxe1 23 Qxd5+ Kh8 24 Rxe1 Bxf3 25 Bxf3 Qxf3 26 Qxc5 Rxa8=) 21 ... Rxf4 22 Be2 Rd8 23 h3 Bxh3 0-1 E.Janev-V.Dimitrov, Bulgarian Ch. 1989. (b) 12 Re1 b5 13 Ne5?! (13 a3=) 13 ... Nxe5 14 Rxe5 b4 15 Na4 Ba7!? (15 ... Bd6=/+) 16 c3 bxc3 17 Nxc3 Bb8 18 Re2 Bxg3 19 hxg3 Ng4 (19 ... Qd6=/+) 20 Qc2 h6 21 Qa4?! Qg5! (threatening ... Qh5) 22 f3 Ne5 23 f4?! (23 Qd4) 23 ... Qxg3 24 Rxe5?, and instead of 24 ... Rxf4! 25 Qc2 Rh4 which is just about winning, Black played 24 ... Qxd3=/+ in Piccoli-Naumkin, Bagnara Calabra 2009.
12 ... b5
Black’s central majority has become the most important feature of the position, and he can also achieve some advantage from 12 ... Nh5; e.g. 13 Rae1 Nxg3 14 hxg3 Qd6 15 Qg5 Rf6 16 Qh5 h6=/+, threatening ... Qxg3 and ... e5, Pogosian-Nikolenko, Podolsk 1992. 13 Rae1 13 a3 Nh5 14 Rae1 Nxg3 15 hxg3 Qf6=/+ Matychenkov-Moskalenko, Podolsk 1990. 13 ... Ra7!? 14 Kh1 Re7 15 Qg5 15 Ne5 Nxe5 16 Bxe5 Ref7!? 17 Qg5 Be7 (17 ... Nd7!) 18 Qg3 Nh5, Sazonova-Nikolenko, Pardubice 1995; and instead of 19 Qh3?! 20 g6, here 19 Qe3 was about equal. 15 ... Bb7 16 Qh4 h6 17 Re2 Qe8! 18 Rfe1 Nh5 19 Qg4 Nxg3+ 20 hxg3 Rf6=/+ V.Onoprienko-Moskalenko, Noyabrsk 1995. 17.212 11 Ne5 Bd7 The traditional move. I analysed 11 ... Qe8!? inPTF2. Another option is 11 ... a6!?, in some cases with the aesthetic idea of ... b5 and ... Ra7; e.g. 12 Qe2 (12 Nxc6 bxc6 13 Be5 Qe7 14 Kh1 Nd7! 15 f4?! Nxe5 16 fxe5, Ungure-Gazis, Aghios Kirykos 2011, and 16 ... Rxf1+ 17 Qxf1 Bd7=/+ obviously favours Black) 12 ... Qe8 (12 ... Nxe5 13 Bxe5 b5!? 14 Rae1 Ra7!? 15 Kh1 Raf7! 16 f4 Nd7!=/+) 13 Nxc6 (13 Na4?? Nd4–+) 13 ... bxc6 14 Na4 Ba7 15 Bd6 Rf7 16 b3 Qd8 (or 16 ... Nd7=) 17 Bg3 Re7 18 Be5 Nd7 19 Bd6 Re8 20 Rae1 e5=/+ 21 Bf5 Nf6 22 Bxc8 Rxc8 23 Bxe5? Nd7 24 Qxa6 Ra8 25 Qxc6 Nxe5-/+ A.Bennett-J.Watson, Seattle 2003.
12 Nxc6 There are several options, none impressive: (a) 12 a3 a6 (12 ... Nxe5 13 Bxe5 Qb6 14 Qe2 Bd6=/+) 13 Nxc6 Bxc6 14 Qe2 Qe7! 15 Rae1 Rae8 16 Be5 Nd7! 17 Qh5 Rf5!! 18 Bxf5 exf5 19 Kh1 (19 Bg3 Qxe1) 19 ... Nxe5 20 f4 Qf7-/+. (b) 12 Qf3!? can be answered by 12 ... Bd4! 13 Nxc6 Bxc6 14 Rae1 e5! 15 Bxe5 Bxe5 16 Rxe5 d4 17 Bc4+ Kh8 18 Nd5 Rc8! 19 Qb3 b5–+. (c) 12 Qe2 Qe7 13 Rae1 Rae8 14 a3 (14 Bg3 is recommended by Baker; then some interesting play can follow 14 ... a6!? 15 a3 b5!, e.g. 16 Nd1 Nxe5 17 Bxe5 Bc6 18 c3 Ne4!? 19 Ne3 Qg5!? 20 f4 Qg6) 14 ... a6 15 Kh1 (15 Bg3 Nxe5 16 Qxe5!? Bc6 17 b4 Ba7 intending ... Bb8 with a small advantage, Drozdov-Hoang Than Trang, Budapest 1994) 15 ... Bd4! 16 Nxd7 Qxd7 17 Bd2 e5=/+ Ljubojevic-T.Petrosian, Las Palmas 1973. 12 ... Bxc6 13 Qe2 Qe7 13 ... Ne4, which I mentioned in early editions, now strikes me as a potentially complete solution, and a forced one at that:
(a) 14 Bg3 Nxg3 15 Qxe6+ Kh8 16 hxg3 Rf6 17 Qg4 (or 17 Qh3 Rh6 18 Qg4 Qe7!) 17 ... Qb6 18 Nd1 Raf8=/+. (b) 14 Be3 Bxe3 15 Qxe3 d4 16 Qe2 dxc3 17 Bxe4 cxb2 18 Rab1 Qf6!? 19 c4? Rab8!-/+ Bellon-Speelman, Amsterdam 1978. 19 Bxc6 bxc6 20 c3 is better; but on the previous move, 18 ... Bxe4 19 Qe4 Qb6 keeps a small advantage. (c) 14 g3 Nxf2 15 Rxf2 Bxf2+ 16 Kxf2, Vogt-Farago, Kecskemet 1979, and now 16 ... g5! 17 Qxe6+ Kg7 is unclear according to Vogt, but this seems to favour Black; e.g. 18 Ne2 (18 Qg4 Kh8!? 19 Qh5 Qe7 20 Re1 Qg7 with advantage) 18 ... gxf4 19 Qg4+ Kh8 20 Nxf4 Qf6 21 Qh5 Qg7=/+. Can White equalize after 13 ... Ne4 even? I’m extremely impressed with this line, which for some reason is hardly ever played. 14 Rae1 Rae8
15 Be5?! 15 Bg3 a6 16 a3 Qf7 17 b4 Bd4 18 Be5 Bxe5 19 Qxe5 Nd7 20 Qg3 e5=/+ Spassky-T.Petrosian, Moscow (m/19) 1966. 15 ... Nd7 16 Nb5 16 Qh5 g6! 17 Bxg6 Nxe5 18 Bxe8 Rf5! 19 Rxe5 Rxh5 20 Bxh5 Bd6 21 Re2 Qh4 22 h3 Qxh5 23 Rxe6 Bb4–+. 16 ... a6!? Or 16 ... Nxe5! 17 Qxe5 Rxf2!? (or 17 ... Qf6!) 18 Rxf2 Bxf2+ 19 Kxf2 Bxb5 20 Bxb5 Qc5+ 21 Re3 Qxb5=/+, but Black probably can’t win. After 16 ... a6, Shilov-Ulibin, Barlinek 2001, went 17 Nd4 Nxe5 18 Qxe5 Bd7 19 f4 Bd6 20 Qh5? (20 Qe3 e5 21 fxe5 Bxe5 22 Kh1=/+) 20 ... g6 21 Qg4 e5! 22 Qg3 and perhaps the simplest course was Bangiev’s 22 ... e4 23 Be2 Qf6 with a winning
game. 17.22
This
whole
line 3
is
a
tribute e5
to
the
power c5
of
the 4
central
majority. b4!?
This is the French Wing Gambit, trying to win the centre by diverting Black’s c-pawn. It has been written about in numerous articles and has a following, but hasn’t yet been taken up by a leading player. White also has some passive alternatives: (a) 4 d3 Nc6 5 c3 (5 Be2 Qc7 6 Bf4 f6) 5 ... f6 6 d4 and Black is a tempo ahead of the Advance Variation. A simple continuation is 6 ... cxd4 (6 ... fxe5 and 6 ... Qc7 are also good) 7 cxd4 Bb4+ 8 Bd2 (8 Nbd2 fxe5 9 Nxe5 Nge7 10 Bd3 0-0) 8 ... Qb6 9 a3 Bxd2+ 10 Qxd2 fxe5 11 dxe5 Nh6! intending ... 0-0 and ... Nf7 or ... Ng4. (b) 4 g3 Nc6 5 Bg2 g5!? 6 h3 Bg7 7 Qe2, J.Gatica-J.Levitt, New York 1994; and among other continuations, 7 ... Nge7 8 Nxg5 Nxe5 secures the advantage. 4 ... cxb4 Of the many other choices, I like 4 ... c4! (recommended by Simon Williams and given some encouraging analysis by others). One key line is 5 c3 a5 6 b5 Nd7 7 d3 cxd3 8 Bxd3 f6 9 Qc2!? (thought to be the best move for many years; 9 Qe2 fxe5 10 Nxe5 Nxe5 11 Qxe5 Nf6 12 Bf4 Bc5 13 0-0 0-0=/+; 9 Bf4 fxe5 10 Nxe5 Qf6=/+) 9 ... Nxe5 10 Nxe5 fxe5 11 Bxh7, NajerMoskalenko, Moscow 1995; and instead of 11 ... Bd6, which has been claimed to favour White (although I doubt it), Black gets a clearly better game from either 11 ... e4! 12 Bg6+ Kd7, or Williams’ suggestion 11 ... Qf6 12 Bg6+ Kd8 13 Bd3 e4 14 Be2 Bc5-/+. If I felt that the Wing Gambit justified two solutions, I would add more details to this one; but I’ll leave that to you instead. 5 a3 5 d4 has been considered inferior by Wing Gambit advocates as well as theoreticians.
Black has several good continuations, among them: (a) 5 ... Bd7 (Harding gives this ‘!’, as does Bosch) 6 a3 (6 Bd3 has several answers; e.g. 6 ... Qb6 intending ... Bb5, or 6 ... Ne7 7 a3 Nbc6 8 axb4 Nxb4=/+ Barendregt-Bronstein, Hamburg 1965) 6 ... Qa5 (or simply 6 ... Nc6 with the idea 7 Bd3 Qb6) 7 Bd3 Bb5 8 0-0 (8 axb4? Qxa1 9 Bxb5+ Nc6 10 Bd2 Qb2 11 c3 a5-/+ L.Day-Hübner, Jerusalem 1967) 8 ... Bxd3 9 Qxd3 Qa6 (9 ... Nc6) 10 Qb3 Nc6 11 Qb2 Qb6=/+ Vedder-Poulton, Gausdal 1990. (b) 5 ... Nc6 6 a3 f6 7 axb4, and 7 ... Bxb4+ 8 c3 Ba5 intending ... Bc7 or 7 ... fxe5 8 Nxe5 (8 b5 Nxd4 9 Nxe5 Qh4) 8 ... Nxe5 (8 ... Bxb4+ 9 c3 Bd6=/+ with the idea 10 Qh5+ g6 11 Nxg6 Nf6 12 Qh6 Rg8 13 Nf4 Qe7) 9 dxe5 Bxb4+ 10 c3 Bc5 11 Qg4 Qe7=/+. 5 ... Nc6 Here 5 ... d4!? has probably been given more attention and been recommended more than other moves. It’s certainly good, but it gives White a great number of options, so I’m going with something which is still complex, but more focused. A solid alternative, which has been pretty popular over the years, is 5 ... Nh6 6 axb4 Bxb4 7 c3 Be7 8 d4 Nf5
9 g4 (9 Bd3 Nc6 10 g4 transposes – this is probably White’s best line, although I don’t think he gets adequate compensation against careful play; instead, 9 g3 Nc6 10 Bd3 0-0 11 h4 prepares g4, but there might follow 11 ... f6 12 g4?! fxe5 13 gxf5 e4 14 Ng5 exd3 15 Qh5 h6 16 Qg6, M.Helin-T.Karlsson, Sweden 1992, and 16 ... Bxg5 17 hxg5 Qe8!? 18 gxh6 exf5+ breaks the attack, or here even 17 ... exf5 18 gxh6 Qe7+ 19 Kf1 Qe2+ 20 Kg2 Qg4+) 9 ... Nh4 10 Nxh4 Bxh4 11 Bd3 Nc6 12 0-0 Be7 13 f4 Bd7 (a known position; 13 ... 0-0 14 g5! gives White a surprisingly dangerous attack) 14 Nd2 (14 Ba3 b5!) 14 ... a5!? (14 ... Qc7 15 Ba3, Mortensen-Farago, Esbjerg 1985, and now 15 ... g6, intending ... h5, or 15 ... a5 should keep some advantage) 15 f5 a4 16 Qf3 Qa5 and White’s attack doesn’t look like it’s worth a pawn. Still, that’s not written in stone and both
sides can have fun here. 6 axb4 Bxb4 7 c3 Be7 This is one of the easier systems to play. Black keeps his bishop near the kingside. 8 d4 f6
Black is consistent with our usual philosophy of attacking the centre. I should mention 8 ... Nh6, when the idea is to play ... Nf5 of course, and notice that a quick ... Nf5 will usually transpose to 5 ... Nh6 above. If White plays 9 Bxh6 gxh6 10 Qc1!, then 10 ... Rg8!? 11 Qxh6 Qb6 aims for ... Qb2, and with the idea 12 Qxh7?! (12 Qd2 a5 with chances for both sides) 12 ... Rg6=/+. 9 Bd3 9 exf6 Nxf6 10 Bd3 0-0 11 Na3 Bd7=/+ 12 Ne5? Nxe5 13 dxe5 Ne4 14 Be3 Qc7 15 Bd4 Nxf2 winning, H.SchulzKindermann, Hamburg 1995; while 9 Bf4 Nh6! 10 Bxh6 (else ... Nf7) 10 ... gxh6 11 exf6 Bxf6 is essentially a tempo up on the main idea. 9 ... fxe5 10 dxe5 After 10 Nxe5?! Reindermann-Glek, Groningen 1992, continued 10 ... Nf6 11 Bg5 Nxe5!? (11 ... 0-0=/+) 12 dxe5 Ne4 13 Bxe7 Qxe7 14 Bxe4 dxe4 15 Qd4 0-0 16 Qxe4 (16 Rxa7 Rxa7 17 Qxa7 Qg5-/+ Buturin-Glek, Belgorod 1989) 16 ... Bd7! 17 Qe3 (17 Qxb7 Qc5!) 17 ... Bb5 18 Nd2 Qc7 19 f4 Rad8-/+. I’m not so sure that Black should get an advantage after 11 ... Nxe5, but Black’s development is freer than it is after 10 dxe5. 10 ... Nh6 10 ... Qc7 is an interesting alternative.
Some sample lines: (a) 11 Qe2 Nh6 12 Bxh6 gxh6 13 0-0 0-0 14 Na3!? Bc5 15 Nb5 Qg7 16 Kh1 Bd7 17 Nbd4 Bxd4 18 cxd4 Be8 and ... Bh5 or ... Bg6 as appropriate. (b) 11 0-0!? Nxe5 12 Nxe5 Qxe5 13 Re1! Qc7 14 Qh5+ (14 Bb5+ Kf7 15 Nd2 Nf6 16 Nf3 Rf8 17 Ng5+ Kg8 18 Nxe6 Bxe6 19 Rxe6 Bc5-/+ Hamed-Garma, Novi Sad Olympiad 1990) 14 ... Kf8 15 Ba3 Nf6 16 Qe2, and 16 ... Bxa3 17 Nxa3 Bd7 is at any rate level, whereas after 16 ... Kf7!, White has some compensation but Black does have two extra pawns. All these positions are hard to assess. 11 Bxh6 gxh6 12 Qd2 The ‘book’ move, but I think White has to improve somewhere, so I’ll assume that 12 0-0 is better, offering some truly arbitrary analysis that can’t be complete or probably even accurate but will hopefully illustrate some key ideas: 12 ... 0-0 13 Re1 (13 Qd2?! Rxf3! 14 gxf3 Nxe5; an instructive and relatively high-level game went 13 Ra4 Rf7 14 Re1 Bd7 15 Nbd2 Kh8 16 Nf1 Qg8 17 Ng3 Raf8=/+ 18 c4? Rxf3! 19 gxf3 Nd4 20 Rxa7 Nxf3+ 21 Kg2 Nxe1+ 22 Qxe1 Bc6 23 Qe3 d4+ 0-1 K.Schmidt-Vaisser, Paris 1990) 13 ... Bd7 (13 ... Rf7!?) 14 Nbd2 (intending Qb1; 14 c4?! Bb4 15 Nbd2 Qe7 16 cxd5 exd5 17 Qb3 Be6) 14 ... a5!? (14 ... Rf7 15 Qe2 Rg7 16 g3 a5; these lines can fairly be described as double-edged, but I lean towards Black in most of them) 15 Qb1 (15 Nd4 Nxd4 16 cxd4 Rf4! 17 Nf3 Bb4 18 Re3 Qf8) 15 ... Rf7=/+ (intending ... a4 and ... a3 or sometimes ... Na5), and Black hopes for 16 Qxb7?? (16 h3 a4 17 Kh1 Na5) 16 ... Rb8 17 Qa6 Rb6. 12 ... Qc7!
13 Qxh6
Others: (a) 13 Qe2 0-0 14 0-0 Rxf3!? (14 ... Bd7) 15 Qxf3 Nxe5 16 Qg3+ Kh8 intending ... Bd6 is unclear, but Black’s bishops and centre have good potential; e.g. 17 Be2 (17 Na3 Bd7 18 Rfe1 Bd6; 17 Bc2 Bd6=/+) 17 ... Bd7 18 Nd2 Rg8 19 Qe3 Ng6!. (b) 13 Qe3 Bg5!? (13 ... Bd8 14 Na3 Qg7 15 0-0 Bb6 16 Qe2 Rg8 17 g3 Rf8, planning ... Qg4) 14 Nxg5 hxg5 15 Bb5 0-0! 16 Bxc6 (16 Qxg5+ Kh8 with the idea 17 Bxc6 Rf5=/+) 16 ... bxc6 (16 ... Rf5!?) 17 Qxg5+ Qg7 18 Qxg7+ Kxg7=; Black has the practical chances in these lines, but I can’t fairly claim a true advantage. 13 ... Nxe5 14 Qh5+ Nf7 15 Bb5+ Kf8 16 0-0 Pedersen says that White has some compensation, but I definitely prefer Black following 16 ... Rg8 (or 16 ... Qf4 17 Ra4 Qf5=/+) 17 Re1 (17 Qxh7 e5 18 Qc2 Bh3 19 Ne1 e4 20 Kh1 Be6-/+) 17 ... Rg7 (or 17 ... Qf4 18 Ra4 Qf6) 18 Ra4 Bf6 19 h4 a6 20 Nd4 Rb8 21 Bd3 e5 22 Nf5 Bxf5 23 Bxf5 e4-/+ with ideas of ... Ne5-d3 or ... Qe5. 17.3
2
d4
d5
3
Be3
This is called the ‘Alapin-Diemer Gambit’ by Harding. The idea after 3 ... dxe4 is to get into a sort of Blackmar-Diemer Gambit (1 d4 d5 2 e4 dxe4 3 f3, or 2 Nc3 Nf6 3 e4 dxe4 4 f3), normally playing f3 early on. The drawback to this strategy is that the bishop on e3 would be poorly placed in a Blackmar-Diemer (an opening of marginal soundness anyway). Nevertheless, the line has its defenders and White, with the advantage of the first move, can get away with more in the opening than Black. A few other third move options are: (a) 3 c4?! dxe4 is a dubious adventure called the ‘Diemer-Duhm Gambit’, although it is not outright awful and has a few happy followers: 4 Nc3 Nf6 (even the superficially weakening 4 ... f5 seems fine; e.g. 5 f3 Nf6 6 fxe4 fxe4=/+), and now White has two tries:
(a1) 5 f3 c5 6 d5 Bd6 (6 ... exd5! 7 cxd5, Heikinnen-Buongiovanni, corr 1999; andECOgives 7 ... exf3 8 Nxf3 Bd6 with a clear advantage, but 7 ... Bd6! 8 fxe4 0-0 9 Nf3 is worse still for White, transposing into the disastrous part of this note) 7 fxe4 0-0 8 Nf3 exd5, which develops and breaks open lines quickly: 9 exd5 (9 cxd5?! Re8 10 Bd3 c4! 11 Bc2 b5! 12 Nxb5 Bb4+-/+; unfortunately for White, this line is one that 6 ... exd5 7 cxd5 leads into), and now, among other methods, there’s 9 ... Bg4 10 Be2 Bxf3! 11 Bxf3 Re8+ 12 Kf2 Nbd7 13 Rb1 Be5 14 g3 Qa5 15 Qb3 Bxc3 16 bxc3 Ne5 17 Rf1 b5! 18 cxb5 c4 19 Qc2 Nxf3 20 Kxf3 Nxd5-/+. (a2) 5 Bg5 Be7 (or 5 ... h6 6 Bxf6 Qxf6 with the idea 7 Nxe4?! Bb4+ 8 Nc3 c5!) 6 f3 (6 Nge2 Nc6 and White is stuck for a good move; e.g. 7 Qd2 Nb4! 8 Ng3 Qxd4!) 6 ... Nc6! 7 d5 (7 Nge2 exf3-/+) 7 ... exd5 8 Bxf6 (8 cxd5 Nxd5-/+) 8 ... Bxf6 9 cxd5 Ne5 10 fxe4 0-0 11 Nf3 c6 12 Be2 cxd5 13 Qxd5 Qb6!-/+. (b) 3 f3 c5 (or 3 ... Nf6 and 4 Nc3 c5! or 4 e5!? Nfd7 5 f4 c5 and the tempo loss by f3-f4 doesn’t inspire) 4 Be3 Nc6 5 c3? (5 dxc5 Nf6 6 Nd2 d4 7 Bf2 e5=/+) 5 ... dxe4 6 fxe4 Nf6 (6 ... cxd4! 7 Bxd4 Qh4+-/+) 7 e5 Nd5 8 Bf2 cxd4 9 cxd4, Stanka-B.Lalic, Oberwart 2001; 9 ... g6!? 10 Nf3 Bh6 11 Nc3 0-0 12 Nxd5 exd5 13 Be2 Qa5+-/+. 3 ... dxe4 4 Nd2 4 f3 Nh6!? (threatening ... Nf5; or 4 ... Nf6 5 Nc3 and 5 ... Nd5 6 Nxd5 exd5 7 Qd2 Bd6-/+ or 5 ... Bb4 6 a3 Nd5! – both lines by Minev) 5 Qd2 (5 Bxh6 Qh4+ is good for Black; 5 fxe4? Qh4+ 6 Bf2 Qxe4+-/+) 5 ... Nf5 6 Bf2 c5!=/+ with the idea 7 dxc5? Qxd2+ 8 Kxd2 e3+! 9 Bxe3 Nxe3 10 Kxe3 Bxc5+-/+. 4 ... Nf6
5 c3!
Others: (a) 5 f3?! Nd5! 6 Qe2 Nc6! (6 ... Nxe3 7 Qxe3 exf3 8 Ngxf3 Be7=/+) 7 c3 exf3 8 Ngxf3 Bd6 and: (a1) 9 Ne4 0-0 10 0-0-0 (10 Bg5 f6 11 Bd2 e5 12 dxe5 Bxe5 13 Nxe5 Nxe5 14 h3 Re8 15 0-0-0 Bf5-/+) 10 ... Rb8! (planning ... b5) 11 c4 Nxe3 12 Qxe3 b5 13 cxb5 Nb4=/+. (a2) 9 Bf2 0-0 10 0-0-0 a6 11 c4 (11 g3 b5 12 Bg2 Qe7 13 Rde1 Bb7 14 Rhf1 Na5=/+ T.Bullockus-U.Drössler, corr 2000) 11 ... Nf4 12 Qe3 Ng6 13 Kb1 b5! (or 13 ... e5 14 d5 Nb4 15 a3 Bf5+ 16 Ne4 Bc5! 17 Qxc5 Bxe4+ 18 Ka1 Nc2+ 19 Ka2 Qd7) 14 cxb5 axb5 15 Bxb5 Nb4 16 Bc4 Nd5-/+ or 16 ... Bb7. (b) 5 Bc4 Nc6 6 Ne2 e5 7 c3 exd4 8 Nxd4 Ne5 9 Qb3 Bd6=/+ Edvardson-J.Watson, Gausdal 1980; or 9 ... a6! 10 a4 Bd6. 5 ... b6!? This has been the most frequent move. 5 ... Be7 is of course a sound alternative. 6 g3 White doesn’t get his pawn back after 6 Qc2 Bb7 7 Ne2 Qd5 8 Ng3 (8 g3 Bd6 9 Bg2 Qf5=/+) 8 ... Nbd7 9 Bc4 Qc6 10 a4 a6 11 0-0 Bd6 12 Be2, L.Szabo-Sinkovics, Frankfurt 1988; and here 12 ... 0-0=/+ is easiest. 6 ... Bb7 7 Bg2
7 ... c5!? 7 ... Be7 8 Qc2 0-0 9 Ne2 (9 Nxe4 Nd5) 9 ... Nbd7 (9 ... Qd5!) 10 0-0-0, C.Ohst-C.Lutz, German League 1987; 10 ... Nd5! 11 Nf4 Nxf4 12 Bxf4 f5=/+. 8 dxc5! Bxc5 9 Bxc5 bxc5 10 Qc2 Qd5 10 ... Qc7! 11 0-0-0 (11 Nxe4?? Nxe4 12 Bxe4 Qe5) 11 ... 0-0 12 Nxe4 Nbd7 13 Nxf6+ Nxf6 14 Bxb7 Qxb7 15 Ne2 c4=/+. 11 Ne2 Nbd7 Black still has some advantage; for example, 12 c4!? Qf5! 13 0-0 (13 0-0-0? Ne5) 13 ... Ne5 14 Nc3 0-0 15 Ndxe4 Nf3+ 16 Bxf3 Qxf3 17 Rfe1 Nxe4 18 Re3 Qh5 19 Nxe4 f5 20 Nd6 f4! and Black’s attack continues. 17.4
2
d4
d5
3
Bd3
A safe move, but it generates an enthusiasm that I find mystifying. Apparently the French Defence is so powerful that White is happy to allow early equality. 3 ... dxe4 3 ... c5 will often lead back into a line of the Exchange after 4 exd5 (4 c3 Nc6 5 Ne2 cxd4 6 cxd4 Nb4! is at least equal for Black) 4 ... exd5. This can be quite a lively variation for Black, often leading to an IQP situation. I can recommend it as an alternative, although it does take play outside of our repertoire (which in section 12.2 is 4 ... Nc6, rather than 4 ... c5). 4 Bxe4 Nf6
5 Bf3! The only active move. (a) 5 Bg5?! c5! (or 5 ... Be7=) 6 Bxf6 (6 c3 cxd4 7 cxd4 Qa5+ 8 Bd2 Qb6 9 Bf3 Nc6 10 Bc3 Be7=/+; White’s bishop on c3 is passive and gets in the way) 6 ... Qxf6 (here 6 ... gxf6! is good; e.g. 7 Nf3 f5 8 Bd3 cxd4 9 Nxd4 Bd7 10 Nc3 Nc6=/+) 7 Nf3 Nc6 8 c3 cxd4 9 cxd4 (9 Bxc6+? bxc6 10 cxd4 Rb8 11 Qd2 c5=/+) 9 ... Bd6 10 Nc3 0-0 11 0-0 Bd7=. (b) 5 Bd3 c5! 6 dxc5 (6 Nf3 cxd4! 7 Nxd4 e5!=/+) 6 ... Nbd7! or 6 ... Bxc5 7 Nf3 Nc6 8 0-0 0-0 9 Nc3 Nb4= Franceschi Rivera-P.Hummel, Oropesa del Mar 1998. 5 ... c5 5 ... Nbd7 6 Ne2 Bd6 7 0-0 0-0 8 Nbc3 e5!= is a perfectly sound way of developing if you want to try something different. Several games have supported that view.
6 Ne2 Or 6 c3 Nc6! 7 Ne2, and now Black takes his pick:
(a) 7 ... cxd4 8 cxd4 Be7 (8 ... Bd6 9 0-0 0-0 10 Nbc3 h6 11 Be3 Bd7 12 Rc1 Ne7 13 Nf4 Qa5 14 Qd2 Bc6=) 9 Nbc3 0-0 10 00, Mamedov-Sa.Martinovic, Aix-les-Bains 2011; and either 10 ... Qb6 or 10 ... Qa5 11 Bd2 Qa6 is equal. (b) 7 ... Be7 8 0-0 0-0 9 Bf4 cxd4 10 cxd4 Qb6 11 Nbc3 Rd8 12 Na4 Qa5 13 Rc1 Qf5=. (c) 7 ... e5 is the most forcing: 8 0-0?! (8 Bxc6+ bxc6 9 0-0 Bd6!? 10 dxe5 Bxe5 11 Qxd8+ Kxd8 12 Bf4 Bxf4 13 Nxf4 Kc7=) 8 ... e4 9 Bxe4 Nxe4 10 d5 Bd7 (10 ... Ne7? 11 Qa4+) 11 dxc6 Bxc6=/+. 6 ... Nc6 The easiest, if unambitious, solution is 6 ... cxd4 7 Nxd4 (7 Qxd4 is a problem if Black is intent on winning: McDonald likes 7 ... Nbd7 8 0-0 Bc5, but then White has a minor advantage after 9 Qf4 0-0 10 Nbc3; here 7 ... Bd6 or even 7 ... Bd7!? appears to be a better try) 7 ... Bc5 8 Nb3 Qxd1+ 9 Bxd1 Be7 10 Bf3 Nc6 11 0-0 (11 g3 Bd7 12 0-0 Nb4 13 Na3 Bc6, at least equal) 11 ... Ne5 12 Be2 b6 13 Bf4 (13 f4 Ng6 14 Bf3 Rb8=) 13 ... Ng6 14 Bg3 Bb7=. 7 Be3
7 ... cxd4 Or 7 ... e5!? 8 Bxc6+ bxc6 9 dxe5 Qxd1+ 10 Kxd1 Ng4= H.Süchting-Alapin, Barmen 1905. The bishop pair makes up for the material and the c-pawn weaknesses. Play might continue 11 Nd2 Be7 12 Kc1 (12 Ne4 Bf5 13 N2g3 Nxe3+ 14 fxe3 0-0-0+ 15 Ke2 Bg4+ 16 Ke1 Rd5=) 12 ... 0-0 13 Bf4 f6! (13 ... Ba6!?) 14 exf6 Nxf6 15 Bg5 (15 Bg3 Nd5=) 15 ... h6 16 Bxf6 Bxf6 17 Ne4 Be5 18 f3 Bf5 19 N2c3 c4! 20 a4 Rab8 and the bishops exert enough pressure to establish full equality.
8 Bxc6+?! 8 Nxd4 Ne5= (Uhlmann); e.g. 9 0-0 Be7 (or 9 ... a6 10 Nc3 Bd6) 10 Nc3 0-0 11 Qe2 (11 Re1 Nc4 12 Bc1 e5) 11 ... a6 12 Rad1 Qc7=. 8 ... bxc6 9 Qxd4 Qxd4 10 Bxd4 c5! 11 Bxf6 gxf6 Black has the bishop pair and open lines, which more than make up for his weaknesses: 12 Nbc3 Rg8 13 Ne4 Be7 14 0-0 Bb7 15 f3 0-0-0 16 Rad1 f5! 17 N4g3 c4=/+ Bezgodov-C.Hanley, Hoogeveen 2002.
Table of Contents Bibliography Introduction 1 Advance Variation: Introduction and Traditional Lines 2 Advance Variation: 6 a3 3 Advance Variation with 5 ... Nh6 4 Tarrasch Variation: Introduction to 3 ... Nf6 with 5 Bd3 5 Tarrasch Variation: Main Lines with 5 Bd3 6 Tarrasch Variation with 5 f4 7 Winawer Variation: Fourth Moves 8 Winawer Variation: Fifth Moves 9 Winawer Variation: Poisoned Pawn Variation 10 Winawer Variation: Positional Main Lines 11 Winawer Variation: Portisch-Hook Variation 12 Exchange Variation 13 Classical Variation with 4 e5 14 MacCutcheon: Introduction 15 MacCutcheon: Main Lines with 6 Bd2 16 King’s Indian Attack 17 Uncommon Early Moves
