Negi - GR. 1 e4 vs The Sicilian I (2015)

Grandmaster Repertoire

1.e4 VS The Sicilian I By

Parimarjan Negi

Quality Chess www.qualitychess.co.uk

First edition 20 1 5 by Quality Chess UK Ltd Copyright © 20 1 5 Parimarjan Negi

Grandmaster Repertoire -l.e4 vs The Sicilian I 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. Paperback ISBN 978- 1 -906 5 5 2-39-8 Hardcover ISBN 978- 1 -906 5 5 2-97-8 All sales or enquiries should be directed to Quality Chess UK Ltd, 20 Balvie Road, Milngavie, Glasgow G62 7TA, United Kingdom Phone +44 1 4 1 204 2073 e-mail: [email protected] website: www. qualitychess.co.uk Distributed in North America by National Book Network Distributed in Rest of the World by Quality Chess UK Ltd through Sunrise Handicrafts, ul. Skromna 3, 20-704 Lublin, Poland Typeset by Jacob Aagaard Proofreading by Daniel McGowan & John Shaw Edited by Andrew Greet Cover design by adamsondesign.com; Cover photo by capture365.com Printed in Estonia by Tallinna Raamatutriikikoja LLC

Contents Series Introduction Preface Symbols & Bibliography

4 5 6

6 .ig5 versus the Najdorf .

4

6 � bd7 Introduction and Rare 6th Moves 7...g6 7...h6 8...g6

5 6 7

6 ... e6 7.f4 Introduction and Sidelines 7.. Y!ic7 The Polugaevsky Variation

81 94 114

8 9

The Gelfand Variation Introduction The Main Line

130 149

10 11 12

Classical Main Line The Gothenburg Variation Three-Piece System with ... h6 Three-Piece System - Main Line

167 184 205

13 14 15

Poisoned Pawn Introduction and 12... tLld5!? 12... tLlfd7 20...E!:d8

225 248 272

16 17 18

12 g5 13.,if2!? 13.exf6 Introduction and 16.@hl 16.E!:bdl

287 310 325

Variation Index

352

.•.

1 2 3

7 26 45 57

. . .

-

Series Introduction Several years ago, Nigel Short once gifted me an opening book with the wry comment, "I expect this doubles your chess library." While that was a slight exaggeration, it is true that I never depended much on opening books for my theoretical knowledge. This scepticism for written material is often rooted in the fact that the modern openings are so fluid, constantly changing, with variations evolving every week. How can a static book keep pace with ideas that are developed, replaced, and forgotten on a weekly basis? In this series, I aim to provide a foundation and structure around which you can develop a lifelong repertoire with 1 .e4. I hope a good number of my recommendations will withstand the relentless assaults of time, but I am realistic enough to know that many of the bright novelties and variations will eventually wither away. However, even if some of the finer details will eventually have to be revised in the future, I believe that the core selection of recommended lines will remain valid for a long time to come. There is a process by which I have developed my own repertoire, as laid out in this series. In several critical positions I have discussed the pros and cons of different options, and explained why I eventually chose one over the other. Apart from remembering the moves themselves, I hope you will also absorb something of this process, so that you will be able to find your own ideas to react to whatever new developments may come along. The selection of recommended systems against Black's various defences has followed a logical pattern, taking into account the structure of the whole repertoire. Throughout the process, I have aimed for active, fluid positions, sometimes sacrificing material but always remaining fundamentally sound. There are a number of long, forcing lines, which are necessary to justifY any suggestion these days, but I have endeavoured to show that even seemingly abstract moves are still based on strategic, human principles. To make the best possible use of this book, I encourage you to pay attention to all such explanations, with the aim of building a framework of inter­ connected ideas in your own mind. Parimarjan Negi New Delhi, July 20 1 4

Preface In my first book, my aim was to present a repertoire built around an interlinked set of ideas, and the moves were often j ust an extension of the explanations and themes. You might notice some changes in style in the present volume - mostly because the opening being tackled here is a completely different type of beast. I have tried a few different set-ups against the Naj dorf, but the aggressive 6 . .ig5 has always been my main weapon. It is one of the most concrete theoretical lines out there and, with the advent of correspondence databases and modern computers, it has been analysed to such astonishing depths that one may wonder if there is anything new to discover at all. Some of the lines are long, and I cannot promise to find a logical explanation for all of them, besides the simple fact that they do work. Despite the forcing, tactical nature of our subject, I am pleased to say it is not j ust about following the computer's suggestions. As you make your way through the book, you will find that the recommendations against different variations will fit together as a coherent whole. For instance, both the 6 . . . ltJ bd7 and 6 . . . e6 7.f4 ltJ bd7 lines are dealt with in a similar manner with 'We2. The queen move may not seem like the most natural response, but once you see the flow of development with 0-0-0 and £2-f4, followed by e4-e5 or g2-g4, you will realize that the two lines have many recurrent ideas, with White essentially trying to achieve the same goals. The Poisoned Pawn stands alone as the most hotly-debated variation in the Najdorf. But despite thousands of games - many of them from the correspondence realm - there are a surprising number of new ideas waiting to be explored, and I have highlighted the most promising of the ones I discovered. It will always be hard to prove an advantage against the most precise, computer-generated defences, but the important thing is to realize that you can always pose new problems, even when the theory seems incredibly far advanced. And while a correspondence player with plentiful time and a powerful computer might navigate the pitfalls, it will be practically impossible for an over-the-board player to do the same. This method of exploration is a constant process that does not have to end with this book. Sooner or later my new suggestions will be tested as well, but if you continue your investigations in a similar vein, you will discover your own original methods to pose problems for your opponents. Several other variations have rich histories, but they are less challenging for us than the Poisoned Pawn. In most cases, White will build a powerful attack, and will be able to choose from a plethora of exciting options. I hope that after reading this book, you will not only understand how to handle the opening, but that you will also have an improved sense for the initiative and how to exploit it. Parimarjan Negi Palo Alto, December 20 1 4

Key to symbols used ± + +-+

iD

� CD

?? !! !? ?! #

White is slightly better Black is slightly better White is better Black is better White has a decisive advantage Black has a decisive advantage equality with compensation with counterplay unclear a weak move a blunder a good move an excellent move a move worth considering a move of doubtful value mate

Bibliography Aagaard & Shaw: Experts vs The Sicilian 2nd edition, Quality Chess 2006 Arizmendi & Moreno: Mastering the Najdorf, Gambit 2004 Emms: Play the Najdorf: Scheveningen Style, Everyman Chess 2003 Ftacnik: Grandmaster Repertoire 6 The Sicilian Defence, Quality Chess 20 1 0 Georgiev & Kolev: The Sharpest Sicilian 2012, Chess Stars 20 1 2 Goh Wei Ming: Chess Developments: The Sicilian NajdorfBg5, Everyman Chess 20 1 4 Palliser: Starting Out: Sicilian Najdorf, Everyman Chess 2006 Rizzitano: Play the NajdorfSicilian, Gambit 20 1 0 Yakovich: Sicilian Attacks, New i n Chess 20 1 0 -

Periodicals New in Chess Yearbooks

Electronic/Internet resources Chess Base Magazine ChessPublishing. com

a

b

c

d

e

f

g

Introduction and Rare 6th Moves Variation Index l .e4 c5 2.� f3 d6 3.d4 cxd4 4.�xd4 � f6 5.� c3 a6 6.J.g5 A) 6...e5 B) 6...J.d7 C) 6...� c6 D) 6 ...� bd7 7.Ve2! D l ) 7...Va5?! D2) 7...e5 D3) 7...b5 D3 1 ) 8.0-0-0 D32) 8.£4! e5 9.fxe5 dxe5 1 0.0-0-O! D32 1 ) 1 0...exd4 D322) 1 0...Vb6N D323) 1 0...h6N D324) 1 0...J.b7N 1 1 .� f5 D324 1 ) 1 l ...Vb6 D3242) 1 l ...Vc7 D l ) after 1 0 . . . h6 8

8 7

6

6

5

5

4

4

3

3

2

2 a

b

c

d

e

f

1 1 .0-0-0!?N

g

D2) after 1 3 Jlc8

D32 1 ) after 1 4 . . .li'lf6

14.�d2!?N

1 5 .li'lxe7!N

..

7

8 9 9 11 11 13 15 15 16 19 19 20 21 22 23

h

h

6 . lLl bd7

8

. .

1 .e4 c5 2.t£Ja d6 3.d4 cxd4 4.t£Jxd4 t£Jf6 5.t£Jc3 a6 6 .ig5 •

8 7

This does not make much sense, as the bishop on gS is perfectly placed to help White exploit the weakened dS-square. The only thing White should not do is misplace his knight by retreating it to b3.

7.t£Jf5

6

This seems simplest, because it more or less forces Black to exchange on fS .

5 4

7....ixf5

3 2 1 a

b

e

d

e

f

g

h

7 . . . .ie6 has been tried by Nakamura in an internet blitz game, but White can j ust play 8 . .ixf6 Wxf6 (8 . . . gxf6 9 . .ie2± leaves Black in a horrible position) 9.lLlxd6t and win a pawn.

We will begin our journey by checking the rare options of A) 6 e5, B) 6 Ad7 and C) 6 t£Jc6, before introducing the rather important topic of D) 6 �bd7, coverage of which will run for a few more chapters. •••

.•.

•••

••.

A database search will show that Black has tried almost every legal move at some point, but most of them can be discarded quickly. A few examples: 6 .. .'\Wb6? does not work without . . . e6 because of 7 . .ixf6 gxf6 8 . lLl d S .

a

b

e

d

e

f

g

h

8 ..bf6! 6 . . . g6?! does n o t make any sense here, since 7 . .ixf6 exf6 8.'!Wd2 leaves Black with an inferior version of the Rauzer with 6 . . . g6. 6 .. .'\WaS is conveniently handled by 7 . .ixf6 (7.Wd2 lLl bd7 can be compared with the 6 . . . lLl bd7 lines) 7 . . . gxf6 8 . lLl b3 Wd8 when Black has lost some time compared with typical positions with this structure. Here it looks good for White to include 9.WhS!;!; before playing 0-0-0 and £2-f4.

A) 6 e5 • • •

This is the easiest choice, giving White a pleasant position without any risk.

8 ...fi'xf6 9.�d5 9.exf5 ! ? is also possible. Mter 9 . . . WxfS 1 O . lLl dS !'1a7 1 1 ..id3 White has excellent compensation for the pawn, but there is no need to sacrifice.

9 ...fi'd8 10.exf5�

White's domination of the light squares gives him a stable advantage, which will remain even if Black swaps down to an opposite-coloured­ bishop middlegame.

9

Chapter 1 - Rare 6th Moves

B) 6 J.d7 •••

This gives Black the option of converting to a line of the Rauzer, but it also has a few tricky points of its own. 7.£4! It is important to get the move order right. 7.W1'd2 e6! is slightly annoying, as both 8 . f4 and 8. 0-0-0 can be met by 8 . . . h6, when retreating the bishop to h4 allows a . . . ltJxe4 trick.

b

a

e

d

e

f

h

g

7... �c6 Black had better play this now, otherwise the queen will go to 8.

1 1 . . .d5!N Creating a strange mess. ( l l . . . W1'c7 1 2 . ltJ de3 e6 1 3 . ltJ g3 favoured White in Mokry - Oltean, Balatonbereny 1 98 5 . ) 1 2.VNxd5 W1'c7 1 3 .W1'd2 e6 1 4. ltJ g3 h5gg

8.VNdl b5

Now I don't see anything better for Black than:

a

b

e

d

e

f

g

h

9.Lf6! Another important detail, avoiding 9 .0-0-0 b4! 1 O.�xf6 bxc3 1 1 .W1'xc3 gxf6 1 2. ltJxc6 W1'c7 when Black will have good compensation in the coming endgame, Kalod - Obsivac, Morava 1 997. 9. . gxf6l0.0-0-0 1 0.ltJf5 is a tempting possibility to exploit Black's move order, but it allows for an interesting possibility: 1 O . . . b4 1 1 .ltJ d 1 .

a

b

e

d

e

f

g

h

lO e6 •••

Finally we transpose to a major branch of the Rauzer, which will be discussed in a subsequent volume.

C) 6 �c6 ...

This is a strange hybrid of the Naj dorf and Classical Sicilians. It has brought Black surprisingly decent results, particularly in the recent games of Ivan Saric, but it should not cause much trouble against a well-prepared opponent.

10

6 . . �bd7 .

7.�d2 Playing in the style of the Classical Rauzer is the simplest way. Exchanging on f6 would lead to a thematic type of position from that system, but not necessarily one of the better versions for White.

7 ... �xd4 This early exchange is the idea behind Saric's set-up. Most other moves, such as 7 . . . e6, 7 . . . �d7 and 7 .. .'lWb6, will lead to variations of the Rauzer that will be covered in a future volume. Here we will concentrate on Black's attempts to keep the game in independent territory.

1 6.0-0 White's space advantage gave him a pleasant position in Oparin - Yuffa, Moscow 20 1 2 .

/

The text move is Saric's idea. Black remains flexible in terms of . . . e6 or . . . e5, and he has already caused White to put his pieces on unusual squares. In most games White has failed to achieve anything special against Black's unfamiliar set-up. However, in 20 1 4 Vladimir Akopian hit upon the right idea. 8 7 6

8.�xd4

5

8

4

7

3

6

2

5

1 a

4

b

e

d

e

f

g

h

lo.lLld5!

3

Exploiting Black's refusal to play ... e6.

2

lo ... lLlxd5 1 l .exd5 �xc2N

1 a

b

e

d

e

f

g

h

8 ...�a5 B . . . e5 has long been abandoned by strong players, and with good reason: 9 .Wa4t �d7 1 0 .�xf6! Wxf6 ( I 0 . . . gxf6 1 1 .Wb3±) 1 1 .�b5 WdB 1 2 .�xd7t Wxd7 1 3 .Wb3 White obtained a dream position with good knight against bad bishop in Kotronias - Kovalev, Debrecen 1 992. 9 ..id2 �c7 With 9 . . . e5 Black j ust accepts a slightly worse position, so it's not much of a concern for White. 1 0.Wd3 �e6 1 1 .tiJ d 5 WdB 1 2.c4 tiJ xd5 1 3 .cxd5 �d7 1 4.Wa3 ! ? �e7 1 5 .�d3 0-0

This move is the only real deterrent towards White's 1 0th move, but it allows White to obtain a huge initiative. 1 1 . . .e5 1 2.dxe6 �xe6 1 3 . 0-0-0;1; gave White a safe, pleasant advantage in Akopian - Ter Sahakyan, Minsk 20 1 4.

12J�cl �f5 13 ..ie2 White can simply finish his development, knowing that Black has no easy way to do the same.

13 ... e5 1 3 . . . �d7 1 4 Jk7;1;

Chapter 1

-

14.dxe6 he6 1 5.0-0i 0) 6 .t2:J bd7 •.

b

e

d

11

8.i.d2!

Black struggles with his basic development.

a

Rare 6th Moves

e

f

g

h

This move became fashionable a few years ago, mainly due to the possibility of 7.f4 Wc7 8.Wf3 h6 9.i.h4 g5!, whereby Black offers a pawn in return for dynamic counterattacking chances. Needless to say, I have no intention of allowing anything like this.

7.'1'e2! Playing in the same spirit as the 6 . . . e6 7.f4 ltl bd7 8.We2 line, as covered later in Chapters 8 and 9. Black can transpose to that variation by playing . . . e6 at various points over the next few moves, but the present move order also offers a lot of independent possibilities. In this chapter we will deal with 01) 7 Wa5?!, D2) 7... e5 and 03) 7 b5. •••

••.

7 ... g6 is the topic of the next chapter. 7 . . . h6 will be covered in Chapters 3 and 4.

01) 7 �a5?! • • •

It seems to me that the black queen will be moving around rather aimlessly in this line, while White's next move actually serves a useful purpose.

The bishop might not appear all that impressive here, but it clears the path for White's kingside pawn expansion without spending extra time. Black's idea is only j ustified against: 8 . f4 h6! (The immediate 8 . . . e5 could be played with similar ideas, but it is worth chasing the bishop away first, since now after 9 . fXe5 dxe5 1 O. ltl fS h 6 White has a n interesting extra option i n 1 1 .i.d2 .) 9.i.h4 e5! Now it makes n o sense for the knight to go to f5 because of . . . g6, so White is essentially forced to play 1 O . ltl b3 Wc7. The resulting position can be compared with the 7 . . . h6 8 .i.h4 g6 variation, as covered later in Chapter 4. Here Black has avoided the weakening move . . . g6, improving his chances in the battle that lies ahead. 8.h4 is similar to the main line, especially if Black reacts with 8 . . . h6, but the inclusion of these moves is not particularly in White's favour, since Black will get some extra options involving . . . ltl e 5 , without having to fear the reply £2-f4 .

12

6 liJbd7 . . .

suspicious version of a Keres Attack.) 1 0 . .ig2!? This seems simplest. ( l 0.g5 b4 leads to some complications, although after l 1 . liJd5 liJ xd5 1 2.exd5 .ib7 13 . .ig2 liJ b6 1 4. 0-0 White is still doing well.) 1 0 ... .ib7 ( l 0 ... b4 l 1 . liJ d5 liJxd5 1 2.exd5±) l 1 .g5 b4 1 2.liJ d l ±

1 3 .g5 liJ g4 things are far from clear.) 12 ... liJe5 1 3 .g5 hxg5 1 4.hxg5 l'!xh l t 1 5 . .ixh l liJ fd7 1 6. 0-0-0 liJ b6 1 7.f4 liJ ec4 1 8 . .ie l .id7 1 9.g6 fxg6 20 . .ih4 l'!c8 White's position remained preferable, but it was still a complicated fight in Ter Sahakyan - Karthikeyan, Chennai 20 1 3 .

9.�b3 e6

1 l %Yc7 •••

9 . . . liJ e 5 ? is met by 1 O .f4± and the knight has no good squares. It would have been a different story had White played 8 .h4, as noted above. 9 . . . g6?! is also poor: 1 0. 0-0-0 .ig7 l 1 .g4 liJe5 ( l 1 . . .h6 1 2. f4±) Now 1 2.h3!± is a further reminder of why the h-pawn should be held back on move 8 .

1 1 . . . .ie7 1 2 .h4 Vlfc7 ( l 2 . . . liJe5? 1 3 .f4 liJ exg4 1 4.e5+-; 1 2 . . . h5 1 3 .g5 liJ g4 1 4 . .ih3±) 13 . .ih3 !?t White's initiative is developing quickly, the immediate threat being g4-g5.

12 ..ig2! Preventing . . . b 5 , which can now be refuted by e4-e5.

10.g4

12 ... �e5

1 0. 0-0-0! ? is also perfectly playable, and should lead to similar play to the main line below.

1 2 . . . l'!b8 and 1 2 . . . .ie7 are both met by 1 3 .h4± , when Black does not have much of an answer to the imminent g4-g5.

10 ... h6

13.f4 �c4 14 ..te1 !? There were certainly other options, since the exchange on d2 would not have been a big deal, but the text move works well. The idea is to open things up with a quick e4-e5.

8 7 6 5 4 3 2 1 a

b

e

d

e

f

g

h

1 1 .0-0-0!?N I I .h4 gives White a decent position, but I don't see any reason to rush with it. 1 1 . . .'�c7 ( l 1 . . .liJe5 1 2. f4 liJ exg4 1 3 .h5!N e5 1 4. f5 g6 1 5 .hxg6 fxg6 1 6.liJd5 liJ xd5 1 7.'�xg4 liJ e7 1 8 . 0-0-0 .id7 1 9.'IMI'h4 strongly favours White.) 1 2 . .ig2 (In the event of 1 2 .l'!h3 ! ? h5

a

b

e

d

e

f

g

h

14.. J1 b8 14 . . . .ie7 is well met by 1 5 .l'!d4! liJ b6 (or 1 5 . . . b5 1 6.e5±) 1 6.e5±.

Chapter 1 - Rare 6th Moves

15.e5! dxe5 16.fxe5 tild7 1 6 . . . lLlxe5 1 7 . .ig3 +,

17 ..tg3± White is well ahead in development, and lLld5 is a nasty threat.

D2) 7 ... e5

13

1 2 . . . b4 1 3 ..ixf6! is a typical idea, since an immediate lLl d 5 isn't possible. 1 3 . . . lLlxf6 1 4 . lLl cd5 lLl xd5 1 5 .lLl xd5 a5 1 6.h4;t Note that the . . . b4 move has weakened Black's queenside and presented White's bishop with the important c4-square.

13.a3!

6

An important move, since . . . b4 was becoming an annoying threat. If White can successfully stifle his opponent's queenside play, his kingside potential will come into its own.

5

13 .. J�c8

8 7

1 3 . . . 0-0 1 4 . b 1 Wfc7 1 5 .g4± gives White an automatic attack.

4 3 2 1 a

b

e

d

e

f

g

h

It seems strange for Black to play this so early, particularly since White's knight is allowed to go to f5 rather than to the more passive b3-square. Nevertheless, it is worth considering how best to exploit White's positional advantages.

8.tilf5 h6 9 ..th4 g6 10.tLle3

Compared to the more popular variation involving 7 . . . h6 8 . .ih4 g6 9 . 0-0-0 e5 1 0.lLl b3, as covered later in Chapter 4, White's knight is ideally placed on e3. True, he has spent one more tempo, but that is a small price for the huge improvement in prospects for this piece. Not only is d5 within the knight's reach, but ideas of jumping towards the kingside are in the air.

a

b

e

d

e

f

g

h

14.�d2!?N

Keeping the bishop on h4 restricts Black's play, since he can't start moving around with his knights yet. 14 . .ie 1 is also somewhat better for White, but it gives Black a bit more freedom, and in N. Kosintseva - Bodnaruk, Moscow 20 1 0, Black eventually prevailed.

10 ....te7 1 1 .0-0-0 b5 12.£3 .tb7

12 . . . 0-0 should be met by 1 3 .a3!, with a possible transposition to the main line in the event that Black develops his bishop to b7.

14 �f8 •••

It seems logical to park the king on g7.

14

6 . . lLl bd7 .

1 4 . . . �c7 should be met by l S . m b l;!; , waiting to see what Black does next.

tD d4 1 B .E:h3 as ( I B . . . mg7 1 9 .hS±) 1 9.mb 1 b4 ( I 9 . . . .ic6 20.gS! tD hS 2 1 .tDedS .ixdS 22.tDxdS E:xc2 23.�e 1 ±)

1 4 . . . tD cS l S . .ixf6! Now that Black cannot take back with the knight, this exchange gives White complete control of the dS-square. 1 S . . . .ixf6 1 6.mb 1 .igS 1 7.g3;!; 14 . . . 0-0 l s .mb 1 tD xe4!? ( I S . . . �c7 1 6.g3!± avoids any ... tD xe4 tricks and intends .ih3 or tD g4 next.) 1 6.tDxe4 .ixh4 1 7.tDxd6 E:c7 1 B .tDxb7 E:xb7 1 9.93 .ie7 20.�e2;!; Returning to the main move, it is important to demonstrate a bit of precision. 8

b

a

e

d

e

f

h

g

20.tDbS! tDxbS 2 1 ..ixbS bxa3 22.tDc4!±

16.j,g3! This is the idea behind g2-g4. White does not have to waste time putting his king on b 1 here, and can proceed with the blunt plan of h4-hS .

7 6 5 4

1 6 ..i f2 d S ! ? 1 7. tD exdS tDxdS 1 B .tDxdS .ixdS 1 9 .exdS .igS 20 . .ie3 .ixe3 2 1 .�xe3 �gS 22 .�xgS hxgS looks like decent compensation for Black.

3 2 1 a

b

e

d

e

f

g

h

1 5.g4! 1 s .mb 1 is inaccurate here due to 1 s . . . mg7!, threatening ... tDxe4. 16 . .if2 (With the rook on hB, 1 6.g3 can be met by 1 6 . . . gS! 1 7. tD f5 t mfB+.) 1 6 . . . tD b6 1 7.g4 d S ! ? 1 B . tD cxdS tD fxdS 1 9.exdS .ics�

1 5 ...
8 7 6 5 4 3 2 1 a

b

e

d

e

f

g

h

16 ... ftlc5 1 6 . . . dS 1 7. tD cxdS tDxdS 1 B .exdS;!; does not give Black any serious compensation.

Chapter 1 - Rare 6th Moves 1 6 . . . �a5 can be met by 1 7. h4, since the exchange sacrifice on c3 isn't really a threat. 1 7 . . . lLlc5 07 . . . �dB I B .h5;!;) I B .�b l lLle6 1 9 .h5;!;

17.h4 tLle6 18.c,1;>bl tLld4 19.�g2 White has an excellent position, for instance:

19 ... a5 20.f4!?± The last move reminds us that h4-h5 is not the only strong plan available.

03) 7... b5 8 7 6

15

Some other games have continued: 9 .�xf6 lLlxf6! 1 0.e5 1 O. f4 b4!? is a forcing roure to equality. ( l 0 . . . g6N is a playable alternative if Black wants to maintain the tension.) I l .lLld5 lLlxd5 1 2.exd5 �xd5N ( l 2 . . . g6! ?N is also probably okay.) 1 3 .lLlb5 e6! 1 4.l'!xd5 �e7 1 5 .lLlxd6t �xd6= 1 O . . . dxe5 I l .lLldxb 5 1 1 .�xe5 �d6= 1 1 . . .lLld7 1 2 .lLla3 This was played in Grabarczyk - Heinemann, Schwetzingen 20 1 3 . If White has nothing better than retreating his knight in this way, it is hard to believe that Black can have any problems. A good continuation is: 12 . . . g6! ?N 1 3 .h4 �g7 1 4.h5 �c7 1 5 .�e3 0-0 1 6.hxg6 hxg6 1 7.�h3 l'!fdB!+

5 4 3 2 1 a

b

e

d

e

f

g

h

This move has become quite fashionable in the last few years. White's most popular continuation has been 03 1) 8.0-0-0. I spent some time analysing this, but was not completely satisfied. I will give a quick summary of this line before moving on to 032) 8.f4!, which I believe to be more accurate.

03 1) 8.0-0-0 I believe Black is doing okay after this move, but it takes some precision to show it. So far nobody has played Black's best line, so I will quickly show you the line that caused me the most trouble.

8 ...�b7 9.f4

a

b

e

d

e

f

g

h

9 .. J�c8! 10.e5 1 0.�xf6N gives Black a choice. 1 O . . . gxf6 is simplest, while the provocative 1 0 . . . lLlxf6!? I l .e5 l'!xc3 1 2.exf6 l'!c5 would have to be checked further.

10 ... dxe5!N 1 0 . . . l'!xc3 was recommended in Grandmaster Repertoire 6. It was a novelty at the time, and has been tested a few times since then, but it is not the best move. 1 1 .exf6 gxf6 1 2. bxc3 fxg5 1 3 .lLlf5 ! 0 3 . fxg5 brought White success in a

16

6 . . . tlJbd7

13 geS 14.Y!Yf2!?

recent blitz game, but matters are not so clear after 1 3 . . . ig7N.) 1 3 . . . e6 1 4.ttJxd6t ixd6 I S .!hd6 V!!c7 1 6.V!!d 2 gxf4 This position was reached in Ziatdinov - Harsha, Chennai 20 1 3, and was also the ending point of Ftacnik's analysis. At this point White should play 1 7.ie2!N intending !l:d 1 , with a powerful initiative.

1 4. fxg7 ixg7 I S .V!! f2 gives White no advantage after: I S . . . V!!b 6 ( I S . . . O-O? 1 6.ttJf5±) 1 6.!l:h3 0-0 1 7.!l:g3 \!lhB I B .ixe7 ( I B .if4 !l:dS 1 9 .!l:xg7 \!lxg7f±) I B . . . !l:xe7 1 9 .!%xg7 V!!h 6t 20.!l:d2 (20.!l:gS f6 2 1 .ttJ fS fxgSFF) 20 . . . V!!xg7 2 1 . ttJ fS V!!f6 22.ttJxe7 V!!xe7 23 J'!xd7 V!!xd7 24.V!!f6t \!lgB=

l 1 .fxeS gxc3 12.exf6

14 gxf6

•.•

•••

1 2. bxc3 ttJ e4! 1 3 .e6 fxe6 1 4 .id2 V!!b 6� is pretty good for Black. After 1 2.ixf6 !l:cS White lacks a decisive combination . I found nothing better than 1 3 . ttJ f5 gxf6 1 4.e6, when Black can, at the minimum, force a draw with 1 4 . . . !l:dS I S .exf7t \!lxf7 1 6.V!!h St \!lgB 1 7.V!!g4t \!lf7, when White is forced to repeat with I B .V!!h S t .

1 4 . . . exf6 I S .if4 !l:dS 1 6.c4 bxc4 1 7.ixc4::!; 8 7 6 5 4 3

8

2

7

1

6

a

5

b

e

d

e

f

g

h

l S .te3 •

4

Black's lack of development seems worrying, but the computer is happy, and his position contains a lot of resources.

3 2

l S gg8 16.gg1CD

1

••.

a

b

e

d

e

f

g

h

12 gcS! •••

The position looks scary for Black, but everything holds together.

13.h4 1 3 . ttJ f3 !l:dS is unclear, while 1 3 . fxe7 iJ.xe7 1 4.ixe7 V!!xe7 I S .V!!x e7t \!lxe7 1 6.ie2 !l:eB 1 7.iJ.f3 ixf3 I B .ttJxf3 ttJf8 reaches an equal endgame.

The position is rich in possibilities for both players, but I do not believe White can claim an advantage.

D32) 8.f4! By forcing the issue in the centre, White provokes . . . eS and avoids the plan of . . . ib7, . . . EkB and . . . !l:xc3 that we saw previously.

17

Chapter 1 - Rare 6th Moves

I l . tD cb 5 ! I l . tD a4 is met b y 1 1 . . .tDd5 when the f4pawn is hanging. 1 1 . . .axb5 1 2.exf6 gxf6 1 2 . . . WI'c5 1 3 .WI'e4 d5 1 4 . .ixb5 t .id7 1 5 . .ixd7t 'it>xd7 1 6. tD b3! dxe4 1 7.tDxc5t 'it>c6 1 8 .tDxe4 exf6;:!; Black has some drawing chances, but White is playing without risk. 1 3 .WI'f3 ! �a5 1 3 . . . .ib7 1 4 . .ixb5t 'it>d8 1 5 .ic6 'it>c8 1 6.0-0 �a6 1 7 . .ixb7t Wl'xb7 1 8 .WI'd3 'it>b8 1 9 .a4t b

a

8

...

e

d

e

f

g

h

eS

This has been played almost exclusively, although there are some other important options to consider. 8 . . . h6 9 . .ih4 transposes to 7 . . . h6 8 . .ih4 b5 9.f4, as covered in variation C of Chapter 3 - see page 48. (9 . .ixf6!? tDxf6 1 0 .0-0-0 .ig4 1 1 .tDf3 is another interesting possibility, but there is no need to spend any time on this.) 8 . . . WI'c7!? 9 . .ixf6! ? 9. 0-0-0 is a good alternative. Now 9 . . . e5 1 0. fxe5 dxe5 is discussed under 8 . . . e5, and 8 . . . b4 9.tDd5 transposes to 8 . . . b4 below. 9 . . . e6 is also dealt with elsewhere in the book - see variation B of Chapter 8 on page 1 33 . Having said that, the text move is rather tempting, as it allows White to fight for the advantage by more forcing means. 9 . . . tDxf6 l O .e5 b4 l O . . . dxe5 1 1 . fxe 5 b4 1 2 .tDa4! tDd5 1 3 .e6t

a

b

e

d

e

f

g

h

e

d

e

f

g

h

1 4 .ixb5 t .id7 1 5 . .ixd7t 'it>xd7 1 6. tD b3 Black's king will be rather unpleasantly stuck in all variations, for instance: 1 6 . . . �a6 1 6 . . . �b5 1 7. 0-0-0± 1 7.0-0-0! �xa2 1 8 .'it>b l �a7 1 9 .WI'h5 'it>e8 20.WI'b5t± A final alternative is: 8 . . . b4!? 9.tDd5 ib7

a

a

b

b

e

d

e

f

g

h

18

6 lLl bd7 . . .

1 0. 0-0-0 The knight can be left where it stands, as any opening of the position should be in our favour. 1 O . .ixf6 lLlxf6 l 1 .lLlxf6t gxf6 may also be better for White, but this structure tends to result in an intense strategic fight. Although Black has created some holes in his queenside, White is not ideally developed either. 1 O . . . lLlxd5 1 1 .exd5 V!!c7 1 1 . . . .ixd5 ? ? 1 2.lLlf5 +1 2.g4!? There is no need to defend the d5-pawn. 1 2 . . . .ixd5 1 2 . . . h6 1 3 . .ih4 g5 1 4. fXg5 hxg5 1 5 . .ixg5 .ixd5 1 6.lLlfS .ie6 1 7.V!!e4;:!; 1 3 . .ig2! 1 3 .lLlfS ! ? is playable, but 1 3 . . . .ie6 blocks the attack for the time being. 1 3 . . . .ixg2

Ideally White would prefer to avoid committing to this exchange, but here it seems to be a necessity. If 9 .lLlfS then 9 . . . V!!a 5!N is quite strong. Some messy complications lie ahead, but a bit of analysis convinced me that Black is out of danger. 9 . 0-0-0! ? doesn't work as smoothly here as in the main line below. Here is a shortened version of my analysis: 9 . . . exd4!? 1 O.e5 .ie7 l 1 .exf6 gxf6 1 2.lLld5 fXg5 1 3 .l"!:e l lLlc5!N 03 ... gxf4 1 4.lLlxe7 lLle5 1 5 .lLlc6 V!!c7 1 6.lLlxd4 .ib7 was also pretty unclear in Soumya A. LAmi, Astana 20 1 3 .) 1 4.lLlxe7 .ie6 1 5 .lLlf5 ! ? V!!f6 1 6.g4 This position could be analysed further, but I find White's play in the main line more convincing.

9 dxe5 ••.

8 7

8

6

7

5

6

4 3

5

2

4 3 a

b

e

d

e

f

g

h

1 4J'!:he l ! lLlc5 14 . . . e6? loses immediately to 1 5 .lLlxe6. 1 4 . . . e5 1 5 .'1Wxg2 l"!:c8 1 6.'i!?b l � Black has no easy way to continue developing. 1 5 .V!!xg2 l"!:c8 1 6.'i!?b l Here too, it is hard for Black to continue his development. The main problem is: 16 . . . e6 1 7.f5 e5 1 8 . f6! Black's position is unpleasant, and delaying . . . e6 will not change much.

2 1 a

b

e

d

e

f

g

h

10.0-0-0! 1 O.lLlfS was played in an amateur game in 200 5 . Castling is much more dynamic, and was my choice when I encountered this variation in 20 1 2. Black has many possible moves here. My own game continued with D32 1) 10 exd4, but White must also consider D322) 1O �b6N, D323) 10 h6N and D324) 10 .ib7N. . ..

9.fxe5

•.•

...

. ..

Chapter 1

-

1 0 . . . �c7 gives White a choice. I I .lLlf5 may well transpose to variation D324 after 1 1 . . . .ib7, ' but White can also consider l 1 .lLld5 ! ? lLlxd5 1 2 .exd5 .ib7 1 3 .g4!?±.

19

Rare 6th Moves

1 9 .�xd4t �d5 20.�xd5t .ixd5 2 1 .�e5± is also possible.

19 g4 20 .td3i •••

•

White has a dangerous initiative for a mere pawn.

D32 1) 10 exd4 l l .e5 !J.e7 •••

I found an internet game where 1 1 . . .�e7 was played, but 1 2.exf6 �xe2 1 3 .lLlxe2 led to an obvious advantage for White.

12.exf6 gxf6 13.c!t:Jd5 fxg5 14J�� e l tLlf6

D322) 10 �b6N •••

8 7 6 5 4 3 2 1 a

b

e

d

e

f

g

h

This sets a devious trap. a

b

e

d

e

f

g

h

15.tLlxe7!N 1 5 .lLlxf6t 'it>f8 1 6.�e4 .ie6! was an unclear mess in Negi - Vorobiov, Leiden 20 1 2 .

l l .c!t:Jd5! l 1 .lLlf5 is not so good because of 1 1 . . .b4! , intending 1 2.lLld5 lLlxd5 1 3 .exd5 ? ? �g6! and Black wins a piece.

l l c!t:Jxd5 12.exd5 �g6!? .•.

15 .ie6 16.tLlc6 �d5 •••

After 1 6 . . . d3 1 7.�xd3 �xd3 1 8 . .ixd3 0-0 1 9 .h4± White has easy play thanks to Black's weak kingside structure.

12 . . . .id6?! 1 3 .lLlf5 0-0 1 4.'it>b l lLlc5 1 5 .g4 gives White an ideal position for this structure.

This was the move I missed in my calculations. Now White gets a pleasant position by force.

1 2 . . . .ic5 is slightly better, as the f5-square is off-limits to the white knight on account of the . . . �g6 trick. However, 1 3 .lLlc6 0-0 1 4.g4 .ib7 1 5 . .ig2± still leaves White with a pleasant position.

17 �xc6

13.h4! .id6

After 1 7 . . . lLlg4 1 8 .�xd4 �xd4 1 9 .1Llxd4 Black is facing a tough endgame.

1 3 . . . h6 1 4 . .id2 .id6 1 5 .h5 looks ugly for Black.

18.�xf6 g;,d7 19.h4!?

14.c!t:Jc6

17.�fl!

.••

6 . . . ttJ bd7

20

This is simplest. 1 4 .g4! ? h6 1 5 .i.d2 ttJ f6 1 6. ttJ f5 i s a n interesting pawn sacrifice, but there is no need for it.

other lines. Fortunately, we can prevent this from happening.

l 1 .ttld5! .ib7

14 ... 0-0 1 5.�e7t

1 1 . . .hxg5 ? 1 2. ttJ c6 is a nice queen trap. Taking the knight is not much better: 1 1 . . .exd4? 1 2.e5! i.e7 1 3 .exf6 ttJxf6 1 4.ttJxf6t gxf6 1 5 .i.xf6+-

1 5 .g4!? ttJ f6 1 6.i.h3;!; is also good.

15 ...he7 16.he7 �e8

12 ..ixf6 ttlxf6 8 7 6 5 4 3 a

b

e

d

e

f

g

h

17.d6 .ib7 18.�bl �ac8 19.h5 �e6 20.�h4;1; The bishop on e7 seriously limits Black's active possibilities.

2 1 a

b

e

d

e

f

g

h

13.tthb5! ttlxd5 1 3 . . . axb 5 ? 1 4.'1Wxb5t wins effortlessly.

D323) 10 ... h6N

1 3 . . . i.xd5 1 4.exd5 axb5 1 5 .Wxe5t We7 1 6.i.xb5 t i>d8 1 7.Wxelt i.xe7 1 8 .d6 i.f8 1 9.a4± The queen side pawns are too much for Black to handle.

8 7 6

14.exd5 .ie7 1 5.d6!

5

White still needs to play accurately. The simplistic 1 5 .ttJ c3 would offer Black excellent compensation after 1 5 . . . 0-0 1 6.Wxe5 �e8 1 7.WfS �c8�.

4 3 2 1 a

b

e

d

e

f

g

h

This is quite a critical move. The point is that if the bishop retreats to h4, Black will get some additional options compared to some of the

15 ....ig5t 16.�bl axb5 17.�xe5t �f8 18.d7! It would be slightly misleading to say that White has excellent compensation, ' because he will not even be materially down for more than another move or two.

21

Chapter 1 - Rare 6th Moves

1 1 .t[}5

8

Black has a number of possible moves here, but with the d-file open, the two most natural choices are D3241) 1 l ...'i!Yb6 and D3242)

7 6

1 l ...'i!Yc7.

5

1 1 . . .!:k8 1 2. tt:l d 5 ! Swapping this piece for Black's light-squared bishop is always welcome. 1 2 . . . ixd5 1 3 .exd5 h6 ( 1 3 . . . g6 1 4. tt:l d4±) 1 4.if4 ( 1 4 .ih4 might give White an edge too.) 1 4 . . . 'lWc7

4 3 2 1 a

b

e

d

e

f

g

h

8

18 .if6

7

Black does not have time for 1 8 . . . b4 on account of: 1 9 .ic4 if6 20 .'lWd6t �g8 2 1 .E!:he l +-

4

•.•

1 8 . . . 'lWe7 1 9 .'lWxb5 ie4 20.id3;!; also gives White excellent chances of success.

19.'i!Yc5t �g8 20 .bb5 .ia6 21 .ha6 gxa6 22.'i!Yc8 ga8 23.'i!Yxd8t .ixd8 24.ghel �h7 25.c4± •

It would be surprising if Black could hold this endgame. .••

b

e

d

e

f

5 3 2

a

b

e

d

e

f

g

h

1 5 .!l:d3! Intending to deploy the rook on c3 or e3 . 1 5 . . . 'lWxc2t ( 1 5 . . . b4 1 6.E!:e3±) 1 6.'lWxc2 E!:xc2t 1 7. �xc2 exf4 1 8 .d6;!; 1 1 . . .h6 1 2.ixf6! 'lWxf6 1 3 .g3 Now ih3 is an annoying threat, and it is hard for Black to get a normal position with short casding. 1 3 . . . ic5 ( 1 3 . . . 0-0-0 1 4.a4;!;) 1 4.�b l

D324) 10 .ib7N

a

6

g

h

This is arguably the most challenging move of all.

a

b

e

d

e

f

g

h

1 4 . . . tt:l b6 ( 1 4 . . . g6 1 5 .tt:lxh6! E!:xh6 1 6.'lWd2 is an important point; 14 . . . E!:d8 is met by

6 . . . tDbd7

22

1 5 .i.h3;!; and Black still cannot castle or play . . . g6.) 1 5 .lthb 5 ! 0-0 1 6. tD bd6;!; 1 1 . . . �a5 It is hard to imagine that this can work now. 1 2 .�f3 ! ? Th e most punishing, although the simple 1 2.'it>b l is also possible, for instance: 1 2 . . . b4 1 3 .tDd5 tDxd5 1 4.exd5 i.xd5 1 5 . b3� 1 2 . . .l:kB 1 2 . . . b4 1 3 .i.c4! �c7 ( 1 3 . . . bxc3 1 4.i.xf7t 'it>xf7 1 5 .:gxd7t tD xd7 1 6. tD d6t 'it>e6 1 7.tDxb7+-) 1 4.:gxd7 �xd7 1 5 .i.xf6 gxf6 1 6.tDd5 i.xd5 1 7.exd5 gives White huge compensation. 8

Another promlsmg continuation is: 1 2.i.xf6!? �xf6 ( 1 2 . . . tDxf6 1 3 .tDxb5! tDxe4 1 4.tDg3 leads to rather complicated stuff, but it looks dangerous for Black.) 1 3 .�b l ! ? ( 1 3 .�d2 O-O-Oloo) It is not easy for Black to continue, and the queen on f6 looks out of place. 1 3 . . . i.c5 ( 1 3 . . . 0-0-0 1 4.a4;!;) 1 4 . h4 g6 1 5 .tD e3 tD b6 1 6.h5;!;

12 ... g6 1 2 . . . :gcB is well met by: 1 3 .i.xf6! �xf6 ( 1 3 . . . tD xf6 1 4.g5 tD d7 1 5 .i.h3+-) 1 4 .�d2! �c6 ( 1 4 . . . tD b6 1 5 .g5 �dB 1 6.�f2 �c7 1 7.i.h3± with a crushing initiative.) 1 5 .g5 White threatens i.h3 , and Black has severe problems to even complete his development.

13.tll e3

7

1 3 .tD h6!? b4 1 4.tDd5 tDxd5 1 5 .exd5 f6 1 6.i.e3 �a5 reaches a position analysed under 1 3 . tD h6!? in the notes to variation 03242 below.

6 5 4 3 2

8 a

b

e

d

e

f

g

h

1 3 .i.xf6! :gxc3! ? 1 3 . . . tD xf6 i s met b y 1 4 .�g3± hitting both e5 and g7. 1 3 . . . gxf6 is ugly, and 1 4.�h3 tD b6 1 5 .�h4 tD d7 1 6.:gxd7! 'it>xd7 1 7.�xf6 is dangerous for Black. 1 4. bxc3 gxf6 1 5 .�d3 �c7 1 6. c4;!;

D3241) 1 l ...YlYb6 The queen is slightly misplaced on b6, and White has more than one pleasant way to continue. He does, however, need to watch for . . . :gxc3 ideas.

12.g4

7 6 5 4 3 2 1 a

b

e

d

e

f

g

h

1 3 ... h6 14.Lf6 tll xf6 15.h4 j,g7 After 1 5 . . . i.c5 1 6.tDed5 tDxd5 1 7.tDxd5 i.xd5 I B .exd5 0-0 1 9.h5;!; White's bishop has the greater attacking potential, while Black's bishop really belongs on g7.

16.h5 0-0

Chapter 1 - Rare 6 th Moves

23

cxb2t 1 5 . 'it>b 1 looks horrid for Black, for instance: 1 5 . . . .ixg7 1 6.ltJxg7t 'it>ffi 1 7. ltJ 5 ltJ c5 I B .'lWc4! .ixe4 1 9 .:B:g l ±) 1 4.ltJd5 ltJ xd5 1 5 .exd5 0-0-0 16 . .ig2;!; Black's queenside is rather exposed.

a

b

e

d

e

f

g

h

17.g5! �xh5

17 . . . hxg5? I B .h6 .ihB 1 9 .h7t wins.

18.�cd5 V9d8 19J�xh5 gxh5 2o.V9xh5 V9xg5 21 .V9xg5 hxg5 22,c�� e7t �h7 23.�d3± Black will soon have to give up the dark­ squared bishop to avoid mate, leaving White with two superb minor pieces for a rook.

1 2 . . . h6 1 3 . .ixf6 ltJ xf6 1 4. ltJ d 5 ! ? ltJxd5 1 5 .exd5 0-0-0 1 6 . .ig2 Compared with 1 2 . . . b4 above, here Black has not weakened his queen side any further. Nevertheless, he still faces concrete problems: 1 6 . . . g6 1 7.:B:d3! ( 1 7 . ltJ e3 .ic5=) 17 ... gxf5 I B .:B:c3 .ic5 1 9 . b4 'it>bB 20.:B:xc5 'lWe7 2 1 .gxf5;!; Let's return to the main line where Black forces the knight away from 5. Here White can choose between an aggressive move and a more solid one. Both options offer him better chances, so the choice depends on personal taste.

D3242) 1 l ...V9c7 8 7 6 5 4 3

a

2

13.tLle3

b

e

d

e

f

g

h

This is the solid option.

1 a

b

e

d

e

f

g

h

12.g4 g6 1 2 . . .:1kB?! is an unfortunate choice due to 1 3 . .ixf6! ltJxf6 1 4.g5 ltJ d7 1 5 . .ih3 when Black is under serious pressure. 1 2 . . . b4 1 3 . .ixf6 ( I 3.ltJd5 may also be better for White.) 1 3 . . . ltJxf6 ( I 3 . . . bxc3 1 4 . .ixg7

1 3 . ltJ h6!? Normally a knight on this square will be extremely annoying for Black, as his king would be stuck in the centre, but here there are some concrete lines that must be considered. 1 3 . . . b4 1 4.ltJd5 ltJ xd5 1 5 .exd5 f6 1 5 . . . .id6 1 6 . .ie3;!;

6 ttJ bd7

24

. . .

1 5 .. .l:k8 1 6.�b l f6 1 7.i.e3 i.c5 1 8 .i.c 1 ! ?;!; The bishop on c5 is in the way of Black's other pieces. 1 6.i.e3 �a5 Without this forcing move, Black's king would have come under heavy fire. 1 7.�b l i.xd5

1 3 . . . i.g7 1 4.i.g2 0-0 1 5 .h4 h6 1 6.i.xf6 ltJxf6 1 7.h5;!; maintains the pressure.

8 7 6 5 4 3 a

2

b

e

d

e

f

g

h

14.hf6

a

b

e

d

e

f

g

h

1 8 J::\ xd5 ! �xd5 1 9 .i.g2 e4 20 .i.f2! White is able to win back the material. 20 . . . i.xh6 2 1 .i.xe4 �e6 2 1 . . . �e5 22 .i.g3;!; 22.l'!e 1 ltJe5 8 7 6

1 4.i.h4!? is another idea that could be explored. 1 4 . . . i.e7 ( 1 4 . . . i.g7 1 5 .i.g2 b4 1 6. ltJ cd5 ltJxd5 1 7.ltJxd5 i.xd5 1 8 .exd5 ± With both o f his bishops still on the board, this structure is excellent for White.) 1 5 .�b l b4 Now White is forced to take on f6, but I would argue that the bishop on e7 is less than ideally placed. ( 1 5 . . . 0-0 1 6.g5t) 1 6.i.xf6 ltJxf6 1 7. ltJ cd5 ltJ xd5 1 8 .ltJxd5 i.xd5 1 9 .exd5 h5 ( 1 9 . . . 0-0 20.h4;!;) 20.i.g2;!;

14 ... 111 xf6 lS.YlYf3!?

5

White's plan includes i.d3 , h2-h4, and a pawn advance to g5 or h5.

4 3 2

a

b

e

d

e

f

g

h

23 .i.xa8 Unless Black is able to solve his problems quickly, White's bishop pair and better pawn structure will give him excellent chances.

13 .. h6 .

1 3 . . . i.e7 1 4.i.h6! keeps the king pinned down in the centre.

1 5 .h4 is slightly less ambitious, but still leads to a pleasant position: 1 5 . . . l'!d8 1 6.ltJed5 ltJxd5 1 7.ltJxd5 i.xd5 1 8 .exd5 h5! ( 1 8 . . . i.d6 1 9 .h5;!;) 1 9 .95 i.g7 20 .i.g2 0-0 Black is solid, and it is not easy for White to break through on the kingside. Nevertheless, he can keep a risk-free edge with 2 1 .i.e4 followed by piling up the heavy pieces on the f-file.

Chapter 1 - Rare 6th Moves

25

Conclusion Starting with Black's rare 6th moves, 6 . . . e5 is clearly dubious, but 6 . . .id7 and 6 ... lL'lc6 both have some tricky points, so it is worth checking both these lines.

a

b

e

d

e

f

g

h

15 ...J.c5

1 5 . . . ig7 also falls short of equality: 1 6.h4 �dB 1 7.id3 b4 0 7 . . . 0-0 I B .g5±) I B .likd5 lL'lxd5 1 9 .1L'lxd5 ixd5 20.exd5 0-0 (20 ... h5 2 1 .ixa6;!;) 2 1 .h5t

16.lLled5 lLlxd5 17.tihd5 hd5 IS.exd5

a

b

e

d

e

f

g

h

White can press for a long time in this opposite-coloured bishop position. His bishop has better potential, and it is not obvious what Black should do with his king, or how he should defend his light squares against the plan of id3 and h4-h5 .

The most important function of this chapter has been to introduce the increasingly popular 6 . . . lL'lbd7 variation, against which I favour nWe2! intending some combination of long castling and a central attack with tL-f4 . 7 . . . "lMfa5 ?! does n o t make much sense; the bishop simply retreats to d2 and White's initiative develops smoothly. 7 . . . e5 seems like a surprising choice, bur it is worth paying attention to some of the details that follow, as White has to play accurately to exploit the weakness of the d5-square. The largest and most important section of the chapter was devoted to 7 . . . b 5 . Then B . f4! e5 9 . fxe5 dxe5 1 0. 0-0-0! is an idea that I tested in one of my own games, and awaits further tests. The position is rich in possibilities for both sides, but more dangerous for Black, whose king is unlikely to feel safe for quite some time.

7 g6 ...

Variation Index l .e4 cS 2.li)a d6 3.d4 cxd4 4.li)xd4 li)f6 S.li)c3 a6 6 ..tgS li)bd7 7.¥Ne2! g6 8.0-0-0 A) 8...¥Nc7 B) 8 eS C) 8....tg7 9.h4!? Cl) 9 ...h6 C2) 9 ...hS 1 0.b l ¥Nc7 l 1 .g4!?N C2 l ) 1 l ...li)b6 C22) l 1 ...hxg4 l 2.hS! gxhS l 3.gxhS gxhS l 4.li)f5 .tm l S ..tg2 C22 l ) l S ...e6 C222) l S ..J�b8 C223) l S ...¥NcS!

27 28 31 31 34 35

•••

A) after 8 ...Wfc7

C2) after 1 O . . . Wfc7

C l ) after 9 . . . h6

8

8

8

7

7

7

6

6

6

5

5

5

4

4

4

3

3

3

2

2

2

a

b

c

d

e

9.h4!N

f

36 38 39 40

g

h

a

b

c

d

e

f

IO.�xf6!N

g

h

a

b

c

d

e

II.g4!?N

f

g

h

27

Chapter 2 - 7 . . . g6

A) 8 VNc7

l.e4 c5 v!i:Jf3 d6 3.d4 cxd4 4.c!thd4 tLlf6 5.c![}c3 a6 6.�g5 tLlbd7 7.VNe2! g6

•••

This position has occurred a few times via a 7 . . . Wic7 move order. Black's set-up is quite interesting; having moved the queen off the d-file, he intends to meet h2-h4 with . . . h6, and ixf6 with . . . It'l xf6, without allowing e4-e5 at the end of the sequence. However, White can throw a stick into the wheel.

a

b

e

d

e

f

g

h

The kingside fianchetto is one of the most important options afforded to Black by 6 . . . lt'l bd7, and in the present chapter he makes use of it immediately. This never used to be a popular choice, but then in 20 1 2, against Alekseev, Grischuk showed that Black could go for the Dragon set-up and simply ignore White's plan off2-f4 and e4-e5 . That game had continued 7.f4 ig7 8.'1We2, but of course the same plan was instantly adopted against 7.Wie2 as well. However, the fact that White has not yet committed himself to f2-f4 provides him with interesting and relatively unexplored additional options.

8.0-0-0

9.h4!N 9 . lt'l d 5 It'l xd5 1 0.exd5 was played in Kosteniuk - Dzagnidze, Beij ing 20 1 3, but now 1O . . . lt'l b6N would have led to an unclear game. 9.f4 e 5 ! ? is surprisingly hard to deal with. For instance, after 1 0. fxe5 dxe5 l l .lt'l b3 b5, the position can be compared with variation B of Chapter 4 on page 5 9 . The key difference is that here Black has improved his chances by avoiding weakening his kingside with . . . h6. 9...h6 9 . . . h5 1 0 .�b 1 ig7 is covered later under the move order 8 . . . ig7 9.h4 h5 1 0.�b 1 Wic7 - see variation C2 on page 34. 8 7 6

8.h4?! is an imprecise move order because of 8 . . . h6!. Now 9 .ie3 h5!? 1 0. 0-0-0 ig7 transposes to 8 .0-0-0 ig7 9.h4 h6 1 0 .ie3 h5! - something we are trying to avoid, as explained later in variation C l . 9 .ixf6 is not much good due to 9 . . . lt'l xf6! when, in contrast to variation C l , the e4-e5 push is toothless. After the text move it is worth considering A) 8 VNc7 and B) 8 e5, but the main move is obviously C) 8 j,g7. •••

•••

•••

5 4 3 2 1 a

b

10.e5!! This leads possibilities.

e

d

to

some

e

f

g

amazing

h tactical

28

6 . lLl bd7 . .

10 ...dxe5 1 O . . . hxg5 l 1 .exf6 ltJ xf6 ( l l . . .e5 1 2.ltJd5 looks horrid for Black.) 1 2. ltJ db 5 ! +- is a nice point.

has good winning chances due to his extra pawn and the exposed black king.

21 .VNg8t @xc7 22.fxe5 VNe6

1 1 ..txf
12.�d5 VNd6 1 2 . . . Wd8 1 3 .h5! undermines the light squares, creating additional squares for the d4-knight too. Despite having all those pawns, Black's position is precarious. 1 3 . . . ltJ b6 1 4.ltJxb6 Wxb6 1 5 .hxg6 fxg6 1 6.We4!± 8

a

7

b

e

d

e

f

g

h

23.@bl ! 2 3 . .id3 Wxa2 24.e6 g fB 2 5 .Wh7 md8 could get really messy.

6 5 4

23 ... fxe5 24 ..td3 .ib4 25J!dl

3

Black is still in the game, but White's extra exchange makes it a mostly one-sided affair.

2

B) 8 e5

1

•••

a

b

e

d

e

f

g

h

13.VNc4 exd4 14.�c7t @d8 1 5.�xa8 b5 16.VNxf7 White's spectacular play has netted an extra exchange, but he still needs to play accurately to save the a8-knight.

16 ... .tb7 17J!e1 .te7 1 7 . . . Wf4t 1 8 .�b l .ic5 1 9.'1Mfg7! ge8 20.gxe8 t �xe8 2 1 .Wxg6t �d8 22.Wg8t ltJ fB 23.ltJb6 Wd6 24. ltJ c4 bxc4 2 5 . .ixc4±

18.�c7! �e5 19.VNg7 gm 20.£4 gf7 Alternatively, after 20 . . . d3 2 1 .gd l Wxc7 22 . .ixd3 ltJ xd3t 23.gxd3t �c8 24.gc3 .ic6 2 5 . gxc6 Wxc6 26.Wxe7 ge8 27.Wf7 White

a

b

e

d

e

f

g

h

In the analogous position with . . . h6 and .ih4 included, the same move is quite good

Chapter 2 for Black. In fact, if you turn ahead to the beginning of Chapter 4, you will see that with the bis hop already on h4 I recommend playing f2-f4 before castling, to avoid this very possibility. However, with the bishop still on g5 White benefits from some extra options.

9.tLld5!?N The most ambitious. 9.liJb3 may lead to a slightly odd version of an English Attack. 9 . . . h6 (9 . . .ie7 l O.ih6! is annoying for Black.) 1 0 .ie3! ( l 0 .ih4 is the same kind of unfavourable transposition as outlined above.) The bishop on e3 makes it harder for Black to develop, and White's g- and h-pawns are free to roam forwards. l O . . . b5 l U3 Now 1 1 . . .liJ b6!?N looks like Black's best chance, and could be explored further. (Instead after 1 1 . . .ib7 1 2.a3 E:c8 1 3 . "Mrd2 the . . . E:xc3 threat is easily parried, and . . . liJ b6 will usually be met by liJ a5 . 1 3 . . . "Mrc7 1 4.Wb l ± Yu - Gordon, Gibraltar 20 1 3 .

-

29

7 . . . g6

IO.�c4! hxg5 1 0 . . . exd4 1 1 .ixf6! liJ xf6 1 2. liJ c7t c:J;e7 1 3 .liJ xa8 ie6 1 4."Mrb4 "Mrxa8 1 5 .e5 gives White a huge attack against the un castled king.

1 1 .tLlc7t �xc7! This leads to a double-edged situation involving three pieces against a queen. 1 1 . . . c:J;e7 is playable, but it leads to more one-sided positions where Black will have to fight for a draw an exchange down. 1 2 . liJ de6 liJ b6 1 3 .liJxd8 liJ xc4 1 4.liJxf7 (White can also consider: 1 4.ixc4 c:J;xd8 1 5 .liJxa8 liJ xe4 1 6.E:he l liJ xf2 1 7.E:d2 liJ g4 1 8 .ixf7 E:xh2 1 9 .ixg6±) 1 4 . . . liJ xb2 1 5 .Wxb2 Wxf7 1 6.liJxa8 liJxe4 1 7.id3±

12.�xc7 exd4

The computer gives a positive evaluation for White, but it is not at all easy for a human player to assess the position. Fortunately there are some concrete ways to pose problems for Black.

13.:1hd4

9 ... h6 9 . . . exd4? 1 0.e5 is horrible. After 9 . . .ig7 1 0.liJb3! h6 I l .ih4± the bishop is misplaced on g7. It won't be able to help the d6-pawn and, even more importantly, the pin will remain a source of great annoyance.

1 3 .ic4 ie7 1 4.E:xd4 allows Black to get organized with 1 4 . . . c:J;f8! 1 5 .f3 c:J;g7 1 6.ib3 liJ e8 1 7."Mrc4 liJe5 1 8 ."Mre2. The struggle goes on, and the computer prefers White, but I think Black's position is acceptable.

13 ... g4 1 3 . . . liJe5 can be met in the same way.

8 7

8

6

7

5

6

4

5

3

4

2

3

1

2 a

b

c

d

e

f

g

h

1 a

b

c

d

e

f

g

h

30

6 . . . �bd7

14.gxd6! Giving up material for the sake of the initiative. After 1 4 . .ic4 .ih6t l S .<.!lb 1 .if4 Black regroups his pieces excellently.

14 ...hd6 15.�xd6 Suddenly Black's position doesn't look so well organized, despite having a rook and two knights against a queen.

15 ... �xe4 16.�b4 1 6.'Wd4 also deserves attention. 1 6 . . . llJ ef6 1 7 . .ic4 0-0 ( l 7 . . . <.!lfB 1 B .ge 1 <.!lg7 1 9 . f3 gxf3 20.g4±) 1 B . h3 g3! ( l B . . . gxh3 1 9.E!:xh3 is too dangerous.) 1 9. fxg3 bS 20 . .ib3 as 2 1 .a3 (2 1 .a4!?) 2 1 . . . .ib7 22.E!:fl a4 23 . .ia2 E!:adB 24.E!:xf6 llJ xf6 2 S .'Wxf6 .idS I think Black should be able to hold.

After 1 7 . . . llJ f8 1 B .'Wc3 rj;;e7 1 9.E!:e l t llJe6 20.f3 gxf3 2 1 .'Wa3t <.!leB 22.'Wxf3 followed by E!:fl , Black's position is hard to keep together.

18.gelt ge5 An interesting position has arisen. Despite being far behind in development, Black is not in any immediate danger. White is ahead in development, but he is running out of pieces and needs to find a plan . 8 7 6 5 4 3 2

8

1

7

a

6

b

c

d

e

f

g

h

19.a4!

5

Preparing to stifle Black's intended development of his queenside pieces.

4 3

19 ... gxelt

2

1 9 . . . b 6 j ust loses the pawn: 20.E!:xeSt llJxeS 2 1 .'Wxb6±

1 a

b

e

d

e

f

g

h

16 ... �ef6!

1 6 . . . llJ df6 1 7 . .id3±

1 6 . . . llJ ecS 1 7 . .ic4 0-0 1 B .h4! Opening up the h-file will give White a strong initiative, with the rook and queen combining to create mating nets.

17.i.c4 gh5! This is the only way to prevent an instant collapse.

1 9 . . . aS gives away the critical bS-square. 20.E!:xeSt llJ xeS 2 1 ..ibSt llJ fd7 (2 1 . . .llJed7 22.'Wd6±) 22.'Wd6 <.!ldB 23.h3 gxh3 24.gxh3 White intends £2-f4, and 24 . . . gS is not much of a defence due to 2 S . f4! gxf4 26.h4 when the h-pawn seems unstoppable.

20.�xel t �f8 21 .�b4t �e8 Black should not allow the queen to go to e7.

22.a5

31

Chapter 2 - 7 . . . g6 How should Black continue development now?

9.h4!? 9 . f4 0-0 leads to a different type of game, as discussed briefly in the introductory comments. Black has achieved good results so far, although there is still a lot to be explored here. I prefer to utilize the fact that White has not yet committed to £2-f4 . Black may react to the advance of the h-pawn with Cl) 9 h6 or C2) 9 ... h5 . ..•

a

b

e

d

e

f

g

h

Neither 9 . . . Wc7 1 0.hS!t nor 9 . . . 0-0 1 0.hS lLlxhS l 1 .lLlf5 ! can be considered acceptable for Black.

Cl) 9 ... h6

22 b5!? ••.

Normal development with 22 . . . lLle5 is difficult since after 23.We l ! lLlfd7 24 . .id5 the b7-pawn is hard to protect. 24 . . . ffi (24 .. .l::!:b 8 2S .h3!±) 2S . .ixb7 !:!:b8 26 . .idS;!;

23.a:xb6 gb8 White has a few ways to maintain some pressure without taking any risks. For instance:

24.Wb3 !:!:xb6 25 .ixf7t �e7 26.�c4 �b7 27.hg6 hg2 28.Wc3 •

The endgame will go on, but Black will have to struggle to establish a defensive fortress.

a

b

e

d

e

f

g

h

This is the more natural-looking response, but it has a drawback.

C) 8 ...Ag7

IO .ixf6!N •

Exploiting the fact that Black cannot recapture with the knight. 1 O . .ie3 looked like a good move after 1 0 . . . b S ? , when l 1 .hS! lLleS 1 2.hxg6 fxg6 1 3 . f4 gave White an ideal position in Durarbayli Georgiev, Dubai 20 1 4 . However, 1 0 . . . hS! i s a stronger move which gives Black an improved version of variation a

b

e

d

e

f

g

h

32

6 . tLl bd7 . .

C2, since the bishop on e3 gets in the way of the f1-f4/e4-eS plan, and also does not support an attack on the e7-pawn like the bishop on gS . When we look ahead to variation C2, we will see that Nakamura chose 9 . . . hS over 1 O . . . h6. There is little doubt that he would have preferred the version with the bishop on e3, so we can infer that he was put off 9 . . . h6 by the prospect of the text move.

lO ....bf6 l O . . . tLl xf6? l 1 .eS! tLl h S ( 1 1 . . . dxeS 1 2. tLl c6+-) 1 2. e6 looks extremely bad for Black.

H .h5 This is the most solid option, aiming for a slight edge. l 1 .tLl d S ! ? Planning h4-hS on the next move i s more ambitious, since Black cannot blockade the position with . . . �xd4 as in the main line, but it is harder to reach an exact conclusion here. 1 1 . . . �g7 1 1 . . .�xd4 1 2J':!:xd4;!; is a dubious choice without the inclusion of h4-hS and . . . gS, as Black's kingside is too vulnerable. 1 2.hS e6! Other moves would give White an ideal position . 8

1 3 .hxg6!? Again this is the most ambitious. 1 3 . tLl e3 gS 1 4.g3, intending �g2 and f1-f4, may well be better for White. However, if I was aiming for this type of position I would prefer I I .h S , since Black is forced to give up his dark-squared bishop there, whereas here it may offer him some counterplay. 1 3 . . . exdS 1 4.gxf7t 'it>xf7 I S .WlhSt 'it>e7 1 6.WlxdS 1 6. tLl fS t 'it>f8 followed by . . . WlgS would enable Black to escape to a drawish endgame. 1 6 . . . tLl b6 After 1 6 . . . tLl f6 1 7.Wlb3� the knight is in the way of Black's pieces, and White will play �c4 next, with excellent compensation. 1 7.Wlb3 The critical position. Black has many ways to continue, but his exposed king should constantly make him uncomfortable. There is no need to analyse all possible moves here, but I will j ust offer a few lines to make my point. 8 7 6 5 4 3 2

a

7 6 5 4 3 2 1 a

b

e

d

e

f

g

h

b

e

d

e

f

g

h

1 7 . . . Wlc7 1 7 . . . �xd4 Black's dark-squared bishop seems too strong to j ust give away. 1 8 .E:xd4 �e6 1 9 .Wla3;!; 1 7 . . . E:f8 is met by the brilliant 1 8 .eS! �xeS 1 9.E:e l and the position gets opened up rather unpleasantly for Black, for instance 1 9 . . . E:f6 ( 1 9 . . . �d7 20.E:xh6±) 20.f4 E:xf4 2 1 .E:xh6 Wlf8 22.E:h7t 'it>f6 23.tLlf3 with a

Chapter 2 crushing initiative. 1 8 .f4!? This seem's the most critical, setting up constant threats involving e4-e5. It's a complicated position, but Black's life seems harder. For instance: 1 8 . . .id7 Or 1 8 . . . 'I&c5 1 9 .e5! dxe5 20.'I&g3! i>f7 2 1 .l!h5! with a strong attack. 1 9 .1!h5!? l!hf8 20.e5 dxe5 2 1 . fxe5 l!ac8 22 .'I&a3t i>f7 23 .'I&g3! Intending id3 with good attacking chances.

-

7 . . . g6

33

will j ust crumple. Black may try to obstruct this plan with a timely . . . g4, but this will bring the dark-square weaknesses into focus. In the meantime, White can j ust continue developing normally, place his pieces on good squares, and see where it leads him. Keeping in mind these ideas, the next few moves will be easy to understand.

8 7 6 5 4 3

a

2

e

d

e

f

g

h

14JWe3! J.e6

1 a H

b

b

e

d

e

f

g

h

...g5

Other moves would obviously result in a hideous kingside structure.

1 4 . . . b5 is not terribly useful, since the bishop is better on e6 than on b7. 14 . . . 'I&a5 can be met by 1 5 .g3! , since taking on a2 will allow a check on c7. ( l 5 .i>b l is too slow: 1 5 . . . ie6 1 6.g3 l!c8 1 7. f4 'l&c5 ! 1 8 .c3 lLlc6=)

12.�d5 hd4 1 2 . . . ie5 is rather awkward, as Black will constantly have to worry about fl-f4 threats. 1 3 .g3! e6 1 4.lLle3 'l&c7 1 5 .ig2;!; intending l!hfl and fl-f4.

15.g3! Mter 1 5 . f3 'l&a5 1 6.i>b l 'l&c5 it will be hard to arrange g2-g3 and f3-f4 .

15 g4 . . .

13.gxd4 �e5 At first Black appears to have a solid position, and the absence of his dark-squared bishop isn't felt immediately. But without that bishop, there is also not much scope for his queenside counterplay. Also it is important to appreciate that if White manages to play g2-g3 and fl-f4, then Black's apparent solidity on the kingside

Black should not allow fl-f4 .

16.�bl gcs 17 .ie2 YlYa5 l S.c3 YlYc5 19.YlYf4� •

Black is pretty solid, but his position contains a number of weaknesses and he has no counterplay.

6 . . . lLl bd7

34

against any lLld5 ideas, while the . . . ixg4 threat spoils White's plans completely. I analysed this a bit further but did not get anywhere, so I began to think of other ways to implement the g2-g4 idea. Ideally, we want to meet . . . hxg4 with h4-h5 , while avoiding the problems caused by a timely . . . lLlb6.

C2) 9 ... h5 8 7 6 5

lO ...�c7

4

1 0 . . . b 5 ? 1 1 .lLlc6! Wc7 1 2 .lLlxe7 is a nasty trap.

3 2

1 O . . . lLlb6? doesn't work either due to 1 1 .e5±.

1 a

b

e

d

e

f

g

h

At first it appears that White can j ust build a quick central initiative with f2-f4, followed by a pawn advance to e5 or f5 , while if Black castles, the weaknesses caused by the . . . h5 move might be felt. Initially the computer seemed to endorse that view, but then Nakamura played this way with Black, and showed that the above assessment might be unfounded. One important point is that f2-f4 will give Black the excellent g4-square for his knight, while any lLld5 threats can be prevented by placing the other knight on b6.

1 o.b 1 The idea of connecting this move with the rather striking novelty on the next turn is not something I came up with directly. First I tried 1 O. f4N, with the idea of meeting 1 O .. .'IWc7 with: 1 1 .g4!? ( 1 1 .W b 1 transposes to 1 1 .f4 in the note to White's next move in the main line below.) The idea was to prevent Black's knights from using the g4-square, and I noticed that 1 1 . . . hxg4 would run into the strong 1 2.h5!. (In this version White's f2-f4 move proves far more useful than Wb 1 , as occurs in the main line.) 50 far so good, but then I realized that Black had a good answer in 1 1 . . .lLlb6!, guarding

10 ... 0-0?! is too committal, and after 1 1 .f3� White's kingside initiative will be swift. 8 7 6 5 4 3 2 1 a

b

e

d

e

f

g

h

1 1 .g4!?N

This seems like the ideal moment to go for the g2-g4 plan. I found no advantage for White after the more obvious attacking try: 1 l .f4 lLlb6! Again this knight move plays a key role in the defence. Now the critical move is obviously: 1 2.e5 1 2. f5 lLlg4! 1 3 .:gh3 Wc5 1 4.lLlb3 Wf2 1 5 .Wd3 lLle5 1 6.Wd4 Wxd4 1 7'.lLlxd4 lLlc6 was fine for Black in 50lak - Nakamura, Rhodes 20 1 3 .

35

Chapter 2 - 7 . . . g6 1 2 . . . dxe5 1 2 . . . 0-0?! , gives White too many options, one example being 1 3 .exf6 exf6 1 4. fS � . 1 3 .fxe5 ttJg4!? 13 ... ttJ fd5 is safer, and 1 4 . ttJxd5 ttJxd5 1 5 .e6 0-0 1 6.Wc4 fxe6 1 7.Wxc7 ttJxc7 1 8 .ixe7 E!:e8 seems playable for Black. The main line also works well though. 1 4.e6 0-0 The position looks dangerous for Black, but it is hard to maintain the pawn on e6. 8

12.gxh5 �xh5 After 1 2 . . . ig4 1 3 . f3 ixh5 1 4.ih3;!; Black's bishop is stuck out of play. 1 2 . . . E!:xh5 1 3 . ttJ fS ! gxf5 1 4.ixf6 ixf6 1 5 .Wxh5 ixc3 1 6.E!:g l ! White invades through the kingside. The changed landscape on the kingside seems to favour White. Kingside castling is out of the question for Black, and White can improve his position in anticipation of plans involving . . . 0-0-0.

7 6 5 4 3 2

a

b

e

d

e

f

g

h

1 5 .g3! ? 1 5 .exf7t E!:xf7oo 1 5 . . . �h8! 1 5 . . . ixd4 is met by 1 6.exf7t E!:xf7 1 7.E!:xd4;!;, but now Black really is threatening to take on d4. 1 6.ih3 1 6.exf7 e5! 1 7. ttJ b3 ifS seems fine for Black. 1 6 . . .ixd4 1 7.E!:xd4 ixe6 1 8 .ixg4 ixg4 1 9 .Wxe7 Wxe7 20.ixe7 E!:fe8 White has no more than a faint initiative in the endgame. Returning to the position after I I .g4!?N, Black may try C21) 1 l �b6 or C22) 1 l ... hxg4. •••

C2 1) 1 l . . ttlb6 .

This is much less effective than it was in the similar position after 1 O. f4N Wc7 I l .g4!? With the pawn back on f2 here, the prospect of a bishop arriving on g4 is far less annoying.

a

b

e

d

e

f

g

h

13.�d5!? White has more than one good way to continue. Another is: 1 3 .E!:gl id7 1 4.ie3! ? Anticipating long castling. 1 4 . . . 0-0-0 After 1 4 . . . E!:c8 1 5 . f4 Black's king is permanently stuck in the centre, and White will prepare to break through, either along the g-file or with e4-e 5 . 1 5 . . . ttJ c4 seems like the obvious try to get counterplay, but White can react powerfully with: 1 6.ttJd5 Wa5 1 7 .E!:g5! Wa4 1 8 .e5! e6 1 9 .ic 1 ! dxe5 20.fxe5 exd5 2 1 .e6 fxe6 22.ttJxe6 @f7 23 .ih3± 1 5 .E!:g5 ! ? This move prepares a possible rook lift t o a 5 ! Another useful point is:

6 . . . lD bd7

36

a

b

e

d

e

f

g

h

l s . . . ih6 1 6.liJf3 ! ? If Black takes on gS, the knight recapture will give White more than enough compensation. This could be an interesting position to explore further, but since the main line gives White an edge by simpler means, there is no need to delve any deeper into it here.

13 liJxd5 14.exd5 �f6 •••

1 4 . . . 0-0? i s too dangerous due t o l S .ig2 followed by if3 , destroying the blockade on the kingside.

a

b

e

d

e

f

g

h

18.tl�e6! fxe6 19.YlYxe6 �b4 1 9 . . . liJ f6 20.ixf6 ixf6 2 1 .W!xf6 leaves Black's king in danger.

20.YlYxe7t YlYxe7 2 I Jhe7t �ffi 22Jhb7 �c6 23J��xd6 �a5 24.i!b4± White is close to winning. He has three pawns for a piece, his rooks are absolutely dominating, and the black knight is close to being trapped.

e22) 1 l ... hxg4

14 . . . ieS l S .W!e3 �fB ( 1 S .. .f6 1 6.id3!±) 1 6.ie2 followed by f2-f4 also looks grim for Black.

15 .tg2 •

It is not easy for Black to develop further, as his king will not be safe on either side of the board.

15 .tg4 16.£'3 .t1t5 •••

The bishop may be misplaced here, but it seems like the best try to stabilize the kingside to prepare castling.

This allows a nice tactical sequence. 1 7 . . . 0-0 is no better due to the simple 1 B .W!xe7± . Objectively Black's best continuation may be 1 7 . . . 'it>fB , although 1 B .ih3;!; leaves White in control.

a

b

e

d

e

f

g

h

12.h5! The key idea. White is looking to create ' entry points for his knights.

12 i!xh5 •••

Chapter 2 - 7 . . . g6 Other moves can be brushed aside with relative ease. 1 2 . . . b5? 1 3 .ig2± creates threats on the long diagonal while defending the rook in preparation for hxg6. 1 2 . . . gxh5 is similar to the main line, but keeping another pair of rooks on the board only serves White's purpose. 1 3 . lLl f5 ifB 1 4.f4!? (White can also play 1 4.ig2 E:bB 1 5 .f4, by analogy with the main line.) 14 . . . gxf3 1 5 .'1Wxf3 lLle5 1 6.Wff4 ixf5 1 7.exf5 White has a useful initiative. In the position without the rooks on h I and hB, Black can get away with 1 7 . . . 0-0-0, but here it runs into I B .ixf6 exf6 1 9 .1Lld5 Wfc6 20.E:h3! and White wins. 12 . . . lLl b6 1 3 .h6! lLl gB ( l 3 . . . ifS 1 4.ixf6±)

37

1 3 . . . E:xh5 1 4.E:xh5 gxh5 1 5 . lLl f5 ifS 1 5 . . . ihB 1 6.lLlxe7t 1 6.ig2 Compared with the main line below, the inclusion of the moves . . . Wfa5 and Wfd2 is clearly in White's favour. It is harder for Black to arrange . . . e6, while White's queen move sets up attacking ideas involving ih6, or possibly even with E:e l . 8 7 6 5 4 3 2

a

8 7 6 5 4 3 2

a

b

e

d

e

f

g

h

1 4.a4! White intends to chase away the knight and plonk his own horse on d5. ( l 4.Wfd2 ie5 also looks pretty good for White, but the position is messier.) 1 4 . . . ixh6 ( l 4 . . . e6? 1 5 .hxg7! E:xh l 1 6. lLl db 5 ! +-) 1 5 .a5 lLld7 1 6.lLld5± 1 2 . . . Wfa5 1 3 .Wfd2!? Other moves can be analysed, but this is the simplest. 1 3 .h6 Wfxg5 1 4.hxg7 E:xh l 1 5 .lLld5 is a fun position to explore, but there is no need for it.

b

e

d

e

f

g

h

1 6 . . . E:bB 1 7.f4 1 7.E:e l ! ? can also be considered. 1 7.e5 does not quite work: 1 7 . . . Wfxe5 I B . lLl xe7 ixe7 1 9 .E:e l Wfc5 20.E:xe7t @xe7 2 l . lLl d 5 t @e6 White has many entertaining ways to give a perpetual, but the win proves elusive. 1 7 . . . gxf3 I B .ixf3 b 5 Otherwise it is hard t o see the point behind putting the queen on a5 and rook on bB. 1 9.Wff4 b4 20.lLld5±

13J:�xh5 gxh5 14.�5 The knight is maj estically placed here.

14 ....1£8 There is nothing else. After 1 4 . . . ihB ? 1 5 . lLl xe7! Black's position collapses, while 14 . . . @fS is not much better after 1 5 .Wfe3!± followed by ih6. We have reached a critical position . Retreating the bishop to fS looks ugly of course, but

6 . . liJ bd7

38

.

White needs to prove if his initiative is worth the flashy double pawn sacrifice. AI; you will see in the following lines, it will be hard for Black to untangle and complete his development, despite the existence of moves like . . . e6 which seem so obvious. And even if Black manages to complete his development, it should not be too hard for White to recover the loose kingside pawns, so we are not taking too big a risk.

C221) 15 e6 .••

8 7 6 5 4 3 2

8

1

7

a

6

b

e

d

e

f

g

h

This is the first logical move to consider. Retreating the knight doesn't seem too inspiring, but we can do better.

5 4 3

16.e5!

2 1 a

b

e

d

e

f

g

h

15 .ig2 •

This calm developing move prevents . . . bS and prepares fL.-f4. It also prepares a nice way to deal with . . . e6. If White could establish a pawn on f4 then Black would be busted, but he can always take en passant of course. For example: l S . f4 gxf3 1 6.�xf3 liJeS 1 7.�f4 .txfS 1 8 .exfS 0-0-0 1 9 . .txf6 exf6 20.liJdS �c6 2 1 .liJ xf6 �f3= I have included this line mainly to show that White is perfectly safe even when Black manages to complete his development, but of course we should make it as hard as possible for him to do this. Returning to the main line, Black has three main tries: e221) 15 ... e6, C222) 15 J�b8 and C223) 15 �c5!. ..

.••

l S ... liJ b6? is more or less refuted by 1 6.liJxd6t exd6 17 . .txf6± followed by liJ d S .

This is particularly effective after Black has committed to . . . e6, as there will be more open lines around his king. 1 6.f4!? is quite a sweet idea too, but it will most probably transpose to the later variation e223: 1 6 . . . gxf3 ( l 6 . . . exfS 1 7.exfS t .te7 1 8 .!!e 1 �d8 1 9 . .tdS! followed by �c4 is an amazing example of domination.) 1 7 . .txf3 �cS This position is covered later via the move order l S . . . �cS 1 6.f4 gxf3 1 7 . .txf3 e6. (Black had better avoid 1 7 . . . liJg4 1 8 .eS! liJ gxeS 19 . .txhS--+, or 17 ... exfS 1 8 .exfSt .te7 1 9 . .txhS--+.)

16 dxe5 17.liJg3 . . •

Now we're three pawns down! But all of our pieces are participating in the game, which certainly cannot be said about the opposition. Black's position is incredibly difficult, and it is hard to play even single moves, while White's tactical options abound. Here 'are a few illustrative lines.

Chapter 2

-

39

7 . . . g6

C222) 1 5 J�b8 .•

8 7 6 5 4 3 2 l �: Cb a

b

e

d

e

f

g

h

17 J�!b8 . •

17 . . . lLlh7 I B .i.h4 f5 1 9 .1Llxh5 �f7 20.f3!± and White opens up the kingside. 17 . . .i.e7 I B .Wd2! Ei:bB 1 9 .i.xf6! lLlxf6 20 .Wh6+- and White just wins the knight. 17 . . .Wb6 can be met by I B .�a l � , avoiding any tricks and asking Black how he intends to develop his pieces.

18.V!Vdl b5 19.Lf
a

b

e

d

e

f

g

h

This seems like a natural way to continue development. Now we have a pleasant choice between a forcing continuation and a more patient one.

16.f4

The more violent option is 1 6.e5!? lLlxe5 ( 1 6 . . . dxe5 1 7.We3 is more dangerous for Black) 1 7.lLlxd6t exd6 1 B .i.xf6 i.e6. White has excellent compensation here, but it wasn't easy to prove an advantage against the computer. The position can be explored further, but I tend to prefer the more measured build-up.

16 gxf3 •••

It is hard to imagine Black tolerating the pawn on f4. For instance, if 1 6 . . . e6 then 1 7.Ei:e l !?t leaves him facing too many tactical threats involving e4-e5.

17 .ixB b5 •

a

b

e

d

e

f

g

h

2o.Ac6t! Itld7 21 ..hd7t V!Vxd7 22.V!Vg5 V!Ve7 23.V!Vxe5 gb6 24.�ce4± White is absolutely dominating.

1 7 . . . e6 allows a familiar response: I B .e5! dxe5 1 9 . 1Ll g3 ± White has an even better version of variation C22 1 above. 1 7 . . . Wc5 is covered under the move order 1 5 . . . Wc5 1 6. f4 gxf3 1 7.i.xf3 Ei:bB - see the notes to Black's 1 7th move in variation C223 below.

6 . . . lLl bd7

40

18.eS!? dxeS I B . . . tLl xeS 1 9 .tLlxd6t exd6 2o.ixf6 ie6 2 l .ixeS dxeS 22.tLldS (22.idS!?�) 22 . . . ixdS 23 .hdS gives White a long-term initiative, which is only enhanced by the presence of opposite-coloured bishops.

19.�dS ttJxdS 2o.ixdS �f6 2 1 .ixf6 22.YfxhS .bf5 23.Yfxf5

8 7 6 5 4 3 2 1 b

e

d

e

f

g

1 6.ih4 e6 1 7J�e l ! ? is an ambitious try. White may even follow up by putting a second knight en prise on dS.

exf6

Compared with the last note, Black has a more compact structure with a pawn on f6 instead of h S . Nevertheless, White keeps a persistent initiative. A possible continuation is:

a

methods as well. I will mention a few of the more interesting possibilities that could be explored more deeply:

h

23 .. J3b6 24 ..bf7t! An important detail.

After 1 6.:!:�e 1 e6 I found nothing better than 1 7.W!d2 transposing to 1 6.Wd2 below. An extremely complicated position results from: 1 6.Wd2!? e6 1 6 .. ':!:�bB 1 7.id WaS I B .ih6� 1 7.i!e l How should Black continue his development? 1 7 . . . tLl gB! A most unnatural defence for a human, but probably necessary since it frees the d7knight and avoids being hit by e4-eS . 1 7 . . . exfS ? I B .exfS t rJidB 1 9.tLldS ig7 20.b4! wins. The logical 1 7 . . . i!bB runs into our typical, beautiful idea of I B .eS!! dxeS 1 9 .i!d l ! Wc7 20.tLlg3 when Black is completely tied down. 8 7

24... �xf7 2SJ::M7t Yfxd7 26.Yfxd7t;t Black will have to work hard to construct a reliable fortress in the endgame.

C223) I S ...YfcS! This seems like the best defensive move, cutting out the e4-eS tricks. The queen is better on cS than on as, since the future plan of .. ':!: � bB and . . . bS will not cut the queen off from the centre.

16.£4 This is the simplest way to continue, but there are a lot of other more sophisticated

6 5 4 3 2

a

b

e

d

e

f

g

h

I B .id Wc7 There are various options that deserve attention here. One possible continuation is: 1 9. f4 gxf3 1 9 . . . tLl gf6 20.id4t

Chapter 2 2o.ixf3 ltJe5 2 1 .Wg2!? ltJ f6 22.ltJg7t ixg7 23 .Wxg7 White has a strong initiative with pressure on the dark squares.

-

41

7 . . . g6

Wf4 2 1 .E:fl Wg5 22.if3 ltJe5 (22 . . . exf5 23.e5!±) 23.ie3 Wg6 24.E:g 1 Wh7 2 5 .ig5 ltJ fd7 26.ih5 The analysis can continue, but White clearly has the initiative. 8 7 6 5 4 3 2

a

a

b

e

d

e

f

g

h

16 ... gxf.3

1 6 . . J''! b B?! gives White a few good options, the simplest being: 1 7.ltJxd6t! exd6 1 B .e5 dxe5 1 9 .fxe5 ltJ h7 2o.if4±

17.i.xf3 e6 This seems like the most logical way, as Black has to try and get rid of the knight. After 17 . . . We5 1 B .ih4i the queen will easily be driven away, so Black's last move doesn't have much point. 17 . . . ltJe5 1 B .ixf6 exf6 1 9 .1tJ d5;!; is excellent for White. 1 7 .. .1''! bB This move prepares . . . b5-b4, but it gives White a lot of time. I like the idea of invading on the kingside with: 1 B .Wg2! 1 B .ixh5!? is also quite dangerous. 1 B . . . e6 1 9 .ie3 ( 1 9.ltJ d5 often runs into the following defence: 1 9 . . . exd5 20.exd5 t ltJ e 5 ! 2 1 .ixf6 ixf5 22.ixe5 Wxc2t 23 .Wxc2 ixc2t 24.i'xc2 dxe5=) 1 9 . . . We5 2o.id4

b

e

d

e

f

g

h

l B. . .b5 Black is not helped by 1 B . . . e6 1 9 .ie3 We5 20.ltJ g7t±. 1 9 .ie3 Wc7 1 9 . . . We5 2o.id4 We6 2 1 . ltJ e2! ib7 22. ltJ f4 Wc4 23.b3 Wc6 24.E:d3± 20. ltJ g7t ixg7 20 . . . i'dB 2 1 . ltJ xh5;!; 2 1 .Wxg7 Now White gets an excellent initiative on the dark squares: 8 7 6 5 4 3 2

a

b

e

d

e

f

g

h

2 1 . . .ltJe5 22.id4! ltJ xf3 Or 22 . . . b4 23 .ixe5 bxc3 24.b3 ie6 25 .id4 intending e4-e 5 . 23.WhBt i' d 7 24.ixf6 exf6 25 .Wxf6 ltJe5 26.Wxe5;!;

42

6 . . CLJ bd7 .

Let's return to the more critical 17 ... e6.

a

b

e

d

e

f

g

1 9 . . . .ie7 1 9 . . . b5 20.CLJc6± 19 . . . lLl g4 20 . .ixg4 hxg4 2 1 .�xg4 It is not easy for Black to complete development while keeping White's pieces from invading through the kingside.

h

1 8.i.e3! Other moves such as 1 8 .i.xh5 and 1 8 .1'%f1 lead to interesting play, but the text move is the most critical.

18 ...Y*le5 1 8 . . . �c7 1 9 .1Llg3 ( 1 9 . lLl h6!? is a dangerous alternative.) 1 9 . . . lLle5 ( 1 9 . . . h4 2o.lLlh5 lLl xh5 2 1 ..ixh5 lLle5 [2 1 . . . lLl f6 22 . .ig5 !t] 22.�f2;!;) 2o.lLlxh5 lLl xh 5 2 1 ..ixh5 .id7 22.�f2! Black still cannot castle due to .ib6, while 22 . . . .ic6 is met by 23.1'%fL t intending .ig5 , maintaining the pressure.

19.111 h 6!? Creating a lot of tactical tricks, so Black is practically forced to give up his crucial dark­ squared bishop. A few other moves are also promising to explore further: 1 9 . .id4 �f4 20.1'%f1 ! ? exf5 2 1 .exf5 t �d8 22.�c4 leads to some crazy complications; White may indeed have excellent compensation, but there's no need to sacrifice a piece. 1 9. 1Ll d4 This nondescript move also gives White decent compensation. Black has a lot of options, so I'll settle for giving a few sample lines.

a

b

e

d

e

f

h

g

20.1'%g l ! ? lLl g4 Or 20 . . . �a5 2 1 .1'%g5 �c7 22 . .ixh5 lLle5 23.lLlf3 and White keeps some initiative. 2 1 ..ixg4 hxg4 22.�xg4 This might look okay for Black, but he is still under some pressure. Particularly the black queen finds it hard to find a safe haven. 22 . . . lLl f6 22 . . . �h8 23 .�g2! leads to strong threats: 23 . . . .if6 24. lLl f5 ! exf5 25.lLld5 lLlc5 26.�g8t �xg8 27.1'%xg8t �d7 28.lLl xf6t �c6 29.exf5 b5 30.1'%f8± 8 7 6 5 4 3 2

a

b

e

d

23 .�g7! .ifS 23 . . . .id7 24.�h8t+-

e

f

g

.

h

Chapter 2 24.'
-

7 . . . g6

43

20 ..bh6 Now Black's dark squares are glaringly weak. A5 long as he doesn't escape with . . . 0-0-0, he will continue to suffer.

20 ... b5 Or 20 . . . '
a

e

d

f

e

g

h

5 4

19 ....bh6

After 1 9 . . . ie7 20.Eig l ! Black's queen is being shooed away: 20 . . . '
3 2 1 a

b

e

d

e

f

g

h

2 1 .Eigl! Most other moves allow Black to complete his queenside development and get castled. 2 1 .a3 ib7 22.'
7 6 5

21 . b4 ..

4

2 1 . . .ib7 22.Eig5 ±

3

22.lll d l gb8

2

22 . . . a5 ? 23.Eig5 '
b

e

d

e

f

g

h

24.e5! '
After 22 . . . lLl g4 23.ixg4 hxg4 24.'
44

6 . . ttJ bd7 .

The text move abandons all hope of castiing, leaving White with several ways to develop his initiative.

26 ... �eg4 26 . . . J.b7 27.E!:gBt!±

27 .txg4 �xg4 28.gh7!? Y;Yb5 29.�d2i •

We have gone a bit deeper than necessary, but the general picture is clear: White has enduring compensation and Black must defend with great care.

Conclusion

a

b

e

d

e

f

g

h

23.gg5

White can also go for 23.E!:g7! ? immediately, for instance: 23 . . . J.b7 24.lt::l £2 Wb5 2 5 .lt::l d3� and there is no easy way for Black to solve his problems.

23 Y;Yd4 24.gg7 �e5 •••

24 . . . J.b7 2 5 .J.e3! We5 26.E!:g5 traps the queen.

25 .ie3 Y;Yc4 26.Y;Yg2 •

8 7 6 5 4 3 2 1 a

b

e

d

e

f

g

h

7 . . . g6 is becoming increasingly popular, and so far Black's position has withstood all of White's attempts to break it. After the natural B.O-O-O, even the sidelines B . . . Wc7 and B . . . e5 forced me to find new and inventive ways to challenge the defence. Most of the time you can expect to face B . . . J.g7, after which 9.h4!? forces Black to make an important choice. 9 . . . h6 is conveniently met by 1 O.J.xf6!N, when Black cannot realistically recapture with the knight, and 1 O . . . J.xf6 leaves White with more than one promising continuation. 9 . . . h5 is more critical, and here long analysis convinced me that 1 0. m b 1 Wc7 1 1 .g4!?N causes significant problems to the defence. There are many ways for Black to lose quickly, but even with best play, I found no clear way for him to defuse the pressure.

8 7 6 5 4 3 2 a

b

e

d

e

f

g

7 h6 ...

Variation Index l.e4 cS 2.liJ a d6 3.d4 cxd4 4.liJxd4 liJ f6 S.liJ c3 a6 6 ..tgS liJ hd7 7.'i'e2! h6 8..th4 A) 8 e6?! B) 8 gS C) 8...hS!? 9.f4 C l ) 9 ...eSN C2) 9 ...gS 1 0.fxgS hxgS l 1 .hgS .th7 1 2.0-0-0 e6 l 3.g4! C2 l ) l 3 ....te7 C22) l 3 ...h4!?N

46 47 48 49 S3 S4 SS

.•.

•••

A) after 1 D . . . hS

Cll) after 1 3 . . . h4!?N

C 1) after 1 1 . . . %lfh6

8

8

8

7

7

7

6

6

6

5

5

5

4

4

4

3

3

3

2

2

2

a

b

c

d

e

1 1 .eS!N

f

g

h

a

b

c

d

e

f

Il.@h l ! ?N

g

h

a

b

c

d

e

1 4. lt\ d S !

f

g

h

h

46

6 . . llJ bd7 .

1 .e4 c5 vlLlf3 d6 3.d4 cx:d4 4.fthd4 c!Llf6 5.c!Llc3 a6 6 .tg5 c!Llhd7 7.�e2 h6 •

Now it seems that the best Black can hope for is to transpose to some other variation. For instance: 9 . . . g6 (9 . . . g5 1 O.i.g3 e6 transposes to variation B below, while 9 . . . b5 1 0.f4 gives Black an unfavourable version of variation C, where the queen has gone to c7 prematurely.) 1 O. f4 e5 1 1 . fXe5 dxe5 This position is covered in the next chapter under the move order 8 . . . g6 9.f4 e5 1 O . fXe5 dxe5 1 1 .0-0-0 Wfc7 see page 59.

A) 8 e6?! .••

a

b

e

d

e

f

g

h

Prodding the bishop has been Black's most popular choice in the position. Though it may seem like a small detail, it can change the character of many of the ensuing variations.

8 ..th4

On the negative side for White, he has lost the chance to meet . . . h6 with i.xf6, as exchanging in the present position would clearly be premature. Also, the bishop has been driven away from the centre; there will be no returning to e3 or d2 for a while. However, Black has paid a price in the slight weakening of the g6-point. In this position Black's most important option is 8 . . . g6, and we will deal with it separately in the next chapter. For the time being, however, it is important to check three alternatives: A) 8 e6?!, B) 8 g5 and C) 8 h5!? •••

•••

•••

8 . . . e5 9 . lD f5 g6 1 O. lD e3 has been covered via the 7 . . . e5 move order - see variation 02 of Chapter 1 . 8 . . . Wfc7 should be met by: 9 . 0-0-0 (9.f4 allows 9 . . . e5!, which is similar to 7.f4 Wfc7 8 .Wfe2 e5, which is considered to be doing fine for Black.)

This makes little sense, since Black will not be able to transpose to normal lines with 6 . . . e6 and 7 . . . lD bd7. Nevertheless, there have been several games with this move order, so it is useful to see where the problem lies.

9.£4 �c7

9 . . . b 5 ? 1 O .e5±

10.0-0-0 h5

a

I I .e5!N

b

e

d

e

f

g

h

An obvious improvement over several existing games, including one between strong grandmasters: 1 1 .f5 lD c5 1 2 .fXe6 fXe6 1 3 .e5 dxe5 1 4.i.g3 b4 was an unclear mess in Potkin - Laznicka, Havana 20 1 2 .

Chapter 3 - 7 h6

47

. . .

1 1 . .. dxe5 12.fxe5 fl:lxe5

9 ... e6

1 2 .. .'I¥fxe5 l 3 .Wff3! Wfb8 1 4.J.xf6 tt:J xf6 ( l 4 . . . gxf6 l S .tt:Jd5! exd5 1 6.l"!e l t+-) 1 5 .tt:J c6 J.b7 1 6.tt:Jd5! White has a serious initiative.

9 . . e5 restricts the bishop on g3, but creates too many weaknesses, and after 1 0 . tt:J f5 tt:J c5 1 1 . 0-0-0 J.xfS 1 2.exfS Wfc7 l 3 .h4 Black was under pressure in Balogh - Savchenko, Dubai 20 1 0 .

13.J.g3 fl:lfd7 14.ti)f3 f6 1 5.fl:lxe5 ti)xe5 16.Wfh5t Wff7

.

10.0-0-0 V!ic7 1 l .h4 gg8 12.hxg5 hxg5 13.V!ie3! This seems like the most effective continuation, but even the more routine 1 3 .f3 should offer White an edge.

13 ... ti)e5

a

b

e

d

e

f

g

h

17.J.e2!

The queen exchange will do nothing to dampen White's initiative.

l 3 . . . b5 1 4.J.xb 5 ! axb5 1 5 . tt:J dxb5 Wfa5 1 6.tt:Jxd6t J.xd6 1 7.J.xd6± gives White an excellent position. Black might succeed in exchanging the queens on b6, but such endgames with three connected passed pawns for a piece tend to favour White. 8 7 6

17 ...Wfxh5 18.i.xh5t ti)f7 19.J.f3 ga7 20.J.c6t
5

Black must lose an exchange, and most probably the game in the long run.

4

B) 8 ... g5 9.J.g3

2

3

1 8

a

7

14.ti)f3!

6

Refusing to give Black time to regroup and develop the queenside.

5

b

e

d

e

f

g

h

14 ... ti) fg4

4

After 1 4 . . . tt:J xf3 1 5 .Wfxf3 there is no way to hold the kingside together. 1 5 . . . J.e7 ( l 5 . . . tt:J d7? 1 6.l"!h7+-) 1 6.J.xd6±

3 2 1 a

b

e

d

e

f

g

h

14 . . . b5 is not much better due to 1 5 .tt:Jxe5

6 ttJ bd7

48

. . .

dxe5 1 6. f3 , when Black's king is stuck rather unpleasantly in the centre. He needs to worry about sacrifices on b5 (for instance, 1 6 . . . .ib7? is refuted by 1 7.ttJxb5!) , but even if he avoids these tricks, his king will have nowhere to go, and his position will remain divided in two parts.

Black's king is far from safe on the queenside, Sethuraman - Hamdani, Ho Chi Minh City 20 1 2 .

C) 8 ... b5!? 8 7

15.�d2 !J.e7 1 5 . . . ttJxf3 1 6.g:xf3 ttJe5 1 7. f4 g:xf4 1 8 . .ixf4 .id7 1 9 . .ie2± White has easy play, with .ie3 and f3-f4 coming soon. Black will evacuate his king to relative safety on the queenside, but he does not have many good squares for his pieces.

6 5 4 3 2

16.i.e2 .id7 1 6 . . . b5 is met by 1 7.ttJxe5 ttJ xe5 1 8 .:gh7 .ib7 1 9 . .ih5 :gfB 20 . .ixe5 dxe5 2 1 .Wfe2 when Black's king is stuck in the centre again. 8 7 6 5 4 3

1 a

b

e

d

e

f

g

h

This move is something of a recent trend, having been used by a few big names including Cheparinov, Wojtaszek and Khismatullin. There are of course some parallels with the 7 . . . b5 line that we encountered in variation D3 of Chapter 1 , but there are also some big differences. One of Black's main ideas here is to be able to meet f2-f4 with . . . g5, taking the game in a completely different direction.

9.f4

2 1 a

b

e

d

e

f

g

h

17.c!ihe5 c!tlxe5 1 8.!J.h5 1 8 .:gh7 is another way to target the f7 -pawn so that Black finds it hard to castle. If 1 8 . . . :gfB then 1 9 . .ixe5 dxe5 20 . .ih5± maintains the pressure. However, it is worth seeing a game where Black managed to castle but still faced problems.

1 8 .. J�g7 19.�e3 b5 20.a3 0-0-0 2 1 ..ie2! The last move prepares a3-a4, showing that

Unlike the 7 . . . b5 line, there is not much difference here between castling and playing f2-f4. The point is seen after: 9 . 0-0-0 .ib7 1 0. f4 :gc8 ? This was a problem in the 7 . . . b5 line, but here it does not work. 1 1 .e5 dxe5 1 2 . fxe5 :gxc3 1 3 .bxc3 ttJ e4 1 4 .e6 Black is dead lost, as capturing the e6-pawn would allow an instant mate. Nevertheless, I will stick with the familiar move order, as it seems like the best way to cut down Black's options. The two main choices are Cl) 9 ... e5N and C2) 9 ... g5.

Chapter 3 - 7 . . . h 6

49

9 . . . b 4 1 0.lLld5 ib7 1 1 .0-0-0! lLlxd5 1 2 .exd5 Wic7 1 3 .E:e l ( l 3 . lLl f5 ! ? lLl b6 1 4.Wie4 is also possible.) 1 3 . . . lLl b6 1 4. f5 ! This is particularly nasry for Black when he has already committed himself to . . . h6. 1 4 . . . lLlxd5 1 5 .Wid2� Black cannot finish his development.

like in the position without . . . h6 and ih4. ( l O . . . g6! ? I l . fXe5 lLlxe5 is also interesting, but the queen move is more convincing.) A possible continuation is 1 1 . 0-0-0 b4 1 2 .ixf6 bxc3 1 3 .ixg7 lLl c 5 ! and the tables have been completely turned.

Cl) 9 e5N

1 0.fXe5 dxe5 1 1 .lLl f5 gives Black a few options, the simplest being: 1 1 . . .Wia5 ! 1 2. 0-0-0 b4 1 3 .lLld5 (Or 1 3 .ixf6 bxc3 1 4.ixg7 lLl b6! 1 5 .ixfB cxb2t 1 6.'kt>b l lLl a4 1 7.E:d3 'tt> xfB and Black is doing well.) 13 . . . lLlxd5 1 4.E:xd5 Wixa2 1 5 .Wic4 Wixc4 1 6.ixc4 ib7 1 7. lLl d6t ixd6 I B .E:xd6 lLlc5 The endgame is double-edged, but Black seems to be doing fine.

...

IO �b7 ••.

Now 1 0 . . . Wia5 ? runs into I l . fXe5 dxe5 1 2.ixf6! lLl xf6 1 3 . lLl c6 and White wins.

a

b

e

d

e

f

g

h

The B . . . h6 line is usually played with the intention of following up with . . . g5 , which explains why the text move has not yet been tested. However, it is a rypical move for the Najdo rf, so it is only a matter of time before someone tries it here. It is interesting to compare the present variation with 7 . . . b5 B .Wie2 e5, as covered in line D32 of Chapter 1 on page 1 6. In that line, a subsequent ... h6 could be met by exchanging the bishop on f6, or even by leaving the bishop temporarily en prise while threatening something else. So could this be a better version for Black? No, because here we have the important option of omitting the pawn exchange on e5.

1O.0-0-0!

It is important to play this before moving the knight away. 1 O.lLlf5 is once again met by 1 O . . . Wia5 ! , j ust

1 0 . . . exd4 I l .e5 highlights the drawback of Black's early . . . h6 move. 8 7 6 5 4 3 2 a

b

e

d

e

f

g

h

With the white bishop on g5, he could play . . . ie7 and take back on f6 with the g-pawn, leaving White with two pieces en prise. With the bishop already on h4, the same idea is impossible, so Black must try something else, for instance: 1 1 . . . lLl xe5 ( l l . . .dxeS may be best, although 1 2. fXeS ie7 1 3 .exf6 lLl xf6 1 4.ixf6 gxf6 1 5 .Wie4 E:a7 1 6.E:xd4± is still unpleasant for Black.) 1 2. fXeS ig4 1 3 .Wie l dxe5 1 4.WixeSt Wie7

6 . tll bd7

50

. .

8

8

7

7

6

6

5

5

4

4

3

3

2

2 b

a

e

d

e

f

h

g

a

1 5 .i.xb 5 t ! axb 5 1 6.Wfxb 5 t i.d7 1 7.Wfb7 dxc3 1 8 .Wfxa8t Wfd8 1 9 J':!: he l t i.e7 20.Wfa3! cxb2t 2 1 .i>b 1 +- Black has no answer to the multitude of pins. 8

b

e

d

e

f

g

h

1 5 .e5! ( 1 5 .ttJxb7 Wfb6 traps the knight, leaving the position unclear.) 1 5 . . . 8 ( 1 5 . . . ttJxe5 1 6.ttJxb7 �e8 1 7.ttJ c5;!;) 1 6.Wfd2 Wfg5 1 7.ttJxbn I would expect White's two pieces to prevail over Black's rook. 1 1 . . . Wfc7 This move is too passive, as there will not be any threat of a . . . �xc3 sacrifice. 1 2.i>b 1 1 2.g3! ? is also interesting, but the king move is fine, and it is difficult to find a good way to continue with Black. Here are a few illustrative lines:

7 6 5 4 3 2

8

1 a

b

e

d

e

f

g

h

l 1 .tll f5 !

7 6 5

Once again, refusing to exchange on e5 reduces Black's options.

4

1 1 ...�b6

2

3

1 1 . . .Wfa5 ?! 1 2J'hd6! is unpleasant. 1 1 . . .exf4 is also rather risky: 1 2.�xd6! i.xd6 1 3 .ttJxd6t i>f8 1 4.i.xf6 Wfxf6 ( 1 4 . . . ttJ xf6 1 5 .ttJ xb7 Wfb6 1 6.e5±; 1 4 . . . gxf6 1 5 .Wfh5 ttJ e5 1 6.ttJxb7 Wfb6 1 7.Wfh4! Wfxb7 1 8 .Wfxf6 ttJ g6 1 9 . ttJ d5�)

a

b

e

d

e

f

g

h

1 2 . . . �c8 1 2 . . . 0-0-0?! 1 3 .fxe5 dxe5 1 4 .ttJd5± 12 ... b4 1 3 .ttJd5 ttJxd5 1 4.exd5 g6 1 5 .fxe5 dxe5 1 6.d6± 12 ... g6 1 3 .fxe5 dxe5 1 4.ttJe3;!; . 1 2 . . . exf4 1 3 .i.xf6! ttJxf6 1 4.ttJd5 ttJxd5

Chapter 3 - 7 . . . h6

SI

l S .exdSt 'kt>dB 1 6.g3! g6 1 7. lD d4 .ixdS ( 1 7 . . . fxg3 , 1 B . lD c6t .ixc6 1 9 .dxc6±) 1 B . .ig2t 1 3 .g3!? b4 1 3 . . . g6 1 4.lDe3 maintains the pressure. 8 7 6 5 4 3

b

a

2

e

d

f

e

g

h

12.�bl !?N a

b

e

d

e

f

g

h

1 4 . .ixf6! lDxf6 l S .lDdS lD xdS 1 6.exdS g6 1 7.lDe3 .ie7 1 7 . . . .ig7 1 B .f5t 8 7 6 5 4

1 2.�d2N E!:cB 1 3 .'kt>b 1 E!:xc3! ? could be interesting to analyse further, but it could be messy: 1 4.�xc3 lD xe4 l S .�e 1 g6 1 6. lD e3 fSoo 1 2. fxeS dxeS 1 3 .g4 E!:cB 1 4 . .ixf6!?N ( 1 4 . .ig2 g6 l S . lD e3 E!:xc3 1 6.bxc3 .ia3t 1 7.'kt>b 1 0-0 Black had enough compensation in Krzyzanowski - Kowalczyk, corr. 20 1 3 .) 1 4 ... �xf6! It is not clear if White has achieved much. This could be analysed in more detail, but I prefer the main line.

12 .. J:�c8

3 2

a

b

e

d

e

f

g

h

1 B . .ih3 E!:dB 1B . . . f5 leaves White with a choice of promising piece sacrifices on f5 . 1 9.f5t Let's now return to 1 1 . . .�b6. This is the most active set-up for Black, as he is aiming for . . . E!:cB and a possible sacrifice on c3 . The last move also transposes to an existing game, but only momentarily, as I would like to take the play in another direction.

1 2 . . . g6 is usually a weakness, and now White can transfer the knight towards d S . 1 3 . fxeS ( 1 3 . lD e3 exf4!=) 1 3 . . . dxeS 1 4. lD e3 .ie7 Now the most logical continuation is arguably: 8 7 6 5 4 3 2

a

b

e

d

e

f

g

h

l S . .ie l ! ? Preparing to put a knight on d S .

52

6 . . . lD bd7

1 5 . . . 0-0 1 6. CtJ cd5 ( 1 6.h4 !!ad8 1 7.h5 is also decent.) 1 6 . . . CtJ xd5 1 7.CtJxd5 .Ld5 1 8 .exd5t White intends g2-g4 and .ig2, and a later h4-h5 could be extremely annoying for Black.

17.g4! e4 18.g5! 1 8 . .ig2!? is also good, but I like the directness of the text move.

18 ... hxg5 19.hg5 �a5!?

Threatening . . . Wi'b4 or . . . .ixb2.

8 7

1 9 . . . Wi'c5 20.c3! Black's counterplay is blocked, and his kingside will soon be ripped apart: 20 . . . .ixc3 (20 . . . CtJe5 2 1 . f6 .ih8 22.h4+-) 2 1 .bxc3 Wi'xc3 22 .Wi'd2+-

6 5 4 3 2 1 a

b

e

d

e

f

g

h

13.c!Lld5

I also considered 1 3 . .ixf6 CtJxf6 1 4. fxe5 dxe5 1 5 . CtJ d 5 , but found the position after 1 5 . . . CtJ xd5 1 6.exd5 Wi'f6 1 7.g4 g6 to be unclear.

13 ... c!Llxd5 14.exd5 g6 1 4 . . . Wi'c7 doesn't change much: 1 5 .g4 g6 1 6. CtJ e3 .ig7 1 7.f5t

15.c!Lle3 .ig7 16.f5! 0-0 1 6 . . . g5 is met by 1 7 . .if2t followed by h2-h4.

a

b

e

d

e

f

g

h

20.�d2! A change of plan. 20.c3 allows 20 . . J'hc3 !, while 20.a3 .ixb2! 2 1 .'kt>xb2 !!c3 22.!!a 1 CtJe5 creates extreme complications.

20 ...�xd2

8

20 . . . b4? 2 1 .fxg6 fxg6 22 . .ih3 is crushing.

7

21 .gxd2

6

Black has no counterplay, while White's initiative continues unhindered in the endgame.

5 4 3

21 ...c!Lle5 2 1 . . . f6 22 . .if4±

2 1 a

b

e

d

e

f

g

h

22.f6! 22.!!g2 CtJ f3 23.f6 .ih8 24.h4t

53

Chapter 3 - 7 . . . h6

22....ih8 23 ..ie2 l:Ue8 24J�lfl Black's bisqop is buried in the corner, and he will continue to suffer for a long time.

C2) 9 ... g5

b

a

e

d

e

f

g

h

As I mentioned previously, this is the main idea behind the set-up with 7 . . . h6 and B . . . b 5 .

10.fxg5 hxg5 1 1 ..ixg5 Ab7 12.0-0-0 e6

It makes sense for Black to prevent any lZ'l d5 ideas. 1 2 . . . b4N is an attempt to force the play. 1 3 .lZ'ld5 lZ'lxd5 1 4.exd5 �a5 ( l 4 . . . lZ'le5 1 5 .lZ'l b3±) Black is making an unusual attempt at immediate counterplay, but he only has an active queen, which doesn't seem enough.

f6 ( l 6 . . . �xa2 1 7 . .txe5 dxe5 I B .�xe5 .th6t 1 9 .1Z'lxh6 E:xh6 20.b3;!;) 1 7 ..txe5 fxe5 I B .�c4 E:cB 1 9 .�b3 E:c5 20 . .tc4 White's pieces are temporarily on awkward squares, but if he manages to consolidate, he will simply be winning. 1 2 . . . E:cB 1 3 . lZ'l d 5 ! 1 3 .�e l E:xc3!+ 1 3 . . . lZ'l xd5 1 4.exd5 lZ'l b6N This seems like the most logical try. 1 4 . . . .txd5 1 5 .lZ'lxb 5 ! axb5 1 6.E:xd5 ± 1 4 . . . e5 was played in Babic - Ottesen, corr. 20 1 4 . White gave up a pawn with 1 5 .h4!? and eventually won, but it is also hard to trust Black's position after the simple 1 5 . lZ'l f5N. 1 5 .h4!? The open posltlon offers huge potential for White's initiative. Here are a few lines illustrating my point: 8 7 6 5 4 3 2

a

8

b

e

d

e

f

g

h

1 5 . . . �c7 1 5 . . . lZ'l xd5 1 6.'kt>b l �b6 1 7. lZ'l f5 e6 I B .E:h3 Black simply cannot develop. I B . . . 'kt>d7 1 9 . 1Z'l d4± Black is in trouble, with E:f3 coming next. 1 6.'kt>b l lZ'l xd5 1 6 . . . .txd5 1 7.�f2� 1 7.�f2! .tg7 I B . .td3 �c5 1 9 .c3� Followed by E:hfl .

7 6 5 4 3 2

a

b

e

d

e

f

g

h

1 5 .lZ'lf5! ( l 5.�b l ? .txd5+) 1 5 . . . lZ'le5 1 6.iJ4

13.g4!

6 . . . lLl bd7

54

14 . . . Vff a 5 1 5 .@b l :gxc3 1 6.j,d2 b4 1 7.bxc3 bxc3 I B . lLl b3 Vff e 5 1 9 .j,e3;!; 1 5 . bxc3 The exchange sacrifice is relatively ineffective here, as the e4-pawn has been protected. 1 5 . . . Vff c7 1 6.:ghfl j,e7 1 7.h4 Black is struggling to get compensation, for instance:

8 7 6 5 4 3 2

8

1

7 b

a

e

d

e

f

g

h

I discovered this move myself before finding a correspondence game, which will be referenced later. 1 3 .a3 has been played a few times, but it is too slow, and after 1 3 .. J'k8! 1 4 .'11g e l :gxc3 1 5 . bxc3 Vff c7 1 6.j,d3 Vff c 5� Black had an excellent position in Rasulov - Khismatullin, Baku 20 1 3 . 1 3 .g3!?N can b e played with similar intentions to the main line. The pawn might actually be better placed on g3 in the event that Black develops with . . . j,e7 and . . . lLl e5 ete., but the drawback occurs in the critical line involving a piece sacrifice: 1 3 . . . b4! 1 4.lLld5 exd5 1 5 .exd5t lLle5 1 6.@b l ! ? ( I 6.j,g2 j,h6!oo is the key difference; with the pawn on g4, White could j ust play j,xh6 and g4-g5, winning the piece back.) 1 6 . . . j,h6 1 7.j,h4 @f8 1 8 .j,g2 White has interesting compensation, but it is less convincing than in the main lines with the pawn on g4.

6 5 4 3 2 b

a

e

d

e

f

h

g

1 7 . . . Vffxc3 1 7 . . . d5 1 8 .j,xf6! lLlxf6 1 9 .95 lLlxe4 20.j,xe4 dxe4 2 1 .g6± I B . lLl b3 Black has no real threats, and will soon face a punishing counterattack. I B . . . Vff c7 1 9 .e5! j,xg2 20.Vffxg2 dxe5 2 1 .j,xf6 lLl xf6 22.Vff a Bt j,dB 23.:gxf6+-

C21) 13 i.e7 ...

After the text move Black has two main ideas: C21) 13 i.e7 and C22) 13 h4!?N. .•.

•••

I am not too worried by: 1 3 . . . :gc8N 1 4.j,g2! :gxc3 14 . . . b4 1 5 .lLld5! exd5 1 6.exd5 t j,e7 1 7. lLl f5 lLle5 1 B . lLl xe7±

a

b

e

d

e

f

g

h

Chapter 3 - 7 . . . h 6

14.J.g2 �e5 Centralizing the knight is a logical idea, and Black needed to do something about the threat of e4-e5 . 1 4 . . . b 4 1 5 .e5! J.xg2 1 5 . . . ltJxe5 1 6 . .txb7 bxc3 1 7 . .txaB cxb2t ( 1 7 ... �xaB 1 B .b3±) 1 B .'it>b 1 �xaB 19 . .txf6 .txf6 20J�hfl .te7 2 1 .g5 Black is clearly worse, and grabbing a pawn back with 2 1 . . . .txg5? leads to disaster after 22.ltJxe6! fXe6 23.l'!xd6.

55

compensation was unconvincing in Schon Dutra, corr. 20 1 3 . 1 5 . . . b 4 1 6. ltJ a4! �a5 1 7.b3 ltJ d3t 0 7 . . . l'!cB 1 B .i>b 1 ±) 1 B .cxd3 �xg5t 1 9 .i>b 1 ltJ d7 20.l'!c 1 �g7 2 1 . ltJ c6 gives White a big advantage.

16.i>bl gg8 1 6 . . . ltJ c4? is refuted by 1 7.e5!, while 1 6 . . . �b6 1 7.l'!hfl ltJ fd7 1 B ..txe7 i>xe7 1 9 .�d2± is also pretty depressing for Black.

17 ..if4 tLlfd7 18.ghfI Y!Va5 1 B . . . b4 1 9 . 1tJ a4 �a5 20.b3± 1 B ... l'!xc3 1 9. bxc3 �c7 20.ltJb3 �xc3 2 1 ..td2 �cB 22.ltJa5±

19.a3 �c4 1 9 . . . l'!xc3 ? is refuted by 20 . .td2!.

20.tLla2� a

b

e

d

e

f

g

h

1 6.�xg2 1 6.exf6!? forces a long variation where Black needs to play accurately, but ultimately it remains unclear: 1 6 . . . .txh 1 1 7. fxe7 �a5 ! 1 B . ltJ cb5 axb5 1 9 .1tJxb5 l'!a6 20.�c4 f6 2 1 .ltJc7t 'it>xe7 22 . .txf6t ltJxf6 23 .�xe6t 'it>dB 24.ltJxa6 ltJ e4!f± 16 . . . ltJxe5 1 7. ltJ c6 ltJxc6 1 B .�xc6t i>f8 19 . .txf6 .txf6 20.l'!xd6;!; Followed by ltJ e4, keeping the upper hand.

Despite his apparent actlVlty, Black has no real threats on the queenside. Meanwhile White remains a pawn up and can look to develop threats in the centre.

e22) 13 ... b4!?N

15.h3 Black is struggling to prove compensation here. White has no problems in the centre, and can create threats either along the f-file, or with various knight sacrifices.

15 ... gc8N This seems like a natural way to deviate from 1 5 . . . �a5 1 6. ltJ b3 �b6 1 7.a3, when Black's

a

b

e

d

e

f

g

h

This forcing move is critical, but it runs into a strong sacrifice.

S6

6 . . . ltJ bd7

19.9he1

14.�d5! exd5 After 1 4 . . . .ig7 I S .ltJxf6t .ixf6 1 6 . .ixf6 'lWxf6 1 7 . .ig2± White consolidates with an extra pawn.

1 5.exd5t �e5

I S . . . .ie7 1 6 . .ig2 0-0 is refuted by 1 7.:ghfl :geB I B .'lWf2 and the f6-knight will soon fall.

1 9 .:ghfl � i s also promising.

19 ....icS

1 9 . . . :gcB 20. 'IW d2 threatens :gxeS, and after 20 . . . .ifB 2 1 .hS :gxgS 22 .'lWxgS White has a clear advantage.

1 5 . . . 'lWe7 1 6.'lWd2! 'lWe5 ( l 6 . . . :ggB 1 7 . .ig2) 1 7 . .ig2 'lWxgS is almost a great way to simplify, but after I B .'lWxg5 .ih6 1 9 .'lWxh6 :gxh6 20.ltJfS ltJ xg4 2 1 .h3± Black loses material.

a

b

e

d

e

f

g

h

20.'lWd2 gxg5 21 .hxg5 �fxg4 22.tihe7 V!!xe7 23.ghl Material is approximately even, but White keeps an obvious initiative. a

b

e

d

e

f

g

h

16.i.g2 fie7 1 6 . . . .ih6? 1 7 . .ixh6 :gxh6 I B .g5 is no good for Black, and 1 6 . . . .ig7 1 7. ltJ fS :ggB I B .:ghfl .ihB 1 9 . .if4 also leaves him under heavy pressure.

17.�f5 ggS l S.h4! White has huge compensation, and Black hardly has any normal moves available. The specific details are not so important at this point, but I will present a few illustrative lines anyway.

lS ... gg6 I B . . . :gcB 1 9 .:ghfl :gg6 ( l 9 . . . :gxgS 20.hxgS ltJ gB 2 1 .g6! fxg6 22.ltJg7t 'it>d7 23.gS!±) 20.:gd4 (20. 'it>b l ! ?�) 20 ... aS 2 1 .:gdf4±

Conclusion The position after 7 . . . h6 B . .ih4 is a relatively fresh theoretical topic, and the existing games have barely scratched the surface. We saw that B . . . e6?! has certain disadvantages, while B . . . gS is playable but extremely committal. Most of the attention has been on B . . . bS!?, when 9.f4 reaches a position rich in possibilities. 9 . . . eS has not yet been tested in practice, but I am sure it is only a matter of time, and I found that White needs to work hard to maintain an edge. The most important line in the current chapter is 9 . . . gS, sacrificing a pawn in an attempt to wrest the initiative. However, after 1 0.fxgS hxgS 1 1 ..ixgS .ib7 1 2 .0-0-0 e6 1 3 .g4!, my analysis indicates that White has the upper hand, although he should be willing to sacrifice a knight on dS if the position demands it.

a

b

e

d

e

f

g

8 g6 ...

Variation Index l .e4 c5 2.tLl a d6 3.d4 cxd4 4.tLlxd4 tLl f6 5.tLl c3 a6 6 ..tg5 tLl bd7 7.�e2! h6 8..th4 g6 9.f4! e5 1 0.fxe5 dxe5 1 1 .0-0-0! A) 1 1 ....te7?! B) 1 1 ...�c7 12.tLl b3 b5 1 3.tLl d5! tLlxd5 14.exd5 B l ) 14....tb7!?N B2) 14....td6 1 5.�d2! B2 1 ) 1 5 ...f5N B22) 1 5 ....it>f8!? B22 1 ) 1 6 ..it> b l B222) 1 6.g4!N B23) 15 ...0-0!?N 1 6.�xh6 .tb7 17.i> b l �fe8 18 ..te2 �ac8 1 9.�c1 !! B23 1 ) 19 ...e4 B232) 1 9 ....tf8 B232 1 ) 20.d6!? B2322) 20..ta!? A) after l l . . .�e7?!

a

b

c

d

e

f

1 2 . tt:l h3!N

g

59 59 60 64 65 67 68 71 74 75 76 76 78

B22) after 1 5 . . . 1tif8!?

h

a

b

c

d

e

f

B222) 1 6.g4!N

g

B23) after IB . . J:!acB

h

a

b

e

d

e

1 9 .Wc l !!

f

g

h

h

6 . ttJ bd7

58

.

.

1 .e4 c5 2.ft:lf3 d6 3.d4 cxd4 4.t£!xd4 t£!f6 5.t£!c3 a6 6 ..ig5 t£! bd7 7.\We2! h6 8 ..ih4 g6

impression is that it does not pose any serious problems. Please note that 1 0 .0-0-0 ?! does not work due to 1 0 . . .exd4 1 1 .e5 dxc3 1 2 .exf6t ttJe5! 1 3 .fxe5 .ie6 when Black has the initiative.

8 7 6

10 ... dxe5

5 4 3 2 1 a

b

e

d

e

f

g

h

This is one of the most critical variations of the Naj dorf these days, having been used by several high-ranking players. Surprisingly, Black's idea is not . . . .ig7, but instead a primitive set-up with . . . e5 and . . . .ie7, having first prevented the knight from j umping to 5. It looks counterintuitive for Black to aim for . . . e5 in the 6 . .ig5 Naj dorf, but White has not played in the most energetic fashion with the early Vf1e2. So far Black's set-up has yielded considerable success, with a score of more than 60%, but I believe this is largely due to the fact that White players have often been caught unprepared as this line came up.

9.£4! The only critical move. Instead 9 . 0-0-0 e5 followed by . . ..ie7 leads to a stable position with good prospects for Black.

9 ... e5 Other moves don't make much sense in conj unction with . . . g6. Besides, Black has to do something about the threat of e4-e5 .

10.fxe5 1 0. ttJ f3 has been played a few times and could be explored in more depth, but my

1 O . . . ttJxe5 has been used successfully by Anna Ushenina in one game, but it seems to hand White a pleasant position after: 1 1 . 0-0-0N ( 1 1 . h3? was too slow, and after 1 1 . . . .ie7 1 2 . .if2 b5+ Black was more comfortable in Kosteniuk - Ushenina, Geneva 20 1 3 .) 1 1 . . . .ie7 ( 1 1 . . . .ig7 1 2.ttJ f3!±) 1 2 .ttJf3 Vf1a5 Trying to maintain the knight on e5. 8

7 6 5 4 3 2

a

b

e

d

e

f

g

h

1 3 J1:xd6! ( 1 3.'it>b l .ie6 looks okay for Black; 1 3 .ttJxe5!? dxe5 1 4 . .ixf6 .ixf6 1 5 .Vf1f3 .ig7 1 6 . .ic4 0-0 1 7.:B:hfU is a quieter route to an edge, pointed out by Roiz in ChessPublishing annotations which appeared late in 20 1 4.) 13 ... g5 ( 1 3 ... ttJxf3 1 4.Vf1xf3 .ig4 1 5 .Vf1d3±) 1 4.ttJxe5! .ixd6 1 5 .ttJ c4 .if4t 1 6.'kt>b l Vf1c7 1 7 . .if2! White has excellent compensation for the exchange. 1 7 . . . .ig4 1 8 .Vf1e l 0-0 1 9 . .id4 Vf1d8 20.Vf1f2� White has a fine position with a lot of promising plans.

1 1 .0-0-0! Another version of what will, now be a familiar sacrificial motif. It is trivial to see that capturing on d4 is a bad idea for Black, and

Chapter 4 so far all games have continued with either

A) 1 l ... .te7? ! or B) 1 1 . .'!Wc7.

,

-

8 . . . g6

59

1 5 . .th3 .tg7 1 6.\f1b U::

•

13.i.xf6 i.xf6 14.c;t>bl b5

A) 1 1 ...i.e7?!

1 4 . . . Wlc7 1 5 .lLld5 Wld6 1 6.h4 also favours White. 8 7 6 5 4 3 2

a

b

e

d

e

f

g

h

This was played by Vishy himself, but my guess is that it was due to a slip of memory. Not long after the game, one of his prominent seconds, Wojtaszek, preferred the more appropriate move that we will deal with shortly. So far nobody has followed in Anand's footsteps, and indeed after correct play Black runs into some problems in completing his development.

12,C!l�b3!N After 1 2.lLl f3 Wlc7 Black continued developing normally and obtained a good position in Hou Yifan - Anand, Wijk aan Zee 20 1 3 .

12 0-0

1 a

b

e

d

e

f

g

h

15.g3 V!!c7 16.c!l)d5 V!!d6 17.h4! A typical way to annoy the defender.

17 c!l) b6 18.h5� ••.

Black faces a difficult decision over what to do about the troublesome pawn.

B) 1 l V!!c7 •••

8 7 6 5

..•

1 2 . . . Wlc7 1 3 .Wlf3! is an important point. Now the misplacement of the bishop on e7 can be dearly seen after 13 . . . b5? 1 4 J'1:xd7!. 12 ... b5 can be met by 13 . .txf6!? .txf6 1 4.g3 when Black's queenside pieces are stuck rather uncomfortably, as he is unable to develop either the queen or the light-squared bishop. 1 4 . . . 0-0 ( l 4 . . .Wlc7 1 5 .lLld5 Wld6 1 6 . .th3 lLl b6 1 7 . .te6! is a nice touch, winning the queen.)

4 3 2 1 a

b

e

d

e

f

g

h

This is the more logical option, removing the queen from danger on the cl-file while maintaining several options for the bishop.

6 . . . lD bd7

60

IV�b3

1 2.ttJ d5N looks tempting, but there is no way to exploit the knight's position on d4. An important line is: 1 2 . . . ttJ xd5 1 3 .exd5 .id6 1 4.g4!? ( 1 4.Wfe3 Wfb6!? is an annoying pin.) This looks like an interesting practical try, with the idea of sacrificing the knight on f5 if Black castles. Unfortunately Black can get the better game with the help of a nice idea:

considered a conservative approach, followed by B2) 14 .. .id6, which has the idea of sacrificing the h6-pawn in certain positions. .

Bl) 14 ....tb7!?N

a b

a

e

d

e

f

g

h

1 4 . . . h5! ( 1 4 . . 0-0 1 5 . ttJ f5 ! is an entirely new set of complications.) 1 5 .gxh5 ( 1 5 .h3 hxg4 1 6. hxg4 g5!+) 1 5 . . J:%xh5 A complicated game lies ahead, but Black's king will be quite comfortable on fS , and most Najdorf players will surely be happy to play such a position. .

12 ... b5 13.lLld5!

This has only been played in a single game so far, but it is definitely the most critical approach. Most games have continued 1 3 .Wfe .ig7 ( 1 3 . . . .ie7?? is impossible due to 1 4 J:%xd7) followed by . . . 0-0, when Black's position is virtually impregnable. By opening up the centre and creating threats like d5-d6, we are aiming to exert instant pressure on the opponent.

13 ... lLlxd5 14.exd5 Despite the lack of practical testing, consider this an absolutely critical position for the future of this variation. We will start by examining Bl) 14 ... .ib7!?N, which can be

b

e

d

e

f

g

h

As we will see later, one of Black's problems after 1 4 . . . .id6 1 5 .Wfd2! is that . . . .ib7 becomes undesirable due to ttJ a5 ideas. Therefore it makes sense to consider Black's possible attempt to get his queenside development out of the way before turning to the kingside. Now White has various ways to continue, but the aim should be to prevent Black from smoothly catching up on kingside development and thus getting the best of both worlds.

15.'lWd2! This is an ideal square for the queen, since it works in all directions. Apart from the blunt d5-d6, the queen supports a possible ttJa5, while eyeing the h6-pawn and thus making it harder for Black to play ... .id6 and ... 0-0 . I explored a few other possibilities, and will j ust give a shortened version of my analysis: 1 5 . .ie 1 ! ? is interesting, but I feel reluctant to remove the bishop from a perfectly decent square on h4. My main line continues: 1 5 . . . b4! 1 6.�b l a5! (I found 1 6 .. J::k 8 and 1 6 . . . .id6 to be less convincing.)

Chapter 4 - 8 . . . g6

8 7 6 5 4 3 2

a

b

e

d

e

f

g

h

1 7.ltJd4 ic5 ! 1 B . ltJ c6 ( l B . ltJ b 5 'lWb6 leads to interesting complications.) Now 1 B . . . ixc6?! 1 9 .dxc6 'lWxc6 20.'lWb5 'lWxb5 2 1 .ixb5 0-0-0 22.c3� gives White the more pleasant endgame, but 1 B . . . 0-0! is more challenging. This position can be analysed in more depth, but I find the main line more convincing. 1 5 .g4!? This seems like a natural way to continue development, but it does not pose Black too many hard questions. 1 5 . . . id6 1 6.ig2 1 6.ltJd2!? ixd5 1 7.ig2 gcB 1 B .ltJ e4 ixe4 1 9 .ixe4 0-0 20.'itb 1 is interesting, but not particularly clear. 1 6 . . . gcB! 1 6 . . . 0-0 1 7.ltJd2! gfcB 1 B .ltJe4;!; shows the ideal way for White to coordinate his pieces. 1 7.�b 1 0-0 1 7 . . . b4 allows 1 B . ltJ d4 0-0 1 9 . 1tJ c6;!;.

61

1 B .c3 a5! The key move to get counterplay. Without it, White will j ust continue playing logical moves and eventually put pressure on Black's dodgy kingside. 1 9 . 1tJ d2 1 9 .'lWxb5 a4! 20.'lWxa4 (20. ltJ a5 gbB�) 20 . . . ga8 2 1 .'lWe4 gfbB followed by . . . ia6 and . . . ic4 offers Black excellent compensation. 1 9 . . . b4 20.c4 ia6 20 . . . b3!? leads to similar positions. 2 1 .gc l At first I thought White had successfully consolidated with a pleasant position, but unfortunately Black has some dazzling tactical resources.

a

b

e

d

e

f

g

h

2 1 . . . b3! 22.a3 22.axb3 a4! 23.bxa4 gbBt gives Black an excellent initiative. 22 . . . ixa3! 23.bxa3 'lWc5 24. ltJ e4 After 24.ltJxb3 'lWxa3 2 5 .'lWb2 'lWd6t White's extra piece does nothing to assist the defence, while Black can easily send reinforcements to the queenside. 24 . . .'lWxa3 2 5 .'lWb2 'lWb4 Without going too deep, it is clear that Black has good compensation.

15 gc8 ...

a

b

e

d

e

f

g

h

1 5 . . . b4? is met by 1 6.d6±, but the text move prevents both d5-d6 and ltJ a5 .

6 tLl bd7

62

. . .

8 7 6 5 4 3 2 1 b

a

e

d

f

e

g

h

so I will present a few lines. It is always satisfying to prove that the machine isn't all­ knowing. 1 B . . . e4 1 B . . . h5 intends . . . �h6 and perhaps . . . 'We3 . 1 9.�d3 �h6 0 9 . . . e4 20.l'!he l ! exd3 2 1 .cxd3 f5 22.�f6t �f7 23.�xg7 cj;lxg7 24.l'!e7t+-) 20.�g5 ! (20.'We 1 'We3=) 20 . . . �g7 2 1 .'We2 0-0 22.�e7 l'!feB 23.l'!hfl �fB 24.g4! �xe7 2 5 .gxh5 White has a strong attack. 1 9 .�e2! e3 1 9 . . . �e5 20.l'!hfl e3 2 1 .'Wd3 �xg2 22.l'!gl ± 20.'We 1

16.�bl

8

1 6. ttJ a 5 ? ? is impossible due to 16 ... �b4!.

7

1 6.c3 has the same idea of preventing any tricks based on the potential mate on c2, but it does not seem appropriate to weaken the queenside at this early stage.

6

16 ... i.d6

2

It seems essential to block the d-pawn. Let's see what may happen if Black refuses to do so: 16 . . . �g7?! 1 7.d6 'Wb6 17 . . . 'WbB 1 B .�e7t is similar. 8 7 6 5

5 4 3

b

a

e

d

e

f

h

g

20 . . . �xg2 20 . . . ttJe5 2 1 .'Wh4t 20 . . . �e5 2 1 .�f3! �xf3 22 .gxf3± 2 1 .l'!gl �b7 22.l'!g3 �e5 23 .�d3! �xg3 24.'Wxg3 Black's extra exchange is irrelevant, and White is completely dominating.

4 3

8

2

7 a

b

e

d

e

f

g

h

1 B .�e7!? The calm 1 B .�f2 'WdB 1 9 .�e3 also looks pleasant for White, but it is hard to resist planting the bishop in front of Black's king. Although the position looks disastrous for Black, the computer is not easily convinced,

6 5 4 3 2 1 a

b

e

d

e

f

g

h

Chapter 4

-

63

B . . . g6

17 ..id3!

18 . e4

By defending the c2-pawn, White frees his queen and thus prevents Black from castling. 1 7.c3 offers Black a lot of extra options such as 17 . . . lLl b6, or j ust 1 7 .. . 'it>f8 and . . . 'it>g7.

I considered a couple of other approaches for Black.

..

17 ... £5 I am always sceptical about this move - it seems to open up Black's position too much, and his set-up will become increasingly unstable as long as White can avoid the instant threats. Although I will admit that such provocative moves can sometimes work. . . 1 7 . . . b4? i s too slow. 1 9 .Wfe3 a4 2o.lLl d4+-) lLld4. 19 . . . e4 20.ixe4 22 .�de 1 Black is routed

1 B .�hfl fS ( l 8 . . . a5 1 9 .Wfe3! Threatening fxe4 2 1 .Wfxe4t lLle5 in the centre.

1 7 . . . WfB is safer, but after 1 8 .�hfl 'it>g7 1 9 .94± White has an ideal position. He can double his rooks on the f-file and continue to manoeuvre around, while Black's counterplay on the queenside looks at best to be a matter for the distant future.

1 B . . . lLl b6 is well met by 1 9 .94!, since Black does not have the defensive plan of . . . e4 followed by . . . lLl e5 blocking the e-file. 1 9 . . . e4 ( l 9 . . . lLl c4 20.Wfe2 g5 2 1 .if2 ixd5 22.gxfS ixh 1 23.�xh 1 � White has overwhelming compensation for the exchange.) 20.gxfS! exd3 2 1 .�he l t 'it>d7 22.fxg6t This looks too dangerous for Black. 1 8 . . . 'it>f7 is safer, but it does not equalize. 1 9 .94! f4 ( l 9 . . . e4?! is too risky: 20.gxfS ! g5 [20 ... exd3 2 1 .fxg6t 'it>g7 22.lLl d4+-l 2 1 .ixe4 gxh4 22.f6! White has a strong attack, with �hfl , if5 and lLl d4 on the way.) 20.g5 h5 2 1 .Wfc2± White has a stable edge. Black's position looks artificial, while White has excellent squares for his pieces - especially e4, which may be used by either the bishop or the knight.

a

b

e

d

e

f

g

h

19 ..ic2 Now there is a severe threat of lLl d4-e6. a

18.c3!

b

e

d

e

f

g

h

The attempt to charge right in with 1 8 .g4?! backfires after 18 ... e4 1 9.�he 1 lLle5, reaching an unclear position where Black's pawn advances seem to make sense.

19 �e5 ...

1 9 . . . Wf7 2o.lLld4 gives White a big advantage.

Mter 1 9 . . . g5 2o.if2 Black's pawn advances in front of his king are not impressive at all, and White is ready to crack open the position.

6 . . ttJ bd7

64

.

Finally, 1 9 . . . b4 is met by 20.cxb4 0-0 2 1 . tt'l d4! intending 2 1 . . . .ixd5 22.tt'lb5 and White wins material.

20.�d4 �c4 21.VBe1 O-O!?

26.J.b3! c2t 27.
2 1 . . .e3 22.g4! sees White's attack get underway in the usual fashion.

B2) 14 .td6 ••.

21 . . . .ie5 does not change much after 22.B:f1 intending g2-g4, when Black's position is resting on shaky foundations. 8 7 6 5 4 3

a

2 1 a

b

e

d

e

f

g

h

2V��e6 VBb6 23.�xf8 gxf8

Black has some compensation for the exchange but White can, at the very least, return the material to regain the initiative.

b

e

d

e

f

g

h

If Black is allowed to castle without repercussion, he should reach a comfortable position, as White is not in much of a position to generate threats on the kingside. With that in mind, and following on from the previous variation, White's next move should come as no surprise.

15.VBd2! The queen hits the h6-pawn, discouraging Black from castling, while also preparing tt'la5 in case the bishop goes to b7. Black has many possible moves, but there are three that deserve particular attention. B21) 15 f5N is ambitious but rather risky. Oddly enough, the two most troublesome continuations are diametrically opposed in their aims. The patient B22) 1 5
24.g4 b4 25.gxf5 bxc3 8 7

•.•

6 5 4

•••

3

•••

2 1 a

b

e

d

e

f

g

h

1 5 . . . .ib7 allows White to demonstrate one

Chapter 4 - 8 . . . g6 of the points of his last move: 1 6.tt:)a5! :B:c8 1 7.tt:)xb7 Wlxb7 I B .'tt> b a It is not smart for Black to gi�e up the light-squared bishop so cheaply. Moreover, it is still not easy for Black to castle, while White can j ust continue developing. 1 5 . . . a5?! is almost a tempting option, but it runs into tactical problems: 1 6.�xb5 a4 8 7 6 5 4 3

65

It seems a bit too early for Black to expand in the centre.

16.g4 The simple 1 6.'tt> b l ! ? is a good alternative. It is hard to give a definitive verdict as there are a lot of possibilities for Black, but in general White's play looks a lot easier thanks to Black's airy king. 1 6 . . . g5 The computer often recommends playing this move early, but it looks suspicious to me. 1 7 .�e 1 tt:) f6 I B .h4 g4 1 9 .h5 �d7 20.�h4 'tt> f7 2 1 .�d3 tt:) xh 5 22.:B:dfl tt:) f4 23 .:B:xf4 exf4 24. tt:) d4 WlcB 2 5 . :B:fl ! ?� White intends to take on f4 with a dangerous initiative, whether or not Black accepts the second exchange sacrifice.

16... £4

2 b

a

e

d

e

f

h

g

1 7.tt:)d4! exd4 I B .:B:he l t 'tt> f8 1 9 .Wlxd4 :B:h7 0 9 . . . :B:g8 20.�e7t! �xe7 2 1 .:B:xe7 Wld6 22.E:de l Wlxe7 23.:B:xe7 'tt> xe7 24.Wle4t 'tt> fB 25.d6+-) 20.�e7t 'tt> g8 2 1 .�xd6 Wlxd6 22.:B:eBt tt:) fB 23.c4!+- White's pawns are too strong, especially with the rook out of play on h7.

Another nice line is: 1 6 . . . tt:) b6!? 1 7.'tt> b l ! 0 7.gxfS ? �xfS is surprisingly unpleasant for White, as he isn't ready to deal with . . . tt:) c4, and the c2-pawn is weak; 1 7.tt:)a5!? f4 I B .h3 might give White the upper hand, but it isn't obvious.) 1 7 . . . tt:) c4 I B .�xc4 bxc4 1 9 . tt:) d4! 8 7 6 5

B21) 15 ... f5N

4 3

8

2

7 6

a

b

e

d

e

f

g

h

1 9 . . . c3 0 9 . . . exd4? 20.:B:he l t 'tt> f7 2 1 .Wlxd4 is crushing.) 20.Wle3 g5 (2o . . . Wlc5 ? 2 1 .gxf5 gxfS 22.Wlg3 ! +-; 2o . . . Wlb6 2 1 . b3;!;) 2 1 . tt:) e6 �xe6 22.dxe6 gxh4 23.gxfSt

5 4 3 2

17..ld3

1 a

b

e

d

e

f

g

h

1 7.g5 ! ? leads to less fancy lines, but also has the potential to be dangerous for Black, despite what the computer seems to think.

6 ltJ bd7

66

. . .

22.gh6! 0-0-0

8

After 22 . . . 'it>f7 23.ltJe6 ,ixe6 24.dxe6t dlxe6 2 5 .,ih7! gh8 26.Wfd3! gxh7 27.Wfh3t White will eventually catch the king.

7 6 5

Another nice line is 22 . . . ,ixd l 23 .gxf6 exd4 (or 23 . . . ,ig4 24.ltJ c6!t and Black is absolutely stuck) 24.Wfe l t 'it>d8 25 .Wfe6 ge8 26.'1Wxd6t Wfxd6 27.gxd6t 'it>e7 28.ge6t 'it>d7 29.'it>xd l gxe6 30.,if5 and White should win the endgame.

4 3 2 1 a

b

c

d

e

f

g

h

17 ... g5 1 8 ..in tLlf6 19.h4 gg8 20.hxg5 hxg5 21.tLld4! 2 1 .gh6 is less convincing due to 2 1 . . .ltJ xg4 22 .gh7 ,id7 23.,if5 ltJ f6 24.,ixd7t ltJxd7 2 5 . ltJ a5 0-0-0 26. ltJ c6 gh8! and Black is surviving. The ensuing lines may appear insane, but White's play is actually quite straightforward. The rest of the analysis j ust proves the natural and simple point that Black cannot get away with pushing all those pawns and allowing the position to be opened up with his king in such a position.

2 1 . ...ixg4 2 1 . . . exd4 22.,ixd4 ltJxd5 (22 . . . ,ie7 23 .gde I ! 'it>f8 24.gh6!+-) 23 .ghe l t ltJ e7 (23 . . . 'it>f8 24.Wfh2+-) 24.,ih7±

23.gxf6 exd4 24.ghl gdf8 8 7 6 5 4 3 2 1 a

b

e

d

e

f

g

h

25.gh7! gxf6 Or 25 . . . ,id7 26.gfu6!� intending 26 . . . g4 27.gxd7 Wfxd7 28 .Wfa5 +- when Black cannot handle the threats of Wfxa6t and gh7.

26.gxe7t .he7 27 ..hd4 Despite Black's two rooks and connected passed pawns, he is let down by his exposed king.

8 7 6

27 ... gh6 28.Wfg2! .ih3

5

28 . . . ,id7 29.d6 wins.

4

29.Wlhl! .id6 30.a4±

3

White will create threats on tHe queenside before Black can land a touchdown with his pawns.

2 1 a

b

e

d

e

f

g

h

Chapter 4

-

67

8 . . . g6

1 7.'kt>b l Neither 1 7 . .ig3 'kt>g7 1 B . .ixd6 Wxd6 1 9 . .ie2 :B:eB nor 1 7.Wd4 .ie5 I B . .ig3 .ixd4 1 9 ..ixc7 .ib6 gives White any advantage.

B22) 1 5 @f8!? •.•

8 7 6 5 4 3 2 1

��������==� a

b

e

d

e

f

g

This was played in the only game so far in which 1 3 .tDd5 occurred. Putting the king on g7 may seem like a primitive plan, but it leads to exceedingly interesting play. Black's plan takes a few moves to accomplish, but once his king gets settled on g7, he will have an appealing choice between starting a que� nside . avalanche with . . . a5-a4 and expandmg m the centre with .. .f5 . White should therefore act quickly.

a

b

e

d

e

f

g

h

1 7 . . . e3 1 B.Wd4 In the event of I B .We l 'kt>g7 1 9 .tD c6 (or 1 9 . .ie2 tDe5 20.:B:f1 .if5 2 1 . tD c6 tD xc6 22.dxc6 :B:heB 23.g4 .ie4) 1 9 . . . :B:eB 20.Wc3t 'kt>h7 Black has good prospects. I B . . . .ie5 1 9 ..ig3 1 9 .Wb4t .id6 leaves White with nothing better than 20 .Wd4 repeating. 1 9 . . . .ixd4 20 . .ixc7 .ib6

I focused on two main options. B221) 16.@b l was played in the existing game, and it is worth analysing it to gain a better understanding of the position. However, my investigation ultimately led me to conclude that B222) 16.g4!N is the strongest move. 1 6.tDa5!?N This is another idea along similar lines to the second main move, i.e. taking immediate action without the prophylactic king move. However, Black gets a good game with a direct reply. 1 6 . . . e4! 1 6 . . . d7g7 1 7.tDc6 f5 1 8 .g4 f4 1 9 . .ig2 is the kind of position White is aiming for.

a

b

e

d

e

f

g

2 1 . .ixb6 2 1 ..id6t 'kt>g7 22.tDc6 .ib7 23 . .ie2 :B:acB 24 . .iB :B:heB is at least equal for Black. 2 1 . . .tDxb6 22 . .ie2 'kt>g7 23 .:B:he l .if5 24 . .if3 :B:acB 2 5 .:B:e2 :B:heB The position remains about equal.

68 1 7 . . .1'!e8 1 7 . . . f5 ! ?N is also possible, when 1 8 .lLlc6 �e8 reaches the main line below.

B221) 16.�bl 8 7 6 5 4 3 2 1 a

c

d

e

f

g

Even though the king move is not objectively strongest, it is certainly not bad either, as it takes the king off the c-file while avoiding any tricks involving . . . e4 and . . . .if4 . I think you will fi n d it useful t o discover the problems I encountered in this line, in order to appreciate the advantages of delaying the king move. As the theory of this variation develops further, you might find it useful to know something about both directions of play.

16 ... �g7 Now White must make an important choice.

b

a

h

e

d

e

f

h

g

1 8 . lLl c6N This seems like a logical attempt to improve over 1 8 .g4 e4, when Black had a nice position in T. Kosintseva - Ushenina, Geneva 20 1 3 . 1 8 . . .fS 1 9.94 f4 2 0 . .ig2 After 20.gS hS 2 1 ..ig2 lLlb8! 22 .�he 1 lLlxc6 23.dxc6 �d8 Black is quite all right. 20 . . . lLlcS 20 ... gS 21 . .ixgS ! ? hxgS 22.h4!--+ leads to some insanely dangerous stuff for Black. 2 1 .gS .ifS ! 2 1 . . .hS 2 2 . .if2t 22.gxh6t mh7co It is hard to assess what is going on here.

17.g4!?N This direct move seems like the most dangerous for Black, even though a clear route to an advantage remains elusive. 1 7 . .id3 is slow, as Black starts his counterplay immediately with 1 7 . . . aS!. The one existing game continued: 1 7. lLl aS Blocking Black's queenside play and entering a complicated strategic fight. This seems like a logical plan, but it is hard to prove anything for White.

a

b

e

d

e

f

g

h

Chapter 4

17 a5! ...

1 7 . . . ib7? I B .ttJa5 would be unpleasant for ' Black. 1 7 . . . b4? wastes time pushing the less important queenside pawn, and after I B .id3 a5 1 9 .Wf2 a4 20.ttJd2 a3 2 1 . ttJ e4!± Black pays the price. 1 7 . . . ttJ b6 This is less energetic than the main move, but it is interesting enough to be worth exploring a bit more deeply. I B .h3 I B .Wg2 and I B .�gl are also possible. Having a piece on the g-file will be useful if Black ever plays . . . f5, but against other moves it is more convenient to have g4 defended by a pawn. I don't think White needs to be afraid of . . . f5 here anyway. I B . . . a5 I B . . . �eB 1 9 . ttJ a5;!; IB . . . ttJ c4 1 9 .ixc4 bxc4 20.ttJa5 c3 (2o . . . id7 2 1 .�hfl ±) 2 1 .Wxc3 Wxc3 22. bxc3;!; I B . . . f5 cannot be refuted directly, but as long as White avoids giving his opponent too much activity, he will have good chances to exploit the weaknesses caused by Black's last move. A good answer is 1 9 .Wg2, preparing for an eventual opening of the kingside, and after 1 9 . . . ttJ c4 2o .if2 White is on top.

a

b

e

d

e

f

g

-

69

8 . . . g6

the tempting 20.�hfl axb3 2 1 .if6t �gB 22.ixhB id7!, followed by . . . Wa7, plays into Black's plans. 1 9 . . . f5 20.gxf5 ixf5 2 1 .id3 �hfB 22.Wg2 White keeps the better chances, although the position is far from one-sided.

b

a

e

d

f

e

g

h

1 8 .ixb5 .

This brave move is the most critical. The position is extremely complicated and I haven't been able to make any final conclusions, but Black seems to be okay. Other moves also lead to exciting play: I B .Wf2 a4 1 9 . ttJ d2 a3 20.ttJe4 axb2 2 1 .ixb5 (2 1 .if6t �gB 22.ixb5 �xa2! 23.ttJxd6 Wa5 24.c4 Wa3 2 5 .Wg3 �al t 26.�c2 b l =Wt 27.�xb l �a2t 2B.�d l Wxd6 29 .ixhB Wb6! is likely to lead to a draw.)

h

1 9 .Wf2! 1 9 .ixb5 a4 is too complicated. For instance,

a

b

e

d

e

f

g

h

70

6 . . itJ bd7 .

2 1 . . J�f8!! A surprisingly cool defensive move. It is necessary to stabilize the kingside before returning to the mess on the queenside. (2 1 . . J'ha2? is refuted by 22.itJxd6.) 22 . .tf6t (22.!%hfl f6!00) 22 . . . 'kt>g8 23 .!%d3 itJxf6! 24.itJxf6t 'kt>g7°o 1 8 .Wfe3 Another way of preparing itJ d2-e4. 1 8 . . . a4 1 8 . . . g5 is the computer's main suggestion after both of the above queen moves, but it seems hugely committal, even though it might prove objectively okay for Black eventually: 1 9 . .tg3 itJ f6 20 . .txb 5 ! a4! The critical move. (20 . . . itJ xg4 2 1 .Wfe2 can be analysed further, but Black's position looks shaky; 20 . . . .txg4 2 1 ..tc6! .txd 1 22.!%xd 1 !%ab8 23.c4:i: White has good compensation, although the position remains unclear.) 2 1 . itJ d2 itJ xd5 22 .WfS .tb7 Again we have a complicated mess on the board. 1 9 . itJ d2 a3 20.itJe4 axb2 2 1 ..txb5 The position of the queen on e3 instead of f2 necessitates a different approach from Black. 8

24. itJ f5 t gxf5 2 5 .!%xd 1 f4+± 24 . . . Wfxd6 2 5 .Wfa3 If White could exchange queens he would have great chances, but Black does not have to comply. 2 5 . . . Wfb6! 26.c4 itJ f6 Black seems to be fine, although the position remains complicated. 8 7 6 5 4 3 2 1 a

b

e

d

e

f

g

h

1 8 a4 19.�c1 •••

19 . .tc6 axb3 20 . .txa8 Wfa7 2 1 . axb3 'Wxa8 is at least equal for Black.

19 J�b8! ••

7

1 9 . . . a3 ?! needlessly blocks the queenside, and after 20.b3 itJc5 2 1 .Wfg2 !%b8 22 ..tc6 White is better.

6 5 4 3 2

a

b

e

d

e

f

g

h

2 1 . . .!%xa2! After 2 1 . . .!%f8 22.!%d3! Black's attack seems hard to continue. The text move leads to an exciting forced line. 22.itJxd6! 22.'kt>xa2? Wfxc2 23.!%b 1 .tb7 wins for Black. 22 . . . !%a1 t 23.'kt>xb2 !%xd 1 24.!%xd 1

1 9 . . . itJc5 20.!%dfl ! intends to provoke . . . f5, when White intends gxf5 followed by sacrificing the exchange to open the g-file. If Black avoids this with 20 . . . g5, he must reckon on 2 1 ..txg5 ! ? itJ e4 22 . .tf6t itJxf6 23.g5 itJh5 24.gxh6t !%xh6 25 .Wfg5 t !%g6 26.Wfxh5 .tg4 27.Wfh4;t.

20.c4 20 . .txa4 itJc5 2 1 ..tc6 .txg4 is extremely complicated.

20 itJc5 21.Wfe2 .•.

Chapter 4 2 1 .E:dfl g5! works better for Black here, as the piece sacrifice isn't as clear: 22.Sixg5 ! ? (Mter 22.Sif2 Sixg4 23 .Sixc5 '.Wxc5 24.h4 f5 ! 25.hxg5 h5 Black is doing surprisingly well.) 22 ... ctJ e4 23 .'.We2 (23 .Sif6t?! ctJ xf6 24.g5 ctJxd5!) 23 ... ctJxg5 24.h4 ctJ h7 2 5 .g5 Again we are in some strange complications, which could be interesting to analyse further.

21 ...f5! 2 1 . . .Sid7 looks more normal, but after 22.E:hfl g5 23 .Sif2 Sixb 5 24.cxb5 ctJ d7 2 5 .h4 White is doing quite well.

-

71

8 . . . g6

8 7 6 5 4 3 2 1 a

b

c

d

e

f

g

h

In the previous variation, Black was j ust in time to generate counterplay with . . . �g7 followed by . . . a5-a4 . This made me wonder if White could benefit by omitting the precautionary i>b 1 . Another possible benefit is that a future . . . a3 might be met by b2-b3, when . . . '.Wc3 might not necessarily lead to mate. Of course the negative aspect is the possibility of . . . e4 and . . . Sif4, so White needs to take care over how he deploys his pieces.

16 ... �g7 22.a3 This is the relatively safe choice, whereas 22.ctJd3!? ctJxd3 23.E:xd3 a3°o is messier. 22.h3!? is a sensible move to maintain the tension. The position is hard to assess, especially with the computer throwing puzzling ideas such as . . . h 5 ! ? into the mix.

22 ... fxg4 23. �al i.f5 24J"�hf1 White has given up a pawn in order to stabilize his position, and he intends Sif2-e3 next. Once again, the "unclear" assessment seems about right.

B222) 16.g4!N

1 6 . . . a5 ! ? This i s a n attempt t o punish White immediately for not vacating the c l -square for the knight. However, White has a far more effective possibility to bring the knight to e4. 1 7.'.We3! Not only enabling the knight to go to d2, but also keeping the c3-square under control, so that . . . a3 can be met by b2-b3. 1 7 . . . g5! ? This somewhat ugly move i s the most challenging. After 17 . . . a4 1 8 .ctJ d2 a3 ( l 8 . . . Sib7 1 9 . ctJ e4 E:c8 2o.Sid3±) 1 9 . b3± it is not clear what Black is doing next. 1 7 . . . h5 1 8 .h3! does not change much, for instance: 1 8 . . . a4 1 9 . ctJ d2 a3 20.b3 Sib7 2 1 . ctJ e4±

6 . . . tLl bd7

72

Finally, in the event of 1 7 . . . .ib7 1 8 . .ixb 5 ! a 4 1 9 . 1tl d2 a 3 20.ltle4! axb2t 2 1 .tj;lb l ± White's pieces are ideally placed to prevent any attack. 8

Let's return to 1 6 . . . tj;lg7. Now, i t i s critical to think about how White may improve on moving the king to b 1 , which would j ust transpose to the previous line. 8

7

7

6 5

6

4

5

3

4

2

3 a

b

e

d

e

f

g

2

h

1 8 .�g3 ltl f6 1 9 . .txb 5 ! a4 After 1 9 . . . ltl xg4 20.'lWe2 Black's position seems too open. 1 9 . . . .ixg4 20.ltld2!? .ixd l 2 1 .E:xd l E:b8 (2 1 . . .ltlxd 5 ? 22.'lWe4) 22 ..ic6 gives White excellent compensation. 20.ltld2 ltl xd5 20 . . . ltl xg4 2 1 .'lWe2 gives White some initiative. 2 1 .'lWe4 .ib7 8 7 6 5 4 3 2

1 b

a

e

d

f

e

g

h

17.Yfg2!

By removing the queen from the f4-c 1 diagonal, White prepares to develop with .id3 and E:hfl , without being hit by . . . e4. He also clears the way for the knight to go to e4 .

17 ...i.b7 1 7 . . . a5 is well met by 1 8 .ltl d2!, as the knight will be ideally placed on e4. 1 8 . . . b4 (Or 1 8 . . . a4 1 9 . 1tl e4 and since . . . a3 can be met by b2-b3, I don't see how Black will generate counterplay. Actually, even 1 9 .�xb 5 ! ? a3 20.b3 is possible, since . . . 'lWc3 can be met by ltl c4, but why go for these forcing lines when we can j ust avoid Black's idea altogether?) 1 9 . 1tl e4 a4 8 7

a

b

e

d

e

f

g

h

22.ltlc4! ltl b6 22 . . . ltl f4 23 .'lWf5 .ixh l 24 . .ixf4 2 5 . E:xd6+23 .'lWxb7! ? 23 .'lWd3 i s also possible. 23 . . . 'lWxb7 24.ltlxd6 'lWe7 2 5 .E:hfl �

6

exf4

5 4 3 2

a

b

e

d

e

f

g

h

Chapter 4 2o.ib5! b3 (20 . . . a3 ? 2 1 .b3±) 2 1 .a3:t Another ide� is: 1 7 . . . lLlb6!? Hitting the d5-pawn. The drawback is that the knight loses touch with the fG-square. I B .id3 I B .lLla5!? is another valid possibility, and after I B . . . id7 1 9 .id3 White stands better, for instance: 1 9 . . . lLl c4 20.'
-

73

8 . . . g6

18.i.d3 a5 Once again, the knight should not leave f6 unattended: I B . . . lLl b6 1 9 .ie4! a5 20.E:hfl and Black is in trouble. I B . . . g5 ! ? is tricky, but White j ust needs a few accurate moves to secure a big positional advantage: 1 9 .ie l ( I 9 .if2? loses the d5pawn after 1 9 . . . lLl f6!+.) 1 9 . . . lLl f6 ( l 9 . . . a5 2o .ixb5 a4 2 l . lLl a5 ! ±) 20.ia5 ! '
19.�d2 g5 It seems to me that Black has to go for this. 1 9 . . . a4 is met by 20.lLle4! a3 (2o . . . ixd5 2 1 .ixb5±) 2 1 .b3 and once again Black's attack is completely stifled.

20 ..ig3 �f6 8 7

8

6

7

5

6

4 3

5

2

4 3 a

b

e

d

e

f

g

h

1 9 .E:hfl ! lLlxd5! 19 ... lLl c4 2o.if6t gives White excellent attacking chances, so Black may as well take a pawn and head for complications. 20 .ie4! lLle3 2 1 .if6t @h7 22.'
2 1 a

b

e

d

e

f

g

h

21.�e4! .bd5 2 1 . . .lLl xd5 ? 22.lLlxd6 '
22J�hf1 .L:e4 23 ..be4i The game goes on, but Black's serious weaknesses on the kingside should provide White with excellent long-term compensation.

6 ttJ bd7

74

. . .

I B .i.g3 intends h4-hS at the right time, but allows I B . . . ltJ f6! 1 9 .i.d3 ( l 9.h4 ltJhSoo) 1 9 . . . e4 20.i.xd6 Wfxd6 2 1 .i.e2 ltJxdS with an unclear position.

B23) 15 0-0!?N •••

8 7

I B .i.gS has the idea of keeping the queen on the h-file after . . . i.f8. The critical line is I B . . . �acB 1 9 .c3 i.fB 2o.Wfh4 Wfd6! 2 1 .i.d3 (2 1 .�d3 ? i.xdS 22.�h3 i.g7+) 2 1 . . . i.xdS when the position is complicated, but my analysis indicates that Black is doing quite well.

6 5 4 3 2

18 .. J:�ac8

1 a

b

e

d

e

f

g

h

Surprisingly, this pawn sacrifice is the computer's top choice, and it was the most troubling move for me to deal with. It turns out that Black can tolerate the queen on h6, and his quick development will compensate for the missing pawn.

16.Yfxh6 i.b7 17.i»b l White has to play this in order to avoid the threat of . . . �f(a)cB, and if c2-c3 then . . . b4. There was also a problem of . . . e4 and . . . i.f4 t winning the queen. I wondered if White could avoid this move with 1 7.i.gS ! ? , but found that after 1 7 . . . �fcB I B .c3 b4 1 9 .c4 as 20.i.e2 (20 .i.d3 i.fB 2 1 .Wfh4 i.xdS+) 20 . . . a4 2 1 .�hfl i.f8 22.'1Wh4 fS !+! Black is in time with his queenside counterplay.

17 .. J:Ue8 With the king safely on b l , 1 7 . . . �acB ? is refuted by I B .i.d3 ! . Instead Black must be ready to answer the bishop move with . . . e4.

1 8 ..te2 It is worth mentioning a couple of other tries, mainly to show some of Black's resources.

a

b

e

d

e

f

g

h

19.Yfc1!!

I feel particularly proud of this move, as I had thoroughly analysed this line twice before eventually noticing this idea! Retreating the queen looks passive, but one advantage is that Black will not be able to play . . . i.f8-g7 with gain of tempo. Another plus point is that White avoids weakening his queenside with c2-c3 . 1 9.Wfd2 ? is a worse version of the idea, as after 1 9 . . . ltJ b6!+ Black will gain a tempo against the queen. 1 9 .c3 is the obvious contmuatlon, but after 1 9 . . . i.f8 2o .Wfd2 Wfd6! Black's play is surprisingly easy: 2 1 .�hfl (2 1 .i.g3 ltJ b6 22 .h4

Chapter 4 �xdS=; 2 1 ..!gS tLl b6 22.h4 �xdS 23 .�c 1 �e6 24.hS b4=t) 2,l . . . .!g7 22.tLlaS .!a8 23.tLlc6 tLl b6 24.tLlb4 tLlc4 2S . .!xc4 �xc4 26 . .!gS as Black wins back the pawn and is certainly not worse. The queen retreat reaches another critical position, where Black's two main ideas are B23 1) 19 ... e4 and B232) 19 ...i.fB. 1 9 . . . tLl b6 makes less sense now, since . . . tLl c4 isn't a threat anymore. 20 . .!f6! tLlxdS (20 . . . tLl d7 is met by 2 1 ..!gSt followed by h4-h S . ) 2 1 .�xdS �e6 (2 1 . . ..!xdS 22.�h6+-) 22.�xd6 �xd6 23.�d l White's two pieces should outwork the black rook. 1 9 . . . b4 does nothing to address White's plans, and allows: 20 . .!g4! Now Black must worry about tLlaS. 20 . . . f5 2 1 ..!xfS ! gxf5 22 .�gSt Wh8 23.�h6t i>g8 24.�g6t i>h8

-

7S

8 . . . g6

With the bishop still on d6, Black is not ready to play 20 . . . f5 on account of 2 1 .�gS ! fxg4 (2 1 . . . tLl f8 22.'!xf5 ± ; 2 1 . . .�xc2t 22.i>a l fxg4 23.�xg6t i>h8 24.�xd6+-) 22.�xg6t i>h8 23.tLld4! with a winning attack.

21 .i.g5 Now it will be hard for Black to hold on to the e-pawn.

21 ...£5 21 . . . e2 22.�d2 �eS 23.�xe2 �xe2 24 . .!xe2 .!xdS 2S . .!f3 .!xf3 26.gxf3t White not only has an extra pawn, but also has chances for a kingside initiative with an eventual h4-hS .

22.i.f3 e2 23J:fd2!

After 23.�de l tLleS 24.�xe2 tLlxf3 2 S .�xe8t �xe8 26.gxf3 .!xdS the bishop pair might compensate for Black's open kingside.

23 �e5 24..be2 �c4

8

•.•

7 6

8

5

7

4

6

3

5

2

4 a

b

e

d

e

f

g

h

25.c4! bxc3 (2S . . . �xc4 26 . .!f6t tLlxf6 27.�xf6t Wg8 28 .�xd6±) 26.�h6t Wg8 27.�d3! c2t 28.Wc 1 e4 29.�g3t .!xg3 30 .�g6t i>f8 3 1 .hxg3+- Mate will soon follow.

B231) 19 ... e4 This looks active, but it allows White to play the same idea as in the last note.

20.J.g4! e3 20 . . . b4 2 1 .tLlaS !t

3 2 1 a

b

25.i.f3! .ie5

e

Or 2S . . . tLl xd2t compensation.

d

e

f

26.Wfxd2�

g

h

with

huge

26J:fd4;t White will be left with two pawns for the exchange, with excellent minor pieces and a safer king.

76

6 . . ltJ bd7 .

B2321) 20.d6!?

B232) 19 ...i.f8 8

8

7

7

6

6

5

5

4

4

3

3

2

2

1

1 a

b

e

d

e

f

g

h

The main aim of this move is to carry out the regrouping plan of . . . Wld6, . . . .ig7 and . . . ltJ b6. Should Black achieve this optimal set-up, his position will have a wealth of potential. White's choice on the next move is of key importance, and I would like to present two interesting options leading to different types of positions. B2321) 20.d6!? gives rise to a complicated middlegame where I prefer White's practical chances, whereas B2321) 20.i.f3!? eventually leads to an endgame (after a series of crazy tactics) where White obtains some winning chances with little risk. 2o.Wld2?! is similar to 1 9.c3, and Black deals with it in the same way: 2o . . . Wld6! (20 . . . .ih6 2 1 .Wlxh6! Wlxc2t 22.�al Wlxe2 23.ltJa5 .ia8 24. ltJ c6±; 20 . . . ltJb6 2 1 . d6 Wld7 22.Wle3 .ixg2 23 .Wlxb6 .ixh l 24Jhh l Wlxd6 2 5 .Wlxd6 .ixd6 26.c3;:!;) 2 1 .�hfl .ig7� Black reaches his ideal set-up, and he has great chances on the queenside with . . . ltJ b6 and . . . e4 coming up. 20 . .ig5 100ks like a tempting try to advance the h-pawn, but Black has a forcing solution with: 20 . . . ltJ b6! (20 . . ..ig7 2 1 .h4-t) 2 1 . h4 ltJ xd5 22.h5 ltJ f4! (22 . . . ltJ b4 23 . .if3!t) 23 . .ixf4 (23 .hxg6 .ie4!?) 23 . . . exf4 24 . .if3 g5°o White has no easy way to exploit the open kingside.

a

b

c

d

e

f

g

h

With this move White returns the extra pawn and takes his chances in a complex middlegame.

20 ...,hd6 21 .i.g4!

This is the idea behind returning the pawn. Now Black must decide what to do a.bout his awkwardly-placed pieces.

21. ..i.e4 2 1 . . . .ixg2 is too greedy, and after 22 .�he 1 Black's poor coordination is more of a factor than his extra pawn. 2 1 . . . f5 ? 22 . .ixf5 ! is an important motif. 22 . . . gxf5 23 .Wlg5 t �h7 24.Wlxf5t mg7 2 5 .�d3 Wlxc2t 26.�a l +2 1 . . . .ic6 22.Wld2 .ie7 avoids the loss of a piece, but after 23 . .ixe7 �xe7 24.Wlg5 �ce8 2 5 . h4! .ie4 26.h5!-t White has a strong attack, and the c2-pawn doesn't matter.

22.�d2 f5 Black shuts out the annoying bishop, but now his kingside has been permanently weakened. Mter 22 . . . .ie7 23 . .ixd7 .ixh4 24 . .ixe8 �xe8 2 5 .�e2;:!; Black's bishop pair offers some

Chapter 4 compensation for the exchange, but it is not enough to eq ualize.

-

77

8 . . . g6

6

White has the superior minor piece, and he j ust needs to get coordinated and find a good place for his queen. 29.g3 29 .El6dS ! ? is another idea. 29 . . . Elxd6 30.Elxd6 Elc6 3 1 .Eld l White keeps the better chances.

5

24.gxf3

4

I was tempted to stop analysing here. There are of course a lot of options, but the stark weaknesses in Black's kingside will remain a nagging problem even if White does not manage to conduct a swift attack. It is worth exploring a few more moves j ust to avoid any immediate counterplay from Black, and to show a few of White's ideas.

8 7

3 2 1 b

a

e

d

f

e

g

h

23.i.f3 J.xf3

23 . . . iaB is a loss of time, as after 24JWd l Ele6 2 S .Ele U Black does not seem to have anything better than taking on f3 anyway. 23 . . . ic6!? This move is slightly more sophisticated than the retreat to aB, but White keeps the more pleasant position all the same. 24.Elhd l 24.Wd l is another possible try. 24 . . . Ele6 2 S .Ele 1 tLlcS (After 2S . . . ixf3 26.gxf3:;l; the earlier retreat to c6 would be pointless.) 26.idB!? WbB 27.tLlxcS ixcS 2B .ixc6 Elcxc6 29 .igS Elcd6 30.g4 White maintains some pressure due to the open kingside. 24 . . . Ele6 2s .ixc6 Wxc6 26.tLlaS Wc7 27. tLl b7 Wxb7 2B.Elxd6 tLl fB 8

8 7 6 5 4 3 2 1 a

b

e

d

e

f

g

h

24 J�� e6! •.

This keeps both the g6-pawn and the d6bishop defended. Now Black can think about counterplay with . . . a s . 24 . . . i fB 2 S .Elg l Ele6 26.Wd l tLl b6 27.idB! ? W b B 2B .ixb6 Wxb6 29.h4:;l; Th e exchange of bishop for knight seems favourable for White.

7 6 5 4

24 . . . tLl b6 2 S . Elg2 Ele6 26.Elhgl 'kt>g7 (26 . . . Wh7 27.Wd2; 26 . . . 'kt>f7 27.Wh6--+) 27.if2 (27.igS ! ? is also possible, intending h4-h S . ) 27 . . . tLldS ( 2 7 . . . tLl c4 2B .h4--+) 2B .ig3 The last

3 2

a

b

e

d

e

f

g

h

78

6 . . ttJ bd7 .

move prevents . . . lD f4, and now the rooks may transfer to the d-file gaining a tempo. White has the more compact position and can shuffie his pieces around, creating threats across the board.

B2322) 20.i.f3!? 8 7 6

25J�el!? Threatening lD d4, and thus forcing Black to commit one of his pieces to defending the rook on e6. Even after he does so, White can always keep in mind the f3-f4 idea to win the d4-square in the future. 2 5 J'l:g l can be met by 2 5 . . . a5! 26.1:l:dg2 lD f8 and it is hard to deal with the . . . a4 threat.

5 4 3 2 1 a

b

e

d

e

f

g

h

Provoking the e-pawn into advancing. 8

20 ... e4

7

Black's choices were limited, as the bishop was ideally placed on f3 and White was threatening d5-d6.

6 5 4

In the event of 20 . . . \Wd6 2 1 ..ig3 .ig7 22.h4! Black is too slow to get counterplay, while the h4-h5 attack is already upon him.

3 2

2 1 .i.g4 f5

1 a

b

e

d

e

f

g

h

25 �f7 •••

After 2 5 ... 1:l:ce8 Black will not be able to exert much pressure on the queenside, so White can play 26.1:l:de2 before looking for ways to improve his other pieces.

26.1:l:de2 Both sides continue to have various options, but it seems to me that Black's exposed kingside means he has more to worry about.

This is more or less forced, as 2 1 . . . e3 22 .1:l:he 1 wins the e-pawn while 2 1 . . . .id6 runs into the unpleasant 22.lDa5!±. 8 7 6 5 4 3 2 1 a

b

e

d

e

f

g

h

Chapter 4

-

8 . . . g6

79

22.d6!

30 ..bg7 J.g5

Removing the excellent dark-squared bishop from the def�nce. It is interesting to note the point behind provoking . . . e4 with the earlier if3 in this line. If the pawn was back on eS, then Black would have the devastating reply . . . 119c4!, winning a piece.

Although we are a long way into the game, Black's last ten moves have been mostly forced. Now White has a couple of good options.

22 ...hd6 22 . . . 119b6 23 .119gS ! fxg4 24.119xg6t ig7 2S .119xg4 ltJeS 26.119g3 ltJ d7 27.igS ! ? gives White terrific compensation.

8 7 6 5 4 3

23.L5 J.f4! 23 . . . gxfS ? 24.119gs t �h7 gives White more than one way to win, with 2 S .119hSt �g7 26.ltJ d4+- perhaps being the simplest.

2 1 a

b

e

d

e

f

g

h

3 1 .J.d4!?

The most ambitious.

a

b

e

d

e

f

g

h

24.gxd7! 24.ig3 ixc 1 2s .ixc7 gxfS 26.gxd7 ie3� reaches a different type of endgame where Black's bishop pair seems more impressive.

24 ...J.xc1 25J�xc7 gxc7 26.J.xg6 ge6 There follows a strange dance involving the rooks and bishops before the position settles into a normal type of endgame.

27.J.g3 gg7 28.J.f5 gf6 29.J.e5 gxf5

3 1 .ic3 is a decent alternative, leading to a slightly favourable endgame. 3 1 . . . e3 32.h4 ih6 33 .g4!? ixh l (33 . . .!:!:f2 34.!:!:e l ±) 34.gxfS ie4 3 S . ltJ d4 (3 S . ltJ cS ixfS 36.lLlxa6 � f7 i s OK) 3 S . . . ig7 36.lLle2 Getting rid o f the bishop pair. 36 . . . ixc3 (36 . . . ih6 37.ieS ixfS 38.if4;!;) 37.lLlxc3 The king will watch the e-pawn, and the knight can start hunting the queenside pawns. Black can probably hold the endgame, but there is still some work to do.

3 1 . .. e3 Now it is a bit harder to stop the pawn, but White can keep the advantage with precise play.

32J�e1 e2 33 ..ic3! gd5! 33 .. .!W 34. lLl d4 ixg2 3 S . lLl xe2± should be close to winning.

6 . . . lLl bd7

80

Conclusion

a

b

e

d

f

e

g

h

34.a4! 34.a3 E:d l t 3 5 . <;t>a2 E:xe l 36.J.xe l J.xg2 37. lLl d4 J.fl reaches a strange endgame, where the e2-pawn makes the situation slightly double-edged.

34... E:dlt 34 . . . bxa4 3 5 . lLl d4±

35. 'it>a2 �he1 36 ..be1 bxa4 36 . . . J.xg2 37.axb5 axb5 3 8 . lLl d4 J.fl 39.lLlxb 5 gives White a serious advantage. 8 7 6 5 4 3 2 1 a

b

e

d

e

f

g

h

37.tC!d4 .bg2 38.tlhe2;t It is not clear if White can convert the extra pawn, but it is a pleasant endgame with two possible results.

In this, our final chapter on the 6 . . . lLlbd7 complex, we have dealt with the topical position occurring after 7.V!le2 h6 8 .J.h4 g6. I advocate the direct 9.f4!, intending to meet 9 . . . e5 with 1 O. fxe5 dxe5 1 1 .D-D-0!, exploiting the fact that the d4-knight is poisoned. Several games between strong players have continued 1 1 . . .V!lc7 1 2 . lLl b3 b5, but at this point I believe White should deviate from the majority of games with the dynamic 1 3 .lLld5! lLl xd5 1 4.exd5. The resulting positions abound in tactical and strategic complexity. It may seem that I went to the extreme in terms of the depth of my analysis of a variation that has hardly ever occurred in a practical game, but I am certain that this line will become more popular in the future. If I am right, then readers of this book will be ready with a full arsenal of ideas.

6 ... e6 7.£4 a

b

e

d

e

f

g

Introduction and Sidelines Variation Index l .e4 c5 2.� f3 d6 3.d4 cxd4 4.�xd4 � f6 5.� c3 a6 6.J.g5 e6 7.f4 A) 7...J.d7 B) 7...� c6? 8.�xc6 bxc6 9.e5 h6 1 0.J.h4 g5 1 1 .fxg5 � d5 1 2.� e4 �b6 1 3.c3! B 1 ) 13 ..J;b8 B2) 13 ...�xb2 B3) 13 ...dxe5 C) 7...J.e7 8.�f3 C l ) 8...�a5 C2) 8 �c7

82 84 85 86 88 90 90 91

•••

B2) afrer l 7 .';!;> f7

A) afrer l2 .. .'�a5

C l ) afrer l6 ... gxf6

. .

8 7 6 5 4 3 2 a

b

e

d

e

1 3 .e5!N

f

g

h

a

b

e

d

e

f

l 8 .�xg6t!

g

h

a

b

c

d

e

f

1 7.'�e4!N

g

h

h

82

6 . . . e6 7.f4

1 .e4 c5 2.ttla d6 3.d4 cxd4 4.lihd4 �f6 5.�c3 a6 6 ..ig5 e6 7.£4

significance. For instance, the 'deferred Poisoned Pawn' with 8 . . . Vfib6 will simply be met by 9 .Vfid2 Vfixb2 l O.i::l b l Vfia3 1 1 .eS, transposing to our main lines.

A) 7 ....id7 This was tested by Bent Larsen a few times in the 1 9 S0s, and is still used now and again by strong players. Black is preparing . . . lLl c6, which may lead to some sort of Najdorfl Rauzer hybrid.

8.�a a

b

e

d

e

f

g

h

From this extremely popular position, Black's main options can be summarized as follows: 7 . . . Vfic7 and 7 . . . bS are important moves which receive dedicated coverage in Chapters 6 and 7 respectively. 7 . . . lLl bd7 is a major system which you can find in Chapters 8 and 9. 7 . . . i.e7 may give rise to three major set-ups: The Gothenburg Variation (Chapter l O) , and the . . . Vfic7/ . . . lLl bd7 set-up, which I call the "Three-Piece" system. The latter can be implemented either with an early . . . h6 (Chapter 1 1 ) or without (Chapter 1 2) . Finally, 7 . . . Vfic7 introduces the ever-critical Poisoned Pawn variation, which is covered in Chapters 1 3- 1 8 . In this chapter we will deal with three topics: the sideline A) 7 ... .id7, the formerly-interesting­ but-now-almost-refuted B) 7 ttlc6?, and finally some miscellaneous variations after C) .••

7....ie7. 7 . . . h6 8 .i.h4 does not have much independent

8.eS dxeS 9 . fXeS VfiaS! is one idea behind Black's set-up. White cannot transpose to a Rauzer with 8 .Vfid2 on account of 8 . . . h6, intending 9 .i.h4? lLlxe4!. 8 .Vfie2 also does not impress me much, since after 8 . . . lLl c6 9 . 0-0-0 the position resembles the more popular 7 . . . Vfic7 8.Vfie2 lLlc6 9 .0-0-0 line, but Black may get some additional options as he has not yet committed his queen to c7. Fortunately, there is nothing wrong with the standard development of the queen to f3 .

8 ... �c6 9.0-0-0 �xd4 9 . . . h6 l O.i.h4 does not change much. 9 . . . Vfib6 can be met by 1 0.lLlxc6 i.xc6 1 1 . fS!?N when Black has problems with the e6-pawn, for instance 1 1 . . .VfiaS ( l l . . .i.e7?! 1 2.fXe6 fXe6 1 3.Vfih3!± is strong) 1 2.i.xf6 gxf6 1 3 .i.c4 VfieS 1 4 .:1l:hfl t and Black is under pressure. 9 . . . Vfic7 1 0 .lLlxc6! ? gives Black a difficult choice. 1 0 . . . bxc6 ( l 0 . . . Vfixc6 is met by 1 1 .i.e2t intending e4-eS ; 1 0 . . .i.xc6 misplaces the bishop, and 1 1 .i.xf6 gxf6 1 2 .f5t followed by i.c4 and 1'l:hfl is promising for White.)

Chapter 5 - I ntroduction and Sidelines Recapturing with the b-pawn is positionally desirable for Black, but his queen and bishop are awkwardly placed on c7 and d7 respectively. 1 1 .�c4 �e7 1 2 .Elhe l e5

83

In the event of 1 2 . . . �c7 1 3 .Elhd 1 h6 ( 1 3 . . . 0-0?! loses an exchange after 1 4.e5 ttJ d 5 1 5 .ttJxd5 �xd5 1 6.�h6±) 1 4.�xf6 �xf6 1 5 .Elxd6 0-0 1 6.e5 White has an extra pawn for hardly any compensation. This position was reached in H . Albert Lombardy, Mar del Plata 1 9 57, and one other game. Here White could have pressed ahead with: 8 7

b

a

e

d

e

f

6

h

g

5

This position occurred in Stan - Nemeth, Bucharest 2003, and now 1 3 . h:e5N dxe5 1 4.�g3;!:: would have been strong.

4

10J:hd4

2

3

1 8

a

7

b

e

d

e

f

g

h

13.eS!N dxeS 14.fxeS .tcS

5

1 4 . . . ttJ d 5 ? ! 1 5 .�xe7 ttJ xe7 1 6.�xg7 ElgB 1 7.�f6±

4

l S .txf6!?

6

•

This exchange sacrifice is not forced, but it is absolutely j ustified.

3 2

l S ...,bd4 a

b

e

d

e

f

g

h

10 ....tc6

This is the logical way to follow on from the earlier 7 . . . �d7.

1 5 . . . gxf6 1 6.Eld3t

16J&xg7 :!:US 17..thS Wfb4 1 7 . . . �xc3 ? is refuted by 1 B .�xf7t! Elxf7 1 9 .�gBt.

1 1 ..te2 .te7 12.'1&g3!?

lSJ:�dl .txc3 19.bxc3 WfcS 20.�d6

12 .. .'IWaS

White maintains extreme pressure. The following illustrative line shows that even a queen exchange does not solve Black's problems.

Renewing the threat of e4-e5, while making it hard for Black to castle.

84

6 . . . e6 7.f4

20 YlYe3t 21 .c;!}dl YlYgI t 22.c;!}e2 YlYxg2t 23.YlYxg2 ,bg2

know a few precise lines to be sure of a big advantage.

White still has excellent compensation, for instance:

8.�xc6 bxc6 9.e5 h6

.•.

9 . dxe5 1 0JWxd8t mxd8 1 1 .fxe5 h6 l 2 .0-0-ot Wc7 l 3 . .ixf6 gxf6 1 4.exf6 leaves White with an extra pawn as well as the better structure. .

.

IO .1h4 g5 •

1 0 . . . dxe5 leads to typically worse positions for Black after 1 1 .�xd8t Wxd8 l 2.fxe5 g5 l 3 . 0-0-ot Wc7 1 4 . .ig3t.

1 1 .fxg5 �d5 12.c!l:)e4 YlYb6 a

b

e

d

e

f

g

h

24.c4! Why hurry to take on e6? This pawn can be captured at any moment, as defending it only leads to greater problems for Black:

l 2 . . . dxe5 is not really challenging. 1 3 . lt'I f6t It'I xf6 1 4 .�xd8t Wxd8 1 5 .gxf6 gives White an extremely pleasant endgame due to all those pawn weaknesses in Black's camp. After the text move we reach the starting position of most of the theoretical debate in this variation.

24 .1h3?! 25 ..16! gb8 26.,bb7± . . •

B) 7 Jt:lc6? ••

13.c3! a

b

e

d

e

f

g

h

This line experienced something of a resurgence in roughly 2006/07, but nowadays it can be regarded as having been buried deep in the ground. Nevertheless, White needs to

A lot of other moves were tried, rather unsuccessfully, before the strength of this move became apparent. White prevents the annoying check on b4 while maintaining various attacking ideas. Black may respond with BI) 13 gb8, B2) 13 YlYxb2 or B3) 13 dxe5. . • .

• • .

• . .

Chapter S - I ntroduction and Sidelines 13 . . . hxg5? is weaker: 1 4.ixg5 !!gBN ( 1 4 . . . dxe5 1 5 .Wxd5! ga:ve White a quick win in Fendrik - Forgacs, Hungary 20 1 0; 14 . . . Wxb2 1 5 .!!b 1 Wa3 has been played a few times, and now 1 6.ic4!N is crushing.) The rook move is Black's only chance, but it is not good enough:

85

8 7 6 5 4 3 2 1 a

b

e

d

e

f

g

h

14 ...fxg6

a

b

e

d

e

f

g

h

1 5 .Wh5! Wxb2 1 6.!!d 1 White has a huge attack. A possible finish is 1 6 . . . !!a7 1 7.ic4! Wxg2 1 B .ixd5 Wxh l t 1 9 .�f2 and wins. The early queen exchange does not help Black either: 1 3 . . . We3t 1 4 .We2 Wxe2t 1 5 .ixe2 dxe5 ( 1 5 . . . hxg5 1 6.ixg5 dxe5 1 7. liJ f6t liJxf6 1 B .ixf6 !!gB 1 9 .ixe5 left White a clear pawn up in Videnova - Avagianos, Rethymno 20 1 1 .) Here White can improve on the game Yilmaz - Keler, Konya 20 1 2, with: 1 6.g6!N fxg6 ( 1 6 . . . f5 1 7. liJ f6t liJxf6 1 B .ixf6 !!gB 1 9 .0-0-0) 1 7.0-0-0 ie7 1 B .ixe7 �xe7 1 9 .if3 White has a clear advantage.

1 4 . . . We3t does not help: l S .We2 fxg6 1 6. liJ f6t �f7 1 7.liJ xd5 Wxe2t 1 B .ixe2 cxd5 1 9 . 0-0t �gB 20.if6 !!h7 2 1 .b3 Even though the material remains equal, Black is too far behind in development.

I S.ttlf6t �f7 In the event of 1 5 . . . liJ xf6 1 6.ixf6 !!gB 1 7.exd6± Black is a pawn down with a weak structure. 8 7 6 5 4

Bl) 13 ... !!b8

3

This was played successfully in a recent Grandmaster game. It makes some sense for Black to avoid the awful structure that occurs after . . . dxe5, but the idea runs into problems against accurate play.

2

14.g6! 1 4 .Wf3 ie7 1 5 .id3 We3t 1 6.Wxe3 liJ xe3 1 7.�d2 liJ f5 was pretty unclear in Hector Ziska, Copenhagen 20 1 4 .

1 a

b

e

d

e

f

g

h

16.ttlxdS exdS

1 6 . . . cxd5 1 7.Wf3t �gB 1 B .id3 is too dangerous for Black.

17.YlYBt �g8 18 ..id3! dxeS 19 ..ixg6N

86

6 . . . e6 7.f4

1 9 .l"i:fl ? allowed Black to escape with a draw in Yazgeldiev - Odeev, Turkmenistan 20 1 1 .

Going after the targets.

16 �f4 .••

This is the only way to defend the g6-pawn.

8 7

1 6 . . .'IWxg2 Taking the kingside pawns doesn't help Black to generate any counterplay. 1 7.l"i:fl ! \Wxh2 ( l 7 . . . tLl e3 leaves White with only one good move, but it is absolutely crushing: 1 8 .\Wa4! +-) 1 8 .\Wg4 White had a decisive attack in Drozel - Wosch, corr. 20 1 3 . If Black tries 1 8 . . . g5 , then 1 9 .ig3 \Wa2 20.exd6 decides the issue.

6 5 4 3 2 1 a

b

e

d

e

f

g

h

19 l"i:b7 20 .ifl YNd8 2 1 .0-0t ..•

•

Compared with the aforementioned game, White has still brought his rook to the f-file, but he has also collected the g6-pawn and brought his king to safety. Black's absolute lack of development makes his position too unstable.

B2) 13 YNxb2 14.gbl YNxa2 •••

1 6 . . . tLl e3 1 7. tLl f6t <;t>d8 ( l 7 . . . <;t>f7 is refuted by 1 8 .ixg6t!, when 1 8 . . . <;t>xg6 1 9 .\Wh5t 'it>g7 20.tLle8t leads to mate in two, while 1 8 . . . 'it>g7 is not much of an improvement, and 1 9 .\wf3 led to a quick win for White in Poulopoulos Papadakis, Athens 20 1 3 .) 1 8 .tLld5t ( l 8 .tLlg4t is the same.) 1 8 . . . g5 1 9 .tLlxe3 8 7 6 5

8

4

7

3

6

2

5 a

4 3 2 1 a

b

e

d

e

f

g

h

1 5.g6! We are already familiar with this idea to create huge weaknesses around the black king.

1 5 ... fxg6 16 .id3 .

b

e

d

e

f

g

h

1 9 . . . gxh4N The only move that makes sense, but it is soon refuted. ( l 9 . . . \Wa5 20.ig3 \Wxc3t 2 l .<;t>e2+- Fier - Matuszewski, Prague 20 1 2.) 20.\Wf3 ig7 2 1 .\Wxc6 ixe5 22.0-0 l"i:a7 23.l"i:b8 l"i:c7 Now 24 .\Wb6 wins comfortably, while 24.l"i:xc8 t achieves the same result in a flashier way.

17.�f6t <;t>f7 17 . . . <;t>d8? loses to 1 8 .tLld5t g5 1 9 .tLlxf4 gxh4 20.\Wf3, threatening both tLlg6 and \Wxc6.

Chapter 5 - I ntroduction and Sidelines

87

cJie8 24 .lLlf6t would have decided the game immediately.

8 7

2 1 .Wfxg2 lLlxg2 22.xg2 dxe5

6

Other moves are no better. Here are a few illustrative lines:

5 4

22 . . . �g6 23.Ei:f3 :ke7 24.Ei:g3t � f7 2 5 .Ei:fl :kxf6 26.:kxf6 Ei:h7 allows a forced win:

3 2

8 a

b

e

d

e

f

g

h

18.�xg6t! This was first pointed out by Kozul in an NIC Yearbook survey, and has resulted in a clean sweep of white victories so far.

7 6 5 4

1. ."""-. ///////

L=.'/uu//,. V///.///' . . . . . . /

3 2

18 ... xg6 1 8 . . . lLlxg6 1 9 . 0-0 leaves Black's king too exposed. A sample line is 1 9 . . . lLl xh4 20.'lWh5t lLlg6 2 1 .lLle8t! �xe8 22.'lWxg6t �d7 23J':1f7t :ke7 24.Ei:xe7t �xe7 25 .'lWg7t cJid8 26.'lWxh8t �c7 27.exd6t and mate will follow soon.

b

20 ...'lWxg2t

e

d

b

e

d

e

f

g

h

27.:kh4t cJie8 28 .Ei:g8t �d7 29.Ei:d8t cJic7 30.exd6t cJib6 3 1 .Ei:ff8 Ei:g7t 32.�f2 cJib7 33 .:ke7+22 . . . :kg7 23.lLlh5t cJig6 24.lLlxg7 Ei:g8 2 5 .:kf6 dxe5 26.Ei:f3 Ei:xg7 27.�xg7 cJixg7 gives Black a temporarily healthy material balance, but White can exploit his superior activity as follows:

19.'lWg4t f7 20.0-0

a

a

e

f

g

h

20 . . . dxe5 was tried in one game, and here 2 1 .lLlh5!N Ei:g8 (2 1 . . .:kc5t 22.cJih 1 doesn't change anything) 22.Ei:xf4t exf4 23.'lWxf4t

a

b

e

d

e

f

g

h

28.Ei:g 1 ! Ei:a7 29.�f2t �h7 30.Ei:f8 Ei:c7 3 1 .�e3 a5 32.�e4 �a6 33.�xe5 +-

6 . . . e6 7 . f4

BB

B3) 13 dxe5

8

•••

7 6 5 4 3 2 1 a

b

e

d

f

e

g

h

23. c;t> h3!?

There is more than one way to intensify White's attack, but I like the idea of opening the g-file for subsequent checks.

23 ... a5 23 . . . e4 runs into 24.liJd5t! �gB 25 .�gl t i.g7 26. liJ e7t �h7 27.liJxcB and White wins a piece.

8 7 6 5 4 3

b

a

e

d

f

e

g

h

14.g6! Again we see this typical idea, which ruins Black's structure and creates new targets for White's pieces.

14 ... fxg6 1 4 . . . '?Ne3t This way Black manages to exchange queens, but Black's problems are far from over in the endgame. 1 5 .'?Ne2 fxg6 1 5 . . . '?Nxe2t 1 6.i.xe2 fxg6 1 7. liJ f6t 1t>f7 I B .liJ g4 transposes to the main line, while I B .O-O:!: is also good. 1 6. liJ f6t 1 6.'?Nxe3 liJxe3 17 . .tf6 �h7 I B .lt>d2 liJxfl t 1 9 .�hxfl is another pleasant endgame. 16 . . . 1t>f7 1 7.liJg4 '?Nxe2t I B .i.xe2

2 1 a

27J�g4!

b

e

d

White wins thanks constructed mating net.

e

f

to

g

a

h beautifully

a

b

e

d

e

f

g

h

Chapter 5 - Introduction and Sidelines 1B . . .ig7 1 9 .0-0t � f4 Or 1 9 . . . i>gB 20.if6 with an obvious advantage. ' 20.g3 g5 2 1 .gxf4 exf4 22 .if2± Azarov - Yilmaz, Rijeka 20 1 0.

B9

1 5 . . . ie7N is another idea, but after 1 6.if2 c5 (or 1 6 . . . 'Wc7 1 7.�d2t) 1 7. � d2!± the g6pawn is under fire and the knight is ready to take up a perfect outpost on c4.

16.0-0-0 J.e7 The text move reaches a critical posltlon. White has a wide choice, but his advantage has not been proven as conclusively here as in some of the previous lines. I suggest a simple, calm approach .

1 6 . . . � e3 doesn't work due to 1 7. � d6t! ixd6 1 B .'Wxg6t i>f8 1 9 .1:!d2 ! ' With b2 safely protected, Black cannot prevent his exposed king from being mated soon. 16 . . . h5 is met by the familiar 1 7. � d2! when it is hard to defend g6. 8 7 6 5 4 3 2

a

b

e

d

e

f

g

h

15.Yfc2!?

1 a

b

e

d

e

f

g

h

17 ..1g3!

Defending the b2-pawn while taking aim at g6. 1 5 .if2!? is another interesting line that gives White excellent compensation, but I don't see the point of giving up the b- and a-pawns here, as the text move manages everything perfectly.

This seems simplest, although there are other good moves, for instance 1 7 . .1f2 'Wa5 1 B .ic4 0-0 1 9 . h4� and White's position clearly has a lot of potential .

15 ... gb8N

This is an awkward-looking computer attempt to keep Black in the game.

The best try. 1 5 . . .'�� e 3t has been played a couple of times, with joyless results for Black. 1 6.ie2 � f4 1 7.�f6t i>f7 1 B .�g4! 'Wxe2t ( 1 B . . . �xg2t 1 9 .�fl h5 20.i>xg2 hxg4 2 1 .gafl t+­ Kanmazalp - Yaramis, Konya 20 1 1 .) 1 9 .'Wxe2 �xe2 Now in Agamaliev - Kovchan, Konya 20 1 1 , the simple 20.i>xe2N± would have left Black with too many weaknesses.

17 .. J3£8

1 7 . . . � f4 1 B . � d2!± is strong. 1 7 . . . � e3 loses in exactly the same way as on the previous move: 1 B . � d6t! ixd6 1 9 .'Wxg6t i>fB 20.l:!d2+17 . . . 0-0 1 B .� d2! The same idea again. 1 B .. J�f5 1 9 .� c4 'Wc5 20.id3± Black has little counterplay and will have to lose material .

6 . . . e6 7.f4

90

8

8

7

7

6

6

5

5

4

4

3

3

2

2 b

a

1 a

b

e

d

e

f

g

h

1 8J�el!?

This is the simplest way to continue. 1 8 . lLl d2 lLl e3! 1 9 .1lMxg6t l'!f7 isn't as clear. 1 8 . .id3 lLl e3 1 9 .1lMd2!? lLl xd l 20.l'!xd l is also quite good according to the computer, but there is no need to sacrifice an exchange.

1 8 l'!f5 19 ..!d3±

e

d

e

f

h

g

Black has problems with the d6-pawn. She tried to solve them with 1 2 . . . lLle8, but after 1 3 ..ixe7 1lMxe7 1 4.e5 White had a clear advantage in Karjakin - Polgar, Bilbao (blind) 2007.

Cl) 8 �a5 ..•

This has been favoured by some strong players in the past, and White will have to play precisely to prove anything against it.

...

White will soon be ready for lLl d2-c4, while Black's position remains uncoordinated and full of weaknesses.

C) 7 ...j,e7 8.�f3 In this section we will deal with a few set­ ups which do not quite fit in to any of the other chapters involving 7 . . . .ie7. They are not at all bad, but at the same time they are not significant enough to warrant their own dedicated chapters. We will start with the unusual Cl) 8 �a5, before checking a sideline after the more common C2) 8 �c7. •••

•••

8 . . . lLl c6 gives White no special problems. It has been played by Judit Polgar, but only in a blindfold game, so we don't need to take it too seriously. 9 . 0-0-0 lLlxd4 1 O .l'!xd4 1lMc7 1 1 ..ie2 0-0 1 2.l'!hd 1

9.0-O-O .id7 Black's set-up looks strange, but he is ready to play . . . lLl c6 with a normal Sicilian position . I recommend an immediate tactical strike before Black gets a chance to complete development. 8 7 6 5 4 3 2 1 a

b

e

d

e

f

g

h

10 ..bf6! .bf6 1 0 . . . gxf6 l 1 .f5 gives White an excellent version of a typical structure.

Chapter 5 - Introduction and Sidelines

91

C2) 8 W!c7 9.0-0-0

1 1 .e5 dxe5 12,c�he6!

.••

1 2.fXe5 is met by the amazing defence 1 2 . . . ic6! 1 3 : tLlxc6 ig5 t 1 4.i>b l tLlxc6 as in Chernih - Le Quang, Yerevan 2006.

12 .he6 •.•

1 2 . . . fXe6 1 3 .W!xb7 0-0 0 3 . . . ic6 1 4 .W!c8t id8 1 5 .W!xe6h-) 1 4.W!xa8 ic6 1 5 .W!a7 exf4 1 6.ic4 left Black with insufficient compensation for the exchange in Sochacki Chabradze, Gorges 20 1 1 .

8 7 6 5 4 3 2

13.W!xb7 exf4 14.W!xa8 0-0

1 a

8

b

c

d

e

f

g

h

From this posltlon the great maJonty of games have continued with . . . tLl bd7, either with or without a preliminary . . . h6 and ih4. In the final part of this chapter we will consider an interesting way for Black to postpone the development of his knight.

7 6 5 4 3

9 ... 0-0

2

a

b

e

d

e

f

g

h

15.tl�d5! It is easy for White to defuse Black's queenside aspirations and consolidate with his extra exchange.

15 W!xa2 16.tLlxf6t gxf6 •••

Now in Startstek - Combatiente, Internet 20 1 2, the most accurate continuation would have been:

This is an older way to handle the position, but it should not be forgotten. The main idea is that the b8-knight can go to c6 or d7 according to circumstances. 9 . . . b5? is premature due to l O.e5 ib7 1 1 .W!g3 .

10.W!g3! 1 0.id3 is not particularly strong because of 1 O . . . tLl c6!, immediately drawing attention to the lack of support for the knight on d4.

17.W!e4!N W!aIt 18.c;t?d2 W!a5t 19.b4 W!c7 20.i.d3 f5 21 .W!d4±

1 0.g4 is also not dangerous due to 1O . . . b5 I l .ixf6 ixf6 1 2.g5 ixd4 1 3 Jhd4 tLl c6 when Black stands at least equal. These two lines show the . . . tLl c6 plan working perfectly.

White is ahead on material and Black's kingside is a wreck.

10 ftlbd7 •••

Black needs to prevent e4-e 5 . 1 O . . . tLl c6 does not make much sense here, as White can either play 1 1 .e5 immediately or exchange on c6 first.

6 . . . e 6 7. f4

92

is: 1 8 . . . :B:ad8 1 9 .cJib l ! ? lLlxe5 20.lLldxb5 axb5 2 1 .'lWxe5 'lWxe5 22.:B:xe5 b4 23.axb4 :B:d4 Black maintains material equality, but both 24.b5 and 24 . .te2 leave White with the more dangerous pawn majority in the endgame.

b

a

e

d

f

e

g

h

1 1 ..id3

1 1 ..ie2 b5 1 2.e5 dxe5 1 3 . fxe5 lLl xe5 1 4 . .tf3 led to an eventual win for White in Archeo Dzwonek, corr. 2002, but after 1 4 . . . .tb7!N 1 5 ..txb7 'lWxb7 1 6.'lWxe5 lLl g4 1 7.'lWf4 .txg5 1 8 .'lWxg5 lLl f2 ! it is difficult to claim much of an edge for White.

l l oo.b5 12J�hel b4

a

b

e

d

e

f

g

h

13.tOd5! Nothing else makes sense.

This has been the usual continuation, forcing the play before e4-e5 comes.

13oo.exd5 14.lLlf5 Ad8 1 5.tOxg7! c,!,>h8!

1 2 . . . .tb7!?N has not yet been tried, but it is worth checking how the game may continue: 1 3 .e5 dxe5 1 4. fxe5 lLl h 5 1 5 .'lWh4 .txg5 t 1 6.'lWxg5 g6

16.exd5

a

b

e

d

e

f

g

Otherwise lLl f5 is killing.

1 6.e5? dxe5 1 7. fxe5 lLl e4 favoured Black in Wittman - Leow, Thessaloniki (01) 1 984. 1 6.lLlfSN is playable, but after 16 . . . lLlh5 1 7.'lWh4 .txg5 1 8 .'lWxg5 lLl df6 1 9 .e5 :B:g8 Black has enough resources to maintain the balance.

h

White does not have an immediate way through, but the calm 1 7.g4 lLl g7 1 8 .a3 maintains an edge. A possible continuation a

b

e

d

e

f

g

h

Chapter 5 - Introduction and Sidelines

Conclusion

16 .. .tik5!N

After 16 .. .l�! g8? 1 7. lLl f5 lLl c5 1 8 . .txf6t Black resigned in Koziak - Buturin, Lvov 1 997, in view of 18 . . . .txf6 1 9 .Wlxg8t �xg8 20J!e8#. The text move enables Black to survive the attack. It was proposed in Sicilian Attacks by Yakovich, who called the position unclear, but it seems to me that White maintains some advantage.

17.YlYh4! tilfe4! 18 ..L.:e4 f5 There is no other defence.

b

e

d

e

f

From the popular tabiya after 6 . .tg5 e6 7.f4, this chapter has dealt with a variety of set­ ups that do not quite fit into any of the other chapters. 7 . . . .td7 is a quirky move, but not a bad one. White obtains a good position with the typical plan of casding, placing the queen on g3 and forcing through e4-e5, although he must keep playing energetically and possibly sacrifice some material if he is to exploit his initiative. 7 . . . lLl c6? was tricky for a while, but the critical continuation of 8 . lLl xc6 bxc6 9.e5 h6 1 0 . .th4 g5 I l . fxg5 lLl d5 1 2. lLl e4 Wlb6 1 3 .c3! can be regarded as a near-refutation of Black's 7th move.

19.i.xd8 YlYxd8 20.YlYxd8 fhd8

a

93

g

h

21 .tilxf5 tilxe4 22J�xe4 i.xf5 23J�xb4;t White has good chances in the endgame, with four pawns against a bishop.

7 . . . .te7 8 .WlS is one of the most important positions in the Najdorf, arguably second only to the Poisoned Pawn in terms of the complexity of the ensuing positions. We will return to this in Chapters 1 0- 1 2, but for now we can say that 8 . . . Wla5 does not work out too well, while 8 . . . Wlc7 9 . 0-0 0-0-0 is trickier, but still quite promising for White in the critical line involving a kingside attack with a forced transition to an endgame.

6 ... e6 7.£4 a

b

c

d

e

f

g

7...11c7 Variation Index 1 .e4 c5 2.tiJ a d6 3.d4 cxd4 4.tiJxd4 tiJ f6 5.tiJ c3 a6 6 ..tg5 e6 7.f4 �c7 B.�e2!? A) B b5 B) B tiJ c6 9.0-0-0 B 1 ) 9 ....te7 B2) 9 ....td7!? B3) 9 ...tiJxd4 1 0J�xd4 B3 1 ) 1 0...b5!? B32) 1 0....te7 l 1 .�b l !N B32 1 ) 1 1 ...0-0 B32 1 1 ) 1 2.g3 B32 1 2) 1 2.h4!? B322) 1 1 ...h6 1 2 ..th4 B322 1 ) 1 2 ...g5? B3222) 1 2 ...0-0

95 96 96 100 101 101 104 105 105 1 06 lOB lOB 109

•••

•••

B3 1 ) note to 1 1 . . .gb8

B l ) after 1 2 . . . Wb6

B32) after 1 0 . . . �e7

8

8

7

7

6

6

5

5

4

4

3

3

2

2 a

b

e

d

e

f

1 3 .�e2!N

g

h

a

b

e

d

e

f

1 5 .Wd2!N

g

h

a

b

e

d

e

f

1 1 .'it>b l !N

g

h

h

Chapter 6 - 7 . . . Vfic7

1.e4 c5 2.tLJa d6 3.d4 cxd4 4.tLJxd4 c!Llf6 5.tLJc3 a6 6.�g5 e6 7.£4 Y!!c7

95

9.e5 looks natural, but after 9 ... dxe5 1 0.dxe5 we reach a line of the Polugaevsky Variation C7 . . . b5 8.e5 dxe5 9 . fxe5 Wic7 1 0 .Wie2) which I prefer to avoid in favour of l O .exf6, as covered in Chapter 7. The text move maintains the parallel with the 7 . . . lLl bd7 8 .Wie2 variation, and I do not see anything better for Black than transposing to that line.

a

b

e

d

e

f

g

h

This was the favourite choice of the young Garry Kasparov, and has been recommended in at least three popular books: Mastering the Najdorf by Arizmendi and Moreno, Play the Najdorf Scheveningen Style by Emms, and Play the Najdorf Sicilian by Rizzitano. It is easy to understand the appeal for Black, as play tends to take on a slower pace, with fewer forcing lines than in most Najdorf variations. Quite a lot of games have continued with an exchange on f6 over the next few moves, leading to a long strategic fight. Even though White has some decent plans there, Black will have achieved his main goal of reaching a stable position with no immediate danger. Instead I would like to propose a different approach which ties in with my recommendation against 7 . . . lLl bd7.

S.Y!!e2!? One of the most natural moves here is 8 . . . lLlbd7, which transposes into 7 . . . lLl bd7 8.Wie2 Wic7 as discussed in Chapters 8 and 9 . This leaves us with two alternatives t o cover in the following pages: the sideline A) S ... b5 and the more popular B) S tLJc6. .••

A) S b5 9.a3! •••

9 . . . �e7 can simply be met by l O .e5 dxe5 l 1 . fxe5 lLl d5 1 2. lLlxd5 exd5 1 3 .�xe7 Wixe7 1 4.g3 0-0 1 5 .�g2 �e6 1 6. 0-0-0;1; with an obvious edge to White. 9 . . . lLl c6 is normally dubious when Black has to recapture with the queen, and White gets easy play after 1 0.lLlxc6 Wixc6 1 1 . 0-0-0 �e7 ( 1 1 . . .�b7?! 1 2.lLld5!±) 1 2.g4 �b7 1 3 .�g2;1;.

10.0-0-0 �e7!?N The only games in my database continued with 1 0 . . . lLl bd7, leading to page 1 34 . 1 0 . . . lLl c6? i s n o t a good idea due t o l 1 .lLlxc6 �xc6 1 2. lLl d 5 ! ± .

6 . . . e6 7.f4

96

9.0-0-0

1 1 .£5! 1 1 .g4 h6 1 2 .ih4 tt'l c6! is an important extra option, when 1 3 .tt'lxc6 ixc6 is okay for Black.

I once played 9 .ixf6, but nowadays if I wanted to change the structure this way I would do it on the previous move. Black has tried three options: BI) 9 ... i.e7, B2) 9 ...Ad7!? and B3) 9 ... tLlxd4. As we will see, there are some reasons why Black may try to delay the knight exchange, but doing so also gives White some extra options.

BI) 9 .. i.e7 .

b

a

e

d

e

f

g

h

lo.tihc6!

1 1 .. e5 12 .ixf6! gxf6 .

This move forces us to take a completely different path to the 9 . . . tt'lxd4 1 O .1'!xd4 ie7 line.

•

1 2 . . . ixf6 allows 1 3 . tt'l dxb5 ! axb5 1 4 .tt'l xb5 when White will get three pawns plus the initiative for the piece.

13.tLlb3;t Black's position is solid for now, but with the d5-square securely under White's control, it is clear that Black will suffer for a long time.

B) 8 ... tLlc6

1 0 .Q;>b l would be fine if Black had to swap knights and transpose to variation B3, but 1 O . . . id7!N is more of a problem. ( 1 0 . . . 0-0 1 1 .g3 id7 1 2.ig2 tt'lxd4 1 3 .1'!xd4 ic6 1 4 .1'!hd l could be unpleasant for Black.) 1 1 .g3 tt'lxd4 1 2.1'!xd4 ic6 1 3 .ig2 h6 1 4 .ih4 1'!c8 This position could be reached via a 9 . . . tt'lxd4 move order, but one that involves 1 1 .g3 instead of my recommended 1 u ttb l ! . Other moves such as 1 O.g4 can be met by 1 0 . . . tt'l xd4, transposing to another line I am not recommending.

IO . bxc6 ..

a

b

e

d

e

f

g

h

This is the usual way to avoid transposing to the . . . tt'l bd7 lines.

Recapturing with the queen seems more awkward: 1 O . . . Wfxc6 1 1 .e5 dxe5 1 2 .fXe5 tt'ld5 1 3 .ixe7 tt'l xc3 After 1 3 . . . tt'lxe7 1 4.Wfh5! White has an excellent version of the structure with the pawn on e5. Black is a long way behind with his counterplay, and can do little to prevent White from building his initiative with id3 , tt'l e4 and so on.

Chapter 6 - 7 . . Y!ic7

97

2 1 . . .Wlxd6 22.exd6t \!lxd6 23.l'!d l t \!lc6 24 ..ixg8 l'!xg8 2 5 .g3 It will take a miracle for Black to hold this ending.

a

b

e

d

e

f

g

h

1 4 . bxc3! 1 4 .Wld2!? is a safe alternative: 1 4 . . .'IWd7 ( 1 4 . . . .id7 1 5 . .ih4! is excellent for White, the main point being 1 5 . . . ltJxd l 1 6.Wlb4!.) 1 5 .Wlxc3 Wlxe7 1 6 . .id3;:!; White intends .ie4, and keeps a pleasant edge without damaging his pawn structure. The text move is more ambitious. 1 4 . . . \!lxe7 1 5 .Wlg4 .id7 1 6 . .ic4! White should not hurry with 1 6.Wlxg7, as after 1 6. ; .l'!hfS it is not so easy to complete development. The bishop move solves that problem, and Black still does not have a convenient way to defend g7. 1 6 . . . Wlc5 16 . . . b5 1 7.Wlxg7 l'!hfS 1 8 . .ib3 Wlxc3 1 9 .\!lb l Wlc5 20.l'!d6+-

a

b

e

d

e

f

g

h

1 1 .e5 dxe5

After 1 1 . . .ltJd5 1 2 . .ixe7 Wlxe7 1 3 . ltJ e4 dxe5 1 4. fxe5 0-0 1 5 .c4;:!; Black's dark squares are too weakened.

12.YlYxe5! A typical way to play against the spoilt queenside structure. The ensuing endgames are always more pleasant for White. 1 2 . fxe5 is not so clear: 1 2 . . . ltJ d5 1 3 . .ixe7 ltJ xc3! 1 4.Wlf3 ! ? ( 1 4 .Wlg4 Wlxe5!) 1 4 . . . ltJ xa2t ( 1 4 . . . ltJ xd l 1 5 . .id6 Wlb6 1 6.\!lxd l �) 1 5 .\!lb l \!lxe7 1 6. \!lxa2 l'!d8°o

a

b

e

d

e

f

g

h

1 7.Wlxg7 l'!hg8 1 8 .l'!xd7t \!lxd7 1 9 .Wlxf7t 'iff c6 20.Wlxe6t \!lc7 2 1 .Wld6t White could keep the queens on: 2 1 .Wlf7t \!lb6 22.Wlf6t 'iff a7 23 . .ixg8 l'!xg8 24.e6± a

b

e

d

e

f

g

h

98

6 . . . e6 7.f4

12 �b6 ...

Every queen move has a drawback of some kind. The problem with this one is that liJ a4 may prove annoying at some point. 1 2 . . .WlxeS 1 3 . fxeS liJ dS 1 4.ixe7 l S . liJ e4;:!; is pleasant for White.

<:!ixe7

1 2 . . . Wlb8 Black has managed to draw two correspondence games after this move, but I found a promising new idea. 1 3 .g3!N Defending the f4-pawn. The significance of this will soon become obvious. 1 3 . . . h6 This is the only real attempt to j ustify the queen's position on b8. 13 ... WlxeS 1 4 . fxeS liJ dS l S .ixe7 <:!ixe7 1 6. liJ e4 gives White a slightly improved version of 1 2 . . . WlxeS, as the extra g2-g3 move may be of some use. 1 3 . . . 0-0 1 4. liJ e4! WlxeS l S . fxeS liJxe4 1 6.ixe7 liJ f2 ( 1 6 . . .l:%e8 1 7.ig2 liJ f2 1 8 .ixc6±) 1 7.ig2 liJ xd 1 ( 1 7 . . . liJ xh 1 1 8 .ixf8 +-) 1 8 .E:xd 1 ib7 1 9 .ixf8 <:!ixfB 20.E:d7 E:b8 2 1 .b4± 8 7 6 5 4 3

A final try is: 1 2 . . . Wla7!?N 1 3 .ie2 1 3 .id3 0-0 1 4. liJ e4?! is a mistake due to 1 4 . . . liJ g4! l S .Wlc3 f6 with excellent chances for Black. 1 3 .E:d3!? is a more interesting alternative, with the blunt idea of transferring the rook to the kingside. 1 3 . . . aS ( 1 3 . . . 0-0 1 4.E:g3! liJ d7 l S .Wle 1 ixgS 1 6. fxgS followed by id3 gives White good attacking chances.) 1 4. liJ e4 liJ xe4 l S .Wlxe4 0-0 1 6.E:g3 White has chances to develop a serious initiative. 8 7 6 5 4 3 2

a

b

e

d

e

f

g

h

1 3 . . . 0-0 1 4. liJ e4! liJxe4 Both 1 4 . . . liJ d7 l S .Wlc3 f6 1 6.ih4;:!; and 1 4 . . . liJdS l S .ixe7 Wlxe7 1 6.g3;:!; leave White with a pleasant advantage. l S .Wlxe4 White is happy to allow the exchange on gS as it will increase his attacking potential. l S . . . E:b8 After l S . . . ixgS 1 6.fxgS;:!; the gS-pawn cramps Black's kingside. White will play id3 next, intending to meet . . . g6 with h4-hS .

2

a

b

e

d

e

f

g

h

1 4.ixf6! ixf6 l S .WlcS The c6-pawn is impossible to defend. l S . . .ie7 l s . . . ib7 1 6.ig2 Wlc7 1 7. 5 ie7 1 8 .Wlc4± 1 6.Wlxc6t <:!if8 1 7.ig2 E:a7 1 8 .Wla4±

a

b

e

d

e

f

g

h

Chapter 6 - 7 . . . 'Wc7

99

1 6.id3 1 6.ixe7 'Wxe7 1 7.id3 g6 I B .h4t is a tempting alternative. 1 6 . . . f5 1 7.'We5 ixg5 1 B . fxg5;!; Black has gained some breathing room on the kingside, but the e6-pawn is backward and his dark squares are weak. 8 7 6

a

5

b

e

d

e

f

g

h

14.�a4!

4

This seems like the right time, as after 1 4.ih4 0-0 1 5 .ltJ a4 'Wa7 the black queen is quite well situated.

3 2

14 ...'Wb7 a

b

e

d

e

f

g

h

13.ie2!N I like this simple developing move, as it maintains the options of both ltJ e4 and ltJ a4. 1 3 .ltJa4 has been played, but the critical reply has not: 1 3 . . . 'Wa7!N ( 1 3 . . . 'Wb4 was seen in Hracek - Woj tkiewicz, Warsaw 1 990, and here both players must have overlooked that after 1 4.a3!N the knight is untouchable. 14 . . . 'WbB [ 1 4 . . . 'Wxa4? 1 5 .:B:d4 traps the queen.] 1 5 .ie2;!;) 1 4.f5 ! ? 0-0 1 5 .ie3 'Wb7 1 6.fxe6 ixe6 1 7.ltJc5 ixc5 I B .'Wxc5 White has the better structure and a nice pair of bishops, but Black has decent prospects for counterplay on the queenside.

13 ... h6!?

This seems like the most stubborn attempt. Mter 13 . . . 0-0 1 4 .ltJe4 ltJ d7 1 5 .:B:xd7 ixg5 1 6.:B:d3! the rook perfectly restricts Black's counterplay, and is also ready to switch to the kingside for an attack. 1 6 . . . ie7 1 7.:B:hd l f6 I B .'Wh5!?t

Now 1 4 . . . 'Wa7 can be met by 1 5 .ixf6! ixf6 1 6.'Wc5;!;. 14 . . . 'Wb4 allows White an intngumg possibility: 1 5 .a3! ? ( 1 5 .:B:d4 is a safe way to maintain a slight edge.) 1 5 . . . 'Wxa4 1 6.:B:d4 'Wxd4 1 7.'Wxd4 hxg5 I B . fxg5 ltJ h7 1 9. h4 0-0 20.id3t Even though Black has a rook and two pieces for the queen, his pieces are disorganized and White has decent attacking chances.

1 5 ..ih4 0-0 16.gd3! 8 7 6 5 4 3 2 1 a

b

e

d

e

f

g

h

6 . . . e6 7.f4

1 00

16 .. J�d8 1 6 . . . tiJ g4 1 7 . .ixg4 .ixh4 1 8 . .if3 also favours White.

17.gb3 Wta7 1 8.�b6 gb8 19 ..ifl c5 This pawn sacrifice is the best practical attempt to free Black's position.

20 ..bc5 .bc5 2 1 .Wtxc5 tLJd5 22.gdl!� Black cannot exploit the pin, and White should be able to untangle himself within a few moves and keep his extra pawn.

B2) 9 ....id7!?

1 1 . .ixf6 gxf6 has been played, but I see no point in giving Black such a fantastic pawn wall. 1 1 .Wi'd2 is similar to the main line, but it allows 1 1 . . . d5!. That explains why we need the queen on the e-file. 1 1 .g3!? is another possible way to develop the light-squared bishop, but I prefer the main line.

H ... a5 1 1 . . J�b8 l2 . .ie2 is similar to the main line, but White could also consider l 2 . .ixa6! ? 1 1 . . . .ie7 looks natural, but l 2.e5! leaves Black with an unpleasant choice: 8 7 6 5 4 3 2

a

b

e

d

e

f

g

h

Here Black tries to be even smarter with his move order than in the previous line. The bishop is not so active on d7, but White still needs to know how to continue.

10.�xc6! The reasoning for this move is more or less the same as in the previous line. 1 O.i>b l can be met by l O . . . .ie7, when White does not get the desired transposition to variation B32.

lo ... bxc6 l O . . . .ixc6?! 1 l .tiJ d 5 ! was unpleasant for Black in Milliet - Stankovic, Basel 20 1 O.

H .Wte3N

a

b

e

d

e

f

g

h

a) l 2 . . . tiJd5 1 3 .tiJxd5 cxd5 1 4 . .txe7 Wxe7 1 5 .exd6t Wi'xd6 1 6 . .ic4! .ic6 1 7.f5t The black king is clearly vulnerable. b) l 2 . . . dxe5 1 3 .Wi'xe5t Compared to the earlier line B 1 , Black has the extra move . . . .id7, but this actually favours White, as the bishop is tactically vulnerable and the . . . tiJd7 retreat is not available. 0 3 .fxe5!? also deserves attention, for instance: 1 3 . . . tiJd5 1 4.tiJxd5 cxd5 1 5 . .ixe7 i>xe7 1 6.Wi'g5 t i>f8 1 7 . .id3 .ib5 1 8 .ghfl t)

12 ..te2 Simplest, although l 2 . .ic4!? is also playable. It is hard to suggest a further course of action

Chapter 6 - 7 . . YJic7 for Black; . . . d5 isn't possible, . . .fie7 runs into e4-e5, while . . . Elb8 and . . . 'lWb7 can always be met by b2-b3.

101

B31) IO b5!? ••.

8 7

12 J�b8 ••

1 2 . . . a4 1 3 .Elhe l a3 1 4.b3 gets nowhere for Black.

6

1 2 . . . fie7 1 3 .e5 dxe5 1 4.'lWxe5 is similar to 1 1 . . .fie7 as mentioned above, and the inclusion of . . . a5 and fie2 must surely favour White. 12 . . . h6 1 3 .fixf6!? ( 1 3 .fih4 is perfectly sensible as well) 13 . . . gxf6 1 4 .fih5t shows another use of the bishop on e2 . 8 7

b

a

e

d

f

e

g

h

It seems odd to try and delay . . . fie7 any longer, but there is a point: with the bishop on e7, Black constantly has to worry about e4-e5, as we will see in variation B32.

1 l .g3

6

l 1 .g4 can be met by 1 1 . . .Elb8 ! ? Now exchanging on f6 makes less sense, and against other plans Black's set-up is j ustified.

5 4 3

1 l J;b8 ••

2 1 a

b

e

d

e

f

g

h

13J�hel YlYb6

1 1 . . .fie7 has been played a few times, but it makes little sense: 1 2.fig2 Elb8 1 3 .e5 dxe5 1 4. fxe5 tLld5 Now in Hector - Hillarp Persson, Borlange 20 1 4 , White missed an opportunity to win a pawn with:

1 3 . . . 'lWb7 1 4.b3t

14.YlYxb6 l3xb6 15 .ih4 •

The endgame is clearly more pleasant for White.

B3) 9 �xd4 IO.Elxd4 •••

8 7 6 5 4 3

Now B3 1) IO b5!? is an interesting sideline, but the most popular move has been B32) •••

IO .ie7.

2

•••

a

1 5 .'lWd2!N

b

e

fixg5

d

e

f

( 1 5 . . . tLl xc3

g

h

1 6.fic6t!+-)

1 02

6 . . . e6 7.f4

1 6.Wi'xg5 1 7.i.xd5±

0-0

( I 6 . . . lD xc3

1 7.i.c6t! +-)

I also considered 1 1 . . . lD d7N to avoid any i.xf6 ideas, but now Black will not be able to play . . . b4 in time. 1 2.i.g2 Eib8 1 3 . lD d 5 ! Wi'a7 1 4.i.h4! Both the knight on d5 and the rook on d4 are untouchable. 1 4 . . . h6 1 5 .g4!± Followed by i.f2 . 8 7 6 5 4 3 2 1 b

a

e

d

f

e

g

h

1 4.Eie l ! ? Setting up lDd5 ideas. 1 4 . . . i.d7 1 4 . . . i.b7 1 5 .lDd5! is a rude shock. 1 5 . . . Wi'd8 ( I 5 . . . exd 5 ? 1 6.exd5t i.e7 1 7.Eie4+-) 1 6.lDe3± White has achieved a favourable regrouping with gain of time. 14 . . . Wi'c5 1 5 .Wi'd2 i.b7 1 6.�b l i.e7 1 7. lD d l �f8 1 8 .lD e3 �g7 1 9 .Eid l i s another possible direction, where White maintains a typical edge. 1 5 . lD d l Even though Black's position i s solid, it does not offer much hope for counterplay. It is not easy to activate the dark-squared bishop, launch a queenside offensive, or find a suitable home for his king. Meanwhile White intends to improve his position with lD e3 and �b l , then continue manoeuvring. 1 5 . . . h4 Black should probably get rid of this pawn before it becomes a target. 8

12.i.g2 1 2.i.xf6!?N This is the safe player's move, and in the long run White's position should be more pleasant. An important point is that Black has committed himself to . . . b5 and . . . Eib8, depriving himself of the opportunity for long castling. 1 2 . . . gxf6 1 3 .i.g2 h5 1 3 . . . i.e7 1 4.�b l i.d7 1 5 . lD d l;!;

7 6 5 4 3 2

a

b

e

d

e

f

g

h

1 6. lD e3 hxg3 1 7.hxg3 The position remains complex, but I prefer White for the reasons outlined above. Black has exchanged off the weak h-pawn, but the opening of the kingside leaves his king with no place other than the centre. Even though the king looks safe there right now, its positioning makes it hard for Black to connect his rooks and arrange any meaningful counterplay. White, on the other hand, can shuffle his pieces

8 7 6 5 4 3 2

a

b

e

d

e

f

g

h

1 03

Chapter 6 - 7 . . :�c7 around in preparation for either a break in the centre or an invasion along the open h-file. On a cautionary 'note, White should keep in mind the solidity of Black's position and the latent potential of the bishop pair, and avoid getting too aggressive too quickly.

1 5 .t.xb2?! Wi'b7 is unclear, but of course White should avoid capturing the b-pawn, as it will ensure the safety of his king - for a while at least. Now it is not easy for Black to choose a square for his queen, and after that he will have problems completing the development of his kingside.

12 ... b4 13.Lf6!?N With Black committed to . . . b4, this exchange is more critical than it was on the previous move. 1 3 . lt:l d 1 It:l d7 1 4 .t.h4 h6 1 5 .i>b 1 a5 is a slightly unusual set-up for Black, but keeping the bishop on f8 gave him good play in Le Bled - Ruemmele, corr. 20 1 0 . I spent some time looking at 1 3 .:B:c4N Wi'a5 ! 1 4.lt:ld5!?, but it does not quite work as desired: 1 4 . . . exd5 1 5 .exd5t t.e7 1 6 . .txf6 gxf6 1 7.:B:e 1 Wi'd8 1 8 .Wi'h5 Wf8 1 9 .:B:ce4 :B:b7 Black is doing fine.

13 ... bxc3!? This is the most testing reply. 1 3 . . . gxf6 1 4 . lt:l d 1 a5 1 5 .lt:l e3;!;; gives Black a much less solid version of the 1 2 .t.xf6 line analysed above, as the . . . b4 and . . . a5 moves have left some holes on the light squares.

1 5 .. .'IWb7 1 5 . . . Wi'a5 ?! 1 6.t.c3 Wi'b5 1 7.:B:d 1 ± is bad for Black, while after 1 5 . . . Wi'd7 1 6.t.d4 a5 ( 1 6 . . . t.b7 1 7.:B:d 1 ;!;; ) 1 7.:B:c3 followed by :B:d 1 , Black's lack of development portends an unpleasant future for him.

16 ..ic3 a5 Guarding the b4-square, but taking up additional time. I also considered: 16 . . . t.d7 1 7.:B:b4 t.b5 1 8 .Wi'e3 Wi'c7 1 9 .a4 t.c6 20.:B:xb8t Wi'xb8 2 1 .Wi'd4!?� Black's kingside is still completely stuck, while White is ready to bring his rook around and get dangerous play in the centre or on the queenside. 8 7 6

8

5

7

4

6

3

5

2

4

1

3

a

2

b

17J�dl .id7

1 a

b

e

d

14Jk4! cxb2t 1 5.c,tJbl

e

f

g

h

e

d

e

f

g

h

The computer's suggestion of 1 7 . . . h5 seems horrifyingly slow, and White can go on the attack with 1 8 . 5 ! e5 ( 1 8 . . . .td7 1 9. fxe6 fxe6 20.t.h3±) 1 9 .Wi'd2;!;; .

1 04

6 . . . e6 7 . f4

1 8.a4!?i This is the simplest move, intending to break through in the centre. ( 1 8 .i.f1 ! ? is more concrete, although I don't see any need for it. For what it's worth, the idea is 1 8 . . . i.b5 1 9 .a4! i.xc4 20.%Vxc4 i.e7 2 1 .i.xg7 l:%g8 22.%Vc3 with a dangerous initiative for the sacrificed material.)

B32) 10 ...ie7 .

This popular move reaches a surprisingly critical position, where I found it necessary to forge a completely new path.

l 1 .�bl!N Before discussing this move, I will offer a shortened version of my analysis of another important line to give you an idea of the difficulties. I l .g4 This blunt attacking move is the other critical line, but Black's position seems to hold up. 1 1 . . .h6 I analysed several other replies, but will focus on the most challenging one. 1 2.i.h4 g5 Judging by the practical results from this position, it looks as though Black has walked right into what White wanted, but despite all the tempting options, it has not been easy to demonstrate an advantage here. 8 7 6 5 4 3 2

a

b

e

d

e

f

g

h

1 3 .e5 1 3 . fxg5 ttJ d7! does not bring much, for example: 1 4.e5 ttJxe5!N An obvious improvement. 1 5 .ttJ e4 hxg5 1 6.i.xg5 i.xg5t 1 7.ttJxg5 %Ve7°o 1 3 . . . dxe5 I also analysed 1 3 . . . gxh4N in detail, but will focus on the best defensive moves. 1 4.l:%c4 1 4. fxe5 gxh4 1 5 .exf6 i.xf6 1 6.ttJd5 i.g5 t 1 7.'if?b l %Vd8 1 8 .%Ve4 0-0 1 9 . ttJ c3 %Vf6 is unclear. 1 4 . . . %Vb8 1 5 . fxg5 ttJ d7! 1 6.ttJe4 This move protects g5 and threatens g5-g6 or gxh6, the point being that . . . i.xh4 will allow l:%xc8t followed by a deadly fork on d6. 8 7 6 5 4 3 2

a

b

e

d

e

f

g

h

1 6 . . . b5! 1 7.l:%xc8 t %Vxc8 1 8 .gxh6! %Vc7 1 8 . . . l:%xh6 1 9 .i.xe7
Chapter 6 Najdorf guides referred to in the introductory paragraph. Black's two most important replies are 1 1 ...0-0 and B322) 1 1 ...h6.

-

1 05

7 . . .vlff c7

varIatIon B3222. Here White's structure is slightly looser with the pawn on g4, but he still has decent chances to convert the extra pawn.

B32 1) B321) 1 1 . .. 0-0

The following alternatives run into concrete problems. 1 1 . . . id7? is a standard developing move, but after 1 2 .e5! dxe5 1 3 .fxe5 ic5 1 4Jk4 lLId5 1 5 .lLIe4± Black is headed for trouble. 1 1 . . .e5 1 2.fxe5 dxe5 1 3 .ixf6 ixf6 1 4.lLId5 \Wd8 1 5 .l'!d3;t gives White a typical edge in what is likely to become an opposite-coloured­ bishop middlegame. 1 1 . . .b5 1 2.e5! dxe5 1 3 . fxe5 lLId5 ( 1 3 . . . ic5 ?! is bad due to 1 4 .l'!d3 lLI d7 [ 1 4 . . . lLI d5 1 5 .lLIxd5 exd5 1 6.g4 also leaves Black under pressure] 1 5 .lLIe4, with the point that 1 5 . . . 0-0? 1 6.if6! lLIxe5 1 7.\Wh5 wins for White.) 1 4.lLIxd5 exd5 1 5 .ixe7 \Wxe7 Now White has a pleasant choice.

a

b

e

d

e

f

g

h

Now you can choose between two promising moves according to your own taste. B32 1 1) 12.g3 is enough for a slight edge, while B3212) 12.h4!? may lead to sharp tactics. 1 2.e5 is too soon. 1 2 . . . dxe5 1 3 . fxe5 lLI d 5 ! This thematic pawn sacrifice offers Black decent compensation: 1 4.lLIxd5 exd5 1 5 .ixe7 \Wxe7 1 6.l'!xd5 ie6�

B321 1) 12.g3

a

b

e

d

e

f

g

h

1 6.g4 ( 1 6.g3 gives Black the option of 1 6 . . . if5 followed by . . . l'!c8; 1 6.\We3!? is perhaps the simplest way to get an edge: 1 6 . . . 0-0 1 7.id3 l'!e8 1 8 .l'!e l White keeps the more pleasant position.) 16 . . . ie6 1 7.ig2 0-0 1 8 .ixd5 ixd5 1 9 .1'!xd5 This can be compared to the 1 5 . . . lLId5 line given in the notes to the later

This resembles - and may transpose to - the later variation B3222, which arises after 1 1 . . . h6 1 2 .ih4 0-0 1 3 .g3 . However, the present move order gives White an intriguing extra possibility.

12 ... h6 13.h4!? 1 3 .ih4 reaches variation B3222 of course see page 1 09 .

13 ... h5 Neither 1 3 . . . id7 1 4.e5!± n o r 1 3 . . . e5 1 4.l'!d2 ig4 1 5 .\Wd3± are acceptable for Black.

6 . . . e6 7 . f4

1 06

16 ... exd5 8 7 6 5 4 3 b

a

e

d

f

e

g

h

14.e5! dxe5

1 4 . . . .ib 7!? leads to interesting complications, but White stays on top : 1 5 . exf6 .ixh l 1 6. fXe7 l'!feB 1 7. f5 ! hxg5 1 B .Wh2! ( l B .hxg5 exf5 1 9 .Wh5 l'!xe7 20 .Wxh l l'!el t 2 1 .l'!d l l'!aeB? is not so clear.) 1 B . . . exf5 1 9 .Wxh l Wxe7 20.hxg5t White stands better. One important detail is:

2 1 a

b

e

d

e

f

g

h

17.i.g2 i.e6 18.he7 V!fxe7 19.,bd5

The position can be compared to the later variation B3222. The inclusion of the h2-h4 move should not make much of a difference.

B32 12) 12.h4!?

8 7 6 5 4 3 2 a

b

e

d

e

f

g

h

20 . . . Wxg5 2 1 . .ixb 5 ! axb5 22.l'!h4 c;t>fB 23.lt'ld5 f6 24.l'!hBt c;t> f7 2 5 . lt'l f4! l'!xhB 26.Wd5t c;t>eB 27.WxaB t c;t>f7 2B .Wd5t! c;t>e7 29 .Wb7t c;t>eB 30.Wxb 5 t +- Having collected the last of the available queenside pawns, White is ready to take the hB-rook.

1 5.fxe5 tll d5 16.tll xd5

1 6.l'!xd 5 ? does not work due to 1 6 . . . .ib7! 1 7 . .ixh6 exd5.

a

b

e

d

e

f

g

h

This move serves an important purpose: by defending the g5-bishop, White makes e4-e5 into a serious threat.

12 .. J!! d 8!

This is the best way to pr�pare for the advance of the e-pawn.

1 07

Chapter 6 - 7 . . W c 7 :

1 2 . . . b5? 1 3 .e5 dxe5 1 4.fxe5 ttJd5 ( 1 4 . . . ttJ eB 1 5 .g4±) 1 5 J'!:cd5 ! shows the significance of the h2-h4 move. 1 5 . . . .ixg5 1 6.hxg5 exd5 1 7.ttJxd5 WdB 1 B .We4 Black is completely busted. 12 . . . e5 1 3 .fxe5 dxe5 14 . .ixf6 .ixf6 1 5 . ttJ d5 gives White an obvious initiative. 1 2 . . . l"!eB 8

1 B .h5! ttJ c6 I B . . . h6 1 9 .95! hxg5 20.h6 f6 2 1 . ttJ e4 gives White a dangerous attack. 1 9 .1"!dd l We5 20.Wf2 h6 2 1 . ttJ e4 f6 22.g5! fxg5 22 . . . hxg5 ? 23.h6 is crushing. 23.l"!hfl Wc7 24.ttJd6 l"!dB 2 5 .l"!d2! White has excellent chances. The last move sets up a beautiful trap: 8

7

7

6

6

5

5

4

4

3

3

2

2 a

b

e

d

e

f

g

h

1 3 .e5! dxeS 1 4.fxe5 ttJ d7 1 4 . . . ttJd5 1 5 .l"!xd5 .ixg5 1 6.l"!d4 .ie7 1 7.We4! .id7 1 B . .id3 g6 1 9 .h5 gives White good attacking chances. 1 5 . .ixe7 l"!xe7 1 6.g4 l"!eB 1 7 . .ig2!? A rather interesting pawn sacrifice. White has a big lead in development and can simply keep pushing his pawns. I will present a few possible lines to show the strength of White's compensation.

a

b

e

d

e

f

g

h

25 . . . ttJ e 5 ? 26.ttJc4!! ttJ xc4 27.Wf7t! Wxf7 2B.l"!xdBt mh7 29.l"!xf7+-

13.e5!? Despite Black's last move, White can still proceed with his main plan.

13 ... dxe5 14Jk4! 'lWd7 16.tLlxd5 ad5

1 5.fxe5

The alternative is: 1 6 . . . Wxd5 1 7.b3!? ( 1 7.c3 .ixg5! I B . hxg5 Wd l t 1 9 .Wxd l l"!xd I t 20.mc2 l"!e l 2 1 .l"!ch4 mfB 22.l"!xh7 l"!xe5 23.l"!hBt me7 24.g6 b5 should be survivable for Black.) 8 7 6 5 4

a

b

e

d

e

f

g

1 7 . . . ttJxe5 17 . . . Wxe5 1 B .Wf2 Wc7 1 9 .95t

h

tLld5

3 2

a

b

e

d

e

f

g

h

1 08

6 . . . e6 7.f4

1 7 . . .'IWd 1 t (After 1 7 . . . 1l«d7 1 8 .1l«e3t it will be impossible to get rid of the g5-bishop, and White j ust continues piling up the other pieces on the kingside.) 1 8 .c;t>b2 1l«xe2 1 9 . .txe2 f6 20.exf6 gxf6 2 1 . .te3;!; White has a pleasant endgame. 8

After 2 1 . . . .te6 22.1l«h5 f5 23 . .txf5 .txf5 24.E!:xf5 Black is under pressure and will soon drop a pawn.

22.i.f4 YlYd7 23.g4! YlYxg4 24.YlYd2-+ Black's position looks horrible, and it will be hard to withstand the impending attack on the kingside.

7

B322) 1 1 . h6 .•

6 5 4 3 2 1 a

b

e

d

e

f

g

h

17J:U4! The usefulness of h2-h4 becomes clear, as the bishop on g5 is rather annoying for Black.

17 YlYe6 •••

1 7 . . . h6? 1 8 . .txh6! gxh6 1 9 .1l«h5 20 . .td3 leaves Black defenceless.

1l«e6

1 8.YlYS! YlYxe5 1 8 . . .f6 can be met by 1 9 . .td3 fxe5 ( 1 9 . . . fxg5 20 . .txh7t+-) 20 . .tf5 1l«c6 2 1 ..txh7t c;t>xh7 22.E!:f7 E!:f8 23.E!:xfB .txf8 24.1l«xf8 and White will maintain the pressure with h5-h6.

a

b

e

d

e

f

g

h

By driving the bishop back to h4 before castiing, Black avoids any plans involving h2-h4 from White. At the same time, . . . h6 is a committal move which weakens the kingside a bit.

12.i.h4 Here we should stop to consider the unsound

B322 1) 12 g5? before addressing the saner B3222) 12 0-0. •••

•••

19.�d3 f6 20.YlYdl !? YlYd6 20 . . . fxg5 ? 2 1 ..txh7t! is winning for White, since 2 1 . . .c;t>xh7? 22.1l«h 5 t c;t>g8 23 .1l«f7t leads to a quick mate.

B3221) 12 g5? 13.fxg5 tLld7 •••

This is a thematic idea for this variation, but it does not work here.

2UHfJ.! 2 1 .1l«h5 f5 22.1l«f3 g6 2 3 . .txe7 1l«xe7 i s not so easy.

14.g6!

21 ...h6

14 hl4

1 4 .1l«d2 ltJe5 would be unclear: •••

Chapter 6 - 7 . . . W1c7 Black had better take the piece, since after 1 4 . . . fxg6 1 5 .i.xe7 �xe7 1 6.Wi'd2 b5 1 7.i.e2 he ' is virtually lo st, as his king is exposed and the d6-pawn is terribly weak.

l O9

20.g3 .ifG 2 1 .�e4 Black is completely stuck. A possible conclusion could be:

21 ...�xf7 22 .tc4 �f8 23.Wih5t �g7 24.�xf6+•

15.gxf7t �f8

The black king will soon perish. 8

B3222) 12 0-0 •••

7 6 5 4 3 2

a

b

e

d

e

f

g

h

16.e5! Wid8

1 6 . . . i.e7 1 7.exd6 i.xd6 1 8 .Wixe6 is horrible for Black, and 16 . . . i.d8 1 7.!l:xd6 lLl b8 1 8 .Wi'h5 also leaves him struggling to survive.

17.Wih5 Wig5 1 7 . . . i.e7 1 8 .exd6 i.f6 1 9.!l:g4 lLle5 20.!l:g3 lLlxf7 2 1 .i.c4 followed by !l:fl leaves Black defenceless.

a

b

e

d

e

f

g

h

13.g3!?

Just as in some of the earlier lines, White is preparing e4-e5 . Th e immediate 1 3 .e5 dxe5 1 4.!l:c4 (or 1 4 . fxe5 lLl d5) 14 . . . Wi'd8 1 5 .ixf6 ixf6 1 6. fxe5 ie7 is less convincing.

18.Widl! 1 8 .Wi'xh4 Wi'xh4 1 9 .!l:xh4 lLlxe5 20.i.e2 �xf7 leaves White with some initiative, but the text move is more ambitious and stronger.

18 ... dxe5 After 1 8 . . . lLlxe5 1 9 .93 White restores material equality while keeping a huge advantage.

19J�d6 Wie7 1 9 . . . Wi'g6 20.i.d3 is winning, while after 1 9 . . . Wi'd8 20.lLle4!± White is completely dominating.

1 3 .g4!? Whereas the main line may lead to slower positions, this alternative sees White going for the throat. 1 3 . . . lLl xe4! This daring move is the critical answer. 1 3 . . . b5 might j ust be playable, but it looks dangerous to allow the pawn to advance: 1 4.g5 lLl h7 ( 1 4 . . . hxg5 1 5 .ixg5t) 1 5 .gxh6! ixh4 1 6.hxg7 !l:e8 1 7.!l:d3! b4 ( 1 7 . . . if6 1 8 .!l:g l ! Wic5 1 9 .!l:g2 followed by !l:h3 and e4-e5 gives White a winning attack.) 1 8 .lLl d 1 if6 1 9 .!l:gl Wic5 20.!l:g2--+

1 10

6 . . . e6 7. f4

1 3 . . . e5 1 4.g5! (After 1 4. fxe5 dxe5 1 5 . .txf6 .txf6 1 6. ltJ d5 �dS 1 7J%d3 .te6 l S . .tg2 .th4 the dark-squared bishop blocks everything for a while.) 1 4 . . . exd4 ( l 4 . . . hxg5 1 5 . fxg5 ltJ h7 [ l 5 . . . exd4 1 6.gxf6 dxc3 1 7.:§:gl +-J 1 6.:§:d2 .txg5 1 7 . .txg5 ltJ xg5 l S . ltJ d5±) 1 5 .gxf6 dxc3 1 6.�g2! .txf6 17 . .txf6 g6 l S . .td3� White will soon develop a crushing attack.

main line seems like a safer route to an edge, but players who enjoy this kind of double­ edged position should be able to play it without much worry. 8 7 6 5 4 3 2 1 b

a

e

d

f

e

g

h

1 3 b5 ...

a

b

e

d

e

f

g

h

1 4. ltJ xe4 .txh4 1 5 .g5 f5 ! 1 6. ltJ xd6 1 6. ltJ d2 ? hxg5 1 7. ltJ f3 e5 is bad for White, and 1 6.gxf6 .txf6 1 7.ltJ xf6t :§:xf6 l S .:§:gl .td7 is also fine for Black. 1 6 . . . hxg5 1 7. fxg5 .txg5 l S . .tg2 :§:bS

a

b

e

d

e

f

g

1 3 . . . :§:eS is too slow, and after 14 . .tg2, intending :§:hd 1 and e4-e5 , Black has no good way to continue. 1 3 . . . .td7!? This move isn't as easy to brush off with the moves . . . h6 and .th4 included. 1 4.e5 dxe5 1 5 . fxe5 g5!? 1 5 ... .ta3 16 . .txf6 �b6 1 7.ltJb5! .txb5 l S .�g4 g6 1 9 .bxa3 .tc6t 20.cj;>c 1 .txh l 2 1 .�h4+The main move looks horrid, but it is actually quite testing.

h

1 9 . h4 .tf6 20 .:§:d3 Both sides have their trumps: Black's dark­ squared bishop is formidable, while White has the potential to cause a lot of trouble on the kingside with h4-h5-h6. I will not endeavour to provide an exact conclusion here, as the

a

b

e

d

e

f

g

h

Chapter 6 - 7 . . . Wic7 1 6.l::k4 ! 16 . .ixg5?! hxg5 1 7.h4 100ks deadly, but it is hard to prove much of an advantage after: 17 . . . liJd5! I B .liJxd5 ( l B .Elxd5 exd5 1 9 .1iJxd5 Wic5 just leads to various draws.) IB ... exd5 1 9.hxg5 .if5 ! 20.g4 Wixc2t 2 1 .Wixc2 .ixc2t 22.�xc2 .ixg5 The computer favours White here, but it seems to me that a draw is the most likely outcome. 1 6 . . . Wb6! 16 . . . Wa5 17 . .ixg5! hxg5 I B .h4 liJ d5 1 9 . hxg5 �g7 20.Elh7t! wins in good style. 1 7.exf6 1 7 . .ixg5 hxg5 I B .h4 ( l B.exf6 .ixf6 1 9 . h4 ElfcBoo) I B . . . liJd5 1 9 .hxg5 liJxc3t 20.Elxc3 �g7 2 1 .Elh7t �xh7 22.Wh5t �g7 23 .Wh6t �gB 24.g6 fxg6 2 5 .Wxg6t �hB= White must take the perpetual. 1 7 . . . .ixf6

111

14.e5 dxe5 1 5.fxe5 g5!?

These ... g5 moves often turn out to be surprisingly annoying for White. 1 5 . . . liJ d5 This allows White to win a pawn. We have already touched on the resulting type of position in a couple of the earlier lines. 1 6. liJ xd5 exd5 1 7 ..ixe7 Wxe7 I B . .ig2 8 7 6 5 4 3 2

a

8 7 6 5 4 3 2

a

b

e

d

e

f

g

h

I B .Wh5 ! .ic6 I B . . . .ig7 1 9 . .ixg5 hxg5 20.Wxg5t 1 9 .Wxh6 1 9 . .ixg5 .ixg5 20.liJe4 EladB! leads to insane complications. My main line continues 2 1 .liJxg5 .ixh l 22.a4 e5! and Black manages to survive. 1 9 . . . .ig7 20.Wxg5 f6 2 1 .Wg6 .ixh l 22.a3 White has excellent compensation for the exchange. For instance: 22 .. .f5 23 . .ig5 Elf7 24 . .id3 .id5 2 5 . Elb4 Wg l t 26 . .ic 1 Intending liJ e2-f4.

b

e

d

e

f

g

h

The d5-pawn will fall, leading to a position of two results. Black's best saving chance is to head for a maj or-piece endgame where it might not be easy for White to convert the extra pawn. I B . . . .ie6 1 9 . .ixd5 .ixd5 20.Elxd5 EladB 2 1 .Elhd l Elxd5 22.Elxd5 EleB 22 . . . f6 23.a4!? (23 .b3 is another possibility.) 23 . . . We6 24.Eld6! Wxe5 2 5 .Wxe5 fxe5 26.Elxa6 leads to a more dangerous endgame for Black: 26 . . . Elfl t 27.�a2 bxa4 2B.Elxa4 Elf2 29.Elc4!± 8 7 6 5 4 3 2

a

b

e

d

e

f

g

h

1 12

6 . . . e6 7. f4

23.b3!? f6 24.�e4 fxe5 2 5 . �b2;!; White has returned the extra pawn but kept a bind. He can improve his position by advancing pawns on both flanks, while Black can only remain passive. Later White can start prodding him with various rook moves.

a

b

e

d

e

f

g

h

19.E!:h4! This seems best. 1 9 .Ag2 �xe5! 20.Wxe5 lLlxe5 2 1 .Axb7 E!:a7 22.Ag2 Axg5 is about equal. a

b

e

d

e

f

g

h

16.i.xg5!? 1 6.Ag2 ( l 6.exf6 Axf6 1 7.!!g4 Ab7 1 8 .Ag2 is the same thing) 16 . . . Ab7 1 7.exf6 Axf6 1 8 .E!:g4 Axg2 1 9 .�xg2 Ae7! (but not 1 9 . . .Axc3 2o .Axg5!) leaves a strange mess on the board. Black's position might actually be rather decent.

16 hxg5 17.h4 .ib7

1 9.E!:e4!? is a funny move, but Black defends cold-bloodedly with 1 9 . . . �g7! 20.E!:h4 Axe4 2 1 .�xe4 E!:h8 22.E!:xg4 E!:ac8°o.

19 �xe5 20.E!:d7 �xe2 21 .,ixe2 .tf3!? ••.

2 1 . . .lLle5 might transpose to the main line after 22.E!:xe7 AB 23 .AxB lLlxf3, but White could also take the other bishop with 22.E!:xb7!?, when 22 . . . Axg5 23 .E!:e4 Af6 24.a4 leaves him with some initiative.

•••

1 7 . . . lLl d 5 ? loses to 1 8 .hxg5 lLl xc3t 1 9 .bxc3 threatening �h5 or E!:dh4.

22 .hf3 ttle5 23J�� xe7 ttlxf3 •

18.hxg5 1 8 .Ag2 Axg2 1 9 . hxg5 is met by 1 9 . . . lLl e4! 20 .�xg2 lLl xg5! and a draw is most likely after 2 1 .E!:d7 �xd7 22.�h2 f5 23 .�h8t �f7 24.�h 5 t .

18 ttlg4! •••

Whi te has various tries here, but they all seem to be simplifying to some sort of endgame.

a

b

e

d

e

f

g

h

1 13

Chapter 6 - 7 . . . Vf1c7

Conclusion

24JU4 tl)xg5 25.a4!�

a

b

e

d

e

f

g

h

The knight on g5 is out of the game for the time being, and White's rooks are well placed to create threats on the queenside.

The 7 . . . V!ic7 system has a lot going for it, as it offers Black a sound position with decent counterattacking potential, while avoiding the crazy complications of some other Naj dorf variants. It is no accident that it has been recommended for Black in several repertoire books. 8 .V!ie2!? is a convenient answer for us, as it not only fits in with my recommendation of 7 . . . ltJ bd7 8 .V!ie2 (see Chapters 8 and 9 for more about that) , but also presents a legitimate threat in the event that Black keeps the game in independent territory with 8 . . . ltJ c6 9 .0-0-0. In the early part of the chapter we saw that 9 . . . �e7 and 9 . . . �d7!? should both be met by 1 0.ltJxc6! , preventing Black from exchanging on d4 at a moment that suits him. The main line sees Black force the issue with 9 . . . ltJ xd4 l OJhd4 �e7, and here my detailed investigation revealed 1 1 .�b l !N to be the strongest move, the main point being to threaten e4-e 5 . Having explored both 1 1 . . . 0-0 and 1 1 . . . h6 in detail, I am confident in White's chances, especially considering the surprise value of our 1 1 th move.

8 7 6 5 4

6 ... e6 7.f4

3 2 a

b

c

d

e

f

g

The Polugaevsky Variation Variation Index l .e4 cS 2.tD a d6 3.d4 cxd4 4.tDxd4 tD f6 S.tD c3 a6 6 .igS e6 7.f4 bS B.eS dxeS 9.fxeS V!fc7 1 0.exf6 V!fest 1 1 .ie2 V!fxgS •

•

1 2.0-0

A) 1 2 V!feS B) 12 J�a7 13.V!fd3 �d7 14.tD e4 B l ) 14 V!fg6!? B2) 14 V!feS I S.tD a B2 1 ) I S V!fc7!? B22) I S V!fxb2 1 6.V!fe3 .ib7 17.�ab l ! B22 1 ) 1 7 V!fxc2 B222) 1 7 V!fxa2!

l IS 1 17 1 17 1 19 1 19 1 24 12S 126

••.

••

. ••

•••

..•

•••

•••

•••

A) note to 1 3 . . . E1a7

B) note to 1 4. tiJ e4

B2 1 ) note to 20 . . . �xb2N

8

8

8

7

7

7

6

6

6

5

5

5

4

4

4

3

3

3

2

2

2

a

b

c

d

e

1 5 . tiJ f5 !

f

g

h

g

1 9 . fxg7!N

h

a

b

c

d

e

f

2 1 .'lWg4!N

g

h

h

115

Chapter 7 - The Polugaevsky Variation

l .e4 c5 2.tila d6 3.d4 cxd4 4.tilxd4 ttlf6 5.tilc3 a6 6 ..ig5 e6 7.f4 b5

This prov� cative move, which seemingly ignores the e4-e5 threat, was developed by the great Soviet grandmaster and theoretician Lev Polugaevsky. Such an idea doesn't seem like such a big deal today, but to play so anti­ intuitively in the 1 9 50s was another matter. Even though this variation has come under heavy fire over the years, Black has been hanging on, and we still see the occasional strong grandmaster employ it.

8.e5 dxe5 8 . . . h6? is not a viable alternative: 9 . .ih4 g5 (9 . . . dxe5 1 0.fxe5 transposes to the note to Black's next move.) 1 O .fxg5 til h7 ( 1 0 . . . dxe5 I l .ttJdxb5±) Now any of I l .exd6, 1 1 .'lWg4 or 1 1 .'lWh5 give White a large advantage.

9.fxe5 'lWc7 Again Black does not have much of a choice. 9 . . . h6? 1O . .ih4 g5 is no good due to: l 1 .exf6 gxh4 1 2.'lWf3 1'!a7 1 3 .0-0-0 'lWb6 ( 1 3 . . . 1'!d7 is swiftly refuted by 1 4 . .ixb 5 ! axb5 1 5 .ttJxe6! .) 1 4 . .id3 .ic5 1 5 .ttJb3 .ie3t 1 6.�b l ttJ d7 1 7.1'!he l Black soon got smashed in Cardenas - Cortes, Santiago 1 97 1 . 8

1 0.'lWe2 is a serious alternative, but I regard the text move as clearly the one that asks Black the more difficult questions. 10

.•.

�e5t 1 1 ..ie2 �xg5 12.0-0

We have reached the first major crossroads, as far as Black's options are concerned. A) 12 �e5 deserves close attention, but B) 12 J�'ia7 is the main theoretical path. •••

••

Other moves are worse, and not worthy of detailed analysis. Here are a few brief examples showing how Black's position may fall apart after one or two inaccurate moves: 1 2 . . . .ib7 1 3 . .if3! 'lWe3t 1 4.�h l .ixf3 1 5 .1'!xf3± Stripunsky - Jaracz, Polanica Zdroj 1 99 5 . 1 2 . . . gxf6 1 3 . ttJ e4! 'lWe3t 1 4.�h l .ie7 1 5 .ttJxf6t .ixf6 1 6.1'!xf6 ttJ d7 1 7.1'!f3 White's advantages in mobility and king safety gave him a winning position in Jakovenko - Smeets, Wijk aan Zee 2007. 12 . . . .ic5 1 3 .�h l .ib7 ( 1 3 . . . 'lWe5 transposes to 1 3 . . . .ic5 in the notes to variation A below.) 1 4 . .if3 1'!a7 1 5 . .ixb7 1'!xb7 1 6. ttJ e4 'lWe5 1 7. fxg7 Black was busted in Fezza - De Toma, Campobasso 2007.

7

A) 12 �e5 •.•

6 5 4 3 2 1 a

1 0.exf6

b

e

d

e

f

g

h

A look at the database would suggest that

2 1 a

b

e

d

e

f

g

h

6 . . . e6 7 . f4

1 16

This used to be considered an important theoretical option but, with the help of the correspondence players, it has more or less been put out of business.

1 3.<.i?hl ! Th e most popular option has been 1 3 . tD f3 (the only move analysed by Rizzitano in this position) , but it is not easy to prove anything here. b

a

13 J�a7 ••

1 3 . . . gxf6? is refuted by 1 4 . .ih5 ! +- followed by E:e 1 and a disaster on e6. 1 3 . . . .ic5 ? ! is not much better: 1 4. fxg7 Wfxg7 ( 1 4 . . . E:gB 1 5 . tD dxb5 tD d7 [ 1 5 . . . axb 5 1 6 . .ixb 5 t followed b y Wff3 is crushing.] 16 . .if3 E:bB 1 7.E:e l Wfxg7 I B . tD d6t White won quickly in Kuntermann - Cornacchini, corr. 2007.) Now White can seal Black's fate with some simple tactics:

a

b

e

d

e

f

g

d

e

f

h

g

1 6.tDe5! Wfxf6 ( 1 6 . . . gxf6 1 7.Wfd4! +-) There are a few ways for White to get a nice position here, but perhaps the simplest is 17 . .ih5 g6 I B . tD d5 Wfh4 1 9 . tD f3 WfdB 20.Wfd4 and White wins material.

1 5.�e4 �d8 1 5 . . . Wfxb2 j ust loses a piece after 1 6.tDd6t .ixd6 1 7.Wfxd6, because 17 . . . tD d7 1 B .tDd4! is crushing.

h

1 5 . tD f5 ! exf5 ( 1 5 . . . Wfe5 1 6. tD h6 E:a7 1 7.tDxf7 E:xf7 I B . .ih5 E:hfB 1 9.E:xf7 E:xf7 20.Wff3 gave White a quick win in Rissanen - Lassen, corr. 2002.) 1 6.Wfd5 0-0 1 7.WfxaB White eventually converted his material advantage in Crosa Coll - Coelho, Ilha Solteira 2003.

14.�f3 Wfxf6

e

After 1 4 . . . Wfe3 White kicks the queen around rather easily: 1 5 .E:e l ! Wfh6 ( 1 5 . . . E:d7? 1 6 . .ixb 5 +-; 1 5 . . . Wfb6 1 6.tDd5 Wfc5 1 7.c4!+-)

a

b

e

d

e

f

g

h

16.�e1! With E:d 1 coming next, White i s much too far ahead in development.

16 ... £5 1 6 . . . .ie7? is refuted by 1 7.Wfg3 ! with a deadly double attack, as in Bazantova .:... Lukasova, em ail 20 1 2 .

Chapter 7

-

1 17

The Polugaevsky Variation

1 6 . . . ll'l c6 1 7.!'i:d 1 V1ic7 1 8 . ll'l fg5 V1ie5 1 9 . .ih5 gave White , a crushing attack in Bartley Kelleher, Parsippany 1 998.

This was first played by Polugaevsky in 1 9 59, and remains the main line today.

1 3.'i;Vd3 gd7 17.!'i:dl gd7 18.'i;Vg3! tLlc6 1 8 . . . fxe4 is met by 1 9 .1l'le5, with the deadly threat of 20 . .ih 5t and mate.

1 3 . . . VIie5 is sometimes played, but after 1 4. tLl e4 Black has no real choice but to play 1 4 . . . !'i:d7, transposing to variation B2 below.

14.t£!e4

8

In this position Bl) 14 'i;Vg6!? is a rare but possibly underrated move, but B2) 14 'i;Ve5 has been the overwhelmingly most popular choice. •••

7

•..

6 5

14 . . . VIid5 ?! poses White no problems. 1 5 .c3 tLl c6 1 6. ll'l xc6 V1ixc6 1 7.VIie3 .ib7 1 8 . .if3 V1ic4? ( 1 8 . . . g6 is better, although 1 9 .!'i:fd l gives White a strong initiative.) Now in Sulskis - Stocek, Port Erin 2002, White could have refuted his opponent's play with:

4 3 2 1 a

b

e

d

e

f

g

h

19.tLlfg5

Not the only way to win, but perfectly good enough.

19 .. J�d5 20 .ih5t g6 21 .gxd5 22.'i;Vc3+•

exd5

Schenning - Czudek, corr. 20 1 1 .

B) 12 ga7 •••

a

b

e

d

e

f

g

h

1 9 . fxg7!N .ixg7 20.ll'lc5!+- Winning material.

Bl) 14 'i;Vg6!? •••

This move has hardly ever been played, but it certainly deserves more attention. Black's idea is to play . . . .ib7 when the pressure against the knight may become awkward, considering that White would prefer to avoid a queen exchange.

a

b

e

d

e

f

g

h

118

6 . . . e6 7.f4

17 ....be4

1 7 . . . �xd4? ? loses to 1 8 .�xd4 �xf4 1 9 .fxg7.

1 7 . . . g6 is not so terrible, but 1 8 .lLlf3;!; leaves White with a stable advantage. Black is behind in development and his king faces an uncertain future.

18.Yfxe4

a

b

e

d

e

f

g

h

1 5.Yfe3

I also considered 1 5 .�ae l N, but found the position after 1 5 . . . .tb7 1 6 . .tf3 lLl c6 1 7.c3 �g8 to be unclear.

1 5 ...J.b7 16.J.d3 1 6. lLl f5 ! ? .txe4! 1 7. lLl h4 �h6 1 8 .�xe4 .tc5 t 1 9 .'kt>h l 0-0 20.fxg7 �fd8°o is another murky line. The text move seems more logical, turning the black queen into a target.

a

b

e

d

e

f

g

h

18 ...J.c5N This seems like Black's best attempt to improve on 1 8 . . . .td6 1 9 .fxg7 �xg7 20.lLlf5 .tc5t 2 1 .'kt>h l �xb2 22.�ffl !± as played in Jarecki - Gagliardi, corr. 20 1 1 . 1 8 . . . gxf6 1 9 . 1Ll f5 �g6 20.'kt>h 1 .tc5 2 1 .�afl ± looks horrible for Black.

19.fxg7 gg8 Bringing in another piece.

a

b

e

d

e

f

g

17JU4!

h

A typical way to avoid the queen exchange, while bringing more pieces into the attack. 1 7.�xh6?! gxh6 seems fine for Black.

1 9 . . . �xg7?! is worse: 20.c3 e5 (20 . . . �g8 2 1 .g3±) 2 1 .�f3 .txd4t (2 1 . . .';!;>d8 22.�f5!±) 22.cxd4 �xd4 23 .�e2± Black's king is too vulnerable.

20.c3 gxg7 21.gafl

2 1 .'kt>h l can be met by 2 1 . . . .t'xd4 22.cxd4 lLl c6! 23.�afl lLlxd4 24.�a8t �d8 25 .�xa6 �g5 and Black is doing all right.

Chapter 1

-

The Polugaevsky Variatio n

The text move defends the rook on f4, and hence avoids the . . . lLl c6 idea, which means Black still h�s to find a way to develop the knight. A possible try is:

1 19

27.a4!? White's chances are preferable in the middlegame, as he can continue to generate threats against the enemy king, although Black is still not out of the game by a long way.

21 ..J�d5 B2) 14 Wle5 •••

8 7 6 5 4 3 2 a

b

e

d

e

f

g

h

22.g 1fl! Removing the pin on the d4-knight.

1 a

b

e

d

e

f

g

h

22.cj,Jh 1 l'!hS! 23.lLlf3 l'!g3 ! would have turned the tables.

If the database is anything to go by, almost all of your games against the Polugaevsky are likely to reach this position if you follow my recommendations.

22 �c6!

1 5.�f3

..•

22 . . . lLl dl 23 . .ic2 lLleS 24 . .ib3± 22 . . . .ixd4 23.cxd4 lLl c6 runs into 24.We l ! followed by Wc3 , and the inevitable queen entry will cause a lot of trouble for Black.

l S .c3 has been played, but I prefer the more energetic way of dealing with the pin. Black has two sensible replies: B21) 1 5 Wlc7!? and the more popular B22) 1 5 ... Wlxb2. •••

B21) 1 5 ...Wlc7!? 23.tilf5! 23.lLlxc6 fS ! will lead to a perpetual after: 24.We2 Wxf4 2S .Wxe6t �fB 26.WcB t �f7 27.lLld4 Wc l t 2B.l'!fl We3t 29.�h 1 Wxd3 30 .We6t=

23 Lf2t 24.gxfl Wlc1 t 25.gfl Wlg5 26.tilxg7t Wlxg7 ..•

Here Black seems to have decent drawing chances in the event that White grabs the hl­ pawn, so I prefer:

This calm move simply aims at keeping Black's position together, without resorting to pawn­ grabbing. This move attracted some followers after Nakamura used it in 20 1 2, so it is one of the key battlegrounds in the debate over the Polugaevsky's resurgence.

16.Wle3 .ib7 1 6 . . . g6? was once played by Polugaevsky himself, but it is practically refuted by the

1 20

6 . . . e6 7 . f4

following strong plan: 1 7.c4! b4 ( I 7 . . . bxc4 1 8 .l'!ac l is horrible for Black.) 1 8 .id l ! (This is much more convincing than 1 8 .<;t>h l , which led to an eventual draw in Novopashin Polugaevsky, Vilnius 1 960.) 1 8 . . . ib7 1 9 .ia4 ic6 20.ixc6 li'l xc6 White's bishop manoeuvre has enabled him to swap off a key defender, and he soon makes intoads along the d-file.

I found this move myself, but it is not quite a novelty, as it was played in an internet game a couple of years ago. 1 7. li'l fg5 h6 led to complications in Wang Hao - Nakamura, Biel 20 1 2. White eventually won that game, but Black seems to be holding his own at this point. 1 7.c4!? is sensible, but after 17 . . . bxc4! it is not easy to find an advantage for White. For instance, after 1 8 .l'!ac l ixe4 1 9 .fxg7 ixg7 20.'lMfxe4 0-0 Black's position was safe enough in Erdelyi - Galyas, Kecskemet 20 1 2 . 1 7.<;t>h l ! ? i s similar to the main line, but there is a key difference after: 1 7 . . . ixe4!N 1 8 .fxg7 ixg7 1 9 .'lMfxe4 ixb2

b

a

c

d

e

f

h

g

2 1 .l'!ad l h5 22.l'!xd7 Wxd7 (22 . . .1.Mfxd7 23 .'lMfb6+-) 23.c5 Wc8 24.'lMfd3 White soon broke through in Tenev - Scacco, Internet 2004. The text move is much stronger, and introduces the potential threat of . . . ixe4 followed by . . . gxf6. At first, White seems to have a few tempting options, but a deeper inspection reveals that things are not so simple.

8 7 6 5 4 3 2

a

e

d

e

f

g

h

Here White would like to play 20.l'!ad 1 , but Black has a safe answer in 20 . . . 'lMfc6!. By aiming for a similar position with the moves a2-a4 and . . . b4, we can eliminate this possibility as the b4-pawn will be left hanging. White can instead play 20 .l'!ab l ic3 2 1 .a4 b4, reaching a position that will be discussed via transposition in the note to White's 2 1 st move in the main line. For a quick verdict, I will say it is not clear if White is better.

8 7 6 5 4 3 2 1 a

17.a4!

b

b

c

d

e

f

g

h

17 ... b4 1 7 . . . bxa4?! 1 8 .l'!xa4 just seems to be inviting trouble.

121

Chapter 7 - The Polugaevsky Variation 1 7 . . . .ixe4 1 8 .Wxe4 gxf6 ( 1 8 . . . Wb7 1 9.Wf4 gxf6 20.axb5 axb 5 2 1 .Wxf6 E:g8 22 .Wh4 leaves White with an excellent initiative without even being a pawn down.) 1 9 .axb5 Wb6t 20.�h 1 axb5

b

a

e

d

e

f

h

g

2 1 .c3! Preparing It'l d4. 2 1 . . . .ic5 (2 1 . . . .ig7 22.Wb4! +-; 2 1 . . . .ie7 22.lt'ld4±) 22 .Wg4!± Once again White has a strong initiative, the immediate threat being Wg7.

1 8 . . . g6 gives White extremely easy play due to the suffocating f6-pawn. 1 9 .E:ad 1 h5 20.lt'lfg5 .ih6 2 1 . lt'l c 5 ± 1 8 . . . .i d 5 seems rather awkward. White has a few pleasant options, for instance: 1 9 .c3!? bxc3 20.lt'lxc3 gxf6 2 1 . lt'l xd5 E:xd5 22 .E:ac l .ic5 (22 . . . Wd7 23.E:fd 1 --+) 23 .Wh6 With an ongoing initiative. 1 8 . . . h6 1 9.E:ad 1 .ixe4 ( 1 9 . . . g6 20.E:fe 1 ±) 20.Wxe4! (20. fxg7 .ixg7 2 1 .Wxe4 Wb7! allows Black to castle next, as It'l g5 has been prevented.) 20 ... gxf6 2 1 .E:xd7 It'l xd7 22 . .ixa6;!; White's life will be much easier. Black's king faces an uncertain future, and the passed a-pawn gives him something else to worry about.

8 7 6 5 4 3

a

2

b

e

d

e

f

g

h

19.fxg7 hg7 20.'ilYxe4 hb2N

1 a

b

e

d

e

f

g

h

18.<;thl! Angling for a n improved version o f 1 7. � h I , as discussed earlier.

18 ....ixe4

This is the most logical reply to the king move. 1 8 . . . Wxc2 1 9. 1t'l fg5 ! followed by E:ac l will give White a crushing initiative.

This seems critical. 20 . . . 0-0? ? is a blunder due to 2 1 .lt'lg5. 20 . . . Wb7 2 1 .Wg4!N Improving over 2 1 . W e3 as played in Gonzaga Grego - Morcin, Internet 20 1 2. 2 1 . . . .ixb2 2 1 . . . 0-0 22.lt'lg5! gives White a strong attack. A brief sample line is 22 . . . h6 23.lt'lxe6 fxe6 24.Wxe6t �h8 2 5 .E:xffit .ixffi 26 . .ic4! and wins. 22.E:ab 1 .ic3

1 22

6 . . . e6 7 . f4 continues: 2 1 . . .ic3 22.ltJg5 h5 (22 .. .'IWe5!? leads more or less by force to an endgame: 23.�xb4 h6 24.1lMxe5 ixe5 25.ltJxf7 �xf7 26.ih5 0-0 27.ixf7t �xf7 28 .�xf7 Wxf7 29 .�b7t mf6 30 .g3 White keeps some winning chances.) Now White has two serious options:

8 7 6 5 4 3 2 b

a

e

d

e

f

h

g

23.ltJg5! �g8 23 . . . ltJ c6 24.ltJe4± 24.1lMh4 Renewing the threat of ltJ e4. 24 . . . ltJ c6 24 . . . �d2 2 5 . ltJ e4 �xg2 (25 . . . �xe2 26.ltJ d6t+-) 26.mxg2 f5 27.1lMh5 t md8 28 .�f4! fxe4 29.1lMa5 t me8 30.mh3+­ White's rooks are ready to go on a rampage. 2 5 . ltJ e4 ltJ e7 26.if3 ltJ f5 27.ltJd6t �xd6 28 .1lMxh7 �h8 29 .1lMxh8t ixh8 30.ixb7± Black can resist for a while, but White's extra exchange should be the decisive factor.

a) 23.ltJxf7!? plunges the game into insane­ looking complications, where best play leads to another difficult-to-assess endgame: 23 . . . �xf7 24.1lMg6 ltJ c6 25 .ixh5 (25 .1lMxe6t �e7 26.1lMg6t md7!oo) 25 . . . ltJe5 26.1lMxe6t mf8 27.ixf7 8 7 6 5 4 3 2

a

b

e

d

e

f

g

h

27 . . . �xh2t! 28.mgl id4t 29.mxh2 ltJ f3t 30.mh3 ltJg5t 3 l .mg4 ltJxe6 32 .ixe6t We8 White can keep playing without much risk, but Black has excellent chances of delivering a perpetual check at some point.

8 7 6 5

b) 23.�f3 1lMe5! is a key resource, forcing White into some concrete variations, and preventing a long-lasting initiative:

4 3 2

8

1

7 a

b

e

d

e

f

g

h

2 1 .gadl!

6 5

As explained earlier, this is the moment where we see the advantage of 1 7.a4! over 1 7.mh l .

4

2 1 .�ab 1 is the computer's top choice, and also warrants consideration. The critical line

2

3

a

b

e

d

e

f

g

h

Chapter 7 - The Pol ugaevsky Variation 24J:hb4 Wxg5 25 J:hbBt �e7 26J!xhB .ixhB 27.Wb4t �eB 2B . .ixa6 l'!d l t� It will be hard for White to �onvert his extra pawn, as Black's pieces are active and he can aim for counterplay against White's king. These lines can of course be explored further, but overall I consider the main line a better bet.

1 23

23.lLld4! White is trying to provoke the exchange of Black's dark-squared bishop, which will lead to the collapse of his position. 23 . . . Wd5 Or 23 . . . Wb6 24.a5! intending 24 . . . Wxa5 2 5 .l'!xf7! winning. 24.lLl b3 Wc6 2 5 .Wxb4t Black's king remains stuck in the centre.

22J�xdl �c5 22 . . . l'!gB 23 .Wxb4 .if6 24.a5± leaves Black struggling to get his king to safety. 8 7 6 5 b

a

e

d

f

e

h

g

21...l'!xdl This seems like the best chance. Black is not helped by: 2 1 . . . Wb7 22 .Wg4! Preventing castling. 22 . . . h6 After 22 . . . h5 23 .Wf4 h4 24.h3± the king will be stuck permanently in the centre. 22 . . . l'!xd l 23.l'!xd l lLl d7 24.lLlg5! leaves Black without a good defence: 24 . . . l'!gB 2 5 .Wf4 .if6 (25 . . . lLle5 26.lLle4+-; 25 . . . lLl f6 26 . .ixa6 Wxa6 27.WbBt �e7 2B .Wc7t+-) 26.l'!xd7 Wxd7 27.Wxf6±

4 3 2 1 b

a

e

d

f

e

h

g

23.�f4 The most straightforward way to get an edge. A more sophisticated try could be: 23.h4!? .if6 (23 . . . h6 is met by 24 .lLld4!±, intending 24 . . . 0-0 25 . .id3 when Black will not be able to defend h7 with . . . Wh5 .) 24.Wf4 �e7 2 5 . lLl g5 .ixg5 26.hxg5 lLl d7 White has a few different options, perhaps the most critical being:

8 7

8

6

7

5

6

4

5

3

4

2

3 2 a

b

e

d

e

f

g

h a

b

e

d

e

f

g

h

6 . . . e6 7.f4

1 24

27.j,xa6!? h6 2S .'Mfh4 'Mfxg5 29 .'Mfxb4t It'l c5 30 .j,e2 !'!dS 3 1 .a5 White keeps the better chances, although I am not sure if this endgame is likely to be winning. I will refrain from going any deeper, as the main move seems like a simpler way to conclude the theoretical discussion.

23 0-0 •.•

After 23 . . . lt'l c6 24.j,xa6;!; the passed a-pawn is an important asset, while the issue of king safety remains a concern for Black.

B22) 1 5 �xb2 .•.

8 7 6 5 4 3 2 1

24 .td3 .tg7 •

Black had better avoid 24 . . . lt'l c6?! 2 5 .'Mfh6!.

25.�e4 �h5 26.�xb4 Black will have to bring the knight to assist with the defence of the weakened kingside, but this will leave the a6-pawn a sitting duck.

a

b

e

d

e

f

g

h

Grabbing the b2-pawn is the main line. Although this line is obviously dangerous for Black, you also need to know your stuff with White, since these forcing lines can be hard to play over the board. Many strong grandmasters have stumbled while trying to prove an advantage - including me!

16.�e3 .tb7 17.!'!abl!

a

b

e

d

e

f

g

h

26 ... tlk6 27.�e4 til e5 Or 27 . . . lt'l e7 2S .j,xa6 It'l d5 29.!,!fl ;!; and White keeps an extra pawn, as well as chances of a kingside initiative.

28.ha6 �xf3 29.�xf3 �xf3 30.gxf3� Black will have to fight for a draw in a one­ sided endgame.

1 7.a4 used to b e considered the main line, especially after Leko won a spectacular game with White against Ghaem Maghami, who played 1 7 . . . b4. For a while that game was considered the refutation of Black's set-up, but then a stronger defensive plan was identified. The critical continuation is 17 . . . 'Mfb4! 1 S .c4 ( l S . lt'l fg5 j,xe4 1 9 .1t'lxe4 'Mfd4 is another key point) 1 S . . . j,xe4 1 9 .'Mfxe4 gxf6 as in Luther - Bromberger, Austria 2006. The position remains messy and there is scope for further investigation, but my analysis indicates that the text move is stronger, a verdict supported by the practical results so far. Black must decide between and B222) 17 �xa2!. ••.

B22 1} 17 ...�xc2

Chapter 7 - The Polugaevsky Variation

1 25

1 8 . . . g6? was played i n my first game i n this line: 1 9 .1'!bc1 Wa4 (Other moves do not change the result: 1 9 . . .Wxa2 20 .Wa7! Black's position collapses; 1 9 . . . Wb2 20.4Jxf7! 1'!xf7 2 1 . 4J d6t .txd6 22.Wxe6t+-)

B22 1) 17 Y;Vxc2 •••

8 7 6

8

5

7

4

6

3

5

2

4 3

1 a

b

e

d

e

f

g

h

This appears critical and has been the more popular choice, but it is refuted spectacularly in the style of the aforementioned Leko Ghaem Maghami game.

2 b

a

e

d

e

f

h

g

20.4Jxe6! fxe6 2 1 . f7t+- Negi - Hermansson, Malmo 2007.

18.c!l)fg5 Now Black must be careful to avoid an instant loss. 8 7 6 5 4 3

a

b

e

d

e

f

g

h

19.a4!

2

1 a

b

e

d

e

f

g

h

18,..Y;Vc7 1 8 . . . .txe4? 1 9.4Jxe4 does nothing to help Black. 1 8 . . . h6? 1 9 .1'!bc 1 Wa4 20.4Jxe6 gave White a quick win in Albers - Euler, corr. 1 97 1 .

Today this move seems obvious with the help of modern engines, but when I analysed it for the first time with Nigel back in 2006, it was quite an impressive discovery. All other moves let Black off the hook. For instance, Rizzitano points out that 1 9 . fxg7 .txg7 20.4Jf6t .txf6 2 1 .1'!xf6 1'!d5 leads to equality, but does not consider the text move.

1 26

6 . . . e6 7.f4

1 9 . . . bxa4 meets with a spectacular refutation: 20.fxg7 .ixg7 2 1 .E:xf7! E:xf7 22.E:xb7!+1 9 . . . b4 transposes to the aforementioned Leko game, which continued: 20.E:xb4! Le4 (Mter 20 . . . lLl c6 2 1 .E:c4 Black's development woes continue: 2 1 . . .V;Ve5 22.lLlf3 V;Va5 23.fxg7 .ixg7 24.'kt> h l lLl e7 2 5 . lLl c 5 lLl fS 26.V;Va3 lLl d6 27.E:cc 1 +- Bakhmatov - Reyes Maldonado, corr. 20 1 3 .) 2 1 .lLlxe4 (2 1 . fxg7 .ixg7 22.V;Vxe4± is also strong.)

2 1 .axb5! hxg5 2 1 . . .axb5 22.lLlf3 V;Va7 sees Black get demolished after 23.V;Vxa7 E:xa7 24 . .ixb5 .ixb5 2 5 .E:xb 5 . 22.bxc6 lLl xc6 Mter 22 . . . V;Vxh2t 23.'kt>f2 Black quickly runs out of checks . 23.h3 V;Va7 24.E:b6+Once again Black is routed.

8 7 6 5 4 3 2

a

b

e

d

e

f

g

h

2 1 . . . .ixb4 22. fxg7 E:g8 (22 . . . .ic3 23.E:c 1 ±) 23.lLl f6t 'kt>d8 24.lLlxg8 .ic5 2 5 . lLl f6! .ixe3t 26.'kt>h l was a spectacular win in Leko Ghaem Maghami, Yerevan 200 1 . A final unsuccessful try for Black is: 1 9 . . . h6 20.E:fc 1 ! .ic6N 20 . . . V;Ve5 2 1 ..ixb5 ! axb 5 22.E:xb5 Gagliardi - Lecha Gonzalez, corr. 20 1 1 . 20 . . . lLl c6 2 1 .axb5 hxg5 (2 1 . . . axb5 22.lLlf3+-) 22.bxc6 .ixc6 (22 . . . V;Vxh2t 23.'kt>f2+-) 23 .g3 +- 'yorkman' - 'JTB' , Internet 2009 .

a

b

e

d

e

f

g

20.axbS as 2 U�bc1! Indirectly defending the highlighted in the next note.

e4-knight,

2 1 . h6 22.tC!f3! ••

Using the same idea again.

22 �xe4 23.E:c8t! E:d8 24.E:xd8t 'kt>xd8 2S.�b6t
7

B222) 17 �xa2!

6

.•.

5 4 3 2

a

b

e

d

e

f

g

h

as

2 1 . . .g6 22.lLlf3! V;Vxe4 23.E:c8t! reveals the point behind White's 2 1 st move. 23 . . . E:d8 24.E:xd8t 'kt>xd8 2 5 .V;Vb6t 'kt>e8 26.V;Vc7!+-

.••

8

h

1 27

Chapter 7 - The Polugaevsky Variation

8 7 6 5 4 3 2 1 a

b

c

d

e

f

g

h

Although this position appears extremely dangerous for Black, it is not trivial to demonstrate a concrete advantage for White.

b

a

e

d

f

e

g

h

18 ...,he4 19.fxg7

1 9 .Wxe4 gxf6 20.cxb5 has never been played, and indeed after 20 . . . �g7! 2 1 .bxa6 0-0 Black is safe.

18.c4 This is the theoretically-travelled path, and probably the strongest move. I B .ltJfg5N looks like another tempting option, but I am not completely convinced by it: I B . . .�xe4 1 9 .1Llxe4 Wd5 20.�bd l We5 2 1 .Wb6 g6!? (After 2 1 . . .Wc7 22.Wxc7 �xc7 23. fxg7 �xg7 24.lLl d6t i>f8 2 5 .�h5 White has a bit of pressure, although I think Black should be okay.) 22.�xd7 lLlxd7 23 .Wc6 �c5t 24.'it> h l 0-0 25.�d l lLlxf6 26.lLlxf6t Wxf6 27.Wxc5 �dB 2B.�fl Wd4 Black has a stable position with three pawns against a bishop. White can certainly try to press for a while, but Black should be able to hold it. I B .Wf4!?N is another completely new idea. I discovered some amazing tactical points further down the line, but it demands a great deal of accuracy from White. Moreover, there are some ways for Black to simplify to a worse but possibly defensible endgame. Although it would have been nice to include this move, the reality is that I B .c4 is a more straightforward route to an advantage.

19 ... .ixg7 20.Yfxe4 gd6!? An interesting defensive idea. Black's plan often revolves around the potential opposite­ coloured-bishop endgames, which tend not to j ustify the computer's excitement for White. 20 . . . h6 has been tested, but it is hard to imagine any human player wishing to play this way. Play continues: 2 1 . lLl e5 �xe5 22 .Wxe5 0-0 23.�a l Wd2 24.�f4 �fdB 2 5 .�afl lLl c6 26.�g4 t i>fB Black is narrowly avoiding being mated, but he will remain under pressure for some time. 8 7 6 5 4 3 2

a

b

e

d

e

f

g

h

27.Wg7t i>e7 2B .�e4 Wg5 29 .Wxf7t i>d6

6 . . . e6 7 . f4

1 28

30.Wff2 �c7 3 1 .1:'!xe6 (3 1 .cxbS axb S 32.1:'!xe6 lLl d4 33 .1:'!e4 is similar.) 3 1 . . .lLl d4 32.1:'!xa6 lLl xe2t 33 .Wfxe2 Wfcst 34.� h l Wfxc4 3 S .WfeS t � b 7 36.1:'!aa 1 I found two computer games that reached this position, one of which was won by White, with the other ending in a draw. For a human player, it is obvious that Black will have to suffer for a long time due to his exposed king.

22 . . . Wfd2 22 . . . 0-0? 23.lLlxf7! 1:'!xf7 24.Wfxe6+23 .Wfxd2 1:'!xd2 24 . .ihS .id4t 2 S .'j{ h 1 8 7 6 5 4 3 2

a

b

a

e

d

f

e

g

h

21 .cxb5! This seems like the simplest route to an advantage. The alternative is: 2 1 . lLl gS 1:'!d4 22.Wfe3 ! ? In m y game I made a tactical oversight with 22.Wfb7? 0-0 23 . .id3 , and after 23 . . . Wfd2! Black was suddenly winning in Negi Shomoev, Moscow 20 1 2. 8

e

d

e

f

g

h

2 S . . . 0-0!N A strong human improvement. After 2S . . . 1:'!f2 26.1:'!xf2 .ixf2 27.1:'!f1 ! l!;>e7! ? 28 .1:'!xf2 f6 29.lLle4 f5 30.lLld2 White went on to convert his advantage in Caire - Reyes Maldonado, corr. 20 1 3 . 26 . .ixf7t �h8 27.cxbS 27.lLlxe6 1:'!c8 28.lLlxd4 1:'!xd4 29.cxbS axbS 30.1:'!xbS lLl d7 reaches a similar endgame. 8 7 6 5 4 3 2

a

7

b

b

e

d

e

f

g

h

27 . . . 1:'!f2 28.1:'!xf2 .ixf2 29 . .ixe6 axbS 30 .1:'!xbS .ie3 3 1 .lLl f3 lLl c6 White can press with absolutely no risk, although ultimately the position is probably a draw.

6 5 4 3 2

a

b

e

d

e

f

g

h

Chapter 7 - The Pol ugaevsky Variation

1 29

33 ... e4 33.�g5 �f6 34.h4 h6 35.�h3! a5 36.E:c8 a4 37.�h2 1-0 M . Rubinstein - Neacsu, Internet 2007.

Conclusion

a

b

e

d

e

f

g

h

21...0-0 22.b6!

22.�h l is met by 22 . . . h6 when Black's king is safe for now.

22 ... E:c6 23.E:fel!? �a5 24.b7 The passed pawn is a huge asset which clearly outweighs Black's extra pawn. In the following Internet game White skilfully converted his advantage by combining threats on both flanks.

24 ... �c5t 25.�hl £5 26.�a4 �h8 27.E:edl e5 28.�h4 E:h6 29.E:dcl �d6 30.�g5 E:g6 31 .�e3 �e6

a

b

e

d

e

f

g

h

32.E:c7! The defence is almost at breaking point. The main threat is ic4.

For more than half a century now, the Polugaevsky Variation has defied all White's attempts at refutation. This is quite remarkable considering that, in the position after 8.e5 dxe5 9 . fxe5 �c7 1 0.exf6 �e5 t 1 1 .ie2 �xg5 1 2 . 0-0, White has developed three minor pieces and castled, while Black has only developed his queen and moved some pawns. From the above position, 1 2 . . . �e5 has been more or less buried by the correspondence players, but the main line of 1 2 .. J::\ a7 is harder to crack. After 1 3 .�d3 E:d7 1 4. lLl e4, I believe that 14 . . . �g6!? may have been unfairly neglected. I found some promising ideas for White, but one could certainly make the argument that Black is at no more of a disadvantage here than in the main lines. The main theoretical path continues 1 4 . . . �e5 1 5 .lLl f3 �xb2 1 6.�e3 ib7 1 7.E:ab l ! , when 1 7 . . . �xa2! is the acid test. The position is complicated, but White can obtain excellent chances by using the last internet game as a model.

8 7 6 5 4

Gelfand Variation

3

�j� ��!!. �� ..

,.=" m..", ,,,,,,,

F'/,//Nh·",JN/////"�,/.,/N"///,·/""/==/,,,d "'",f·""··/"."F� ,=//�'

2 a

b

c

d

e

f

g

Introduction Variation Index l .e4 c5 2.� f3 d6 3.d4 cxd4 4.�xd4 � f6 5.� c3 a6 6.i.g5 e6 7.£4 � bd7 8JWe2 Ylfc7 9.0-0-0

A) 9 ...i.e7 B) 9 b5 1 0.a3 B l ) 1 0...i.b7 B2) 10...i.e7 1 l .g4 B2 1 ) 1 1 ...i.b7 B2 1 1 ) 1 2.f5!? B2 1 2) 12.i.h4

1 32 1 33 134 1 36 1 37 137 140 141 142 143 145

•••

B2 1 2 1 ) 12 ...� c5 B2 1 22) 1 2 ...� b6 B2 1 23) 1 2 ...g6!? B22) 1 1 ...h6

A) after 1 5 , . J!bS

82 1 22) after 1 4 . . . :i"1cS

822) after I S . li:l g4 . .

8 7 6 5 4

, ...... /a"O·/... • . . . .

3 2 a

b

e

d

e

f

1 6.�e3! N

g

h

a

b

e

d

e

f

1 5 ,\t>b l !N

g

h

a

b

c

d

e

1 9 .96!?N

f

g

h

h

Chapter 8 - I ntroduction

1.e4 cS 2.�f3 d6 3.d4 cx:d4 4.c!LJxd4 �f6 S.c!LJc3 a6 6.�gS e6 7.£4 c!LJbd7

131

8 7 6 5 4 3 2 1 a

a

b

e

d

e

f

g

h

This is one of the most natural ways to develop. It is often referred to as the Gelfand Variation due to the Israeli icon's colossal contribution in the main lines after 8 .Wf3 . Even though 6 . . . liJ bd7 has recently taken over as the most fashionable alternative to the Poisoned Pawn, no particular problems have been found for Black in the main theoretical lines of the present variation. Its only drawback is a slight lack of flexibility (especially when compared to 6 . . . liJ bd7) , and the considerable volume of theory that Black needs to remember.

8.We2 8.Wf3 is a big option of course, but I believe the text move to be at least as dangerous. The queen supports a possible e4-e5 break, while leaving the long diagonal free for the light­ squared bishop, for instance after g2-g4 and ig2 . Before we get into the business of discussing different moves and plans, I should mention that this variation has a lot of different move­ order possibilities. To help make sense of it all, I would like to draw your attention to the following position, which often arises after Black's 1 2th move.

b

c

d

e

f

g

h

Both sides have arranged their forces logically but, along the way, certain move orders will offer independent possibilities for both sides. The timing of the . . . h6 move may also prove significant. Another peculiar point is that while . . . h6 provokes ih4, retreating the bishop to h4 may also provoke . . . h6 due to the threat of g4-g5 , which only increases the number of possible transpositions. The above position will serve as the starting point for the next chapter. Before then, we will examine the various ways in which the game may veer off in a different direction.

8 ...%Yc7 8 . . . h6 9 .ih4 gives Black nothing better than converting to one of the usual paths. 9 . . . e5 has been played in several games, but after l O.O-O-O ( l 0 . fxe5 should also be pretty good) the position is the same as variation C l of Chapter 3 , except that Black i s missing the . . . b5 move. 8 . . . ie7 does not have much independent value. 9 .0-0-0 h6 (9 . . . Wc7 reaches variation A below.) 1 0.ih4 Wa5 Black is trying to prepare . . . g5 , but the queen may turn out to be misplaced. I l . liJ f3! Preventing . . . g5 . 1 1 . . .e5 1 2.�b l ! Now the queen looks awful on a5 . 1 2 . . . exf4 1 3 .liJd5 liJxd5 1 4.exd5 Wd8 1 5 .l"1e1 g5 1 6.if2� Kurnosov - Vorobiov, Moscow 20 1 2.

1 32

The Gelfand Variation

10 ... h6

9.0-0-0

l O . . . bS 1 1 .a3 takes us to variation B2.

10 ... 0-0 is not great, and 1 1 .�g2! makes it hard for Black to develop at all. Perhaps the best attempt is 1 1 . . . lD b6, but this is far from an ideal solution. Shirov showed an easy way to deal with Black's plan: 1 2.h4! lD c4

a

b

e

d

e

f

g

h

We will start by considering A) 9 ....ie7, before moving on to the main move B) 9 ... h5.

A) 9 ....ie7 10.g4

b

a

White proceeds with his normal plan. Now Black has nothing better than converting into one of the lines discussed under 9 . . . b S . 1 O . lD f3 ! ? used t o b e considered quite strong, but things aren't so easy after 1 0 . . . h6 1 1 .�h4 e S . There is a lot more to explore here, but White has absolutely no reason to go down this path.

e

d

e

f

h

g

1 3 .:gh3! Since . . . bS isn't possible, Black does not have many useful moves, while White is ready for b2-b3 . 1 3 . . . �h8 1 4.b3 ltJ aS l S .eS± Shirov - Fercec, Kallithea 2008.

1 1 ..ih4 g5!? If Black is determined to avoid transposing to normal paths with . . . bS, this is the only serious option.

8 7 6 5 4 3 2 1 a

b

e

d

e

f

g

h

a

b

12.fxg5 hxg5

e

d

e

f

g

h

1 33

Chapter 8 - Introductio n 1 2 . . . tLlh7 1 3 . .ig3! hxg5 1 4. tLl f5 ! tLle5 1 5 .tLlxe7 Wlxe7 1 6.Wlf2 is excellent for White.

13.,hg5 b5 14.a3 Black has a bit of compensation for the pawn, but White's play is quicker.

while they may have posed some troubles initially, most of them have been successfully repelled. By nudging the a-pawn forwards, White shows that he is in no hurry to break through in the centre. Instead he tries to limit Black's counterplay before preparing the g4-g5 advance.

14 ... tt:Je5 1 5.l:�gl :ab8 Over the next few moves Black usually develops both of his bishops, with . . . h6 thrown in somewhere. We will examine Bl) 10 ....tb7 and B2) 10 ... i.e7 in turn. The first move has been more popular, but I believe the second is more accurate in terms of limiting White's options, as well as maintaining maximum flexibility for Black.

8 7 6 5 4 3 2 1 a

b

e

d

e

f

g

h

16.i.e3!N

By threatening g5-g6, White improves on 1 6.'�b 1 b4°o as played in Yilmaz - Aveskulov, Kharkov 20 1 1 .

16 ...i.b7 1 6 . . . b4?! is unsatisfactory due to 1 7.axb4 Eixb4 1 8 .g5 tLl fd7 1 9 .96!, but now, with the b-file blocked, Black's queenside counterplay is at a standstill.

This would not be a great time for Black to play 1 0 . . . h6?!. White can either play 1 1 . .ih4, staying on course for the target position shown on page 1 3 1 , or try to exploit Black's move order with 1 1 ..ixf6!? tLl xf6 1 2.g4, with favourable prospects. 1 0 . . . Eib8 has been played a few times, but is not something White needs to worry about. 1 1 .g4 ( l 1 . tLl d 5 ! ? is interesting but unnecessary.) 1 1 . . .b4 Rather risky. ( l 1 . . .i.e7 transposes to 1 1 . . .Eib8 in the notes to variation B2 - see page 1 36.) Now in Chirila - Nastase, Baile Olanesti 20 1 0, White could have obtained an improved version of the sacrifice with:

17.i.f4 White is definitely better.

B) 9 b5 •••

This is the most natural way to continue there is no need to commit the bishop to e7 just yet.

10.a3 White has tried a few other alternatives but,

a

b

e

d

e

f

g

h

The Gelfand Variation

1 34

1 2.tLld5!N exd5 1 3 .exd5 t tLl e 5 ! ? ( 1 3 . . . .ie7 1 4. tLl c6 tLlffi 1 5 . .ixf6 gxf6 1 6.axb4 gives White a great position.) 1 4. fxe5 dxe5 1 5 . .ixf6 gxf6 1 6.a4 Material is level, and White has a promising initiative with great prospects on the light squares.

8 7 6 5 4

Bl) 10 ...i.b7

3 2

8

1

7

a

6

b

e

d

e

f

g

h

12 ..ixf6! gxf6

5

1 2 . . . tLlxf6 looks normal, but after 1 3 .g5 tLld7 1 4.h4t Black's position is rather unpleasant. There is no clear counterplay on the queenside, while White's kingside advance already looks imposing. White also has the option of :gh3 to restrain any counterplay.

4 3 2 1 a

b

e

d

e

f

g

h

This is the normal way to continue, but it gives White a few additional options besides converting into the main line considered in the next chapter.

l 1 .g4

1 1 ..ixf6!? tLlxf6 1 2.g4 is another interesting direction, which makes sense when you consider that Black often recaptures with the bishop in the 1 O . . . .ie7 lines. However, I will focus on the normal continuation.

1 1 J�c8?! ••

Black should prefer 1 1 . . . .ie7, leading to a position discussed below under the 1 O . . . .ie7 move order. The text move looks natural, but it gives White a relatively simple path to an advantage.

13.h4 tLlb6

Sending the knight towards c4 looks like Black's most logical idea. Other ideas include: 1 3 . . . h6 1 4.:gh3! The rook is perfectly placed here. 1 4 . . . :ggB 1 5 .<;t;>b l �b6 1 6.f5 e5 1 7.tLlb3t Grischuk - Wang Hao, Beij ing 20 1 3 . 1 3 . . . d 5 N has not been tested, with good reason in my view. 1 4.exd5 (I might also be tempted by 1 4. f5 ! ? , when Black's position looks shaky.) 1 4 . . . �xf4t 1 5 .�b l :gxc3 1 6.bxc3 ixd5 1 7.ig2 ic4 This was given by Palliser, who suggested that Black had compensation for the exchange, but Goh Wei Ming shows that I B .�e l ie7 ( 1 B . . . tLle5? 1 9 .tLlc6+-) 1 9 .ic6 is good for White. 13 . . . �b6 1 4 .:gh3 h5!?N is an interesting suggestion from Goh Wei Ming. ( 1 4 . . . tLlc5? 1 5 . f5 e5 1 6.tLld5 was excellent 'for White in Ortiz Suarez - Becerra, Silvania 20 1 1 .)

Chapter 8

-

135

Introduction

15Jlh3 �b6 16.£5 e5 17.lll d5 �d8 Here we can improve on the analysis of the Singaporean IM.

b

a

e

d

e

f

g

h

1 5 .f5!? (Goh Wei Ming's idea was 1 5 .g5 fxg5 1 6.hxg5 ig7 with counterplay for Black.) 15 ... e5 0 5 ... hxg4 gives White a pleasant choice between 1 6.Wfxg4 and 1 6. fxe6! ? ) 1 6.ttJb3 hxg4 1 7.Wfxg4;!; Black has managed to swap off his isolated h-pawn, but White has good control over the light squares and the h-pawn might become a factor if more pieces are exchanged.

a

b

e

d

e

f

g

h

18.g5!

1 B . ttJ b3 E:gB leads to a tense game according to Goh Wei Ming, but the knight does not have to retreat.

18 ... exd4

1 B . . . fxg5 1 9 .ttJ e6! is crushing.

19.t£!xf6t @e7 20.E:hd3! White has a huge initiative for the sacrificed piece. Here is j ust one illustrative line:

20 ... lll e3 2 1 .E:xd4! 23.exd6t @f8 a

b

14.@bl t£!c4N

e

d

e

f

g

h

This seems like the most energetic move, and an obvious attempt to improve on the slower 14 . . . h6 1 5 .E:h3 E:gB which was played in Rui ­ Molner, Santa Clara 20 1 4 . In the game White went for an immediate attack with 1 6.g5 ! ? , but Goh Wei Ming's 1 6.E:g3N ttJ c4 1 7.Wfe l i s a simpler way t o fortifY the queen side, when White remains in control and is ready for f4-f5 .

lll xdl

22.e5 iog7

8 7 6 5 4 3 2 1 a

b

e

d

e

f

g

h

1 36

The Gelfand Variation

24.d7 gaS 25.YlYe5! Threatening a deadly check on cS.

25 YlYe7 26.YlYc7 gdS 27.gxd1+•••

White will soon pick up the b7-bishop as well. Material will thus be about equal, but the mighty d7-pawn is crippling for Black.

B2) 10 i.e7 •••

8 7 6 5 4 3 2 1 a

b

e

d

e

f

g

h

This move order has a few advantages for Black. Firstly, it enables him to recapture on f6 with the bishop if White decides to exchange there. Secondly, it gives him the option of . . . E:bS, delaying the development of the cS­ bishop in order to threaten . . . b4. Finally, even if Black simply intends to enter the main line of the next chapter, the text move enables him to reduce White's options.

better than 1 2 . . . h6, transposing immediately to variation B22 below. However, the text move is even more flexible. An interesting independent possibility is: 1 2.i.xf6!? i.xf6 1 3 .gS i.e7 ( 1 3 . . .i.xd4 1 4.E:xd4 would be fine for Black if he had played . . . i.b7 instead of . . . E:bS, as he could just play . . . eS, but here it is not so easy for him.) 1 4.h4 b4 The only justification for the rook on bS. ( 1 S . . . W!b6 is possible, but it provides some additional options for White as well.) l S .axb4 E:xb4 Black may be okay here, but he has to worry about various attacking ideas such as f4-fS and tt:l fS . Another good reason for Black to avoid this move order! 1 2 . . . h6 1 2 . . . b4?! 1 3 .axb4 E:xb4 1 4.eS! leads to a quick conclusion, as after 1 4 . . . dxeS I S .tt:lc6 the f4-pawn is defended, so Black cannot take it with his rook. I S . . . i.b7 1 6.tt:lxb4 i.xb4 1 7. fxeS i.xc3 I S .exf6 gxf6 1 9 .bxc3 ± Ivanchuk - Van Wely Wijk aan Zee 20 1 0 . 1 3 .h4!?N My official recommendation would have to be 1 3 .i.h4, transposing to variation B22. However, I cannot resist mentioning the text move. Even though Black is objectively okay in the end, the sacrifice is worth considering for its shock value.

l 1 .g4 Now we will analyse B21) by B22) 1 l h6.

1 1 ...i.b7 followed

••.

1 1 . . . E:bS This has been a more common move order, but it gives White some additional options, without offering any real perks for Black. 1 2 .i.g2 Another simple reply is 1 2.i.h4, when the threat of g4-gS means that Black has nothing

a

b

e

d

e

f

g

h

1 3 . . . b4! 1 3 . . . hxgS 1 4.hxgS E:xh l I S .E:xh l tt:lgS 1 6.E:hS WfS 1 7.g6! b4! I S .gxf7 ( 1 S.tt:ldS

1 37

Chapter 8 - Introductio n exdS 1 9 .9xf7 \t>xf7 20.eS is another possible direction.) ,1 8 . . . 'it>xf7 1 9 .tLlxe6! The analysis goes on, but Black is certainly not having an easy time of it. 1 4.axb4 l'!xb4 l S . .txf6 .txf6 1 6.eS dxeS 1 7.tLlc6 l'!xf4 1 8 .tLle4 8 7 6 5 4 3 2 b

a

e

d

e

f

g

h

1 8 . . . l'!xe4! 1 8 . . . 0-0 1 9 .1'!h3! .tb7 20.l'!c3 .txc6 2 1 .l'!xc6 �aS 22.l'!xa6 �c7 23 .�bS tLl b8 24.l'!b6 leaves White with some initiative. 19 ..txe4 .tb7 20.l'!h3 tLlcS! 2 1 .l'!c3 tLla4 2 1 . . .tLlxe4 might be okay as well. 22.l'!c4 tLl b6= This was a shortened version of my analysis. Ultimately it is up to you whether you wish to stick to the main lines or keep the possibilities of 1 2 . .txf6!? and 1 3 .h4!?N as potential surprise weapons in your back pocket.

B21) 1 1 .. ..ib7

This position is almost the same as the main line shown in the earlier diagram on page 1 3 1 , the only difference being that the moves . . . h6 and .th4 have not been played here. Before going any further, it is worth highlighting one important point: if White continues developing with the natural .tg2, he will have to consider the annoying possibility of . . . �c4. Queenless positions are generally okay for Black, while if the white queen moves away, then . . . b4 will give Black attacking chances. It is worth considering two approaches for White. B21 1) 12.f5!? is an interesting attempt to exploit the fact that Black has been slow to play . . . h6, while B212) 12 ..ih4 invites Black to play . . . h6 and transpose to the main line, which will be analysed in the next chapter. However, Black is not forced to comply, and we will examine his attempts to keep the game in independent territory. It is worth mentioning that 1 2 . .tg2 is not completely pointless here, because after 1 2 . . . �c4 1 3 .�d2 the absence of . . . h6 and .th4 is definitely in White's favour. However, Black still seems to be doing fine, so I will not go into any more details here. 1 2 . .txf6 is also ineffective due to 1 2 . . . .txf6! 1 3 .gS .txd4 1 4.l'!xd4 eS l S . fxeS tLl xeS when Black is fine.

B21 1) 12.f5!?

8 7

This is a unique possibility when Black develops his bishops and neglects to play an early . . . h6.

6 5 4

12 ... e5

l . . . .. . ./'·//m·.. . . . . . . .

12 . . . tLlcSN has never been played, and indeed after 1 3 . fxe6 fxe6 14 ..tg2 0-0 l S .\t>b U White has plenty of targets to attack.

3 2

I

.

/,=/,

a

b

1 e

d

e

f

g

h

1 38

The Gelfand Variation

13 ..ixf6 1 3 . lLl b3 ? loses a pawn after 1 3 . . . lLl xe4! 1 4.lLlxe4 i.xg5 t .

13 ... �xf6 13 ... i.xf6?!N 1 4. lLl dxb5 axb 5 1 5 .lLl xb5 �b6 1 6.lLlxd6t±

14.�b3 At first, Black seems to be doing rather well here, especially j udging by the only over-the­ board game. However, if White takes the correct measures to stop Black's counterplay on the queenside, then his kingside pawn avalanche may prove dangerous.

White intends Eih3, renewing the threat of g4-g5 .

1 5 ... Eic8 1 5 . . . lLl d7 1 6.Eih3 EicS is covered under 1 6 . . . lLl d7 in the notes to the main line below. (This was the actual move order from the Balogh - Gabrielian game, as referenced there.) 1 5 . . . 0-0-0 brings the king to safety, but leaves White with a considerably more pleasant position. 1 6.i.g2 c;t>bS 1 7.c;t>b l EicS I S .g5 ( I S .Eih3!? is also possible, not being in a hurry to advance the g-pawn.) l S . . . hxg5 1 9 .hxg5 lLl h7 ( I 9 . . . lLl d7!? 20 .�d2;!;) 20.g6 fxg6 2 1 .fxg6 lLl f6 22 .i.h3 EicfS

8 7 6 5 4 3 2 1

a

a

b

e

d

e

f

g

h

14 ... h6 1 4 . . . 1:'kS?N 1 5 .g5 lLl d7 1 6. h4 gives White a dream middlegame, so it is understandable that the text move is the only one to have been tested so far. 1 4 . . . g5! ?N is an unusual method, but perhaps it deserves attention. 1 5 .h4 ( I 5 .i.g2 is a sensible alternative.) 1 5 .. .:1%gS 1 6. h 5 ! ? 0-0-0 1 7.i.g2 I prefer White in this strange middlegame. His king is more secure and, after doubling on the d-file, he may look to send the b3-knight towards d 5 .

1 5.h4

b

e

d

e

f

g

h

23 .i.e6! White showed exceptional technique to outplay his opponent in the following game: 23 . . . i.dS 24.�d3 �e7 2 5 .i.f7 Eixh l 26.Eixh l i.b6 27.lLld5 lLlxd5 2S .i.xd5 �f6 29 .i.f7 cj;; a7 30.lLl c l i.cs 3 1 . lLl a2 i.a5 32.lLlb4± Schubert - Bubir, corr. 2009. 1 5 . . . �b6!? This was played in another correspondence game. 1 6.Eih3 1 6. lLl d2!? can be compared with the main line below. The idea is to prevent . . . b4, although White does not have to be too worried about it for the momellt. 1 6 . . . EibS 16 . . . b4 1 7.axb4 �xb4 I s .lLl d2;!;

1 39

Chapter 8 - I ntroduction

20 . . . a5 2 1 .ltld5 .ixd5 22.exd5 b4 2 3 . ltl c4 �d8 24.f6!? gxf6 Or 24 . . . .ixf6 2 5 .g5 with a strong initiative. 2 5 .E!:fl � Black i s under continuous pressure.

8 7 6 5 4 3 2

v.=z. . ..,.,,.,)'" ,.,.,,.

a

b

c

d

e

f

g

h

1 7.ltld2!N Preventing . . . b4 due to the reply ltl c4. Meanwhile the g4-g5 threat is becoming increasingly strong. 1 7.g5 hxg5 1 8 .hxg5 E!:xh3 1 9 . .ixh3 ltl h7 20.g6 .ig5t 2 1 .�b 1 ltl f6 was unclear in Kain - Avotins, corr. 20 1 3 . 1 7 . . . ltld7 1 8 .�b 1 ltlc5 1 9 .�g2! White is in no hurry to play g4-g5 . Black has no clear counterplay, since . . . b4 is still hard to arrange. 1 9 .95 is slightly premature, and 1 9 . . . hxg5 20.hxg5 E!:xh3 2 1 ..ixh3 .ixg5 22.�h5 �d8 is unclear. 1 9 . . . .ic6 Intending . . . a5 and . . . b4. 1 9 . . . �d8 20 .�g3±

b

a

e

d

f

e

g

h

16J:!h3 Preparing g4-g5 , while guarding against potential exchange sacrifices on c3 .

16 ...Y;Yb6 Threatening both . . . b4 and . . . E!:xc3 . 1 6 . . . ltl d7 transposes to a grandmaster game a few years ago. 1 7.�b l !N ( l 7.ltld5?! is too soon. 1 7 . . . Ld5 1 8 .exd5 ltl b6 1 9 .�f2 ltl a4°o Balogh - Gabrielian, Plovdiv 20 1 2.) 1 7 . . . ltl b6

8 7

8

6

7

5

6

4

5

3 2

4

v·= ,·....... /'·c=

3 2

a

b

e

d

e

f

g

I':�'%. . ."'''' /'l.''' ;

h

20 . .ie2 20.b4!? ltld7 2 1 . ltl b3 is also favourable for White, although it seems a bit too 'computerish' .

a

b

e

d

e

f

g

h

1 8 .g5! hxg5 ( l 8 . . . ltl c4 1 9.�h5t) 1 9 . hxg5 E!:xh3 20 . .ixh3 .ixg5 2 1 .f6 .ixf6 22 . .ixc8 .ixc8 23.ltl c l ±

1 40

The Gelfand Variation

16 . . .Wfc4N could be considered, but after 1 7.Wfxc4 bxc4 1 B . ltJ a5 iaB 1 9 .1tJxc4 ltJ xe4 20.ltJxe4 ixe4 2 1 . ltJ e3 ib7 22.ltJ d5;l; it's a rather one-sided game.

b

a

e

d

e

f

g

h

b

e

d

e

f

g

1 B . . . ltJ c5 1 9 .95t

a

It is worth adding that going into an opposite-coloured-bishop endgame would only make things worse for Black: 22 .. .l'k5 23 .ig2 ixd5 24.ixd5 0-0 2 5 . b4±

a

18.�bl Wfd8

h

17Jl�d2! As we have already seen, this is the perfect way to prepare for . . . b4. Now White j ust needs a few more preparatory moves before playing g4-g5 .

17 ltJd7 .••

1 7 . . . b4 is met by I B .axb4 Wfxb4 1 9 . 1tJ c4;l; intending ltJ d5.

b

e

d

e

f

g

h

19.Wffl �c5 20.g5! hxg5 21 .£6! With promising play for White.

B2 12) 12 .ih4 .

a

b

e

d

e

f

g

h

This invites a transposition to the next chapter with 1 2 . . . h6, which is by far the most common continuation. However, Black can also consider deviating with B2 121) 12 ltJc5, B2122) 12 �b6 or B2 123) 12 g6!? •••

.•.

••.

The tactical operation 1 2 . . . ltJx�4?! does not work due to 1 3 .ltJxe4 ixh4 1 4.ltJxe6! fXe6 1 5 .ltJxd6t
141

Chapter 8 - I ntroductio n excellent position for White, Rotella - Aykent, corr. 20 1 O.

B2 121) 12 ... tLlc5

1 7. tLl c6! Wfc7 (After 1 7 . . . Wff4t 1 B .c;t>b 1 Black probably has to go back to c7 anyway.) 1 B . tLlxe7 Wfxe7 1 9 .1'!he 1 Wfb7 ( 1 9 . . . Wfc7 20.ixf6 Wff4t 2 l .c;t> b 1 Wfxf6 22.tLld5 WfdB 23.g5!±) 20.Wff2 Wfb6 (20 . . . tLl fd7 2 l .tLld5 is also unpleasant.) 2 l .Wffl Black is under pressure.

a

b

e

d

e

f

g

h

14.g5!?

13.i.g2 0-0 This seems scary for Black, but I am not sure if there is anything better. 1 3 . . . tLlh5?! is a funny idea, but it doesn't work. 1 4.ixe7 ( l 4.tLlcxb5 ! ? axb5 1 5 .Wfxb5t Wfd7 1 6.ixe7 �xe7 1 7.gxh5 is a messier variation which also favours White.) 14 . . . tLlxf4 1 5 .Wff2 Wfxe7 1 6.Wfxf4 e5 1 7.Wfg3 exd4 1 8 .l'!xd4 White has a solid advantage. 1 3 . . . 1':kB 1 4.e5 ixg2 1 5 .Wfxg2 dxe5 1 6. fXe5 Wfxe5

1 4.e5 is a good alternative which leads to a safe edge. 14 . . . dxe5 ( 1 4 . . . ixg2 1 5 .Wfxg2 dxe5 1 6. fXe5 tLld5 1 7.tLlxd5 exd5 1 B .ixe7 Wfxe7 1 9.1'!he l;!;) 1 5 . fXe5 tLl fe4 1 6.tLlxe4 ixh4 1 7.tLlxc5 ( 1 7. tLl d6!?) 1 7 . . . ixg2 1 B . tLl cxe6 fXe6 1 9 .Wfxg2 Wfxe5 20 .Wfc6 Black faces a tough defence.

14 ... c!Llfd7 1 4 . . . tLl fXe4? tries to be too smart. The idea is 1 5 .tLlxe4 ixe4 1 6.ixe4 d5, regaining the piece due to the threat of . . . Wfxf4t. However, after 1 7.ixh7t c;t>xh7 1 B .c;t>b l !± intending f4-f5 , it is obvious that the trick has backfired. ( 1 8 . . . Wfxf4 ? ? is impossible due to 1 9 .96t winning the bishop on e7.)

15.f5N Ponomariov has played 1 5 .c;t>b 1 , but the text move is more to the point.

1 5 ... e5 16.c!Llf3 a

b

e

d

e

f

g

h

White has good chances of success on the kingside.

1 42

The Gelfand Variation 1 4 . . . lLl c4N I S .�h3 Already �xe6 is quite a strong threat. I S .�e l ! ? is a more solid move, which can be compared with the main line below after 1 4 . . . E:c8.

B2122) 12 ... �b6

b

a

e

d

f

e

g

h

13.g5 �fd7 14J:�gl ! 1 4. 5 has been played by Shirov, but the position after 1 4 . . . eS I S . lLl f3 E:c8 is not so easy. For instance, after 1 6.�h3 Wc4 1 7.Wxc4 E:xc4 the position was double-edged in Sapozhnikov - Amanov, Philadelphia 20 1 1 .

14 .. J�c8 After 1 4 . . . lLl cS ?!N I S . 5 eS 1 6. lLl f3t the knight on cS is completely misplaced, and fS-f6 comes next. 1 4 . . . lLl a4N is also not helping Black: I S .lLlxa4 bxa4 1 6. fS eS

a

b

e

d

e

f

g

h

I S . . . dS I S ... 0-0 1 6.�xe6!? ( 1 6.g6!? �xh4 1 7.gxf7t E:xf7 1 8 .�xe6 lLl f8 1 9 .1LldS is a slightly more restrained alternative. 1 9 . . . �xdS 20 .�xdS E:e8 2 1 . lLl fS mh8 22.lLlxh4 E:xf4 I prefer White, due to his attacking chances and the excellent bishop on dS.) 1 6 . . . fxe6 1 7.lLlxe6 Wc8 1 8 .fS White has a dangerous initiative, the main threat being 5-f6. 1 6.exdS �xa3 1 6 . . . Wxf4t 1 7. m b l is too dangerous for Black. 1 7. bxa3! The position after 1 7.dxe6 �xb2t 1 8 .mb l 0-0 is exceptionally crazy. 1 7 . . . Wd6! This insane idea almost works. 8 7 6 5

a

b

e

d

e

f

g

h

1 7.f6! A typical idea since the c3-knight isn't hanging. 1 7 . . . exd4 1 8 . fxe7 lLl cS 1 9 .mb l t

4 3 2

a

b

e

d

e

f

g

h

1 43

Chapter B - I ntroduction l S . til b 1 Wxf4t 1 9 J!d2 Wxh4 1 9 . . . til xd2 20.tilxe6!± 20.dxe6 0-0 i 1 .exflt �hS 2 1 . . .Eixfl 22 . .ie6± 22 . .ixd7 tilxd2 23.Eig4±

13.f5!? The direct approach seems quite appealing. 13 . .ih3 Ied to success for White after 1 3 . . . EicS 1 4.g5 til h 5 1 5 .Eihf1 ! ± in Stanciu - Brunner, Aix-les-Bains 20 1 1 , but the improvement 1 3 . . . h6!N would completely block White's play.

8 7

1 3 .�b l is not a bad alternative. 1 3 . . . e5 1 4. fxe5 til xe5 l S .Eig l ! 0-0 1 6.gS til fd7 ( l 6 . . . til h 5 1 7 . .ie l EifeB 1 S . til d 5 .ixd5 1 9.exdS .ifS 20 .Wf2;!;) 1 7 . .ie l EifeS l S . h4 .ifS 1 9 .h5 .ig7 20.hxg6 hxg6 2 1 .Wh2t White had the better chances in Najer - Idani, Baku 20 1 1 . However, I like the main move even more.

6 5 4 3 2

a

b

e

d

e

f

g

h

15.b l!N White has no need to hurry. 1 5 . .ie 1 d5 1 6.e5 g6 was unclear in Pucher Carrillo, Hayange 20 1 3 .

15 ... �c4 16.J.el! With the knight on c 3 securely defended, Black has no good way to develop his queenside play. White, on the other hand, it j ust getting started on the kingside.

13 ... e5 14.�b3 1 4 . .ixf6?! tilxf6 1 5 . til b3 gxf5 1 6.gxf5 0-0-000 is less convincing.

14 .. J�c8 1 4 . . . gxf5 ? !N l S .gxfS seems rather dodgy; the king will be permanently stuck on eS, and once White prevents against any immediate tricks, he can utilize the g-file. 1 5 . . . EigB 1 6 . .ig2 EicS 1 7 . .ig3 Wc4 l S .Eid3;!;

16 ... �db6

1 6 . . . 0-0 1 7.f5 ! e5 l S . til b3±

17J�g3 Black has a highly unpleasant position.

B2123) 12 ... g6!? This rare but surprisingly interesting move is intended to prepare . . . e5. We will see a similar idea in variation C2 of the next chapter on page 1 5 5 .

a

b

1 5 ..ixf6!?N

e

d

e

f

g

h

l S ..ig2 is a sensible continuation, but it is hard to decide whether or not White should

1 44

The Gelfand Variation

allow 1 5 . . . lLl d 5 ! ? He was successful after 1 6.exd5 ( 1 6.lLlxd5 Axd5 1 7Jl:xd5 Lh4 is equal) 1 6 . . . Axh4 1 7. lLl d2! 0-0 I B . lLl de4 g5 1 9.f6 %Vb6 20.!!d3 in Krauss - Gonzalez Alcon, corr. 20 1 3 . Black may be able to improve somewhere, but it seems to me that he is under some pressure here too, so take your pick between this and the novelty given in the main line.

8 7 6 5 4 3

v.='-. . .-:-:,..,//.-,,= .

2

1 5 . . ti)xf6

b

a

.

1 5 ... Axf6 1 6.h4!? gxf5 Mter 1 6 . . . g5 1 7.'i!
e

d

e

f

g

h

1 9 . . . %Vb6 1 9 . . . h6 runs into 20 .AfS ! hxg5 2 1 .hxg5 when the opening of the kingside looks dangerous for Black. 1 9 . . . !!dB 20.AfS lLl fB 2 1 .lLld5 Axd5 22.exd5;t; 20.lLld5 Axd5 2 1 .!!xd5 2 1 .exd5!? also looks interesting. 2 1 . . .!!c7 22 .Af5� Black has little chance of counterplay, while White can continue to shufHe his pieces around and exert pressure.

4 3 2

v.=z. . ..,..,,.., ;//:�= ;

a

b

e

d

e

f

g

h

1 7 . . . Ae7 1 7 . . . Ag7 I B . .ih3 f4 1 9.!!xd6 %Vxd6 20.!!d l %Vb6 2 1 .Axd7t 'i!
a

b

e

d

e

f

g

h

16.,tg2 I doubt that Black can just sit around and allow g4-g5 and fS-f6, so he needs to make a decision.

16 ... gxf5

Chapter 8 - I ntroduction None o f 1 6 . . . g 5 1 7.h4, 1 6 . . . 0-0 1 7.h4 or 16 . . .Wi'c4 1 7. � d3 are much help to Black. 16 . . . h6 is not an ideal solution, as Black's structure will be rather loose after an exchange on g6. 1 7.Wb l ! ? Keeping many options open. 0 7.fXg6 fXg6 I B .h4 0-0 1 9 .Wb l also looks more pleasant for White.) 1 7 . . . g5 0 7 . . . 0-0 I B .h4i ; 1 7 . . . gxf5 1 8 .exf5;!;) I B .h4 l:!gB 1 9.hxg5 hxg5 20. tLl cl ! ?;!;

145

dangerous, but Black can equalize with precise defence: 22 . . . Wi'c6! (22 . . . WfB 23 .Wi'h3! l:!xg5 24.Wi'h6t l:!g7 2 5 .l:!g l +-; 22 . . . h6 23.h4 Wi'c6 24.Wi'h2! hxg5 2 5 . hxg5 l:!xg5 26.l:!e l i) 23 .Wi'xc6t l:!xc6 24.gxf6 '!xf6 2 5 .l:!e l t Wffi 26.l:!d2 l:!g5 =

20".h6 The attempt to counterattack with 20 . . . .!xa3 comes unstuck after 2 1 .tLlxd5 tLl xd5 22.l:!xd5 .!xb2t 23.Wb l ! .!c3 24.l:!ed l ± .

17.exf5! 1 7.gxf5 is less dangerous after White has given away the dark-squared bishop.

20 . . . d4 is well met by 2 1 .tLlxd4! exd4 22.l:!xd4 h6 23 .h4 Wffi 24.Wi'd2 hxg5 2 5 . hxg5 tLl eB 26.g6!± intending Wi'h6.

17".h:g2 18.VNxg2 gg8 I B . . . Wi'c6 1 9 .Wi'xc6t l:!xc6 20.g5 tLl g4 2 l . f6;!;

21 .g6 d4 2 1 . . . fXg6 22.tLlxd5 tLl xd5 23.l:!xd5 +-

19.95! 1 9.h3 Wi'c6°o

22.tLle4 ttlxe4 23.VNxe4

19".d5 1 9 . . . h6 20.h4 does not help Black. This is especially true if he continues with 20 . . . d5?, because of 2 1 .tLlxd5 tLlxd5 22 .l''!:xd 5, when the inclusion of . . . h6 and h2-h4 prevents Black from taking on g5 with a piece, and after 22 . . . hxg5 23.Wb l Black is in deep trouble.

Black is under considerable pressure, as his king is unsafe and White dominates the light squares.

B22) 1 l ".h6 8 7 6 5 4 3 2 1 a

a

b

e

d

e

f

g

h

e

d

e

f

g

h

It makes sense for Black to insert this move, regardless of whether he intends . . . l:!bB or . . . .!b7 next.

20.ghel! 20.@b l d4 2 1 .tLlxd4 exd4 22.l:!xd4 looks

b

12 ..th4

1 46

The Gelfand Variation

Black's last move was well timed. One important point is that 1 2.�xf6 can be met by: 1 2 . . . �xf6! 1 3 . ttJ dxb5 axb5 1 4 .ttJ xb 5 �c5 ( l 4 . . . �xb2t! ? 1 5 .�xb2 �b8 is possible, but there is no special reason for Black to go for this.) 1 5 .ttJxd6t �f8 I spent a while analysing this wild position, but did not find anything special for White. The main line of my analysis runs: 1 6.e5 :B:xa3! 1 7.�b l :B:a4 1 8 .exf6 �a5 1 9 .c4 :B:a l t 20.�c2 �a4t 2 1 .�d2 �b4t 22.�e3! �b3t 23.:B:d3 �b6t 24.:B:d4 gxf6 The position remains a crazy mess, but Black does not seem to have any complaints.

13 ... b4 14.axb4 :B:xb4 1 5 ..iel! 1 5 .e5 dxe5 1 6. ttJ c6 :B:xf4! 1 7.�g3 :B:xg4 doesn't really work for White. I once tested this variation with Black. My opponent continued with the unimpressive 1 5 .�f2?!, reaching a position that Black has even been ready to play a move down in some games. 1 5 . . . g5! 1 6. fxg5 ttJxg4! 1 7.�xg4 �xg5t 1 8 .�b l �xc3 1 9 . ttJ b3 �c7°o I went on to win in Yu Yangyi - Negi, Ho Chi Minh City 20 1 2 . 1 5 .:B:he I ! ? is a fascinating alternative, intending either a sacrifice with ttJd5 or ttJf5, or a breakthrough with e4-e5. I analysed this in some detail, but did not find anything conclusive for White.

1 5 lYb6 •••

1 5 . . . �c4 1 6.h4! intending g4-g5 leaves Black with an unpleasant position, regardless of the queen exchange. 8 7 a

b

e

d

e

f

g

h

12 J3b8 •.

If Black intends a set-up with the rook on b8, this is the best move order, as the inclusion of . . . h6 and �h4 reduces White's options. (Compare the earlier note with 1 1 . . .:B:b8 on page 1 36.) 12 ... �b7 is the most common move, reaching the main-line position as covered in the next chapter.

13 ..ig2 The only additional option for White via the present move order is 1 3 .�e l , but this does not seem to present much of a problem after: 1 3 . . . g5 ! ? 1 4.f5 ( 1 4. fxg5 hxg5 1 5 .h4 ttJe5 seems fine for Black.) 1 4 . . . ttJ e5°o

6 5 4 3 2 1 a

b

e

d

e

f

g

h

16 .ifl! •

Mter 1 6. ttJ b3 a5!? 1 7.�f2 ttJc5 Black is doing quite well.

16 lYc7 •••

Losing a move like this is certainly not Black's dream scenario, but his position has held up in several correspondence games (as

1 47

Chapter 8 - Introduction well as Yu Yangyi - Negi where it was Black to move here!) . After 1 6 . . Jhb2 1 7. lD db5! Black loses an exchange for insufficient compensation. 16 . . . lD c5 1 7.e5!? led to a fine win for White in a correspondence game: 17 ... dxe5 ( 1 7 . . . l'!xb2 1 B .lDf5! lD fd7 1 9 . 1Dxe7 rJ;; xe7 20.exd6t±) 1 B .lDc6 l'!xf4 1 9 .1'!hfl Wic7 20.i.xc5! (20 .lDxe5 0-000) 20 ... l'!xfl 2 1 .i.xe7 l'!xd 1 t 22.Wixd 1 i.b7 23 .i.d6 Wib6

19.96!?N I found some correspondence games where White played more slowly, but it is hard to resist the allure of the direct attacking move. Black's position seems to be on the edge, but things are still not easy.

19 ... 0-0 Also after 1 9 . . . i.f6 20.gxf7t cj;Jxf7 2 1 .i.h3 White keeps the upper hand, although the position remains complicated.

20 ..tel ! Avoiding 20.gxf7t l'!xf7 2 1 . lD xe6 Wia5 when Black is suddenly doing well.

8 7 6

20 ... tile5

5

20 . . . l'!bB 2 1 .gxf7t l'!xf7 22.lDxe6±

4 3

8

2

7

1 b

a

e

d

e

f

6

h

g

24.i.a3! Wic7 2 5 .Wih l ! Black resigned in Cipka - Siigur, corr. 20 1 2 . Most humans would play on a bit longer, but it is obvious that Black will not be able to evacuate his king from the centre without suffering heavy losses.

5 4 3 2 1

17.h4 h5 18.g5 tilg4

a

b

e

d

e

f

21 .�d5! exd5 22 ..ixb4 dxe4

g

h

22 . . . lD xe4 23 .i.e 1 also favours White.

23.gxf7t gxf7 24.gh3! 24.l'!hfl ? d5+

24... gxf4 24 . . . d5 2 5 .l'!c3!t

a

b

e

d

e

f

g

h

1 48

The Gelfand Variation

Conclusion

8 7 6 5 4 3 2 1 a

b

e

d

e

f

g

h

25J�c3 tJf6 26 ..bc5 dxc5 27.�c4t i>h7 28.tJe6 .be6 29.�xe6 Material is roughly balanced, but the opposite-coloured bishops give White excellent prospects of generating an attack.

7 . . . lLl bd7 is a flexible move, as evidenced by the bewildering number of move orders that may all end up at the same position (as featured in the next chapter) . Even though 1 O . . . �b7 has been the most popular so far, I do not believe it is the most accurate, as after 1 1 .g4 �e7 (or 1 0 . . . �e7 1 1 .g4 �b7) White has the tempting option of 1 2 .f5!? Whether or not you prefer this over 1 2 .�h4 - which invites a transposition to the next chapter with 1 2 . . . h6, but also leaves Black with the independent possibilities of 1 2 . . . lLlc5, 12 ... lLl b6 and 12 ... g6!? - is up to you, but if I were playing Black, I would not give my opponent the choice. For this reason, 1 O �e7 1 1 .g4 h6 makes a good deal of sense. After 1 2.�h4, we ended the chapter by considering the interesting possibility of 1 1 . . J�b8. This is quite challenging, although Black's main line involves shuffling the queen to b6 and back to c7, a tacit draw offer which might in itself dissuade a lot of players from choosing this line. Obviously we have some ways to keep the game going, and I am satisfied with White's chances in the main line I analysed. • • •

Having concluded our examination of the different move orders and sidelines, we may turn the page and consider the main line after 1 2 . . . �b7.

8 7 6 5 4

Gelfand Variation

3 2

",=['C!//.,,//, '//////',,;OWJ//// b",J'///a//. ,,'l[-='C'=z.�=/.'l=' " �"'m"//r"cc'

1 a

b

e

d

e

f

g

The Main Line Variation Index l .e4 c5 2.�f3 d6 3.d4 cxd4 4.tDxd4 � f6 5 . tD c3 a6 6.Jtg5 e6 7.f4 � bd7 8.�e2 �c7 9.0-0-0 b5 1 0.a3 Jte7 1 l .g4 h6 1 2.Jth4 Jtb7 1 3.c�b l ! A) 13 ... g5!?N B) 13 .. J�b8 C) 13 .. J�c8 14J�gl C l ) 14 ...�b8 C2) 14... g6 C3) 14 ... g5 1 5.fxg5 hxg5 1 6.Lg5 �b6! 17.�e l C3 1 ) 17 ... � e5?!N C32) 17 . . J�xc3 1 8.Jtxf6 �xf6 1 9.�xc3 tDxe4 20.�e l C32 1 ) 20... Jth4 C322) 20...Jtf6

8) after l 4 . , ,�a8

C322) after 24 . . . E1h4

C l ) after l 4 . . . %Vb8

8

8

8

7

7

7

6

6

6

5

5

5

4

4

4

3

3

3

2

2

2

g

l 5 .�e l !N

h

g

l 5 .E1g3!N

1 50 1 52 153 153 1 55 1 57 1 58 1 59 1 60 1 63

h

a

b

c

d

e

2 5 . g6!N

f

g

h

h

1 50

The G elfand Variation

1 .e4 cS 2.c�:lf3 d6 3.d4 c:x:d4 4.t£!xd4 t£!f6 S.t£!c3 a6 6.J.gS e6 7.£4 c!LJbd7 S.ti'e2 ti'c7 9.0-0-0 bS 10.a3 !i.e7 1 1 .g4 h6 12.J.h4 J.b7

castle on either side, so his main idea will be to prepare a timely . . . g5 . We will analyse the untested A) 13 gS!?N, followed by the rare but quite logical B) 13 J�bS, and finally the most popular C) 13 !!cS. •••

••

••.

A) 13 gS!?N •••

Even though this move has never been played, it is an obvious candidate for Black to consider, so it is important for White to know how to react.

14.fxgS t£!h7

a

b

e

d

e

f

g

h

This position can be considered the main line after B .W/e2. It has occurred more frequently via the 1 O . . . .ib7 move order, but my investigations in the previous chapter indicated that the bolded move order is the most logical for Black, as it limits White's options while keeping White guessing for a few moves.

13.
a

b

e

d

e

f

g

h

I S.!i.g3

1 5 . ttJ f3 hxg5 1 6 . .if2 ttJ hfB 1 7 . .id4 !!gB is unclear. 1 5 .g6?! does not work: 1 5 . . . .ixh4 1 6.ttJcxb5 ( 1 6.gxf7t? cj;>xf7 does not give White anything.) 1 6 . . . axb 5 1 7.ttJxb5 W/b6 I B .ttJxd6t cj;>e7 1 9 .9xh7 ttJ e5+

I S hxgS 16.h4! .••

We need to create instant problems, otherwise Black's excellent structure will promise him a fine game . 1 6. ttJ dxb5 ! ? axb5 1 7.ttJxb5 W/c6 I B . .ixd6 !!cB 1 9 . .ixe7 ( 1 9 .W/d2? ttJhf6-+) 1 9 . . . �xe7 20.ttJd6 !!cdB 2 1 .W/f2 ttJ hf6 is rather messy.

151

Chapter 9 - The Main Line

1 6 . . . liJe5 1 7.hxg5 0-0-0 I B .g6! liJxg6 1 9 .1iJdxb5 � b5 20.liJxb5 Wlb6 2 1 ..if2 Wla5 22 ..id4 liJe5 23.b4+-

1 9 . . . Wlc5 20.liJxd6t fB 21 ..ixh4 .ic6 22 . .if2 Wlg5 (22 . . . Wla5 23 . .ie l Wlg5 24 . .id2 Wlg7 2 5 .Wlc4±) 23 . .ie3 Wle7 24.Wlc4 .ia4 2 5 . b3 liJ hf6 26 . .ig5 ±

Objectively, the best move is perhaps 16 . . . liJhfB, but after 1 7.h5t the establishment of a protected passed pawn makes Black's position considerably less appealing.

1 9 . . . .ixg3! ? is an interesting try: 20.liJxc7t e7 2 1 . liJ xaB !!xaB At first, Black seems to have decent compensation, but White will continue to cause him trouble:

16 ... gxh4

b

a

a

b

e

d

e

f

g

h

17J�xh4! .ixh4 1 7 . . . b4 I B .liJd5! exd5 ( 1 B . . . .ixd5 1 9 .exd5 .ixh4 20 ..ixh4 e5 2 1 .axb4±) 1 9 .exd5 liJe5 20J%h5 0-0-0 2 1 .axb4t

18.�dxb5! IB . .ixh4 Wlc5 19 . .if2 Wlg5 isn't very clear.

f

g

h

7 6 5

Black has a number of options, but White's play is similar in all cases. Despite being a rook down, his initiative is tremendous.

3

1 9 . . . Wlb6 20.liJxd6t fB (20 . . . e7 2 1 ..ixh4t liJhf6 22 . .if2 Wlc7 23.e5 liJ d5 24 . .id4t) 2 1 ..ixh4 !!gB (2 1 . . .liJe5 22.Wld2) 22.Wld2 !!xg4 23 . .if2 Wlc7 24.Wlb4±

e

8

4

1 9 . . . Wla5 20.liJxd6t fB 2 1 ..ixh4 .ic6 22 . .ie l transposes.

d

22 .Wle3 .ie5 23.g5 ! ? !!gB (23 . . . .ic6 24 . .id3 liJ hfB 2 5 . !!fU) 24 . .ib5 ! liJxg5 2 5 .!!gl f6 26.Wla7 liJ c5 27.a4±

18 ... axb5 19.tlhb5

19 ...�d8

e

2 1 �� oo a

b

e

d

e

f

g

h

2o.c!ihd6t �f8 21 .�xb7 �b6 2 1 . . .Wlg5 22 ..ixh4 Wlxh4 23.!!xdn White will soon regain his coordination, and the queenside passers will give him great chances in the endgame.

1 52

The Gelfand Variation

22Jhd7 tilf6 22 . . . .ixg3 23 .Wf3 It'l g5 24.Wf6 �h7 2 5 .We7t (25 .Wxg5 ! ? also looks good.) 2 5 . . . �gB 26.Wxg5t �g7 27.Wf6;:!; 8

1 5 ... c!Llc5 1 5 . . . b4 1 6.axb4 �xb4 1 7.Wxa6± 1 5 . . . g5 1 6.h4! gxf4 1 7.g5 looks too dangerous for Black.

16 ..1g2

7

White should not be too hasty to launch his attack, for instance:

6 5

1 6.e5 dxe5 1 7.fxe5 It'l fd7 1 B .b4 O-O!oo creates a mess.

4 3 2 1 a

b

e

d

e

f

g

h

1 6.g5 hxg5 1 7.fxg5 It'l fxe4 1 B .lt'lxe4 .ixe4 ( I B . . . lt'l xe4 1 9 .96i) 1 9 .b4 White wins a piece, but once again the game becomes too complicated after 1 9 . . . d5!00.

23 . .ic7! Wc6 24 . .id6t 'i!/g8 25J�c7 Y:Vxe4 26 . .ie5 Y:Vxe2 27.,ixe2 The queenside pawns ensure that Black will continue to suffer in the endgame.

B) 13 ... l:1b8 8 7 6 5 4 3 2

16 . . . 0-0 1 7.e5 dxe5 1 B .fxe5 It'lh7 19 . .ixaB �xaB 20.h4 gives White a promising initiative.

1 a

b

e

d

e

f

g

h

14.l:1g1 .ta8 1 5 ..tel !N 1 5 . .ig2 has been played, but the thematic 1 5 . . . Wc4! would have been okay for Black. ( I 5 . . . b4? 1 6.axb4 �xb4 1 7.Wxa6 WbB 1 B .lt'ldb5± was great for White in Sutovsky ­ Efimenko, Kallithea 200B.)

17.h4 tilb6 18.b3!?t Preventing Black's counterplay. White's kingside initiative will continue uninterrupted, while Black does not seem to have a clear plan.

Chapter 9

-

C) 13 J�c8 14J�gl

1 53

The Main Line

1 5J:�g3!N

••

Rizzitano co�ers S.�e2 rather briefly in his chapter on the Gelfand Variation, and in a note covering this position he j ust mentions that 1 4.�g2 �c4 is equal. The text move is stronger, and fulfils several purposes. White conveniently keeps the bishop on f1 to deter . . . �c4, while also avoiding the threat of . . . lLlxe4 and supporting a future g4-g5 . Black has made all the obvious developing moves, and now needs to show his plan . We will analyse the rare options of Cl) 14 �b8 and C2) 14 g6 before examining the most popular and combative move, C3) 14 g5. •••

•.•

•••

1 4 . . . �b6?! is an inferior version ofvariation C l , because the queen is more exposed and does not restrain e4-e5 . A good answer is 1 5 .�f2!? ( 1 5 .e5 is also strong) 1 5 ... E!:xc3 1 6.lLlxe6 with complications that favour White.

This is the most logical move for a human player to understand, so I have given it main­ line status. Both 1 5 .�e l lLl xe4! and 1 5 .�e l lLl c 5 ! ? are interesting to analyse, but I found no clear route to an advantage in either case . 1 5 .�g2 led to an extremely impressive win for White in an email game, but this style of play would be much harder to emulate without the aid of a computer. You can use the game as a starting point for your own investigation if you wish: 1 5 . . . E!:xc3 1 6. bxc3 lLlb6 1 7.�d3 lLl c4 ( 1 7 . . . �c7 is better, but again I S .e5 seems good for White.) l S .e5! dxe5 1 9 .�xf6 gxf6

14 . . . lLl b6 1 5 .�xf6! forces 1 5 . . . gxf6 ( 1 5 . . . �xf6? allows 1 6.lLlxb5) and now both 1 6. f5 and 1 6.E!:g3!? are promising for White.

Cl) 14 �b8 ••.

a

b

e

d

e

f

g

h

20.lLlxe6! fxe6 2 1 .g5!! hxg5 22.fxg5 f5 23 .�d7t i>f7 24.g6t i>f6 2 5 .�xb7 E!:gS 26.g7 �a7 27.h4 Black was busted in Cunha - lames, email 20 1 O .

1 5 ... g5 Trying to exploit White's last move.

a

b

e

d

e

f

g

h

I only found one example of this move, but it is a natural idea which threatens . . . E!:xc3 .

1 5 . . . g6 prepares . . . e5, rather like variation C2 below. I like 1 6.E!:h3 ! ? here. Even though the rook looks weird, it serves a useful purpose in discouraging Black from castling. Now f4-f5 is possible at any point, so Black had better not delay his central action: 1 6 . . . e5 1 7. fxe5 lLl xe5 ( 1 7 . . . dxe5 l s . lLl b3 ± Black's position is stuck.)

The Gelfand Variation

1 54

1 8 .i.g2t White intends ttJ d5, while Black does not have an obvious way to continue.

8 7

8

6

7

5

6

4

5

3

4

2

3

1

2

a

b

e

d

e

f

g

h

21 .�d5!

1 a

b

e

d

e

f

g

h

16.fxg5 tilh7 17.B:h3!

2 1 .h4 b4! is unclear.

21 ...�xe4

1 7.E:e3 hxg5 1 8 .i.g3 ttJe5 is unclear.

Maybe Black should prefer 2 1 . . . exd5 22.exd5 f6, although after 23.h4 ttJf7 24.fie@, intending ttJ f5 , he continues to suffer, while White has a dominating initiative.

17 ... �xg5

22.�xe7 �xe7

Discouraging Black from recapturing with the h-pawn.

1 7 . . . hxg5 is positionally desirable, but after 1 8 .i.e l t the pin along the h-file is troublesome. Black will struggle to untangle his pieces, for instance: 1 8 . . . ttJe5 ( 1 8 . . . 0-0 may be a better try, but after 1 9 . ttJ f3 Black's king can hardly feel safe.) 1 9 . ttJ f3 fic7 20.ttJxe5 dxe5 2 1 .fif3± Intending i.d3 and ttJ e2.

22 . . . ttJxg3 23.hxg3 cj:;xe7 24.ttJ f5t±

23.�h4t �f8

18.B:e3 tile5 19 ..tg3 Intending h2-h4.

19 ... �c4 1 9 . . . E:g8 20.h4 E:xc3 (20 . . . ttJh7 is refuted by 2 1 .ttJxe6! fxe6 22.i.xe5 dxe5 23.g5+- and the queen gets to h 5 . ) 2 1 . bxc3 ttJ h7 22.g5! hxg5 23 .i.xe5 dxe5 24.ttJxe6!± a

20.B:ed3 .ta8

20 . . . ttJe5 is met by 2 1 .i.xe5 dxe5 22.ttJ b3t when White's ideas include E:d7 and h2-h4 .

24.�xe6t!

b

e

24.i.g2 ttJ c5ao

d

e

f

g

h

1 55

Chapter 9 - The Main Line

24 ... fxe6 25 ..ig2 tLlc5 26Jlfl t cj{g8 27.b4! ixg2 28.bxc? tLle5 29.cxd6 .ixfl 30.\Wxe5+-

about the fluid state of the kingside. 1 8 . . . lLl c4 1 9 .h4 Wfb6 20.�g3;!;

C2) 14 ... g6

1 6 . . . lLl xe4 1 7.lLlxe4 .ixe4 1 8 .Wfxe4 gxf5 1 9 .WfxfS .ixh4 20 . .ig2� occurred in Sherwood - Konstantinov, Internet 20 1 2 . White's position looks extremely pleasant due to the light squares. Finally, after 1 6 . . . gxfS 1 7.exfS Black has tried two moves: a) 1 7 . . . d5 1 8 . .ig2 e4 1 9 . .ig3 lLle5 Now in Langer - Zakharov, Internet 20 1 1 , White could have secured an edge with:

a

b

e

d

e

f

g

h

A completely different concept from the more popular 14 . . . g5. Black takes away the f5square from the knight and prepares to play . . . e5. It looks a little slow, but White hasn't been plotting any quick combinations either.

15.f5! This is the only move to cause Black any problems. 1 5 ..ig2 e5 1 6.fXe5 dxe5 1 7 . .ig3 Wfc4 is equal. 1 5 .g5 hxg5 1 6 . .ixg5 Wfb6 is comfortable for Black.

15 ... e5 16.�b3 Exchanging the bishop serves no purpose: 16 . .ixf6 lLlxf6 1 7.lLlb3 g5 and Black is fine. After the text move we arrive at a typical blocked position.

a

b

e

d

e

f

g

h

20.h3!N Now 20 . . . .id6 2 1 ..ifl;!; favours White, while 20 . . . .ixa3 runs into the nasty 2 1 . .ixe4! dxe4 22.Wfh2 . b) 1 7 . . . lLl e4!? 1 8 .lLlxe4 .ixe4 1 9.Wfxe4 .ixh4 occurred in Depasse - Kolar, Internet 20 1 3 . Here White should have played: 8 7 6

16 ... g5!?N

5

Blockading the kingside is logical, as Black's more combative attempts do not work.

4

After 1 6 . . . lLl b6N 1 7. fXg6 fXg6 1 8 . .ie 1 White has a pleasant position, and Black has to worry

2

3

v,=·"" ,�-;-;,

a

b

e

d

e

f

g

h

1 56

The Gelfand Variation

20J':!:g2!N Overprotecting c2 and preparing to centralize the rook. 20 . . . ttJ b6 2 1 .E:e2 ttJ c4 22.ttJ c l ! ? The knight starts making its way towards d 5 . Black is weak on the light squares and his king lacks a safe home.

17 ..iel ! 1 7.i.f2 seems less accurate due to 1 7 . . . h 5 I S .h3 ttJ b6 1 9 . ttJ aS i.aS , intending 20.i.g2 ttJ a4!. By putting the bishop on e l , White overprotects the knight on c3 while preparing to support a knight on a5 in the future. a

b

e

d

e

f

g

h

8

20 ..ifl

7

20.E:g3 is another plan, but it seems more artificial to me. For instance: 20 . . .'JifB 2 1 .�g2 Planning ttJ d S . 2 1 . . . hxg4 22. hxg4 mg7 23.ttJdS ttJxdS 24.exdS White has ideas of E:c3 and i.f2, although it's difficult to assess the position. Most probably it should be a balanced fight.

6 5 4 3 2

20 ...�e7 21 ..ig2 hxg4

1 a

b

e

d

e

f

g

h

17 ... h5 1 8.h3 tile5 I S . . . ttJ b6 is met by 1 9 .ttJaS! i.aS 20.i.g2. White's last move prevents a . . . dS break, and now he can j ust continue playing patiently, and slowly improving his position. Here is an illustrative line: 20 . . . ttJ c4 2 1 .ttJxc4 �xc4 22 .E:d3 as 23.b3 �cS 24.E:h l hxg4 2 S . hxg4 E:xh l 26.i.xh l E:bS (26 . . . i.c6 27.E:h3) 27.i.f2 �c7 (Or 27 . . .�xa3 2 S . ttJ a2! and the queen is stuck.) 2S .E:h3 i.c6 29 .i.e3;!;

19.tlhe5 I also considered 1 9.i.g2!?, but after 1 9 . . . ttJxb3 20.cxb3 hxg4 2 1 .hxg4 b4!? Black obtains some counterplay: 22.axb4 �b6 23 .i.d2 �xb4 24.i.xgS �xb3 2 S .E:d3 �b6 26.i.e3 �b4f!

19 ...�xe5

2 1 . . .�fB 22.E:h l ! hxg4 24.E:xh 1 mg7 2 S .i.e3 ±

23.hxg4

22.hxg4 �e4! Black has to find counterplay.

23J�d3 a5 23 . . . b4 is met by 24.b3! of course. 8 7 6 5 4 3 2 1 a

b

e

d

e

f

g

h

E:xh l

1 57

Chapter 9 - The Main Line

24.h3 �c6 25.tLld5 c!tlxd5 26.exd5 �a6 27J�hl The alternative is 27.ifl b4 28.a4, intending W'e4, :B:d 1 and so on. We have a similar type of closed position where both sides can continue manoeuvring for a long time. White's chances appear a little better, although Black's position remains hard to crack.

27... :B:xhl t 28 ..bhl White has chances t o generate pressure along the h-file, while Black does not seem to have any concrete plan for now.

C3) 14 ... g5

b

a

e

d

e

f

g

h

1 8 .g5! ixg5 1 9 .ih3 lLl dffi 20 .ig4 I!;>e8 2 1 .e5 dxe5 22 .ixe6 lLl g6 23.W'h5 lLl hffi 24. lLl f5 :B:d8 2 5 .:B:xd8t W'xd8 26. lLl g7t I!;>e7 27.:B:xg5 ± Morokova - Chiru, corr. 20 1 2 .

16 ..bgs �h6!

8

Black has serious prospects for counterplay, with both . . . b4 and . . . :B:xc3 on the cards.

7 6 5

V//h'/•/, • :=

4 3 2 1 a

b

e

d

e

f

g

h

This is the most critical move, and it has been played in the great majority of games from this position, although as far as I can see, every single one of them has been a correspondence or internet encounter.

15.fxg5 hxg5 1 5 . . . lLlh7?! has turned out badly for Black in a few games: 1 6.g6 ixh4 1 7.gxf7t I!;>xf7 ( l 7 . . . 'it>e7 1 8 .g5 lLlxg5 1 9 .:B:g4 :B:cf8 20.:B:xh4 :B:xf7 2 1 .:B:g4± R. Bauer - Gerola, corr. 20 1 2.)

a

b

e

d

e

f

g

h

17.�el 1 7.:B:g3 is another logical try. If you wish to explore this in more detail, I suggest using the following game as a starting point: 1 7 . . . :B:c4 1 8 .:B:gd3 lLle5 ( l 8 . . . lLl c5 1 9 .1Llxe6! ?) 1 9 . 1Ll f5 ! lLl xd3 20.lLlxe7 lLl xb2! 2 1 .I!;>xb2 b 4 22.I!;>a 1 :B:d4!! 23 .ixf6 bxc3 24.:B:b l W'c5 2 5 . I!;>a2 :B:d2 26.:B:xb7 :B:xe2 27.:B:b8t I!;>d7 28 .:B:b7t I!;>e8 29.:B:b8t I!;>d7 lh-Y2 Walsh - Weber, corr. 2009. This line featured some insane complications,

1 58

The Gelfand Variation

and there are a number of ways to deviate, but ultimately I found no advantage for White.

20.�xb5t �f8 22.�xe5 +-)

At this point we will analyse the tempting but inaccurate C31) 17 ... tLle5?!N, followed by the more critical C32) 17 �xc3.

C31) 17 ... tLl e5?!N

2 1 .e5!

lLlxe5

8 7 6

•••

A playable but less exciting try is 1 7 . . . b4!?N, after which 1 8 .axb4 �xb4 1 9 . h3 E:xc3! ? 20.�xf6 lLl xf6 2 1 .�xc3 �xc3 22. bxc3 lLl xe4 23.�b2 �f6 reaches a slightly better endgame for White. Black of course has decent compensation, but it is not enough to equalize completely.

(20 . . . �d8

5 4 3

V."""'h. . . . . . .

2

b

a

e

d

e

f

h

g

2 1 .�xg4! lLlxg4 22.�g l ! +- A brilliant win. 8 7 6 5 4 3 2 1

a

b

e

d

e

f

g

h

This computer suggestion does not end well for Black.

IS.h4! tLlfxg4 1 8 . . . E:xc3 1 9 .�xc3 lLl xe4 20.�e3 lLl xg5 2 1 .hxg5 �g8 22.�g2 �xg5 23.�g3± is a typical situation where Black has less than full compensation for the exchange. 1 8 . . . b4 is possible, but Black will j ust remain a pawn down, while White consolidates: 1 9 . axb4 �xb4 20.E:g3 �b8 2 1 .�e2 �a8 22.b3± Finally, 1 8 . . . lLlexg4 runs into a spectacular refutation: 1 9 .it:k xb5! axb5 ( 1 9 . . . lLle5 2o.lLlc3±)

a

b

e

d

e

f

g

h

19.J.e2

1 9 .�xe7 �xe7 20 .�e2 lLl f6 is murky.

19 ... tLlf6 1 9 . . . E:xc3 ? does not work: 20.�xe7 E:e3 2 1 .�xd6! �xd6 22.lLlf5 �b6 23.lLld6t �f8 24.lLl xb7 �xb7 25 .�b4t +1 9 . . . �xg5 20.hxg5 leaves the knight short of squares, and after 2o . . . lLl e3 2 1 .�xb5t axb5 22.�xe3 White is clearly better.

20.�f1 ! tLled7 20 . . . lLlxe4 2 1 .lLlxe4 �xe4 22 .�xb5t! axb5 23 .�xe4±

Chapter 9

-

1 59

The Main Line

so much harder to play over the board. I would definitely argue that White is better here, norwithstanding the fact that it is difficult to overcome the precise defence of a silicon brain .

8 7 6 5

From this position C321) 20 J.h4 is a sensible move, but C322) 20 .if6 has been the overwhelming favourite among the correspondence players. •••

4

•••

3 2

20 . . . E1xh2N is not much of a concern, and White has a pleasant choice:

1 a

b

e

d

e

f

g

h

21 .�dS! Out of nowhere!

21. exdS 22.exdS± •.

Intending ig4, with a powerful attack.

C32) 17 J�xc3 •.

a) 2 1 .E1g2 ih4 (2 1 . . .E1h6 22.E1e2 ih4 23 .�b4 ie7 24.i.g2 � g3 2 5 .E1f2;!;) 22 .�b4 E1xg2 (22 . . . t2k3t? 23.�xc3 ixg2 24.ixg2 E1xg2 2 5 .�h3! wins material.) 23 .i.xg2 �c5 24.�xc5 dxc5 2 5 . � f3;!; With only one pawn for the exchange, Black faces a tough endgame. b) 2 1 .g5!? should lead to a better version of variation C322: 2 1 . . .liJ xg5 22.�g3 E1h l (22 . . . E1h6 23 .ig2;!;) 23.E1xh l ixh l 24.ig2 ixg2 2 5 .�xg2 d5 26. liJ f3 liJxf3 27.�xf3

8 7 6 5 4 3 2 1 a

b

e

d

e

f

g

h

Following the wisdom of the crowds . . .

18 ..ixf6 �xf6 19.\Wxc3 �xe4 20.\We1 Amazingly enough, almost all games have ended in draws from here. Does this position look like 0.00 to you? This kind of situation highlights the danger of using correspondence games to build a repertoire: there are some positions which are objectively defensible, yet

a

b

e

d

e

f

g

h

Now White is ready to infiltrate with his rook through the kingside. Black will remain under pressure, for instance: 27 . . . a5 (27 . . . 5 2B.E1h l �d7 29 .�g3;!;) 2B.E1h l b4 29.E1hBt �d7 30.axb4 axb4

1 60

The Gelfand Variation Another ugly move, but Black's pieces also lack harmony.

8 7 6

22 ... �e7!?

5

Intending . . . a5.

4

22 ... 'Wc5N is less challenging, as it leads to a worse endgame for Black, though a lot of fight remains.

3 2 b

a

e

d

e

f

g

h

3 1 .�a8! (3 1 .'Wxf7 'Wg 1 t 32.@a2 b3t! 33.@xb3 'Wb6t=) 3 1 . . . 5 (3 1 . . .'Wb7 32.�a4 f5 33 .'We3±) 32.'Wh5t

8 7 6 5

C321) 20 ....th4

4 8

3

7

2

6

1

5

a

4

23.tLlb3

3

Restraining Black's a-pawn and intending g4-g5 .

2 1 a

b

e

d

e

f

g

h

This has been tried successfully in a few correspondence games, and it seems quite logical since White is forced to play a couple of odd-looking moves.

2 1 .�b4 .if2 2 1 . . .@e7?! is too slow: 22.i.g2 i.f2 (22 . . . �c8 23.i.xe4 i.xe4 24.�gf1 ±) 23.i.xe4 i.xe4 24.�gf1 a5 2 5 .'Wc3 �xh2 (25 . . . b4 26.axb4 axb4 27.tt'lc6t 'Wxc6 28.'Wxc6 i.xc6 29 .�xf2±) 26.�xf2 �xf2 27.'We 1 �f4 28 .'We3 i.xc2t 29.@xc2 'Wc5t 30.'Wc3±

22J�!g2

b

e

d

e

f

r

h

Mter 23.g5 a5 24.'Wxb5 i.xd4 25.�xd4 tt'lc3t 26. bxc3 'Wxb5t 27.i.xb5 i.xg2 28 .�a4 White seems a bit better, but Black should have enough counterplay to equalize.

23 ... e5! This has been played in all games so far. The point is that Black has no real way of stopping g5-g6, and you will see in the main line that the advancing e-pawn is an important source of counterplay in the ensuing endgame.

24.g5! The most straightforward attempt, and since it's only playing for two results, r see no reason to avoid it. 24.c4 has been played a few times, but after

Chapter 9 - The Main Line the inevitable exchanges Black seems to be able to make a draw: 24 . . . bxc4 2 5 . lt'l a5 'Wxb4 26.axb4 i.a8 27.lt'lxc4 i.d4 28 J�e2 :gh4 29.lt'le3 i.xe3 30.:gxe3 lt'lf2 3 1 .:gc 1 It'lxg4 32 .:gc7t
24 .. J::! h4! Again, the only move to have been played. 8 7

161

8 7 6 5 4 3 2 b

a

e

d

e

f

g

h

3 1 .h3!? The simplest. 3 1 .i.xa6 g4 3 2 . lt'l c 5 ! ? would be more combative, but the text move is enough for a slight plus. 3 1 . . . lt'l h 5 32.i.xa6 lt'l f4 33 .i.c8! 33 .i.xb5 lt'l xh3 34.i.e2 may also be better for White, but it gives Black some counterplay. 33 . . . e4 34.�c 1 e3 3 5 .� d l �d6 36.i.g4±

26Jlxg6 �f7

6 5 4 3 2 1 �� OD a

b

e

d

e

f

g

h

25.g6!N 25 .h3 :gf4! 26.:gg4 :gxg4 27.hxg4 It'lxg5 was okay for Black in Dunlop - Petersons, corr. 20 1 1 . a

25 ... fxg6

25 . . . lt'l f6 This may lead to a similar type of endgame with a slight edge to White. 26.:gxd6! :gxb4 27.:gxb6 i.xg2 28 .i.xg2 i.xb6 29.axb4 fxg6 30.i.b7 g5 After 30 . . . lt'l g4 3 1 .i.xa6 It'lxh2 32.i.xb5± Black's pawns are much harder to push.

b

e

d

e

f

g

h

27.gg2!? 27.:gg3 is a tempting option, but 27 . . . i.xg3 28.hxg3 :gh l ! leads to some forcing lines where I found no more than a draw for White. Also after 27.i.d3 �xg6 28 .i.xe4t :gxe4 29.:gxd6t �g5 30.:gxb6 :gxb4 3 1 .:gxb7 :gh4 Black has enough counterplay for equality.

The Gelfand Variation

1 62

27.:!::l gxd6! ? is possible, but I think White gets more winning chances in our main line. After 27 . . . ltJxd6 28 .:!:hd6 :gxb4 29.:gxb6 ixb6 30.axb4 White is definitely trying, but the two bishops may well be enough to neutralize the extra pawn.

27 ... �f6 28J�xd6 28 .�xd6 �xd6 29.:gxd6 ixg2 3o.ixg2 gives Black additional options.

28 ... :gxb4 29.:gxb6 .bg2 29 . . . ixb6 30.axb4 ixg2 3 1 .ixg2 e4 32.c3 is obviously better for White.

30 ..ixg2 gg4 3 1 Jlb7t g6 Usually I would stop analysing after reaching an endgame, but it is worth seeing how White should keep the e-pawn under control.

a

b

e

d

e

f

g

h

35.c3! 3 5 .ixa6? ltJ d 5 ! 36.:ge6t \t>f7 37.:ge5 ? �f6! wins for Black.

35 ... �d5 3 5 . . . :gxh2 36.ltJd4 shows White's ideal set-up. 36 . . . ltJd5 37.if5t \t>g5 38 .:ge5 ltJc7 39 .id3t \t>f6 40.:gc5;!; Now that the bishop has secured the blockade of the e-pawn, White can continue exerting pressure on the queenside.

8 7 6 5

36.ge6t

4

36.:ge5 gives more options for Black, although it should lead to a similar position.

3 2

36 ... f7

1 a

b

e

d

e

f

g

h

32.�h3! 32.ic6 :gc4! solves Black's problems. Now both 33 .ig2 :gg4 and 33.ltJa5 :ga4! 34. ltJ b3 :gc4 j ust repeat the position, while after 33 .id7 ltJ xd7 34.:gxd7 :gc6 Black is certainly not worse due to the activity of his pieces.

8 7 6 5 4

32 .. J!h4 33 ..ic8 e4 34J!e7 e3

3

At first glance, Black seems to be close to a draw, but once we are able to stop the e-pawn, Black's queenside weaknesses will not be so easy to defend.

2 1 a

b

e

d

e

f

g

h

Chapter 9

-

The Main Line

1 63

37.tild4 gxh2 38.ge5! �c7 39 ..ig4� Once agai� the pawn has been successfully halted, resulting in a one-sided endgame. The bishop will go to e2, after which White can start poking around on the queenside or targeting the e-pawn .

C322) 20 .if6 .••

8 7

a I'''/,/��//. ;;;;;;

6 5 4 3 2

a

"b

c

d

e

f

g

h

This is the most popular and logical move. Black is provoking c2-c3 , which can often be considered a weakness in the upcoming lines.

21 .c3 gxh2 2 l . . .id5?N 22.g5! til xg5 23 .ig2 ixg2 24.�xg2 j ust gives White an additional h-pawn compared to the main lines. At first, White doesn't seem to have a way to untangle, until you notice the following idea.

22.g5! �xg5 Mter 22 . . . ie5 ? 23.g6 Black's position would collapse. 22 . . . ixd4N Exchanging doesn't seem logical, but White should still be quite precise here. 23.cxd4 b4! ? Ruining White's structure.

b

e

d

e

f

g

h

24.axb4 24.�xb4 til d2t! 2 5 . m a l �xb4 26.axb4 if3 27.ge 1 id5 !� 24 . . . id5 2 5 .�g2! 2 5 .ig2? �c6 would turn the tables. 2 5 .id3 Wic6 26. b 5 ! ? axb5 27.Wia5 me7 28 .Wia7t Wib7 29 .Wixb7t ixb7 30.mcl leads to an interesting endgame where White is pressing, but I 'm not sure if it's enough. 25 . . . gh l 8 7 6 5 4 3 2

a

b

e

d

e

f

g

h

26.g6! fxg6 27.gc2! A nice switch! Now White gets good play along the c-file. 27 . . . me7 28 .Wie3 Wixb4 29.l::i dc a White maintains some initiative.

23.�g3 gh6! 23 . . . �h8 This retreat has been played more frequently, but it is less precise. 24.ig2 ie4t

1 64

The Gelfand Variation

24 . . . .ixg2 2 S .Wxg2 dS 26.�gf1 .ie7 27.lLlf3 leads to an unpleasant defence for Black. 24 . . . lLl e4 2 S .We l highlights the drawback of the rook on hB instead of h6. 2S . . . dS (After 2S . . . �gB 26.l!?a l the knight cannot move, since the rook is unprotected on gB .) 26 . .ixe4 dxe4 27.Wg3 White has a clear initiative. 2 S .l!? a l dS 26.�df1 eS Or 26 . . . .ie7 27.he4 lLl xe4 2B .Wg7 �fB 29.�h l and Black is suffering.

Certainly a natural move, although we have some other options here as well. 24.Wa l ! ?N is quite logical, but I won't analyse it in detail since there are just too many options for both sides. White keeps the ideas of .ig2 and �e3 intact, and the computer indicates that Black must play the bizarre 24 . . . Wd7 in order to transfer the king to a safer place. A brief line is 2S ..ig2 (2S .�e3 �cSoo) 2S . . . dS 26.�dfl Wd6 27.We l �eS and Black seems okay. 24 . .id3!?N is another way of playing, intending to regroup with moves like �de 1 and .ic2. I think practically it looks harder for Black to continue, even if objectively his position is okay. A few brief possibilities: 24 . . . l!?d7 (24 . . . lLl e4? 2 S .We3 �h4 26.�df1 ±; 24 ... .ie7 2 S . .ic2 �h3 26.Wg4 WcS 27.�de l .idS 2B . .ib3 .ixb3 29.lLlxb3 �c4 30.lLld4t) 2 S . .ic2 I!?c7°o

b

a

e

d

e

f

g

h

27.�xf6! Wxf6 2B .�f1 Wg7 29 . .ixe4 dxe4 3 0 . lLl fS Wf6 3 1 . lLl e3 lLl f3 32.Wg4! Wc6 33.�d l ± This position has been reached i n several correspondence games, with a 1 00% score for White.

24.We3 is the most popular move, which also deserves attention. 24 . . . WcS 2S . .ig2 .ixg2 26.�xg2 WdS ! If Black can exchange queens, he will be completely comfortable.

8 7 6 5 4

a

3 2 1 a

24 .ig2 .

b

c

d

e

f

g

h

b

e

d

e

f

g

h

27.Wf2!?N (I found nothing special for White after 27.�g4 �h l ! or 27.Wg I N .ixd4!.) 27 ... lLl e4 There are some other moves, but I just want to show White's idea. 2B.Wg l ! Avoiding the queen exchange. 2B . . . l!?d7 29.l!?a l WcS 30.We l White maintains some pressure.

Chapter 9

-

The Main Line

1 65

26 ... �g3!?N

24 �e4 .••

24 . . . ie4t? 2 5 .ixe4 It:l xe4 26.WlgBt rJle7 27.WlaB+-

Another move which is barely surviving, but it seems necessary.

24 . . . d5?! is risky; there is no need for Black to allow the queen to go to bB.

26 . . . d5 This was played in the only game here. 27.ixe4! E:xg l 2B .E:xg l dxe4

24 ... ixg2 also makes White's life easier than in the main line. 25 .Wlxg2 d5 26.Wld2 E:g6 27.Wle3 a5 (27 . . . e5 2B .Wle2;l;) 2B.E:dfl ie7 29 .Wld3 (29.b4!?) 29 . . . b4 30.cxb4 axb4 3 1 .a4;l;

25.Wle1 25 .WlgBt rJle7 does not help White. 25 .Wle3 E:h4 26.rJla l e5 27. lt:l f5 Wlxe3 2B.lt:lxe3 should be okay for Black, although his coordination is not great. Black has drawn a couple of correspondence games from this position, but White might have some chances to press in an over-the-board game.

25 .. J�1g6! Another machine-generated idea which keeps Black in the game. 25 . . . E:h4? 26.rJla l ± leaves Black poorly coordinated.

26.'lWe2 26.rJlal unfortunately runs into 26 . . . lt:l xc3! 27.Wlxc3 ixd4 2B.E:xd4 ixg2 when White has no more than a draw.

a

b

e

d

e

f

g

h

a

b

e

d

e

f

g

h

29 .E:gBt!N 29 .Wlf2 e5! 30.E:gBt rJld7! is the reason why White needs to check on gB a move earlier. 3 1 .Wlf5 t rJle7 32.lt:lc2 id5 Black managed to hold in Kupsys - Oates, corr. 2009. 29 . . . rJle7 30.Wlf2 Now Black's position is stuck, and White has multiple ways to develop his initiative. Without the computer's resourcefulness, Black's position would be collapsing almost instantly, but even with it, he is barely staying alive. 30 . . . Wlc7 30 . . . ixd4 3 1 .cxd4± 30 . . . Wld6 3 1 .E:gl ixd4 (3 1 . . .Wle5 32.lt:lc2t; 3 1 . . .ie5 32.E:fl f6 33.E:h l ±) 32.cxd4;l;

a

b

e

d

e

f

g

h

1 66

The Gelfand Variation

3 1 . lLl c2 Regrouping the knight. 3 1 .E:g l ! ? is another logical move, intending to attack along the f-file. Black barely survives with 3 1 . . .i.e5 ! 32.E:fl f6! . Even here, after 33.E:d I intending lLl c2-e3, Black's position looks unenviable. 3 1 . . .a5 3 1 . . .i.e5 32.Wfh4t i.f6 33 .Wfh6+32.E:g1 i.e5 3 3 . lLl e3 33 .Wfh4t �d6 34.E:d I t i.d5 3 5 .Wfxe4 f5 36.Wfe2;!;

27.Wfd3 and 27.Wff2 are sensible alternatives. The text move has the advantage of keeping the queen in contact with the g2-bishop, while also threatening to take on e6.

27 d5 •••

27 . . . i.xg2 28 .Wfxg2 d5 29 .Wfh3;!; leaves Black under pressure. 27 . . . i.e5 tries to maintain the tension, but White has many ways to keep a pull. I won't go into too much depth here, but will just show a few sample lines to make my point: 28.E:de 1 �e7 29 .Wfd3 i.xd4 30.cxd4 lLlf5 (3o . . . lLlh5 3 1 .i.xb7 E:xg 1 32.E:xg 1 Wfxb7 33.E:c 1 ;!;) 3 1 .Wfxf5 i.xg2 32.Wff2 Wfc6 33.';!;>a I Wff3 34.Wfd2 Wfb7 3 5 .E:c 1 ;!;

28.i.a

b

a

e

d

e

f

g

A battle of slow manoeuvring lies ahead, but White can keep poking around with his queen, or going after the g3-knight, while Black just has to wait and defend.

h

33 . . . b4 33 . 5 34.Wfh4t! �d7 3 5 .Wfh7t 36.Wfg6 Wfc8 37.lLlxf5 ± 34.axb4 axb4 3 5 . cxb4;!; White remains a slight favourite. . .

8 7 6 5 4 3 2 1 a

27.V!fc2

b

c

d

e

f

g

h

28 V!fd6 29J�de1 gg7 30.V!ff2 i.e5 31 .lLlb3 •••

�c6

3 1 .Wfh2!? is another idea. I think it's safe to say that White has the better chances.

Conclusion We have now analysed the most complex paths of the 8 .Wfe2 variation against 7 . . . lLl bd7. Like many of the more difficult Najdorf lines, it is mostly the domain of correspondence players. The main line continues 1 3 .�b l ! E:c8 1 4.E:g 1 g5 I 5 . fig5 hxg5 I 6.i.xg5 Wfb6! , although there were some deviations along the way. Now after my recommended I 7.Wfe 1 , Black can sacrifice an exchange on c3 and pick up at least one extra pawn; sometimes two. Despite being materially okay, he still suffers from a slightly exposed king, and I think I have demonstrated how White can maintain enduring pressure.

8 7 6 5 4

Classical Main Line

3 2 a

b

e

d

e

f

g

The Gothenburg Variation Variation Index l .e4 c5 2.�f3 d6 3.d4 cxd4 4.�xd4 � f6 5 . � c3 a6 6.i.g5 e6 7.f4 i.e7 8.'i'f3 h6 9.i.h4 g5 1 0.fxg5 1 0... � fd7 A) 1 1 .'i'h5!? B) 1 1 .�xe6! fxe6 1 2.'i'h5t @f8 1 3.i.b5! gh7! 14.e5!?N B 1 ) 14 ... axb5 B2) 14 ... �xe5 1 5.0-0t @g8 1 6.g6! B2 1 ) 1 6 ... �xg6 B22) 1 6 ... gg7 B23) 1 6 ...hh4!

B) after 1 3 . . . �h7!

B) note ro 1 3 . . . �h7!

a

b

e

d

e

f

1 7.�c4!N

g

1 68 170 171 174 175 176 177

h

a

b

e

d

e

1 4 .e5!?N

f

g

B2 1 ) after 16 ... tl'l xg6

h

a

b

e

d

e

1 7.�xe7!

f

g

h

h

1 68

Classical Main Line

A) 1 1 .YlYh5!?

l .e4 c5 2.�a d6 3.d4 cxd4 4.�xd4 �f6 5,c!l�c3 a6 6 ..1g5 e6 7.f4 �e7 8.YlYa h6 9.�h4 g5 8 7 6 5 4 3 2 a

1 a

b

e

d

e

f

g

h

You see before you the Gothenburg Variation, named after the famous Gothenburg tournament of 1 9 5 5 , where the Soviet trio of Geller, Spassky and Keres scored spectacular wins with White against, respectively, Panno, Pilnik and even Miguel Najdorf himself. Since then, this variation has regularly been proclaimed dead, only to be revived by some new defensive twist. Still, the Gothenburg has never been the most popular choice for Black. That is possibly due to the first impression you get from the main lines: with his whole kingside stripped apart, Black appears to be on the verge of getting mated. It is amazing that Black's audacious scheme has survived such sustained analytical assaults; I even felt sufficiently inspired to try it myself in 20 1 1 , and went on to draw without too many problems against Balogh.

10.fxg5 tll fd7

Here I will present two options for White. A)

1 1 .YlYh5!? has recently become popular, and can

be considered an unpretentious alternative to the swashbuckling sacrifices, but I was unable to make it work in quite the way I would have wished. This led me to focus on B) 1 1 .tll xe6! as my main weapon.

b

e

d

e

f

g

h

Normally allowing the knight to establish itself on e5 would give Black an excellent position, but here White has some concrete ways to challenge for this key square.

1 l ... �e5 12.�g3 This is the latest move to cause Black some worries. 1 2.t.f2 t.xg5 1 3 .h4 t.e7 1 4.0-0-0 \Wa5 1 5 .g4 tt'l bd7! 1 6.g5 tt'l g6 1 7.tt'lb3 \Wc7 1 8 .gxh6 tt'l f6 1 9 .\We2 has occurred in a few correspondence/ email games. So far Black has always recaptured the h6-pawn, but I think he can ignore it in favour of 1 9 . . . b5!?N, with good compensation.

12 ...hg5 13.�a!? tll bd7 Black should maintain a knight on e5, even if it means losing the bishop pair. Mter 1 3 . . . tt'lxf3t 1 4 .\Wxf3 Black's position is hard to hold together due to the open kingside and pressure on d6. A good example is: 1 4 . . . tt'l c6 1 5 .h4 t.f6 1 6.0-0-0 t.e5 1 7.\Wf2 E!:g8 1 8 .t.xe5 tt'lxe5 1 9 .t.e2 b5 20.E!:d2! (20.E!:d4 is also possible.) In De Filippis Yordanova, corr. 20 1 1 , White ' obtained a pleasant advantage after doubling on the d-file.

Chapter 1 0

-

1 69

The Gothenburg Variation

1 8 .gxh3 ttJ f3 t 1 9 .r;t>d 1 ttJ de5 looks too dangerous for White. The text move has a 'computerish' , correspondence feel to it. White has indeed been victorious in two correspondence games so far, but to me the position seems unclear.

18 b5! •.•

Improving over 18 . . . :gg8 1 9 .'iNd4 b5 20 . .tg3 .tb7 2 1 ..te2 hxg2 22.:gxg2 ttJ c6 23 .'iNf2 b4 24. ttJ a4;!; as seen in H. Kolev - Yordanov, corr. 20 1 1 .

19.a4 1 9 . .tg3 ! ? We3t 20.ttJe2 ttJ c4 2 1 .'iNd4 e5 22.'iNxe3 ttJ xe3 leads to an endgame where White can try to make something of his bishop pair, but Black should probably be okay.

15 ... h5! A key resource for the defence.

19 b4!N •••

16.'iNxd6 h4 17 .tfl •

A more human try is 1 7 . .txe5 ttJxe5 when White remains a pawn up, but Black controls the dark squares. The likely plan of . . . .td7 -c6 and . . . :gd8 may lead to an endgame where it will be hard for White to capitalize on his extra pawn .

This seems more logical than 1 9 . . . bxa4 as played in Czegledi - Cardoso, Internet 20 1 2.

20.ttle2 a5 Alternatives such as 20 . . . ttJ f6 and 20 . . . b3!? could also be explored.

17 h3 ..•

a

a

18J�gl !?

b

e

d

e

f

g

h

b

e

d

e

f

g

h

2 1 .�d4 'iNf4 22.�b5 ga6 23.'iNd2 'iNxd2t 24. �xd2 �e7i The position remains double-edged, despite White's extra pawn. There is a lot of room for

1 70

Classical Main Line

exploration in the 1 1 .'lMfh5 line, but overall I find it less convincing than recent results might suggest. That is why the bulk of the chapter has been devoted to the following piece sacrifice.

B) 1 1 .tLlxe6! fxe6 12.Y!YhSt c;!;>f8 8 7 6 5

13 . . . lLle5 1 4.i.g3! E!:h7 14 . . . i.xg5? 1 5 .0-0t rJie7 1 6.i.xe5 'lMfb6t 1 7.rJih l dxe5 I B .'lMff7t rJid6 1 9.E!:ad l t Black was soon mated in the stem game Geller Panno, Gothenburg 1 9 5 5 . 1 5 .i.xe5! dxe5 1 6.E!:d 1 i.d7 1 6 . . . lLl d7 1 7.'lMfg6 E!:f7 I B .'lMfxh6t �gB 1 9 .96 E!:g7 20 .i.c4 'lMfb6 2 1 .lLld5 was winning for White in Ninov - Spasov, Bulgaria 1 9B9. The text move leaves White with a few promising continuations, but the most convincing looks to be: 8

4

7

3

6

2

5

1

4 a

b

e

d

e

f

g

h

13 .ibS! •

The bishop is essential on b 5 to challenge Black's control over the e5-square. If he could establish a knight on e5 then all would be well, but now . . . lLle5 can be met by i.g3 , and the b5-bishop is ready to eliminate the other knight if necessary. Compare the line 1 3 .i.c4 lLle5 1 4.i.g3 lLl bc6! when Black defends successfully. The text move was played in all three of the Gothenburg 1 9 5 5 games (credit goes to Geller for being the first to find it over the board) , and was an amazing idea for its time. The constant mating threats make the bishop immune from capture for the time being, but Black has a lot of resources too.

13 3h7! •• J

This was used by Fischer to draw with Gligoric in 1 9 5 B , and remains Black's best, and indeed only, hope to make this variation work. The two alternatives can be dealt with quickly, and are mainly of historical interest:

3 2

a

b

e

d

e

f

g

h

1 7.i.c4!N rJig7 Neither 1 7 . . . 'lMfb6 1 B .E!:d3! nor 1 7 . . . i.xg5 I B .i.xe6 offers much hope for Black either. 1 B .g6 E!:hB 1 9 .i.xe6 White should win. 1 3 . . . rJig7 This leads to a long sequence of forced moves (on Black's side at least) , culminating in a winning position for White. 1 4. 0-0 lLle5 1 4 ... 'lMfgB 1 5 .g6 i.xh4 (Or 15 ... lLl f6 1 6.i.xf6t i.xf6 1 7.e5 i.xe5 Stein - Zheliandinov, USSR 1 9 56, I B .i.d3N followed by E!:f7t winning.) 1 6.i.xd7 lLlxd7 1 7.'lMfxh4 'lMfdB 1 B .E!:f7t rJixg6 1 9 .E!:e7 +- Vasilchuk - Suetin, USSR 1 9 56. 1 5 .i.g3 lLl g6 1 6.gxh6t E!:xh6

Chapter 1 0

-

The Gothenburg Variation

171

accurate than h4 on move 23, although White should eventually win either way. Let's return to 1 3 . . . l:'!h7!. This position has been reached in well over a hundred games and analysed for more than half a century, so I was especially happy to discover a promising novelty.

a

b

e

d

e

f

g

h

1 7J'!f7t! A key detail. 17 . . . lt>xf7 1 8 .�xh6 axb5 1 9 .1:'!f1 t i>e8 20.�xg6t It>d7 2 1 .l:'!f7 lD c6 22.lDd5 l:'!xa2 23.h3! 23.h4 �h8? 24.lDxe7 lDxe7 2 5 .�g5 also led to a quick win for White in Keres - Naj dorf, Gothenburg 1 9 5 5 . The text move is slightly more accurate, although it should not matter too much.

a

b

e

d

e

f

g

h

14.e5!?N

a

b

e

d

e

f

g

h

23 . . . l:'!xb2 23 . . . �h8? 24.lDxe7 lD xe7 2 5 .�g5 l:'!al t 26.lt>h2 �d8 27.�xb5t i>c7 28 .�c5t was another rout in Spassky - Pilnik, Gothenburg 1 9 5 5 . 2 3 . . . exd5 24.�xd6t i>e8 2 5 .�g6 i>d7 26.exd5 l:'!a6 27.�e4 leads to the same final result. 24.lt>h2 l:'! b l 25 .�f6+Followed by lDxe7 and �h4; the latter move shows why moving the pawn to h3 was more

The natural 1 4.0-0t has been overwhelmingly the most popular choice. I could fill several pages with analysis of this move, but it will suffice to say that I am not altogether convinced by White's chances. My new move has the simple idea of opening further lines of attack, while trying to delay the check along the f-file until the most inconvenient moment for Black. There are two logical replies: BI) 14 ...axb5 and B2) 14... �xe5.

BI) 14 ... axb5 15.exd6! We should continue to delay castling for the moment. 1 5 .0-0t It>g8 1 6.g6 100ks tempting, but Black has a decent reply in 1 6 . . . ixh4! ( 1 6 . . . l:'!g7 1 7.l:'!f7! gives White a powerful initiative) 1 7.gxh7t It>h8. The position is rather messy, but Black's king is safe for now.

1 72

Classical Main Line 2 1 . . .:gf7t 2 1 . . .:ga4? is refuted by 22.tDg5!' 22.<;!;gl c;!{g7 23 ."lMg4t <;!;h7 24."lMxe6 c;!{g7 2 5 ."lMg4t <;!;h7

1 5 ....bg5

1 5 . . . .ixd6? 1 6.g6 wins trivially.

16.,L:g5 YlYxg5 17.0-0t! Perfect timing.

8

17 ... tLlf6

Black has little choice, a s 1 7 . . . <;!;g7 I B ."lMeB! "lMg6 1 9."lMxcB :ghB 20."lMxb7 :ga6 2 1 .:gf3 :gb6 22."lMc7 "lMxc2 23 .:gg3t <;!;h7 24.:gfl gives him little hope of survival.

7 6 5 4 3

8

2

7 a

6 5 4 3 2 1 a

b

e

d

e

f

g

h

1 8J�xf6t! YlYxf6 19J:Ul YlYxfl t

1 9 . . . tD d7 should be met by 20.tDe4! (20.:gxf6t tD xf6 2 1 ."lMg6 :gf7 22."lMxh6t <;!;eB! is good enough for equality) 20 . . . "lMxfl t 2 1 . <;!;xfl , transposing to the note to Black's next move in the main line below.

20.�xfl In material terms Black is more than okay, with two rooks and a piece for the queen, but his king is awful and his development nonexistent. White's queen and knight will work together in perfect harmony, enabling him to gobble up a lot of pawns and ultimately reach a promising endgame.

20 . :gf7t .

.

20 . . . tD d7 2 1 . tD e4 2 1 . <;!;e2!? is an interesting alternative, aiming to keep the rook stuck on h7.

b

e

d

e

f

g

h

26.h4!? White has a few ways to exert pressure, and there is certainly no need to go for the forcing lines as suggested by the computer. Nevertheless, from a theoretical point of view the main line seems most clear-cut. 26."lMh4 <;!;g7 27."lMg3t <;!;h7 2B.h3!? :gxa2 29 ."lMd3 <;!;gB 30.<;!;h2 is a decent alternative. White maintains a more pleasant position, and Black will have to struggle for a draw. 26 . . . :gxa2! 26 . . . tD e 5 ? 27."lMh5! :gfS 2B ."lMeB wins. 26 . . . b4 27."lMh5! <;!;g7 (27 . . . :gg7 2B.tDg5t <;!;gB 29."lMxh6±) 2B.g4! tD f6 (2B ... :gxa2 29.g5 hxg5 30.tDxg5 +-) 29.tD xf6 :gxf6 30."lMeB! :gf7 3 1 .h5 The queen is absolutely dominating the position. 3 1 . . . b6 32.d7! :gxd7 33."lMg6t <;!;hB 34."lMxh6t The kingside pawns will be unstoppable. 27.h5 :gg7 2B ."lMfSt c;!{hB 29 ."lMe6!? Black's minor pieces are both stuck, while his rooks remain uncoordinated. Obviously Black still has chances for a draw, and we should remember that fortress defences are often underestimated by computers. Nevertheless, it seems that Black will have to suffer for a while longer, while White gets to choose between pleasant alternatives.

Chapter 1 0 - The Gothenburg Variation

22 .tl�d7

8

1 73

••

After 22 . . . ttJ c6 possibilities:

7 6

23.g5

Black

has

two

a) 23 . . . hxg5 24.�xg5t �f8 Now we get to see a nice example of some effective zig-zag queen manoeuvring.

5 4 3

8

2

7

1

6 a

b

e

d

e

f

g

h

21 .�e2! 2 1 .�gl also looks interesting, but after 2 1 . . .mg7 22.�e5t �h7! Black continues to hang by the edge. (22 . . . �g8 23.ttJe4! is more dangerous for him.) 23 .�e4t (23 . ttJ e4 ttJ c6 24.�h5 mg7°O) 23 . . . �g7 24.�d4t �g8 ! (24 . . . mh7? 25 .h4!±) 2 5 .�d3 �g7 26.h4 ttJ d7! 27.h5 �f5oo

21. .. �g7 22.g4! The simple plan of g4-g5 makes White's position easy to handle. Black is still surviving, but he is clearly a long way from equality. In the upcoming variations the computer suggests various alternatives for White, but I will just present a few illustrative lines to show how White can continue to exert pressure.

5 4 3 2

a

b

e

d

e

f

g

h

2 5 .�h6t �e8 26.ttJxb5 �xa2 27.�h5! ttJ d8 28 .�h8t �d7 2 9 . ttJ c7!± b) 23 ... ttJ d4t 24.�d3 ttJ f5 2 5 . gxh6t ttJ xh6 26.�g5t �h7 27. ttJ e4 8 7 6 5 4 3 2

a

b

e

d

e

f

g

h

27 . . . .id7 (27 . . . ttJ g8 28 .�xb5±) 2 8 . ttJ f6t �xf6 29 .�xf6;!; The game goes on, but White's pawns give him the better chances.

23.g5 hxg5 24.Yfxg5t �h7 24 . . . �h8 2 5 . ttJ e4 �a4 (25 . . . �f5 26.�h6t �g8 27.�xe6t +-) 26.�h 5 t �g7 27.ttJg5+a

b

e

d

e

f

g

h

1 74

Classical Main Line

B2) 14 ... �xe5

25.�h5t! White still requires some precIsion to maintain the pressure. 2 S . lD e4 !!fS!f± is less convincing.

25 ... �g7 26.�g4t �f8 27.�e4 �f6

27 . . . !!a4 can b e met b y 28 .'1Wg6, threatening lDgS. 28 . . . !!g7 (28 . . . !!xa2 29 .Wfh6t i>g8 30.Wfxe6+-) 29 .Wfxe6 Black's position remains unpleasant.

8 7 6 5 4 3 2 1 a

b

e

d

e

f

g

h

1 5.0-0t �g8

l S . . . i>g7? leaves Black awfully placed on the kingside, and he is swiftly punished: 1 6 . .ig3 ! axbS 1 7 . .ixeSt dxeS 1 8 .!!f7t i>h8 1 9 .!!xh7t i>xh7 2o .Wfxh6t i>g8 2 1 .!!f1 +-

16.g6! a

b

e

d

e

f

g

h

28.�h4! �g8 If Black has to play such a move, he clearly isn't having a good time.

29.�d8t �g7 30.tl�c5 30.h4!? also deserves attention.

30 ... b6 3 1 .�xe6t 3 1 .Wfxb6 eS!1X> allows Black's bishop to become active.

3 1 .. ..ixe6 32.�xa8 �h6 33.�c6 �f5 34.�d2 gd7 35.�c1 gxd6 36.�xb5;t White can continue to press in the endgame without any risk.

I also considered: 1 6 . .id3 ! ? I t i s tempting t o remove the annoyingly strong eS-knight, but it only leads to some nice tactical lines where Black manages to hold on with precise defence. 1 6 . . . lD xd3 1 7.g6 !!g7 1 7 . . . .ixh4? is impossible now that we have diverted the knight, and 1 8 .gxh7t wins easily. 1 8 . .ixe7 Wfxe7 1 8 . . . Wfb6t!? 1 9 .i>h l lDf2t 20.!!xf2 Wfxf2 2 1 . lD e4 Wff4 22. lD f6t i>h8 23.lDg4 !!xe7 24.lDxh6 lD c6 is another wild line, where it seems that White will eventually have to settle for j ust a draw. (However, 24 . . . c;t>g7? would be a fatal error due to 2S.lDg4! c;t>g8 26.Wfh4 when White wins the queen and the game.)

1 75

Chapter 1 0 - The Gothenburg Variation

1 7.Wxg6t E:g7 1 8 . .ixe7 Wb6t! 1 9.�h l E:xg6 20.E:f8 t �h7 2 1 .E:f7t �g8 is only a draw.

17 ... tihe7 Both 1 7 . . . E:xe7 1 8 . lLl e4! and 1 7 . . . Wxe7 1 8 . lLl e4! axb 5 1 9 .Wxg6t �h8 20.E:f3 are hopeless for Black.

18 ..id3 b

a

e

d

e

f

g

h

1 9 .1Lle4 lLl d7 20.cxd3 d5 20 ... e5 2 1 .Wxh6 We6 22.l''U7 ! Wxg6 23 J:%xg7t Wxg7 24.We6t �h8 2 5 . lLl xd6 is dangerous for Black. 2 U !f7 E:xf7 22.gxf7t Wxf7 23 .Wxh6 Wg7! Mter 23 . . . dxe4 24.E:f1 Wg7 2 5 .Wxe6t �h8 26.We8t Wg8 27.Wxe4 Black is in trouble, despite the two extra pieces. 24.Wxe6t Wf7 2 5 .Wg4t Wg7= White has nothing more than a draw.

Now White j ust needs to play natural moves to increase his advantage. There are multiple ways of doing this, so I will j ust include a few illustrative lines.

1 8 .. JIg7 1 8 . . . lLl f5 1 9 .E:ae l .id7 20.E:xe6 .ixe6 2 1 .Wg6t �h8 22.Wxe6 lLl d7 23 .Wxf5 ± Material is almost level, and Black's king is too exposed.

Returning to the main line, we have reached a critical position where Black has three plausible moves: B21) 16 .. .tihg6, B22) 16 .. J!g7 and B23) 16 ... hl4!. Only the last option offers serious resistance.

B2 1) 16 . . �xg6 .

8 7 a

6

b

e

d

e

f

g

h

19.YlYxh6 d5 20J:�ae1 ttlf5 2 1 ..ixf5 exf5 22J!f3 ttlc6 23J�g3!

5

Eliminating the key defensive piece.

4 3

23 .. J!xg3 24.hxg3 .id7

2

24 . . . �f7 2 5 .Wh5 t �f8 26.g4 lLl e7 27.Wh6t � f7 28 .g5 is crushing.

1 a

17.Le7!

b

c

d

e

f

g

h

25.ttlxd5

1 76

Classical Main Line

18 ..id3

8

Safeguarding the bishop and overprotecting the g6-pawn. White will soon develop a lethal attack around this tower of terror.

7 6 5

18 ... c;!?g7 I B . . . ltJxd3 1 9 .cxd3 �g7 20.E:f4+-

4 3 2 1 a

b

e

d

e

f

g

h

25 ...Yff8 26.�f6t c;!?f7 27.Yfxf8t gxf8 28.�xd7±

If Black gives up the h6-pawn without a fight then he will soon be crushed, for instance: I B . . . E:g7 1 9 .Wfxh6 ltJ bd7 ( l 9 . . . ltJg4 2o.Wff4 ltJe5 2 1 .i.e4 ltJ bd7 22.E:ad 1 +-) 20.E:f4 ltJfB (20 . . . ltJ f6 2 1 .E:afl ltJ eg4 22.Wfh4+-) 2 1 .E:afl ltJ Exg6 22.i.xg6 ltJxg6 23.E:g4 WfeB 24.ltJe4+-

Black has avoided being mated, but his endgame prospects are dismal.

B22) 16 .. J3g7

a

b

e

d

e

f

g

h

19.9f4! i.d7

a

b

e

d

e

f

g

h

This move looks obvious but it is too slow, and White easily builds a strong attack.

17.he7 gxe7 1 7 . . . Wfxe7?! goes down more quickly: I B . ltJ e4! E:xg6 ( l B ... ltJ bd7 1 9 .i.xd7 ltJ xd7 2o.Wfxh6+-) 1 9 .i.eB! Wfg7 ( l 9 . . . E:g7 20.ltJ f6t+-) 20 .i.xg6 Wfxg6 2 1 .Wfh4 Material is close to equal, and White still has a crushing initiative.

1 9 . . . ltJ bc6 20.E:afl i.d7 2 1 .Wfh4 should end in the same result. If Black tries 2 1 . . .d5 to stop the knight from going to e4, then 22.ltJe2! sets it on another path towards the kingside.

20.gafl 20.ltJe4 is also promising.

20 ....ic6 21 .gf8 Yfxf8 Mter 2 1 . . .Wfb6t 22.�h l Wfe3 23 .Wfh4 Black will have to give up his bishop to force a queen exchange, and the pin on the eighth rank will decide the game.

Chapter 1 0

-

The Gothenburg Variatio n

8 7 6 5 4 3 2 1 a

b

e

d

e

f

g

h

1 77

Black can hardly afford to exchange it, as the loss of the knight on e5 will give White too many squares, especially fl. 1 8 . . . ttJ bd7 1 8 . . . .te7 1 9 .9ae l gives Black nothing better than 1 9 . . . ttJ bd7, when 20.ttJe4 transposes to 1 8 . . . ttJ bd7 1 9 . ttJ e4 .te7 20.gae l as given below. 1 8 . . . .tg5 1 9.9ae l ttJ bc6 runs into 20.h4! .td2 (20 . . . .txh4 2 1 .�xh4! is one example of why the h7-pawn matters) 2 1 .ge2 .txc3 22.bxc3 .td7 23 .ge3 ttJ e7 24.gg3 and White wins cleanly.

24.�h4! gd7 25.�f6 ti)xd3 26.�xe6t �g7 27.cxd3+Black's queenside pieces are stranded, and he has no good answer to ttJ e2-g3 etc.

B23) 16 ...hl4!

a

a

b

e

d

e

f

g

h

AI; mentioned before, this is the only move to present a real challenge.

17.gxh7t ci>xh7 1 7 . . . Wh8?! Sometimes the pawn on h7 can act as a hindrance to White's attack, but here it only helps him. 1 8 . .td3! The bishop is perfectly placed here, but

b

e

d

e

f

g

h

1 9 . ttJ e4 .te7 1 9 . . . d5 20.ttJd6 .te7 2 1 .ttJflt ttJxfl 22.gxfl is crushing. 20.gae l d5 2 1 . ttJ f2 ! The knight blocks the f-file, b u t only temporarily. 2 1 . . . .td6 2 1 . . .ttJ xd3 ? 22.ttJxd3 wins more quickly. 8 7 6 5 4 3 2

a

b

e

d

e

f

g

h

1 ?8

Classical Main Line

22.ltJg4! ltJ xg4 23.Wixg4 ltJ e5 24.Wid4 Wie7 2 5 . <;!f h 1 2 5 . b4 Wig? 26.1'he5 .id7 27J!fe 1 ± i s also strong. 25 . . . Wig7 26.l:!xe5 .id7 27.l:!f6! A nice touch, with the simple tactical point: 8 7 6 5 4

pragmatic choice, but also the strongest in an absolute sense. White can go for an all-out attack with 1 9.93?!, but it ultimately falls short: 19 ... .ie7! (But not 1 9 . . . .ig5 ? 20.l:!ae l ! ltJ bc6 2 l .h4 .id2 22.l:!e4± and White dominates.) 20.l:!f7t (20.l:!f2 ltJ bd7! 2 l .ltJe4 b5 22.l:!afl WigB defends everything.) 20 . . . ltJxf7 2 l .Wig6t <;!ffB 22.l:!fl Wib6t! 23. <;!fg2 <;!feB 24.l:!xf7 ltJ c6! 25 . .ifS <;!fdB! 26 . .ixe6 .ixe6 27.Wixe6 Wixb2 2B.ltJd5 Wie5+ Black defends successfully, and emerges with some winning chances.

3

19 ... ltJbc6

2 b

a

e

d

e

f

g

1 9 . . . .ie7? 20.<;!fh l ! is too dangerous, for instance 20 . . . ltJ bc6 (or 20 . . . .id7 2 l .l:!f3! ltJxf3 22.gxf3 .ieB 23.l:!gl t .ig5 24.Wih4+- with a crushing attack) 2 l .l:!ae 1 ltJ d4 22.c3 ltJf5 23.g4+- and White wins material.

h

27 . . . Wixf6 28.l:!xd5 Wixd4 29.l:!xd4 Simplifying to a winning ending.

20.t£\xd6 White should take the opportunity to capture a valuable pawn. Instead after 20.<;!fh I ? d 5 2 l . ltJ d6 Wixd6 22 .Wixh4 Wie? his initiative is dead.

2o ...Wixd6 21 .VNxh4

a

b

e

d

e

f

g

h

18.j,d3t

1 8 . .ieB is a cool-looking move, but after 1 B . . . Wig5 1 9 .Wie2 Wie7°o the position gets messy.

18 ... �g7 1 B . . . ltJxd3 1 9 .1:!f7t is the usual idea.

19.t£\e4! Preparing ltJxd6 is not only the most

a

b

e

d

e

f

,g

h

2 1 ...VNd4t Black had better exchange queens of course.

Chapter 1 0

-

1 79

The Gothenburg Variation

After 2 1 . . .Wle7 22 .Wlg3t Wlg5 23.WlfL. .id7 24 . .ie4t Bla�k faces an uncertain future, as the a I -rook will soon make its way towards the kingside.

23 . . . ttJ e2t 24.i>h l does not achieve anything for Black. 23 . . . ttJ dc6 24.l:'l:ae l .id7 transposes to 23 . . . .id7 24.l:'l:ae I ttJ dc6 as analysed below.

22.\Wxd4 �xd4 It looks as though Black has escaped the immediate problems, and can settle into a safe endgame, but. . . not so fast! Actually White is still able to generate a lot of threats due to Black's lack of development and still­ vulnerable king. Even without those factors, the material balance of rook and pawn versus two knights is not so bad, and if White gets a chance to pick up the e6-pawn, he will obtain a clear, long-term advantage.

23 . . . ttJ f7 24.l:'l:ad l ttJ f5 2 5 .l:'l:d3 ttJ g5 (25 . . . ttJ 7d6?! 26 . .ixf5 ttJ xf5 27.g4 ttJ e7 2B.l:'l:df3 ± ; 2 5 . . . l:'l:bB 26.l:'l:c3t Once again White finds a way to infiltrate. The rook will be hard to deal with once it arrives on c7.) 8 7 6 5 4

8

3

7

2

6 5

a

b

e

d

e

f

g

h

26 . .if3!t It is hard for Black to complete his development, and the knights are stuck in strange positions on the kingside. (26 . .ixfS exfS 27.l:'l:dB looks tempting, but after 27 . . . ttJ f7 2B .l:'l:eB ttJ d6 i t is impossible to maintain the rook on the eighth rank.)

4 3 2 1 a

b

e

d

e

f

g

h

23.i.e4! 23.l:'l:ad I is less convincing: 23 . . . � dc6 24 . .ie4 .id7 25 . .ixc6 (25 .c3 is playable, but it does not stop Black from consolidating and equalizing.) 25 . . . .ixc6 26.l:'l:de I ttJ c4 27.b3 ttJa3 The knight looks slightly awkward, but after 2B.c4 l:'l:dB ! ? Black gets enough counterplay.

23 ... �f5! Preventing the desirable l:'l:ad 1 . I considered several other moves, none of which offer a complete solution to Black's problems.

23 . . . ttJ c4!? This is an active approach, but Black does not seem to be completely ready for it. Particularly annoying for him is the threat of penetrating with a rook to dB. 24.l:'l:ad l ! 24.l:'l:f2!? is another way of maintaInIng pressure: 24 . . . ttJ f5 (24 . . . ttJ d6 2 5 .l:'l:e a) 2 5 .l:'l:e l ttJ xb2 (25 . . . ttJ cd6 26.l:'l:d2 Black's pieces are still awkward and his development remains hard.) 26.l:'l:f4! Threatening to win a piece with .ixfS and l:'l:b l . 26 . . . ttJ d6 27.l:'l:g4t i>f6 2B.l:'l:g6t i>e5 29 . .if3t i>d4 30.l:'l:xh6t The kingside pawns will be hard to stop.

1 80

Classical Main Line

8 7 6 5 4 3 2

a

b

e

d

e

f

g

h

24 . . . e5 24 . . . ltJ e2t 2 5 . \t> f2 ltJ f4 26J�d8± 2 5 .l'!d3 i.g4 26.i.xb7 l'!a7 27.i.d5 ltJ xb2 28 .l'!e3 ltJ d l After 2 8 . . . ltJ e2t 29.\t>h l ltJ f4 30.i.b3;!; the knight is out of play on b2. 29.l'!xe5 ltJ c3 30.i.b3;!; White keeps both of his extra pawns safe. 23 . . . i.d7 This looks like the most obvious developing move, but it also leads to some problems. 24.l'!ae l ! Now Black has to make a difficult choice.

25 . . . ltJ f7 2 5 . . . i.e8 26.i.xc6 ltJxc6 27.l'!xe6 reaches a depressing endgame for Black. Typically, scenarios involving rook versus two minor pieces, with three pawns against two on one flank and two against one on the other, can be considered practically lost for the minor pieces (with the possible exception of rook against a pair of bishops) . Here White will have to spend a few moves consolidating, while making sure that Black's rook does not combine with the pieces to create counterplay, but once that is done, he can simply go about improving his position at Black's expense. 8 7 6 5 4 3 2

a

8 7 6 5 4 3 2 a

b

e

d

e

f

g

h

24 . . . ltJ dc6 This seems best. Neither 24 . . . i.c6 2 5 .c3! i.xe4 26.cxd4± nor 24 . . . i.b5 2 5 .l'!f2 ltJ g4 26.l'!d2 l'!d8 27 .i.xb 7 ± is any good for Black. 2 5 .l'!e3! Threatening a nasty check on g3 , while Black's pieces remain hard to untangle.

b

e

d

e

f

g

h

26.l'!d l ltJ ce5 26 ... i.e8 27.i.xc6 i.xc6 28.l'!xe6 reaches the same type of endgame as discussed above. 27.l'!c3! i.c6 After 27 . . . l'!b8 28Jk 7 i.c6 29.i.xc6 bxc6 30.b3 Black's knights are clumsily placed, and he will find it hard to hold on to the many weak pawns. 28 .i.xc6 ltJxc6 28 . . . bxc6 29.b3;!; is similar to the last note. 29 .l'!d7 l'!b8 30.l'!e3 ltJ cd8 3 1 .l'!g3t I1If8 32.l'!c7;!; The rooks are tremendously active, and it will be hard for Black to do anything without losing another pawn. Let's return to the more challenging 23 . . . ltJ f5 ! .

Chapter 1 0

-

IBl

The Gothenburg Variation

8

8

7

7

6

6

5

5

4

4

3

3

2

2 a

a

b

e

d

e

f

g

h

24J::� ael!

The alternatives 24.E:fe l lLl f7 2 5 .E:ad l lLl 5d6 26.if3 E:bB and 24.E:fd l lLl f7 2 5 .E:d3 @f6 26.E:e l lLl 5d6 27.E:f3t @e7 both left me unsatisfied.

24 lLlfl ...

24 . . . lLl c4 25 .E:f4! lLl cd6 (25 ... lLlxb2 26.ixf5 exf5 27.E:b l +-) 26.ixf5 lLlxf5 (26 . . . exf5 27.E:d4 lLle4 2B .E:dB±) 27.E:c4;!; leaves Black with typical problems for this line. 24 . . . @f6!? This looks rather provocative, but there is no immediate refutation. 25 .E:d l 25 .E:e2!? id7 26.E:fe 1 could be an interesting alternative. 25 . . . id7 26.ixb7! ? Black is in time to refute 26.g4? with 26 . . . E:gB!. 26.h3 is another possibility: 26 ... h5 27.E:fe l ic6 2B .ixf5 �xf5 29.E:d2! lLl g6 30.E:f2t @g5 3 1 .E:xe6 h4 For the time being, the fixing of White's kingside makes it hard for him to make a passed pawn. However, it should be possible to continue improving his pieces and eventually play g2-g3 after suitable preparation. In short, White has reasonable winning chances. 26 . . . E:bB 27.ixa6 E:xb2

b

e

d

e

f

g

h

2B .g4! Now that . . . E:gB is impossible, this wins material. 2B . . . E:xa2 2B . . . lLlxg4 29.E:xd7± The simplest reply to 2B . . . ic6 is 29 .ie2. (29.gxf5 ! ? lLlf3t 30.�f2 E:xc2t 3 1 .�e3 E:xa2 32.E:a l E:xh2 33.fxe6 �xe6 34.E:fd l ! reaches an extremely interesting endgame. The black king is forced to go to the last rank after 34 . . . lLle5 3 5 .icBt �e7 36.E:a7t �eB and, with both of his rooks on the board, White has plenty of scope to create mating threats. 29.ie2 E:xc2

a

b

e

d

e

f

g

h

30.E:f2! E:c5 3 1 .gxf5 exf5 It will not be easy for White to make progress from here, but he can certainly keep trying for a while, for instance by slowly improving his king. A rook exchange may help to facilitate this. I do not see any fortress, so Black will have to defend carefully and react to White's ideas.

1 82

Classical Main Line

After the text move White must decide whether to start poking around with his rook immediately, or to play more slowly.

a

b

e

d

e

f

g

26.:!'!c3 :Bb8 I also considered 26 . . . tLlxe4 27.:Bxe4 e5 28 .:!'!c7 1i7f8, when Black intends . . . b5 and . . . i.e6. For the moment Black has solved his development problems, but there are still quite a few weaknesses in his position, and White's rooks are rather active. The passed e-pawn is not too dangerous as the white king can always blockade it. In the event of a rook exchange, Black's pawn weaknesses are likely to come into sharp focus.

h

25.lH3!?

Aiming to put maximum pressure on the opponent with :!'!c3-c7.

2 5 . b3 is the other main idea, preparing c2-c4 at the right time. (The immediate 2 5 . c4 is met by 25 . . . tLl 5d6 26.i.d3 b6, when Black is able to develop his queenside pieces more easily than I would like.) Black has to resort to some sophisticated manoeuvring, but he seems to be okay after 25 . . . tLl 5d6! 26.i.f3 :!'!a7!, while preparing to move the b-pawn, and ensuring that a white rook will not be able to invade on c7. White can continue probing, but the position seems tenable for Black.

25 ... ttl 5d6 25 . . . i.d7? would be a big mistake due to 26.i.xf5 exf5 27.:!'!e7±. A couple more short but instructive lines are 25 . . . tLl 7d6 26.i.xf5 ! tLlxf5 27.:!'!c3t and 25 . . . tLl g5 26.:!'!d3! tLlxe4 27.:!'!xe4t, both of which see White swapping off his bishop in order to invade with a rook to either the seventh or eighth rank.

a

b

e

d

e

f

g

h

29.:Bb4 The most direct idea, intending to weaken Black's queenside. (White should keep an edge with other moves too, for instance 29.:Be3 b5 30.1i7f2 i.e6 3 l .b3, intending to meet 3 1 . . . :BdB ?! with 32 .:!'!d3! :!'!xd3 33.cxd3 ±.) 29 ... b5 30.a4 bxa4 3 1 .:!'!xa4t

27 ..ia

Safeguarding the bishop and introducing the possibility of i.h5 combined with :!'!c7.

27.:!'!c7?! is not much good due to 27 . . . tLlb5, but 27.:!'!g3t Ii7fB 2B .i.h7!? li7e7 29.:BgB is interesting, as Black will have to go to some trouble to escape the pin.

27 f6 •••

After 27 . . . tLlb5 2B .:Bd3 tLl bd6 29.:!'!ed l tLl f5 30.i.h5 tLle5 3 1 .:!'!c3t Black's shuffling of the knights has not helped him.

Chapter 1 0 - The Gothenburg Variation

1 83

27 . . . b5 also falls short of equality: 28.1'' k 7 'iil g6 29 .ig4!? e5 ,<29 . . . 'iil f6 3o.ih5!? ttJ e5 3 1 .�d l �b6 32.�xd6 �xd6 33.�xc8::!;) 3o.ixc8 �xc8 3 1 .�xc8 ttJxc8 32.�e3!::!; The black pawns are sitting ducks, and knights are well known to be hard to coordinate against a rook in the endgame. Moreover, the passed e-pawn is not much of a worry with the white king so close.

29J�c7t .id7 30.c4!? rJJ d8 3 1 .�xd7t rJJ xd7 32.c5 gc8 33.cxd6 �xd6 34. rJJ f2

28 ..ih5!?

Even by the standards of the Naj dorf, the Gothenburg Variation is rather provocative, yet so far it has defied all of White's attempts at outright refutation. Recently I I .Wlh 5 ! ? has been used by some strong players. Although Black does not have an easy time, I was not completely satisfied with White's chances. For this reason, I devoted most of my efforts on the sacrificial main line with I l .ttJxe6! fxe6 1 2.Wlh5t 'iil fS 1 3 .ib5 ! �h7!, after which my novelty 1 4 .e5!?N poses serious problems to the defence. 1 4 . . . axb5 is one obvious answer, but then the rather forcing 1 5 .exd6! ixg5 1 6.ixg5 Wlxg5 1 7. 0-0t! ttJ f6 1 8 .�xf6t! Wlxf6 1 9 .�fl Wlxfl t 20.'iil xfl reaches an endgame where White's queen and extra pawns enjoy the upper hand over Black's uncoordinated mob of pieces. 1 4 . . . ttJ xe5 can arguably be considered the main line, when 1 5 . 0-0t 'iil g8 1 6.g6! again forces Black's hand. The best defence is 16 . . . ixh4!, when 1 7.gxh7t 'iil x h7 1 8 .id3t 'iil g7 1 9 . ttJ e4! ttJ bc6 20.ttJxd6 Wlxd6 2 1 .Wlxh4 Wld4t 22.Wlxd4 ttJ xd4 reaches another type of favourable endgame, where White's early advantages in development and mobility continue to make themselves felt.

This is the most forcing route, though not the only good one. 28 .�c7 ttJb5 29.�c4 is another way to break Black's co-ordination, while the white pieces can continue to j ump around and exert pressure, for instance: 29 . . . id7 (or 29 . . . e5 30.�fl ) 30.m4t 'iil e7 3 1 .c4 ttJ c7 32.ih5::!;

28".rJJ e7

28 . . . ttJe5 29.�d l ! soon gets nasty for Black after 29 . . . 'iil e7 30 .�g3 ttJ f5 3 1 .�g8 ± .

a

b

e

d

e

f

g

h

Material is level, but White keeps certain advantages in the form of the better minor piece, the outside pawn majority and a target on h6.

Conclusion

Classical Main Line Three-Piece System with ...h6 Variation Index l.e4 cS 2.tL!a d6 3.d4 cxd4 4.tL!xd4 tL!f6 S.tL!c3 a6 6..igS e6 7.f4 .ie7 8.Y;Ya h6 9..ih4 Y;Yc7 10.0-0-0 10...tL!bd7

A) 11.5!? A I ) l I tL!eS A2) 11. 0-0!?N A3) l I eS

18S

B) 11..ie2

192

186

••.

187

•.

189

•..

B l ) 11..J:�g8?

192

B2) 11. gS?!

193

B3) 11..J:�b8?!

19S

B4) l I

198

••

••.

bS

BS) 11. 0-O!?

200

•.

B2) after 16 ...�d7

a

b

e

d

e

f

17.%Vel!N

g

B3) note to 12 .. j"1g8

h

B3) after 16 ...lt>d8

8

8

7

7

6

6

5

5

4

4

3

3

2

2

a

b

e

d

e

19.93!N

f

g

h

���?�1;1���Jli\?!;!ii�j F""��;;;:;;/"",,� m,�o 'm"JOO"'Oo_', ,../"".' r":?" �,,"o�[-7 m'o/N:?�' MM, o'�:?N••,,",. 1

a

b

c

d

e

f

17.%VO!N

g

h

185

Chapter 1 1 - Three-Piece System with . . . h6

l.e4 cS 2.«�f3 d6 3.d4 cxd4 4.tihd4 c!tlf6 S.c!tlc3 a6 6.AgS e6 7.£4 Ae7 8.%Yf3 h6 9.Ah4 %Yc7 io.o-O-o 8 7 6 5

1 2 . . . 4Jxc3 (Or 1 2 . . . �xe7N 1 3 .4Jxe4 d5 1 4. 4J c3 and White wins a pawn.) 13 . .ixd6 4Jxa2t (After 1 3 . . . �c6 1 4.�xc3 �xc3 1 5 .bxc3 ± Black has severe problems with the dark squares, and he is also a pawn down.) 1 4.Wb l 4J c3t 1 5 .bxc3± Black's dark squares are too weak and he went down in Blanes Conde Poderoso, email 1 99 8 . 8

4

7

3

6

2

5

1 b

a

e

d

f

e

g

h

My "Three Piece" terminology refers to Black's set-up on the seventh rank after his next move. He can play this way either with or without the moves . . . h6 and .ih4 included, the latter scenario being the focus of the next chapter. Both versions are somewhat out of fashion these days, although they can still be seen in grandmaster games from time to time.

IO ... c!tlbd7 1 0 . . . 4Jxe4? has been played in quite a few games, but hardly anyone managed to find the clear refutation: 1 1 .4Jxe6!N ( I 1 .�xe4 .ixh4 1 2 .4Jf5 .ie7 1 3 .4Jxg7t Wd8 was not so simple in Deepan Chakkravarthy - Wang Yue, Kuala Lumpur 2002.) 1 1 . . . .ixe6 ( I 1 . . .fxe6? 1 2.4Jxe4 .ixh4 1 3 .�h5t+-) 1 2 ..ixe7

4 3 2 1 a

b

e

d

e

f

g

h

The present position can be considered the real starting point of the chapter. Black is combining the three-piece set-up with the thematic plan of . . .g5, which gave him excellent results in early outings. However, once White figured out the correct plans, Black's prospects diminished considerably. I will start by presenting A) l l .f5!? as an interesting sideline, but my main recommendation is B) l l .ie2. •

A) 1 1 .f5!? Apart from its surprise value, this line has the major advantage of being equally playable without the moves . . . h6 and .ih4 included. See variation A of the next chapter for coverage of that scenario. The inclusion of . . . h6 should help Black slightly, but he still faces plenty of challenges. We will consider three replies: AI) l l til eS, A2) l l O-O!?N and A3) l l eS.

8 7 6 5 4 3 2

•••

•••

a

b

e

d

e

f

g

h

•••

1 86

Classical Main Line 13 . . . exf5 1 4.�xf6 �xf6 1 5 .ttJd5 �g5t 1 6.mb 1 'l&c5 1 7.ttJxf5±

AI) 1 1 ...tZleS 8

1 3 . . . ttJ c6 1 4.g4 ttJxd4 1 5 .E:xd4 Black has no clear counterplay, while White has the whole position under control and g4-g5 will follow soon.

7 6 5

1 3 . . . 'I&d7 1 4.g4! With the moves �e2 and . . . 'I&d7 included, this leads to a strong attack. 1 4 . . . b5 1 5 .g5 hxg5 1 6.�xg5 b4

4 3 2

8

1 a

b

e

d

e

f

g

7

h

6

12.Y:Vh3! O-O!?N

5

Black appears to be castling into the fire, but it is actually harder for White to prove something than I had imagined initially.

4 3 2

1 2 . . . exf5 ? 1 3 .�xf6 �xf6 1 4 . ttJ d5 looks awful for Black. 1 2 . . . ttJ fg4 has been played a few times, but the strong reply 1 3 .�e2! shows that the inclusion of . . . h6 and �h4 is not without benefits to White.

1 3.i.e2!

d

e

f

g

h

8 7 6

1 3 . fxe6 fxe6 transposes to variation A2 below. Although White can fight for the advantage there, he should not hurry to exchange on e6 if he can avoid it. The text move is an ideal solution which keeps all options open.

2

Black has problems after other moves as well:

e

1 7.�h6! ttJ e8 ( l 7 . . . gxh6? 1 8 .'I&xh6+-) 1 8 .E:hg l White has a strong initiative.

My first thought in this position was that 1 3 .g4 looked dangerous for Black, but after 1 3 . . . b5 it turns out that White's attack is not so special, while Black's counterplay is coming fast.

13 ... hS

b

a

5 4 3

1 a

14.fxe6!

b

e

d

e

f

g

h

We have seen that the g2-g4 plan would be too slow, but the inclusion df the last two moves makes this a more favourable moment to capture on e6.

1 87

Chapter 1 1 - Three-Piece System with . . . h6

14 fxe6 1 5.tl�xe6 YlYb7 ..•

This time 1 5 . . .'IMfa5 1 6.'kt>b l 'kt>f7 runs into a tactical refu�ation: 1 7.i.xf6! i.xe6 ( I 7 . . . i.xf6 1 8 J'hd6+-) 1 8 .�h5t g6 1 9 .�xh6 i.xf6 20.�h7t 'kt>e8 2 1 .:B:xd6+-

16J�� hf1

lH7

17.a3!?

White has time to stabilize on the queenside, and there is no obvious way for Black to exploit the pin on the knight. The following line is critical, but it works out well for White. 8

This is a computer-esque move, but it can be characterized as an attempt to reach an improved version of the previous line.

12.fxe6 1 2.g4 ltJe5 can be compared with 1 1 . . .ltJe5 1 2.�h3 0-0, when it was noted that 1 3 .g4 was not the best move. True, White can put his queen somewhere other than h3, but I was unable to find anything convincing. One important point is that 1 3 .�g3 can be met by the annoying 1 3 . . . g5 ! '

12 tL'Je5 13.YlYh3 •••

7

Grabbing a pawn is risky: 1 3 .exf7t :B:xf7 1 4.�e2 ( I 4.�g3 g5Ft= is an important point.) 14 . . . i.g4 1 5 . ltJ f3 :B:c8 White is a pawn up, but Black has a lot of activity and there are too many tactical tricks hanging in the air.

6 5 4 3

13 fxe6 •.•

2

a

b

e

d

e

f

g

h

17 g5!? 18.i.g3 g4 19.YlYxh6 Le6 20.Le5 dxe5 21 .YlYg5t �h8 22.YlYxe5 .••

White has three pawns and a strong initiative for the piece.

A2) 1 1 . 0-O!?N

Compared to variation A I , White has been forced to commit to fxe6, which considerably reduces his attacking potential. Nevertheless, there are still ways to put the opponent under pressure. 8 7 6

••

5 8

4

7

3

6

2

5

1

4

a

b

e

d

e

f

g

h

3

14.i.e2

2

1 4.ltJxe6?! �a5 ! 1 5 .i.e2 'kt>f7! does not work for White, but the bishop move renews the threat.

a

b

e

d

e

f

g

h

1 88

Classical Main Line

14 YlYd7 •.•

1 4 . . . b5 transposes to variation A I , where we saw that White could simply take on e6. 14 . . . �d7!? seems playable thanks to some computer tactics. The critical line is: 1 5 .ttJxe6 1Mra5 ( 1 5 . . . 1Mrc8 1 6.ttJd5 ttJ xd5 1 7.exd5 �xh4 I B .1Mrxh4 �xe6 1 9 .dxe6 1Mrxe6 20.'it>b l ;!;) 1 6.'it>b l El:acB 1 7.ttJd5 ttJ xd5 I B . exd5 �xh4 1 9 .1Mrxh4

is 1 5 . . . b5 1 6.a3 �b7 1 7.El:hg l ttJxe4 I B .ttJ xe4 �xe4 1 9 .95, reaching a complicated position where White certainly has some initiative for the pawn.

1 5 .tl�fg4 . •

Supporting the crucial knight on e5. Mter 1 5 ... ttJxf3?! 1 6.gxf3 the open g-file gives White an automatic attack. 1 5 . . . 1Mrc7 1 6.ttJxe5 dxe5 1 7.g4!? b5 I B .�d3! looks a bit unpleasant for Black. White intends g4-g5 , and if . . . b4 or . . . �b4, White can just play ttJ e2 and ttJ g3 .

8 7 6 5 4 3 2 b

a

e

d

e

f

g

h

1 9 . . . 1Mra4! 20.1Mrxa4 �xa4 2 1 .b3 El:f2 22.El:de l �eB 23.El:hfl El:xg2 24.El:f8t 'it>h7 2 5 . h4 Black has regained the pawn, but White keeps a slight plus thanks to his active pieces. 8

a

7

b

e

d

e

f

g

h

16.El:d4!�

6 5 4 3 2 1 a

1 5.ttJf3

b

e

d

e

f

g

h

1 5 .g4! ? could also be explored. A possible line

1 6.�xe7 1Mrxe7 1 7.1Mrg3 1Mrc7 I B .h3 ttJf6 1 9 .ttJxe5 dxe5 seems okay for Black, since there is no easy way to advance the g-pawn. Instead White should just maintain the tension and build some pressure against the d6-pawn . Obviously he should be in no hurry to exchange on e5, as the black knights are rather awkward. Black's position remains rather solid, but he has no obvious counterplay and White can continue to create mini-threats while gradually improving his pieces.

1 89

Chapter 1 1 - Three-Piece System with . . . h6 A3)

where the same position is reached with the pawn on h7 instead of h6. Here the position may be even more favourable for White, as the pawn on h6 enables White to open files more easily with g4-g5 .

1 l ... e5

8 7 6

8

5

7

4

6

3

5

2

4 3 a

b

e

d

e

f

g

h

This is the most obvious reply, and the choice of Nakamura.

12.tll de2 h5 13 ..ixf6 This exchange is a key part of White's plan.

13 ... Eilxf6 14.tll d5

a

b

e

d

e

f

g

h

2

a

e

d

e

f

g

h

1 6 . . . b4 After 1 6 . . . l"k8 1 7.�d3 White's ideas include �b l , lD h 5 and eventually f5-f6, as pointed our by Marin in his annotations to the Smeets - Nakamura game. 16 . . . 0-0 seems dangerous after 1 7 .�d3 intending �e4, lD h 5 , g2-g4 and so on. 16 . . . 0-0-0 puts the king in a safer spot, but now Black has no counterplay, and eventually White can open up the queenside as well. 1 7.�b l �b8 1 8 .lD e4 l"!:c8 1 9 .c3 �b6 20 .�e2 b4 2 1 .c4:;!; 1 7. lD h 5 ! 1 7 .�d3 �a5 ! is Black's idea. Now it is not easy to defend the g7-pawn. 1 7 . . . l"!:g8 1 7 . . . 0-0? 1 8 . f6 is too dangerous. 8

14 ...�c5!

7

This may look like a surprising choice, but it avoids giving White the easy play he gets after the exchange on d5.

5

1 4 ... lDxd5N 1 5 .exd5 �b7 1 6. lD g3 This structure will be discussed at greater length in variation A1 of the next chapter,

b

6 4 3 2 a

b

e

d

e

f

g

h

1 90

Classical Main Line

I B .i>b l With the rook committed to gB , Black's position looks awkward. White intends to secure everything with .td3-e4 before advancing on the kingside. I B . . .'IWc5 Stopping the aforementioned plan, but White can make a simple adj ustment. 1 9 . 1tJ g3! Followed by ltJ e4 with a fine position .

�b6 20.�d3 (20.i>b l ! ?N seems like an improvement, stopping the idea of a quick . . . b4. It should be similar to our main line, but I'd still prefer not to commit to c2-c3 .) 20 . . . :8:c6 2 1 .a3

b

a

a

b

e

d

e

f

g

h

1 5.tLJec3

e

d

e

f

g

h

2 1 . . . b4!N It is important for Black to open a file for counterplay. (2 1 . . .�e7 22.�b l h5 23 . .te2 h4 was played in Smeets - Nakamura, Wij k aan Zee 20 1 0 , and now 24.�f3!N would have stopped the queen from j umping to fL. . White i s slightly better, for instance: 2 4 . . . :8:bB 25 . .td l a5 26 . .tb3;!;) 22.axb4 �xb4 23.i>b l �b6 24.h4 O-Of!

After 1 5 .ltJxe7 i>xe7 1 6.g4 .tb7 1 7. ltJ g3 d5! Black is helped by the inclusion of the ... h6 move, and White's play is too slow.

1 5 ....ib7 16.cj{b l !N I don't see any reason for White to exchange on f6 yet, and he will save time if Black exchanges on d5. I guess White's idea was to be able to counter . . . :8:cB with c2-c3 , but it seems to me that this pawn move is what allows Black to get counterplay in the opposite-coloured­ bishop middlegame that is likely to arise. To illustrate these points, it is important to see what happened in the game: 1 6.ltJ xf6t .txf6 1 7.ltJd5 :8:cB! I B . c3 .txd5 ( l B . . . b4!?N is also possible, and after 1 9 .c4 .txd5 20.:8:xd5 �b6 I don't like the fact that White has already weakened the queenside.) 1 9 .:8:xd5

a

16 .. J3c8

b

e

d

e

f

g

h

1 6 . . . ltJxd5?! makes things easy for White . 1 7.ltJxd5 ( l 7.exd5? :8:cB) 1 7 . . :.txd5 I B .:8:xd5 �b6 1 9 .h4 White has an even more convenient version of the main line below.

191

Chapter 1 1 - Three-Piece System with . . . h6 1 6 . . . b4 1 7.tLla4 'IWa5 1 B . tLl ab6 tLlxd5 1 9 .tLlxd5 ( 1 9.tLlxaB? tLl f6!) 19 . . . .ixd5 20 .l'hd5 'lWb6 2 1 .g4 results In a similar type of position.

17.'lWe2! I want to get a similar position to the Smeets - Nakamura game, but without committing to c2-c3 . The queen might seem misplaced on e2, but this is only temporary. After a subsequent h2-h4 and g2-g4, the queen may move to g2, where it will be ideally placed to support White's kingside play. 1 7 . .id3 would squander the advantage: 17 . . . .ixd5 I B .tLlxd5 tLlxd5 1 9 .exd5 .if6 20 . .ie4 a5 Without the white knight, it will be hard to get the initiative going on the kingside.

17 ...tihd5 18.t£!xd5 hd5 19J:hd5 �b6 The computer assessment is close to equal here, but White has definite potential on the kingside. Black does not have much scope for counterplay, as the following lines demonstrate.

21 ...a5 2 1 . . . 0-0 seems too dangerous after 22.g4 l'!fcB 23.g5 . Black should therefore leave his king in the centre, but in that case the hB­ rook will remain out of play, and his queenside counterplay will not carry much weight.

22.g4 a4!? 22 . . . b4 makes the c4-square available for White's bishop. This can be exploited by: 23.b3! (White can also start by moving the queen first.) 23 . . . 'lWc6 (23 . . . a4 24.bxa4!?;!; ; 2 3 . . . l'!c3 24.'lWg2;!;) 24.'lWg2! a4 2 5 . .ic4 White has an obvious edge.

23.�e3 �a5 After 23 . . . 'lWb7 24.l'!hh2 b4 2 5 .g5 White's initiative is developing nicely. 23 . . . 'lWc6 is well met by 24.a3! intending c2-c3 , now that the . . . b4 plan is off the table.

8 7 6 5 4 3 2 a

1 a

b

e

d

e

f

g

h

b

e

d

e

f

g

h

24.l'!hh2! Preparing to push the g-pawn.

20.h4 l'!c5 20 . . . h5 leaves the rook stuck on hB. 2 1 .g3 l'!c5 22 . .ig2;!;

21 .l'!d2! By avoiding the rook exchange White keeps more prospects of an initiative.

24 a3 .••

24 . . . f6 restrains the kingside advance, but is an obvious positional concession. 2 5 .c3! b4 If Black does not play this, a2-a3 will block the queenside, leaving White in complete control. 26.cxb4 'IWxb4 27.a3 'IWb7 2B.l'!c2±

1 92

Classical Main Line

White is on top, as the following line illustrates:

This has been White's most popular move, as well as the highest-scoring. We will consider five main responses from Black: Bl) 1 l E:g8?,

25 axb2 26.f6!

B2) 1 l g5?!, B3) 1 1 ...E:b8?!, B4) 1 1 ...b5 and B5) 1 l O-0!?

25.g5

•••

..•

•••

•••

Opening up the kingside. Black's bishop is particularly poor.

1 1 . . . lD b6 does not seem to help Black's cause at all. White has a few plans, but the simplest is to prepare e4-e5: 1 2 . .ig3 0-0 1 3 .�b l lD c4 ( 1 3 . . . .id7N allows 1 4.e5 followed by lD e4 with excellent play for White.) 1 4 . .ixc4 Wxc4 1 5 .e5 dxe5 1 6.fxe5 lDd5 1 7.lDxd5 exd5 I B .b3 Wc5 1 9 . .if2 Wa5 20.g4;!; Cheparinov - Dominguez Perez, Linares 2002.

Bl) 1 l E:g8? .••

Preparing . . . g5 in this way is too slow.

12.E:hfl a

b

e

d

e

f

g

h

Anticipating Black's next move.

12 g5

26 ... gxf6

•••

Mter 26 . . . .iffi 27.fxg7 .ixg7 2B.E:xd6 Black's position is collapsing.

1 2 . . . b5N is met by the typical 1 3 . .ixf6! lD xf6 1 4.e5 .ib7 1 5 .Wg3 lD e4 1 6.lDxe4 .ixe4 1 7 . .if3 .ixf3 I B .Wxf3 with a clear advantage.

27.gxf6 .ifS 27 . . . .ixf6 2B.E:xd6 wins.

28.E:hg2 White is dominating.

B) 1 l .ie2 •

8 7 6 5

a

4

13 ..ig3!

b

e

d

e

f

g

h

This counterintuitive move is an important discovery from the correspondence arena.

3 2

1 3 .fxg5 works out fine after 1 3 . . . hxg5 ?!, when

1 a

b

e

d

e

f

g

h

1 93

Chapter 1 1 - Three-Piece System with . . . h6 1 4.�g3 ttJeS l S .�xeS! dxeS 1 6.ttJdbS! axbS 1 7.ttJxbS WlcS 1 8 . ttJ d6t c;;t> f8 1 9 .WlhS ! led to an impressive victory for White in Robson Lu, Ningbo 20 1 3 . However, Black can avoid this with 1 3 . . . ttJeS!N, after which 1 4.Wlf2 hxgS I S .�g3 �d7 transposes to the unclear game Mathes - Luethke, corr. 1 983.

8 7 6 5 4

13 ... gxf4

3

1 3 . . . g4 1 4.Wlf2± gives White an excellent position, since the entire purpose of the . . . gS push - namely, to wrest control of the eS­ square - has been thwarted.

2 1 a

b

e

d

e

f

g

h

16,tlhe6! fxe6 17.e5 bxc3 1 8.exf6 cxb2t 19.�bl

14.Wfxf4!

White had a crushing attack in Oseledets Palladino, corr. 2009.

B2) 1 1 ...g5?! 8 7 6 5 4 a

b

e

d

e

f

g

h

14 ... h5

The tactical point of White's play is revealed after 1 4 . . . eSN I S .Wlf2! exd4 1 6.ttJdS Wld8 1 7.E:xd4+- when Black can do nothing about the threat of ttJxe7 followed by �xd6. 14 . . . ttJeS l S .Wlxh6 E:g6 1 6.Wlh8t E:g8 1 7.Wlh4 ttJg6 1 8 .Wlh6 �d7 1 9 .E:f2± was not much better for Black in Vittore - Patzereli, Internet 20 1 2. White is a pawn up and the eS-square isn't all that helpful for Black.

15.Wfxh6 b4 This looks like a reasonable try for counterplay, but it is refuted spectacularly.

3 2 1 a

b

e

d

e

f

g

h

Initially this was considered one of the main ideas here, but the results have been clearly in White's favour, which has led to a loss of interest in the pawn sacrifice. White j ust needs to place his pieces on the right squares.

12.fxg5 t£Je5 13.Wfe3! 1 3 .Wlf2 ttJ fg4 forces White to either take on g4 or place his queen on the awkward f4square.

Classical Main Line

1 94

13 ... til fg4 1 3 . . . tLl h7 has been played more often, but the knight is too passive on this square. White simply continues development and puts pressure in the centre: 1 4. tLl f3 hxgS ( 1 4 . . . tLl xf3 I S .Wxf3 tLl xgS 1 6.Wd3 bS 1 7.�g3 gave White a big advantage in Larino Nieto - M. Gomes, Barcelona 20 1 1 .) I S .�g3 �d7 White has more than one good option here, but the simplest was played in the latest game.

is not easy at all. There is no reason for White to commit to the exchange on g4 so soon.

14 ... hxg5

1 4 . . . WcS can be met by l S .tLla4!?N Wc7 1 6.h3± and White is in control.

14 . . . :B:gB l S .�xg4 tLlxg4 1 6. tLl f3 was also excellent for White in Grabarczyk - Dinger, Germany 2009.

1 5.i.xg5 f6 16.i.f4 i.d7

8

1 6 . . . tLl f2 occurred in Bernal - Sridharan, Internet 200 S , and now the simple 1 7.�bSt!N axb S 1 B .tLldxbS Wb6 1 9 .�e3 would have given White a large advantage.

7 6 5 4

After the text move I found a simple improvement over the game Ootes Zherebukh, Leeuwarden 2009, where 1 7.:B:hfl was played.

3 2 b

a

e

d

e

f

g

h

1 6.tLlxeS ! ? ( 1 6.Wd4 has been more popular, but it seems pointless to allow . . . f6.) 1 6 . . . dxeS 1 7.:B:hfl Preventing long castling and preparing �hS . 1 7 . . . tLl fB I B .Wf2 f6 1 9.h4!± Li Chao - ]i, Xinghua 20 1 0 . 8 7 6 5 4

a

b

e

d

e

f

g

h

17.W1el !N

3

Surprisingly, Black's position is on the verge of collapsing.

2 1 a

b

e

d

e

f

g

h

14.W1d2! 1 4 .�xg4 tLl xg4 I S .Wd2 has been played several times, but after l S . . . tLl e S ! ? the position

17 ... 0-0-0 1 7 . . . bS 1 B .Wg3 embarrasses the knight on g4.

Chapter 1 1 - Three-Piece System with . . . h6 1 8 . . .if8 1 9 .ixg4 lZlxg4 20.lZlxd6t ixd6 2 1 .ixd6 give, White an extra pawn plus a huge initiative.

1 95

12.'V!fg3 gg8 Other moves also fail to solve Black's problems: 1 2 . . . b5 runs into 1 3 .e5! dxe5 1 4.lZlxe6 fxe6 1 5 .Wixg7 gg8 1 6.ih5 t 'kfid8 1 7.Wixf6+- as in Adhiban - Mustafaev, Tashkent 2007. 12 . . . 0-0 1 3 .if3 makes it hard for Black to continue developing. 1 3 . . . lZl b6 ( l 3 . . . b 5 ? 1 4.e5! dxe5 1 5 . fxe5 ltJ xe5 1 6. lZl c6+-) 8 7 6 5 4

19.�d5!? Definitely the flashiest move, although 1 9 .1Zlxe7t Wixe7 20.h3 lZl h6 2 l .Wif2 'kfib8 22 .Wib6 is also pretty conclusive.

19 ... exd5 20.exd5 .ie8 2 1 .�xe7t 'V!fxe7 22 ..ixg4t <.t>b8 23.'V!fe4 White is winning.

B3) 1 l J�b8?! .•

3 2

a

b

e

d

e

f

g

h

1 4.ghe l ! ? Continuing to strengthen the position. (The immediate 1 4.e5 dxe5 1 5 . fxe5 ltJ fd5 1 6.lZlxd5 lZl xd5 1 7.ixd5 ixh4 1 8 .Wixh4 exd5 1 9 .9he l ge8 20.ge3± was also pleasant for White in Kotronias - Kulaots, Heraklio 2007.) 14 . . . id7 1 5 .e5 dxe5 1 6. fxe5 ltJ fd5 8 7 6 5 4 3 2

a

b

e

d

e

f

g

h

This has been played in a lot of games but it is rather slow, and White has scored excellently.

a

b

e

d

e

f

g

h

1 7.ixd5!?N ixh4 1 8 .Wixh4 ltJ xd5 1 9 .1tJxd5 exd5 20.gd3±

1 96

Classical Main Line

The last of Black's alternatives is: 12 . . . g5 1 3 . fxg5 hxg5 1 3 .. .l:!gB 1 4. ltJ xe6! fxe6 1 5 .gxf6 was great for White in Weiler - Stanek, Donaueschingen 1 9B 5 , since 1 5 .. .l:!xg3 ? would have lost instantly to 1 6.ih 5 t . 1 4.ixg5 b 5 1 4 . . . l'!gB 1 5 .h4 b 5 1 6.'1We3 b 4 1 7. ltJ b l ib7 I B .ltJ d2± gave White an excellent position in Nunn - Nemet, Biel 1 9B3.

to stay in the game in Planinec - Nemet, Zagreb 1 977. 1 9 . . . l'!xh2 20.ltJe4 d5 2 1 .ltJg5 With a raging initiative. Let's return now to 1 2 . . . l'!gB . This has been the most popular choice, but it seems such an awkward set-up for Black. White j ust prepares the typical e4-e5 break while ensuring he is suitably placed to meet . . . g5 .

8 7 6 5 4 3 2

a

b

e

d

e

f

g

h

1 5 .e5! White obtains a powerful initiative by opening up the centre. 1 5 . . . ltJ xe5 1 6.ixf6 ixf6 1 7.l'!hfl ie7 I B .'IW g7 l'!h4 I B . . . l'!xh2 1 9. 1tJ e4 leaves Black without a good move, as 1 9 . . . ib7 20.ltJxe6! fxe6 2 1 .ltJ f6 t is crushing.

b

a

e

d

f

e

g

h

13J3hfl ! b5 1 3 . . . g5 1 4. fxg5 gives Black an unpleasant three-way choice: a) 14 . . . hxg5N 1 5 .ixg5± b) 14 ... Wfa5 runs into a spectacular refutation: 1 5 . ltJ xe6! fxe6 1 6.gxf6 l'!xg3

8 7 6

8

5

7

4

6

3

5

2

4 3 a

b

e

d

e

f

g

h

1 9 .93!N 1 9.WfgB t i>d7 20.l'!xf7 WfdB enabled Black

2

a

b

e

d

e

f

g

h

1 97

Chapter 1 1 - Three-Piece System with . . . h6 1 7.Eld5 !!+- Weinstein - De Fotis, Chicago 1 973. c) 14 . . . lt:le5 1 5 .lt:lf3 hxg5 16 . .ixg5 This position first occurred in a correspondence game from 1 9B 1 , and Black has still not found a good answer. A recent example: 1 6 . . . lt:l h 5 1 7.�h3 .ixg5t 1 B .lt:lxg5 Elxg5 1 9 . .ixh5 b5 8 7 6

a

5

b

e

d

e

f

g

h

17.'i;Yf7!N

4 3 2 b

a

e

d

e

f

g

h

20.lt:le2 �e7 2 1 .lt:l f4 Elg7 22.�b 1 .id7 23 . .ie2 Elg8 24.lt:lh5 �dB 2 5 . g4 White was in full control in Sanchez Rodenas - Vinchev, em ail 2009. 8 7

Not everything is good here. For example, 1 7. f5 ? ! exf5 1 B ..ixf6 .ixf6 1 9. 1t:l d 5 Wfc6 20.Elxf5 led to a win for White in Lombardy - Quinteros, Manila 1 973, but the result might have been different had Black found 20 . . . Wfe6!+ with . . . .ib7 coming next 1 7 . .ixf6 .ixf6 1 B .Wff7 ElhB 1 9 . 1t:l e4 .ie7 20.Wfxg7 was a better continuation which gave White excellent play in Barczay - Ribli, Budapest 1 976, but the text move is even more conclusive.

17 ... �U'8 18.'i;Yxg7 exf4 19.�e4 gg8

6

1 9 . . . Wfe5 20.WfxfBt! .ixfB 2 1 ..ixf6t is a nice detail.

5 4

8

3

7

2

6

1 a

b

e

d

e

f

g

h

14.eS!

Having made all the necessary preparatory moves, it is time to start the attack.

14 ... dxeS l sJihe6 fxe6 16.'i;Yg6t i>d8

5 4 3 2 1 a

b

e

d

e

f

g

h

1 98

Classical Main Line

20.YlYxh6! This is stronger than 20.'Wf7 :B:fB 2 l .'Wxe6 :B:b6 22.'Wf5 .ib7 when White does not seem to have a knockout blow.

White must insert this exchange, since the immediate 1 2.e5? is ineffective due to 12 . . . .ib7 1 3 .'Wg3 dxe5+.

12 �xf6 ••.

20 �xe4 2 1 ..be7t
Threatening :B:xf4 t . Black only has one response.

1 2 . . . .ixf6 1 3 .e5 dxe5 1 4. ttJ dxb5 axb5 1 5 .'WxaB 0-0 1 6.ttJxb5 'Wc5 1 7.'Wa3 ± Groszpeter - Karner, Zalakaros 2006.

13.e5 .ib7 14.YlYg3 dxe5 1 5.fxe5 t£!d5

23 e5 24.YlYfSt
There is nothing better:

•.•

25 . . . �hB 26.'Wh6 is mate.

26.YlYh6t :B:g7 26 . . . �e7 27.'Whlt �fB 2B . .ih5 would not change much.

1 5 . . . ttJ e4 runs into the forcing sequence 1 6.ttJxe6! 'Wxc3! 1 7.ttJxglt �fB I B .ttJ e6t \t>eB 1 9 .'Wxc3 ttJxc3 20.ttJclt \t>fB 2 l .bxc3 :B:cB 22.ttJd5, when Black will have to suffer in a one-sided endgame.

27.:B:xd7! YlYxd7 28.YlYh8t White is winning.

1 5 . . . ttJ d7 This allows White to launch a powerful attack. 1 6.ttJxe6! fxe6

B4) 1 1 ...b5 8

8

7

7

6

6

5

5

4

4 3

3

2

2 1

a

a

b

e

d

e

f

g

h

This is Black's old main line in this variation, but everything has been well established. A relatively forcing sequence leads to an opposite­ coloured-bishop endgame where Black has chances to make a draw a pawn down, but must suffer endlessly.

12 .bf6! •

b

e

d

e

f

g

h

1 7.'Wg6t! 1 7 . .ih5t is also strong, but the text move is virtually winning by force. 1 7 . . . �dB I B .'Wxe6 :B:eB 1 9 .'Wg6! ig5t 19 ... 'Wc6 20 .'Wg3 \t>c7 (20 ... 'Wxg2 2 1 .'Wd3 'Wc6 22.ig4) 2 1 .if3 'We6 22.ig4 'Wxe5 23.:B:xd7t 1 -0 ] . Williams - Agulnick, corr. 20 1 3 . 20.�b l

1 99

Chapter 1 1 - Three-Piece System with . . . h6

8

8

7

7

6

6

5

5

4

4

3

3

2

2 b

a

e

d

e

f

g

h

b

a

20 . . . E!e7 20 . . . E!xe5 was played in Figuero Toro Damaso, Evora 2008. 2 1 . .ig4N E!e7 22.lLld5 .ixd5 23.E!xd5 +2 1 .h4 Wic6 22.E!d6 Black resigned in Bernal - Deghose, Internet 20 1 1 .

e

d

e

f

g

h

23 . . . .if6 24.Wif7t i>d6 25 . .if3 E!ab8 26.Wia7 b4 27.Wixa6t i>e7 28 .Wia7t i>d6 29.a3 bxa3 30 .Wixa3t i>d7 3 1 .c3 E!hc8 32.Wia7t i>d6 33.i>c2± In Sebag - Kokarev, Plovdiv 2008, White went on to convert her advantage without giving Black much of a chance.

20.�xgS hgst 2 1 .@bl hdS 22J�xdSt @c7

8 7 6 5 4 3 2 1 a

b

e

d

e

f

g

h

16.tihe6! The next few moves are forced.

a

b

e

d

e

f

g

h

23J::�cSt! 16 ... fxe6 17.�g6t @d7 18 ..ig4 �xeS 19.1LlxdS �gst!

The only move. After 19 . . . .ixd5? 20.E!xd5 t Wixd5 2 1 .E!d 1 .ig5t 22.i>b l Wixd l t 23 . .ixd l Black might be doing okay materially, but the presence of opposite-coloured bishops amplifies White's attacking chances against the weak black king. An instructive example of White's play continued:

By nudging the king to b6, White hopes to get a slightly better version of the typical endgame. This move is not mentioned by Ftacnik in Grandmaster Repertoire 6, but it has been played in almost all of the recent correspondence games.

23 ... @b6

23 . . . i>d6 holds onto the e-pawn, but after

Classical Main Line

200

24.:!:k3 gac8 2 5 .ga3 ! gc6 26 . .if3 gb6 27.gd 1 t Black is under pressure. That is why most players have kept the king on the queenside.

of Ftacnik has not really taken off. I will leave you with an instructive example of White's model technique from a correspondence game.

24J:!e5 ghe8

29.a3 a5 30.c3 ge7 3 1 .ge6 .tf6 32.q;a2 i.e5 33.a4 bxa4 34.h3 ge7 35.ge6 i.f6 36.ge4 q;b6 37.gxa4 ge7 38.q;b3 gd7 39.gf4 ge7 40.q;a4 gel 41 .gg4 q;e5 42.q;b3 q;b5 43.ga4 gel 44.ge4 gal 45.i.e6t q;b6 46.q;e2 q;a6 47.ge4 q;b6 48.ge6 q;a7 49.q;b3 q;b8 50J:�e4 q;a7 5 1 .q;e4 ga2 52.ge2 q;b6 53 ..id5 ga4t 54.q;b3 gal 55.ge6t q;e5 56 ..te6 gdl 57.i.f3 gd6 58.ge4 q;b5 59.ge4 i.e5 60.ge7 a4t 61 .q;a3 gd2 62.e4t q;a6 63.ge6t q;a7 64.ge8 ge2 65. q;xa4 gxb2 66.e5

24 . . . gae8 2 5 .ghe 1 ghffi 26.gxe6t gxe6 27.gxe6t leads to a similar type of situation. See the game Lepage - Teichmann, corr. 20 1 2, for one example where White successfully overcame Black's resistance.

25.ghel 8 7

White has made steady progress, and he went on to win in Fleetwood - Stalmach, corr. 20 1 2 .

6 5 4

B5) 1 l 0-0!? .••

3 2 1 a

b

e

d

e

f

g

h

25 ... gae8 25 . . . b4! ? leads to a change of structure, but after 26 . .ih5 ge7 27.gxe6t gxe6 28 .gxe6t �c5 29 . .if3 ga7 30.gc6t �b5 3 1 .c4t bxc3 32.gxc3 the position remained similarly one­ sided in Devour - Amartne, Internet 20 1 1 .

26 ..tf3 ge7 27.gxe6t 27.a4 gce7 28 . .ih5 gc8 29 ..if3 gce8 30 . .ih5 gc8 3 1 .gxe6t gxe6 32.gxe6t gc6 33.a5t �c5 34.ge4 was also better for White in Robson - Le Quang, Lubbock 20 1 1 , although Black managed to hold on.

a

b

e

d

e

f

g

h

In some ways this is the most critical line, since it poses White a direct challenge: can he exploit such provocative play? This move has recently become rather trendy, so I would suggest paying close attention to what follows.

27... gxe6 28.gxe6t q;a7

Black has decent defensive chances, but he faces a long and thankless task, so it is not surprising that this particular recommendation

12.%Yg3 1 2 .g4! ? is an exciting alternative which I analysed in some detail. I will just show the

Chapter 1 1 - Three-Piece System with . . . h6 main variations to show why I was not quite satisfied wit� it: l 2 . . . b5 1 3 .g5 hxg5! ( l 3 . . . lt::l h7 1 4.1&h3!N hxg5 1 5 .lt::l xe6 fxe6 1 6.1&xe6t cj;Jh8 1 7.1&xe7 gxh4 1 8 .:B:hg l :B:g8 1 9.1t::l d5 1&c5 20.1&xh4±) 1 4.i.xg5 b4! 1 5 .lt::l d5!? An attempt to refute Black's play outright. ( l 5 .i.h6 g6! is complicated.)

a

b

e

d

e

f

g

a

b

e

d

e

f

g

h

1 3.e5!N

1 3 .i.f3 e5! is Black's idea, when he seems to be doing alright.

13 ... ttlfe4

8

1 3 . . . dxe5 1 4. fxe5 It::l fe4 ( l 4 . . . lt::l d5 1 5 . lt::l x d5 exd5 1 6.i.xe7 1&xe7 1 7.i.f3 :B:d8 1 8 .:B:he l;!;) 1 5 .lt::l xe4 It::l xe4 gives White more than one decent continuation: 1 6.1&f4 (There is also 1 6.1&e3 , transposing to 1 5 .1&e3 dxe5 1 6. fxe5 as covered in the next note.) 16 . . . lt::l c5 1 7.i.xe7 1&xe7 1 8 .i.f3;l; Intending to attack with h4 and g4-g5 .

7 6 5 4 3 2

a

b

e

d

e

f

g

h

20.1&xg4 1&xe5 2 1 .i.xf6 1&xf6 22.:B:xd5 :B:fe8 23.:B:xc5 1&£2! Black seems to be close to a draw. •••

1 2 . . . b5 is best met by 1 3 .i.f3 ! . (After 1 3 .lt::l d5 It::l xd5 1 4.exd5 i.xh4 1 5 .1&xh4 It::l b6 1 6.i.f3 It::l xd5 1 7.i.xd5 exd5 1 8 .:B:he 1 White eventually prevailed in Swiercz - Banikas, Rijeka 20 1 0, but White's attacking ideas of e4-e5 or It::l d5 will be much stronger on the next move.) 1 3 . . . b4 ( l 3 . . . i.b7 1 4.e5! dxe5 1 5 . fxe5 wins material.) 1 4.lt::l d5! It::l xd5 ( l 4 . . . exd5 1 5 .lt::l f5 +-) 1 5 .exd5 i.xh4 1 6.1&xh4±

h

1 5 . . . exd5 ( l 5 . . . lt::l xd5 1 6.:B:hg l !! was good for White in Solovyev - Nordal, corr. 20 1 2.) 1 6.lt::l f5 !N ( l 6.exd5 has been played, but Black seems to be okay after the forcing line: 16 . . . i.b7!N 1 7.:B:hg l i.xd5 1 8 .i.xf6 It::l xf6 1 9 .1&h3 g6 20.f5 It::l h7 2 1 . fxg6 fxg6 22.:B:xg6t Wh8°o) 1 6 . . . lt::l c5! 1 7.lt::l xe7t 1&xe7 1 8 .e5 dxe5 1 9 .fxe5 i.g4

12 c!Llc5

20 1

14.ttlxe4 ttlxe4 1 5:�f3 It is hard to decide if it is more precise to put the queen on f3 or e3 . 1 5 .1&e3 It::l c3 ! ? This i s a n extra possibility which i s not available when the queen goes to f3 .

Classical Main Line

202

1 5 . . . dxe5 allows White to go for 1 6. fxe5, since the black knight will not hit the queen from g5. ( 1 6.Wfxe4 transposes to the main bolded line below.) 1 6 . . . lik3 ( 1 6 . . . tLlg5 1 7.ig3;t) 1 7.ixe7 tLl xe2t I B .tLlxe2 Wfxe7 1 9 .tLlg3 Intending tLl h 5 with some initiative.

b

a

e

d

f

e

g

h

1 5 ... dxe5

I consider this the best try.

a

b

e

d

e

f

g

h

1 6.ixe7 1 6.Wfxc3 Wfxc3 1 7. bxc3 ixh4 I B .exd6 :1!dB gives Black decent compensation. 16 . . . tLl xd l 1 7.exd6 tLl xe3 1 7 . . . Wfxe7 I B .dxe7 tLl xe3 1 9 .exf8=Wft c;t>xfB 2o.iS :1!bB 2 1 .:1!e l tLl f5 22.tLlxf5 exf5 23.c4;t I B .dxc7 tLl xg2 1 9 .id6 :1!eB White is a full exchange down, but the c7pawn and d6-bishop are monstrous.

a

b

e

d

e

f

g

1 5 . . . tLl c3 ? ? 1 6.ixe7 just wins now that the queen is on f3 . 1 5 . . . ixh4 This is the computer's first suggestion, but I think the machine is underestimating White's attacking prospects. 1 6.Wfxe4 :1!dB ! ? This looks odd, but it i s not easy t o refute. 1 6 . . . ie7 1 7.id3 g6 I B .h4--+ 1 6 . . . dxe5 1 7.fxe5 ie7 I B .:1!hf1 !? id7 1 9.id3 g6 20.h4± 16 ... d5 1 7.Wfe3 ie7 I B .g4± White has an excellent version of this structure since h6 is a severe weakness.

h

20.:1!g l tLl h4 2 1 .c4 id7 22.c;t>d2 It is hard to give a definite assessment, but I prefer White.

a

1 7 .g4! ?

b

e

d

e

f

g

h

Chapter 1 1 - Three-Piece System with . . . h6 1 7.id3 g6 1 8 .g3 ie7 1 9 .h4 dxe5 20.fXe5 b6! is uncl\!ar. 1 7.exd6 �xd6 1 8 . llJ f3 ie7 1 9 .id3 f5 20.'I&e2 gives White some attacking chances based on g2-g4, although he will have to watch out for possible counterplay involving Black's strong dark-squared bishop. I 7 . . .id7 I7 . . . dxe5 I 8 . llJ f3! ie7 ( 1 8 . . . if6 I 9 .g5--+) I 9 .llJxe5 followed by id3 gives White some initiative. I 8 .id3 g6 8

203

16.%Vxe4 I 6. fXe5 is met by I 6 . . . llJg5! I 7.'I&e3 b5 intending ... ib 7, when Black is in time to avoid the attacking plan of ig3, h2-h4, 'l&e4 and id3 .

16 exd4 17 .lxe7 %Vxe7 18J:hd4 �a7! •••

•

The only way to continue developing. Both 1 8 . . . �d8 I 9.�xd8 t 'l&xd8 20.�d l and I 8 . . . �b8 I 9 .id3 g6 20.'I&e5 leave Black under heavy pressure. 8

7

7

6 5

6

4

5

3

4

2

3 b

a

e

d

e

f

g

2

h

1 9 .f5! dxe5 I 9 . . . d5 20 .'I&e2± 20.fXe6 ie8 20 . . . exd4 2 1 .exf7t �f8 22.'I&xg6 looks extremely dangerous. 2 1 .llJ f3 if6 22.� b I Black has held his kingside together, but White's position remains more pleasant.

1 a

b

e

d

e

f

g

h

19.h4!?

This virtually forces Black to play . . . f5 in the near future. Even though h2-h4 might not be an ideal move in that structure, if White can consolidate, he will have the better chances. Mter I 9.id3 g6! ( 1 9 . . . f5 20 .'I&e5;!;) it is surprisingly hard to break through on the kingside: 20.h4 h5 2 1 .g4 (2 1 .f5 gxf5 22 .'I&e2 b5 23.�h3 f6 24.�g3t �h8 2 5 .'I&xh5t 'l&h700) 2 1 . . .b5 22.'I&e3 hxg4 23.h5 f5 ! 24.hxg6 'l&f6oo

19 b5 20 .td3 f5 •••

•

20 . . . g6 2 1 .h 5 ±

2 1 .%Ve2 The queen is better here than on e3, as it can help to support a future g2-g4. a

b

e

d

e

f

g

h

204

Classical Main Line

Conclusion

8 7 6 5 4 3 2 1 a

b

c

d

e

f

g

h

2 1 ..J�d7 22J:!:xd7 �xd7 22 . . . ixd7 23.g3 ic6 24.:8:fl is similar; White will prepare g3-g4, perhaps after tidying up his king's position with i>b 1 and a2-a3 .

23J�dl �f7 24.i>bl .ib7 25.a3� The g2-g4 idea is in the air, although White may also keep it in reserve and look to improve his position in other ways.

Against the . . . h6 version of the Three-Piece System, we have covered two promising approaches for White. The rare 1 1 .f5 ! ? packs a definite punch, which is only enhanced by its surprise value. The untested 1 1 . . .0-0!?N may be Black's best bet, although even here he must overcome plenty of pitfalls. A nice bonus of the early f4-f5 is that it works just as well (or perhaps even better) in the version without . . . h6, making it an especially convenient line to have in your repertoire. We also covered the tried-and-tested 1 1 .ie2 , which remains in excellent theoretical shape. We looked at five responses in detail, of which the first three - 1 1 . . .:8:g8 ?, 1 1 . . .g5 ?! and 1 1 . . .:8:b8?! - all suffer from clear drawbacks. 1 1 . . . b 5 , on the other hand, is fairly reliable, but there cannot be many players who would relish the prospect of being tortured in a pawn­ down endgame. Black's most interesting option is 1 1 . . . 0-0!?, although my new idea of 1 3 .e5!N introduces plenty of fresh problems for the defence. I would encourage you to make up your own mind as to whether to put the queen on f3 or e3 a few moves later; either way, I like White's chances.

8 7 6 5

Classical Main Line

4 3 2 a

b

e

d

e

f

g

Three-Piece System - Main Line Variation Index l .e4 c5 2.�f3 d6 3.d4 cxd4 4.�xd4 � f6 5.� c3 a6 6.i.g5 e6 7.£4 i.e7 8.�f3 �c7 9.0-0-0 A) 10.£5!? A I ) 1 0... � e5 A2) 1 0...e5 B) 10.g4 B 1 ) 1 0... h6 1 1 .i.xf6 i.xf6 1 2.@b 1 !? B 1 1 ) 12 ... g5 B 1 2) 12 ... �b6 B2) 10 ... b5 1 1 .i.xf6 B2 1 ) 1 1 . ..i.xf6 B22) 1 1 ...�xf6 1 2.g5 � d7 1 3.£5 B22 1 ) 1 3 ...hg5t?! 14.@ b 1 B22 1 1 ) 14...� e5 B22 1 2) 14 ... 0-0 B222) 1 3 ... � c5 B223) 1 3 ... 0-0!? 14.h4!? b4 1 5.� ce2 B223 1 ) 1 5 ...i.b7!? B2232) 1 5 ...e5

206 206 207 208 209 209 211 212 214 215 215 215 217 218 22 1 22 1 223

h

Classical Main Line

206

1 .e4 c5 2.c�::J f3 d6 3.d4 cxd4 4.�xd4 c!Llf6 5.�c3 a6 6 ..lg5 e6 7.£4 .le7 8.'1Wf3 Y!!c7 9.0-0-0 �bd7 8 7

to the previous chapter by retreating to h4, but an easier solution is I I .fxe6! ltJ e5 ( l 1 . . .hxg5 1 2. ltJ d 5 ! Wa5 1 3 .exd7t .ixd7 1 4 . .ic4±) 1 2 .Wh3 fxe6 1 3 . .id2 when Black is under pressure. The threat of ltJxe6 is real, and if Black defends suitably then the simple plan of .ie2 and g4-g5 looks promising.

6

AI) 10 ... �e5 1 1 .Y!!h3

5

I played the pointless I 1 .Wg3 ?! here, but the lack of pressure on the e6-pawn meant that Black was comfortable in Negi - Wen Yang, Mashhad 20 1 1 .

4 3 2 1 a

b

e

d

e

f

g

h

This is the starting position for the Three­ Piece System without . . . h6, which used to be considered one of the absolute main lines of the Naj dorf. We will analyse the rare but promising A) 10.5!? followed by the more traditional B) 10.g4.

1 1 ... �fg4 1 1 . . .0-0 ?! 1 2 . .ie2! threatens to take on e6, and the upcoming g4-g5 plan renders Black's position shaky. In the previous chapter we analysed a similar position with . . . h6 and .ih4 included, but here it's simply a worse version for Black, so I won't go into further details.

12 ..lxe7 Y!!xe7 13.Y!!g3

A) 10.5!? This attacking plan is even more appealing here than it was in the previous chapter, as Black's options are considerably reduced without the moves . . . h6 and .ih4 included. I therefore feel it is a fully valid alternative to the huge theoretical lines after 1 0 .g4, even though White is in good shape there too. The text move considerably reduces the number of lines White needs to know; not j ust because there have been fewer games played, but also because the play tends to be less forcing in general.

a

b

13 ... 0-0

e

d

e

f

g

h

Black has two choices: maintaining the pawn on e6 with AI) 10 ... �e5, or blocking the centre with A2) 10 ... e5.

1 3 . . . Wg5 t 1 4.mb l We3 has been played a few times, but after 1 5 . ltJ f3!N Black is struggling due to the pressure on the d6-pawn .

After 1 0 . . . h6N White can, of course, transpose

14 . .le2 �f6

207

Chapter 1 2 - Three-Piece System - Main Line We have reached a complicated middlegame with a lot to explore. The most interesting direction seems to be: 8 7

12 ..bf6! This has hardly ever been played, but it gives White a slightly improved version of the Smeets - Nakamura game from the previous chapter, as Black's kingside is slightly less solid with the pawn on h7 instead of h6.

12 ... tLlxf6 1 3.tLld5 tLlxd5

6

This is the only move to have been tested.

5 4 3 2

a

b

e

d

e

f

g

h

15.�f4! 1 5 .fxe6N fxe6 1 6. tLl f3 �c7 does not seem like much for White. Instead we will prepare g4-g5, while also overprotecting the e4-pawn for the future.

1 3 . . . �c5 !?N may be objectively better, but it gives White a pleasant choice. 1 4. tLl ec3 is a simple option which can be compared with variation A3 of the previous chapter. With the pawn on h7 instead of h6, White will find it easier to advance on the kingside. Alternatively, 1 4.tLlxe7!? mxe7 1 5 .g4 is an ambitious attempt, when the extra tempo required by Black to prevent g4-g5 might prove crucial.

14.exd5 8

15 ... h5 This position was reached in Nguyen Kim ­ Le Thi, Vietnam 1 994. Here I propose:

7

16.a3N

5

Holding up the queenside play, and forcing Black to come up with a good counter to the g4-g5 plan.

4

6

3 2

A2) 10 ... e5 1 1 .tLlde2 h5

1 a

8

b

e

d

e

f

g

h

A very interesting strategic situation. The computer assesses it as comfortable for Black, with an evaluation veering towards the dreaded "+", but that is incomprehensible to me. Black's king lacks a safe haven, and he has virtually no counterplay on the queenside; or at least it will be extremely slow to materialize. I f he ever opts for kingside castling, as the computer so cheerfully does in many lines, White will j ust

7 6 5 4 3 2

a

b

e

d

e

f

g

h

208

Classical Main Line

plonk his bishop on e4 to secure the d5- and c2-pawns, then put his knight on h5 before starting a terrifying kingside pawn assault. Even in the centre, the black king cannot feel too comfortable, since White can either advance on the kingside in the same way, or perhaps even play Wb 1 followed by opening the c-file.

not seem very stable to me, and White gets a good position with either 1 8 . lLle4!?, or 1 8 . .ie4 followed by lLl h 5 . Th e text move intends . . . .ig5 followed by . . . We3 , but White can easily avoid the queen exchange and maintain his positional advantages.

14 ...i.b7N Trying to delay or avoid . . . 0-0 altogether. 1 4 . . . b4 1 5 . lLl g3 was good for White in Grabarczyk - Matuszewski, Poland 2007. Black faces the same type of problems, and the premature advance of the b-pawn only leads to a loss of flexibility on the queenside. Another possible line is 14 . . . 0-0N 1 5 .lLl g3 Wd7 1 6 . .id3 f6 1 7. lLl h 5 .id8 1 8.g4 .ib7 1 9 .h4 WO 20 ..ie4 E!:c8 2 1 .W b l with good prospects for White. Some of Black's moves may appear strange, but the computer approved of all of them. Even now the machine seems to think Black is fine, but the inevitable g4-g5 break, which White can prepare in his own sweet time, makes me strongly question this assessment.

a

b

e

d

e

f

g

h

17.i.e4 1 7.Wg4!? g6 1 8 . .ie4 is another idea, avoiding the . . . .ig5 plan altogether.

17 ... i.g5t 18.Wb l �e3 19.�h5! White will soon drive the enemy pieces away with E!:d3 and h2-h4.

8

B) 10.g4

7 6 5 4 3 2 1 a

b

e

d

e

f

g

h

1 5.�g3 !!c8 16.i.d3 �c5 After 1 6 . . . .id8 1 7.h4 f6 Black's set-up does a

b

e

d

e

f

g

h

209

Chapter 1 2 - Three-Piece System - Main Line This is the big main line. Now BI) IO ... h6 is a significant option, but by far the most important move is B2) IO ... h5.

BI) IO ... h6 1 1 ..txf6 Lf6 IV;t>h l !? 1 2 .h4 leads to similar variations, and it is useful to look at the following line to understand why I prefer the text move. 1 2 . . . �b6 1 3 . ttl b3 ( 1 3.ttlde2 deserves attention, but that is another story.) 1 3 . . . �c7 ( 1 3 . . . ttl c5 led to a nice victory for me a long time ago: 1 4.ttlxc5 �xc5 1 5 .e5! dxe5 1 6.ttle4 �a5 1 7. ttl d6t \t>m 1 8 .g5--+ Negi - Efimenko, San Marino 2006.) 1 4.e5 dxe5 1 5 .f5 ttl b6! 1 6. ttl e4 exf5 1 7.gxf5 id7 8 7

BI I) 12 ... g5 8 7 6 5 4 3 2 1 a

b

e

d

e

f

g

h

This possibility seems to be the main argument for preferring 1 2.h4 over 1 2.\t>b l , but here I found a new and exciting idea:

13.e5!N dxe5 14.�xe6!

6

It is hard to resist attaching two exclamation marks to such an idea!

5 4 3

1 4. fxe5 'lWxe5 1 5 . ttl e4 gives White decent compensation too, but it is much less exciting.

2

a

b

e

d

e

f

g

h

As we will see a bit later, this same position offers White a lot more scope with \t> b l played instead o f h2-h4. As things stand, after 1 8 .ttla5 0-0-0 1 9 . ttl d6t \t>b8 20.�xb7t 'lWxb7 2 1 .ttl dxb7 �c8 22.ttld6 �c7 Black equalized fairly easily in Shirov - Dominguez Perez, Sofia 2009. At this point Black has two main options: B I I ) 12 ...g5 and B12) 12 ...�h6. 12 . . .ie7 1 3 .h4 b5 1 4 .g5 gives White a nice position with an automatic attack, for instance: 1 4 . . . b4 1 5 . ttl ce2 ttl c5 1 6.f5 e5 1 7. f6 gxf6 1 8 .gxf6 if8 1 9 . ttl f5 ± Kloster - Petukhov, corr. 20 1 1 .

14 ... fxe6 1 5.fxg5 Despite being a piece down, White's pieces are almost perfectly placed.

1 5 ....bg5 1 5 . . . hxg5 1 6.�xd7± 8 7 6 5 4 3 2 1 a

b

e

d

e

f

g

h

210

Classical Main Line

16.tLle4 Threatening a nasty check on d6. The following analysis is certainly not exhaustive, but it clearly demonstrates that White has a huge initiative.

16

••

JH'8

After 1 6 . . . �e7 1 7.�g2 it is virtually impossible for Black to continue development. 1 7 . . .!! ffi 1 8 .'\Wh3� 16 . . . �f4 runs into another amazing idea: 1 7.g5!

23.l'!xd7 �xd7 24.�xfBt 'it>a7 25 .�c5 t, when White simplifies to a favourable endgame.

19.h4 .ie3

1 9 . . . �e7 20 .�xh6± 8

7 6 5 4

8

3

7 6

2

5

1

4

a

3

b

e

d

e

f

g

h

20.g5 ci>c7 2 1 .gxh6 ti'b4

2

a

b

e

d

e

f

g

h

1 7 . . . tLl b6 ( l 7 . . . hxg5 I B .h4!! I hardly even know how to explain this, but the kingside is being torn open, and Black has no pieces to bring to its defence.) I B .gxh6 �d7 ( l B . . . tLld5 1 9 .1'!gl �d7 20.�h5t 'it>dB 2 1 .c4�) 1 9 . tLl d6t 'it>fB 20.�g4 �a4 2 1 .b3t

17.ti'h3 ci>d8

Trying to run away via c7.

After 2 1 . . . b5 22.tLlg5! �xg5 23.hxg5 Black will not be able to cope with the connected passed pawns.

22.i.g2 ti'e7 23.l'!h3 i.f4 24.h7 ti'g7 25.�d6 ci>b8 8 7 6 5

17 . . . �a5 I B .�h5t 'it>e7 ( l 8 . . . 'it>dB transposes to the main line below.) 1 9.�g6 tLl f6 20.tLld6 tLld5 2 1 .�g7t 'it>xd6 22 .�xfB t 'it>c6 23.c4±

4

1 8.ti'h5 ti'a5

1

I B . . . �b6 1 9 .h4 �e7 20 .�xe5 �c7 2 1 .�xe6 �c6 22.�e5 'it>eB 23 .�d4 �b6 24.�d2 gives White a continuing initiative. I also considered I B . . . �c6 1 9.�g2 �f4 20 .�h4t 'it>c7 2 1 .l'!d6 �c4 22.�e7 'it>bB

3 2

a

b

e

d

e

f

g

h

26.l'!c3 �f6 27.ti'e2 �d5 28.l'!xc8t l'!xc8 29.tLlxc8 ci>xc8 30.ti'c4t ci>b8 3i .i.e4± Black's position is collapsing.

21 1

Chapter 1 2 - Three-Piece System - Main Line

A couple of examples are 1 5 . . . �c6 1 6.�xc6 bxc6 1 7. fxe6 fxe6 I B . ltJ e4;!;:, and 1 5 . . . 0-0 1 6. fxe6 fxe6 1 7.i.d3;!;:.

B12) 12 YNb6 •••

8

16.tLle4 exf5 17.gxf5 .id7

7

In the Shirov - Dominguez game mentioned earlier, White had the pawn on h4, which is pretty useless in the majority of lines. It is not immediately obvious why 'it> b l might be useful, but one concrete point will become clear in the main line below.

6 5 4 3 2

a

b

e

d

e

f

g

h

This is a normal move in these positions, but it enables White to get an improved version of the Shirov - Dominguez game mentioned in the note to move 1 2 .

13.tLlb3 YNc7

After 1 3 . . . ltJc5?! 1 4 .e5 dxe5 1 5 .ltJxc5 �xc5 1 6.ltJe4 �c6 1 7.i.g2 Black was in trouble in Sulskis - H. Hunt, Warsaw 200 5 . 1 3 . . . i.e7N just gives White easy play: 1 4.h4 ltJ c5 1 5 .f5 i.d7 1 6.g5;!;:

14.e5!N

This is the only way to justifY the placement of the knight on b3. Black was threatening . . . i.xc3 , and was also ready for the usual queenside play with . . . b 5 . Although the text move is technically a novelty, the same idea has been played more than once after 1 2.h4, and for the next few moves we will follow one of the critical games from that line.

14 dxe5 ..•

1 4 . . . i.e7 1 5 .exd6 i.xd6 1 6J:M4 i.e7 1 7.h4;!;:

1 5.£5 tLlb6

Other moves give White a clear edge, as he will be able to ruin Black's structure after fxe6.

a

b

e

d

e

f

g

h

18.tLla5

I B . ltJ d6t deserves attention, but forcing the king to f8 does not give us any immediate returns, so there is no need for this. I B . ltJ bc5!? makes some sense, as after I B . . .i.c6 Black cannot play . . . 0-0-0 anymore, but I prefer the text more.

18 0-0-0 19 .ig2!? •••

•

I believe this is the right time to deviate from the parallel Shirov - Dominguez game. 1 9 . 1tJ d6t 'it>bB 20.�xb7t �xb7 2 1 .ltJ dxb7 �cB would be like the game, and even though White has a better version here, it is not obvious if he can claim an advantage. Another idea is 1 9 .�d3!?, forcing 1 9 . . . i.c6 20.ltJxc6 �xd3 2 1 .cxd3 {this would not be

Classical Main Line

212

possible with the king on c l ) 2 1 . . . Wxc6 22 . .ig2 . White seems to have terrific compensation for the pawn. His opposite-coloured bishop is clearly superior, but it is still not easy to break through. 22 . . . c;t>b8 23.l'k l Wd5 24.l'k5 Wd7 2 5 .l'k2 Wd5 Black is j ust holding on, although White still has the better prospects.

8 7 6 5 4 3

8

2

7

1

6

a

5

b

e

d

e

f

g

h

22J�d3!

4

Black is unable to exchange rooks because the opening of the c-file will prove fatal. This would not have been the case with the king on c l !

3 2 1 a

b

e

d

e

f

g

h

19 .ic6 20.tLlxc6 Y;Vxc6 •••

20 . . . bxc6?! 2 1 .Wa3 regains the pawn while keeping a clear advantage.

21 .Y;Vh3 Y;Vc7 2 1 . . .c;t>b8 This move gives White a pleasant choice. 22.lt'l xf6 22.Wxf7 gxd l t 23.gxd l gc8 24.Wb3 also offers White better chances due to the higher potential of his light-squared bishop. 22 . . . Wxf6 22 . . . Wxg2 23 .Wxb6 gxd 1 t 24.gxd l gxf6 2 5 .Wxf6 ge8 26.Wxf7 gc8 27.Wb3 Wxh2 28 .a4! Even though the material is equal, White has a lot of scope to create threats against the enemy king. 23.a4!? The a4-a5 threat forces Black to weaken his queenside. 23 . . . a5 24.Wb5 24 . .ie4!? is also good. 24 . . . gd6 2 5 .ghe U :

22 tLlc4 23.tLlcS ••.

23 .gc3 also makes sense, but the text move leads to an excellent position.

23 tLld2t 24J�xd2 gxd2 2S.i.xh7t 'it>d8 26.tLle4 gd7 27.i.dS; •••

Black is still holding his position together, but it looks extremely unpleasant to play, while White can continue improving his position in various ways.

B2) IO hS •.•

8 7 6 5 4 3 2 1 a

b

e

d

e

f

g

h

213

Chapter 1 2 - Three-Piece System - Main Line This i s the main theoretical path, which has been tested in , thousands of games.

1 1 ..ixf6 White exchanges his bishop to make way for the g-pawn. Here we will consider the underrated B2 1) 1 l ... .ixf6 followed by the standard B22) l l ... tLlxf6. 1 1 . . .gxf6?! This has been played in a lot of games, with surprisingly okay results, but I believe this is mostly because a lot of White players have not found the right way to punish it. 1 2.f5! It is possible to take a slower approach, but it makes sense to launch an immediate attack as Black will not get enough time to recover. Notice how quickly we can zone in on the weak e6-pawn, with plans such as It:l ce2-f4 and %Vh3/g4-g5 hanging in the air. 1 2 . . . lt:le5 12 ... lt:l c5 has been played successfully by Movsesian, but White can punish it immediately: 1 3 .g5! fxg5N ( l 3 . . . b4 1 4. fxe6 bxc3 1 5 .exf7t gave White a crushing attack in Casallas Abril - Camargo, Bogota 20 1 0 .) 1 4 . fxe6 ixe6 ( l 4 ... fxe6 1 5 .e5 followed by %Vh5t destroys Black's position.) 1 5 .lt:ld5 ixd5 1 6.exd5± Followed by It:l c6 and ih3 .

13 . . .id7 is met by 1 4.lt:lce2±, when the threats created after It:lf4 are too hard to deal with. The text move has been the most popular, but Black seems to be castling into fire. 1 4.E:gl 1 4. lt:l ce2 has also done well for White, but one good line is enough. 1 4 . . . mhB 1 5 .lt:l ce2 E:gB 1 6.E:g3 ib7 1 7.ig2 8 7 6 5 4 3 2

a

b

e

d

e

f

g

h

1 7. . .d 5 ! ? 1 7 . . . E:acB I B . fxe6 fxe6 1 9. 1t:l f4± Yilmaz Darini, Kemer 2009. I B . lt:l f4! ic5 ? Black should prefer I B . . . E:acB as played in Gaida - Jedrzejowski, corr. 1 992, although 1 9 .exd5 N still gives White the better chances. The text move allows a thematic tactical refutation. 8

8

7

7

6

6

5

5

4

4

3

3

2

2 a a

b

e

d

e

f

g

h

1 3 .%Vh3 0-0 1 3 . . . b4 1 4. lt:l ce2 just helps White.

b

e

d

e

f

g

h

1 9 .%Vxh7t! mxh7 20.E:h3t mg7 2 1 .lt:l dxe6t fxe6 22.lt:lxe6t m f7 23 .E:h7t+Hamalainen - Gullulu, Arnhem 1 9B9.

214

Classical Main Line 17 . . . �b6 1 8 .l'!xd6 �c7 1 9 .�d4;!; leaves Black a pawn down with little counterplay.

B21) 1 1 ...hf6 8

8

7

7

6

6

5

5

4

4

3

3

2

2

1 a

b

e

d

e

f

g

h

This move definitely deserves more attention than it has been given.

12 ..ixb5!

White had better go for the forced stuff, as after the more popular 1 2 .g5 Black seems to get a fully playable position.

1 a

b

e

d

e

f

g

h

18.�g3!N 1 8 .�e3 has been the usual choice, as well as the computer's suggestion, but after the accurate 1 8 . . . �a3 ! White does not seem to have anything better than 1 9 .l'!b 1 0-0 20.tiJxb5 �a5t 2 1 .�c3 l'!xb 5 22 .�xa5 l'!xa5 with a drawish endgame.

12 J�b8 ••

The critical move, threats against b2.

creating

immediate

12 ... axb5 ? 1 3 . tiJ dxb5 �a5 1 4.tiJxd6t rJ:Je7 1 5 .e5± is j ust horrible for Black.

18 0-0 •••

The main point of White's last move is that 1 8 . . . �a3 ?? is refuted by 1 9 .tiJxb5 �xg3 20.tiJxd6t.

19.a4 .id7 With 1 2 . . . 0-0 Black shows he is willing to play a pawn down, but he does not seem to have enough compensation for it. 1 3 .ixd7 ixd7 1 4.g5 ixd4 1 5 .l'!xd4 l'!ab8 1 6.�d3± Fier - Williams, Reykjavik 20 1 1 . The next few moves are logical and relatively forced.

H.g5 .ixd4 14 .ixd7t .ixd7 1 5Jhd4 �b6

1 9 . . . ic6 20.l'!b 1 �a3 2 1 .l'!xb8 l'!xb8 22 .�d3 d5 23.exd5 exd5 24.tiJxd5 also favours White. Despite the open nature of the position, I do not see how Black can really exploit the open king, particularly with the knight around. If Black retreats with the queen, White should j ust move the knight away, probably to e3, and keep a solid edge.

•

1 5 . . . l'!xb2 1 6.rJ:Jxb2 �b6t 1 7. tiJ b 5 ! is a nice trick.

16.�d3 �xb2t 17.c;td2 .ib5

20J�bl �a3 21 .l'!xb8 �hb8 22.�d3�

White's king might seem vuln'erable, but it is actually pretty safe on d2, particularly since Black's bishop has limited prospects. In the

215

Chapter 1 2 - Three-Piece System - Main Line likely event o f an endgame - for instance after . . . 1'l:c8 and 1'l:c4 - the centralized king will be a great asset. Therefore it seems to me that Black is destined to suffer.

B22) 1 1 ...ttlxf6

B221) 1 3 .ixg5t?! 14.c.!?bl •..

I find it hard to explain the popularity of this variation. Most of the lines end up in an opposite-coloured-bishop middlegame where Black has few prospects and White keeps a constant, nagging initiative. We will analyse B22 1 1) 14 ttle5 and B22 12) 14 0-0. .•.

•••

B22 1 1) 14 tLle5 •.•

8 7 6 5 4 3

This is the main move of course.

2 1

12.g5 ttld7 13.£5 Leaving the g5-pawn hanging is standard practice for this line. We will analyse B22 1) 13 .ixg5t?!, B222) 13 ttlc5 and the currently fashionable B223) 13 ... 0-0!? • . •

. • .

13 . . . tt:le5?! places the knight on a superficially good square, but it neither exerts any pressure on e4, nor defends e6. 1 4 .%I[g3 0-0 ( l 4 . . . b4 0nly helps to improve White's knight: 1 5 . tt:l ce2 i.b7 [ 1 5 . . . 0-0 1 6. tt:l f4±] 1 6.fxe6 i.xe4 1 7.i.g2±) 1 5 .'iit b 1 ( l 5 .f6!? also deserves attention. The bishop will be awkward on d8, and White may launch a full-throttled kingside attack with h4-h5 or tt:l ce2-f4.) 1 5 . . . 1'l:e8 1 6.fxe6 Not the most ambitious choice, but it keeps things simple and is more than enough for a slight edge. 16 . . . fxe6 1 7. tt:l ce2! i.d8 1 8 . tt:l f4 %l[e7 1 9 .h4t Rowies - Maia, corr. 20 1 1 .

b

a

e

d

f

e

g

h

This is the most popular choice, but White only needs a few accurate moves to secure an edge.

1 5.%Yh5 %Ye7 1 5 . . . %I[d8 1 6.h4 The most precise way. 1 6.tt:lxe6 would be slightly less accurate here, as a subsequent tt:l d5 will not hit the queen. 1 6 . . . i.f6 1 7. fxe6 0-0 1 8 .i.h3 8 7 6 5 4 3 2 a

b

e

d

e

f

g

h

216

Classical Main Line

1 8 . . . 'kt>h8 1 8 . . . fxe6 1 9 . .txe6t 'kt>h8 20.tLld5 is the same thing. 1 8 . . . g6 1 9 .We2 does nothing, apart from giving White the attacking option of h4-h 5 . 1 9 .tLld5 fxe6 20 . .txe6 tLl c4 20 . . . .tb7 2 1 .E:hfI g6 22.We2 .txh4 23.E:h l gave White a powerful initiative in Lj ubicic - Karacsony, corr. 20 1 2.

1 8.exf7t
8 7 6 5 4 3 2

a b

a

e

d

e

f

g

b

e

d

e

f

g

h

2 1 ..te2

h

2 1 . tLl c6 We8 22 .Wxe8 E:xe8 23.tLl xf6 E:xe6 24.tLld5 E:xe4 2 5 . tLl c7!± Luther - Vink, Wijk aan Zee 200 1 . I remember playing through this game as a kid and being impressed by the way White trapped the rook.

Funnily enough, the above position was reached in a recent GM game with Black to move, as White reacted to 1 5 . . . Wd8 with the imprecise 1 6.tLlxe6. Despite the lost tempo, White still had little trouble demonstrating his long-term initiative. Play continued: 2 1 . . . E:a7 This is not the most accurate, but it is still useful as a free move. 22.Wg3 h5 23.E:g l .th6 24 . .te2 'kt>h7 2 5 . E:dfl h4 26.Wf2± Van Kampen - Fedorowicz, Montreal 20 1 4. 2 1 .Wc3!? is arguably the most precise way to continue, and has been popular among correspondence players. However, I don't see the need for anything fancy, as White keeps an edge with simple developing moves.

2 1 ...E:f8 22.E:hf1 ;1; With a one-sided game. a

b

e

d

e

f

g

h

16.lihe6 .ixe6 17.fxe6 g6 1 7 . . . 0-0 1 8 .tLld5 Wd8 1 9 . .th3 is also good for White.

Chapter 1 2 - Three-Piece System - Main Line

B22 12) 14 ... 0-0

217

19.%Yxg5 exd4 20.i.d3 g6 Black is barely hanging on, but now he plans . . . .ifS to simplifY the position.

8 7

2 1 .�hf1

6

2 1 .�hg 1 is met by 2 1 . . .Wlg7 intending . . . .ifS .

5

21 ...%Yg7

4 3 2 1 a

b

e

d

e

f

g

h

This is a more forcing way, which at least prevents White from playing with an easy initiative. At some point I thought this might be an interesting way to revamp this line, but White emerges on top with a bit of a precision.

15.fxe6 c!£lb6

2 1 . . . .ih3 22.�g1 ! Wlg7 23 .Wlg3 shows the point behind the �f1 -g 1 finesse. 23 . . . .ifS 24 . .ixfS �xfS 2S .Wlxd6 �eB 26.WlcS �dB 27.�xd4 White's superiority has been demonstrated in a few correspondence games, one example being Kevicky - Raidaru, corr. 20 1 0.

22.�xfSt %YxfS 23.i.xg6! The critical move. 23.�f1 WleB 24.Wlh4 �a7 2 S . a4 is a more subtle try, but it is not easy to prove an advantage.

23 ... hxg6 24.%Yxg6t %Yg7 25.%Ye8t i>h7

A series of logical moves follow.

16.c!£ld5 c!£lxd5 17.exd5 fxe6 18.%Yg4

a

a

18 ... e5

b

e

d

e

f

g

h

After 1 B . . . .if6 1 9 .1L'lxe6 .ieS 20 . .ig2 Black will have to exchange on e6 sooner or later, leaving White with much the better bishop, not to mention a safer king.

b

e

d

e

f

g

h

26.�f1 i>h6 26 . . . .ib7? runs into 27.WlhSt �gB 2 B .Wld l ! with the deadly threat of �g l .

27.%Yel %Yg5 27 . . . � h S ? is refuted by 2B .Wle2t (2B.�g1 WleS !oo) 2B ... �h6 29.Wlf2! followed by �gl .

218

Classical Main Line

35 b3!?

2SJ:H8

•..

35 . . . bxa3 36.Wxd6 would also lead to suffering for Black.

Continuing the domination theme.

2S b4 .•.

28 . . . .ib7? 29 .We6t wins quickly.

36.%Yd7t 'it>g6

28 . . . i>g7 29.h4 We5 30 .Wf2 maintains a big advantage, for instance 30 . . . a5 (30 . . . ga7 3 1 .gxc8±) 3 1 .a3+- and the queen is ready to move.

36 . . . i>g8 loses to 37.Wc8 t followed by Wb7t and Wxb3.

37.%Yxd6t 'it>h5 Now White can neutralize his opponent's activity with:

8 7 6 5 4 3 2 1 a

b

e

d

e

f

g

h a

29.%Yf2!N Improving over 29.Wfl We7 30.gh8t i>g7 3 1 .gh5 i>g6 32.gh8 i>g7 when White could only draw in Motylev - Sj ugirov, Rijeka 20 1 0.

29 J�a7 .•

29 . . . We7 30.gh8t i>g7 3 1 .Wxd4t We5 32.Wh4! +- is the point behind putting the queen on f2.

30.gxcS gf7

e

d

e

f

g

h

3S.%Yg3! bxc2t 39.'it>xc2 %Yxd5 40.h4 With excellent winning chances. Note that if the pawn had gone to a4 instead of a3 on move 3 5 , Black could have given an annoying check on b4 at this point.

B222) 13 tLlc5 •••

8

White has a few ideas here, but the simplest is to convert to a favourable queen endgame.

7

31 .ghSt 'it>g7 32.gh7t 'it>xh7 33.%Yxf'7t 'it>h6 34.%Ye6t 'it>g7 35.a3!

5

White can also start with 3 5 .Wd7t. However, 3 5 . a4 is less accurate as 3 5 ... b3! gives Black chances to resist.

b

6

4 3 2 1 a

b

e

d

e

f

g

h

Chapter 1 2 - Three-Piece System - Main Line

14J�gm This is an old idea of Boleslavsky, which ' has recently been taken up successfully by correspondence players. White isn't in a hurry to define the kingside structure, and Black does not seem to have a good plan to fight against the impressive pawn behemoth. 1 4.f6 gxf6 1 5 .gxf6 .ifB 1 6.l"i:g l has traditionally been considered dangerous, but in recent years correspondence players have been holding the fort for Black in this line.

219

1 5 . . . .ib7 also falls short of equality: 1 6.fXe6 0-0 ( l 6 . . . fXe6 1 7. lLI f4±) 1 7. lLI f4 .ixe4 ( l 7 . . . lLI xe4 1 B .Wi'h5 d5 1 9.exf7t l"i:xf7 20.lLI de6 Wi'e5 2 1 .g6 Wi'xh5 22.lLIxh5 hxg6 23.l"i:xg6±) This position arose through a slightly different move order in Tidman - D. Coleman, Sunningdale 2009. Here White's simplest path to an advantage would have been: 8 7 6 5

14 ... b4 The only way for Black to offer resistance is by trying hard for counterplay. I will show a bad move to illustrate White's idea in this line: 1 4 . . . l"i:bB? 1 5 .fXe6! Now of course Black should not take with the knight and allow lLId5, while taking with the bishop allows lLI f5 . So 1 5 . . . fXe6 is more or less forced, but then 1 6.g6! fXg6 1 7.l"i:xg6 l"i:h7 I B . .ih3 leaves Black without a good defence.

4 3 2

a

b

e

d

e

f

g

h

1 B .exf7tN l"i:xf7 1 9 .Wi'g3 Wi'b7 20.g6 hxg6 2 1 .lLIxg6 .if6 22.Wi'h3 .ixg6 23.l"i:xg6 Wi'e4 24.l"i:g4 Wi'e5 2 5 .Wi'g3;!; Despite the simplifications, White maintains some pressure.

16.�b3! 1 6. f6? exd4 1 7. fXe7 d3! I B .cxd3 b3 was seen in the stem game Boleslavsky - Aronin, Moscow 1 9 57, where White was lucky to escape with a draw.

8 7 6 5

16 ... g6

1 6 . . . d5 can be met by the blunt 1 7. lLI g3 lLIxb3t I B .Wi'xb3 .ixg5t 1 9.<;t>b l d4 20.lLIh5 with good attacking chances, Penzler Migliorini, corr. 20 1 2.

4 3 2 1 a

b

c

d

e

f

g

h

15 ... e5 This has been the usual choice in correspondence games, but now the kingside pawns are extremely strong.

1 6 . . . lLI d7 1 7.<;t>b 1 a5 (After 1 7 . . . .ib7? 1 B . lLI g3 0-0 1 9 . f6 Black resigned in Fajs - Achermann, corr. 2009.) I B .lLI g3 a4 1 9. 1LI d2 The queenside attack is obviously much less impressive than White's kingside play. The one example on my database ended as follows: 1 9 . . . b3 20.axb3

220

Classical Main Line

.ixg5 (20 . . . axb3 2 1 .ltJxb3 .ixg5 22.ltJh5±) 2 1 . ltJ c4 .ie7 22.ltJh5 axb3 23.ltJxg7t �f8 24.cxb3

a a

b

e

d

e

f

g

h

24 . . . .ib7 2 5 .'Wg4 'Wc6 26.�d5 ltJ f6 27.ltJe6t fxe6 2S .'Wg7t 1 -0 J. Fernandez - Flores Ramos, corr. 20 1 0.

17.�bl a5

1 7 . . . .ib 7 has also been played in some correspondence games, but White has scored heavily. A few examples: l S .ltJxc5 'Wxc5 ( l S . . . dxc5 1 9. f6 .if8 20.h4 h5 2 1 ..ih3 .ic6 22.c4± Valinova - Gromotka, corr. 20 1 3.) 1 9 . 1tJ c l h6 ( l 9 ... gxf5 N 20.'Wxf5 .ics 2 1 .'Wf3 a5 22.b3! .ie6 23 . .ih3±) 20.'Wh3 i.xe4 2 1 ..id3

e

d

e

f

g

h

lS.h4

White has a much more pleasant position. The following game is a good illustration:

lS ...a4 19.tlhc5 %Vxc5 20.£6 AdS 21 .h5 Ab7 22J�hl �f8 23.ti:lc1 .tb6 24.%Vg4 %Vc6 8 7 6 5 4

8

3

7

2

6

b

1

5

a

4

b

c

d

e

f

g

h

25.gh2

3

Preparing to centralize the rook. Black cannot take the e4-pawn because the check on b5 would be devastating.

2

a

b

e

d

e

f

g

h

2 1 . . . gxf5 (2 1 . . . .ixf5 22 . .ixf5 gxf5 2 3 . ltJ d3 'Wc4 24.gxh6 also favoured White in Barlow - Joao, corr. 200S.) 22.gxh6 .if6 23.h7 'Wc6 24 ..ixe4 fxe4 2 5 .�dfl �e7 26.'Wh6 d5 27.�g2!+­ Barlow - Santana Penate, corr. 2009.

25 ....id4 26.ghd2 gxh5 27.%Vxh5 0-0-0 2S.tl:le2� Black was able to castle, but his position still ' contained numerous weaknesses in Cremasco - De Filippis, corr. 20 1 2 .

22 1

Chapter 1 2 - Three-Piece System - Main Line

1 5 . . . ltl e 5 ? misplaces the knight, and 1 6.W'g3 ± leaves Black with big problems.

B223) 13 ... 0-0!? 8

B223 1) I S ....ib7!?

7 6 5 4 3 2 1 a

b

e

d

e

f

g

h

Despite being nowhere near as popular as the previous two moves, this is perhaps the most critical continuation of all. Castling looks extremely provocative, and White has to be careful not to be sucked into the trap of trying to play for mate too quickly. Instead we will aim for a pleasant version of a typical structure.

a

b

e

d

e

f

g

h

Black tries to keep the tension for as long as he can, while hinting at ideas like . . . d5.

16 ..ig2!? Anticipating the central pawn break.

14.h4!? Several direct attacking ideas have been tried, but I did not find any of them fully convincing.

Mter 1 6. fxe6 ltlc5! Black regained the pawn with a decent game in O' Brien - Wood, email 20 1 2.

1 4.l'!gl has been the most popular choice, and can be played with similar ideas to the text move. Mter 14 . . . b4 1 5 . ltl ce2 e5 1 6.f6 exd4 1 7.fxe7 l'!eB I B .ltlxd4 ltle5 the position is similar to the main B232 line as featured below, with a few different details. The basic premise - that White should be better in this structure once he neutralizes Black's immediate threats - should hold, but I prefer to aim for the position with h2-h4, as a subsequent h4-h5/g5-g6 is a typical plan, whereas the rook on gl does not serve an immediate purpose.

1 6.ltlg3 is a natural move, but now 1 6 . . . d5! creates a mess. ( l 6 ... e5 1 7.f6! gives White excellent attacking chances.) There are many ideas to explore here, but overall I find the main line more appealing.

14 ... h4 IS.tll ce2 Now it makes sense to consider IS ....ib7!? and B2232) IS ... eS.

B223 1)

16 ... gac8!? Another noncommittal move, escalating the tension even further. A few other ideas: 1 6 . . . d5 1 7.exd5 .ixd5 1 B .W'h3t is not too worrying here. 1 6 . . . e5 1 7.f6 exd4 ( l 7 . . . .ixf6 1 B .ltlf5 .idB 1 9 .1tlxd6 .ic6 20.g6! fxg6 2 1 .W'b3t <;t>hB 22.ltlf7t l'!xf7 23 .W'xf7±) 1 B .fxe7 l'!feB 1 9 .1tlxd4

222

Classical Main Line 1 9 . . . �xd6 2o.Wfxd6 Wfxd6 2 1 .:B:xd6 lDc5 22.lDg3 lDxe4 23 .�xe4 �xe4 24.:B:e l �c6 24 . . . �b7?! 2 5 . :B:xe8t :B:xe8 26.:B:b6± 2 5 .:B:xe8t �xe8 26.h5± The pawn is going to h6 next, and White keeps a definite plus in the endgame.

8 7 6 5 4 3 2

a

b

e

d

e

f

g

h

1 6 . . . :B:ac8 ! ? keeps the position in flux for even longer; there are many interesting possibilities, but I like the following line, which is surprisingly almost forced on Black's side.

1 9 . . . lD c5 ( l 9 . . J�xe7 2o.lDf5±) 2o.lDf5 �xe4 2 1 .Wfh3 �xf5 22 .Wfxf5 White has obtained the typical structure for this line. The queenside seems to be well under control, and the h 5-g6 attacking plan is on the cards. Another sensible continuation is: 16 . . . exf5 1 7.lDxf5 :B:fe8 1 8 .Wfg3 I also considered 1 8 . ';t>b l �f8 1 9 .Wff4 :B:ac8 , and now the aggressive 20.lDed4! ? lDc5 2 1 .h5 lDxe4 22.g6 is interesting, bur White does not seem to have more than a draw. 1 8 . . .�f8 1 8 . . . lDe5 1 9 .'kt>b l :B:ac8 20.lD ed4± Black's counterplay is at a standstill, and the simple h5-g6 plan gives White clearly better chances.

a

b

b

e

d

e

f

g

h

1 9 .1Dxd6 This leads to heavy simplifications, but the upcoming endgame is not as equal as you might first think.

d

e

f

g

h

17.fxe6 fxe6 18.Wfh3 e5 19.Wfe6t �h8 20.�xe7 exd4

After 20 . . . :B:fe8 2 1 .Wff7 exd4 22.lDxd4 :B:f8 23 .Wfb3± Black does not have enough compensation for the pawn .

2 1 .�xd4 gee8

a

e

Chapter 1 2 - Three-Piece System - Main Line

22.�e6! Y;Yxc2t 23.xc2 �Hlt 24J�d2 gxd2t 25. xd2 gxe7 After so m:iny exchanges it may seem that we are heading for a draw, but in fact White continues to exert pressure after:

26.i.h3 i.xe4 27.gel! An excellent example of domination in the endgame.

27 ... �b6 27 . . . lLlcS 2s.lLlxcS dxcS 29.!!xcSt obviously favours White due to his more active pieces, especially the king.

28.gfl !? h6 2s . . . lLld7 29.lLldS!?t

223

16 ... exd4 17.fxe7 ge8 1 8.�xd4 �e5 19.Y;Yg2 1 9.'1Wf2 is also possible, but I am not sure if White is better in the upcoming line: 1 9 . . . !!xe7 20.hS oig4! Trying to generate quick counterplay. (2o . . . oib7 2 1 .oig2t gives White an ideal position.) 2 1 .oie2 oixe2 22.Wxe2 I would normally be happy with this structure, but here Black has the annoying 22 . . . Wc4! 23 .Wxc4 lLl xc4 and it's hard to assess how much better White is in this endgame. It could be worth exploring in further detail, but I like the main line even more.

19 ... gxe7 1 9 . . . Wxe7 is met by 2o.lLl fS oixfS 2 1 .exfS as (2 1 . . .!!ecS 22.f6 Wc7 23.hS--+) 22.oib S ! !!ecS 23 .oia4t when Black is stuck without any real counterplay.

29.e3 i.g6 30.d4!?� Black's extra pawn is irrelevant, and White's excellent pieces give him lasting pressure.

B2232) 1 5 ... e5

1 9 . . . oig4 is not much good here, because Black will not be in time to exchange the queens later: 20.oie2 oixe2 2 1 .Wxe2 !!xe7 22.lLlfS !!e6 23.hS White has typical, comfortable play for this structure. 23 . . . !!cS (23 . . . Wc4?! 24.Wxc4 lLlxc4 2 S .!!d4±) 24. lLl d4 !!eeS 2 S . g6 fxg6 26.hxg6 h6 27.i>b a 8 7 6 5 4 3

a

b

e

d

e

f

g

h

This is certainly a logical try for Black, but it runs into a thematic reply.

16.f6!N

Technically a novelty, although it has been played many times in the similar position with Elg1 played instead of h2-h4.

2 1 a

b

e

d

e

f

g

h

20.�f5! Preventing any counterplay with . . . oig4.

20 ...hf5

224

Classical Main Line

20 .. .!!e8 2 1 .h5 .te6 22.mb 1 gives White a pleasant position.

2 1 .exf5 gb8 It is a similar story after other moves: White's advanced kingside pawns give Black plenty to worry about. 2 1 . . J�ae8 22.f6 ge6 23.h5± 2 1 . . Jk8 Black does not have many threats. 22.f6 gee8 23.h5 IlNc6 (23 . . . lD g4 24.g6!) 24.'1Wxc6 gxc6 2 5 . .tg2 gb6 26.fXg7 mxg7 27.h6t±

22.�bl a5 23.£6 ge6 24.h5 24.b3 ! ?± is another idea, preventing . . . a4.

24 ... a4 25.fxg7 a3 25 . . . mxg7 26.g6! gives White a dangerous attack. 25 . . . b3 also proves insufficient after 26.cxb3 axb3 27.axb3 gxb3 28.gc l ± .

a

b

e

d

e

f

g

h

26.g6 hxg6 27.hxg6 gxg6 28.YlYh2 �xg7 29.J.g2 Black is in considerable danger.

Conclusion Just as in the previous chapter, the early 1 0.f5!? is a useful option which offers good chances to put the opponent under pressure without having to memorize too many complicated variations. Nevertheless, despite the appeal of this approach, the traditional main line with 1 O.g4 was too interesting a topic to ignore. 1 O . . . h6 is a reasonable move, but we saw that 1 1 . .txf6 .txf6 1 2. mb I ! ? poses a lot of problems to the defence, usually in conj unction with a sacrificial attack based on a timely e4-e5. The usual continuation is 10 ... b5 1 1 . .txf6 lDxf6 1 2.g5 lD d7 1 3 . f5 , when Black must make a pivotal choice. 1 3 . . . .txg5 t?! does not make much sense to me, as White soon regains his pawn while keeping the more active position. 1 3 . . . lDc5 is the old main line, but here 1 4.gg l ! ? seems like an excellent way to strengthen White's position without yet committing to f5-f6. Black seems to be under real pressure here. 1 3 . . . 0-0! ? is the most important option nowadays, but I still like White's chances after 1 4.h4!? b4 1 5 . lD ce2. Black's 1 3th move is rather provocative, and White should resist the temptation to throw everything into a hasty kingside attack. Instead, if he keeps an eye on the centre, he should be able to bring all his pieces into play and gradually build an attack from a position of strength.

Poisoned Pawn a

b

e

d

e

f

g

Introduction and 12 �d5!? ...

Variation Index 1 .e4 e5 2.� f3 d6 3.d4 cxd4 4.�xd4 � f6 5.� e3 a6 6 ..tg5 e6 7.f4 YlYb6 8.YlYd2 YlYxb2 9J�b 1 YlYa3 1 0.e5 A) 1 0...� fd7 B) 10...dxe5 (without . . . h6) C) 10...h6 1 1 .i.h4 dxe5 1 2.fxe5 � d5!? 1 3.�xd5 exd5 1 4.e6!? he6 1 5.�xe6 fxe6 1 6.i.d3 i.e7 17.i.g6t �d8 1 8.he7t YlYxe7 1 9.0-0 C l ) 1 9 ...b5? C2) 1 9 ...� d7 20J�f7 YlYe5t 2 1 .�h1 C2 1 ) 2 1 ..J�e8 C22) 2 1 ..JUS 22Jhg7 gf6 23.i.d3! �e7 24.e4 d4 25.YlYe 1 ! gaf8 26.YlYe4 C22 1 ) 26 ...g6f7!?N C222) 26 ...b6

B) afrer 2 l . . .1t>f7

C2 l ) afrer 2 5 " .l"1g8

226 227 23 1 233 234 234 238 239 24 1

C222) afrer 2 8 " .lt>d6

2 a

b

e

d

e

f

22.l"1b l !N

g

h

a

b

e

d

e

f

26.%Vh4t!N

g

h

a

b

e

d

e

f

29 .1i.c2!N

g

h

h

Poisoned Pawn

226

1 .e4 c5 2.c�f3 d6 3.d4 cxd4 4.�xd4 �f6 5.�c3 a6 6.i.g5 e6 7.f4 �b6 8 7 6 5 4 3 2 1

X •' ; .!. �.�

��

�£i � �� 'i�_' "" i %� � '� . �'5� �"""�� ' ' "z �"",I'�z '�· � � �.��!n·n'� � ""' ; "", z "" ' ;� � � � � ��' � ��' 0 � 8 �� 8 • • 8 � � " d" %� ;�r.� � � . �_�

a

b

e

d

e

f

g

h

The Najdorf Poisoned Pawn is one of the most notorious opening variations in chess. It is also one of the most theoretically challenging, and a favourite among correspondence players who know they will have enough time (not to mention computer assistance) to navigate the many pitfalls. I cannot promise to refute it, but I will certainly show you some interesting ways to cause trouble for your opponents.

8.�d2 �xb2 If 8 . . . lLl c6 then 9 .il.xf6 transposes to the Rauzer which I will cover in a later volume.

9J�bl �a3 10.e5 This is the most direct attacking plan, and I prefer it over the other main options of l O. f5 or, on the previous move, 9 . lLl b3 . In this chapter we will deal with the alternatives A) 10 ... c!LJfd7 and B) 10 ...dxe5 (focusing on lines where Black does not follow up with a quick . . . h6) , before beginning our investigation of the main line C) 10 ... h6.

A) 10 ... tlHd7 I spent a while analysing this variation back in 2006, but nowadays it's j ust a matter of

following correspondence games where White's advantage has been proven beyond a doubt. 8 7 6 5 4 3 2 1 a

b

e

d

e

f

g

h

1 1 .£5 c!LJxe5 1 1 . . .dxe5? 1 2 .fxe6! exd4 1 3 .exf7t @xf7 1 4.il.c4t wins quickly.

12.fxe6 fxe6 13.J.e2 Almost every game from this position has been won by White.

13 ... d5 Another noteworthy game continued 1 3 . . . lLl bc6 1 4.lLlxc6 bxc6 1 5 .lLl e4 d5 1 6.0-0 il.e7 1 7.�b3 Wfa4 1 8 .il.xe7 Wfxe4 1 9 .il.d6 lLl f7 20 .il.c5 and Black was busted i n Anand Polgar, Moscow (blitz) 2006.

14.0-0

1 4.�b3 ?! has been played several times, but simple main line is much more convincing.

14 ...J.c5 Both 14 . . . lLlf7 1 5 .il.h5 ! and 1 5 .�be 1 ! are winning for White.

14 . . . Wfd6

1 4 . . . lLl bc6 1 5 .il.h5t g6 1 6.lLlxd5! exd5 1 7.lLlxc6 lLlxc6 1 8 .�be l t il.e7 1 9 .Wfxd5± also leaves Black under heavy pressure:

Chapter 1 3 - Introduction and 1 2 . . . ttJ d 5 ! ?

227

12.�e4 Y;Yxal

8

Once again Black can insert 1 2 . . . h6, when 1 3 .ih4 leads to the next chapter. In the present variation we will see how the absence of these two moves may lead to some important nuances.

7 6 5 4

I3.gdl

3 2 1 a

b

e

d

e

f

g

h

16.�xd5! exd5 17.Y;Yxd4 Y;Yd6 lS ..th5t g6 19JU6 �bc6 20.gxd6 �xd4 2 1 .gdSt c,!,>f7 22.gxhS gxh5 23.gfl t �f5 24.gxf5t hf5 25.gxaS± White emerged with a technically won endgame in Gerasimov - Sekretaryov, corr. 20 1 0.

B) 10 ... dxe5 1 l .fxe5 8 7 6 5 4 3

1 3 .:1!b3 ?! This used to be considered interesting, but it has been more or less refuted by: 1 3 . . . h6! Black needs to drive the bishop back to h4 for his defensive idea to work. 1 3 . . . ttJ xe5 ? 1 4. ttJ b 5 ! is certainly not advisable for him. 13 . . . Wal t?! 1 4.�f2 Wa4 was first played in the famous game Tal - Tolush, Leningrad 1 9 56, which resulted in a spectacular win for White after 1 5 .ib5 ! ? Interestingly, Tolush repeated this position and won a subsequent game with the help of an improvement he found, but this is a moot point since 1 5 .ttJxe6! fxe6 1 6. ttJ d6t hd6 1 7.Wxd6 gave White a winning attack in Fekih - Tavinski, cOff. 2007. 1 4 .Y;Yc3 1 4.ih4 Y;Ya l t 1 5 .�f2 Wa4! is a fairly well­ known line that has been established as good for Black. 14 . . . ttJ c6! 1 5 .ttJ xc6 hxg5!

2

8

1

7 a

b

e

d

e

f

g

h

1 l ... �fd7 1 1 . . .h6 1 2.ih4 transposes to Variation C. 1 1 . . .ttJd5? is unplayable without the earlier . . . h6 and ih4 included. The problem is seen after 1 2 .ttJxd5 exd5 1 3 .:1!b3 Wa4 ( 1 3 . . .Wxa2 1 4.Wc3! is no better) 1 4.ttJ b5!+- when the bishop isn't hanging on h4.

6 5 4 3 2 1 a

1 6. ttJ a7

b

e

d

e

f

g

h

Poisoned Pawn

228

White's attack seems to be fading away so he tries to win some material , but Black has a lot of resources too. 1 6 . . . WI'a4 1 7.i.d3 �h4! 1 8 . lLl d6t i.xd6 1 9 .exd6 Wl'f4! 20.@d l Wl'xd6! 2 1 . lLl xc8 Wl'c5 This position has been reached in several correspondence/internet games. Black stands better, and has achieved a heavy plus score. Let's return to the main line, and to the question of who will benefit from the absence of . . . h6 and i.h4. Practical results seem to indicate that the difference favours White, but it is important to be aware of some details.

8 7 6 5 4 3 2

a

b

e

d

e

f

g

h

1 5 .WI'f4! i.b4t 1 6.c3 i.b7 1 6 . . . axb5 1 7. 0-0 0-0 1 8 .i.f6! lLlxf6 1 9 .exf6 g5 20 .WI'g4!± 8

8

7

7

6 5

6

4

5

3

4

2

3 2

a

1 a

b

e

d

e

f

g

h

1 3 Yfd5 ...

Apart from 1 3 . . . h6, when 1 4.h4 transposes to the next chapter, Black does not have many sensible alternatives. 1 3 . . . b6? This is a fairly pointless move, and White could possibly continue with a slow initiative, but it's nice to see how soon things can go bad for Black after: 1 4.i.b 5 ! ? Wl'd5 14 . . . WI'a5 ? 1 5 .c3! axb5 1 6. 0-0! +1 4 . . . h6 1 5 . 0-0! hxg5 1 6.WI'f2+14 . . . axb5 1 5 .lLlxb 5 100ks terrible for Black. 1 4 . . . i.b7 1 5 .i.xd7t! lLl xd7 1 6. lLl b 5 ! +-

b

e

d

e

f

g

h

1 7.lLlxe6! 1 7. lLl c6!? 17 ... WI'xe6 1 8 .i.c4 Wl'xc4 1 9 .1Ll d6t i.xd6 20.WI'xc4 i.xe5 2 1 . 0-0 0-0 22 .i.e7 b5 23 .WI'a2± 1 3 . . . WI'b2 Black might try to play along the lines of Radjabov - Karjakin (see variation A of the next chapter on page 249) without including . . . h6. 1 4 .WI'e3N 1 4.i.e2 i.b4 1 5 .c3 Wl'xd2t 1 6.�xd2 i.e7+ is no good for us. 1 4 . . . i.c5 14 . . . lLl c6 1 5 .lLlxc6 bxc6 1 6.i.e2i 1 5 . c3! This is a lot stronger without . . . h6 and i.h4 included.

Chapter 1 3 - Introduction and 1 2 . . . ttJ d S ! ? I S . .ie2?! tLl c6 1 6.c3 �a3 gives White a worse version of the Radjabov - Karjakin game. On � illustrative line is 1 7. 0-0 tLl xd4 I B .l"�xd4 .ixd4 1 9 .�xd4 0-0 20.tLlf6t gxf6! when there is no �g4 t, and Black threatens to exchange queens with . . . �cS . I S . . . �a3 1 5 . . . tLlc6 1 6.tLlxc5 tLl xcS 1 7.tLlxc6+1 5 . . . h6 1 6.tLlxc5! tLl xc5 ( l 6 . . . hxg5 1 7.tLlcxe6+-) 1 7. tLl fS ! hxgS I B .tLlxg7t wins for White. 1 5 . . . �b6 16 . .ie2 tLl c6 1 7. 0-0 0-0 I B . tLl f6t! tLlxf6 19 . .ixf6 gxf6 20.exf6 :B:dB 2 1 .�g5 t r,f{fB 22 . .ih5 :B:d7 23.�h l ! is also crushing.

229

I B . .ie2 has been played, and also looks to be a good move, but I am happy to proceed by analogy with the . . . h6 line. I B . . . �xeS I B . . . �a2 1 9 . .ie2 �b l t 20.�f2 �xh l 2 1 ..ih6! �xh2 22.�gS g6 23 . .ixfB �xfB 24.e6! is winning for White. 8 7 6 5 4 3

8

2

7 6

a

5 4 3 2

a

b

e

d

e

f

g

h

1 6.�g3! In the analogous position with . . . h6 and .ih4 included, Black could simply play . . . g5 here. Now .id is a serious threat, and Black can hardly castle. 1 6 . . . .ixd4 1 6 . . . g6 1 7 . .ih6+- sucks for Black as well. 1 7.tLld6t �fB I B .:B:xd4 White is winning.

14.YlYe3 YlYxeS 1 4 . . . .ic5 The analogous position featuring . . . h6 and .ih4 will be analysed extensively in variation B 1 of the next chapter. White should proceed in exactly the same way here: 1 5 .tLlxe6! .ib4t 1 6.c3 �xe6 1 7.cxb4 0-0 I B .:B:d6N

b

e

d

e

f

g

h

1 9 . .ic4! With the moves . . . h6 and .ih4 included, Black's best continuation is . . . �hS, so he faces a difficult choice. 1 9 . . . tLl c6 This seems like the best chance. 1 9 . . . :B:eB 20 . .ixf7t �xf7 2 1 .�b3t :B:e6 22.0-0t tLl f6 23.tLlxf6 gxf6 24.,ixf6 �xd6 2S . .ieS t wins for White. 1 9 . . . bS 20 . .if4! In the . . . h6 variation the bishop's retreat to g3 is crushing (see page 249 for details) , so here we go to f4 instead. 20 . . . Wb2 (20 . . . Wa I t 2 1 .:B:d l Wxd I t 22.�xd l ±) 2 1 ..ib3t 20.0-0 tLl b6 2 1 ..ib3 .ifS 2 1 . . .�hB 22 . .if4 �bS 23.tLlgS!± 22.tLlf6t gxf6 23.�xeS tLl xeS 24.:B:xfS fxgS 2 S .:B:xeSi Black has avoided being mated, but the endgame will certainly not be a picnic for him.

l S .ie2!? .ics .

1 5 . . . fS does not change much: 1 6. 0-0 ( l 6.tLlxe6! ? is possible, but the text move is simpler, as it should j ust transpose to our main

230

Poisoned Pawn

line.) 1 6 . . .'
16.0-0!? 1 6.i.f4 i.xd4 1 7.l'!xd4 Wla5t 1 8 .l'!d2 0-0 1 9 .i.d6 can be compared with variation B22 of the next chapter. White's pieces are on identical squares, but Black's h-pawn is on h7 instead of h6 here. In whose favour is this? One the one hand White cannot break open the kingside with g4-g5 , which is quite an important difference. On the flip side, he has some extra possibilities involving the g5square, for instance after 1 9 . . . l'!d8 20.Wlg3 Wlf5 2 1 .i.d3, when tD g5 is a threat. Although it would be interesting to analyse this more deeply, I actually do not think there is any need for us to get into this complicated position . Castling seems like the most attractive idea, especially since it represents a clear and substantial improvement over the position with . . . h6 and i.h4 included. The reason is explained in the next note.

1 7.tDxc5! as the queen is defended. A nice detail is 1 7 . . . Wlxc5 1 8 .l'!xf7! when Black is toast. 16 . . . i.xd4? 1 7.l'!xd4 0-0 1 8 .i.f6! is crushing, for instance: 1 8 . . . tDxf6 1 9 .tDxf6t gxf6 20.l'!e4 Wlg5 (20 . . . Wlb2 2 1 .l'!f3! mh8 22.l'!g3 f5 23 .Wlh6+-) 2 1 .l'!g4 Wlxg4 22.i.xg4 tDd7 23 .Wlh6 I doubt that Black can survive. 1 6 . . . 0-0 1 7.i.f4 Wld5 1 7 . . . i.xd4 1 8 .l'!xd4 Wla5 1 9 .\Wg3 Black is under pressure. 1 8 .c4 \Wc6 1 8 . . . i.xd4 1 9 .1'!xd4 \Wa5 20.\Wg3t Now White must find (or know) some strong moves to break through. 8 7 6 5 4 3 2

8

a

b

e

d

e

f

g

h

1 9 .i.d6! \Wb6 20.i.xf8 i.xf8 2 1 .i.h5! f6 2 1 . . . g6 22.tDg5! i.h6 (22 . . . gxh5 23.l'!xf7 +-) 23 .\wf2 f6 24.tDgxe6 gxh5 25 .c5 tDxc5 26.tD f5+-

7 6 5 4

8

3

7

2

6

1

5 a

b

e

d

e

f

g

h

16 ... £5 1 6 . . . tD c6? gives Black a near-decisive advantage in the . . . h6 line, but with the bishop still on g5, White has a simple refutation in

4 3 2

a

b

e

d

e

f

g

h

Chapter 1 3 - I ntroduction and 1 2 . . . ltJ d S ! ? 22.cS! Forcing the exchange of a valuable defensive piece. If the queen moves away then the e6pawn drops, so Black has to capture on cS. 22 . . . i.xcs 22 . . . ltJxcS 23.ltJ xf6t gxf6 24J%xf6+23.ltJxcS ltJ xcS 24.ltJb3 ltJ bd7 2 S . ltJ xcS g6 In material terms Black is j ust about hanging on, but his position will be left in tatters. 26J:k l gxhS 27.cJih l +White has a winning position, as Black's king is too exposed. Many years ago when I analysed this variation with an older version of Fritz, it was hard to convince the engine of White's superiority. Nowadays Houdini is quick to grasp the truth of the position.

23 1

21 J�d6 �f7 2 1 . . .i.d7? 22.!!e l ! leaves Black unable to cope with the numerous threats. 8 7 6 5 4 3 2 1 a

b

c

d

e

f

g

h

22.l:� b l !N Improving over 22.i.c4, which forced a draw in 'Auditora' - 'Bevo Xiv' , Internet 2009.

8 7

22 ... b5 22 ... ttJ d7 23.!!xb7 !!hd8 24.i.xa6 reaches an endgame with equal material and ongoing problems for Black.

6 5 4

23.J.h5t!

3

Taking the opportunity to weaken a key diagonal.

2 1 a

b

e

d

e

f

g

h

17.t£Jxc5! ti'xe3t 18.1i.xe3 tLlxc5 19.t£Jxf5 exf5 20 ..hc5 White's fantastic bishops and huge lead in development give him a powerful initiative for the two pawns.

20 ...J.e6 20 . . . ttJ d7 2 1 .E!:fe l 'kt>f7 22 . .ic4t 'kt>g6 23 .E!:d6t ttJ f6 24.E!:e7 f4 (24 . . . E!:fB 2S.E!:clt Eshuis - Wantoch Rekowski, Internet 20 1 3.) 2S .i.f7t 'kt>h6 26.E!:eS i.g4 27.h3 E!:hf8 28.hxg4 !!xf7 29.gSt+- Filev - Roeder, Ohrid 2009 .

23 ... g6 24.J.a tLld7 25.J.d4� White will pick up an exchange, and his active pieces will enable him to pick up one of Black's queenside pawns, leaving him with good winning chances.

C) 10 ... h6 Black can also nudge the bishop a move or two later, but I will group all of the . . . h6 lines under the present move order for the sake of simplicity.

1 1 ..ih4

Poisoned Pawn

232

We have reached an important branching point. 1 2 . . . lD fd7 is the subject of the next two chapters, and the most popular 1 2 . . . gS will be covered in Chapters 1 6- 1 8 . In the following pages we will concentrate on a move which can be considered a sideline in terms of popularity, but which has achieved the best statistical score of all Black's options.

12 ... ttld5!? This is Andriasyan's proposal in Winning with the Najdorf Sicilian. Its forcing nature b

a

e

d

f

e

g

h

Black's move order has the small practical advantage of preventing the bishop from retreating to f4, although in most cases after . . . dxeS and fXeS the bishop should retreat to h4 anyway.

makes it an appealing choice for Black, who needs to worry about far fewer possibilities than in most other variations of the Poisoned Pawn.

1 l ... dxe5 l 1 . . .gS ? has hardly ever been played, and for good reason: 1 2. fXgS dxeS 1 3 .g6! Hammering the nail into Black's coffin. 1 3 . . . .te7

a

b

e

d

e

f

g

h

13.tihd5 exd5 14.e6!?

a

b

e

d

e

f

g

h

1 4 . .tbS t ! .td7 l S .E:b3 �cS 1 6 . .txf6 .txf6 1 7.�f2 �e7 1 8 . lD e4 .th4 1 9 . .txd7t lD xd7 20.gxf7t cj;>d8 2 1 .lDxe6t+- Barnsley - Tritt, corr. 2009. White had some good alternatives along the way, but this game was as good an example as any to show why l 1 . . .gS ? belongs on the scrapheap.

12.fxe5

Most other moves allow Black to complete his development quickly with . . . .te7 (or . . . .tcS) and . . . 0-0 .

14 ....he6 Other moves have hardly ever been tried, but I will briefly mention a few alternatives. 1 4 . . . fXe6? l S . .td3! and White's initiative is too strong. 1 4 . . . �xa2?! has been played in a few correspondence games, almost all of which

Chapter 1 3

-

I ntroduction and 1 2 . . . ltJ d 5 ! ?

have been won by White. Mter l S .exf7t 'kt>xf7 1 6J'!d 1 White has massive compensation for the two paw� s.

233

Cl) 19 ... b5?

1 4 . . . .ie7N l S .exf7t 'kt>xf7 16 . .ixe7 'ffx e7t 17 . .ie2 is mentioned by Andriasyan, and indeed White has excellent chances here too.

a

b

e

d

e

f

g

h

This used to be the top suggestion of some analysis engines, so it is hardly surprising that it has been tested in a lot of correspondence games. However, Black is ignoring his development completely, and the last move can now be considered dead and buried thanks to the brilliant reply:

20.c4! 'fIc5t 20 . . . bxc4 2 1 .l:!f7 'ff c st 22.'kt>h l is the same.

16 ...J.e7 Black's next few moves are virtually forced.

17 ..igGt c,t;dS IS.hen 'fIxe7 19.0-0 Now we need to consider Cl) 19 ... b5? and C2) 19 ... ttld7. Preventing l:!f7 with 1 9 . . . l:!fB? is punished swiftly: 20.l:!xfBt 'ffxfB 2 1 .'ff a St c;;t> c 8 22.'ff b 6 'ff e7 23.c4!+1 9 . . . 'ff c St 20.c;;t> h l lLl d7 has been played a few times, but White can, at the minimum, play 2 1 .l:!f7 transposing to variation C2 below. (2 1 . .if7!? is an alternative try to exploit Black's move order, but there is no reason to spend time on this.)

2 1 .c,t;hl bxc4 22J:U7! ttld7 23J�b7 The rooks are too strong.

23 ... ttle5 24J:hg7 ttlxg6 8 7 6 5 4 3 2 1 a

b

e

d

e

f

g

h

234

Poisoned Pawn

25.ti'el! This has been played successfully in several correspondence games.

25 .. JU'8 26Jlxg6 ti'c6 27J'�gg7 c3 28.h4 ti'd6 29.g;,gl d4 30.ti'e4 White was winning in Dibley - Lelenko, corr. 2008, and several subsequent games.

C2) 19 ... tll d7 8 7

on the queenside, either with . . . �c7-b8 or, more likely, . . . �c7 and . . . �c8-b8. So far White has not demonstrated a clear path to an advantage, even in the latest correspondence games.

22.ti'el! 22.�xb7 �c7 23.�xc7 �xc7 24.�xg7 �f8! is okay for Black, so White must come up with something a bit more subtle. The text move not only hits the e6-pawn, but also prepares a potentially annoying check on h4.

22 ... ti'c6 22 . . . Wd6 places the queen on a better square, but the drawback is seen after: 23.�xb7 �c7 24.�xc7 �xc7 2 5 .�xg7

6 5 4

8

3

7

2

6

1

5 a

b

e

d

e

f

g

h

Of course this developing move is better.

4 3 2

2oJ:H7 20.�xb7 is met by 20 . . . Wc5 t 2 1 .�h l �f8 and Black is doing quite all right.

20 ...ti'c5t 2 1 .g;,hl The first critical divergence. Let's first consider C21) 2 1 ..J:�c8 before moving on to the more common C22) 2 1 ..JU8.

C21) 2 1 . .. �c8 This was Andriasyan's proposal, making it an especially important move for us to consider. It's a logical method of dealing with the �xb7 threat: Black simply plans to play . . . �c7, and his king is not feeling too unsafe in the centre for now. At the same time, Black gives himself the option of bringing his king to a safe haven

a

b

e

d

e

f

g

h

Unlike the 22.�xb7 line, White has no problem with his back rank here, so he can develop threats against Black's king without having to make any real concessions. 25 . . . �b8 (25 . . . �f8 26 . .id3;!;) 26.h3 �c6 27.c4t 22 . . . �f8N is an attempt to simplifY the position and bring the king to safety immediately, but Black loses his extra pawns and White can strive for a long-term edge: 23 .Wxe6 (23.�xg7? Wc3!+) 23 . . . �xf7 24.Wxf7 �c7 (24 . . . We7 2 5 .Wxd5 We5 26.Wf7;!; is a line given by Andriasyan.) 25 . .if5 �d8 26.Wxg7 Wd6 Black is ready for . . . �b8, but he cannot claim full equality, for instance:

Chapter 1 3

-

Introduction and 1 2 . . . ltJ d S ! ?

235

has a chance of analysing it in great depth and making a draw. My recommendation in the main line offers White a more fluid position that is harder to analyse to a definite conclusion. At the same time, the variations are interlinked, and I must mention my concerns in this line as well: 24 . . . i>dB 2 5 J �bf1 l'!c7 26.l'!xg7 �c3 26 . . . i>cB 27.l'!ff7i 27 . .id3

8 7 6 5 4 3 2 b

a

e

d

e

f

g

h

27.l'!e l (27 . .ig4!? i>bB 2B.c4 intending .if3 also deserves attention.) 27 . . . i>bB 2B .h3 tLlc5 29 .�d4 White keeps a positional edge, despite the reduced material.

8 7 6 5 4 3 2

a

a

b

e

d

e

f

g

h

23.�h4t Q;c7

Now White has various permutations of queen checks, and it is hard to find the correct plan to either prevent the enemy king from running to bB, or create pressure elsewhere.

24.�f4t!

This is not mentioned by Georgiev and Kolev.

The more popular continuation is: 24.�g3t White gets some pressure here, but the ensuing lines are a bit too forcing for my taste, the main problem being that Black

b

e

d

e

f

g

h

27 . . . l'!c6! 2B.h4 2B.l'!ff7 �a l t 29 . .if1 i>cB= 2B ... i>cB 29.l'!ff7 �e5 ! ?N 29 . . . l'!dB 30.l'!h7 l'!b6 3 1 .i>h2 �a l 32.�g7 �xg7 33.l'!fXg7 i>c7 34Jhh6 i>d6 seems dangerous for Black to me, even though the computer is less worried. 30.�xe5 tLl xe5 3 1 .l'!xb7 l'!gB This endgame requires further analysis to reach a clear conclusion. A few possible lines are: 8 7 6 5 4 3 2 a

b

e

d

e

f

g

h

Poisoned Pawn

236

32.h5 a5 33.g4 33.E:bc7t E:xc7 34.E:xgBt @d7 3 5 .g4 4J f7 is safe enough for Black. 33 . . . E:xg7 34.E:xg7 E:c7 3 5 .E:gBt Or 3 5 .E:xc7t @xc7 36.g5 hxg5 37.h6 4J f7 3B . h7 @d6= and the king goes to g7. 35 . . . @b7 36.g5 Maybe 36.@g2!? is the best chance to achieve something. 36 . . . hxg5 37.h6 4J xd3 3 B . cxd3 @c6 39.@g2 @c5 Black has enough counterplay to hold the draw.

26.Wb4 @c7 is the only line given by Andriasyan. Black can also play 26 . . . E:c7 intending . . . @cB-bB with a safe position. 26.E:b6 has been played in one correspondence game, bur after 26 . . . Wc3 White found nothing better than 27.E:xd7t forcing a draw.

26 ... c7 27.Yfg3t d8 28J�bfl 2B.W£2!? E:c7 29 .E:b6 4Jxb6 30.E:fSt @e7 3 1 .E:xgB md6 32.E:dBt 4J d7 33.h3 could be an interesting position to explore further, but the computer insists on an evaluation of 0.00, and it may well be right.

24 ... @d8 25.Yfd4! This is the main idea behind the check on f4. White is aiming to lure the enemy rook to gB . It is hard to see why this should be in White's favour, but you will see the idea in the next few moves.

25 .. J�g8 25 . . . @c7?? 26.E:b6! wins instantly. 2 5 . . . b5 opens up the queenside, and after 26.E:xg7 E:fS (26 . . . Wc3 27.Wh4t @c7 2B .E:e7!±) 27.Wh4t @c7 2B .�d3;!:: the black king will be under pressure for a long time.

28 .. J�c7 2B . . . Wc3 ?? 29 .Wd6 wins immediately. 2B . . . Wc5 seems awkward. After 29.c4! E:fB 30.E:xfSt 4JxfB 3 1 .cxd5 exd5 32.�f5 E:c7 33.h4� White's initiative continues, and the black king is unlikely to feel safe for a long time.

8 7 6 5 4 3

a

b

e

d

e

f

g

h

2

29.�h7!

1

Now the reason for provoking . . . E:gB becomes clear.

a

b

e

d

e

f

g

h

26.Yfh4t!N Having provoked . . . E:gB , White switches back to checking the black king.

29 .. J�e8 30.Yfxg7 Most of the following lines are similar, in that White will capture both of the kingside

Chapter 1 3

-

I ntroduction and 1 2 . . . ltJ d 5 ! ?

pawns, while Black takes his king to safety. In the ensuing positions White aims to restrain his opponent's counterplay, while gradually pushing his kingside pawns. Piece exchanges will tend to favour White, whose kingside pawns will become more dangerous, while Black's pawns are not really intimidating. Finally, although Black's king might be safer on b8 in the short term, once the endgame arises, it will be too far away to help deal with the g- and h-pawns.

237

pieces. 32 . . . tOe5 33.E!:xc7t (33.E!:fB E!:xfB 34.E!:xf8t �d7 3 5 .E!:n reaches a similar type of position.) 33 . . . Wfxc7 34.h4 (34.E!:fB!? E!:xfB 3 5 .WfxfBt Wfd8 36.Wfxd8t �xd8 37.ie2 is another interesting endgame where the strong kingside pawns make White the favourite.) There is no easy way for Black to equalize, for instance: 8 7 6

30 �c8 .••

30 . . . h5 only postpones the inevitable: 3 1 .ig6 cj;>c8 (3 1 . . . h4? 32 .E!:e7! is a pretty win.) 32.ixh5± 30 . . .Wfd6 3 1 .Wfxh6 tOe5 32.Wfg5 t �c8 33.E!:xc7t Wfxc7 34.id3 Wfc3 3 5 .Wff4 also favours White.

5 4 3 2

a

b

e

d

e

f

g

h

34 . . . tOxd3 (34 . . . tO g4 3 5 .Wff4±) 3 5 . cxd3 Wfg3 (35 . . . �b8 36.h5±) 36.E!:f3 Wfe5 37.Wff6±

32J�f8 32 . . .E!:xc7t would lead to a similar position, but the queen will be better placed on c7, so I prefer to exchange the rooks a different way.

32 ... Eke7 The only move.

33.�xe8t gxe8 34..td3 a

b

e

d

e

f

g

h

31 .Wfxh6 ltJe5 There does not seem to be much sense in delaying this move, as the following lines demonstrate: 3 1 . . .cj;>b8 32 .id3 Wfd6 33 .h4! ? White's play is straightforward. 33 . . . Wfe5 (33 . . . tOe5 34.E!:fB±; 33 ... E!:g8 34.E!:g7t) 34.Wfg7± 3 l . . .Wfd6 32 .id3 White just regroups his

White's play seems unassuming, but we are simply following the broad plan outlined in the comments to White's 30th move. Black has many options of similar value, but no forcing way to solve his problems.

34 ...Wfc3 34 . . . Wfc7 advantage.

3 5 .h4!?

maintains

White's

Exchanging pieces with 34 . . . tOxd3 3 5 . cxd3 is no problem for White, particularly since it is

238

Poisoned Pawn

e22) 21 ..JU'8

hard for Black to create a passed pawn. Black's exposed king may also enable White to gain a couple of tempos by checking at the right time. 35 . . . 'it>b8 36.h4!;!; Simply pushing the pawns. (36Jl:f8 l:!xfB 37.�xf8t 'it>a7 38.h4 also leads to an unpleasant endgame for Black, but I don't see any reason to exchange the rooks j ust yet, as the pure queen endgame gives Black more checking possibilities.) 8 7 6

b

a

5

e

d

f

e

g

h

3

This is the more direct way to deal with the l:!xb7 threat. Black aims to solve all his problems by generating quick counterplay.

2

22.l:!xg7 l:U6

1

Black almost always plays this, probably because 22 . . . 'it>c7 23 . .if7! causes immediate problems. A recent example continued: 23 . . . �e7 24.�c3t 'it>b8 25 .�g3t 'it>c8?! Ibarra Chami - Hevia Alejano, Villahermosa 20 1 4 . Here White could have made excellent use of his advanced bishop with:

4

a

b

e

d

e

f

g

h

35.�f4! White does not mind an exchange on d3, as his pieces are ready to launch a quick along the c-file.

35 .. J�g8

8

3 5 . . . lD xd3 36.cxd3 j ust increases White's initiative, while 35 . . . l:!h8 36.h3;!; does not really change the position.

7

36.h3 �c7

4

After 36 . . . 'it>b8 37.g4! 'it>a7 38 .�f6! �c7 39.'it>g2± White's play continues smoothly.

3

6 5

2

37.g4;!; White continues with his plan, and Black has no clear way to ease the pressure.

a

b

e

d

e

f

g

h

26 . .ig8!N �f6 27.h3 Black is badly tied up and it will be hard to keep his position together.

23 ..id3! Georgiev and Kolev do not consider this.

Chapter 1 3 - Introduction and 1 2 . . . llJ d 5 ! ? 23 .l'!gBt l'!fB 24.l'!g7 was played in many games, and at some point the variation was even written ' off as a draw, since exchanging rooks on fB promised White little. But of course Black's life should not be so easy with his king stuck in the centre.

239

27.l'!xb7 l'!fl t 2B . .txfl Wc3 29 .Wxc7t Wxc7 30.l'!xc7
24.c4 d4 Black has to keep the centre blocked.

8

25.'1Wel !

7

A fine regrouping move. Another important idea is to prevent Black from exchanging the excellent rook on g7 with . . . l'!afB-f7. 2 5 .We2 is similar, but the queen j ust looks more elegant on e l .

6 5 4 3

25

••

.l:�af8

2 5 . . . b6 26.We4 l'!afB transposes to variation C222 below.

2 1 a

b

e

d

e

f

g

h

The text move prepares c2-c4, and has injected a hew lease of life into this variation. When I analysed it several years ago, I felt it was an unpleasant position to play for Black, since he seems to be constantly passive and barely hanging on to the draw. However, since then a huge explosion of interest in correspondence games has resulted in Black finding forced draws in many variations. The basic premise for White remains the same as in some earlier lines: he tries to create practical problems for Black by aiming at the weak light squares and, most importantly, restricting his counterplay. In most of the following lines the computer indicates that Black is okay, but there is a huge difference between a computer (or correspondence player) demonstrating equality and a human player doing it over the board. Our main aim will be to avoid any forced draws, and preserve a nagging initiative.

23 ... �c7 Mter 23 . . . WfB Andriasyan gives a nice line: 24.l'!h7 l'!cB 25 .Wa5t l'!c7 26.l'!hB WxhB

26.�e4 Here it is necessary to consider the surprising C221) 26 .. J�6f7!?N followed by the normal

C222) 26 ... b6. C221) 26 .. J�6f7!?N 8 7 6 5 4 3 2 1 a

b

e

d

e

f

g

h

This untested move was my biggest worry initially, since the position after the other move j ust looked like a free initiative for White - but first things first. Here Black tries to exchange some pieces and challenge White's plan with

Poisoned Pawn

240

a more provocative but active scheme for counterplay.

27.Wfxb7t c;t>d8 28.gxf7

White can hardly avoid the rook exchange: 28 .gg3?! Wb6! favours Black, while after 28.gg4 ttJ e 5 ! White has nothing more than a repetition .

Another possible line is: 30 . . . ttJ c5 3 l .Wa5 t! 3 1 .Wa8t is possible, but the queen is better on a5 . It keeps the c5-knight in a pin and prepares ge l , and the queen may come back to d2 or e l in some lines. 3 1 . . .�e7 32.ge l Wd6 33 .ig6 gf2 Both 33 . . . gf8 34.a4 and 33 . . . gf6 34.ih5 favour White.

28 gxf7 29.Wfxa6 •••

Unlike the lines you will see after 26 . . . b6, it is White who is materially ahead here! At the same time, Black's king is in less danger following the rook exchange, and he has excellent posts for his pieces. However, if White can consolidate his position then Black will fall short of full equality.

8 7 6 5 4 3 2

29 Wfe5 •••

Intending . . . ttJ c 5 .

a

After 29 . . . � e 7 30.Wb7! gf2 3 1 .a4;!; White's queen is excellently placed for both active and defensive purposes. 8 7 6 5

e

d

e

f

g

h

34.Wa8 ! Preventing . . . Wc6 and threatenirig a check on e8. 34.ih5 Wc6 forces White to go passive. 34 . . . gfB 3 5 .Wa7t ttJ d7 36.gd l gf4 36 . . . e5 37.ih5;!; 37.Wa8 gfB 38 .WbTt: White's coordination is improving and he will soon be ready to advance the a-pawn.

3 1 .Wfb7

4

Pinning the knight and protecting g2.

3

3 1 . gm 32.gdl ! gb8

2

••

1 a

b

e

d

e

f

g

h

30.h3! This is stronger than 30.a4 ttJ c5 ! 3 1 .Wa5t �e7� when Black's active pieces and strong passed pawns give him good chances.

30 c;t>e7 •••

b

32 . . . Wb8 meets with an instructive reply: 33 .We4! Wf4 34.We l !;!; White should avoid the queen exchange for the moment, as Black's central pawns would be too strong in the endgame.

33.Wfa7 Wfc5 33 . . . h5?! 34.ic2! is excellent for White.

Chapter 1 3

-

I ntroduction and 1 2 . . . ltJ d S ! ?

34.YNa4 YNe5

24 1

C222) 26 ... b6

Or 34 . . . hS 3 S .Wlc2;!; successfully r�grouped.

and

White

has 8 7 6 5 4 3 2 1 a

a

b

e

d

e

f

g

h

35.ig6! 3S .Wlc2 lLIcS 36.a4 �b3!� is far from clear. The text move deals with the . . . lLIcS threat, while also preparing to transfer the bishop to a more useful square such as B. At the same time, the rook becomes useful.

b

e

d

e

f

g

h

27.YNh7!

This important move prevents . . . �f7 and forces Black to adopt a more passive set-up.

27 J�d8 .•

27 . . . Wld6? is refuted by 28.cS!. As you will see, opening files with this type of pawn sacrifice is a thematic idea. 28 . . . bxcS 29.�xd7t Wlxd7 30.�b7t+-

35 lLIc5 •••

Another illustrative line is 3 S . . . �g8 36.Wlb4t WlcS (36 . . . lLIcS 37.ie4!;!;) 37.Wlb l ! Wlxc4 38 .ihS;!; when Black's light squares are glaringly weak.

28.h3 It is hard to imagine not playing this move. 28.a4 �d6 29.h3 transposes to 2 8 . h3 �d6 29 .a4, and while there are some other options, none of them really work out.

36.YNa5 36.Wla3 �f6 37.ic2 is another way to maintain a slight plus.

36 cbf6 •••

36 . . . �g8 is met by 37.ihS;!; intending iB . Again, White seems to be quite well coordinated.

37J�1el YNd6 38.ih5;t White intends Wld2, and he may even start playing against the black king.

8 7 6 5 4 3 2 1 a

b

e

d

e

f

g

h

242

Poisoned Pawn

We have reached the most critical position for the assessment of the 1 2 . . . tLld5 variation. The computer constantly shows "0.00" here, but don't let that deter you, as things are far from simple. Most importantly, as you will see in the coming variations, Black's play mainly revolves around moves such as . . . 'it>d6-c7, waiting for White to do something rather than carrying out his own active ideas. Moving the queen might allow c4-c5 , while all his other pieces are stuck for now. Meanwhile, White faces a struggle to open up lines on the queenside. He would be happy to sacrifice at least the c-pawn, and sometimes even the a-pawn before bringing the queen back from h7 to utilize all those open files on the queenside. One major problem is the relative weakness of the white king. If White tries to enter with his second rook through the kingside, the weakness of his first rank will give Black ample opportunity for counterplay. In short, Black's pieces are well placed but it is hard for him to improve his position, apart from a few ideas like . . . 'it>d6 and . . . tLl fS . The challenge for White is to find a way to break through. The c4-c5 break is a constant theme, but White needs to make sure the rest of his pieces are sufficiently well placed for it to work.

28 �d6 •..

Sometimes Black has preferred: 2S . . . Wfe5 In all five of the examples on my database, White has responded with 29.a4, inviting 29 . . . 'it>d6 with a position normally reached after 2S . . . 'it>d6. However, since I am not recommending this path for White in the main line, we need to find an alternative. Fortunately there is another idea, which should appear obvious in light of my earlier comments. 29.c5! Wfxc5 29 . . . bxc5? 30 . .ixa6 wins immediately.

b

a

e

d

e

f

g

h

30.Wfe4! White has the simple plan of bringing the queen back to e l , which will leave Black worrying about several ideas, including :ge l , .ie4 and .ixa6. 30.a4 is met by 30 . . . Wfd5! when White's queen is temporarily cut off on h7. 30 . . . Wfd5 30 . . . a5 3 1 ..ib 5 ± 3 1 .Wfe l Wfd6 3 1 . . .a5 32.Wfh4t In the event of 3 1 . . .e5, both 32 ..ie4 and 32 . .ixa6 100k dangerous for Black. 32 . .ixa6 :gfS 33 . .ib5 :gd5 34.Wfe l t c;t>b7 3 5 .Wfb2� With ongoing pressure on the black position. 8 7 6 5 4 3 2 1 a

b

e

d

e

f

g

h

29 ..ic2!N An important improvement, the point of which will be discussed shortly. Firstly, I will

Chapter 1 3 - Introduction and 1 2 . . . ttJ d S ! ? briefly show why I was not satisfied with the alternatives. Moving the queen back with 29.�e4 almost always runs into 29 .. J'!df8!, so this manoeuvre is of no use for the time being. 29.a4 has been played in almost every game on my database from this position. Preparing the a4-a5 break is an obvious idea, and White was successful with it in several games. Eventually, however, the correspondence guys found an incredible and forcing way to hold the position together: 29 . . . �e5! 30.c5t (In the event of 30.a5 bxa5! White has more than one drawing line, but I was unable to find a way to increase the pressure.) 30 . . . ttJxc5!

243

will have to make a hard choice on every turn, while constantly having to worry about tactical motifs. For White, the main aim is to keep the initiative going without allowing a forced draw. Obviously he must also take care to keep Black's central pawns in check. Although my analysis is not completely conclusive, I believe I have fulfilled those objectives.

29 �f8 ...

This i s the reply that you are most likely to face should you ever encounter this variation in a game, as Black avoids the main threat without weakening his position or committing himself in a serious way. The alternatives all involve clear drawbacks or risks of some kind. 29 . . . 'tt> c7? ? 30 . .ia4 is killing, and 29 . . . �e5 ? 30 . .ia4 is also bad for Black.

8 7

29 . . . �xc4 should not be too worrying, as White wanted to sacrifice this pawn anyway. 30 . .id3 �d5 (30 . . . �c5 3 1 ..ixa6;!; ; 30 . . . �c8 3 1 .a4� The open files clearly favour White.) 3 1 . .ixa6 White stands better. An important point is that 3 1 . . . �f5 ? is refuted by:

6 5 4 3 2

a

b

e

d

e

f

g

h

3 1 .:B:xb6t (Mter 3 1 .:B:a7 'tt> d 5! 32.a5 :B:f2 33.axb6 �g3 34 ..ifl d3 Black is doing well, despite the king on d5!) 3 1 . . .'tt> d 5 32 . .ic4t 'tt> xc4 33 .�c2t 'tt> d 5 34.�a2t 'tt> e4 3 5 .:B:b l :B:fl t 36.:B:xfl �xg7 37.:B:e l t 'tt> f5 38 .�c2t 'tt> f6 Black has held this position in several correspondence/internet games. Let's return to my new suggestion of29 . .ic2!N. The main point of this move is that White now threatens .ia4, so Black cannot j ust sit around and wait. I do not claim this to be a refutation of Black's set-up, and there will be a lot of possibilities for Black at every step - and that is actually part of the problem for him, since he

8 7 6 5 4 3 2

a

b

e

d

e

f

g

h

32.:B:xd7t! :B:xd7 33.:B:xb6t 'tt> c 5 34.:B:b5 t and wins. 29 . . . �c6 30.a4!? a5 (30 . . . �c5 3 1 .a5! �xa5 32.�e4 �c5 33 . .ia4t) Black's last move has prevented an immediate breakthrough, but

Poisoned Pawn

244

left a hole on the b5-square, meaning that

the . . . 'WcS defence will probably never work. White has a few tempting options, for instance:

a

b

e

d

e

f

g

h

3 1 . 'W d3 (3 1 .i.e4? unfortunately runs into 3 1 . . . tLl f8 ! , but 3 1 .i.d3! ? is possible.) 3 1 . . .1'% f4 32.'Wg3 �df8 33 .i.d3� 29 . . . b S ! ? Though it may look cooperative t o open the queenside, it is not easy to prove anything here for White.

a lot of fine details and nuances. Interested

readers may conduct their own research, but I will focus on the more straightforward option. 30 ... axbS 3 1 .i.d3 b4 32.�d l ! ? The idea is to prevent . . . tLleS in a creative way. . . and then the game just continues. 32.i.c2 is a sensible alternative, but I was not quite able to make it work the way I wanted. The key line continues: 32 . . . 'WbS! Rather artificial, but it prevents i.a4 while threatening a check on fl , so it makes sense. 33.'it>gl 'We2! Preventing the white queen from returning to the centre. 34.�d l 'We3t 3 S .'it>h l tLlf8 36.'Wh8 'WeS The position remains quite messy, but White's pieces are not so flexibly placed, so Black should be okay. 32 . . . tLl b6 32 . . . tLleS?! 33.i.bS! is extremely unpleasant. Now there are many possibilities, but we are already a long way into a rather weird line, so I will j ust give a few more logical moves to illustrate how the play might go. 8 7 6 5 4 3 2

a

b

e

d

e

f

g

h

30.cxbS This is certainly the most natural move, and the easiest to handle. 30 .'We4?! would have been a fun move, but 30 . . . �df8! ruins it by threatening a powerful check on fl . 30.�d l ! ? is a fascinating alternative. I analysed this in considerable depth and found it a worthy alternative, but there are

a

b

e

d

e

f

g

h

33 .i.c2!? �df8 Or 33 . . . �d7 34.�xd7t tLlxd7 3 S .i.a4 tLlb6 36.'Wh8 and the pressure continues. 34.'Wd3 'Wc4 3 S .'Wg3t eS 36.i.d3 White maintains excellent compensation for the pawn, and Black must watch out for !'1b7 ideas.

Chapter 1 3 - Introduction and 1 2 . . . ltJ d 5 ! ?

8

8

7

7

245

6

6

5

5

4

4

3

3

2

2 b

a

1 a

b

e

d

e

f

g

h

3o.VNd3

30.'lWg8 E:d7!=

30 J�M7 31 .VNg3t ••

I also analysed 3 1 .E:g4!? and found it interesting, but I prefer the text move.

31. e5 32.E:g8 ••

The more pieces that remain on the board, the greater our chances of an initiative. Now Black has a wide choice, but White also has a lot of ways to improve his position: a prophylactic mh2; attacking with a4-a5 ; and various ways to poke Black's position on the light squares using the bishop.

32 E:df7 .••

32 . . . mc7 should be met by 33 . .ie4, when the bishop will be ideally placed on d5. 33 . . . E:df7 34.mh2 E:f4 35 . .id5;!; 32 . . . b5 33 . .id3! b4 Black has blocked the queenside for the moment, but White can prepare a2-a3 . 33 . . . bxc4 34 . .ixc4! is excellent for White, since 34 . . . 'lWxc4? loses to 3 5 .E:b6t . 34.�h2 Again Black has a lot of possible moves, but in general he seems to be suffering due to a lack of counterplay. A nice illustrative line is: 34 . . . E:df7

e

d

e

f

g

h

3 5 .a3! bxa3 36 . .ie4! Threatening 'lWb3. 36 ... a2 37.E:a l 'lWxc4 38 .'lWa3t md7 39.E:xa2 'lWfl White has a few good options here, but j ust one line to show an advantage is: 40.'lWa4t me7 4 1 ..if3! E:xf3 42.'lWb4t md7 43 .'lWb7t md8 44.'lWb6t md7 4 5 . gxf3 ± 8 7 6 5 4 3 2 1 a

b

e

d

e

f

g

h

33.cJih2! As usual for this line, there is no need to hurry! It is too soon for 33 . .ie4?!, as Black gets counterplay with: 33 . . . b 5 ! (33 . . . E:fl t 34.E:xfl E:xfl t 3 5 .mh2 favours White, who threatens 'lWg7.) 34 . .id5 (34.cxb5 E:fl t 3 5 .E:xfl E:xfl t 36.mh2 'lWc l 00 ) 34 . . . E:fl t 3 5 .mh2 d3! This works because the bishop on d5 prevents White from taking on d3 with check. 36.E:xfl

Poisoned Pawn

246

:gxfl 37.'lWxd3 'lWg l t 38.�g3 'lWflt 39.�g4 'lWd4t White will have to fight for a draw after the queens come off.

repetition. White maintains some pressure, although a long fight lies ahead.

34.a4 33 .. �f4! J

The rook seems slightly awkward here, but it is important to prevent i.e4. 33 . . . tiJ d7 and 33 . . . tiJ e6 are both met by 34.:ga8 with pressure against Black's queenside. 33 . . . b5 34.'lWb3 ! ? 34.i.e4 bxc4 3 5 .:gb8 might give some compensation, but after 34 . . . d3 Black has got more counterplay than I want to allow. Some of the following lines might seem difficult, but the main point is that the positions are much harder to play for Black. With White you do not need to find many 'only moves' - there are multiple ways to maintain a tense position, while Black does not have that luxury. 8

Black has no threat, so White can improve his position and force Black to consider a4-a5 on every turn.

34 .. rJlc7 .

34 . . . tiJ d7 as usual runs into: 3 5 .:ga8 ! rJJ e7 (35 . . . �c7 is met by 36.'lWg6! intending i.e4. 36 . . . :g4f6 37.'lWe4±) 36.:gxa6 'lWxc4 37.i.d3 'lWc7 38 .a5 bxa5 39.:gxh6 With an ongoing initiative. We have reached another critical posltlon. White has played just about all the useful moves he has available, so it is time to make something happen. 8 7 6

7

5

6

4

5 4

3

3

2

2

1

1

a a

b

e

d

e

f

g

h

34 . . . 'lWxc4 34 . . . bxc4 3 5 .'lWb8t �d5 36.:gg4! :gf4 37.:gxf4 :gxf4 Now there follows a nice zigzag manoeuvre by the queen: 38 .'lWd8t 'lWd6 39 .'lWa5t 'lWc5 40 .'lWxa6 d3 4 1 .:gb5 dxc2 42.:gxc5 t �xc5 43 .'lWa5 t± The queen will pick up the c2-pawn. 3 5 .'lWa3t 'lWc5 36.'lWxa6t 'lWc6 37.'lWa3t 'lWc5 38 .:gb4!� An unusual but effective way to avoid the

b

e

d

e

f

g

h

35 ..ig6! Normally I would hate to exchange my bishop for such a knight, but here the exchange would grant an entrance for the white queen - an unpleasant scenario for Black, especially when combined with the rook on g8 . If Black does not take the bishop, then it will go on to annoy him in other ways.

35 .. �d7 J

35 . . . :g7f6 36.i.h5t leaves Black facing

Chapter 1 3 - I ntroduction and 1 2 . . . lD d 5 ! ? the direct threat of Wi'g7, as well as the more patient .ig4 o r .iB. , Here is some proof of why taking the bishop does not help Black: 3S . . . ttJxg6 36.Wi'xg6 :B:4f6 (36 . . . :B:7f6 37.Wi'eB+-) 37.Wi'g4! :B:ffi 8 7 6 5

247

thought that 37.:B:hB might be a good try, bur 37 . . . e4! 3B . .ixe4 Wi'eS solves all Black's problems.

36 .. J�e7 37.i.h5 White can continue improving his pieces with .if3 , .ig4, or even :B:hBxh6. Black's position has not yet been breached, but he remains under pressure, with no clear way to simplify or obtain counterplay.

Conclusion

4 3 2 b

a

e

d

e

f

g

h

3B.aS! bxaS (3B . . . :B:xgB 39 .Wi'xgB Wi'ffi 40.Wi'dS±) 39 .:B:g7t :B:Bf7 40.:B:xf7t :B:xf7 4 1 .Wi'g6! Amazingly, Black's position is collapsing. 4 1 . . .Wi'xc4 (4 1 . . .:B:d7 42.Wi'xa6+-) 42.Wi'xh6+- Intending :B:c 1 !

a

b

e

d

e

f

g

h

36.i.e8! A cool idea! The rook is excellently placed to help launch Black's counterplay with . . . d3 , so the bishop takes a short detour to displace it. In the event of 36 . .id3 :B:df7 it is hard to improve White's position further. I briefly

This is only the first of many Poisoned Pawn chapters, and already we have encountered some insanely complicated positions. At least it did not take long to establish that 1 O . . . ttJ fd7 does not have much of a future, while after 1 0 . . . dxeS 1 1 . fxeS, Black has little to gain by avoiding . . . h6. The real fun begins after 1 0 . . . h6 1 1 ..ih4 dxeS 1 2 . fxe S . The bulk of this chapter was spent analysing 1 2 . . . ttJ d S ! ? , a favourite move among correspondence players as well as the recommendation in Andriasyan's book. After the relatively forcing sequence 1 3 .ttJxdS exdS 1 4.e6!? .ixe6 l S .ttJxe6 fxe6 16 . .id3 .ie7 1 7 . .ig6t mdB 1 B . .ixe7t Wi'xe7 1 9. 0-0, we arrive at a position where Black has tried solving the problem of his uncastled king in various ways. The biggest challenge came in the final variation C222, where I had to find a novelty on move 29 (!) to pose new problems to the defence. Most readers will never have to prepare their openings in this much depth, and if you play through some of the lines you should get enough of a feel for the available plans and piece manoeuvres to cause your opponents plenty of problems.

Poisoned Pawn a

b

e

d

e

f

g

Variation Index l .e4 c5 2.� f3 d6 3.d4 cxd4 4.�xd4 � f6 5.� c3 a6 6.J.g5 e6 7.f4 �b6 8.�d2 �xb2 9J�b l �a3 1 0.e5 h6 1 l .J.h4 dxe5 12.fxe5 � fd7 13.� e4 �xa2 14J�dl 249 252 252 254 255 255 257 261 261 262 265

A) 1 4...�b2!? B) 14 ...�d5 15.�e3 B l ) 1 5 ...J.c5 B2) 1 5 ...�xe5 1 6.J.e2 J.c5 17.J.g3 B2 1 ) 17...�d5 B2 1 1 ) 2 1 ..J�d8 B2 1 2) 2 1 ...f5! B22) 17 ...hd4 1 8J�xd4 �a5t 19J�d2 0-0 20.J.d6 B22 1 ) 20..J�e8? B222) 20...� c6?! B223) 20...f5

8223) after 28 . . . liJ fd7

82 1 2) after 23 . . . liJ bd7

A) after 25 . . . b 5 8

8

8

7

7

7

6

6

6

5

5

5

4

4

4

3

3

3

2

2

2

a

b

c

d

e

f

26.�xd8 !?N

g

h

a

b

c

d

e

24.�f3!N

f

g

h

a

b

c

d

e

f

29.l'� d 1 !N

g

h

h

Chapter 1 4 - 1 2 . . . ttJ fd7

l.e4 c5 2.'!l�f3 d6 3.d4 cxd4 4.c!ihd4 �f6 5.�c3 a6 6 .ig5 e6 7.f4 %Vb6 8.%Vdl %Vxb2 9.gbl %Va3 lO.e5 h6 1 l ..ih4 dxe5 12.fxe5 �fd7 13.�e4 •

8 7 6

249

worth taking seriously, but the most popular move by far is B) 14 %Vd5. •••

1 4 . . . lLl c6? is not a serious option: 1 5 .lLlxc6 ( 1 5 . .ie2 .ib4! ends all the fun.) 1 5 . . . bxc6 1 6 . .id3! ( 1 6 . .ie2 �a4 1 7.�f4 �a5t 1 8 .c3 g5 was a mess in Nataf - Perunovic, Herceg Novi 200 5 . ) 8

5

7

4

6

3

5

2

4 3 a

b

e

d

e

f

g

h

This can be considered the old main line. It has been tested in a few thousand games over the years, sometimes through different move orders such as 1 0 . . . dxe5 1 1 .fxe5 lLl fd7 1 2. lLl e4 h6 ( 1 2 . . . �xa2 was covered in variation B of the previous chapter) 13 . .ih4.

2 b

a

e

d

e

f

g

h

1 6 . . . �d5 1 7. 0-0 �xe5 1 8 .fj{ h l White had an obvious initiative, which soon turned into a winning attack after 1 8 . . . .ie7 1 9 . .ig3 �d4 20.c3 �b6 2 1 . lLl d6t .ixd6 22 . .ixd6 in Oesterman - Weber, corr. 2008.

13 %Vxa2 •.•

A) 14 %Vb2!?

It is not worth spending much time on other moves. Here are a few brief examples:

•••

13 . . . lLlxe5? allows 1 4. lLl b 5 lLl d3t 1 5 . .ixd3 axb5 1 6 . .ixb5t lLl d7 1 7.0-0 with a huge initiative for White. 1 3 . . .�a4 14 . .ie2 lLl c6? 1 5 .lLlxe6 g5? 1 6. lLl f6t 1-0 was a fiasco for Black in Radjabov Anand, Rishon Le Zion (blitz) 2006. 13 . . . g5 14 . .ig3 �xa2 has been played a few times, but after 1 5 .!!d 1 it is hard to believe that Black will benefit from having weakened his kingside.

14.gdl In this position the rare

A) 14 ... %Vb2!? is

a

b

e

d

e

f

g

h

This was first played by Karjakin against Radjabov in 2006, but White won the game and the move has never really caught on. Apart

250

Poisoned Pawn

from threatening . . . .ib4, it is hard to see much purpose for the queen on b2, although some of the ensuing lines are still not easy.

15.Yfe3 i.c5 1 5 . . . lD c6 1 6.lDxc6 bxc6 1 7 . .ie2 'iWxe5 1 8 . 0-0 gives White a typical initiative thanks to his huge lead in development.

1 8 . . . lDxd4 1 9J'hd4! .ixd4 20.'iWxd4 0-0 2 1 .lD f6t is similar to the main line below. 2 1 . . .gxf6 (2 1 . . .lDxf6 22 . .ixf6 transposes exactly to Radjabov - Karjakin.) Now White has a forced win: 8 7 6

16 ..ie2 1 6.c3!? is an interesting alternative. This move is less effective when the moves . . . h6 and .ih4 have been included (compare 1 3 . . . 'iWb2 in the notes to variation B of the previous chapter, on page 228) , but it could still be explored further in this variation as well. However, since the main line seems to be in excellent health, there is no need for White to deviate.

16 ...Yfa3! 1 6 . . . lD c6? was played in Radjabov Karj akin, but this is the wrong move order. White missed the chance to play 1 7.lDxe6! , instantly getting a winning advantage, as 1 7 . . . .ixe3 ? ? 1 8 . lD d6 is mate.

5 4 3 2 b

a

e

d

e

f

h

g

22.'iWg4t �h8 (22 . . . �h7 23 . .id3t+-) 23.'iWf4! 'iWc5t 24.�h l �h7 (24 . . .'iWxe5 25 .'iWxh6t �g8 26.�f3+-) 25 . .ixf6 lDxf6 26 . .id3t! �g8 27.'iWg3t! �h8 28 .�xf6 Mate is unavoidable.

8 7 6 5 4 3

a

2

19.t2H6t! �xf6

1 a

b

17.c3 ttlc6

e

d

e

f

g

h

We have now transposed back to the aforementioned game.

18.0-0 0-0

b

e

d

e

f

g

h

1 9 . . . �h8 20.'iWe4 lDxf6 (20 . . . g6? 2 1 .'iWf4+-) 2 1 .exf6 g6 22 . .if2 White has good attacking chances.

20 ..ixf6 �xd4 Also after 20 . . . lD e7 2 1 .'iWg3 lDg6 22.h4 Black is in trouble.

Chapter 1 4 - 1 2 . . . ttJ fd7

25 1

2lJhd4 .bd4 22JWxd4 ge8! The stem game quickly turned painful for Black: 22 . . . gxf6? 23.exf6 Wfa5 24.h4! \t>h7 25 .�d3t+- Radjabov - Karjakin, Cap d'Agde 2006. 22 . . . �d7? also fails to offer much resistance: 23 .Wfg4 g6 24.Wfh3 \t>h7 2 5 .�g5 ! +-

23.�g4 �f8! 23 . . . g6?! 24.�d3! \t>h7 (24 . . . Wfc5 t 2 5 . \t> h l 'it>h7 26J:l:f3+-) 25 .h4 Wfxc3 26J:l:f3 Wfc 1 t 27.'it>h2+-

a

b

e

d

e

f

g

h

25.gB!

2 5 .�xd8? Wfxd8 26.Wfe4 g6 leaves White struggling to j ustify the two-pawn deficit.

8 7 6

25 ... b5! 26.hd8!?N

5

Black's last move prepared to meet 26.!!g3? with 26 . . . !!xd3 27.!!xd3 �b7, stopping !!d8 .

4 3 2

a

b

e

d

e

f

g

h

24.J.d3!

I also checked 24.l:'! f3 ! ?N, but after 24 . . . \t>h8! Black can defend: 25 .Wfh5 (25 .�d3 gxf6 26.l:'! xf6 !!d8 27.Wfh5 !!xd3 28 .!!xh6t Wfxh6 29 .Wfxh6t \t>g8=) 25 . . . gxf6 26.!!xf6 Wfg7! 27.�d3 \t>g8 28.!!xh6 !!d8! White only has a choice of draws.

24.. J:�d8! 24 . . . �h8 ?? loses instantly to 25 .Wfe4! Wfg8 26.Wff4 (or 26.Wfh4) . 24 . . . a5? 2 5 .!!f3! leaves Black defenceless. 24 . . . �d7?! 25 .!!f3 �b5 26.�c2 also puts Black in big trouble, with !!g3 coming next.

26.�e4 has been played in all the games so far (all of them internet/correspondence encounters) , but 26 . . . �b7! 27.�xb7 !!d l t 28.\t>f2 !!ad8 29.�xd8 Wfc5 t 30.\t>e2 !!xd8 3 1 .Wff4 Wfc7 gives Black some compensation, as shown in Sadowski - Cardelli, corr. 2009. Still, with a bishop for two pawns, White is pretty much playing for two results here, so you may still consider this if you find the main line too double-edged.

26 ...J.b7 26 . . . Wfxd8 gives White a much-improved version of the note to move 2 5 above. With the rook already on the third rank, he can create serious threats with: 27.!!g3! g5 (27 . . . Wff8 28 .Wfe4! threatens mate while hitting the rook on a8 .) 28 .Wfe4 Wfb6t 29.\t>f1 �b7 30.Wfh7t \t>f8 3 1 .Wfxh6t \t>e8 32 .Wfxg5±

27.J.b6 ha 28.�xf3-+

Material is roughly level, but White has excellent chances to create an attack.

252

Poisoned Pawn 1 7 . . . g5 1 8 .i.g3 1 8 .i.f2! ? is also dangerous. The text move may give Black the extra option of . . . i.c5 at some point, but it defends e5 and leaves the f-file open. Either way, White can develop a massive attack. 1 8 . . . Wfxc2 1 8 . . . i.c5 1 9 .Wff3 and 1 8 . . . i.e7 1 9 .h4 both give White a strong initiative. 1 9. 0-0 i.c5 20.tLlxc5 Wfxc5 2 1 .E:d4! 0-0 22.h4-+ It would be amazing if Black could survive such an attack.

B) 14 .. .\Wd5

b

a

e

d

f

e

g

h

Bl) 1 5 ....ic5

This is the most natural and popular move, centralizing the queen and attacking a few white units.

15.YlYe3 Defending the e4-knight and opening the d-file. Now we will analyse Bl) 1 5 ... .ic5 followed by the main move B2) 1 5 ...YlYxe5. 1 5 . . . tLl c6N Now White must decide whether to win the enemy queen or continue his brewing attack. 1 6.tLlxc6! 1 6. tLl b 5 is the materialistic option, but 1 6 . . . Wfxd l t 1 7.�xd l axb5 will lead to a long, slow fight. 1 6 . . . Wfxc6 1 7.i.e2 White has excellent compensation , and Black will have to weaken his kingside j ust to complete his development. 8

a

b

e

d

e

f

g

h

16.�xe6! Nothing else will do! 1 6.i.e2 ? tLlc6 1 7.c3 0-0 is clearly better for Black, while 1 6.c4? Wfxe5 1 7.i.g3 i.xd4 1 8 .E:xd4 Wfa5 t leads to a considerably worse version of the main line for White.

16 ....ib4t

7

Black has no choice, as 1 6 . . . Wfxe6? 1 7.Wfxc5 tLl c6 1 8 .Wfa3 , 1 6 . . . i.xe3 ? 1 7.tLlc7t �f8 1 8 .E:xd5 and 1 6 . . . fxe6? 1 7.tLlxc5 Wfxc5 1 8 .Wfxc5 tLlxc5 1 9 .E:d8t c;t;f7 20.E:xh8 all leave him close to defeat.

6 5 4 3 2

17.d YlYxe6 IS.cxb4 0-0 a

b

e

d

e

f

g

h

2S3

Chapter 1 4 - 1 2 . . . ttJ fd7 The last couple of moves were forced on both sides. In this position White has tried a few moves, b � t the strongest is:

19J�d6! �xe5

19 ... Wa2 Ieads to a short but intense tactical sequence, which is worth remembering: 20 . .ie2 Wb l t (20 . . . Wa l t 2 1 ..id l ! WxeS 2 2 . 0-0±) 2 1 .�f2 Wxb4 (2 1 . . .Wxh l ? 22J::!:x h 6! +-) 22.Wf4!N This suggestion of Goh Wei Ming seems to be winning. (22.E:hd l tO c6 23.E:xh6 tO cxeS 24.E:d4 Wb l 2 S . .if6! tOxf6 26.tO xf6t gxf6 27.Wg3t Wg6 2 8 .E:xg6t fxg6 29.Wb3tt as has been played in a few correspondence games.) 22 . . . WaS (22 . . . aS 23.E:hd l tO cs 24.E: l d4 tO xe4t 25 .Wxe4 Wb2 26.�g3! Threatening .id3 . 26 . . . .ifS 27.WxfS Wxe2 2 8 .E:e4 Wc2 29.E:xh6 Wc3t 30.�g4! gxh6 3 1 . � h S ! +-) 23.E:xh6 gxh6 24 . .if6 Wb6t 2 S . �g3 tO xf6 26.tO xf6t �g7 27.E:d l E:h8 (27 ... .ifS 28.E:d6) 2 8 . .ic4! .ifS 29.h4!! Goh Wei Ming's analysis ends here, and indeed after 29 . . . tO c6 (or 29 . . ..ig6 30.hS) 30 .E:d7! E:af8 3 1 .�h2! Black is defenceless.

tactical ideas, and understand why the queen move is necessary. 20 .. J::!: e 8? is instantly refuted by 2 1 ..ixf7t! �xf7 22.Wb3t. 20 . . . tO b6!? is not so terrible, but it leads after 2 1 .E:xb6 .ifS (2 1 . . .Wa l t? 22.�f2 Wxh l 23.E:g6!+-) 22.tOf6t gxf6 23 .Wxe5 fxeS 24.0-0 to a position where White has a pleasant endgame initiative. 20 . . . bS 2 1 ..ig3! Wb2N (I wanted to see if Black could improve on 2 1 . . .Wa l t 22.E:d l Wxd l t 23.mxd l bxc4 24.E:fl ! tOc6 2S.tOd6 ttJxb4 26.tOxf7! tOd5 27.tOxh6t with a winning position for White, as seen in some correspondence/internet games.) 22 . .ib3! Black does not seem to have a good move, for instance: 8 7 6 5 4 3 2

a

b

e

d

e

f

g

h

22 . . . tOeS (22 . . . .ib7 23.0-0! .ixe4 24.E:d2! +-) 23.0-0 tO c4 24 . .ixc4 bxc4 2S . .ie l !+- The bishop will be monstrous on c3 .

a

b

e

d

e

f

g

h

2o.Ac4! �h5! The most popular choice, and the strongest. Several other moves have been tried, and it is worth playing through the lines, not to memorize them but rather to grasp the various

A final alternative is: 20 . . . tO c6 2 1 . 0-0! tO b6N 2 1 . . .Wh5 meets with a beautiful refutation: 22 . .if6!! (A year later Shirov missed this move and played 22.Wf4 against Wang Hao, but he won anyway.) 22 . . . tO xf6 23.tOxf6t gxf6 24.E:fxf6 tOeS 2 S .E:xh6 tOxc4 26.We7 1 -0 Sj ugirov - Brodsky, St Petersburg 2008.

Poisoned Pawn

254

24 . . . Wle6 resulted in yet another rout: 2 5 .l'!f6! tL\ xb4 26.l'!h5 1 -0 Sikorsky - Fenwick, corr. 2008.

8 7 6 5

25.tl�f6t c;!;>h8

4 3 2 b

a

e

d

e

f

g

h

22.tL\f6t! This leads to a winning position, but even if this tactic was not available, the simple 22 .i.b3 would leave White with a great deal of activity. 22 . . . gxf6 22 . . . �h8 23 .Wld3+23 .Wlxh6 tL\ xc4 24.l'!d3! Black is defenceless.

a

b

e

d

e

f

g

h

26.c;!;>hl ! �e7 27.l'!h5 �g8 28.�xg8 c;!;>xg8 29.hf7t �lxf7N 29 . . . �h7 30 .Wlg5 was winning for White in Ter Sahakyan - Kotanj ian, Yerevan 20 1 1 .

8 7

30.Wfe8t C;!;>h7 3 1 .Wfxf7±

6

Despite improving slightly on the above game, Black is still in serious trouble.

5 4

B2) 1 5 Wfxe5

3

..•

2 1 a

b

e

d

e

f

g

h

21 ..ig3!

2 1 .Wlf4 had been my original intention, but 2 1 . . .Wle5 ! is still quite complicated.

21 ...�c6 22.0-0 �de5 23.he5! Wfxe5 23 . . . tL\xe5 24.Wlg3 i.g4 2 5 .l'!d5 l'!ae8 26.h3 wins material, and White has won almost all games from this position.

24J:�d5 Wfb2

a

b

e

d

e

Eating the pawn is normal .

f

g

h

255

Chapter 1 4 - 1 2 . . . lLl fd7

16 ..ie2 Developing another piece while incidentally preventing . : . f5 , which would allow an embarrassing mate from h 5 .

16 ....ic5

1 6 . . . j,b4 t has never been played, and White gets a clear advantage with precise play: 1 7.c3 0-0 1 B .cxb4 f5 1 9 .j,g3! ( 1 9 . li'l f6t l:!xf6 20.'Wxe5 li'lxe5 2 1 .j,xf6 gxf6 22.li'l xf5 li'l bc6�) 19 .. .f4 ( 1 9 . . . 'Wxe4 20.'Wxe4 fxe4 2 1 .li'l xe6 leaves White with a powerful initiative.) 20 .'Wb3 l:!eB 2 1 . li'l f5 ! fxg3 22.li'led6±

17 ..ig3

Now we should consider the bizarre B2 1) 17...'Wd5 for the sake of completeness, but the great majority of games have continued with

B22) 17 ....ixd4. B2 1) 17 .. .'�d5 This is one of the weirdest ideas I have ever seen . Black j ust provokes c2-c4, and then plays exactly like in the main line after 1 7 . . . j,xd4. This has been played in over a hundred correspondence games, but it is hard for a normal player to make sense of Black's idea. Finally, however, I was able to find a key line where the c2-c4 move may hamper White's play, forcing us to modify our plans slightly.

18.c4

1 B .j,f3 looks harmless due to: 1 B . . . 0-0! 1 9 .'Wc3 a5! 20.0-0 e5 2 1 .li'l f6t li'l xf6 22.j,xd5 j,xd4t 23.l:!xd4 exd4 24.'Wxd4 li'l xd5 2 5 .'Wxd5 li'lc@

18 ....ixd4

1 B . . . j,b4t has never been played, for good reason: 1 9 .<;!;>f1 'Wa5 20.li'ld6t j,xd6 2 1 .j,xd6 ....

19.1:!xd4 'i;Ya5t 20.l:!d2 0-0 21 ..id6

a

b

e

d

e

f

g

h

So far everything is proceeding in parallel with the main line which will be examined later, except for the position of the pawn on c4 instead of c2. Here B21 1) 2 1 ...Eld8 has been Black's most popular choice, but B212) 2 1 ...£5! is the best attempt to utilize the difference in placement of the white pawn.

B21 1) 2 1 ..JM8 8 7 6 5 4 3 2 1 a

b

e

d

e

f

g

h

22.g4! The same plan that we will see later. Instead 22.'Wg3 'Wf5 23 .j,e5 'Wg6 24.'Wh4 li'l c6 2 5 .j,h5 'Wf5 26.j,xg7 li'l c5 ! 27.li'lxc5 'it>xg7 reaches a messy position.

22 ... c!L)c6

2S6

Poisoned Pawn

22 . . . f5 ? is too dangerous: 24.gxh6 lDeS

23.gS!

fXe4

8 7 6 5 4 3 2

a

b

e

d

e

f

g

h

move with 2B.cS!?�, obtaining excellent compensation. Instinctively I think White should be doing well after 23.g5 !?N, and you may wish to play this way, using the analysis in the next chapter as a starting point. However, for the purposes of analysing in this book, it seems a bit excessive to check every possible line to see if the c2-c4 move might help Black somewhere, as the entire 1 7 . . . WdS variation is extremely rare in over-the-board play. 8

2 S . 0-0! E:xd6 26.E:xd6 g6 27 . .ihS ! .id7 2B .Wf4 This was Borwell - Sueess, corr. 20 1 3 , which was won by White in brilliant style, which is rather typical of the correspondence games in this line.

7

23.0-0!?

3

This is the simplest way, which does not require too much analysis.

2

23.gS!?N is the maximalist try, when play may continue: 23 . . . hxg5 24.E:gl f6 2 5 .h4 lD de5 26.hxg5 f5 27.lDf6t i> f7 8 7 6

6 5 4

1 a

b

e

d

e

f

g

h

23 . . tll de5 .

23 . . . f6 enables White to make good use of the c-pawn: 24.c5! bS 2 S .g5 hxgS 26.Wh3 f5 27 . .ihS lD ce5 2B.E:df2 E:a7 29.lDxg5 lDf6 30 . .ixe5 1 -0 Selen - Oorer, corr. 20 1 2 .

24.h4

5

24.cS!?� has been played, and also deserves attention.

4 3 2

24 .. J:�xd6 a

b

e

d

e

f

g

h

In the position with the pawn on c2, White does best to avoid this situation by playing c2-c3 instead of hxg5 - see variation 032 of the next chapter on page 2B3. However, in the current scenario he can exploit the extra

24 . . . f6 seems more critical, but it meets with a forceful reply: 2 S . lD xf6t! gxf6 26.E:xf6 lDf7 27.E:g6t i>h7 2B.hS! E:xd6 29.Wf4! Wcs t 30.i>g2 lD cdB 3 1 .E:xd6 Wxd6 32.Wf6 We5 33.E:xh6t i>gB 34.E:hBt lDxhB 3 S .Wxe5+­ Sommerbauer - Reinstadler, COrI-. 20 1 4 .

257

Chapter 1 4 - 1 2 . . . ttJ fd7 White is heading for a more pleasant endgame, although Black has some drawing ' chances.

25 ... b6 26.g5 �c5 27.�xc5 bxc5 28 ..ih5 hxg5 29.hxg5 .id7 30.tLlxf7 .ie8 3 1 .�xe5 tLlxe5

This is the line in which Black can, rather amazingly, make use of the fact that White's pawn is on c4 instead of c2. The main idea is that the pawn can become a weakness, enabling Black to simplify to certain types of endgame that would be unfavourable with the pawn on c2. There is even a line where White's inability to put a piece on c4 enables Black to get away with an otherwise dubious line of play.

22.hf8 Other moves don't promise much.

22 ... �xfB 23.�d6 �bd7

a

b

e

d

e

f

g

h

32 ..idl!?

32 .ixe8 l'!xe8 was agreed drawn in Walter - Kain, email 20 1 2 . After the text move Black was able to hold in Sferle - Gusan, corr. 20 1 2, but in a game without engines, Black will be under pressure for a while.

B212) 2 1 . .. 5!

a

b

e

d

e

f

g

h

24 ..i6!N

White is angling for an improved version of a thematic endgame, and intends to meet . . . �e5 or . . . �c5 with 'it>f2 and 'it>e2 respectively, letting Black exchange on e3 rather than taking on e 5 . Also, White is not in a hurry to play g2-g4, since that move will not be required in the upcoming endgames.

8 7 6 5 4

24.g4 is the recommended move with the pawn on c2, but here Black has two ways of exploiting the small change in the position:

3 2 1 a

b

e

d

e

f

g

h

a) 24 ... fxg4!? 2 5 . 0-0 Wfe5 If you turn ahead to page 267, you will see that with the pawn on c2 White can get a great position with ttJ c4

258

Poisoned Pawn

here. Here it is not possible, and after 26.Wlf2 ttJ f6 27.J.xg4 Wlg5 2 B .'kt) h I the position was balanced in Ntirlis - David-Bordier, corr. 20 1 4. b) 24 . . . Wle5 leads to a version of the endgame which I prefer to avoid. 2 5 .Wlxe5 (Ideally White would like to keep the queens with 2 5 .Wlf2, but Black can launch a series of checks: 25 . . . Wla I t 26.J.d I [After 26Jl:d I Wlc3t! there is no way to escape the checks, since 27.i>f1 ? Wlh3t 2B.i>gI ttJe5 looks bad.] 26 . . . Wle5t If White is to escape the checks he must play the ugly 27.i>f1 ttJ f6, when Black is fine.) 2 5 ... ttJ xe5 26.gxf5 exf5 27.0-0 g6 2B.l!b I Later on page 266 we will encounter the almost-identical position with the pawn on c2, when Black faces a rather unpleasant endgame. The difference here is felt after:

8 7 6 5 4 3 2 1

������

b

a

e

d

f

e

h

g

24 Wlal t ••.

Trying to get an improved version of the endgame by luring White's rook to a slightly worse square. If the rook stays on d2, then the h I -rook can instantly go to b I , leaving both rooks excellently placed. If the rook is on d 1 , it will cost White a tempo to reach the same configuration. 24 . . . Wle5 ? makes no sense, and after 25 .lt>f2! Black's position can quickly spiral out of control. 2 5 . . . Wlxe3t (25 . . . l!bB 26.Wlxe5 ttJxe5 27.l!b l t) 26.i>xe3 ttJc5 27.l!b I l!bB (Mter 27 . . . ttJ fd7 2B .J.xb7! J.xb7 29.ttJxb7 Black does not have much of a blockade against the c-pawn.)

a

b

e

d

e

f

g

h

2B . . . J.e6! Since White cannot take on b7 with the knight, it is hard for him to control the a-pawn. (2B . . . b 5 ! ? is probably close to equality as well, but it is not required.) 29.c5 (Or 29.l!xb7 a5 and Black is not worse.) 29 ... a5 30.l!a I a4 3 1 .l!a3 A draw was agreed here in Yu - Wei, China 20 1 4, although Black's position looks more comfortable to me. 24. i>f2!?N is another interesting move which I analysed in detail, but the quick version is that I prefer the text move.

a

b

e

d

e

f

g

h

2B .J.c6! Not strictly necessary h'e re, but still rather strong. We will see a similar idea in the main line, but here White has an extra tempo.

Chapter 1 4 - 1 2 . . . tt'l fd7 2B . . . g5 29.'kt>d4 b6 30.tt'lxcB E!:xcB 3 l .E!:xb6 Black's positi ? n is collapsing. 24 .. .'�c5 is slightly better than the above line, as Black does not have to take on e3 so quickly, but after 25 .'kt>e2!, which avoids the pin on White's queen, it is hard for Black to do anything, and he always has to worry about White's queen suddenly moving away. 25 . . . E!:bB (25 . . . a5 26.E!:b U) 26.E!:b l

259

27 . . . tt'l xe5 2B .E!:he l lLl xf3 2B . . . lLl ed7 allows 29 . .ixb7 .ixb7 30.tt'lxb7 a4 3 l .c5 E!:cB 32.E!:c l E!:c6 33.E!:ed l and Black is under pressure. 29.gxf3 8 7 6 5

8

4

7

3

6

2

5 4

b

a

3 2

a

b

e

d

e

f

g

h

26 . . . Wh7 (26 . . . g5 is met by 27.h4! intending 27 . . . g4 [27 . . . Wxe3t 2B.'kt>xe3 tt'l c5 29.hxg5 hxg5 30.E!:db2t] 2B .Wxh6! gxf3t 29.gxf3 and suddenly White gets a crushing attack.) 27.g4!? lLlg6 (27 . . . fxg4 2B . .ixg4t) 2B.gxf5 exfS 29 . .ih 5 Wxe3t 30.'kt>xe3 f4t 3 l .'kt>f2 White can continue to press without any worries.

e

d

e

f

g

h

29 . . . a4 30.E!:b l ! ? 'kt>h7 If 30 . . . a3 3 l .E!:b3! the pawn will soon fall. Also after 30 . . ..id7 3 l .E!:xb7 White's chances are preferable. 3 l .E!:b4 lLl d7 32. 'kt>e3 Intending E!:a l . The fight is still on, but I am not sure how Black can create counterplay.

25.E!:dl ti'e5 25 . . . Wa5t does not make sense for Black after 26.Wf2! intending g2-g4.

26. n ti'xe3t 26 . . . a5!? This way Black can try to keep some tension, and not make a concrete decision about the queens. At the same time, it invites the E!:b l -b5 plan in the future. Again, there are multiple ways to continue, but perhaps the simplest decision is: 27.Wxe5 I also considered 27.E!:he l but found that Black was okay.

a

27 tl:k5

b

e

d

e

f

g

h

...

27 . . . a5 ?! is premature: 2B.c5 ! ? g5 (After 2B . . . lLlxc5 29.E!:c l lLl fd7 30.E!:hd l !± Black loses a piece.) 29.c6 Black is in trouble.

Poisoned Pawn

260

27 .. .!!b8 28 .�d2 leads to the same type of positions as the main line. We now reach an important moment, where White has to decide where to place his rooks: on aI and b I , bI and d I , or d2 and b I . The first set-up does not achieve much, but choosing between the other two is not easy.

28J�� d2

28 . . . � b 8 i s met b y 29.�b I intending .ic6 and �d4. Most probably Black should go for 29 . . . g5, which converts to 28 . . . g5 below. 28 . . . g5 29.�b I �b8 (29 . . .f4t 30.�d4 lLl fd7 3 I .g3;l;) 30 . .ic6! Preventing . . . lLl fd7, while threatening �d4. 30 . . . .id7 (30 . . . �g7 3 1 . �d4! b6 32.lLlxc8! �xc8 33.�xb6 is good for White since the d2-rook isn't hanging.) 3 I ..ixd7 lLl fxd7

I analysed 28.�b I �b8 29.�hd I in some detail, but will j ust show the main line where I found a slight problem: 29 . . . g5 30 . .ic6!? �g7 3 I .�d4 b6 Now if 32.lLlxc8?! �xc8 33.�xb6 �d8t! the position of the rook on dI instead of d2 makes all the difference. A better try is Alternatively 32.�xb6, but after 32 . . . �xb6 33.�xc5 �b8 34.lLle8t �f7 3 5 . �d6 �b6!? 36.c5 �xc6t 37.�xc6 �xe8 3 8 .�c7 .id7 39.c6 .ixc6 40.�xc6 �e7 Black draws comfortably. 28 .g4!? is another idea that deserves attention, but I will focus on the main line.

8 7 6 5 4 3 2

a

b

e

d

e

f

g

h

32.lLlxb7! f4t 33.�e2;l; With one ofthe knights exchanged, Black's blockade on c5 doesn't appear as strong, and White is definitely on top.

29J�� b l Th e rook manoeuvre to b5 works well, despite the loss of time. Black will not be able to push the pawn beyond a4, as it will only become a weakness then.

29 a4 30J::1 b s lLlfd7 31 .g4!?� •••

a

28 aS!?

b

e

d

e

f

g

h

..•

Trying to exploit the fact that White did not put his rook on b I on the last move. Now he will have to waste time if he wishes to bring the rook to b 5 .

It is hard to analyse precisely as there are many possibilities, but White's position is certainly preferable. He can exert pressure on b7 or perhaps try a g2-g4 break at some point, simply bringing the king over towards the queenside.

Chapter 1 4 - 1 2 . . . ltJ fd7

26 1

Practical results have demonstrated that this is an inferior cousin of 20 .. .!l:d8 .

B22) 17 ....hd4 18J�xd4

2 1 .0-0! 5 If Black's rook was on d8 instead of e8, White's last move could have been met by 2 1 . . .'lWb6. Thanks to the poor placement of the black rook, the same move now runs into 22.'lWxb6 lD xb6 23 . .th5 ! when Black has no good way to defend f7 .

a

b

e

d

e

f

g

h

We are heading down the main theoretical path of the chapter. The next couple of moves are not completely forced, but they are clearly the most logical, and alternatives are virtually never played.

2 1 . . . lD c6 22.'ilYg3 also leaves Black without much of an answer for the coming attack. The statistical score has been a massacre for White. One nice example continued: 22 . . . 'ilYb6t (22 . . . lD e7 23 . .tc7 b6 24.gxd7 .txd7 2 5 . lD f6t 'it>h8 26.lDxd7±) 23.'it>h l lD e7 24 . .tc7 'ilYc6 8 7 6

18 ...'ilYa5t 19J�d2 0-0 20 ..td6 Now we reach an important branching point. In the rest of this chapter we will look at B221) 20 J��e8?, B222) 20 ... tLlc6?! and ••

B223) 20 ... 5. The most popular move has been 20 . . J�d8 , which will be covered in the next chapter.

B22 1) 20 ... ge8? 8 7

5 4 3 2

a

b

e

d

e

f

g

h

2 5 .!l:xf7! 'it>xf7 26.lD d6t 'it>g8 27.lDxe8 lD fS 28.'lWg6 'lWc3 29.'ilYxe6t 'it> h 8 30.'lWxfS 'lWxd2 Black is doing his best to hang on, but it j ust takes a few more accurate moves to break through to his king. 8

6

7

5

6

4

5

3

4 3

2

2

1 a

b

e

d

e

f

g

h

a

b

e

d

e

f

g

h

262

Poisoned Pawn

3 1 .Wi'f7! Wi'e l t 32 . .ifl Wi'a l 33 . .id6 a5 34.c3! Wi'xc3 3 5 . .ifs lLlxfB 36.Wi'xfSt �h7 37 . .id3t+­ Kubicki - Marek, corr. 2009. 8

26 b5 27.c3 Ab7 28.cxd4 fxe4 29.,be4 ,be4 30.YNxe4 gc8 31 ..be7 YNxe7 32.Wi'g6± •••

Tinture - Schmidt, corr. 2007. Black has avoided being mated, but the position should be technically winning for White.

7

B222) 20 �c6?! •.•

6 5 4 3 2 1 a

b

e

d

e

f

g

h

22.Wi'g3! �c6

Another game ended after 22 . . . fxe4 23 .Wi'g6! gdS 24.gf7 Wi'c3 25 . .ib4 Wi'e5 26 ..ig4! �hS 27.gxg7 Wi'xg7 2S ..ic3 with mate to follow, Siefring - Leontiev, corr. 20 1 0.

a

b

e

d

e

f

g

h

6

This used to be considered a decent option for Black, but recent games and analysis indicate that it is imprecise. The point is that White is in no hurry to swap his mighty bishop for the rook on fS , as this will only help Black to get coordinated. The bishop is so strong on d6 that Black usually plays . . . f5 a move or two later in order to force the exchange on fS , but if is going to play . . . f5 anyway, he should do it on move 20 (as in variation B223 below) , in order to force White's hand. In short, Black's last move can be considered an inferior version of 20 . . . f5, and I will mention a few places where the two lines can be compared with one another.

5

21 .0-0!

4

2 1 ..ixfS ? lLlxfS 22.0-0 Wi'e5 23 .Wi'fL f6 is much more solid for Black. It is important to understand that the bishop on d6 is exerting a much greater influence than the rook on fS , and White should not exchange those pieces unless there is a compelling reason.

23.YNg6 YNd8 24J:�d3! ge7 24 . . . fxe4 2 5 .gf7 is killing.

25.gg3 �d4 26.Ad3 White has a deadly attack. My database shows a long list of correspondencelinternet games, with White scoring close to a hundred percent. Probably the best defensive try is: 8 7

3 2 1 a

b

e

d

e

f

g

h

Chapter 1 4 - 1 2 . . . ltJ fd7 It is worth mentioning that 2 1 .g4 is not the most effective choice here. Black has a precise defence with: 2 1 . . .'lWb6! 22 .'lWc3!? (22.'lWh3 5 ! was better for Black in Smirnov - Nepomniachtchi, Novokuznetsk 200B; 22.'lWg3 ltJ d4! is another key idea. Now 23.g5 is met by 23 . . . ltJ f5 24.'lWg2 h 5 ! when White's attack is not really dangerous.) The text move is a good try, but Black has a good answer: 22 . . . 'lWa5 ! Possibly the only move. Now 23 .'lWg3 can be met by 23 . . . ltJ d4 24.i.d l f5 ! 25 .i.xfB fxe4 and Black is doing well.

263

27.i.f4! This posltlon has occurred in several correspondence games, with White maintaining a perfect score so far. Here is one brief example: 27 . . . :Bd7 2B .i.g6! 'lWb6t 29 .i.e3 'lWdB 30.i.xh7t 1 -0 Pheby - Paleckis, corr. 20 1 1 . As mentioned previously, 2 1 . . .f5 leads to a worse version of the more respectable 20 . . . 5 line. 22.i.xfB ltJ xfB 23.ltJd6 8 7 6

2 1...ltJce5 Trying not to commit to . . .f5 makes sense, as playing this move will j ust lead to a worse version of 20 .. .f5 . However, White is not in a hurry to take on fB either. 2 1 . . .'lWb6 is not really a viable option due to 22.'lWxb6 tlJxb6 23.i.xfB �xfB 24.ltJd6 f6 25.ltJxb7! (25.:Bb 1 is also promising) with a clearly better endgame for White. 2 1 . . .:BdB meets a horrible fate: 22.'1Wf4! 'lWf5 23 .'1Wh4 'lWa5 (Another game continued 23 . . . g5 24.'lWxh6 'lWxe4 2 5 .'lWxg5t 'lWg6 26.'lWh4! 'lWh7 27.i.h5! 5 2B .:Bd3 1 -0 Blank - Lounek, corr. 200B.)

5 4 3 2

a

b

e

d

e

f

g

h

Black finds it hard to exchange queens here, and even if he does manage it, he gets an unfavourable version. It may still be possible to hang on and defend, especially in a correspondence game, but I can't imagine anyone wishing to do so in a tournament. I will show a few example lines after a) 23 . . . ltJ e 5 , b) 23 . . . � h 7 and c ) 23 . . . 'lWe5 . a) 23 . . . ltJe5 24.:Bb l �h7 (24 . . . b5 ltJ fg6 26.:Bal 'lWc7 27.ltJxb 5 'lWxc2 Moore - Ryan, corr. 2009.) i.xb7 26.:Bxb7 :BcB 27.h3 ltJ fg6 Laffranchise - Price, corr. 20 1 1 .

8 7 6 5 4 3 2 a

b

e

d

e

f

g

h

24.:Bd3! White has too many pieces in the attack. 24 . . . ltJfB 2 5 .i.h5 5 26.:Bg3 ltJ h7

2 5 .:Bdd l 2B.:Bac 1 ± 2 5 . ltJ xb7 2B .i.h5;!;

b) 23 . . . �h7 This is a strangely popular move in correspondence games. The simplest reply seems to be: 24.g4 Another interesting plan is 24.h4!? 'lWe5 2 5 .'lWxe5 ltJ xe5 26.:Be l ltJ fd7 27.h5 ltJ f6

264

Poisoned Pawn

2B .�d l lLl c6 29 .�f3 a5 30.c3 as in Magalhaes - Rosas, corr. 2009. White maintains his extra exchange, and will slowly collect some of the pawns back while retaining long-term pressure. 24 . . . lMre5 24 . . . lLl g6 2 5 . gxf5 exf5 26.�h5 lLl ge7 27J!g2t 2 5 .lMrxe5 lLl xe5 26.gxf5 exf5 Fischer - Spitz, corr. 20 1 2. When you compare this position with the 20 . . . f5 endgames, it becomes obvious that White has an improved version here. c) 23 . . . lMre5 This gives White a pleasant choice: he can either simplifY to a slightly better endgame, or continue to press with queens on the board. 24.lMrxe5 24.1Mrf2 ! ? �d7 (24 . . . l'!bBN 2 5 .g4!?) 2 5 . lLl xb7 lLl b4 26.lMrc5;!; Kozlowski - Mamonovas, corr. 200B. 24 . . . lLl xe5 2 5 .l'!b l There is not much point in gIVIng exact variations here, since both sides have many options, but a few games to show the general ideas are: 8 7 6

27.l'!d4!? l'!bB 2B.l'!a4 a5 29.lLlxcB l'!xc8 30.l'!xa5 lLl fd7 3 1 .l'!a2 <;t>f7 32.l'!xb4 also gave White the better chances in Walter Teeriaho, corr. 20 1 1 . 27 . . . lLl fd7 2B.�b5 lLl f6 29.l'!e2 lLlg6 30 .g3 <;t>f8 3 1 .ic4 Black's pawns soon started dropping in Auzins - Teimer, corr. 200B. 8 7 6 5 4 3 2 1 a

b

e

d

e

f

g

h

22.l'!ddl! 22.c4 b5 i s not s o clear.

22 f5 •••

22 . . . l'!eB is seldom a good idea in these positions, and this is no exception. 23.�h5! lLl g6 24.1Mrf3 1Mrf5 2 5 .lMre2 lMra5 26.l'!xf7! c;f{xf7 27.lMrg4 lLl dfB 2B.l'!f1 t +- Sutkus - Guevara, corr. 2009. 22 ... lLl g6 23.c4 l'!eB 24.c5± White was dominating in Yeremenko - Enricci, corr. 200B.

5 4 3 2

23 .lxm •

a

b

e

d

e

f

g

h

25 . . .b5 25 . . . lLl fd7 26.lLlxcB l'!xcB 27.l'!xb7 lLl c5 2B.l'!a7;!; Santos - Flor Trejo, corr. 20 1 1 . 26.l'!a l b4 27.l'!a5

23.lLld2!? is an interesting alternative, but one good line is enough.

23 c;txf8 24Jt�d6 1Mrc5 •••

24 . . . lLl f6 can be met by 25.c;f{h U when, with the queens on the board for the foreseeable future, White has a pleasant position.

265

Chapter 1 4 - 1 2 . . . tt:l fd7

The latest trend, and the recommendation in The Sharpest Sicilian. Apart from being an objectively good move, the text has the advantage of forcing White's hand, making it an appealing choice for over-the-board players. Compared to the previous line where Black played 20 . . . tLlc6 and followed up with . . . 5, here Black is more flexible, as he keeps the useful option of developing the bB-knight to d7.

2 1 .hf8 �xf8 22.�d6 a

b

e

d

e

f

g

h

No other move makes any sense.

22 ... �bd7!

25J�d4!N A rather obvious novelty, preventing a queen exchange and preparing g2-g4. This move was mentioned by Georgiev and Kolev in The Sharpest Sicilian 2012 as part of the explanation for why they abandoned 20 . . . ttk6. 25.tLlxf5 ?! 'lWxe3t 26.tLlxe3t �e7 27.:B:d4 b5 was quite solid for Black in Karj akin - Kokarev, Sochi 20 1 2 .

25 ... b6 26.g4 tLlf6 27.gxf5 exf5 28.�xf5;t Black is under pressure. The white rooks have a great deal of power on the open board. Having won back a pawn, White may also consider a queen exchange at a suitable moment.

B223) 20 ... 5

This is the best square for the knight. Now Black can offer a queen exchange from either c5 or e5. 22 ... 'lWa 1 t ? is premature, as after 23 .id 1 tLl bd7 24.c3 the queen is stuck. 8 7 6 5 4 3 2 1 a

8

b

e

d

e

f

g

h

23.g4!

7

The most critical.

6

23 .ic4 'lWa 1 t 24.'�;>f2 was a bizarre line played in a couple of games. Even though there is no obvious win after 24 . . . 'lWxh 1 , Black could also have played 23 . . . 'lWe5 leading to a safe endgame.

5 4 3 2

23.h4N intending to meet . . . 'lWc5 with :B:h3

1 a

b

e

d

e

f

g

h

266

Poisoned Pawn

seemed like an interesting idea, until I realized that Black can simply avoid the queen exchange and continue developing with 23 . . . lLl f6! 24.0-0 Vff c 7, when White is too slow to generate any initiative.

8 7 6 5 4

23.0-0! ? Vff c 5 24Jl:d3 is quite an interesting alternative, when 24 . . . b6 2 5 .!l:b l Vff c7 26.g4 has been played in a few correspondence games. White has decent chances here, although I am a bit mystified by Black's plan of . . . b6 and . . . Vff c 7. The rook on d3 seems rather awkward, and I would be worried about proving an advantage in some versions of the many possible endgames. Still, this could be a possible direction to explore in the event that the main line gets exhausted. 8 7 6 5 4 3 2 1 a

b

e

d

e

f

g

h

23 fxg4 •••

This has been played in almost all games so far. Black can try to exchange the queens immediately, but it only seems to lead to slightly worse endgames where Black will have to suffer in the hope of drawing.

3 2

a

b

e

d

e

f

g

h

26.0-0 g6 (26 . . . b5 27.lLlxf5t) 27.!l:b l (27.!l:el intending .id l is another possibility.) 27 ... b5 28.!l:aU White will pick up the b5-pawn. Even though Black will still have two pawns for the exchange, White's rooks are active and the c-pawn will be strong. 23 . . . Vff a l t ! ? This improves slightly on the above line, as White is forced to misplace either his bishop or his rook. However, this does not greatly alter the assessment of the endgame. 24 . .id l Vff e 5 (24 . . . fxg4 ? 2 5 . 0-0±) 2 5 .Vffxe5 lLlxe5 26.gxf5 exf5 27. 0-0N Going after the f5-pawn seems simplest. 27. �f2 was played in a correspondence game. White managed to create some pressure, although it wasn't sufficient in the end: 27 . . . .ie6 28.lLlxb7 !l:c8 29.!l:e l lLlf7 30 . .if3 lLl g5 3 1 ..ig2 !l:c4 32.!l:a l �h7 33.lLld6 !l:f4t 34.�e l lLl g6 3 5 .!l:xa6 lLl h4 and Black went on to hold in Schinke - Turati, corr. 20 1 2 . Obviously it's a lot easier to defend like this when you have the assistance of a computer. 8 7

23 . . . lLl f6 24.!l:g l gives White a promising initiative. 23 . . . Vff e 5 is the most convenient version of the endgame for White: 24.Vffx e5 lLlxe5 2 5 . gxf5 exf5

6 5 4 3 2

a

b

e

d

e

f

g

h

267

Chapter 1 4 - 1 2 . . . ttJ fd7 27 . . . id7 27 . . . g6 2B.,E:e l ± 2B. ctJxf5 2B .E:e l is also possible, for instance 2B . . . ctJ fg6 29.h4!? ctJg4 30.ttJxb7t and White has no cause for complaint. 2B . . . E:eB 29.c3t Obviously Black has drawing chances, but he is a long way from being completely safe.

8 7 6 5 4 3 2

8

1

7

b

a

e

d

f

e

g

h

25J�d4! b6

6

25 . . . b 5 ? would be a mistake. The fact that the queen remains unprotected will come back to haunt Black after 26.ixg4 ttJ f6 27.E:xf6! gxf6 2B .if3!+- threatening E:g4t.

5 4 3 2 1 a

b

e

d

e

f

g

h

24.0-0!

This move, intending to stay in the middlegame, is not mentioned in The Sharpest

Sicilian. 24.ixg4 has been the main try, but it will be hard to prove anything in the upcoming endgame where Black has three pawns for an exchange. 24 . . . 'Wa l t (A familiar idea, forcing the bishop back to d 1 , although the immediate 24 . . . 'We5 also seems playable.) 2 5 .id l 'We5 26.'Wxe5 ttJxe5 27.0-0 This position was reached in Yu - Wei Vi, Xinghua 20 1 4 , and now after 27 . . . a5N Black should be fine.

24 ...�c5

This is the only move to have been played. 24 . . . 'We5 would be a mistake due to 2 5 .'Wxe5 ttJxe5 26.ttJc4! ttJ ed7 (26 . . . ttJxc4 27.E:dB+-) 27.ixg4 with a clear advantage to White.

2 5 ... e5 26.E:d3! (26.ic4t i>h7 27.id3t i>gB 2 B . ttJ f5 b 5 ! [avoiding 2B . . . exd4 29.'WeB +-l may look tempting for White, but it is hard to find anything conclusive.) White will get an improved version of the endgame, as the pawn will be weaker on e5 than on e6. For instance: 8 7 6 5 4 3 2

a

b

e

d

e

f

g

h

26 . . . h5 (26 . . . 'Wxe3t 27.E:xe3 ttJ f6 2B.E:xe5 b 5 29.E:a l;:\;) Defending the g-pawn seems logical, but after 27.'Wxc5 ttJ xc5 2B .E:d5 ttJ fd7 29.E:f5 g6 30.E:f7 ttJ e6 3 1 .E:d l b 5 32.E:e7 ttJ df8 33 .id3t White maintained the pressure in Siefring - Zarate, COff. 20 1 0.

268

Poisoned Pawn

27.�e4 V!le7

Black could consider deviating with: 27 . . . Wc7N This makes some sense, since ttJ d6 can now be met by . . . Wc5 . 28 .Wg3! .ib7 29 . .ixg4 cj;>h7 The critical position.

a

b

e

d

e

f

g

h

26.l:H2! In the event of 26.'!Mff2 ttJf6 27.ttJe8, Black can play 27 . . . ttJ 8d7! 28.ttJxf6t ttJ xf6 29 J�� d 8t �f7 30 .Wxc5 bxc5 31 . .ixg4 �e7 defending everything. 26.:B:f7 was played in one of the few over­ the-board games on my database, but Black immediately got a good position after 26 . . . ttJ e 5 ! 27. ttJ e4 W a 5 28JW .ib7+ in Akshayraj Sandipan, Kolkata 20 1 2.

26 ... �e5

26 . . . h5 feels too artificial. I like White's position after 27 . .id3 ! ? , intending 27 . . . ttJe5?! 2 8 . ttJ e4! We7 29.ttJg5 ± . 26 . . . We5 seems a logical move, but once again White can exploit the weakness of the 8th rank to force Black into a passive endgame: 27.Wxe5 ttJ xe5 2 8 . ttJ c4! ttJ ed7 (28 . . . ttJ xc4? 29.:B:d8+-) 29 ..ixg4 Black has a lot of pawns, but it is hard for him to develop. 29 . . . :B:b8 (29 . . . b5 3 0 . ttJ b6! ttJ xb6 3 1 .:B:d8 ttJ bd7 32 . .ixe6t+-) 30.:B:d6 b5 3 1 . ttJ a5 :B:b6 32.:B:fd2 The a6-pawn will eventually fall, and Black remains under pressure. The following line is not forced, but it illustrates White's plan: 32 . . . :B:xd6 33.:B:xd6 ttJe5 34 . .ih3 �f7 3 5 .:B:d8 .id7 36.:B:a8 ttJ fg6 37.:B:xa6 ttJ f4 3 8 . .ifl;!;

a

b

e

d

e

f

g

h

30.h4! 30 . .ih5 was the machine's first suggestion, but it leads to a drawish endgame after: 30 . . . .ixe4! 3 1 .:B:xe4 ttJ fd7 32.:B:fe2 Wc5 t 33.�g2 :B:f8 34 .:B:xe5 ttJxe5 35.:B:xe5 Wxc2t 36 . .ie2 Wc6t 37 . .if3 Wd6 38 .:B:e2 Wxg3t 39. hxg3 :B:f6 It is hard to imagine White winning from here. 30.ttJd6 Wc5! 3 1 .c3 .ic6 32.ttJf7 ttJeg6 100ks shaky, but Black is surviving. 30 . . . .id5 ! ? A strange computer idea to survive. 30 . . . :B:e8 3 1 ..ih5± 30 ... b5 runs into 3 1 .ttJd6 Wc5 32.c3 .ic6 33.ttJ f7!± and now h4-h5 is an important resource against a knight on g6. 3 1 .c4 .ib7 Black's idea is that ttJ d6 can be met by . . . Wc5 , since c2-c3 cannot be played now. However, White can switch plans while conveniently making use of the free c2-c4 move. 32 . .id l ! With a dangerous initiative.

28.tl�d6 �fd7

Chapter 1 4 - 1 2 . . . lLJ fd7 Another game continued: 2B . . . ttJ fg6 Here Black is able to simplify more easily, but it is still not enough to kill the game. 29 ..td3 �gS 30.�xgS hxgS 3 1 ..txg6 3 1 .ttJ f7 leads to the same position. 3 1 . . .ttJ xg6 32.ttJf7 .tb7 33.ttJxgS as 33 . . . ttJeS 34Jl:e2 ttJ f3t 3 S . ttJ xf3 .txf3 36J:l:xe6 bS 37.!':!:e7 (or 37.!':!:b6!?) is not yet a draw. 34.!,:!:d7 .tf3 3 S . ttJ xe6 ttJ fB!

a

b

e

d

e

f

g

269

37 . . . ttJ d7 3 B . ttJ e6 .tdS 3 9 . ttJ c7 .tc4 40.!':!:f4 bS 4 1 . ttJ eB! a4 42.ttJd6 ttJ b6 43.!':!:xg4 a3 44.�f2 a2 4 S . !':!:g l ± 3B .!,:!:d2 The safe choice. 3B .c4!? �fB 3 9 . ttJ e6t �e7 40.ttJxg7 �d6 4 1 .h4 is more of a mess, although White may well be better here too. 3B . . . �fB 3 9 . ttJ e6t 39.!':!:d7;!; 39 . . . �e7 40.ttJxg7 ttJ gS 4 1 .c4;!; Obviously Black has chances to make a draw in all these lines, but his task is not easy.

h

36.!':!:dB!N 36.!':!:xg7t 'kt>hB 37.!':!:e7 ttJxe6 3B.!':!:xe6 bS (3B . . . a4 39 .c4! a3 40.!':!:e l ±) 39.h3 a4 Yz-Yl was Waiter - Johnson, corr. 20 1 1 , one of those typical insane draws which would be hard to imagine in an over-the-board game. The continuation might be 40.hxg4 a3 4 1 .!':!:xf3 a2 42.!':!:e l a l =� 43.!':!:xa l !':!:xa l t 44.'kt>f2 !':!:c 1 when Black seems to b e able to force the exchange of his b-pawn for White's c-pawn. Perhaps White can look for a finesse somewhere in this line, but I find the text move a lot more logical. 36 . . . !':!:xdB 37.ttJxdB Exchanging a pair of rooks makes perfect sense. The a-pawn is not really dangerous, and it will be hard for Black to hold on to his pawns once White has activated his rook. A few illustrative lines: 37 . . . ttJh7 37 . . . a4 3B .!,:!:d2;!;

a

b

e

d

e

f

g

h

29J�dl!N The point of this move is to threaten �c3 , as the reply . . . ttJ cS will not be possible due to the hanging knight on e S .

I found a few games where White played rather insipidly with 29 .�c3 ttJ cS 30.�g3 and draws were soon agreed. 30 . .txg4 is a more ambitious try, but it leads to a messy position after Black completes development: 30 . . . .td7 3 1 .�g3 ttJ c6 32.!,:!:d l !':!:fB 33 .!':!:xfBt �xfB 34.!':!:f1 �e7°o Matei - Kruger, corr. 20 1 3 . 29.h3!?N gxh3 30.�h2 i s a n interesting idea, intending to attack along the g-file, but we don't need to hurry with it.

270

Poisoned Pawn

29 J�b8 ••

This is the computer's top suggestion. Black prepares . . . i.b7 in some lines, and removes the rook from its slightly exposed position on the long diagonal. 29 ... a5!? introduces the possibility of ... i.a6. Here White can switch plans with 30.�e4 1'!b8 and now the amazing 3 1 . 'kt>f1 ! , j ust making sure that the knight will not come to f3 with check. It's still a messy position, but more dangerous for Black, who has to worry about i.b5 .

32 . . . b4 33.�g3 �g5 34.tDe4! �g6 3 5 .tDc5 �g5 36.i.c4!±

33.g5!? 33.�g3 should lead to a pleasant endgame: 33 . . . a4 (33 . . . �g5 34. tD e4±; 33 . . . b4 34.i.c4±) 34.i.xb5 �g5 (34 . . . a3 ? 3 5 .g5 �xg5 36.�xg5 hxg5 37.tDxc8+-) 3 5 .i.xa4 tD f6 36.�f4 �xg4t 37.�xg4 tD fxg4 38 .1'!f4 White is pressing, but I prefer the attacking continuation shown in the main line.

30.Y*ic3 Y*id8 The only move. 8 7 6 5 4 a

3

b

e

d

e

f

g

h

33 ...Y*ixg5t

2

33 . . . b4 34.�d4 �xg5t 3 5 .1'!g2 reaches the note to Black's next move in the main line.

1 a

b

e

d

e

f

g

h

3 1 .h3! 3 1 .1'!g2!? deserves attention as well, but I like the text move more.

3 1 . b5 .•

3 1 . . .a5 32.hxg4 b5 is the same thing. 3 1 . . .gxh3 32.�xh3 �e7 33 .�c3 �d8 34.1'!g2! gives White a powerful initiative, for instance: 34 . . . b5 3 5 .i.g4! ? (35 .i.h 5 ! ?) 3 5 . . . tD xg4 36.1'!xg4t

32.hxg4 a5

34.1'!g2 Y*id8 34 . . . b4 3 5 .�d4! (35 .�h3 �e7 36.�xh6 i.b7 gives Black a slightly improved version of the main line after 34 . . . �d8 .) 3 5 . . . �e7 36.1'!f1 ! g5 This ugly move is forced. 37.i.h5 White has a powerful initiative, although it's not conclusive yet.

35.Y*ih3!

I also considered 3 5 .�b3 , when 35 . . . 'kt>h7? allows White to perform a nice dance with the queen: 36.�g3! �g5 (36 . . . �e7 37.1'!f1 i.b7 38.1'!f7 �g5 39 .�d3t!+-) 37.�c3 �d8 38.tD f7+-

27 1

Chapter 1 4 - 1 2 . . . lLl fd7 However, a better defence is 3 5 . . . �e7 36.lLlf5 �c5t 37.�d4 liJ f3t 38 . .ixf3 �xf5 39 . .ig4 �c5 40 . .ixe6t Wh8 4 1 .�e3 when White keeps some pressure, although the position is simplifying a bit faster than I would like.

a

b

e

d

e

f

g

h

35 ...�e7 35 . . . Wh7? 36.�g3! transposes to 3 5 .�b3 Wh7? 36.�g3! in the previous note.

36.�xh6 36.cj;>h2!? also seems to offer White a strong initiative.

36 ....ib7 37J3g3t White keeps creating threats, while Black does not seem to have a clear way out.

Conclusion The 1 2 . . . lLl fd7 variation is one of the most important in the Poisoned Pawn. After the usual continuation of 1 3 . lLl e4 �xa2 1 4.�d l �d5 1 5 .�e3 �xe5, Black has deposited three extra pawns in the bank, but White has a formidable lead in development. Of the many variations examined in this chapter, there are rwo that stand out as especially interesting. Variation B2 1 with 1 7 .. :�d5 sees Black deliberately lose a tempo to provoke c2-c4. In the maj ority of positions this free move is more likely to benefit White, but there are a few specific cases where it can help Black. I have shown how White can modify his plans to use the extra move advantageously, although I should add that this line has almost exclusively been the domain of correspondence players so far. The last section of the chapter introduced the main line beginning with 1 7 . . . .ixd4. After the standard moves 1 8 .�xd4 �a5 t 1 9.�d2 0-0 20 . .id6 we reach an important branching point where, after dismissing a few questionable alternatives, we examined 20 . . . f5 in detail. My analysis departs from that in The Sharpest Sicilian 2012 at move 24, which is actually quite early by the standard of the Poisoned Pawn nowadays. With the help of an improvement over a correspondence game a few moves later, I believe I have been able to keep the defence under pressure. In the next chapter we will deal with the most popular 20 . . . �d8.

8 7 6

L=P'= =. V"�� •.=J····�w=.

5 4

Poisoned Pawn

3 2

bm/,···N/m/,jmm ,=Fm'//N,j·mN1 .mm. ......r....

a

b

c

d

e

f

g

20 J�d8 ..

Variation Index l .e4 c5 2.tlJ a d6 3.d4 cxd4 4.tlJxd4 tlJ f6 5.tlJ c3 a6 6.i.g5 e6 7.f4 �b6 8.�d2 �xb2 9J3b l �a3 1 0.e5 h6 1 l .i.h4 dxe5 1 2.fxe5 tlJ fd7 1 3.tlJ e4 �xa2 14J:!: d l �d5 1 5.�e3 �xe5 1 6.i.e2 i.c5 1 7.i.g3 i.xd4 1 8J:!:xd4 �a5t 1 9J:!:d2 0=0 20.i.d6 �d8 2 1 .g4!! 274 274 275 277 278 279 280 281 283

A) 2 1 ,..�b6 B) 2 1 ,..f6 C) 2 1 ,..tlJ f6 0) 2 1 ,..tlJ c6 22.g5 hxg5 23J:!: g l ! 0 1 ) 23,..g6? 02) 23,..tlJ f6!? 03) 23,..f6 24.h4!? 03 1 ) 24,..tlJ rs 032) 24,..tlJ de5

C) after 22 . . . gxf6

D32) after 27 . . . f4!?

D 3 1 ) after 36 .. . f4 8 7 6 5 4 3

.�.... ·.;C;;;J·'..'..··m/../......

.,�......../'m.. b///j"N/NN/"WNNm ,'-;;:;;"'=/'///,,]/NN/1

2 a

b

e

d

e

23.c3!?N

f

g

h

a

b

e

d

e

f

37.@c l !N

g

h

a

b

e

d

e

f

2 8 . tiJ f6t!N

g

h

h

273

Chapter I S - 20 . . . Ei:d8

l.e4 c5 2.�f3 d6 3.d4 cxd4 4.�xd4 �f6 5.�c3 a6 6 ..ig5 e6 7.f4 'i'b6 8.'i'd2 'i'xb2 9J�b l 'i'a3 '1O.e5 h6 1 1 .J.h4 dxe5 12.fxe5 �fd7 13.�e4 'i'xa2 14J�dl 'i'd5 15.'i'e3 'i'xe5 16.J.e2 J.c5 17.J.g3 hd4 18J�xd4 'i'a5t 19JM2 0-0 20 .id6 gd8

29 .�d2 �xd2t 30.Ei:xd2 4JeS 3 1 . 0-0 gS (3 1 . . . .ic6!? 32.Ei:xf4 4J fd7 also gives decent drawing chances.) 32.4Jxb7 as With double­ edged play, Busemann - Zambor, corr. 20 1 0 .

•

It seems strange to start a chapter twenty moves into the game, but the Poisoned Pawn is - to put it mildly - not like other opening systems. In terms of popularity Black's last move is the most important option in the position, although it was untested until 2006. Like many of the lines examined in this and the previous chapter, it has mostly been used in correspondence games due to the dangers it involves. Still, White needs to play extremely precisely ifhe is to prove anything. I remember analysing this line in great detail in 2006, when there was not a single example in the database. No correspondence games to kill all the fun, and even the engines were not too strong in those days; Maybe I used to be a good player back then, because eventually I discovered the following idea.

21 .g4!! I remember analysing this line in something of a daze for a few days - it j ust felt too surreal. I finally got a chance to play it four years later, and now, with more than thirty subsequent games (mostly correspondence) , the idea still seems quite potent. 2 1 .�g3 has been played in many hundreds of games, without any clear conclusion in White's favour. I will j ust give an extremely short summary of the theoretical picture. 2 1 . . .4J c6! (2 1 . . .�f5 used to be the main line, but this was refuted spectacularly by 22 . .ieS ! �g6 23 .�h4 4J c6 24 . .ihS �f5 2 S . .ixg7!, which has brought White hundreds of wins in correspondence play.) 22 ..ic7 �al t 23 .Ei:d l �b2 24.c3!? The critical line. 24 . . . Ei:fB 2S . .id6 f5 ! 26 . .ixfB 4Jxf8 27.4Jd6 .id7 28 .�e3 f4

b

a

e

d

f

e

g

h

I have analysed four main moves here, of which the last is by far the most important: A) 2 1 . 'i'b6, B) 2 1 . f6, C) 2 1 ...�f6 and D) .•

••

2 1 ...�c6. The untested 2 1 . . . f5 ? may seem like a natural reaction, but it j ust opens up the position for White. 22.gxfS ! exfS (22 . . . �al t 23.Ei:d l �aS t 24.c3 exfS 2 S J'!g l +-) 8 7 6 5 4 3 2

a

b

e

d

e

f

g

h

23.Ei:g l ! i>h8 (23 . . . fxe4 24.'i'd4! 4JeS 2 S . Ei:xg7t i>xg7 26 . .ixeSt +-) 24.�d4 Ei:g8 2 S .Ei:g6! i>h7 26 . .ic4! i>xg6 27 ..ixg8 With mate to follow.

274

Poisoned Pawn

29 ...Wfe5 30.i.e3!+-

A) 2 1 . ..Wfb6

White has too many threats, and a queen exchange will not help Black.

B) 21. .. £6 8 7 6 5 4 a

b

e

d

e

f

g

h

This move has the idea to disturb White's pieces, but White has a convenient answer.

3 2

1

Renewing the threat to push the g-pawn. The queen only seems to be improved now. 22.tD f6t is not so clear after 22 . . . �h8! (22 ... gxf6 23 .Wfxh6-+) 23 .i.cS WfaS 24.i.e7 tD c6 with complications.

b

a

22.Wfg3!?

e

d

f

e

g

h

Trying to hold back the g-pawn? No chance!

22.g5! fxg5 22 . . . hxgS allows White to carry out a dreamlike onslaught: 23.h4! g4 24.hS tDc6 2 S .h6 g6 26.i.xg4 f5

22 ... �c6N This seems like the most logical attempt to improve on 22 . . . Wfb l t 23 .i.d l tD c6 24.gS hxgS 2 S . h4! g4 26.Wfxg4 Wfb2 27.E!:gl g6 28.hS when Black resigned in Dolgov - Tarrio Ocana, corr. 20 1 3 . 22 . . . f6? is no good due to 23.gS! hxgS 24.h4 and Black's position is collapsing.

8 7 6 5 4 3 2

23.g5 hxg5 24.Wfxg5 24.h4!? looks quite strong as well.

24 ...Wfbl t 25J��dl Wfxc2 26J�gl g6 27J:�g4 Wfb2 2S.E!:h4 geS 28 . . . bS 29 .i.e7+-

29.i.c5! Threatening E!:xd7.

a

b

e

d

e

f

g

h

27.WfgS! mh7 28 .i.hS!+-

23.h4 g4 23 . . . gxh4 24.E!:xh4 tD c6 (After 24 . . . WffS 2S .i.c4+- every white piece is coming into the attack.) 2 S . E!:xh6! Black will be annihilated on the kingside, for instance:

275

Chapter 1 5 - 20 . . J%d8 29 .�e7t! \t>g8 30.0-0! +-

25J:�gl The attack is too strong, and White probably has multiple ways of winning. A few examples:

25 ... �h8 26.�g3 �al t O r 26 . . . tD de5 27 . .id 1 ! :9:g8 2 8 . .if8 �c7 29 .:9:£2+- threatening .ixg7t . b

a

e

d

e

f

g

h

25 . . . �a l t (25 . . . gxh6 26.�xh6 tD d4 27 . .ih5 +25 . . . tD d4 26.�xd4 gxh6 27 . .ic4+-) 26J%d 1 �b2 27J�h8t! \t>xh8 28 .�h3t \t>g8 29.�xe6t c;iJh8 30.�h3t c;iJg8 3 1 ..ic4 mate! 8 7 6 5 4 3

a

2

b

e

d

e

f

g

h

28.�g5! �d8 29 ..tc7

1 b

a

e

d

f

e

g

h

Black resigned in Hohne - Permykov, Internet 20 1 3 .

24 ..txg4 �c6

Another beautiful winning line is: 24 . . . tD b6 25.tDf6t! c;iJh8 26.�e4 gxf6 (26 . . . �a1 t 27 . .id 1 �xf6 28 . .ie5 :9:xd2 29 . .ixf6+-) 27.�g6 :9:xd6 28.�xf6t \t>h7

7 6 5 4 3 2

a a

b

e

d

e

f

g

h

b

e

d

e

f

g

h

276

Poisoned Pawn

This provocative move has been played twice in email games. White tried to refute Black's play with a blunt attack, but in both cases Black survived with seemingly miraculous defence. The chances of anyone doing the same over the board are minimal, but j ust in case, I found a way for White to build his initiative more slowly. It seems Black's best-case scenario is to simplify to an exchange-down endgame and fight for a draw.

similar variations. 23 . . . <;i.>g7 is met by 24.h4! threatening g4-g5 .

22.tihf6t gx:f6

b

a

e

d

f

e

g

h

24.h4 gxd6

24 . . . <;i.>h7 is a computer suggestion, but it seems artificial. White gets a strong initiative after: 2 5 .g5 f5 26.gg 1 ! ? I'm sure there are other good moves. 26 . . . gxd6 27.gxd6 W'e5 28 .W'd2 W'h2 8

a

b

e

d

e

f

g

h

23.c3!?N

Avoiding 0-0 completely. My idea is that the rook on h I can be well placed to support an attack with h2-h4 and g4-g5 , while in the likely endgames after a .. J::\xd6 exchange sacrifice, it will be useful to have the king in the centre. 23.0-0 llJ c6 leads to some insane variations where Black just seems to hold on. See the game Tesic - Shablinsky, corr. 20 1 3 , where Black survived the middlegame and held a suspicious­ looking endgame with rook and knight against queen. From a practical point of view this would not be much fun for Black either, but I like my recommended line even more.

The immediate 23 . . . gxd6 would lead to

7 6 5 4 3 2

a

b

e

d

e

f

g

h

29. Wfl ! Black is in trouble. After some thinking, the only way the computer manages to survive is by entering a piece-down position after: 29 . . . llJe5 30.W'd4 hxg5 3 1 .hxg5 b6 32.g6t fxg6 33J%g2 W'h l t 34.Wf2 j,b7 35 J%d7t llJxd7 36.W'xd7t Wh8 37.W'xb7t Materially Black has enough pawns for the bishop, but he is in obvious danger due to his exposed king.

25.gxd6 �e5

277

Chapter 1 5 - 20 . . . E!:d8 25 . . . �g7 just provokes 26.g5! fxg5 27.hxg5 hxg5 2B.E!:h5 ,with a strong attack. 25 .. Y/lc7 26.E!:d2 �d7 is a logical attempt to develop, but it is too slow. 27.E!:h3! Threatening �xh6. (27.g5 tD e7! creates a mess.) 27 . . . �g7 (27 . . . tDe7 2B .�xh6±) 2B.g5 With an attack. 25 . . . tDe7 tries to create quick counterplay, but Black is far too behind in development. 26 J:l:f1 ! The king will be useful in the centre after the queen exchange. 26 . . . �e5 (26 . . . tDd5 27.!'!xd5! is the key idea. 27 ... �xd5 2B.!'!xf6 �d7 29 .!'!xh6± Threatening !'!h5.) 27.�xe5 fxe5 2B.g5 hxg5 29 .E!:dBt �g7 30.hxg5 White intends �d3 and !'!h 1 , while Black is absolutely stuck.

26.�xe5 �xe5 Mter 26 . . . fxe5 27.g5 ! White will be able to create an entry for the rook either along the h- or the g�file, and Black's a-pawn is not really dangerous yet. 27 . . . hxg5 (27 . . . a5 2B.gxh6;!;; ; 27 . . .'J::i g7 2B.gxh6t �xh6 29 .!'!g l;!;; ) 2B.hxg5

a

b

e

d

e

f

g

h

28.
The endgame will be unpleasant for Black. 2B.�d2!? also makes sense, intending to use the king to blockade the a-pawn in the future.

28 ... b5 2B . . . a5 29.!'!hd l a4 30.!'!eB! followed by !'!ddB will create mating threats as well as planning to take on cB. 30 . . . tD c6 3 1 .�f3 +Both 2B . . . tD c6 29.!'!eB;!;; and 2B . . . b6 29.!'!b l;!;; favour White as well.

29J3a1� Black will drop a pawn on the queenside.

D) 21 ... �c6

a

b

e

d

e

f

g

h

2B . . . e4 (2B . . . a5 is met by 29.�f3+- followed by �xc6 and !'!dB{t) with mating nets.) 29.�d l ! ? WfB 30.g6 Wg7 (30 . . . fxg6 3 1 .!'!h7±) 3 1 .gxf7 Wxf7 32.!'!h4;!;; Black will continue to struggle with development for a while, so the position remains one-sided.

27J;d8t
a

b

e

d

e

f

g

h

278

Poisoned Pawn

Developing the knight is the obvious move, and it has been by far the most popular choice.

22.g5 hxg5 22 . . . �a l t?! has never been played, and after 23 J::i d l �aS t 24.c3 the change in position helps White as Black is deprived of the check on a I , which can be an important resource in some lines. 22 . . . hS?! This second untested move is also relatively easy to refute. 23 . .txhS �a l t 23 . . . g6 24 . .td l ± 23 . . . f5 24.gxf6 �xhS gives White several ways to win, for instance 2 S . fXg7 +­ intending 0-0. 24JM I 24.�f2? �xh l 2 S . .txf7t does not work due to 2S . . . �xf7 26.�f4t �g8 27.g6 eS 28 .�f7t �h8 when there is no mate and Black wins. 24 . .td l ! ? is interesting, but the text move is more forcing. 8

turns out that 23 . . . g4! 24.hS tt:l f6! is a fairly straightforward refutation. After 2S.tt:lxf6t gxf6 White still has some compensation, but his attacking prospects are virtually nil. We now come to a final dividing point, where Black has tried 01) 23 ... g6?, 02) 23 ... tLlf6!? and 03) 23 ... f6.

Dl) 23 ... g6?

a

b

e

d

e

f

g

h

This allows White to continue the attack unhindered.

7

24.c3!

6

24.l'l:xgS may also be good, but I like the logic of preventing the queen from assisting in the defence via a l -h8.

5 4 3

24 ... �g7

2

a

b

e

d

e

f

g

h

24 . . . �aS t 2 S . c3 f5 26.0-0!? tt:l ceS 26 . . . tt:l deS? 27. tt:l f6t! wins. 27. tt:l d2±

23.ggl ! 23.h4?! is more romantic, and I spent considerable time analysing it in 2006, but it

24 . . . �b6 2 S . tt:l f6t �g7 26.tt:lhSt 1-0 was the quick end to the game Kozma2 - Bereza, Internet 20 1 3 .

25.gxg5 Y*ib6 26.Y*if4 e5 27.Y*if3 I found one internet game where Black gave up in this position, and another where he stumbled on for a few more moves.

279

Chapter 1 5 - 20 . . . �d8

28 .'lWd3 e5 29 .ig5 +-) 27.E!:h5! E!:xc7 28 .'lWh3 Now Black is forced to run with 28 . . . �f8, but after 29.�xd2 he faces considerable trouble with his poor development and vulnerable king.

2SJ�Ml I briefly checked 2 5 .id l ? CLlxe4 26.ie5, but it is refuted by 26 . . . 'lWxd l t 27.E!:xd l E!:xd l t 28.�xd l CLl xg5 29.'lWxg5 CLl xe5 30.'lWxe5 id7 with a winning endgame for Black. a

b

e

d

e

f

g

h

27 W1bl t 2S.Adl W1cl 29JUS W1xd2t 30.ci!?xd2 gx5 31 .W1g3t ci!?hS 32.W1h4t ci!?g7 33.W1gSt ••.

Black resigned in misterija - chesspernik5 80, Internet 20 1 3 .

D2) 23 .. .tl�f6!? 8 7

b

a

6

e

d

f

e

g

h

2S ...W1xdlt

5

Another game continued: 25 . . . CLlxe4 26.E!:xa l CLl xg5 27.ic7! (27.E!:d l f600) 27 . . . E!:d7 (27 . . . �d5 28.c4 �d7 29.'lWxg5 E!:xc7 30.E!:a3!-t) 28.'lWxg5 E!:xc7

4 3 2 1 a

b

e

d

e

f

g

h

At first, this move seems to enable Black to simplify things considerably, but White ends up winning the enemy queen for some pieces while keeping a raging initiative, as demonstrated in several games.

24.�xgS WIal t No better is: 24 . . . CLlxe4 25 J::\ x a5 CLl xd2 26.ic7 E!:d7 (26 . . . CLlxa5 27.ixd8 CLl ac4

a

b

e

d

e

f

g

h

29.E!:a3! White quickly brings his last piece into the attack. 29 . . . f5 30.E!:g3 b5 3 1 .'lWg6 E!:e7

280

Poisoned Pawn

32J'%h3 1 -0 Szczepankiewicz - Kolodziejski, corr. 20 1 1 .

2B . . . e5 29.h4 also gives White a strong attack.

26.�xdl!

29 ..th5 �f5 30.YlYe5 gd7 31 ..tg4 gd5 32.YlY£6 �f8 33.YlYc3 g6 34 ..tf3 gb5 35.h4-+

In the stem game where I introduced 2 1 .g4!!, I played 26.�xd 1 which was also quite dangerous for Black. After 26 . . . lLlxe4 27.'lWxe4 :B:xd6 2B .h4 f6 29.:B:g3

Black did not manage to find a suitable defence in Ulbig - Fortune, corr. 20 1 3 .

D3) 23 ... £6 8 7 6 5 4 3 2 b

a

e

d

e

f

g

h

I eventually won a crazy game in Negi Huschenbeth, Chotowa 20 1 0, but the text move gives White a better version.

26 ... tihe4 27.YlYxe4 gxd6t 28.�c1 Compared with the previous note, White's king can help to block the a-pawn, and the bishop is well placed to j oin in the attack via d3 or h5.

a

b

e

d

e

f

g

h

This is the most critical variation for the assessment of White's 2 1 st move. All the other lines were unpleasant for Black, and most of White's moves were direct and easy to understand. Now he must make an important decision.

24.h4!?

8 7 6 5 4 3 2 1 a

28 ... �d4

b

c

d

e

f

g

h

My original intention was: 24.lLlxg5 ! ? Unfortunately, some correspondence guys found an unbelievable forcing line, where White does not appear to have more than a draw. I will present the game with a heavily shortened version of my notes. 24 . . . lLlc5! 2 5 .c3!? 2 5 . lLl f7 'iit xf7 26.:B:xg7t cj;Jxg7 27.'lWg3t 'iit h B 2B .'lWh4t cj;JgB 29 .'lWg4t cj;JhB was agreed drawn in Berg - Williamson, corr. 20 1 3 . 2 5 . . . :B:xd6 26.:B:xd6 lLl e 5 27.:B:d4!?

Chapter 1 5 27.ttJh3!? ttJ f7 2B.E!:d l intending ttJ f4-h5 deserves further attention, but I think Black should be okay. 27 . . . id7 2B.ttJf3 ttJxf3t 29.'lWxf3 ic6 30 .'lWh3! 30.E!:xg7t leads to a forced draw. I also analysed 30.'lWe3 and 30 .'lWg3 , but in both cases Black can hold with precise defence. 30 . . . ie4! 3 1 .E!:g3! Intending 'lWh5 and E!:h3 . In the key correspondence game Black found what is probably the only way to defend, which even the computer initially fails to give as its top choice. 8

-

20 . . .l:!d8

2B l

45 .'lWdB 'lWf3t 46.\!
D32) 24 �de5.

7

•••

6

03 1) 24... �fB 25.c3 Ei:xd6

5

25 . . . ttJxg6 transposes to variation 032 below, but the text move is an important independent possibility.

4 3 2

26.�xd6 �e5 a

b

e

d

e

f

g

h

3 1 . . .if5! 32.'lWh5 'lWb6! 33 .E!:b4 'lWc6 34.E!:h4 'lWh 1 t 3 5 .ifl g5! 36.'lWhBt \!
4

b

e

d

e

f

g

h

27.�h3!

3

Preventing any . . . 'lWh2 j umps, and preparing to attack along the h-file.

2

a

b

e

d

e

f

g

h

42.E!:h3 g4 43.E!:h7t \!
27... 5 After 27 . . . ttJ h7 2B.hxg5 ttJ xg5 29.'lWh5! Black's position collapses.

Poisoned Pawn

282

28.hxg5

31 .g6t!?

This seems like the natural choice, preparing a future g5-g6.

There are other possibilities, but this seems the most tempting. At several of the following moves, White has different options, but the main correspondence game seems good enough - right up to the moment where I found a clear improvement.

28Jhg5 is an interesting alternative. The critical line continues: 28 . . . .td7 29.�d3 l:iJ e7 30.h5 I:iJ h7! 3 1 J!gl .tb5 ! Taking away the c4-square from the knight. 32.c4 .tc6 White seems to have the initiative, but I couldn't find anything concrete.

28 ....td7 The computer's "0.00" is deceptive, as Black is barely hanging on. White has multiple ways to create problems, which at the very least make Black's position practically hard to play. Moreover, having dug a bit deeper, I believe Black's position to be problematic even after correct defence.

31 ...�xg6 32.«!i::J d6 �f6 33.gfl �h6 34.�g3 g6 35.�dl!? Removing the king from various checks. It is hard to see how Black is going to improve his position, while White has a lot of possibilities : 3 5 . .td l ! ?N could also be considered, with the idea of deploying the bishop on b3.

35 ... e5 Trying to get some counterplay. 35 ... We7 36.l:iJxb7 leaves Black under pressure.

29.ti)c4 By driving the queen away, White takes away . . . g6 as means of covering the h8-square.

36 ..tc4 f4

29.�h 1 g6 30.l:iJc4 V!1g7 3 l .'kt>f2 is another interesting position that could be explored further. The queen seems rather stuck on g7, but White's initiative has been neutralized for now, so the position seems balanced.

29 ...�f4 30.ghl �f7 8 7 a

6

b

e

d

e

f

g

h

37.�cl!N

5 4 3 2 1 a

b

e

d

e

f

g

h

37J!h2 reduced the pressure on Black, and after 37 . . .'tie7 38 .V!1f2 V!1xh2! 39.V!1xh2 Wxd6 he was able to hold in Simmelink - Boehme, Internet 20 1 3 . After the text move Black is rather stuck; he has little counterplay and will find it tough to survive from move to move.

Chapter 1 5

37 <.!.>g7 .•.

37 . . . <.!.>e7 38.Wfgl ! is excellent for White. 37 . . . Wfh3 3 8 . ttJ e4t! avoids the queen exchange, and after 38 . . . cj,Jg7 39 .Wfe l Black must worry about :B:h 1 .

38.Wffl 38.Wfe l is similar. In either case, Black's position requires computer accuracy to defend. A final few brief lines:

-

283

20 E!:d8 . . .

2 5 . . . g4? is refuted by 26.ttJxf6t! 27 . .ixg4 with a winning attack.

gxf6

25 . . . :B:xd6! ? has been played only once, but deserves further attention. 26.ttJxd6 (After 26.:B:xd6 g4! 27 ..ixg4 fS! 28 . .ixfS exfS 29.:B:xg7t cj,Jxg7 30.Wfg5 t cj,Jf7 White only has a perpetual.) 26 . . . g4 8 7

38 .. .'IWh3 38 . . . Wfg5 ? 39.:B:g l Wfe7 40.Wfh2 wins.

6 5 4

39.:B:d3 �h6 39 . . . Wfh5 40 . .id5 ±

3 2

40.:B:fdl 40 ..id5 g5! attacks the knight on d6, but the text move maintains White's initiative.

D32) 24 ... c!Llde5

a

b

e

d

e

f

g

h

a) 27.h5 .id7! 28.h6 g6 29.ttJxb7 (29 . .ixg4 cj,Jh7!oo; 29.h7t! ? cj,Jh8°o) 29 . . . Wic7 30.ttJc5 .ie8 3 1 ..ixg4 cj,Jh8 32 . .ie2 This insane position was reached in Simmelink - Wurtinger, corr. 20 1 3 . White is materially ahead, but it is hard to be certain about anything without much deeper analysis. The game eventually ended in a draw after a long fight. b) 27.Lg4!?N ttJ xg4 28.:B:xg4 Wfe5 29.Wfxe5 (29.:B:e4!? Wih5 looks extremely dangerous for Black, but it's not easy to prove anything against the computer.) 29 . . . ttJ xe5 8

a

b

e

d

e

f

g

h

25.c3 25.hxg5 ? is insufficient due to 25 . . . f5 ! 26.ttJ f6t (or 26.c3 fXe4) 2 6 . . . cj,Jf7 when there is no mate. White can try the tricky 27.c3 ! ? , but Black keeps a big advantage with 27 . . . ttJ g6! intending . . . :B:xd6 and . . . Wfe5 .

25 ... c!Llg6

7 6 5 4 3 2

a

b

e

d

e

f

g

h

Poisoned Pawn

284

30.E:g3 We have reached a pretty one-sided endgame. Materially, Black is more than okay, but White's rooks are extremely active, making it hard for Black to hold on to all those pawns. He certainly has drawing chances, but he will need to play accurately.

3 1 . . . ttl e7 3 1 . . .'Wa4 32 ..tf3 'Wh4t 33.�g2 f4 34.'Wc5t �f7 3 5 . .txc6 'Wxg5t 36.�f1 'Wxf6 37 . .te4 �eB 3B .E:e 1 maintains heavy pressure. 32 . .te2 ttl f7 8 7 6 5 4 3 2

a

a

b

e

d

e

f

g

h

26.hxg5 f5 27 ..th5! 27 .E:h 1 is an interesting option, but Black has far too many options to draw any conclusions.

27 ... f4!? This seems like a critical line to consider. However, if my analysis proves correct, it is not objectively best. 27 .. .'tif7? 2 8 . ttl c5 is j ust bad. 27 . . . E:xd6 2B.E:xd6 ttl ge5 Black's play is not terribly logical, and is quite easy to refute, despite the computer's initial insistence that Black is okay. 2B . . . ttl ge7? is certainly not an improvement: 2 9 . ttl f6t �fB 30.�f2 e5 3 1 .E:gd l ! +A better idea is 2B . . . fXe4, when 29 . .txg6 transposes to 27 . . . fXe4 below. 29.ttl f6t �f8 30.�f2! g6 Both 30 . . . E:bB 3 1 .E:gd 1 ! and 30 . . . b6 3 1 .'Wd2! give Black big problems as well. 3 1 .E:dd l ! With the knight on f6, Black's position is ready to fall apart.

b

e

d

e

f

g

h

33.E:h 1 ! e5 33 ... 'Wc7 34.E:h4 e5 3 5 .E:c4!± 34 ..tc4 f4 3 5 .'We4 'Wc5 t 36.�e2 ttl f5 37.E:d7!± 27 . . . fXe4! is objectively best, only move to survive. At the voluntarily enters a slightly so White should not have 2B ..txg6 E:xd6 29.E:xd6 ttle5

and possibly the same time, Black worse endgame, any complaints. 30 . .txe4 ttl c4

8 7 6 5 4 3 2

a

b

e

d

e

f

g

h

3 1 ..th7t! A nice touch. 3 1 . . .�fB (3 1 . . .�xh7? 32.'Wd3h-) 32.'Wg3 'We5t 33 .'Wxe5 ttlxe5 34 ..te4 �e7 3 5 .E:d4 White kept the better chances in Lj ubicic - Muller, corr. 20 1 2. In the game Black was able to hold a draw with precise defence, but over the board this would not be at all easy.

Chapter 1 5

b

a

e

d

f

e

g

-

h

28.c!tlf6t!N

28.ixf4 l:'i:xd2 29 .ixg6 c!tl e7! 30 .ih7t was agreed drawn in Malashenkov - Noth, corr. 20 1 4 . After 30 . . . �xh7 3 1 .g6t �g8 32 .'lWxd2 'lWd5 the position looks quite dangerous for Black, but it's hard to prove anything. I also considered 28 .'lWe2!?N 'lWf5 with a strange mess, as well as 28 .'lWf3 'lWfS 29.ixg6 'lWxg6 30.'lWxf4 e5 3 1 .'lWf3 if5 with equality, before ascertaining that the main line was the best.

285

20 . . j::� d 8

a

b

e

d

e

f

g

h

30.'ilYhl !!

Can a move be both ridiculous and brilliant at the same time?

30 ... f5 30 . . . 'lWe3t 3 1 .�d l l:'i:xd6 32.l:'i:xd6 'lWxc3 33.gxf6 �f8 (33 . . . 'lWa l t 34.�e2+-) 34.ixg6 'lWxf6 3 5 .ie4 �e7 36.l:'i:d2+-

3 1 ..hg6 gd7 32.hf4 32.l:'i:h2? 'lWe3t 33.�fl 'lWc l t 34.�f2 'lWb2t is only a draw.

28 ... gxf6 29.'lWh3! 'lWb6 Another nice variation is: 29 . . . e5 30 .'lWd3 f5 3 1 .ixg6! ie6 8 7 6 5 4 3 2

a a

b

e

d

e

f

g

h

32.ih7t! �xh7 33 .'lWh3t �g8 34.l:'i:h l 'lWa l t 3 5 .l:'i:d l 'lWxd l t 36.�xd l l:'i:xd6t 37.�e l ±

b

e

d

e

f

g

h

32 ...'ilYbl t 32 . . . l:'i:g7 33 .'lWh2! l:'i:xg6 34.l:'i:h l leaves Black without a defence.

33.�f2!

Poisoned Pawn

286

33.�e2? vgb5t 34.�f2 vgc5 t 3 5 .�e2 vgb5t is a draw.

33 ...'We4

36 . .ih7t <.!?g7 37 ..ie5t tll xe5 38.!!xd8 b5 39.!!g8t <.!?f7 4OJ�al ± White should win accurate play.

the endgame with

Conclusion

a

b

e

d

e

f

g

h

34.i.d6!! The analysis is stretching a bit too far, but I cannot resist putting in the complete variation. 34.vgxe4 fxe4 3 5 .i.xe4 e5 should be about equal.

34 ...'Wxhl 34 . . . ltJe5 3 5 .vgxe4 fxe4 36.i.xe5 !!xd2t 37.�e3 !!d5 38.i.f6±

35J�xhl !!d8 Neither 35 . . . !!g7 36.!!h6± nor 35 . . . !!a7 36.�g3 ± help Black's cause.

a

b

e

d

e

f

g

h

From the chapter's starting position after 20 . . . !!d8, I am pleased that my attacking plan involving 2 1 .g4!!, which I discovered in 2006 and introduced as a novelty in 20 1 0, is still going strong. We started by analysing three sidelines, all of which are playable, while at the same time suffering from clear defects. Black's most popular and reliable move is 2 1 . . .ltJ c6, after which 22.g5 hxg5 23.!!g l ! leads to another branching point, where 23 . . . f6 24.h4!? leads to the most complicated positions of the chapter. Both 24 . . . ltJ f8 and 24 . . . ltJde5 lead to the same type of story: Black's position cannot be refuted outright, but he is constantly in danger. The main challenge for White is to avoid certain forcing lines which might be analysed all the way out to a draw, which seems a bit crazy, but it could happen nowadays. I am satisfied that I have accomplished this goal. I cannot promise that my recommendations will break the defences of correspondence players armed with computers, but I am confident that they will cause problems to anyone who takes up Black's position over the board.

12 g5 ...

a

b

e

d

e

f

g

13.if2!r Variation Index l.e4 cS 2.� a d6 3.d4 cxd4 4.�xd4 � f6 5.� c3 a6 6.i.g5 e6 7.f4 �b6 8.�d2 �xb2 9J�b l �a3 1 0.e5 h6 1 1 .i.h4 dxe5 1 2.fxe5 g5 1 3.i.f2!? tD g4 14.i.g3 A) 14 ...i.g7!? 1 5.� e4 A I ) 1 5 ...�e3t!?N A2) 1 5 ...0-0 B) 14...�d7 15.i.e2 � gxe5 B l ) 1 6.�xe6!? B2) 1 6.0-0 i.g7 17.�xe6! fxe6 1 8.� e4 B2 1 ) 1 8...� f7!?N B22) 1 8..J3f8 1 9.c4!!N B22 1 ) 1 9 ...b6 B222) 1 9 ...b5!

A2) after 1 9 . . . lihf6

288 290 292 294 295 298 299 303 304 306

822) after 1 8 . . J:'l fB

8 1 ) after 1 7 . . . <;!Ie7

8

8

8

7

7

7

6

6

6

5

5

5

4

4

4

3

3

3

2

2

2

a

b

c

d

e

f

2o.lihf6t!N

g

h

a

b

c

d

e

f

1 8 .'�e3!?N

g

h

a

b

c

d

e

1 9.c4!!N

f

g

h

h

288

1 2 gS . . .

1 .e4 c5 2.C!l�f3 d6 3.d4 cxd4 4.tlhd4 tLlf6 5.tLlc3 a6 6.J.g5 e6 7.f4 VlYb6 8.VlYd2 VlYxb2 9.gb 1 VlYa3 10.e5 h6 1 l .J.h4 dxe5 12.fxe5 g5

a

b

e

d

e

f

g

h

The biggest theoretical debate going on in the entire 6.i.gS Naj dorf has been centred around this variation. Black has done surprisingly well in the games here, with ambitious White players often seeing the tables turned. My aim is to demonstrate some interesting tries which have not yet been analysed out to a finish, and those in which I feel Black will not have the easy life that he gets in many of the main lines. 1 3 .exf6 has been the overwhelmingly most popular move, and we will cover it in the final two chapters of the book. Before then, however, I would like to draw your attention to a rare move which I believe has been unfairly neglected.

will be to prove something concrete for his two-pawn investment. 1 3 .i.g3 is the poor cousin of the text move, and Black is doing rather well after: 1 3 . . . tLlhS 1 4. tLl e4 (Trying to keep the bishop alive with 1 4.i.fL. is pointless, since a future . . . tLl f4 jump will be annoying: 14 . . . tLld7! I S .i.e2 tLl f4 1 6. 0-0 i.g7 1 7. tLl e4 0-0+) 1 4 . . . 'lWxa2 I S J::!: b 3 ( 1 S J::!: d l is usually the ideal move, but here after I S . . . 'IW dS Black will most likely be able to eliminate both the dark-squared bishop and the eS-pawn.) I S . . . tLl d7 1 6.'lWc3 b6! 1 7.tLld6t i.xd6 1 8 .exd6 tLl xg3 1 9 .tLlxe6 'lWaS 20.tLlc7t �d8 2 1 .l::!: b 4 'lWeS t 22 .'lWxeS tLlxeS= Revuelta Capablanca - Biedermann, corr. 20 1 1 .

13 tLlg4 14.J.g3 ••.

8 7 6 5 4 3 2 1 a

b

e

d

e

f

g

h

White is threatening to trap the knight, so Black only has two serious options: A) 14 .ig7!? and B) 14 ... tLld7. •..

13.J.fZ!? This move has been played in barely a dozen games out of nearly four thousand on my database! Superficially it may seem that White is taking too much time, but there is definite logic behind the decision to preserve the dark­ squared bishop. Black will soon be forced to capture the eS-pawn, leaving all of the central files open. Obviously the challenge for White

A) 14 ....ig7!? This move has only occurred in a single intern et game. It has the advantage of avoiding the tLlxe6 ideas that may occur after the knight goes to d7, and it is also useful for Black to have the option of castling at a moment's notice.

Chapter 1 6 - 1 3 .if2!?

1 5.�e4 I s .ibSt!?N I also anaiysed this direct attempt, but did not find anything clear. I will present a shortened version of my analysis which you can use as a starting point for your own investigation should you wish. I S . . . lLld7! I S . . . axbS ? 1 6. lLl dxbS works out badly for Black. I s . . . id7 is met by the clever 1 6.ie2! lLlxeS 1 7.lLle4, when the bishop is misplaced on d7. Now the knight is unable to go to d7 to prevent lLl f6t, and the b7-pawn may be left hanging in some lines. 1 6.ixd7t ixd7 1 7.E:b3 WfaS 1 8 J::!:xb 7 lLlxeS 1 9 .0-0 E:c8 Mter 19 . . . 0-0 20.lLlb3 lLl c4 2 1 .Wfxd7 Wfxc3 22.E:xf7 E:xf7 23 .Wfxf7t Wh8 24.Wfxe6 Wfxc2 2S .h3t White has chances to develop a serious initiative.

a

b

e

d

e

f

g

289

b

a

e

d

e

f

g

h

29 . . . E:xc2t! If it were not for this resource, Black could resign. 30.Wfxc2 Wfd4t 3 1 .We2 WfeS t 32.Wd2 Wfd4t= There is no way to escape the checks. It is no use moving the king to cl in order to block a check with the b7-rook, as this would release the pin on the f7 -pawn. Let's now return to l S .lLl e4.

h

20.lLlfS! Black must play accurately to survive. 20 . . . exfS 2 1 .ixeS 0-0 22.ixg7 Wxg7 23 .Wfd4t �h7! 24.Wfxd7 E:xc3 2 S . E:xfS Wfxa2 26.E:f6!? A nasry idea, but Black is j ust in time with a brilliant defence. 26 . . . �g8 27.E:g6t Wh7 28 .WffS Wfa l t 29.Wf2

a

b

e

d

Black can try either

A2) 1 5 0-0.

e

f

g

h

AI) 1 5 'We3t!?N or •••

•••

I S . . . lLl xeSN 1 6.E:b3 Wfe7 1 7. lLl f3 gives White a strong initiative.

290

1 2 g5 . . .

AI) I S ...Yfe3t!?N It is easy to ignore this move when analysing with the computer, as it will tell you that castling is stronger, but any human player can understand the logic for Black in exchanging queens. It is tempting to assume that Black must be completely okay, but White can still put his activity to use.

16.Yfxe3 �xe3 17.�d6t q;e7 18 ..ie2! 1 8 .lLlxb7?! f5 ! favours Black.

1 9 . . . lLlxg2!? This would be a hard move to make, but Black is surviving on the edge here. 2o.lLlxc8t l"laxc8 2 1 .l"lxb7 lLl f4 22 . .1f3 lLlg6 8 7 6 5 4 3

1 8 . .1d3 is nothing special: 1 8 . . . lLl d7 1 9.'it>e2 lLl d 5 ! ( 1 9 . . . lLl g4 might seem more natural, but 20.l"lhfl lLl gxe5 2 1 . .1e4! gives White excellent compensation.) 20.lLlxc8t l"lhxc8 2 1 .l"lxb7 lLl c3t 22.'it>d2 lLl a4 Black intends . . . lLl ac5 with a comfortable position. 8 7 6 5 4 3

2

a

b

e

d

e

f

g

h

23.l"ld l 23 . .1c6 l"lhd8 24 . .1a4 lLl gxe5 25 . .1xe5 .1xe5 26.lLlc6t l"lxc6 27 . .1xc6 'it>d6 28 . .1f3 'it>e7 29.l"ld l l"lc8 (intending . . . l"lc7) leads to a similar type of endgame where White has the exchange for two pawns. The main line seems like a slightly favourable version though. 23 . . . lLl gxe5 24.'it>c l ! 24.'it>e2 lLl xf3 2 5 . lLl c6t 'it>f6 26.l"ldxd7 l"lhf8 27 . .1d6 l"lxc6 28 ..1xf8 .1xf8 29.l"lxf7t 'it,lg6 30 .l"lxf3 .1d6 is equal. 24 . . . lLlxf3 2 5 . lLl xf3 l"lhd8

2 8

1 a

b

e

d

e

f

g

h

18 ... �d7 If 1 8 . . . lLl xg2t 1 9 .'it>f2 lLl f4 20.lLlxc8t l"lxc8 2 1 .l"lxb7t 'it>f8 22 . .1xf4 gxf4 23 . .1h5 White maintains some pressure.

7 6 5 4 3 2

19.q;d2 In the event of 1 9 .'it>f2 lLl d5 20.lLlxc8t l"lhxc8 2 1 .l"lxb 7, Black equalizes with 2 1 . . . lLl c3! intending either ... l"lab8 or ... lLl e4-c5.

19 .. .tt:ldS

a

b

e

d

e

f

g

h

26 . .1c7 'it>e8 27 . .1xd8 l"lxd8 28.h3 I prefer White's chances in this complicated endgame. Materially Black is fine, but he has

Chapter 1 6 - 1 3 .i.f2! ? some coordination problems and White's rooks are ext�emely active. 8 7

24.ha6 gxb7 25 .L:b7 ge7 26.tLle6t �e8 27.gbl tile5 •

After 27 . . . tLl xe5 2 8 . tLl xe5 i.xe5 29.a4 White's pawns seem more dangerous, for instance: 29 . . . �e7 30.gb3 i.d4 3 1 .�d3 i.e3

6

8

5

7

4

6

3

5

2

4 3

1 a

b

e

d

e

f

g

2

h

20.tihe8t ghxe8 21 .gxb7 tLlf4 2 1 . . .�e8?! is unsatisfactory due to 22.i.h5!, while 2 1 . . .:gab8 22.:gxb8 tLlxb8 23.:gb 1 tLl d7 24.:gb7 also preserves a plus for White.

22.Lf4 22.i.f3 looks tempting, but it allows 22 . . . :gc4! 23.c;;t> e3 :gac8 when Black regroups.

b

a

e

d

e

f

g

h

32.:gb5! (32.a5 :gd7t 33.�c4 :gc7t=) 32 ... :gd7t 33.�c4;!; Black should probably hold, but he still needs to be careful.

28.i.a8 ge8 8

22,..gxf4 23.gdl gab8 23 . . . i.xe5 24.�c1 :gab8 2 5 . tLl c6t :gxc6 26.:gdxd7t �f6 27.:gxf7t �g5 28.:gxb8 i.xb8 29.c;;t> d 2 a5 30.i.d3 i.e5 3 1 .�e2 reaches a one-sided endgame. Despite the considerable simplifications and opposite-coloured bishops, Black is still a long way from securing a draw.

7 6 5 4 3 2

8

1

7

a

6

b

e

d

e

f

g

h

29.gb5! gxa8 30J�xe5 gc8!?

5

After 30 . . . :gxa2 3 1 . tLl b4 Black should probably hold, but White can keep playing on by targeting the f4-pawn.

4 3 2

29 1

3 1 .ge4 'it>d7 32.tila5 he5 33.gxe8 'it>xe8 34.'it>d3

. , =0-//" ////'

1 a

b

e

d

e

f

g

h

292

1 2 . . . g5

Black's kingside pawn mass is not doing much for now, and White's knight can j ump around and create weaknesses. Obviously there is a fair chance of a draw, but it is clearly Black who will need to work harder to achieve it.

19.exf6 �xf6

A2) 1 5 ... 0-0 8 7 6 5 a

4

b

e

d

e

f

g

h

20.�xf6t!N

3

This was played in the one game involving 14 . . . .tg7. It looks dangerous to castle having already weakened the kingside, but it is not easy to prove anything for White.

20.�d6 runs into the annoying 20 .. :�Va5t!, when White is more or less forced to exchange queens: 2 1 .'(Nd2N (2 1 .�d2? lLl c6 left White with insufficient play in 'Vigon777' 'Otshelnik' , Internet 20 1 2.) 2 1 . . .'(Nxd2t 22.�xd2 Although White has decent compensation for the two-pawn deficit, it seems hard to imagine him converting it into more than equality.

16.h4!

20 ....ixf6 21 .Yfd3!

Of course, otherwise there is no way to question Black's last move.

Preventing Black from capturing on d4 due to the check on g6. We are left with an interesting position where Black has no immediate way to kill the initiative.

2 1 a

b

e

d

e

f

g

h

16 ...Yfxa2! Black has to play concretely, otherwise White will get an overwhelming initiative. 1 6 . . . '(Ne3t 1 9 .�h5;!;

1 7.'(Nxe3

lLl xe3

1 8 .hxg5

hxg5

1 6 . . . �d8 is too slow. 1 7.hxg5 lLl c6 1 8 .c3t Also after 1 6 . . . 5 1 7.exf6 lLl xf6 1 8 .lLlxf6t .txf6 1 9 .c3� White has good attacking chances.

17.�dl Yfd5 1 8.Yfe2 5! Again, the most concrete method is required.

2 1 .. .tg7 •

This natural move is safest. 2 1 . . .c;!;>g7 22 ..td6! gives White good prospects. 2 1 . . .e5 22.hxg5 ! exd4 23.gxf6 �xf6 24 . .te2 '(Na5 t 2 5 .c3 gives White an ongoing initiative. 2 1 . . . lLl d7 is also risky for Black: 22 .'(Ng6t .tg7 23 . .td3 �f7 (23 . . . �f6 24 .'(Nh7t �f8 25 . .te4 '(Na5 t 26.c;!;>e2!±) 24 . lLl f3 gxh4 25.�xh4

Chapter 1 6 - 1 3 .i.f2! ?

293

22 �xg5 23 ..id6 E:f7

8

•••

23 . . . E:dB 24.lLlf3 Wf6 2 5 .We4 Wc3t 26.\t>e2! reaches a weird position with good chances for White, despite the misplaced king. The immediate threat is i.e5 .

7 6 5 4

8

3

7

2

6 b

a

e

d

e

f

g

5

h

25 . . . lLl fB 26.Wg4 e5 27.Wh5 Wa5t 2B.\t>f1 i.e6 29.\t>g l � White's king problem has been solved and his pieces remain extremely active.

4 3 2 b

a

e

d

e

f

g

h

26 . . . Wc6 (26 . . . lLl c6 defends the rook, but is met by 27.i.e5! all the same. Black's lack of development is starkly visible after 27 . . . i.xe5 2B.E:xdBt lLl xdB 29.lLlxe5 when there is no defence.) 27.Wf4 (27.Wd3 E:xd6 2B .Wxd6 Wxd6 29.E:xd6 lLl c6=) 27 . . . Wxc2t 2B .E:d2 Wc3 29 .Wg3 (29.E:d3 Wc2t=) The position remains extremely complicated, but White is not the one who is taking the risks. a

b

e

d

e

f

g

h

22.hxg5!?

This is the most ambitious continuation. There are many interesting lines ahead, but considering how unexplored this whole variation is (remember 14 . . . i.g7 has only been played in one obscure Internet game) , I will not analyse it too exhaustively. Instead I will present a few variations to show that White is at least not objectively worse, while highlighting what I consider the more promising possibilities for future analysis and testing. It is worth mentioning that 22.Wg6 should lead to a draw: 22 . . . E:f6 (22 . . . Wa5 t 23.E:d2 Wal t 24.\t>e2 Wa5 might also lead to equality.) 23 .WeBt E:fB 24.Wg6=

a

b

e

d

e

f

g

h

24.tLlB �a5t 24 . . . Wf5 2 5 .We3 Wa5t 26.\t>f2 lLl d7 27.i.d3 White has dangerous compensation,

294

1 2 . . . g5

with a comparatively safe king and a huge lead in development.

Both 26 . . . lLl f6 27.'it>gl and 26 . . . Wd5 27.We2 maintain White's initiative.

25.�fl tLl d7

26 . . . Wb6t 27.'it>g3 e5 (27 . . . lLl f6 28.lLle5±) 28 .'it>h2! also seems quite promising.

25 . . . Wb6t is met by 26.'it>g3! when the king is completely safe. 2 5 . . . .id7 26.l'!h4 .ic6 27.'it>gl lLl d7 2 8 . lLl d4 maintains White's initiative. 2 5 . . . lLl c6 26.Wg6! ? Wf5 (26 . . . lLl e7 27 . .ixe7 l'!xe7 28.l'!xh6t) 27.l'!xh6 Wxg6 28.l'!xg6 8

27J3f4 27.l'!g4?! runs into 27 . . . lLl f6! 28.l'!f4 Wb6t when it is not possible to put the king on g3 due to the fork on h 5 . However, the immediate 27.'it>g3!? could be considered.

27 l'!xf4 28.J.xf4i

7

•••

The posltlon remains exceptionally complicated, but I feel happy about White's prospects. At the very least, it would seem dangerous to play this way with Black in a practical game.

6 5 4 3 2

B) 14 tLld7 ..•

b

a

e

d

e

f

g

h

The endgame is interesting, but White is at least not worse due to his extreme piece activity. A possible continuation is 28 . . . a5 29 . .id3 a4 30 . .ia3! l'!a5 3 1 ..ie4 l'!d7 32.l'!h l with some initiative. 8 7 6 5

a

4

b

e

d

e

f

g

h

To date, this has occurred in all but one of the games in which 1 3 . .if2 was played.

3 2

15 ..ie2 lLlgxe5

1 a

b

26.l'!h4!? %¥d8

e

d

e

f

g

h

We have reached a critical crossroads. So far White's development has been logical, and the stage is set for various tactical ideas. White actually has a few interesting tries here, and I

295

Chapter 1 6 - 1 3 .�f2! ? decided to present both B2) 16.0-0 ,in detail.

BI) 16.�xe6!? and

There is even a third idea: 1 6.lLle4 �g7 1 7.h4!? Forcing some weakening on the kingside. Black also needs to react precisely after 1 7.lLlxe6!? fxe6 1 8 .�b3 �a4 1 9. 1Ll d6t @e7 20.0-0 �f8 2 1 .�d l , but after 2 1 . . .�c6! I found no more than a draw. 1 7 . . . g4! 1 7 . . . gxh4? is refuted spectacularly by: 1 8 .lLlxe6! fxe6 1 9.�b3 �a4 20.�h5t @d8 2 1 .�xh4t The point. 2 1 . . .@c7 22.�d6# 1 7 . . . lLl c5 seems rather shaky for Black after 1 8 .lLl d6t @f8 1 9.c3 lLl cd3t 20.�xd3 �xd6 2 1 . 0-0�.

1 8 . . . �e7 is the computer's suggestion, but after 1 9 . 0-0 lLl c5 20.lLlg3 ! things are far from easy for Black. 1 9 .1Llxf6t �xf6 20.0-0 Now 20 . . . �e7 2 1 .�bd l �g7 22.�e3t would be good for White, but instead Black can play: 8 7 6 5 4 3 2 b

a

8 7 6 5 4 3 2

a

b

e

d

e

f

g

h

1 8 .�f4N 1 8 .0-0 0-0 1 9 .�xe5 occurred in Staratorzhsky - Jandek, corr. 2009, and now 1 9 . . . �xe5!N leaves White with nothing better than forcing a draw: 20.lLlxe6 fxe6 2 1 .�xh6 �e7 22.�xg4 �xfl t 23.�xfl lLl f8! 24.�xf8t �xf8 2 5 .�xe6t �xe6 26.�xe6t �f7 27.�xe5 �f8= The text move prevents Black from castling. Initially I thought it was promising for White, but unfortunately Black has a strong defence: 1 8 . . . lLl f6! 18 . . . h5? is pointless, as 1 9. 0-0 0-0 20.�xe5! just wins.

e

d

e

f

g

h

20 . . . �c5 ! An incredibly annoying computer idea. It is hard to fathom the reasoning behind it, but the general idea is to misplace White's pieces a bit, before retreating to e7. It is possible to analyse this further, especially as the computer has a tendency to underestimate White's initiative in such positions, but I don't see any great need to do that right now since we have plenty of other interesting ideas to explore.

BI) 16.�xe6!? fxe6 17.i.h5t q;e7 8 7 6 5 4 3 2 1 a

b

e

d

e

f

g

h

1 2 . . . g5

296

White's piece sacrifice is the bluntest approach, and is not mentioned in either The Sharpest Sicilian 2012 or Andriasyan's Winning with the Najdorf. Compared with the variation B2 below, White has a few interesting ways to try and exploit the fact that the bishop has not yet gone to g7.

2 1 .�f3N 2 1 .�b3 Wfa4 22.�d l Wfxe4! 23 .Wfa5t b6 24.Wfxb6t 'it>e7 2 5 . �xd7t .ixd7 26.Wfc5 t was only good enough for a draw in 'grasuI2009' - 'guardio' , Internet 20 1 3 . 2 1 . . .Wfa4 2 1 . . .Wfxa2?! 22.�d l Wfa4 23.lDd6 IS becoming increasingly dangerous for Black.

1 8."e3!?N Another tricky continuation is: I B . .ixe 5 ! ? lD xe5 1 9. 0-0 lD d7! Intending to walk the king to dB and c7. After 1 9 . . . Wfc5t ? 20.'it>h l Wfc7 2 1 .�be l Black is collapsing. 1 9 . . . .ig7 does not lose instantly, but Black comes under heavy pressure after: 20.�bd l ! Wfc5 t 2 1 .'it>h l a5 (2 1 . . .Wfc7 leads to similar lines but gives White extra options, such as putting the knight on b5 instead of a4.) 22.lDa4!? (22 . lD e4 is also interesting.) 22 . . . Wfc7 23 .Wfe3 .id7 (23 . . . b 5 ? 24.Wfxe5! and mate follows.) 24. lD b6 �afB 2 5 . �fe l The initiative continues, and Black does not have an easy way to untangle. 20.lDe4 20.�f7t? j ust drives the king where it wants to go: 20 . . . 'it>dB 2 1 .�d l 'it>c7+ 8 7 6 5 4

8 7 6 5 4 3 2

a

b

e

d

e

f

g

h

22.�d l ! ? 22.�d3 .ie7 23.�d l �f8 24.lDd6 .ixd6 2 5 . �xd6 Wfb5 is nothing special, . so White should keep the first rook on the f-file. 22 . . . .ie7 22 . . . 'it>c7? 23.�xfB! +22 . . . e5 23.Wfd5 Wfb5 24.Wfe6± 22 . . . Wfc6 23.'it>h l .ie7 24.�f7! �f8 25 .�xe7 'it>xe7 26.Wfb4t lD c5 27.Wfc3! (27.lDxc5 Wfb5-+) 27 . . . e5 2B.lDxc5t 23.�f7 White is playing for the same idea as after 22 . . . Wfc6 above, but here Black has another defence available.

3 8

2

7 a

b

e

d

e

f

g

h

20 . . . 'it>dB! Now it is difficult to find an advantage, although White's compensation is enough to recover the material and maintain the balance.

6 5 4 3 2

a

b

e

d

e

f

g

h

297

Chapter 1 6 - 1 3.if2!? 23 . . . �b4! 23 . . . �xe4?. loses to 24.�a5 t . 24.�f2 24.c3?? allows 24 . . . �xe4 as the check on a5 is unavailable. 24 . . . mc7 25 J'M4 �b l t 26J!d l �b4= White has nothing better than to repeat moves.

23.i>h l ! (23.:1%0 :1%fBoo) 23 . . . e5 (23 . . . :1%fB 24.:1%xfBt ixfB 2 5 . llJ xg5! hxg5 26.�xe6+-) 24. llJ f6! J.xf6 2 5 .:1%xf6 i>c7 26.:1%0 With a strong attack.

20.�e4 �U8 2U�fdl! Intending the sneaky J. e 1 .

2 1 . .. �c4! Although the above line resulted in a draw, it was anything but boring. Now we return to I B .�e3!?N.

2 1 . . .�c7? is no good due to 22.�a3t.

22.'ilYe2 22.�d3 ?! leads to absolute mayhem; Black barely survives, but with accurate play he ends up with the better position: 22 . . . �d5 23 .�e2 J.d4t! (23 . . . �c6 24.J.d6t llJ xd6 2 5 . :1%xd6 �xd6 26.llJxd6 i>xd6 27.:1%d l t i>c7 2B.�xe6 i>bB 29 .J.eB;t) 24.i>h l 8 7 6 5 4

b

a

e

d

e

f

g

h

3 2

18 ...'ilYa5

Forced, as I B . . . J.g7? 1 9 .1lJd5t wins the queen.

19.0-0 Ag7 1 9 . . . i>dB is too slow here. 20.llJe4 (20.mh l ! ?) 20 . . . llJ c4 2 1 .�e2 b5 22.:1%bd l J.e7 (22 . . . �b6t 23.i>h l J.e7 24.:1%Ot)

a

b

e

d

e

f

g

h

24 . . . llJ e3 ! 2 5 .J.d6t (25 .:1%d3? �xe4-+) 2 5 ... i>dB 26.J.xfB �xe4 27.:1%d3 llJ xfB 2 B . c3 J.d7 29.J.f3 �f5 30.cxd4 J.b5 3 1 .:1%xb5 axb5 32.�xe3 g4+

22 ... h5!

8

Amazingly the only move again! (22 . . . �a4 does not lose instantly, but White's initiative seems overwhelming.) Now White does not have many options against the intended regrouping with . . . �b6.

7 6 5 4 3

23.
2

23 . . . J.b7? 24.llJd6!± a

b

e

d

e

f

g

h

1 2 . g5

298

.

24.�xc4 This is more or less forced. Black was threatening . . . Wfe3 , and if 24.�d6t lD xd6 2 5 .!l:xd6 Wfa7 26.1:'!xe6t i>xe6, Black's king is surprisingly safe, despite the possibility of a discovered/double check.

.

8 7 6 5 4

24 ... bxc4 25J:�xb6

3

8

2

7

1 a

6

b

e

d

e

f

g

h

29J'hd7t!

5

Without this sacrifice White would be worse in the endgame.

4 3

29 ... tLlxd7 30J�xe6t �d8 31.c!ll xa6 :5bl t 32.�h2 tLlb6 33.tLlc5

2 1 a

b

e

d

e

f

g

h

25 ...J.e5!

25 . . . lD xb6? 26.�d6t was White's idea, but the text move is enough to save Black. Now White must even be a bit careful.

Despite the simplifications, the position remains complicated but equal according to the computer. Black has an extra exchange, but his kingside pawns are weak and White may be able to develop threats against the king.

B2) 16.0-0

26.J.xe5! 26.�g4!? lDxb6 27.�xe5 is perhaps a bit too ambitious for White, and Black can at least give back the material with 27 . . . :55, when he is in no danger at all. After 26.:5b4 �xg3 27.lDxg3 :5a7 28.:5xc4 �b7 29.:5cd4 White has regained the missing pawn and the endgame should probably be a draw, but Black has a positional edge.

26 ... tLlxe5 27.h3 J.d7 28.tLlc5 �Hb8 a

16 ...J.g7

b

e

d

e

f

g

,

h

1 6 . . . �e7 is not so good in view of 1 7.lDxe6! fxe6 1 8 .�h5 t i>d8 1 9 .�xe5.

299

Chapter 1 6 - 1 3 .�f2! ?

17.ltJxe6! White should not allow the black king to get castled of course.

The first critical position, where White has many possible ways to continue. 8

17 ... fxe6 1S.ltJe4 In this position it is worth considering the untested B2 1) lS ... ttlf7!?N as well as B22)

lS .. JUS.

7 6 5

B2 1) lS ... �f7!?N

4 3 2 1 b

a

e

d

f

e

g

h

2 1 .�f7t This seems like the most dangerous. Certainly 2 1 .
a

b

e

d

e

f

g

h

This was mentioned briefly by Palliser in Chesspublishing. By threatening to castle, Black forces White to make immediate decisions rather than continue his attack on a slow burner. At the same time, it requires extreme precision for Black to pull this off, and even after perfect defence, it is not clear if he can fully equalize.

2 1 .Ei:fd 1 is the only move mentioned by Palliser. After 2 1 . . . .teS 22.ltlxc8 �cs t 23.
19.�d6! It is important to note that 1 9 .1tld6t ltlxd6 20 .�xd6 does not work, since after 20 .. .'�c3! the queen exchange cannot be avoided. 1 9 .Ei:b3 �a4 20.Ei:b4 �c6! forces matters, and after 2 1 .�f3 0-0 22.ltlf6t ltlxf6 23 .�xc6 bxc6 Black's three pieces seem quite enough to deal with the queen.

19 ... ltJxd6 20.ltJxd6t dS

8 7 6 5 4 3 2

a

b

e

d

e

f

g

h

1 2 . . . g5

300

22.ltJxb7t 22.Wfd3 ltJe5! 23.ltJxb7t 'kt>eB 24.ltJ d6t 'kt>dB is j ust a draw. 22.Wfd l Wfc5t 23.'kt>h l Wfd4 24.Wfxd4 .txd4 2 5 J:!: d l :B:fB 26.:B:xd4 :B:xf7 27.ltJxf7t 'kt>e7 2B.ltJxh6 ltJf6 29.ltJg4 b5 reaches a comfortable endgame for Black. 22 . . . 'kt>eB 23.ltJd6t 'kt>dB White obviously has a draw in his pocket, but can he achieve anything more? 8

26.:B:xfBt 26.:B:h7 Wfc5t 27.Wfxc5 ltJxc5 2B . .tB :B:xB 29.gxB .td7= 26 . . . WfxfB 27.Wfa5 t 'kt>e7 2B .Wfb4t 'kt>dB White has no more than a draw.

21 ...�e7 2 1 . . .'kt>eB is worse due to 22 . .th5 �c5t 23.'kt>h l :B:fB 24.:B:bd l ! �e7 25 .�d3 !± intending Wfh7. 8

7

7

6 5

6

4

5

3

4

2

3 a

b

e

d

e

f

g

h

24.Wfe3 24.Wfd3 ltJe5! transposes to 22.Wfd3 above and forces a draw. 24.Wfd l :B:a7! (24 . . . Wfc5 t 2 5 .'kt>h l Wfd4 26.Wfxd4 .txd4 27.:B:d U ;) Black's position looks incredibly dangerous, but somehow he is surviving: 25 .:B:B (25 . .tc4 .te5 26.ltJxcB 'kt>xcB 27 . .txe6 Wfd6=) 25 . . . g4 26.:B:d3 Wfc5t 27.'kt>h l .te5 2 B . ltJ f7t 'kt>e7 29.ltJxhB .txhB 30.Wfd2 .tg7 Black is doing rather well. 24 ... Wfxd6 2 5 .Wfxc3 :B:fB!

2 1 b

a

e

d

f

e

g

h

22J�b4! 22.ltJxhB .txhB favours Black.

22 ... �f6 22 . . . .te5 leads t o a n insane but relatively forcing variation: 23.ltJxe5 ltJxe5 24.:B:d l ltJf7 2 5 .:B:c4! :B:dB 26.:B:c7t 'kt>eB 8 7

8

6

7

5

6

4

5

3

4

2

3 2

a

a

b

e

d

e

f

g

h

b

e

d

e

f

g

h

27.WfxdBt! ltJxdB 2B . .th5 t ltJ f7 29.:B:xf7 �c5 t

Chapter 1 6 - 1 3 .if2!?

30 1

30.�h 1 Wfd5 ! 3 1 .E:xd5 exd5 32.E:h7t �d8 33.E:xh6;!; White keeps some pressure in the endgame.

23.�d6t
24
The king must dodge all sorts of mating nets here. One example is: 24 . . . �g8 2 5 .Wfd8t! �h7 26.id3t Wfxd3 27.Wfxd3t �g8 28.Wfg6+a

25.c3! Mter 25 .i.d3t ? Wfxd3 26.cxd3 E:fB+ Black is consolidating.

b

e

d

e

f

g

h

26.i.d3t

25 .E:b3 Wfa4 26.E:xf6t �xf6 27.E:f3t Wff4! is also okay for him.

26.E:xfG gives Black chances to force perpetual check: 26 . . . Wfc 1 t 27.�f2 Wfc2! 28.�e3 Wfc 1 t 29.�f3 i.d7 30.id3t �g8 3 1 .�e2 (3 1 .Wfxd7 Wfd l t 32.�f2 Wfd2t=) 3 1 . . .ixf6 32.Wfxd7 Wfxc3 33 .E:xb7 Wfe5 t =

25 ...
26
25 . . . E:fB is a smart defensive idea, but White has a strong answer: 26.i.d3t lD e4 27.i.xe4t E:f5 28 .E:b3! (28.g4 Wfxc3 29 .Wfxc3 i.xc3 30.E:c4 i.e5 3 1 .gxf5 t �f6 32.fxe6t i.f4 is rather drawish.)

a

b

e

d

e

f

g

h

29J:�d4! �f8! a

b

e

d

e

f

g

h

28 . . . Wfa4 (28 . . . Wfxa2 29 .E:b6 �h7 30.E:d6! Wfe2 3 1 .i.xf5 t exf5 32.Wff7! Intending Wfg6. 32 . . . ie6 33.E:xe6±) 29.E:e 1 ! g4 (29 . . . Wfd7 30.E:xb7 Wfxc7 3 1 .E:xc7+-) 30.i.xf5 t exf5 3 1 .E:b6t �h7 32.E:f6! Wfd7 33.E:f7 Wfxc7 34.E:xc7 �g6 3 5 .E:e8±

After 29 . . . id7 30.E:xd7 WffB 3 1 .Wfe5! White continues to dominate. Black can avoid an immediate loss with a few precise moves, but he will continue to suffer: 3 1 . . .E:e8 32.E:xb7 E:e7 33 .E:b8 E:e8 34.ic4 E:h7 3 5 .E:b7!? (35 .Wfxe6t E:xe6 36.ixe6t �h8 37.E:xfBt ixfB 3 8 .E:xfG ic5t 39.�f1 E:b7 40.ib3;!;) 35 . . . �h8 36.ib3�

302

1 2 . . . g5

33 ...h5

30J:�d8 .id7 3 I J�xffit gxffi

33 ... a5 34.Wfc7 a4 3 5 .c4 h5 36.c5 g4 37.h4 favours White.

Another critical position.

After 33 . . . .td7 34.Wfxa6 tLlg4 (34 . . . tLld5 3 5 .Wfd6!±) 3 5 .a4! the a-pawn may prove troublesome, for instance: 8 7 6 5 4 3 a

b

e

d

e

f

g

h

2

32.Wfxb7!

Even though Black has plenty of material for the queen, not to mention the computer's approval, I think his position remains quite tricky, especially in a practical game. 32J'hf6 makes Black's life easier, and after 32 . . . .txf6 33 .Wfxd7 1:'!f7 34.Wfxe6 mg7 he manages to regroup. The opposite-coloured bishops can still be used to create some pressure, but Black should enough resources.

b

a

e

d

e

f

g

h

3 5 . . . .te5?! (35 . . . 1:'!xfl t 36.mxfl is objectively better, although White keeps an edge here too.) 36.Wfb7! .txh2t 37.mh 1 1:'!xfl t 38 . .txfl ± The advancing a-pawn is more important that the small loss on the kingside.

32 .ic8 •••

32 . . . a5 33.gxf6 gives White an improved version of the above note. 32 . . . 1:'!d8 33 . .tg6!? (33.c4 tLl g4 34.Wfb6 1:'!c8 3 5 .mh 1 °o; 33 . .tc2 mm 34 . .ta4 me7°o) 33 . . . 1:'!f8 (33 . . . h5 34.Wfxa6 1:'!h6 3 5 .Wfa5 1:'!b8 36 ..td3 g4 37.a4±) 34.Wfxa6 White keeps an edge, and is ready to meet 34 . . . tLld5 with 3 5 .Wfb7!.

33.Wfc6

33 .Wfc7 leads to similar positions, but the queen seems better on c6 as she is ready to capture some pawns.

a

b

e

d

e

f

g

h

34.a4 a5 34 . . . h4 3 5 . .txa6 (35 .h3!?) 36.Wfxe6t mh7 37.Wfxa6±

35.Wfc7 g4 36 .ic4 •

36.h4! ? is also possible.

35 . . . .txa6

Chapter 1 6 - 1 3 .if1!?

36 h4 37.Wfxa5� •••

It will not be easy for Black to coordinate his forces and restrain the a-pawn.

B22) IS .. JUS

303

again. The main idea is to cut off the black queen from the defence with c4-c5 . Once I realized the importance of this idea, I began investigating the same possibility a move earlier. You can find my main recommendation analysed below, but first I will j ust show a few lines demonstrating the robustness of Black's defence in the present line. 8 7 6 5 4 3

a

b

e

d

e

f

g

h

We have reached another critical position, where a deep investigation led me to conclude that a novelty was needed.

19.c4!!N

Let us briefly note that after 1 9 .ih5t �e7 20.:!'1:b3 1':1:xfl t 2 1 . �xfl lD c4! 22.1':1:xa3 lDxd2t 23.lDxd2 ie5 White was struggling for compensation and Black eventually converted his extra pawn in Anand - Nepomniachtchi, Mainz (rapid) 2009. To understand the real value of the text move, we should compare it to the following line: 1 9 .c,!;h l ! ? This was played i n a correspondence game. Here I was unable to find an advantage after: 1 9 . . . b5!N 1 9 . . . b6 was played in Drossler - Dobrica, corr. 20 1 3 , and now 20.c4!N would have been strong. You can find this position analysed in variation B22 1 below, under the move order 1 9 .c4!! b6 20.�h l ! . I considered a few other moves, including 19 . . . c,!;e7, when 20.c4! is strong once

2

a

b

e

d

e

f

g

h

20.ih5t 20.if2 hopes for land a lethal pair of checks on h5 and c5, but after 20 . . . 1':1:xf2! 2 1 .1':1:xf2 �d8! White must force a draw with: 22.1':1:b3 Wlxa2 23.lDc5 �e7 24.lDe4 �d8= 20.lDd6t �e7 2 1 .1':1:fd l 1':1:b8! 22.1':1:b4 1':1:b6 23.c3 seems like a romantic position, but I don't think Black has any objective problems. 20 . . . �e7 2 1 .1':1:xfS �xfS 22.1':1:fl t �g8! 22 . . . �e7 23 .Wld4 100ks extremely dangerous for Black.

a

b

e

d

e

f

g

h

1 2 . . . g5

304

23 . .ixe5 Le5 23 . . . liJ xe5 24.VMd8t mh7 2 5 . liJ f6t .ixf6 26.VMxf6 VMd6 27 . .ig6t mg8 28 . .if7t mh7 might be another way to make a draw. 24.liJxg5 After 24 . .if7t mg7 2 5 . .ixe6 liJ f6 26 . .ixc8 1:%xc8 27.liJxf6 VMd6! Black is doing well . 24 . . . VMe7 2 5 . liJ f7 .ig7 26.liJxh6t mh8= I found nothing more than a perpetual for White. Let's return now to the stronger 1 9 .c4!!N. Depending on Black's response, White can either play mh 1 having avoided the defensive line given above, or do something else. Black has two main tries: B221) 19 b6 is playable but leaves Black under pressure, so B222) 19 b5! seems like a better try. •••

20 J:�xf1 t ••

20 . . . VMe7? loses quickly after 2 1 .liJd6t md8 22.c5 ! , and 2o . . . ib7? 2 1 .liJd6t �e7 22.liJxb7± is also unpleasant for Black. 20 . . . md8?! 2 1 .1:%fd l ! gives White an improved version of the main line below, since the king cannot escape via f7 now. 2 l . . .1:%b8 (2 l . . .mc7? 22.c5!+-; 2 l . . .me7 22.1:%b3! VMa4 23 .VMd6t mf7 24.1:%fl t mg8 2 5 .1:%xfBt liJxf8 26 . .ixe5+-) 22.1:%b3 VMe7 23.c5 White has a powerful initiative.

2I J3xf1 d8! 22J3dl Mter 22.c5 Black can defend with 22 . . . .ib7!, so White should prevent the bishop from developing.

••.

1 9 . . . liJc6 20.�h 1 gives White a strong initiative.

B221) 19 b6 •••

a

b

e

d

e

f

g

h

22 Y*fe7! ••.

a

b

e

d

e

f

g

h

20.hl !

Transposing to a favourable version of 1 9 . m h I where White has bypassed the resilient 1 9 . . . b5!N.

20.c5 leads to some interesting complications, but the fact that the king can come under check proves to be important in several variations.

The only way to offer serious resistance. 22 . . . mc7? 23.c5 is crushing, while after 22 . . . 1:%a7 23.c5! Black's position is collapsing again: 23 . . . bxc5 24.liJxg5! hxg5 25 .VMxg5t mc7 26.VMxg7+-

23.�d6 gb8! 23 . . . mc7 24.c5! liJ xc5 2 5 .VMe3 mb8 26.liJxc8 mxc8 27 . .ixe5 maintains some pressure. In the main line the computet continues to find obscure defensive ideas, but Black's position still remains on the edge.

Chapter 1 6 - 1 3 .�f2! ?

305

27.YlYb2! 27.cxd7 !!xd7 is perfectly safe for Black, but there is no need to release the tension so soon.

27 ... gxc6!

Mter 27 . . . Wfc5 2B . .if3! g4 29 . .ie4! (29 .cxd7 !!xd7 30 . .ifl Wfb5 3 1 ..ixb6t �eB 32.Wfxb5 axb5=) the two bishops will continue to provide terrific compensation against Black's exposed king, for instance:

a

b

e

d

e

f

g

h

24.cS 24.�h5!? gives White excellent compensation too, while avoiding too many forced lines. This could be relevant in case Black finds an exact way to equalize after the text move.

24 ...�b7 2S.tihb7t gxb7 26.c6 26.�xa6?! !!a7 27.�xe5 �xe5 2B .c6 �d6! leads to a slightly worse endgame for White.

26 ... gc7

No better is: 26 . . . lD xc6 27.�f3 Wfc5 (27 . . . lDa5 2B .�d6 WfeB 29 .Wfd3!±) 2B.!!c1 (2B ..ifl!? is also tempting.) 2B ... Wfd4 29 .Wfc2 lDde5 30.!!d l !!d7 3 1 .!!xd4 lD xd4 32.Wfb U Black has enough material for the queen, but the queen is in her element with targets spread across the board.

a

b

e

d

e

f

g

h

b

a

e

d

e

f

g

h

29 . . . Wfb5 (29 . . . �cB 30.cxd7t !!xd7 3 1 .!!c1 lD c4 32.Wfe2 b 5 33 . .id3t; 29 ... �eB 30.!!c 1 Wfb5 3 1 .cxd7t �xd7 32.Wfd2t �cB 33.!!b l Wfc5 34.a4�) 30 .Wfd2! Wfa4 3 1 ..ih4t �cB 32.cxd7t !!xd7 33.!!c1 t �bB 34.Wfe l �

28.heS heS 29.YlYxeS The position looks rather unpleasant for Black, although the computer continues to find defensive resources.

a

b

e

d

e

f

g

h

306

1 2 . . . g5

29 Y!YcS! 30.Y!Yh8t r!Jc7 31 ..ba6 Y!YdS! 32.Y!Yal!? •••

Avoiding the forced draw that may occur after 32 . .te2 'We5 33 .'WeB E!:d6 34.E!:c l t E!:c6.

20 gxf1 t •..

20 . . . ltJ c4? can be dispatched surprisingly easily. 2 1 ..txc4 bxc4 22.E!:bd l ! (22.c6 'Wd3! 23.ltJd6t �e7 24.ltJxcBt E!:axcB 25 .'Wb4t @eB 26.cxd7t 'Wxd7=)

32 Y!YeS 33.Y!Ybl E!:d6 34.E!:f1 •••

The computer would have you believe that Black is absolutely fine after every move, but I would argue that White maintains nagging pressure. Of course Black's position is defensible with computer precision, but there is no forcing route to a draw, which is rather annoying for the player with an open king. Black's extra pawn is not too significant, as the major pieces are still on the board. Also, while Black needs to choose berween lots of relatively indistinguishable options, White has a number of easy moves such as .tb5 , a2-a4, and various ways of poking around with the queen and rook.

B222) 19 bS! •••

8 7 6 5 4 3 2 b

a

e

d

e

f

g

h

22 . . . 'Wd3 (22 . . . E!:xfl t 23.E!:xfl 'Wd3 [23 . . . ltJxc5 24.ltJ d6t+-J 24.ltJ d6t @dB 25 .'Wa5t+-) 23.E!:xfB t ltJxfB 24.'We l .td4t 25 .@h l 'We3 26.ltJ d6t �e7 27.'Wb4 .txc5 2B.ltJxcBt E!:xcB 29 .'Wb7t+-

21 .gxf1 r!Jd8!

8

The only way to survive. Even though the king cannot run to c7 yet, he at least avoids the immediate threats.

7 6 5 4 3 2 1 a

b

e

d

e

f

g

h

The best try, but White has some extra resources compared to the 1 9.�h l b 5 ! line mentioned in the notes on page 303.

20.cS

20.ltJd6t?! �e7 2 1 .E!:fd l E!:bB (2 1 . . . bxc4 22. �h 1 -+) Black intends . . . E!:b6, and after 22.cxb5 .tb7! White's initiative is stalling.

a

b

e

d

e

f

g

h

22.gdl Keeping up the pressure. A move like 22.@h l would be too slow, and after 22 . . . .tb7 White

307

Chapter 1 6 - 1 3 .if2!? will struggle to demonstrate compensation.

22 ... lLlc6!? Somewhat counter-intuitive, but Black creates important resources like . . . 'lWb2 or . . . 'lWb4 to threaten a queen exchange, as well as the possibility of putting a piece on d4 to block White's attack. 22 . . . a5?! intends . . . 'lWb4, but White can easily deal with this: 23 .ixb5 'lWb4 24.'lWe2 ctJe7 25 .ie 1 ! (2 5 . ctJh 1 looks good too, but Black has an incredible defensive line beginning with 25 . . . E:b8!, when I found no more than a draw for White.) 25 . . . 'lWa3 26.if2t Black is under heavy pressure, with no obvious way of simplifying. A more resilient attempt is: 22 . . . E:a7 23.ctJh l 23.lLlxg5 does not work due to 23 . . . hxg5! 24.'lWxg5 t rJlc7 2 5 . E:xd7t ixd7 26.ixe5t (after 26.'lWxg7 'lWxc5t Black is at least equal) 26 . . . ixe5 27.'lWxe5t ctJb7, when White must force a draw with 28.if3t ctJc8 29 .'lWh8t ctJc7 30 .'lWe5 t . Now lLlxg5 i s a huge threat, a s White will be able to capture the bishop on g7 without the c5-pawn falling with check. Black's position is on the verge of collapse, and he only has one defence.

This keeps Black in the game, although it feels rather awkward. 2 5 .ixe5 2 5 . lLl e4!? 'lWa4 26.'lWf4� is another possibility, when there are a lot of threats hanging in the air. The text move is a safer alternative which leads to a forcing line. 25 . . . ixe5 26.if3 8 7 6 5 4 3 2

a

b

e

d

e

f

g

h

26 . . . ic3! Another only move to survive. 27.'lWe3 27.'lWe2 'lWxc5 28 .ixc6 hxg5 29.'lWxe6 id4 30.'lWg4 Ieads to a balanced endgame. 27 ... 'lWxc5 28.lLlxe6t E:xe6 29 .'lWxe6 ctJc7 30.'lWxh6 It is hard to assess this position, even with a computer at your disposal. Although the fight will continue for three possible results, Black's decisions are likely to be harder, considering his open king. Another significant point is that a queen exchange may not be bad at all for White, as he will then be able to advance his kingside pawns more easily.

23.i.f3!? I also analysed 23 .'lWd6 'lWe3t, but found that Black could maintain the balance.

a

b

e

d

23 . . . E:c7! 24.lLlxg5 E:c6

e

f

g

h

Another interesting try is: 23.ctJh l ! ? 'lWb2 23 . . . lLl d4 24.id6!+-

1 2 . . . g5

308

24.'1Wd3! ? 24.'?ge3 .id4! 2 5 .'?9f3 .ie5 26. liJ d6 Iflc7+ 8

23 liJd4 •..

23 . . . g4? is refuted by 24.'?9d6! gxf3 25 . .ih4t lfle8 26.'?9xc6, and 23 . . . .ie5 24.lflh I ! .ixg3 2 5 . hxg3 also heavily favours White.

7 6 5 4 3 2

a

b

e

d

e

f

g

h

24 . . . .ie5 There are many possibilities, but I will focus on the computer's top choice. Mter 24 . . . liJ d4 25 . .if3 g4 26 . .ixg4 '?9c2 27.'?9xc2 liJ xc2 28.c6 liJ e3 29 . .ih5 ! Black remains under pressure; the rook cannot be taken because of the brilliant mating net with c6-c7t and .id6. 29 . . . e5 30.l'!d3� 2 5 . liJ d6 .ixg3 26.hxg3 liJ ce5 27.'?9e3 '?9c2 Or 27 . . . l'!b8 28 . .if3� and the fight goes on. 28.l'!c1 '?9b2 28 . . . '?9xa2 29.c6!±

a

b

e

d

e

f

g

h

24.c6 ga7

24 . . . liJ xc6 2 5 . liJ d6±

25 .td6 •

The most obvious human way. 2 5 .'?9£2 and especially 25 .h4!? are interesting alternatives which can be checked. 8 7 6 5 4 3

a

b

e

d

e

f

g

h

29.c6 liJ f6 30.'?9b6t lfle7 3 1 .'?9c7t .id7 32.cxd7 liJ fxd7 33 . .ih5 '?9a3 33 . . . '?9xc 1 34.'?9xc 1 Iflxd6 will also leave Black on the defensive for a long time. 34.liJc8t l'!xc8 3 5 .'?9xc8 White can continue trying without risk.

2 1 a

b

e

d

e

f

25 �xf3t .

.

.

2 5 . . . '?9a4? 26.liJxg5 ! is nasty.

g

h

Chapter 1 6 - 1 3.i.fL.! ?

Conclusion

26.gxf3 YlYxf3 27.YlYc2!?i 27.E:e l leads to similar stuff.

a

b

e

d

e

309

f

g

h

There is no way for Black to force a draw, and White's initiative is likely to continue with moves like ig3 , lD d6 and possibly a rook lift. White is a few pawns down (the piece does not really count, as the d7-knight can be removed at any moment) , but the position is much trickier to play for Black.

Considering how heavily the Naj dorf Poisoned Pawn has been analysed, it seems almost inconceivable that an almost-new move such as 1 3 .if2!? could actually be any good. Having studied it in detail, I believe it to be a fully viable alternative to the 'automatic' 1 3 .exf6. After the logical 13 ... lD g4 1 4.ig3 , we began by looking at 1 4 . . . ig7!? 1 5 . lD e4 . At this point the most obvious move for a human is the untested 1 5 . . . YlYe3t , but we saw that White's tremendous piece activity offered promising compensation even after the queen exchange. The computer prefers 1 5 . . . 0-0, but then 1 6.h4! leads to interesting complications where both kings may become somewhat exposed, but Black's position remains trickier to handle. 1 4 . . . lD d7 has been the most common choice in the few games in which the position has been reached. Then after 1 5 .ie2 lD gxe5 we analysed 1 6. lD xe6! ? as a tricky sideline, but I believe the acid test of Black's position to be 1 6. 0-0 ig7 1 7.lDxe6! fxe6 1 8 . lD e4. Black has several options in the battle ahead, but in all lines White maintains a potent initiative for the sacrificed piece. Crucially, Black does not seem to have any way to force a draw, either by perpetual check or heavy simplification . Considering the objective dangers in the position, and factoring in the considerable surprise value of 1 3 .if2! ? , I regard this hitherto-ignored move as a serious practical weapon.

12 g5 ...

a

b

e

d

e

f

g

13.exf6 Introduction and 16.�hl -

Variation Index l .e4 c5 2.tLl f3 d6 3.d4 cxd4 4.tLlxd4 tLl f6 5.tLl c3 a6 6 .ig5 e6 7.£4 V;Yb6 8.V;Yd2 V;Yxb2 9J�b l V;Ya3 1 0.e5 h6 1 1 .ih4 dxe5 1 2.fxe5 g5 1 3.exf6 gxh4 •

•

1 4 .ie2 •

311 313

A) 14 tLl d7?! B) 14 V;Ya5 1 5.0-0 Bl) 1 5 h3!? 1 6 .if3!? tLl d7 17.tLl d5 V;Yc5! 1 8J�be l !N .id6 1 9.@ h l ! Bl l ) 1 9 tLl e5 B 1 2) 1 9 .ie5 B2) 1 5 tLl d7 1 6.@h l V;Yg5 17.V;Yd3!? B2 1 ) 17 tLl c5 B22) 17 tLlxf6!? .••

•••

•••

•

313 315 316 318 319 32 1

•••

•••

•••

•••

•••

B2 1 ) afrer 1 7 . . . ii:l c5

B 1) afrer 17 ... 'lWc5!

B22) afrer 1 7 . . ii:l xf6!? .

8

8

7

7

6

6

5

5

4

4

3

3

2

2 a

b

e

d

e

f

l S .:i:'lbe l !N

g

h

a

b

e

d

e

f

l S .'lWc4!N

g

h

,"",""',,;W;J"" "=[���NNN' ,mm "'mJ""

_�w�N�f;<{\ffir _'�N�N ��h�

a

b

e

d

e

f

l S .ii:l f3!?N

g

h

h

311

Chapter 1 7 - 1 3 . exfG - Introduction and I G. @ h l

1.e4 c5 2.c!LJa d6 3.d4 cxd4 4.t£Jxd4 c!LJf6 5.c!LJc3 a6 6 .ig5 e6 7.£4 Wtb6 8.Wtd2 Wtxb2 9J::!: b l Wta3 ' 10.e5 dxe5 1 1 .fxe5 h6 12 .ih4 g5 13.exf6 gxh4 •

switching to a queenside initiative, but it didn't yield anything.

•

Mter the text move Black sometimes tries

A) 14 t£Jd7?!, but B) 14 Wta5 is the more •••

•••

reliable move.

8 7

1 4 . . . h3 can be ignored by 1 5 . 0-0, with a likely transposition to one of the later lines involving . . . h3.

6 5

A) 14 t£Jd7?!

4

•••

3 2 1 a

b

e

d

e

f

g

h

Although 1 3 . .if2!? is a fascinating possibility, it would seem wrong not to also consider the established main line. Fortunately there are plenty of new paths to explore here as well.

14 .ie2 •

Considering that 1 4 . .ie2 lD d7 1 5 . lD e4 is a promising line for White, it is logical to ask if 1 4 .lDe4 might be a good idea a move earlier. It prevents . . . Wa5 and was indeed considered interesting a while back, but eventually Black found a good antidote that did not involve committing himself to . . . lD d7. Here is a brief summary: 14 . . . Wxa2 1 5 .�d l ( 1 5 .�b3 lD d7 I G.Wc3 is met by 1 6 . . . b6! intending . . . Wa5 ; this move has more or less killed the 1 5 .�b3 variation. 1 7.lDxe6 fxe6 I B .f7t
a

b

e

d

e

f

g

h

A fair number of players, including Areshchenko, have apparently not noticed any difference between this and the second main variation. In fact the text move gives White a powerful extra option.

1 5.t£Je4! I should mention that White is not obliged to play this way, and can easily transpose to the main line with 1 5 . 0-0 when Black almost invariably plays 1 5 . . . Wa5 . I will not analyse the text move in great detail, but I do consider it a promising attempt to exploit Black's chosen move order.

1 5 Wtxa2 16J:!:dl Wtd5 •••

1 6 . . . �gB 1 7. lD b3 ! ? (After 1 7. 0-0 Wd5 I B .We3 h3 1 9.93 .ic5 the position is more

312

1 2 . . g5 .

complicated as ltJ xc5 isn't possible, although I still prefer White.) 1 7 . . .l'hg2 ( I 7 . . . Wfa3 I B .O-Ot) I B .'kt>f1 :B:xe2 ( I B . . . :B:g6 1 9 .ih5t) 1 9.Wfxe2 Wfb2 20.:B:gl ltJxf6 2 1 .:B:d4 ltJ xe4 22.Wfxe4 White keeps some initiative. 1 6 . . . h3 can be ignored. 1 7. 0-0 hxg2 This is not particularly scary, but there is nothing better. ( I 7 . . . :B:gB I B . ltJ b3 ! ? [ I B .if3 hxg2 1 9 .:B:fe l seems quite dangerous for Black as well.] I B . . . :B:xg2t 1 9 .'kt>h l As usual, losing the g2 pawn is no big deal. 1 9 . . . Wfa3 20.:B:f3!±) 1 B .:B:f2 Wfd5 8 7

Both 1 7 . . . :B:gB and 1 7 . . . h3 are met by 1 B . ltJ b 5 ! with a clear advantage to White. Mter 1 7 . . . ic5 1 B .0-0 it will not be easy for Black to finish development. l B . . . h3 ( I B . . . e5 1 9 .1tJxc5 Wfxc5 20.:B:f2!±) 1 9 .:B:f3!? hxg2 20.ltJxc5 (20.:B:h3! ?�) 20 ... Wfxc5 2 1 .l'!f5 e5 22 .Wfg3--+

18.c3 ltJe5 This prepares . . . id7 and . . . ltJ c6, but Black will end up two tempos down on the 1 4.ltJe4 line mentioned above. Therefore it will be easy for White to generate counterplay against the intended long castle.

19.0-0 .id7 20.:B:al lYe7 21.�hl :B:g8 22.h3 0-0-0

6 5 4

8

3

7

2

6 b

a

e

d

e

f

g

5

h

1 9 .Wfe3 ! Wfe5 ( I 9 . . . ic5 20.ltJxc5 Wfxc5 2 1 .:B:f5 ! Wfb6 22.ih5 +-) 20.ltJf5 ! exf5 2 1 .:B:xf5 Wfe6 22.ic4 Wfc6 23 .:B:d6 ixd6 24.ltJxd6t 'kt>dB 2 5 . ltJ xf7t 'kt>c7 26.Wff4t 'kt>b6 27.ltJxhB±

4 3 2 1 a

b

e

d

e

f

g

h

23 .ba6! bxa6 24.:B:xa6 .ie6 25.tihc6 �xc6 26J:�bl �d7 •

White was ready to increase the pressure with moves like, Wfe2 , Wff3 or :B:ab6.

27.lYd3t �e8 28.lYb5 gb8 29.lYxe6t lYxe6 30.gxb8t �d7 3U�xe6 �xe6 32.e4

a

b

17.lYe3 lYa5t

e

d

e

f

g

h

This was the wonderfully instructive game Hebels - De Carlos Arregui, corr. 20 1 2. White has achieved a dominating position, and he went on to win by exploiting the pin along the eighth rank.

Chapter 1 7 - 1 3 . exf6 - Introduction and 1 6. @ h l

B) 14 'lNa5 •••

8 7 6 5 4 3 2 1 a

b

e

d

e

f

g

h

313

the queen is ideally placed on e3 to support further tactical motifs in the centre. 1 6 . . . .ic5! 1 7.@h l ! ? 1 7 . .if3 has been much more common, but Black seems to have nullified all of White's attempts here. The king move has hardly ever been played, but is a useful precaution against the annoying . . . It'l c6. 1 7 . . . hxg2t 1 B .@xg2 Losing time with the king is certainly a concession, but Black's pieces are not active enough to utilize the g-file, and White might actually be able to make better use of it after a few moves. Unfortunately, Black has a surprisingly effective defence.

This is the usual move, effectively ruling out a quick It'l e4 on account of the queen exchange.

8

15.0-0

6

7

Now we will consider Bl) 1 5 .. h3!? followed by the more popular B2) 1 5 �d7. .

•.•

Bl) 1 5 h3!? ••.

This move has largely been ignored by over­ the-board players as it offers White plenty of attacking possibilities, whereas developing the knight to d7 is much easier to understand as a plan. Nevertheless, a number of lines have been deeply explored in correspondence play.

16 ..tf3!? It seems to me that this is a good moment to steer the game away from the heavily trodden paths of the correspondence players. 1 6.g3 is another possible direction to explore, but I would prefer to activate my pieces more quickly. The most natural and popular move has been: 1 6.�e3 Apart from preventing the . . . �g5 threats,

5 4 3 2

a

b

e

d

e

f

g

h

I B . . . .ixd4! With the simple idea of completing development as quickly as possible. The open king on g2 gives Black enough counterplay to make up for the loss of the dark-squared bishop. All other moves seem to be bad for Black, but I will not take up space with the details. 1 9.�xd4 It'l c6 20.�d3 .id7! 2 1 .gxb7 0-0-0 22.gb2 @c7 Black was doing fine in Rasmussen - Siigur, corr. 20 1 2. Let's now focus on my recommended move. Placing the bishop on f3 is a typical way to meet . . . h3, since . . . hxg2 is never a real threat. White also anticipates . . . �g5 .

314

1 2 g5 . . .

1 B . . . lLl b6 seems senseless. Both 1 9 .1Llxb6 'lWxb6 20.'it>h 1 � and 1 9 . 1Ll e7 lLl c4 20.'lWc3� (or even 20.'lWf2) give White a pleasant position with a typical initiative.

a

b

e

d

e

f

g

h

16 ... tLld7

This has almost always been played.

1 6 . . . hxg2 only gets in the way of Black's future ideas, and after 1 7.E:f2! White has a fine position. A short illustrative line is 1 7 . . . lLl d7 1 B .lLld5! 'lWc5 1 9 . 1Ll xe6 fxe6 20.J\h 5 t 'it>dB 2 1 .lLl f4 and wins.

An additional option, when compared to 1 B .E:fe 1 , could be: l B . . . 'it>dB!? This hardly looks like the most worrying line for White, even though it is not easy to refute. 1 9 . 1Ll e7 Plonking the knight in the far reaches of the enemy position is a typical idea. It often proves annoying for Black, as exchanging it for the dark-squared bishop would open up his whole position. 1 9 .1Lle3!? is also possible, intending a slower attack against the enemy king.

17.tLld5 Obviously the critical move! The correspondence players continue the calm defence with:

17 .. :�c5! Not 1 7 . . . 'lWxd2? 1 B . lLl c7t 'it>dB 1 9 . 1Ll dxe6t fxe6 20.lLlxe6t 'it>eB 2 1 .J\h5 mate! Now for some reason everyone has placed the wrong rook on the e-file!

1 8.E:bel !N Frankly the most logical move. Moving the other rook to e1 seems intuitively the wrong attacking option to me, even though it has the advantage of keeping an active rook on b 1 .

1 8 ....id6 This has been the normal response in the analogous position where the other rook has gone to e l .

a

b

e

d

e

f

g

h

1 9 . . . e5 1 9 . . . J\xe7 20.fxe7t 'if;; xe7 2 1 .E:f2 maintains White's initiative. 1 9 . . . lLlxf6 is unsatisfactory due to 2o.lLlxcB 'it>xcB 2 1 .J\xb7t! 'if;; x b7 22.E:xf6 and White is totally dominating. 20.c3 20.E:e4! ? also deserves attention. This could all be analysed considerably further, but it doesn't seem too theoretically relevant. I think White has a strong initiative here" and there was also a safer alternative on move 1 9, so take your pick.

Chapter 1 7 - 1 3 . exf6 - Introduction and 1 6. � h l

315

of options. 2 1 . . .Wc6 22.lt:la5 Wc5 23 .:!'!e4! The idea is surprisingly j ust :!'!d l . Black has no way to continue development. 23 . . . :!'!g8 24.lt:l e3± Black's position seems to be on the verge of collapse.

8 7 6 5

Returning to the main line, White has a few ways to keep the initiative going.

4 3 2 1 b

a

e

d

e

f

g

h

19.�hl!

By escaping the p i n , White prepares for all sorts of knight jumps. 1 9 .1t:lf4 is too slow, and after 1 9 . . . lt:lxf6! 0 9 . . . hxg2 20.lt:lfxe6!) 20.�h l .ie7! (20 . . . .id7? 2 1 .lt:l dxe6! wins) 2 1 .lt:lxh3 e5 22.lt:lb3 Wc7 Black seems safe. I also considered: 1 9 .:!'!f2 .ie5 ! ? Not the only move, but the most forcing. 20.c3 exd5! 2 1 . .ig4 Wd6! (2 1 . . .�d8 22 . .ixd7 seems unpleasant for Black.) 22 . .ixd7t �xd7 23.lt:lf5 We6 24. lt:l g7 Wd6= White has no more than a repetition, despite Black's awkward pieces. We will now analyse B l l) 19 �e5 and B12) 19 i.e5. The second has been the universal choice in the parallel position with a rook on b 1 instead of f1 , but the first is certainly playable as well.

a

b

e

d

e

f

g

h

21.�e3!? This is the simplest move, j ust aiming for a long-term initiative, particularly since Black's development woes continue. After 2 1 . lt:l a5 Wc5 there is nothing better than repeating with 22.lt:lb3.

••.

..•

19 ... hxg2t?! 20.Wxg2 does not help Black at all.

Bl l) 19 ... �e5 20.�b3 %Yc6

2 1 . lt:l e7!? is another tempting possibility. 2 1 . . .Wc7 (2 1 . . .lt:lxf3 22.:!'!xf3 hxg2t [22 . . . Wc7 23.:!'!c3! +-; 22 . . . .ixe7 23.fxe7--+] 23 .Wxg2 Wc7 24.lt:ld4t seems dangerous for Black.) 22.lt:ld4!? Increasing the pressure. 22 . . . .id7 (22 . . ..ixe7 23.fxe7t ; 22 . . . lt:lxf3 23.:!'!xf3t) 23 . .ih5 ! ? The position is unclear but certainly dangerous for Black.

21. %Yb6 ••

20 . . . Wb5 2 1 .a4 would not really help Black. After 20 . . . lt:lxf3?! 2 1 .gxf3 there will no longer be any threat of . . . exd5, so White gets a lot

2 1 . . .Wc7 22.c4! .id7 23.c5 gives White a powerful initiative. 2 1 . . . lt:l xf3 22.gxf3 Now the knights are

1 2 . . . gS

316

jumping, and it already seems impossible to hold Black's position together. 22 . . . �f4 (22 . . . �d7 23.tDaS! Vlff c7 24. tD d S +-; 22 . . . Vlff c7 23.tD fS ! ±)

B12) 19 ....te5

a b

a

e

d

e

f

g

h

23.Vlff d 4! Vlff d 6 24.tDdS �gS 2 S . tD c S ! White wins.

b

e

d

e

f

g

h

This appears awkward, but it has been the universal choice in the 1 8 .�fe l line, and is also the computer's top choice.

20.tilb3! In the position with the rook on b 1 instead of n , White has always played 20.tDxe6 fxe6 2 1 . tD b6, but after 2 1 . . .�b8 the defence has held firm, and the database shows a long list of draws. I would like to steer the game in a different direction; one which enables us to make use of the rook on n .

2o ...V9d6 a

b

e

d

e

f

g

h

22J:�dl!?

22.c4 would run into 22 . . . �b4, so the rook steps aside.

22 ....tc7 None of the other squares look convenient.

23.gxh3t The position looks awfully dangerous for Black. Note that it is not possible for him to continue developing with . . . �d7 on account of tD c4.

2o . . . Vlff c4 2 1 .tDaS Vlff c S is covered under 2 1 . . .Vlff c S in the note to Black's next move.

21.tila5!

A typical re-routing of the knight.

21 ...b5 After 2 1 . . .hxg2t?! 22 .�xg2 bS 23.c4 the opening of the g-file only helps White. 2 1 . . .Vlff c S This is a more critical response, challenging White to come up with something better than repeating the position with tDb3. 22.�xe S ! ?

317

Chapter 1 7 - 1 3 .exf6 - Introduction and 1 6 . � h 1 Challenge accepted! 22.c4!? might offer White a decent initiative too, but it will require a lot more analysis. 22 . . . ttJxe5 23 .�f4 exd5 24.�xe5t i.e6 25.ttJxb7 The struggle remains complicated, but Black is under pressure and has no forcing way to equalize.

22 ... exd5

22 . . . 1'!gB is met by 23.gxh3 exd5 (23 . . . 1'!g5 24.c5) 24.i.xd5 1'!a7 2 5 .1'!f5! and Black's position is collapsing.

22 . . . 1'!bB runs into a spectacular refutation: 23.c5! ttJ xc5 (23 . . . �xc5 24.1'!c l +-) 24.1'!xe5! �xe5

8 7 6 5 4 3 2 b

a

e

d

e

f

g

h

b

a

25 . . . �b6 25 . . . �b4 �d7!) 26.i.xd5 (26.ttJ d6t 26 . . . �f8 27.i.e4 maintains excellent compensation. 26.ttJd6t �f8 26 ... �d7? 27.ttJxf7! i.xf7 2B .�e7t wins. 27.gxh3 1'!dB 2B.ttJf5 White continues to create problems.

e

d

e

f

g

h

2 5 . ttJ e7! i.d7 (25 . . . i.b7 26.ttJ ac6 �c7 27.1'!c l +-; 25 . . . ttJ e4 26.�e l ±) 26.ttJ ec6+-

23.c5!? 23 .i.xd5 �f8 could also be explored further.

23 ... lthc5 23 . . . �xc5 24.i.xd5± 8 7 6 5 4 3 2 1

a

b

e

d

e

f

g

h

22.c4! Continuing to add firepower to the attack.

a

b

e

d

24.Y;Yxd5 Y;Yxd5 24 . . . ttJ e6 2 5 .�xaB;!;

e

f

g

h

1 2 . g5

318

. .

25 ..hd5 ga7 26.gxe5t tile6 27.gdl

17.�d3!?

Despite the simplifications, the f6-pawn remains annoying for Black. If he castles to bring his rook into play, then his king may come under fire.

It is obvious that we need to avoid the queen exchange, but this is the rarest way of doing it, having been played in only about twenty games out of a few thousand. It has been ignored by most sources, but I believe it is a good practical try.

B2) 1 5 ... til d7

I used to think 1 7.�e l ! ? was dangerous as well, but by now the correspondence players have figured out that 1 7 . . . ltJxf6! leads more or less by force to a drawish position.

8 7 6 5 4 3 2 1 a

b

e

d

e

f

g

h

From this theoretical tabiya I have examined j ust about every plausible try for White, and will present what I consider to be the two most promising directions.

16.�hl

1 7J'!f4 is the well-known theoretical line. Once upon a time nobody knew much about this position, but then after a lot of correspondence and computer testing it was played in a wild battle between Nisipeanu and Woj taszek in 20 1 1 . At the time this ushered in a new chapter of the Poisoned Pawn, but by now there isn't much excitement left, so I will cut down on the analysis and j ust mention the main line with a few brief notes: 17 . . . e5 ( 1 7 . . . ltJxf6? 1 8 . ltJ db5!t) 1 B .ltJd5 exd4 1 9 .�xd4 @dB! 8 7

This has been by far the most popular choice, but I will soon be adding a fresh twist.

6

1 6J�bd 1 ! ? is a rare but potent alternative which will be analysed in detail in the final chapter.

4

16 ...�g5

2

This is one of the most heavily debated positions in the Poisoned Pawn, and it is crucial for the viability of the 1 2 . . . g5 variation. Unfortunately, the majority of games have followed the sad trend of repeating well­ established lines from correspondence games and reaching forced draws. At some point even I got the feeling that this line was close to being 'dead' for White, but then I realized there is an oasis of possibilities that have yet to be worked out in detail.

5 3

a

b

e

d

e

f

g

h

20.ltJe7! There are other moves, but they have also been analysed to death in correspondence games. 20 . . . �e5! (Amazingly, 20 . . . �c5 ?! 2 1 .�b2!! is rather good for White. It's an exciting position to check further, but since it has little relevance to the theoretical debate, I won't go any deeper.) 2 1 .�d2 ixe7 22.fxe7t �xe7 23 .ig4 f5 ! 24 .!!xf5 !!eB Many games

319

Chapter 1 7 - 1 3 . exf6 - Introductio n and 1 6. m h l have been drawn from here. Here are a couple of examples showing the general pattern: 25.1"1:d1 (2 5 .1"1:d5 b6 26.h3 ilc7 27.Wfxh6 1"1:bB 2B .Wff4t ilb7 29 .Wfd4= Guseinov Areshchenko, Antalya 20 1 3 .) 25 . . . ilc7 26.1"1:f7 Wfxf7 27.Wfd6t <;!;dB 2B .Wfb6t <;!;e7 29 .Wfd6t <;!;dB= Yu - Wei, Sharjah 20 14.

b

a

e

d

f

e

g

h

18 ... b5

This is the first move that comes to mind, but it looks extremely scary for Black. Still, the sacrifice on b5 is not winning outright, and we should check it to make sure Black does not have a refutation or a forcing path to equality. a

b

e

d

e

f

g

h

Returning to my recommended move, we will start by looking at the natural B21) 17 �c5 before checking B22) 17... c!ihf6!? .••

B21) 17 ... �c5 18.�c4!N This unusual-looking move may offer considerable shock value. As you may have noticed, it is often hard for White to find a safe square for his queen in these positions, so in that sense c4 is not a bad square at all. A more important point is that Black cannot easily prepare long castling as his knight will be caught in a pin. He can put his rook on cB to chase the queen, but then she will just retreat to the ideal e2-square and Black's king will be stuck in the centre. 1 B .Wff3 is the only move to have been played, but it gives Black a few ways to make a draw, for instance: 1 B . . . .id7 1 9 .1'!xb7 ttJ xb7 20.ttJxe6 .ixe6 2 1 .Wfc6t .id7 22 .Wfe4t .ie6= LePage ­ Lifanov, corr. 20 1 2 .

I B . . . h3 is met by 1 9 . .if3 when, as usual, the opening of the g-file is of little concern for White. 1 9 . . . hxg2t 0 9 . . . 1"1:gB 20.g3;!;) 20 . .ixg2 .id7 2 1 .Wfe2! ?� Black still cannot castle, and White is ready to j ump all over the centre with ttJ f5 , l'!f5 or ttJ f3 . This can be analysed further of course, but it is clear that White has the initiative. 1 B . . . .id7 1 9 . .if3 .id6 0 9 . . . 1"1:cB 20 .Wfe2 only improves the position of the white queen, and after 20 . . . .id6 2 1 . ttJ e4 ttJxe4 22 . .ixe4 Black lacks a safe place for his king.) 20.ttJe4!? (20 .Wfe2 Wff4!) 20 ... ttJ xe4 2 1 ..ixe4 0-0 (2 1 . . . l'!cB 22.Wfd3±) 8 7 6 5 4 3 2

a

b

e

d

e

f

g

h

320

1 2 . . g5 .

22.1Mfe2!? The pawn on f6 gives White a constant initiative, and he can patiently build up his position, for instance: 22 .. J'!:fd8 23J'!:be l ! ? (Avoiding 23 . .td3 .tc5 24.tLlf3 1Mfe3! when Black equalizes.) 23 .. J'!:ac8 24 . .td3! i>f8 2 5 .1Mfe4-+

a

b

e

d

e

f

g

h

19.c!Llcxb5 axb5 20.c!Llxb5 Y!Ve5 2 1 .�fdl! Black seems t o b e i n considerable danger, but he can survive as follows.

23 ... h3!

23 . . . .txb 5 ? 24.1:'!:xb5 is winning for White.

23 . . . 1:'!:b6! ? This does not lose by force, but it looks too strange and gives White a lot of possibilities. At least one way to continue pressing is: 24.1Mfxh4 24J�d2 .txb5 25 .1Mfxb5t 1:'!:xb5 26 . .tc6t tLl d7 27.1:'!:xd7 1Mfxf6 28 .1:'!:dd l t rtie7 29.1:'!:d7t is only a draw. 24.h3!? is another possibility, but I don't see any need to go too deep here. 24 . . . 1:'!:g8 (24 . . . 1:'!:d6? 25 .1:'!:e l +-; 24 . . . .txb5 25.1:'!:xb5 .td6 26.i>g l !±) 25 .1:'!:d4!? With extreme complications. 24 . . . 1:'!:g8 2 5 .1Mfb4 .txb5 The only defence, but White has an attractive reply. 8 7 6 5

2 1 . J.d7 22.J.f3 �b8 ••

22 .. .:!':'!:a6 23 .1:'!:d2! 1:'!:d6 24.tLlxd6t .txd6 2 5 .1Mfxh4;!;

4 3 2

23.a4

Black seems to be on the verge of losing, but he can stay in the game with: 8 7 6 5

a

b

e

d

e

f

g

h

26.1Mfxb5t! 1:'!:xb5 27 . .tc6t tLl d7 28.1:'!:xd7 1Mfxf6 28 . . . 1:'!:xb l t 29.1:'!:d l # is a beautiful line. 29.1:'!:bd l ! .td6 30.1:'!:7xd6t rtif8 3 1 .axb5 Black will have to fight for a draw. White's pawns are dangerous, and he does not take any risk.

24J�e1

4

24.g3 also looks tempting, but there are some beautiful defensive ideas: 24 . . . 1:'!:g8! (24 . . . h5 2 5 .a5 h4 26.a6! hxg3 27. hxg3 1Mfxg3 28 .1Mfc3! +-) 2 5 .a5 1:'!:g4! 26.1Mfxg4 1:'!:xb5 27.1:'!:xb5 .txb 5 and Black should survive.

3 2 1 a

b

e

d

e

f

g

h

32 1

Chapter 1 7 - 1 3 . exf6 - Introduction and 1 6. 'it> h l

30.�xc5t .id6 3 1 .�h5

8

3 1 .Wfe3 Wfh4 32.h3 �gB�

7

3 1 ...�g5 32.�h3

6

Despite the computer's assessment of 0.00, the outcome is still uncertain and I would definitely consider Black's position harder to play.

5 4 3

B22) 17 .. .tl:Jxf6!?

2

a

b

c

d

e

f

g

h

24 .. J�xb5! Lines like this are great fun to analyse. Fortunately, Black's amazing defensive resources do not guarantee a draw. 24 . . . hxg2t 25 . .ixg2 would lead to the same position in a few moves.

25.axb5 Yfxf6 Black seems to be getting away, but White can cash in his passed b-pawn to win the knight by force.

26.b6 i.d6 27.b7 hxg2t 28.i.xg2 'it>e7 29.b8=� .ixb8 29 . . . �xbB 30.�xbB .ixbB 3 1 .Wfxc5t .id6 32.Wfc4 Wff4 33.Wfxf4 .ixf4 34.c4 also gives White risk-free chances to press.

a

b

e

d

e

f

g

h

a

b

e

d

e

f

g

h

Even though the analysis of the previous line was not exhaustive, we saw clearly that White's initiative persisted, so let us now look at a blunt, even artificial, attempt to force things.

1 8.tJf3!?N I B .WfS looks natural, but unfortunately a relatively forced drawing line has already been demonstrated here: I B . . . �gB 1 9 .�bd 1 ( 1 9 . tiJ e4!? looks exciting, but it still only leads to a draw: 1 9 . . . tiJ xe4 20.Wfxf7t WdB 2 1 ..if3 tiJ d6 22.tiJxe6t .ixe6 23.Wfxe6 �g6 24.Wfb3 �cB [24 . . . �bB also leads to the same outcome.] 2 5 . .ixb7 .ie7 26.Wfb6t �c7 27 ..ic6 h3 2B .g3 Wfe5 White will have nothing better than an eventual WfbB-b6 perpetual.) 1 9 . . . �g7! An amazing defensive move. Despite his big lead in development, White found nothing better than 20.tiJxe6 .ixe6 2 1 .Wfxb7 �cB 22.�xf6 Wfxf6 23 . .ib5 t when a draw was agreed in Jensen - Pellen, corr. 20 1 1 .

322

1 2 . . . g5

18 YlYg7 •••

It would be too risky to post the queen on the queenside with the black king so exposed.

once you compare the lines below, you will see that Black is suffering here. 8

19.�e4 tLlxe4 1 9 . . . h3 is not a good idea, since after 20.gxh3 the open file will only benefit White.

7

1 9 . . . i.e7 20.�d6t! i.xd6 2 1 .Wlxd6 � e4 22 .Wlb4 � c3 looks scary for Black after both 23 J%be l and 23.�d2!?, when 23 . . . tLlxe2? would lose to 24.tLlc4.

5

6

4 3 2

20.YlYxe4 .ie7! 20 . . . fS 2 1 .Wlc4 i.e7 22.tLld4! is too shaky for Black. The computer's initial suggestion was: 20 . . . i.c5 2 1 .1:%bd l ! Black's idea is: 2 1 .tLlxh4 Wld4 (2 1 . . .0-0 22.1:%f3!-+) 22 .WlB 0-0 23 .Wlh5 White seems to have some initiative, but in fact Black is fine after 23 . . . i.e7 24.tLlB Wlg7, when it is hard to avoid a queen exchange. 2 1 . . . 0-0 22.tLlxh4 i.e7 22 . . . i.a7 23 .i.h5 looks dangerous.

a

b

e

d

e

f

g

1 a

b

e

d

e

f

g

h

21 .tLlxh4 2 1 .Wla4t!? i.d7 22.i.b5 is a strange idea. After 22 . . . i.xb5 23.1:%xb5 0-0 24.1:%xb7 White is not taking any real risk, as he can easily pick up the h4-pawn, but perhaps it is also true (as the computer claims) that Black is too active to be worse here.

21. 0-0 22.�f5! exf5 23.YlYxe7 ••

A critical position. The computer is happy with Black's extra pawn, but it's clear that White is in absolutely no danger. If he can j ust coordinate his forces, Black may feel increasingly uncomfortable.

h

2 3 . tLl fS ! exfS 24.Wlxen White has a considerably improved version of the main line analysed below. The rook is on d 1 instead of b 1 , so the a2-pawn will not be captured with tempo. The computer still indicates that Black is not doing too badly, but

a

b

e

d

e

f

g

h

323

Chapter 1 7 - 1 3 .exf6 - I ntroduction and 1 6. @ h l

24 ....ie6 25.�xb7 .ba2

23 �g5! •••

Gaining a ,useful tempo. 23 . . . .ie6 seems natural, but now White can arrange the optimal piece placements: 24.'Wxb7 .ixa2 25 .:!::\ a l .ie6 (25 . . . 'We5 26 . .id3 .id5 27.'Wb6 .ie6 2S.l'hel;!;) 26 . .id3

25 . . . gadS is the computer's suggestion, but I don't see why it is so nice for Black. After 26.gbd l .ixa2 (26 . . . a5 27.a3 is similar) 27 . .ixa6 White has an 'extra' pawn on c2, since Black's f-pawns are not so useful. Of course it's not a clear advantage, but nor does it look like 0.00 in a practical game.

26J�a1 .ie6 8 7 6 5 b

a

e

d

e

f

g

4

h

26 . . . :1'UbS (26 . . . a5 27.ga4 gfdS 2S .'Wf3 .id5 29 .'Wf2;!;) 27.'Wf3 'Wg4 2S .'We3 Wg5 29 .'We2� The a I -rook is ready for its lift to the kingside.

3 2 1

24.�c7

a

24.'Wb4 is a more solid approach, and a decent practical try because Black's 25th move isn't easy to find. 24 . . . i.e6 2 5 .gf3 @h7! The only clear way to equalize. (25 . . . i>hS 26.'Wxb7! and Black cannot capture on a2 due to 'Wb2t.) 26 . .id3 ggS 27.gg I b5 With a likely draw after White takes the f5-pawn.

b

e

d

e

f

g

h

27.i.d3!? If 27 . .ixa6 f4 2S .'Wf3 .id5 White's pieces are somewhat awkward. Also after 27.gxa6 gxa6 2S . .ixa6 .id5 Black is okay.

27 a5 •••

27 . . . gfdS can be met by 2S .ga5 ! when the fight goes on. White can continue improving his position with moves like h2-h3 , while keeping the option of suddenly switching over to the kingside. Black will have to keep choosing between many okay-looking moves, without quite making a draw yet.

8 7 6 5 4 3

28JJ:a4 �Ud8

2 1 a

b

e

d

e

f

g

h

2S . . . gfbS 29.'Wf3 ga7 30.h4! is suddenly unpleasant for Black: 30 . . . 'Wg7 (30 . . . 'Wg6 3 I .g4!±) 3 I . .ixf5;!;

1 2 g5

324

. . .

Conclusion

8 7 6 5 4 3 2 1 a

b

e

d

e

f

g

h

29.V9B EM5!

29 . . . .id5 30.�f2�

30.c4 gdd8 3 1 .h3 The game goes on, although Black should be all right.

1 3 .exf6 gxh4 1 4 . .ie2 can be considered the 'ultimate main line' of the Poisoned Pawn. We started by considering the inaccurate 1 4 . . . ltJ d7?!, when White can either allow a transposition to the main line or attempt to punish Black's last move with 1 5 .ltJ e4!, the latter being fully j ustified according to my analysis. 14 . . . �a5 is the correct move order, and after 1 5 . 0-0 we analysed a couple of paths. 1 5 . . . h3!? is a bit risky, but nowadays you never know if your opponent may have spent hours memorizing some long forced drawing lines discovered by correspondence players. For that reason I chose the slightly unusual 1 6 . .if3! ? , and after 1 6 . . . ltJ d7 1 7.ltJd5 �c5 my new suggestion of 1 8 .�be l !N poses a lot of problems to the defence. The more natural and popular choice is 1 5 . . . ltJ d7, when 1 6.'kt>h1 �g5 has been tested in thousands of games. Here I chose to investigate a rare but surprisingly venomous option in 1 7.�d3! ? Then 1 7 . . . ltJc5 1 8 .�c4!N is extremely dangerous for Black, although he seems to survive after correct defence. The alternative is 1 7 . . . ltJxf6!?, but this leads to a relatively simplified position where Black's weakened kingside enables White to test the defence while taking virtually no risk: a nice scenario in a practical game. In the final chapter we will investigate a fascinating alternative for White on move 1 6.

12 g5 ...

a

b

e

d

e

f

g

16J�bdl!r Variation Index l.e4 cS 2.�f3 d6 3.d4 cxd4 4.�xd4 �f6 5.�c3 a6 6..ig5 e6 7.f4 'Wb6 8.'Wd2 'Wxb2 9J�b1 'Wa3 10.e5 dxe5 11.fxe5 h6 12..ih4 g5 13.exf6 gxh4 14..ie2 'WaS 15.0-0 �d7 16J�bd1!?

A) 16....ib4N B) 16...'Wg5!? 17.'Wd3!N �c5 18.'Wc4! .id7 19..if3 B1) 19...h3!? B2) 19...'We3t! 20.i>h1 b5 B21) 21.'Wb4 B22) 21.�d5!? C) 16...h3!? Cl) 17.g3 C2) 17..ig4!?N C21) 17...hxg2 C22) 17...'Wg5!? C23) 17...�e5! C3) 17.'Wd3!?N C31) 17...�e5 C32) 17...'We5 C33) 17...hxg2! 18J�f4!? C331) 18...'Wc7 C332) 18....id6!

326 328 329 331 331 333 336 337 339 339 340 343 345 345 346 347 347 348

h

326

1 2 g5 . . .

1 .e4 c5 2.tiH3 d6 3.d4 cxd4 4.tihd4 �f6 5.�c3 a6 6.i.g5 e6 7.£4 Yfb6 8.Yfd2 Yfxb2 9J�bl Yfa3 10.e5 dxe5 1 l .fxe5 h6 12 ..th4 g5 13.exf6 gxh4 14.i.e2 Yfa5 1 5.0-0 �d7

A) 16 ...i.b4N 17.Yfe3!

We began our examination of this position in the previous chapter, where I noted that the most popular continuation of 1 6.'it>h l �g5 was holding up fairly well for Black - even though 1 7.�d3! ? is an excellent practical try. In this chapter I would like to shift the spotlight to an exceedingly rare move.

16J:1bdl!? b

a

8

e

d

f

e

g

h

Threatening tDxe6; a recurring motif in this chapter.

7 6

17 ...Yfg5

5

1 7 . . . ixc3 ? meets an unpleasant fate: 1 8 .tDxe6! fxe6 1 9 .�xe6t 'it>d8 20.Wfe7t 'it>c7 2 1 .�d6t 'it>d8 22.f7 ig7 23 .ig4 �b5 24. fB =�t �xfB 2 5 .�xfBt ixfB 26 .WfxfB t 'it>c7 27.�d6t 'it>dB

4 3 2 1 a

b

e

d

e

f

g

h

I find it staggering that this move has only been played in four games. Presumably most players have felt reluctant to move the rook from its seemingly active post on the b-file, but my analysis indicates that it is more purposeful on d 1 , while White may yet be able to do without @h 1 . The first important question is how White should respond to the critical A) 16 ... ib4N. Once we have dealt with that move, we will consider B) 16 ... �g5 and C) 16 ... h3!? 1 6 . . . tDe5N can be compared with 16 . . . h3 1 7.g3 tD e 5 , as covered on page 337 in the note to Black's 1 7th move in variation C L White has good prospects there, and the absence of . . . h3 and g2-g3 only helps White, so I will not go into further details here.

8 7 6 5 4 3 2

a

b

e

d

e

f

g

h

2B .h3! �c6 29.�fBt 'it>c7 30 .�f4t! 'it>dB 3 1 .�e5 +- Black's pieces are absolutely stuck, and White can continue improving with @h l before finally taking the rook with ixd7 and �h8t . 1 7 . . . �e5 I B . tD e4! As I have explained before, White's key objectives are to avoid a queen

327

Chapter 1 8 - 1 6.�bd l exchange and increase the activity of his pieces. Here everything is ideal: 1 8 . . . .tc5 1 9 J:'!f3! Unpinning the e4-knight. 1 9 . . . b6 ( 1 9 . . . .ta7 is met by 20.@h l , intending �a3 , when White has as good a position as he could hope for in this line.)

avoid the queen exchange, although White's position might still be preferable.

lS ...h3 1 8 . . . liJ xf6? is impossible due to 1 9 . 1iJ e4!. 18 ... .txc3 1 9 .�xc3 0-0 cannot be refuted immediately but, without the dark-squared bishop, Black's position can't be good. 20.liJf3 �xf6 2 1 .:B:d4 �g7 22.liJxh4 With a clear initiative.

8 7 6 5 4

19JHl!

3

A typical way to respond to the pawn's arrival on h3.

2 b

a

e

d

e

f

g

19 ... hxg2 20.ttl e4 Y;Ye5 2 1 .Y;Yb3!

h

20.liJxc5! A rare case when White's initiative carries over into the endgame: 20 . . . �xe3t 2 1 .:B:xe3 liJxc5 (2 1 . . . bxc5 22.liJc6 .tb7 23 . .th5! +-) 22.liJ f5 .td7 23.liJd6t @f8 24 .liJc4 White has a great deal of activity for a mere pawn.

A nice touch, although a simple move like 2 1 .c3 would also give White a promising position .

2 1 .. ..if8 2 1 . . .a5 22.c3± does not change much. 2 1 . . . .tc5 22.liJxc5 liJ xc5 23 .�b4 leaves the knight paralysed due to the potential mate on e7. 23 . . ..td7 24.liJb3 :B:c8 2 5 . liJ a5 Black has no defence. 8 7 6 5 4

a

b

e

d

e

f

g

h

lS.�f3!

Intending liJ e4.

1 8 .�e4 .tc5 ! ( 1 8 ... .txc3 allows 1 9 .1iJxe6! . ) 1 9 .mh l h 3 ! ( 1 9 . . . �e5 ? 20.�xh4±) Now Black is ready for . . . �e5 when it will not be easy to

3 2 1 a

b

e

d

e

f

g

h

22.ttlxe6! fxe6 23 ..ih5t! cj;>dS 23 . . . �xh5 24.�xe6t @d8 2 5 .:B:d5 wins.

1 2 . g5

328

. .

After checking the above lines, I initially thought this was the end of the road for 1 6.�bd 1 . However, after analysing the 1 6. � h 1 1Mfg5 1 7.1Mfd3! ? line from the previous chapter, I suddenly realized White could put the queen on d3 here as well. It is always nice to find a promising move that does not appear in the top three suggestions of the computer - something that happens quite often in these positions. When comparing the present variation to the similar line in the previous chapter, you will notice that the rook turns out to be better placed on d 1 .

24.YlYb6t YlYc7 25.YlYxe6 Black is defenceless.

B) 16 YlYg5 •••

8 7 6 5 4 3 2

17 tLlc5

1

1 7 . . . lDxf6? doesn't work of course: 1 B .1Mff3 .ic5 ( 1 B . . . �gB 1 9 .�f2! +-) 1 9 .�h l 0-0 20.1Mfxf6±

•••

a

b

e

d

e

f

g

h

This is the most natural continuation, and usually White's biggest problem in similar positions.

17.YlYd3!N Initially, I had considered Black's last move to be the 'refutation' of 1 6J�bd 1 , as neither of the two correspondence games showed much promise. 1 7. lD e4 was played in Nilsson - Mendl, corr. 20 1 2. White should not be in too much danger after the queen exchange, but I can't envisage him fighting for an advantage, and I prefer to avoid trading queens unless it involves a real concession from Black. 1 7 .�f4 seems tempting too, but 1 7 . . . e5 proved perfectly fine for Black in Kasperek - Domanov, Internet 20 1 0. It is worth mentioning the following brilliant idea: 1 8 . lD e6!?N (Also after 1 8 .lDd5N .ic5! 1 9 .cj;Jh 1 exd4 20.lDc7t cj;Jf8 2 1 .lDxa8 lD xf6 Black is doing well.) 1 B . . . fxe6 1 9 .1D e4 This is such a devilish idea, but after 1 9 . . . exf4! ( 1 9 . . . 1Mfxf4? 20 . .ih5 t cj;JdB 2 1 .1Mfa5t leads to a quick mate) 20.lDxg5 hxg5 2 1 .1Mfc3 �f7+ Black is safe.

1 7 . . . �gB 1 B .�f2! does not change anything, and after 1 7 . . . lD c5 1 9 .1Mfc4! .id7 20 . .if3 the play is similar to the main line below. 1 7 . . . lDe5 places the knight rather awkwardly. Now 1 B .1Mfh3� intending lD e4 is the simplest, but I B .1Mfe4! is even better and leads to an advantage by force: I B . . . h3 1 9 .�f5! ( 1 9 .g3�) 1 9 . . . exf5 20 .1Mfxe5t .ie6 2 1 ..if3 f4 22 .1Mfc7 .ic5 23.lDe4 �c8 24.1Mfxb7 1Mfe5 25.lDxc5 1Mfxc5 26. cj;Jh 1 ±

a

b

e

d

e

f

g

h

329

Chapter 1 8 - 1 6.!!bd l

ls.Wlc4! An amazingly effective square for the queen! Now it is s�fe against immediate attempts to exchange, and most importantly it prevents . . . 0-0-0. Without the option of long castling, it will be hard for Black's king to feel safe.

lS .id7 •••

l S . . . b 5 ? was barely playable in the 1 6. h 1 line, and here it is an outright mistake. 1 9 .1Llcxb5 axb5 20.lLlxb5 We5 2 1 .if3 1':1:a6 22.1':1:f2 1':1:d6 23.lLlxd6t ixd6 24.Wxh4± l S . . . h3 can be ignored by 1 9.if3!, for instance: 19 .. .'IWe3t?! (Black should prefer 1 9 . . . id7 transposing to variation B 1 below; 1 9 . . . hxg2? should definitely be avoided, as 20.1':1:fe l ! id7 2 1 .ixb7! lLlxb7 22.lLlxe6 wins outright.) 20.1':1:f2! id7 2 1 . lLl f5 ! We5 (2 1 . . .exf5 22.lLld5 We5 23.1':1:e2+-) 22.lLlg7t d S 23 .1':1:e2 Wc7 24.lLl f5±

1 9 . . . h5 20. h 1 1':1:h6 is an attempt to eliminate the f6-pawn, but it is refuted by 2 1 .ixb7! lLlxb7 22.lLlxe6! fxe6 23.f7t d S 24.Wc6 We7 2 5 .Wxb7 and White wins. 1 9 . . . 1':1:cS 20.We2 sees the queen settle on an ideal square, while Black has been prevented from castling. It is hard to suggest a plan for Black, for instance: 20 . . . h5 (20 . . . id6 2 1 .lLl f5 if4 22.lLlg7t f8 2 3 . lLl e4 We5 24. h l !±) 2 1 . h 1 1':1:h6 (2 1 . . .ih6 22.lLld5±) 22.lLld5 Black is not yet losing by force, but the position looks horrifyingly dangerous for him.

Bl) 19 h3!? ...

8 7 6 5 4 3 2 1 a

b

e

d

e

f

g

h

Considering how hard It IS for Black to develop, trying to create a mess on the kingside makes sense. a

b

e

d

e

f

g

h

19 .i6 .

After 1 9 .1':1:f2 1':1:cS! the queen has no place to hide, but the text move frees the e2square. Here I analysed two main moves: Bl) 19 h3!? and B2) 19 Wle3t!. •..

•••

19 ... 0-0-0? 20.1':1:b l ! is too dangerous for Black, despite the loss of time with the rook, for instance: 20 . . . b S 2 1 .1':1:xb7t! lLl xb7 22.Wxa6+-

20J:tf2! The only way to fight for an advantage. This avoids a check on e3 while keeping the f6pawn protected. 20.g3 We3t! 2 1 . h 1 b5 is perfectly fine for Black, for instance: 22.Wb4 (22.lLld5 We5 23.lLlxe6 ixe6! 24.Wb4 id6 2 5 . lLl e7 ixe7 26.ic6t fS 27.fxe7t xe7 2S.1':1:fe 1 a5 !+) 22 . . . lLl e4 23 .Wb3 Wlxc3 24.ixe4 1':1:cS!=

1 2 . . g5

330

.

22 �g8

20 �d8

••J

•• J

The rook is better placed here than on c8.

20 . . . hxg2 2 1 .J.xb7! �a7 (2 1 . . .lLlxb7? loses to 22.lLlxe6) 22.J.f3t 20 . . . �c8 ?! 2 1 .Wfe2 h5 (2 1 . . .�g8 22.g3 �g6 23.lLldSt; 2 1 . . .md8 22.J.h S !t) 22.mh 1 md8 Trying to j ustify the placement of the rook on c8. 23.gxh3 mc7

22 . . . J.h6 23 .�e2!+8 7 6 5 4 3

8 7

2

6

1

5

a

4

b

e

d

e

f

g

h

23.�hl gg6

3

23 . . . J.d6 24.gxh3;!;

2

24.gxh3! b

a

e

d

e

f

g

h

24 .h4! Wfxh4 2 S .WfeSt mb6 26.�b 1 t ma7 27. lLl e4 The attack is irresistible.

24.lLldS J.d6 2 S . gxh3 J.c6 26.lLlxc6 bxc6 is not so clear.

24 J.d6 •••

24 . . . �xf6 2 S . lLl d S ! is winning. 8 7 6 5 4 3 2 1 a

b

e

d

e

f

g

h

2 1 .Wie2 h5 2 1 . . .�g8 22.g3 ±

22.Wiel ! Anticipating the plan of . . . J.h6 and . . . Wie3 . Black's position is on the verge of crumbling.

a

b

e

d

e

f

g

25.h4! Wie5

2S . . . Wfxh4 26.lLl f5 +-

26.ge2 Wixf6 27.hh5 White has a formidable initiative.

h

33 1

Chapter 1 8 - 1 6J':'!:bd l

B2) 19 ti'e3t!

8

•••

7 6 5 4 3 2 b

a

a

b

e

d

e

f

g

h

Since the other lines all leave White with a strong initiative, this direct attempt to exchange queens seems best. But even here, Black's defence is incredibly hard, even with a computer!

e

d

e

f

g

h

25 . . .:!:k7 26. ltJ d6t i.xd6 27J:!:Xd6 0-0 (27 . . ':!::!:xc 2? 2 8 .�ed3+-) 28.�d4 �fc8 29.�g4t �f8 30.�xh4 �xc2 3 1 .h3t White has serious threats against the black king, in addition to the long-term potential of the h-pawn.

22.ti'b3 ti'xc3 22 . . . ltJ c5 is well met by 23 .Wb l ! intending ltJ d 5 .

20.�hl b5

23 .be4 ti'xb3

Now White has the pleasant scenario of being able to choose between B21) 21 .ti'b4, which leads to a safe endgame with two possible results, and B22) 2 1 .c!LJd5!? leading to wild complications.

24.cxb3 0-0-0

•

23 . . . �c8 24.ltJf3;!;

Now White has a few decent ways to continue.

2 1 .ltJ dxb5 axb5 22.ltJxb5 is almost winning, but after the accurate 2 1 . . .:!::i a4! 23.ltJc7t �d8 24.:!::!:xd 7t �xd7 2 5 JWb5t �xc7 26.Wc6t c;t>d8= White is left to choose between many different perpetuals.

B21) 21 .ti'b4 This simple move, without trying anything spectacular, is a decent practical try.

2 l . . .�c8 22.�fe l ltJ a4 23.�xe3! i.xb4 24.ltJxa4 bxa4 2 5 . ltJ f5 leads to a one-sided endgame, for instance:

a

b

e

d

e

f

g

25.tLlf3 This is perhaps the most logical try.

h

332

1 2 . . . g5

2 5 .a4 looks tempting, as Black cannot afford to open up the queenside, but the computer indicates that 25 . . . i.a3! is not easy to crack. 26.lLlc2 i.c5 27J�d3 b4 28Jk l Wb8 2 9 Jkd l Wc7 Black is holding everything together.

8 7 6 5 4 3

A more promising attempt could be: 2 5 Jk l t Wb8 26. lLl c6t i.xc6 27J'hc6

2

8

b

a

e

d

e

f

g

h

32.Wfl E:d2 33.E:a5 Wc7 34.E:xb5 i.d4 35.a4 Black should be able to hold, but White can keep pressing with no risk.

7 6 5 4

25 rJ:Jc7 26.E:f2! •••

3 2

a

b

e

d

e

f

g

h

Swapping down to an opposite-coloured­ bishop position doesn't look ambitious, until you notice that White can win the queenside pawns by force. I have not analysed this endgame in much depth since it is too deep, with many options for both sides. Black should be able to equalize with correct play, but it is obvious that White's queenside passed pawns have the potential to be more dangerous than Black's disjointed kingside pawns. 27 .. J�d6 28J�fc 1 E:xc6 29.E:xc6 i.b4 29 . . . a5 30.E:a6:t 30.E:xa6 E:d8 3 1 .Wgl 3 1 .E:a8t Wc7 32.E:xd8 Wxd8 33 .i.c6 i.c3 34.i.xb5 i.xf6 is equal. 3 1 .g3!? intending Wg2 is similar to the main line. Here too, I believe Black can equalize with accurate defence, but a lot of play remains. 3 1 . . . i.c5 t Also after 3 1 . . .E:d4 32 .i.£3 i. c 5 33.Wfl Black still has to prove the draw.

A sweet way to bring the rook into play. The less accurate 26.E:d2 can be met by 26 . . . i.a3 , stopping the second rook from coming to the c-file.

a

b

e

d

e

f

g

h

26 Aa3! •••

Amazingly enough, this is the only move.

27J�c2t rJ:Jb6 28.E:cd2 rJ:Jc7 29.�e5 i.c8 30.E:xd8 30.g4!? This is another possible try. 30 . . . hxg3 30 . . . E:xd2 3 1 .E:xd2 E:f8 32.
Chapter 1 8 - 1 6.Ei:bd 1 up the h-file. 32 ... id6 33.Ei:c2t 'itbB 34.lt:k6t 'itc7 35 .'ith3 'itb6 36.'itxh4 ib7 37.'ith5;!; 3 1 .Ei:c2t! 3 1 . hxg3 Ei:xd2 32.Ei:xd2 Ei:ffi 33.'kt>g2 id6 34.Ei:c2t 'kt>bB 3 5 .�c6t 'kt>b7 is equal. 3 1 . . .'kt>b6 32.Ei:xd8 Ei:xd8 33.�xf7 Now Black needs to be precise.

32 ... �e8 33.�xh6 Ab2 34.�g4 �f7 35.g3!? hxg3 36.h4 .ixf6 37.�xf6 �xf6 38.b4! Black is still suffering.

B22) 21.�d5!? 8

8

7

7

6

6

5

5

333

4

4

3

3

2

2

1 b

a

e

d

e

f

g

h

a

33 . . . Ei:d l t! 33 . . . Ei:fB 34.�e5± 34.'kt>g2 gxh2 3 5 . 'kt>xh2 ib7 36.ixb7 'kt>xb7 37.'kt>g2 iffi 38 .Ei:e2 Black should be able to hold, but the advanced f-pawn makes the position a bit unpleasant for him.

b

e

d

e

f

g

h

This is the most critical move, and it leads to some amazing ideas.

21. .'!We5 •

2 1 . . .bxc4 22.�xe3 0-0-0 23.�xc4;!; gives White a pleasant endgame.

22.�xe6!!

30 .. Jhd8 3U�xd8 �xd8 32.c�� xf7t It is possible to analyse this endgame in more detail but, since this is only one out of many options available to White, I will j ust show a possible line where White maintains some pressure.

A beautiful idea, although Black has an equally miraculous defence.

8 7 6 5 4 3 a

2 1 a

b

e

d

e

f

g

h

b

e

d

e

f

g

h

1 2 . . . gS

334

22 ....be6! Both 22 . . . lLlxe6 23.'lWxh4 icS 24.l%fe l 'lWb8 2S.lLlf4 and 2 2 . . . bxc4 23.lLldc7t 'lWxc7 24.lLlxc7t \t>d8 2S.lLlxa8 are winning for White.

23.�xh4 O-O-O! Another possible pitfall for Black is: 23 . . . l%d8 24.l%fe l 'lWb8

27. lLl f4! l%xd l 28.l%xd l 'lWb6 28 . . . id6 29.lLlxe6 lLl xe6 3o.ic6t \t>d8 3 1 .'lWb4+29.lLlxe6 lLlxe6 30.'lWe4 id6 3 1 .ihS \t>f8 32.l%a l !± The rook will hurt Black on the eighth rank. 8

8

7

7

6

6

5

5

4

4

3

3

2

2

1 a

b

e

d

e

f

g

h

2 S . a4!! In the event of2 S . lLl b4 l%d6 Black is surviving somehow, while after 2 S . lLl f4 l%xd l 26.l%xd l id6! Black is ready to castle. Amazingly, the text move leaves Black absolutely paralysed, and the opening of a queenside file is enough to break the defence. 2S . . . l%g8 2S . . . lLl xa4 26. lLl c7t! followed by l%xe6t is crushing. 2S . . . bxa4 26.l%xe6t! lLl xe6 27.'lWxa4t l%d7 2 8 . lLl b6 wins in great style. 2 S . . . hS also does not help: 26.axbS axbS 27. lLl b4 l%d6 28.l%a l ! +26.axbS axbS

a

b

e

d

e

f

g

h

24.fll b 6t rJlc7 25.l%xd8 rJlxd8 26J�e1 �c7 27.fll a8!? An amazing, insane-looking way to continue. After 27.l%d l t lLl d7 28.lLlxd7 ixd7 White does not seem to have any more than a perpetual, despite the many tempting possibilities. 29 .We4 id6 30 .'lWa8t ic8 Now there are various ways to make a draw, for instance: 3 1 .WdS l%e8 32.'lWxd6t 'lWxd6 33.l%xd6t \t>c7 34.l%c6t c,td7=

8 7 6 5 4 3 2 a a

b

e

d

e

f

g

h

b

e

d

e

f

g

h

Chapter 1 8 - 1 6.!!bd 1

27 Y;Ya5 28J�dl t �c8 • • •

We have reached a fascinating position . It is hardly p�ssible to analyse every possibility here, but the general evaluation is that Black is walking on the edge, while White can consider a few plans and ideas. At the very least, this doesn't seem a pleasant position to play over the board for Black, as a single misstep could prove costly; for instance, if he allows a file to be opened on the queenside. He will therefore have to spend most of his time playing waiting moves, or perhaps aiming for a conservative level of kingside counterplay.

29.Y;Yg3 29 .Wf4 is similar and may transpose.

335

Mter 3 1 .c4 b4! it is hard for White to improve, although Black also does not have much to do besides waiting. 3 1 . h3 would be desirable, but Black is j ust in time with 3 1 . . . E:gB! intending . . . E:g3 .

3 1 . h3 •.

3 1 . . .b4? 32.E:d4! 33.E:xb4+-

j ust

wins:

32 . . . E:h6

3 1 . . .llJxa4? is met by 32.We4! and the queen entry will be lethal. 3 1 . . .E:h6 32.h3 E:hB (32 . . . E:g6 33.axb5 axb5 34.E:a l +-) 33 .Wd4! White intends axb 5 and E:a l .

29 h5! .••

29 . . .ixa2 30J'h l ! is awkward for Black.

8

If Black plays too slowly then White will play We5 and, sooner or later, break open the queenside with a2-a4.

7

30.Y;Ye5 h4

4

30 . . . E:gB can be met by: 3 l .c3 ! ? (3 l .a4 E:g4! is an annoyingly strong idea!) 3 1 . . . E:g4!? Otherwise White can play h2-h3 and later a2-a4. 32.id5�

6 5

3 2 1 a

8

b

e

d

e

f

g

h

32.g4 E:g8

7

32 . . . E:h6 leads to a similar situation after 33 .Y;Yf4!? intending g4-g5 . Black is limited to shuffiing around with his rook, while White has many possible moves, but no clear way to break through.

6 5 4 3

33.g5 E:h8

2 1 a

31 .a4!?

b

c

d

e

f

g

h

Black's position appears to be on the edge, but White has no obvious way through. The game is in something of a deadlock, although White can shuffie around a bit with c2-c3 , E:d4, Wf4/g3 and so on. I guess it's possible to

1 2 g5

336

. . .

explore this line in more detail, in the hope of finding a nuance to improve White's prospects. For the time being through, the above line is enough to cause serious practical problems for the defender.

weapon. However, as the theory unfolds in the future, some lines might become neutralized, or at least better known. Before discussing these three options, I will mention a few other moves that I discarded.

C) 16 ... h3!?

White's final option can be understood when you consider the variation 1 7.'lWe3 'lWeS ! , when it is hard to hold on to the f6-pawn. 1 8 .'lWxh3 (Or 1 8 . lLl e4 lLlcS! 1 9 J!f4 id7 20.lLlf3 'lWc7 intending . . . 0-0-0 and Black is doing just fine.) This was Isaev - Tritt, email 20 1 0, and now 1 8 . . . lLlxf6 1 9J!fe l 'lWgS , intending . . . ie7 and . . . 0-0, seems to be okay for Black.

a

b

e

d

f

e

g

h

This is already a good sign for White - we will easily avoid an early queen exchange, and we are entering a whole new world of variations and ideas. There are only two correspondence games in the database here, so there is a great deal to explore. At the same time it will be difficult to draw many firm conclusions due to the sheer amount of possibilities for both sides. Whenever I discovered an interesting idea for White, there always seemed to be an ingenious defensive idea in the position, and the complicated battle raged on. Instead of recommending a single move for White, I will present a few different tries and identify what I think are the most critical lines that may be developed with new games and analysis. We will begin by analysing the blunt Cl) 17.g3, before considering the more fluid options of C2) 17 .tg4!?N and C3) 17.YlYd3!?N. Considering the dearth of games after White's 1 6th move, any one of these moves could serve as an excellent practical •

1 7.E:f4N is a crazy version of the next line below, but here Black has a forcing defence: 1 7 . . . ib4! All other moves give White an excellent initiative. 1 8 .'lWe3 'lWgS ! 1 9 .93 lLlxf6 20.lLle4 lLlxe4 2 1 .'lWxe4 icS Intending . . . 'lWdS , and Black i s doing well. 1 7 .E:f2N The idea of this move is to be ready for . . . hxg2, but Black has no reason to play that. 1 7 . . . lLl e S ! Now it will b e hard for White t o stop the plan of . . . id7 and . . . 0-0-0. 1 7 . . . hxg2? 1 8 . lLl b3! gives White a crushing initiative, but I will spare the further details as Black can do much better than this. 17 . . . ib4?! 1 8 .'lWe3! is strong, and 1 7 . . . hS 1 8 .'lWe3 'lWeS 1 9 . 1Ll e4 ics ( 1 9 . . . ih6 20 .'lWd3) 20.if3 also favours White. 8 7 6 5 4 3 2

a

b

e

d

e

f

g

h

Chapter 1 8 - 1 6J:i:bd 1 I B .lLl b3 I B .'lWe3 �d7 1 9 .'lWxh3 is a possibility, but Black's king will be safe now, and his bishop pair will offer him better chances. I B . . . 'lWc7 1 9 .i.b 5t! lLl d7! 19 . . .i.d7 20.i.xd7t lLlxd7 2 1 .lLld5 'lWd6 22 .'lWe3 gives White some initiative. 20 .i.e2 l:l:gB!? 20 . . . lLle5= repeats the position, but the text move tries for more than equality. Either way, this is not what I want to achieve from the opening.

Cl) 17.g3 8 7 6 5 4

337

1 7 . . . 'lWg5 ? is bad due to: I B .l:l:f4 lLl xf6 ( 1 B . . . e5 1 9. 1Ll e4+-) 1 9.1:l:xf6 'lWxf6 20.lLle4 'lWdB 2 1 .'lWe3 i.e7 22.lLl f5 ! 'lWxd l t 23 .i.xd l exf5 24.lLld6t @f8 2 5 .'lWc3! ± 1 7 . . . lLle5 This has been played in several correspondence games, so it might well be the choice of an over-the-board opponent too. I B . lLl f3 I B .'lWe3 i.d7! 1 9 .1Llxe6 ( 1 9 .lLl e4 0-0-000) 1 9 . . . fxe6 20.f7t lLl xf7 2 1 . lLl e4 0-0-0 22.l:l:xf7 i.eB 23.l:l:xdB t 'lWxdB seems okay for Black. 1 8 . . . lLl c6 I B . . . i.b4? does not work due to 1 9.1Llxe5 i.xc3 20.'lWd6 i.b4 2 1 .'lWd3 'lWd5 22.i.f3 'lWxd3 23.l:l:xd3 0-0 24.i.e4!, intending l:l:f4 with a winning attack in the endgame. However, after the text move White really does have to worry about the bishop's arrival on b4. 8

3

7

2

6

1

5 a

b

e

d

e

f

g

h

This could be considered the most natural reply - even though we have seen several instances where White j ust ignores the pawn on h3 . The real question now is who will benefit from the inclusion of . . . h3 and g2-g3 . Incidentally, this position has been reached via transposition a few times, via the move order 1 5 . . . h3 1 6.g3 (in variation B 1 of the previous chapter I recommended 1 6.i.f3!?) 1 6 . . . lLl d7 1 7.l:l:bd l .

17 ..i.b4!N .

Surprisingly, all of the existing correspondence games - which are usually known for their precision - missed this idea.

4 3 2

a

b

e

d

e

f

g

h

1 9.a3!N 1 9 . 1Ll e4 was played in three correspondence games, but the queen exchange obviously makes Black's life easier. 1 9 .'lWd3 lLl b4! and 1 9.1:l:b l i.d7! both seem fine for Black. 1 9 . . . i.xa3 1 9 . . . b5 j ust gives White easy play, while Black will struggle to complete development: 20.@h l l:l:a7 2 1 .'lWe3! l:l:d7 22.l:l:xd7 i.xd7 23.l:l:d l t

1 2 . . g5

338

.

20.V9d3 Now the . . . tLl b4 idea is less effective as the bishop will be stuck on a3 . In most lines Black will have to waste time returning his bishop to f8, and in such positions a tempo may prove far more important than a pawn. 20 . . . V9c7 20 . . . e5 2 1 . tLl e4 .ie6 22.Vge3 .if8 23.c4� Black's king is stuck in the centre, and White has excellent prospects of an initiative. 8

24 . . . 0-0-0 24 . . . V9a7 2 5 .Vge4t 2 5 . tLl a4t White has various attacking ideas including :!:!c3 , :!:!db3 and a possible bishop sacrifice on b 5 . (25 . tLl xb 5 ! ? axb5 26.:!:!c3 looks quite dangerous as well.) Let's now return to the more accurate 1 7 . . . .ib4!N, which can be compared with the earlier variation A. As we will see, the inclusion of . . . h3 and g2-g3 proves useful for Black here.

7 6 5 4 3 2

a

b

e

d

e

f

g

h

2 1 .Vge3! ? 2 1 . tLl e4 .iffi is similar, but White does not need to commit the knight j ust yet. 2 1 . . . .id7 22.:!:!d3 .iffi 23.:!:!b l ! b5 Both 23 . . . 0-0-0 24.tLla4 and 23 . . . :!:!d8 24.tLld5 V9a5 2 5 .Wh l .ic8 26.:!:!bd l see White maintaining some initiative. 8 7 6

a

b

e

d

e

f

g

h

1 8.Vge3 .bd! Without the aforementioned pawn moves this would be losing, but here Black can get away with it. Instead 1 8 . . . V9g5 1 9 .Vge4 would give White plenty of attacking options.

19.t2he6 fxe6 20.%Yxe6t Wd8 21 .%Ye7t Wc7 22.%Yd6t Wd8 23.f7 .ig7 24 .ig4 %Yb5 25.f'8=%Yt :!:!xf8 26J�xf8t .ixf8 27.%Yxf8t Wc7

5

•

4 3 2

a

b

e

d

e

f

g

h

24.Wfl !! 24.tLld5 V9a7!+ is annoying, while after 24.Wh l 0-0-0 the following sacrificial ideas do not work.

White's problem is that his king is too exposed: he can never go after the a8-rook since Black will have at least a perpetual, if not more. Therefore he has to content himself with:

Chapter 1 8 - 1 6J'�bd 1

339

17 ... .ib4 18 . .ixe6! .ixc3 dangerous for Black.

8 7

1 9.'?Me3

is too

C2 1) 17 hxg2

6

.•.

5

8

4

7

3

6

2

5 b

a

e

d

e

f

g

h

2S.Y;Vd6t c,t;.>dS 29.Y;VflJt= We may conclude that 1 7.g3 is a reasonably safe move that could easily catch an unprepared opponent, but it is not too difficult for Black to find a path to safety. In the remaining part of the chapter we will focus on a couple of surprising but extremely dangerous options available to White.

C2) 17 .ig4!?N •

4 3 2 1 a

b

e

d

e

f

g

h

ISJU4! Of course the pawn has to be ignored, but this looks insane. Still, White's pieces come together perfectly. 1 8 .1'%£2 is less precise due to 1 8 . . . lLle5 1 9 . .ih5 .id7 when Black will be able to castle.

I S tLle5 •••

1 8 . . . .ib4 is met by 1 9.'?Me3 '?Me5 20.1'%e4 '?Mg5 2 1 .lLld5! '?Mxe3t 22.1'%xe3 and Black will lose material.

19.ttlf3!

Challenging the strong knight. 8 7

b

a

e

d

e

f

g

h

With this move White shields the g-pawn against . . . '?Mg5 ideas, while setting up various sacrifices on e6. Black may consider C2 1) 17 hxg2, C22) 17 ... Y;Vg5!? or C23) ••.

17 ... ttle5!.

6 5 4 3 2 1 a

b

e

d

e

f

g

h

340

1 2 . . gS .

19 tihf3t! •••

Mter 1 9 . . . ttJ xg4 20J'hg4-+ it is impossible for Black to develop. 1 9 . . . ttJ c6 20 J::k4 ! also looks dangerous. A few variations to illustrate my point: 20 . . . .ib4 (20 . . . Wfb6t 2 1 .xg2 l:!g8 22.h l ± ; 20 . . . l:!g8 2 1 .Wfe3 .id7! 22.ttJdSt)

Black's position looks so shaky, but he can survive with a crazy-looking line.

24.fxe7 �hg4t 8 7 6

8 7

5

6

4

5

3

4

2

3

1

2

a a

b

e

d

e

f

g

h

2 1 .Wfe3! .id7 (2 1 . . . .ixc3 22.l:!xc3 .id7 23.l:!xc6 bxc6 24. ttJ e S +-) 22.ttJdS With a dangerous ini tiative.

b

e

d

e

f

g

h

25. c;t>f3! a!h4! The only move, as is Black's next.

26.ttJe4 a!h3t! 27.c;t>e2 a!xh2t 28.tLlfZ Black is still in some danger.

20.l:!xf3

e22) 17 .'I�Yg5!?

20 . .ixf3 can also be considered. After the text move it is hard for Black to develop, so he has to try and make use of the pawn on g2.

•.

20 l:!g8 .•.

20 . . . .icS t 2 1 .�xg2 l:!g8 22.l:!g3 ±

2 1 .l:!g3 a!g6 2 1 . . .l:!gS can be met by 22.h4.

22.c;t>xg2 It would be possible to spend a lot of time analysing this position, but I will j ust present a few variations to demonstrate that White's initiative is real.

22 Wfc7 •••

22 . . . l:!xf6 23 . .if3�

23.a!d3! .le7!

a

b

e

d

e

f

g

h

Chasing the queen is a typical idea in these positions.

18.'iNe2 'iNe5 19.'iNfZ! Other moves are not so dangerous:

34 1

Chapter 1 8 - 1 6J�bd l 1 9 .tLle4 tLlc5 20.if3 id7 2 1 .Wh l 0-0-0 is safe for Black.

24.h4!±) 24.l'!d4! White's initiative continues. 8

1 9 .�d3 is met by 1 9 . . . ic5! when it is hard to untangle for White. 20.tLle4 ib6! (20 . . . ixd4t 2 1 .�xd4 �xd4t 22.l'!xd4 hxg2 23.l'!fd l;:!;) 2 1 .ixh3 tLlc5 22.tLlxc5 ixc5 again . . . ixd4 is a threat. 23.Wh l (23 .c3 id7! 24.W h l 0-0-0+) 23 . . . ixd4 24.�xd4 �xd4 2 5 .l'!xd4 id7=

7 6 5 4 3 2

19 ...i.c5!? A slow move such as 1 9 . . . ib4? will be blown away by 20.ixe6! fxe6 2 1 .l'!fe l �h5 22.tLlxe6 cj;Jf7 23 .�d4! with a winning attack. 1 9 . . . hxg2 This just gives White a typical initiative. 20.l'!fe l �xf6 2 1 .�e3 2 1 .if5 ! ? could be dangerous as well. The text move intends tLl e4, tLl e6 or ie6, and Black will have to defend carefully to hold his position together.

b

a

e

d

e

f

g

h

23 .ixe6! fxe6 23 . . . 0-0 24. tLl e4 ixd4t 2 5 . l'!xd4 �e5 26.l'!xd7 ixd7 27.ixd7± 24.�xe6t Wffi There is no mate, but White can win back the piece. 2 5 . tLl e4 ixd4t 26.l'!xd4 '?;Ve5 27.tLld6! '?;Vxe6 2B.l'!xe6 tLl f6 29.l'!xf6t± White's active pieces give him the more pleasant endgame.

8 7

8

6

7

5

6

4

5

3

4

2

3 a

b

e

d

e

f

g

h

2 l . . .�g5 22.�e2! 22.tLld5 �xe3t 23.l'!xe3 id6 24.tLlxe6 tLle5 is just a draw. 22.�h3! ? is also interesting, but after 22 . . . tLle5 23.tLle4 �f4 24.ixe6!? fxe6 25 .�h5t tLl f7 26. tLl f5 ie7! I found nothing more than a forcing line leading to an equal endgame. 22 . . . ic5 After 22 . . . tLlc5 23.tLlf5 l'!gB (23 . . .id7

2 1 a

b

e

d

e

f

g

h

20J�fe1! 20.ixh3 would b e thwarted b y 20 . . . '?;Vxf6 (20 . . . tLl xf6? 2 1 . tLl a4! ia7 22.tLl b2+-) 2 1 .'?;Ve3 '?;Vg5 ! and we cannot avoid the queen exchange.

20 ...'i1Yxf6 2 1..1£3 Threatening tLl e4 .

1 2 . g5

342

. .

21 ...0-0

24 ... �d7 25.�f4 e5

2 1 . . .�g6? is bad due to 22.lLle4 0-0 23.lLlxc5 lLl xc5 24.1:'! e 5 ! lLl d7 2 5 .l:!h5! hxg2 26.l:!d2 with a winning attack. 2 1 . . . .ie7 leads to some fireworks after 22.lLle4 �g6 23.lLlxe6! fxe6 24. lLl d6t �dB . Here is an illustrative line, where all of Black's moves are forced, though White has some options along the way:

25 . . . lLle5 26.l:!e3! leads to a shockingly swift and deadly attack: 26 . . . hxg2 (26 . . . lLlxf3t 27.l:!xf3 �h7 2B.l:!g3! �f6 29.�e3+-; 26 . . . lt>hB 27 ..ie2 lLl g4 2B . .ixg4 �xg4 29.l:!dd3!±)

25 . . . lLlf6 can be met by 26.�h l ! ?, intending 26 . . . hxg2t (26 . . . e5 27.gxh3t) 27.lLlxg2 �hB 2B.l:!gl ..... and Black's position is collapsing.

8

8

7

7

6

6

5

5

4

4

3

3

2

2 b

a

e

d

e

f

g

h

b

a

2 5 .�d2!? e5 26.lLlxcB .ic5t 27.�h l l:!h7 2 B . lLl e7 l:!xe7 29.l:!xe5 hxg2t 30 . .ixg2 �d6 3 1 .l:!d5 �c6 32.l:!f5 �d6 33 .�xd6 .ixd6 34.l:!xd6;l;

e

d

e

f

g

h

27 . .ie2! �f6 (27 . . . lLl g6 2B .l:!g3 ..... ) 2B .l:!g3t �h7 29 .�d4+8 7

8

6

7

5

6

4

5

3

4

2

3

1

2

a

1 a

b

e

d

e

f

g

h

22.�e4 Y!!g7 23.�xc5 c!Llxc5 24.�e2

24.�h l ! ? will lead to similar positions. Black has two extra pawns, but White has excellent attacking prospects, especially since Black does not have his dark-squared bishop anymore.

b

e

d

e

f

g

h

26.�h5! 26.lLlxh3 lLl f6 27.�h4 also gives White compensation but the attack is slowed down. 26.lLld5 lLl f6 seems okay for Black.

26 ...Y!!g5 26 . . . �g6 27.�h4! maintains the attack.

Chapter 1 8 - 1 6J�bd l

343

27. c;f? h 1!?

C23) 17... �e5!

Mter 27.l'!e3 lLl f6 Black can defend. 8

27 hxg2t •••

Now 27 . . . lLl f6? is refuted by 28.gxh3 lLl xh5 29.l'!g 1 .

7 6 5

8

4

7

3

6

2

5

1

4

a

3

b

e

d

e

f

g

h

This seems to be the only way to avoid coming under a powerful attack.

2 1 a

b

e

d

e

f

g

h

2S.c;f?gl ! 2B . .txg2 �xh5 29.l'!gl c;f?h7 30 . .tf3 �f5 3 1 . �g2 � g5 leads to a draw, but the text move is an amazing way to keep the initiative.

2s lLlf6 ••.

Black must take great care, for instance: 2B .. .';t>hB 29.l'!d6! f6 30.h4 �g6 3 1 .l'!e2 wins. 2B . . . l'!eB 29.l'!d6! e4 30.l'!e3 with a dangerous attack.

l S .bh3 .i.b4! •

A good way to find defensive ideas in such positions (I mean the more subtle ideas that are not immediately suggested by the computer) is to check the most forcing continuations, as they do not allow time for the initiative to blossom. Here White is forced to simplifY to an endgame. I B . . . .td7 gives White a typical strong initiative after: 1 9. 1Llxe6! .txe6 20 . .txe6 .tc5t 2 U ��h l fxe6 22.�f4 lLl f7 23.lLle4 l'!gB (23 . . . l'!cB 24.�g3t) 24.l'!d6h 8

2B . . . �g6 29 .l'!e2! c;f?hB 30 .�e3 Black remains under pressure, for instance 30 . . . �b6 3 1 .�xb6 lLlxb6 32.l'!d6! and wins.

7

29.h4 YlYg6 30J::lxe5 ttlg4

5

30 . . . lLlxh5 3 1 . .txh5 +-

4

31 ..bg4 .ixg4 32.ttlf6t c;f?hS 33.l'!d4± This line was not completely forced, but it certainly highlights some of the dangers for Black. The next move is the most resilient of all.

6

3 2 1 a

b

e

d

e

f

g

h

1 2 . . . g5

344

19.�b3 V!ic7 20.�b5 V!ib6t 2 1 .V!id4 V!ixd4t 22J:�xd4 axb5 23.gxb4 Black seems to be almost out of the woods, but he still has to be careful. 8 7 6

27.�c4! Forcing the knight to go to c5 , where it will be further away from the kingside defence. 27.1L1 b5 0-0 28 .�g4t �h8 29.�h4 cj{h7 is not easy to break. 27 . . . 1L1 c5 28.1L1b5 0-0 29.�g4t �h8 30.�h4 Black is under pressure, especially considering the following idea: 30 . . . �h7 3 1 . 1L1 d6 .ta6

5

8

4

7

3

6 5

2

4

1 b

a

d

c

f

e

3

h

g

2

23 ...i.d7! 23 . . . �xa2 The computer indicates that this is j ust as good, but a closer investigation shows that it does not fare as well. 24J:hb5 lLl d7 After 24 . . . 1L1 c6 2 5 J:k5 the pawn on f6 ensures White's easier play. 2 5 . 1L1 d4 b6 26J:l:b4!? Preparing to switch the rook to the kingside in anticipation of castling.

b

a

e

d

e

f

g

h

32 . .tf5 t ! ! �g8 32 . . . exf5 33.1L1 xf5 �e8 34.�xh6t �g8 3 5 . 1L1 e7t �xe7 36.fxe7 wins. 33.c4 exf5 34.1L1xf5 .tb7 3 5 . 1L1 xh6t �h7 36.1L1xf7t White continues to press without any risk. 8 7

8 7

6

6

5

5

4

4

3

3

2

2

1 a

b

e

d

e

f

g

h

26 . . . i.b7 26 . . . h5 27.1L1b5 0-0 28 .i.g4! �xc2 (28 . . . hxg4 29.�xg4t �h8 30.�f3 +-) 29 . .txh5 t

a

b

e

d

e

f

24J�e1 24.1L1d4 1L1 c6! is fine for Black.

g

h

Chapter I B - 1 6.l:!bd 1

24 c!tlc4

34S

where it could be harassed by . . . WleS , as noted earlier on page 336.

•..

24 . . . c!tlg6 , 2 S . tLl cSt

25.c!tld4 �ha2 2S . . . tLld6 gives White various options including tLl f3-eS, or transferring the b4-rook to the kingside. Black might still be okay, but it is not an ideal situation for him.

Once again, Black has three main candidates, and we will analyse C3 1) 17 ttl e5, C32) 17 VNe5 and C33) 17 ... hxg2! in turn. •••

.••

1 7 . . . .icS leaves the queen cut offon as. I B . tLl e4 eS 1 9J�fS ! exd4 (After 1 9 . . . .ixd4t 20.Wlxd4 exd4 2 1 .gxaSt White's initiative continues in the endgame.) 20.tLlxcS tLl xcS 2 1 .Wlxd4 .ixfS 22 .Wld6 WldB 23 .WleSt tLl e6 24.gxd8 t gxd8 2 S .WlxfSt The f6-pawn is strong.

C3 1) 17 ttle5 •••

8 7 6 a

b

e

d

e

f

g

h

5

26.c!tlxb5 .L:b5 27J�xb5 gxc2 28J��xb7 0-0

4

There is still some play, but it should be a draw.

3

C3) 17.VNd3!?N

2 1 a

b

e

d

e

f

g

h

This is quite likely to be the reaction of a surprised opponent, as it enables Black to castle on the queenside in the near future. Nevertheless, it does not solve all his problems, as White can grab the h3-pawn before starting an initiative on the queenside.

1 8.VNxh3 .td7 19.�hl

a

b

e

d

e

f

g

h

This is arguably the most logical move of all, as it avoids the threats of ... .ib4 and ... WlgS . White also avoids putting the queen on e3

1 9.Wle3 leads to similar posltions, for instance 1 9 . . . 0-0-0 20.tLle4 �b8 2 1 .c4 Wlc7 22.cS with good prospects for White.

19 0-0-0 20.ttle4 gg8 •••

20 . . . hS 2 1 .c4!? h4 22.cS �b8 23 .gd2 �a8 24.gb l also gives White some initiative.

1 2 . g5

346

. .

18 ... hxg2? is bad as usual, and after 1 9.E:fe l White has an overwhelming initiative. 1 8 . . . .ic5 is met by 1 9 . tD e4 when exchanging on d4 gives Black an unpleasant endgame, and otherwise White's initiative j ust gets stronger. 1 8 . . . tDc5 1 9 .Wfc4! can be compared to the earlier variation B with 1 6 . . . Wfg5 , but this is a much better version for White. 1 9 . . . .id6 20.g3 .id? 2 1 oE:fe l Wfxf6 22.tDd5t a

b

e

d

e

f

g

h

19.tll e4! White can simply ignore the threat to the h2-pawn, as his pieces are so active. Instead after 1 9 .93 tDxf6 Black might be okay.

2 1 J:�d2!? c;t>b8 2 1 . . .Wfc? 22.E:b l tD c6 23 .Wfe3t

22J�bl c;t>a8 23.c4 The position is complicated, but I like White's chances on the queenside.

19 YlYxh2t 20.c;t>f2 hxg2

C32) 17 Wfe5

21 .E:gl .ifS 22J:!xgl YlYf4

•••

a

b

e

d

e

f

g

18 .if3!

h

•

1 8 .g3 tD c 5 ! allows Black to regroup in time to arrange castling. ( 1 8 . . . tD xf6 1 9.E:xf6! Wfxf6 20.tDe4 Wfd8 2 1 oWfe3 ±) 1 9 .Wff3 .id? 20.E:fe l Wfc? 2 1 otD f5 0-0-0 Black i s doing well.

18 ... .id6 1 8 . . . tDxf6? is refuted by 1 9 . tD c6!.

•.•

20 . . . .ie5 2 1 oE:h l Wff4 22.g3 wins.

a

b

e

d

e

f

g

23.tll e2!

h

23.tDxe6! ? fxe6 24.Wfd4 leads to some insane stuff, but I was only able to find a draw here, so there is no need to force the issue.

23 ...YlYe5 23 . . . Wfc? 24 . .ih5 Wfe5 2 5 . tD 2g3t·

24.c4!?

347

Chapter 1 8 - 1 6.1:!bd l White has an ongoing initiative. Black is surviving f�r the moment, but his position does not look like much fun.

C33) 17 hxg2! .••

If Black tries 1 B . . . .tc5 then 1 9 .Wxg2 makes more sense, escaping the pin and preparing all sorts of knight j umps. Black does not have the time to exploit the open king, for instance: 1 9 . . . ll'l e 5 20.'?Mg3 .td7 8

8

7

7

6

6

5

5

4

4

3 2

3 2

b

a

1 a

b

e

d

e

f

g

h

In most cases Black does best to postpone this move. The present case is an exception to the rule, although it would be practically impossible to figure this out over the board.

e

d

e

f

g

h

2 1 . ll'l b3 (2 l .ll'l xe6! ? leads to some forcing lines, and also favours White according to the computer.) 2 1 . . .'?Mc7 22.1'!e4 .td6 23.1'!xd6 '?Mxd6 24.1'!xe5;!;

C331) 1 8 flc7 •••

18JU4!? This looks strange at first, but it makes sense when you consider 1 B .1'!f2 ll'l e 5 ! 1 9 .'?Me3 .td7 intending . . . 0-0-0 with a safe position for Black. Now the prospect of a rook lift along the fourth rank makes it harder for Black to castle. l B . Wxg2 is not bad according to the computer. This could be investigated further, but I have chosen the option that does not involve opening up my king.

a

b

e

d

e

f

g

h

Here we have a final branching point, with C331) 18 '?Mc7 and C332) 18 .td6! the main candidates.

This is the most natural move, but it leads to trouble for Black.

1 B . . . ll'le5? is no good here due to 1 9 .'?Me3! .td7 (or 1 9 . . . '?Mc7 20.1'!e4+-) 20.1'!e4! .td6 (20 . . . ll'l c6 2 1 .ll'lxe6+-) 2 1 .ll'lxe6 and White is winning.

1 9 .'?Mg3 is possible, but I prefer not to have the rook pinned.

.••

•••

19.fle3!

1 2 gS

348

. . .

2 1 ..ih5! tLlf8

19 ....ic5 Intending . . . WfeS . 1 9 . . . eS 20.tLle6! fxe6 2 1 .:B:c4 WfaS 22 . .ihSt 'it>d8 2 3 . tLl a4! threatens tLl b6, and if 23 . . . bS 24.WfxeS! White wins.

Both 2 1 . . .eS 22.tLle4 and 2 1 . . .'it>f8 2 2 . .ixf7 tLl eS (22 . . . .ixd4 23.:B:xd4 'it>xf7 24.:B:f2±) 23.:B:xeS 'it>xf7 24.tLldS maintain White's initiative.

1 9 . . . tLl b6 gives White a pleasant choice. 20.tLlfS ! ? (There is also 20 . .ihS ! ? .ics 2 1 .Wfg3!, when it is hard for Black to develop further, as . . . .id? will allow .ixf7.) 20 . . . .id? (20 . . . .icS 2 1 . tLl g?t 'it>ffi 22.:B:d8t Wfxd8 23 .Wfxcst 'it>g8 24.:B:g4+-) 2 1 . tLl e? There is no clear win yet, but it looks horribly dangerous for Black. 8 7 a

6

b

e

d

e

f

g

h

22.C!l�d5 'Wd6 23J�e5! .id7 24.c!Lle7

5

Total domination has been achieved, and Black will have to give up his queen to avoid an even worse fate.

4 3 2

24 ... .ib8 25.tLldc6 'Wxdl t 26.hdl bxc6 27.'Wc3

1 a

b

e

d

e

f

g

h

White should win.

20J�f5!!

C332) 1 8 ....id6!

Amazingly enough, Black is rather stuck, and White j ust has to throw pieces at his king!

20 ...i.a7 Other moves also fail to solve Black's problems: 20 . . . bS 2 1 . tLl e4! .ib6 22.Wfa3 is winning. 20 . . . tLl b6 2 1 .:B:xcS WfxcS 22.tLle4 tLldS 23 .Wfg3 Wfc? 24.Wfa3! ± 20 . . . 'it>f8 2 1 .:B:xcS! WfxcS 22.tLle4± 20 ... eS 2 1 . tLl e4 It will be hard for Black to contain White's activity.

a

b

e

d

e

f

g

h

349

Chapter 1 8 - 1 6.!!bd l This may be the only way for Black to survive.

19JHl 1 9 .tLlxe6? ixf4 20.tLl xf4 does not work due to 20 . . . 'Wc5t 2 1 .�xg2 'Wg5t+, but after the text move the bishop on d6 is a genuine target.

19 ... ttle5 1 9 . . . ib4 20.tLle4 is obviously dangerous for Black, while 1 9 . . . 'Wc7 is well met by 20.tLlxe6! ixh2t 2 1 .�xg2 fxe6 22 .'Wg6t �d8 23.gf3± intending 'Wg7. Now White has a choice between a sound yet tricky continuation which leads to draw after accurate defence by Black, and a wildly complicated line with unclear consequences.

2 1 . tLl b3 2 1 . tLl db5 axb5 22.'Wc5 tLl c6! 23.if3 id7 24.ixc6 0-0-0! is okay for Black. 2 1 . . . 'Wb4! 2 1 . . . 'Wa3 22.'We4!t 2 1 . . .'Wb6 22.tLlc5 id7 23.gb ! 'Wc6 24.gxb7 (24 . tLl 3e4 b6 2 5 . tLl xd7 tLlxd7 26.gxg2°o) 24 . . . ic8 2 5 . gxc7! 'Wxc7 26.tLl 3e4� looks dangerous for Black. 8 7 6 5 4 3 2

a

a

b

e

d

e

f

g

h

20.'Wg3 I have put this as the main line, as it is impossible to determine what is happening after the alternative. 20 .'We3!? This leads to a completely out-of-control position. 20 . . . ic7 I have a feeling that White may be able to conj ure a strong initiative here, but so far I have not been able to demonstrate it.

b

e

d

e

f

g

h

22.gd4 22.tLlc5? b6=t is not much good for us. 22.tLle4 id7 23.c3 'Wa3 is a strange mess, but Black is ready to castle so he should be doing well. 22 . . . 'Wb6! 22 . . . 'Wffi 23.tLl c5� 23.tLla4 'Wc6 24.gxg2 id7 2 5 . tLl ac5 0-0-0 26.gg7°o The position remains unclear, but Black seems out of danger.

20 ....ic7! Black does best to avoid both 20 . . . ib8 2 1 . tLl e4t and 2o . . . id7 2 1 .tLlxe6! ixe6 22.gxd6 tLl d7 23.tLle4±.

21.ttlcb5!

A nice attacking idea, but unfortunately Black can hold with brilliant defence.

2 1 ... axb5!

This leads to a forced draw after precise play.

1 2 . . g5

3 50

.

Black is in danger after: 2 1 . . . i.b8 22.a4! 22.i.h 5 ! ? is fun to analyse, but it should lead to an exotic draw: 22 . . . axb5 23 .l:!e2 ttJ g6! (23 . . .'IWb6 24.l:!xe5±) 24.ttJxe6! Wfb6t 2 5 .Wfe3 Wfxe3t 26.l:!xe3 i.xh2t (26 . . . i.xe6? 27.l:!xe6t ne6 28 .i.xg6t �f8 29.l:!d8#) 27.�xh2 i.xe6 28.l:!xe6t ne6 29 .i.xg6t �fB 30.l:!d7= 22 . . . Wfb4 Both 22 . . . i.d7 23 .Wfa3! and 22 . . . axb5 23 .i.xb5 t i.d7 24.Wfg7 l:!fB 2 5 . ttJ xe6! are winning for White.

22 ..bb5t .id7 23.�xe6 bb5 23 . . . ne6 is a slightly more complicated way to make a draw: 24.i.xd7t ttJxd7 25 .Wg6t �d8 26.Wg7 8 7 6 5 4 3 2

8

b

a

7

e

d

e

f

g

h

26 . . . i.xh2t! 27.�xh2 Wh5t 28 .�xg2 Wxd l 29.Wxh8t �c7 30.Wxa8 Wg4t= With a perpetual.

6 5 4 3

8

2

7 b

a

e

d

e

f

g

6

h

23.c3 Wfc5 24.ttJb3 WffB 2 5 .Wfe3! ttJ d7 2 5 . . . axb5 26.Wfb6!+26.i.c4! axb5 27.i.xb 5 ± Black faces serious threats including Wfb6 and Wfd3 (or Wfd2) .

5 4 3 2 1

8

a

7

b

e

d

e

f

g

h

24.ti�g7t �f8 25.tLle6t= Neither side can avoid the repetition by now.

6 5 4 3 2 1 a

b

e

d

e

f

g

h

Chapter 1 8 - 1 6.�bd l

Conclusion In such a heavily analysed variation as the Poisoned Pawn, it would seem almost miraculous to find a fresh idea, which has been played in just a handful of games out of a few thousand, and discover that it is actually rather good. This exactly sums up my experience of analysing the intriguing option of 1 6.�bd l ! ? in this most critical of variations. We started by analysing the crude attempt to punish White's 1 6th move with 16 . . . .ib4. Once you appreciate that White can disarm the threat with 1 7.WI'e3!, threatening to smash through on e6, White's position starts to make more sense. We then considered the natural 1 6 . . . WI'gS ! ? , when my new suggestion of 1 7.WI'd3!N ttJcS 1 8 .WI'c4! poses Black plenty of problems. Even if Black defends optimally, White will still have some freedom at move 2 1 to steer the game towards the kind of position that suits him. The most complicated reply of all is 1 6 . . . h3!?, when I analysed three main replies for White. All of them are sound and, although none of them offer a forced route to an advantage, each of them poses different problems to the defender.

35 1

1 7.g3 is probably the easiest move for Black to equalize against, although even here he must avoid some pitfalls while White does not risk much. 1 7 . .ig4!?N leads to a tactical minefield, where 1 7 . . . ttJ e S ! seems best, with Black j ust about maintaining the balance with precise defence. 1 7.WI'd3!?N was our final option, after which Black's only fully satisfactory defence seems to be the counterintuitive 1 7 . . . hxg2!, then meeting 1 8 .�f4!? with 1 8 . . . .id6!. Even here White has a draw in his pocket in the main line, with the option of going for unclear complications along the way if he wishes. Throughout my coverage of the Poisoned Pawn, I have been under no illusions about refuting it. Instead I have looked for ways to avoid the many forced drawing lines that have been mapped out by correspondence players, proposing a variety of new and theoretically sound ways to pose problems to the defender. With the analysis in this chapter, combined with the ideas in the previous couple of chapters, you will have a whole array of weapons with which to trouble your future opponents.

Variation Index Chapter 1 1.e4 c5 2.�f3 d6 3.d4 cxd4 4.�xd4 �f6 5.�c3 a6 6.J.g5

A) 6...e5 8 B) 6....id79 C) 6...lik69 D) 6...�bd7 7.Wfe2!11 D1) 7...'lWa5?!11 D2) 7...e513 D3) 7...b515 D31) 8.0-0-015 D32) 8.f4! e5 9.fxe5 dxe5 1O.0-0-0!16 D32l) 10...exd419 D322) 10...'lWb6N 19 D323) 1O...h6N 20 D324) 1O....ib7N l1.tiJf5 21 D324l) 1l...'lWb6 22 D3242) 1l...Wfc7 23

Chapter 2 1.e4 c5 2.�f3 d6 3.d4 cxd4 4.�xd4 �f6 5.�c3 a6 6.J.g5 �bd7 7.�e2! g6

8.0-0-0

A) 8...'lWc7 27 B) 8...e5 28 C) 8....ig7 9.h4!?31 Cl) 9...h631 C2) 9...h5 10.mb1 Wfc7 11.g4!?N 34 C2l) 1l...tiJb635 C22) 1l...hxg4 12.h5! �xh5 13.�xh5 gxh5 14.tiJf5.ifS 15..ig236 C221) 15...e638 C222) 15...�b839 C223) 15...'lWc5! 40

Variations I ndex

353

Chapter 3 1 .e4 c5 2.tLia d6 3.d4 cxd4 4.clll xd4 clll fG 5.clll c3 a6 6.i.gS clll bd7 7.YlYe2! h6 8.i.h4 A) 8 . . . e6?! 46 B) 8 . . . g5 47 C) 8 . . . b 5 ! ? 9.f4 48 C l ) 9 . . . e5N 49 C2) 9 . . . g5 1 0. fxg5 hxg5 1 1 .ixg5 ib7 1 2.0-0-0 e6 1 3 .g4! 53 C2 l ) 1 3 . . . ie7 54 C22) 1 3 . . . b4!?N 55 Chapter 4

1 .e4 c5 2.clll a d6 3.d4 cxd4 4.clll xd4 clll fG 5.clll c3 a6 6.J.g5 clll bd7 7.YlYe2! h6 8.J.h4 g6 9.£4! e5 1 0.fxe5 dxe5 1 1 .0-0-0! A) l l . . .ie7?! 59 B) 1 l . . .'?9c7 1 2. tLl b3 b5 1 3 . tLl d 5 ! tLl xd5 1 4.exd5 59 B l ) 1 4 . . . ib7! ?N 60 B2) 1 4 . . . id6 1 5 .V9d2! 64 B2 l ) 1 5 . . . f5N 65 B22) 1 5 . . . 1t> f8 ! ? 67 B22 1 ) 1 6.lt>b 1 68 B222) 1 6.g4!N 71 B23) 1 5 . . . 0-0!?N 1 6.'?9xh6 ib7 1 7.lt>b 1 gfe8 1 8 .ie2 gac8 1 9.'?9c l !! 74 B23 1 ) 1 9 . . . e4 75 B232) 1 9 . . . if8 76 B232 1 ) 20.d6!? 76 B2322) 20.if3 ! ? 78

Chapter 5 1 .e4 c5 2.clll a d6 3.d4 cxd4 4.clll xd4 clll fG 5.clll c3 a6 6 ..tgS e6 7.£4 A) 7 . . . id7 82 B) 7 . . . tLl c6? 8 . tLl xc6 bxc6 9.e5 h6 1 0.ih4 g5 1 1 . fxg5 tLl d 5 1 2. tLl e4 '?9b6 1 3 .c3! 84 B 1 ) 1 3 . . . gb8 85 B2) 1 3 . . . '?9xb2 86 B3) 1 3 . . . dxe5 88 C) 7 . . . ie7 8 .'?9f3 90 C l ) 8 . . . '?9a5 90 C2) 8 . . . '?9c7 91

354

1 .e4 v s The Siclian I

Chapter 6 1 .e4 c5 2.tl:H3 d6 3.d4 cxd4 4.�xd4 �f6 5.�c3 a6 6 ..ig5 e6 7.£4 �c7 8.�e2!? A) 8 . . . b5 95 B) 8 . . . tLl c6 9 .0-0-0 96 B 1 ) 9 . . . !e7 96 B2) 9 . . . !d7!? 1 00 B3) 9 . . . tLl xd4 1 O.l'l:xd4 1 01 B3 1 ) 1 O . . . b 5 ! ? 1 01 B32) 1 0 . . . !e7 1 1 .'it>b 1 !N 1 04 B32 1 ) 1 1 . . . 0-0 1 05 B32 1 1 ) 1 2.g3 1 05 B32 1 2) 1 2.h4!? 1 06 B322) l 1 . . .h6 1 2.!h4 1 08 B322 1 ) 1 2 . . . g5? 1 08 B3222) 1 2 . . . 0-0 1 09 Chapter 7 l .e4 c5 2.�f3 d6 3.d4 cxd4 4.�xd4 �f6 5.�c3 a6 6 ..ig5 e6 7.£4 b5 8.e5 dxe5 9.fxe5 �c7 1 0.exf6 �e5t 1 l ..ie2 �xg5 1 2.0-0 A) 1 2 . . . �e5 1 15 B) 1 2 . . . l'l:a7 1 3 .�d3 l'l:d7 1 4. tLl e4 1 1 7 B 1 ) 1 4 . . . �g6!? 1 1 7 B2) 1 4 . . . �e5 1 5 . tLl f3 1 19 B2 1 ) 1 5 . . . �c7!? 1 19 B22) 1 5 . . . �xb2 1 6.�e3 !b7 1 7.l'l:ab 1 ! 124 B22 1 ) 1 7 . . . �xc2 125 B222) 1 7 . . . �xa2! 126 Chapter 8 l .e4 c5 2.�f3 d6 3.d4 cxd4 4.�xd4 �f6 5.�c3 a6 6 ..ig5 e6 7.£4 � bd7 8.�e2 �c7 9.0-0-0 A) 9 . . . !e7 132 B) 9 . . . b5 1 0.a3 133 B 1 ) 1 O . . . !b7 134 B2) 1 O . . . !e7 1 1 .g4 136 B2 1 ) l 1 . . .!b7 137 B2 1 1 ) 1 2.f5 ! ? 137 B2 1 2) 1 2 .!h4 140 B2 1 2 1 ) 1 2 . . . tLlc5 141 B2 1 22) 1 2 . . . tLl b6 142 B2 1 23) 1 2 . . . g6! ? 143 B22) l 1 . . .h6 1 45

355

Variations Index Chapter 9

l .e4 c5 2.�a d6 3.d4 cxd4 4.�xd4 �f6 5.�c3 a6 6.i.g5 e6 7.£4 � bd7 8.YlYe2 YlYc7 9.0-0-0 b5 10.a3 .le7 I l .g4 h6 1 2 .lh4 .lb7 1 3.<.!?b l ! •

A) 1 3 . . . g5!?N 150 B) 1 3 . . . Eib8 152 C) 1 3 . . . Eic8 1 4 .Eig l 153 C 1 ) 1 4 . . . �b8 153 C2) 1 4 . . . g6 155 C3) 1 4 . . . g5 1 5 . fxg5 hxg5 1 6.i.xg5 �b6! 1 7.�e l 157 C3 1 ) 1 7 . . . lt:le5 ?!N 158 C32) 1 7 . . . Eixc3 1 8 .i.xf6 lt:l xf6 1 9 .�xc3 lt:l xe4 20.�e 1 159 C32 1 ) 20 . . . i.h4 1 60 C322) 20 . . .i.f6 1 63 Chapter 1 0

1 .e4 c 5 2 . � a d 6 3.d4 cxd4 4.�xd4 �f6 5.�c3 a6 6 ..lg5 e 6 7.£4 .le7 8.YlYa h 6 9 ..lh4 g 5 1 0.fxg5 �£d7 A) 1 1 .�h5 ! ? 168 B) 1 1 .lt:lxe6! fxe6 1 2.�h5t \t>f8 1 3 .i.b5 ! Eih7! 1 4 .e 5 ! ?N 1 70 B 1 ) 1 4 . . . axb5 1 71 B2) 1 4 . . . lt:lxe5 1 5 .0-0t \t>g8 1 6. g6! 1 74 B2 1 ) 1 6 . . . lt:l xg6 175 B22) 1 6 . . . Eig7 1 76 B23) 1 6 . . . i.xh4! 1 77

Chapter I I 1 .e4 c5 2.�f.3 d6 3.d4 cxd4 4.�xd4 � f6 5.�c3 a6 6 ..lg5 e6 7.£4 .le7 8.YlYa h6 9 .lh4 YlYc7 10.0-0-0 � bd7 •

A) l U5 ! ? 185 A 1 ) 1 l . . .lt:le5 186 A2) 1 l . . .0-0!?N 187 A3) 1 l . . .e5 189 B) 1 1 .i.e2 192 B 1 ) I l . . .Eig8? 192 B2) 1 l . . .g5?! 193 B3) I l . . .Eib8?! 195 B4) l l . . . b 5 198 B5) 1 l . . .0-0!? 200

356

l .e4 v s The Siclian I

Chapter 1 2 1 .e4 c 5 2.c�� a d 6 3.d4 cxd4 4.tLlxd4 tLl f6 5.tLlc3 a6 6.J.g5 e 6 7.f4 J.e7 8.ti'a ti'c7 9.0-0-0 tLl bd7 A) 1 0. f5 ! ? 206 A I ) 1 0 . . . tLl e 5 206 A2) 1 0 . . . e5 207 B) 1 0.g4 208 B 1 ) l O . . . h6 1 1 .ixf6 ixf6 1 2.<;!;>b 1 !? 209 B 1 1 ) 1 2 . . . g5 209 B 1 2) 1 2 . . . Wfb6 21 1 B2) l O . . . b5 1 1 .ixf6 212 B2 1 ) l l . . .ixf6 214 B22) 1 l . . .lLlxf6 1 2.g5 lLl d7 1 3 .f5 215 B22 1 ) 13 ... ixg5 t?! 1 4 . <;!;> b 1 215 B22 1 1 ) 14 ... lLle5 215 B22 1 2) 1 4 . . . 0-0 217 B222) 13 ... lLlc5 218 B223) 13 ... 0-0!? 1 4.h4!? b4 1 5 . lLl ce2 221 B223 1 ) 1 5 . . . ib7!? 221 B2232) 1 5 . . . e5 223 Chapter 13 1 .e4 c5 2.tLla d6 3.d4 cxd4 4.tLlxd4 tLlf6 5.tLlc3 a6 6.J.g5 e6 7.f4 ti'b6 8.ti'd2 ti'xb2 9':!�b l ti'a3 1 0.e5 A) 1 0 . . . lLl fd7 226 B) 1 0 . . . dxe5 (without . . . h6) 227 C) l O . . . h6 1 1 .ih4 dxe5 1 2. fxe5 lLl d 5 ! ? 1 3 .lLlxd5 exd5 1 4.e6!? ixe6 1 5 .lLlxe6 fxe6 1 6.id3 ie7 1 7.ig6t <;!;>d8 1 8 .ixe7t Wfxe7 1 9.0-0 231 C l ) 1 9 . . . b 5 ? 233 C2) 1 9 . . . lLl d7 20.l"i:f7 �c5t 2 1 .<;!;>h 1 234 C2 1 ) 2 l . . .l"i:c8 234 C22) 2 l . . .l"i:f8 22.l"i:xg7 l"i:f6 23 .id3! <;!;>c7 24.c4 d4 2 5 .Wfe 1 ! l"i:afB 26.Wfe4 238 C22 1 ) 26 . . . l"i:6f7!?N 239 C222) 26 . . . b6 241

Variations Index

357

Chapter 14 l .e4 cS 2.tll a d6 3.d4 cxd4 4.tLlxd4 tLlf6 S.tLlc3 a6 6 ..tgS e6 7.f4 �b6 8.�d2 �xb2 9.gb l �a3 10.eS h6 1 l ..ih4 dxeS 1 2.fxeS tLlfd7 13.tLle4 �xa2 14.gdl A) 14 ... �b2!? 249 B) 1 4 . . . �d5 1 5 .�e3 252 B 1 ) 1 5 . . . .ic5 252 B2) 1 5 . . . �xe5 1 6 . .ie2 .ic5 1 7 . .ig3 254 B2 1 ) 1 7 . . . �d5 255 B2 1 1 ) 2 1 . . .l'!d8 255 B2 1 2) 2 1 . . . 5 ! 257 B22) 1 7 . . . .ixd4 1 8 .l'!xd4 �a5t 1 9 .1'!d2 0-0 20 . .id6 261 B22 1 ) 20 . . . l'!e8? 261 B222) 20 . . . ttJ c6?! 262 B223) 20 . . . 5 265 Chapter I S 1 .e4 cS 2.tLla d6 3.d4 cxd4 4.tLlxd4 tLlf6 S.tLlc3 a6 6 ..igS e6 7.f4 �b6 8.�d2 �xb2 9.gb l �a3 10.eS h6 1 1 ..ih4 dxeS 1 2.fxeS tLlfd7 13.tLle4 �xa2 14.gdl �dS I S .�e3 �xeS 1 6 ..ie2 .icS 17 ..ig3 hd4 1 8.gxd4 �aSt 1 9.9d2 0-0 20 ..id6 gd8 2 1 .g4!! A) 2 1 . . .�b6 274 B) 2 1 . . . f6 274 C) 2 1 . . . ttJ f6 275 D) 2 1 . . . ttJ c6 22.g5 hxg5 23.l'!g l ! 277 D 1 ) 23 . . . g6? 278 D2) 23 . . . ttJ f6!? 279 D3) 23 . . . f6 24.h4!? 280 D3 1 ) 24 . . . ttJ f8 281 D32) 24 . . . ttJ de5 283

1 .e4 v s Th e Siclian I

358

Chapter 1 6 1 .e4 c 5 2 . tiH3 d 6 3.d4 cxd4 4.�xd4 � f6 5.�c3 a6 6.J.g5 e 6 7.f4 Wb6 8.Wd2 Wxb2 9 J'� b l Wa3 1 0.e5 h6 1 l .ih4 dxe5 1 2.fxe5 g5 13.J.f.2!? �g4 14 ..ig3 •

A) 1 4 . . . i.g7!? 1 5 . lt'l e4 288 A 1 ) 1 5 . . . We3t!?N 290 A2) 1 5 . . . 0-0 292 B) 1 4 . . . lt'l d7 1 5 .i.e2 lt'l gxe5 294 B 1 ) 1 6. lt'l xe6!? 295 B2) 1 6.0-0 i.g7 1 7.lt'lxe6! fxe6 1 8 . lt'l e4 298 B2 1 ) 1 8 . . . lt'l £7!?N 299 B22) 1 8 . . J'1f8 1 9. c4!!N 303 B22 1 ) 1 9 . . . b6 304 B222) 1 9 . . . b 5 ! 306

Chapter 17 1 .e4 c5 2.�f3 d6 3.d4 cxd4 4.�xd4 �f6 5.�c3 a6 6.J.g5 e6 7.f4 Wb6 8.Wd2 Wxb2 9J'�bl Wa3 1 0.e5 h6 1 1 .J.h4 dxe5 1 2.fxe5 g5 13.exf6 gxh4 14.J.e2 A) 1 4 . . . lt'l d7?! 3 1 1 B ) 1 4 . . . Wa5 1 5 .0-0 313 B 1 ) 1 5 . . . h3! ? 1 6.i.f3 ! ? It'l d7 1 7. lt'l d 5 Wlc5! 1 8 .l:!be 1 !N i.d6 1 9 .@h 1 ! 313 B 1 1 ) 1 9 . . . lt'le5 315 B 1 2) 19 ... i.e5 316 B2) 1 5 ... lt'l d7 1 6.@h 1 Wlg5 1 7.Wld3! ? 318 B2 1 ) 17 ... lt'lc5 319 B22) 17 ... lt'l xf6!? 321

Variations Index

359

Chapter 18 1 .e4 c5 2.�f3 d6 3.d4 cxd4 4.�xd4 �f6 5.�c3 a6 6 ..ig5 e6 7.£4 VNb6 8.VNd2 VNxb2 9J�b l VNa3 10.e5 dxe5 I l .fxe5 h6 1 2 ..ih4 g5 13.exf6 gxh4 14 ..ie2 VNa5 1 5.0-0 � d7 1 6J:�bd l !? A) 1 6 . . . .ib4N 326 B) 1 6 . . . '?;Ifg5 ! ? 1 7.'?;Ifd3!N tLlc5 I B .'?;Ifc4! .id7 1 9 . .if3 328 B l ) 1 9 . . . h3!? 329 B2) 1 9 . . . '?;Ife3t! 20.Wh 1 b5 331 B2 1 ) 2 1 ."Wb4 331 B22) 2 1 . tLl d 5 ! ? 333 C) 1 6 . . . h3!? 336 C l ) 1 7.g3 337 C2) 1 7 . .ig4!?N 339 C2 l ) 1 7 . . . hxg2 339 C22) 1 7 . . . '?;Ifg5 ! ? 340 C23) 1 7 . . . tLl e 5 ! 343 C3) 1 7."wd3!?N 345 C3 1 ) 1 7 . . . tLle5 345 C32) 1 7 . . . "We5 346 C33) 1 7 . . . hxg2! I B .E!f4!? 347 C33 l ) I B . . . '?;Ifc7 347 C332) I B . . . .id6! 348