Pawn Structure Chess [Batsford Chess ed.] 9781849941150, 9781849940702

Citation preview

First published in the United Kingdom in 2013 by Batsford 10 Southcombe Street London W14 0RA An imprint of Anova Books Company Ltd Copyright © Batsford 2013 Text copyright © Andrew Soltis 2013 The moral right of the author has been asserted. 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, mechanical, photocopying, recording or otherwise, without the prior written permission of the copyright owner. First eBook publication 2013

eBook ISBN: 978-1-84994-115-0 Also available in paperback

Paperback ISBN: 978-1-84994-070-2 The print edition of this book can be ordered direct from the publisher at the website www.anovabooks.com, or try your local bookshop.

Contents

Introduction Chapter One: The Caro-Slav Family Chapter Two: The Slav Formation Chapter Three: The Open Sicilian/English Chapter Four: Chain Reactions Chapter Five: The e5 Chain Chapter Six: The King’s Indian Complex Chapter Seven: The Queen’s Gambit Family Chapter Eight: The Panov Formation Chapter Nine: Stonewalls And Other Prisons Chapter Ten: The Nimzo-Gruenfeld Formation Chapter Eleven: The Lopez Formation Chapter Twelve: The Closed Sicilian/English

Introduction

Mastering pawn structures comes down to acquiring three traits: (a) Being able to recognize what good and bad pawn structures look like, (b) Understanding how to exploit a good one, and (c) Knowing how to change a structure favorably. Some good structures are obviously good. But it’s not obvious why they are good. For example, Polugayevsky – Dokhoian, Belgrade 1988: 1 f3 f6 2 c4 b6 3 g3 b7 4 g2 e6 5 d4 b4+ 6 d2 xd2+7 xd2 0-0 8 c3 e4 9 d3 f5 10 e5. Black used tactics to release White’s pressure on the long diagonal, 10 … c5 11 dxc5 xg2. Then came 12 g1 c6 13 0-0-0 f6 14 f4! bxc5.

White to play

A pawn ahead, Black was understandably reluctant to give up two pawns. So when White played 15 xc6 he rejected 15 … xc6 (16 xd7 d4 17 xc7 ab8 with Black counterplay) in favor of 15 … dxc6. An inexperienced player – particularly one who peeks at the ‘Black resigns’ at the end of the game – will be tempted to say, “Of course, he loses. He tripled his pawns. They’re weak.” But that’s not the reason. Black went rapidly downhill after 15 … dxc6?? but not because his pawns could be captured. Play continued 16 e4! a6 17 e5! and White had achieved what annotators like to call a ‘positionally won game.’ He severely restricts enemy pieces and can maximize the power of his own without interference: 17 … e7 18 f3 e8 19 g2 f7 20 gd2.

Black to play Black is lost because his pieces can’t play: His knight has only two moves. One is bad (20 … b8? 21 d8) and the other is useless (20 … b4 21 a3). His rooks can’t go to the d-file, the only one that matters, and he wants to keep one of them on his second rank to avoid d7. If his queen moves away he allows xc6 (That’s the only major way in which the vulnerability of his pawns is felt.). Black’s predicament is not an accident. His pieces can’t play because the pawn structure won’t let them. If we could make a tiny change – like picking up the pawn at c7 and moving it to d5 or d4 – Black’s fortunes

change dramatically. But that’s illegal. As it stands, Black’s pawns are fairly well protected. Yet they cost him the game. White took his time to find a winning breakthrough. That’s what a very favorable pawn structure allows you to do – take your time. Black eventually ran out of ideas: 20 … h6 21 a4 h7 22 a3 b8 23 c3 c8 24 a5 b3 25 d3 xd3 26 xd3 b7 27 c2 e7 28 d8! g6 29 d2 f7 30 d1 g6?! 31 g4! fxg4 32 xg4+ h7 33 f5 exf5 34 xf5+ g6 and Black resigned before White could play 35 f8 and deliver mate. A good pawn structure is one that allows you to carry out desirable plans and tactics. A bad structure won’t let you do that. It’s as simple as that. Doubled, even tripled pawns are not always bad. True, they may be more vulnerable to capture. But it’s what the pawn structure allows you to that matters most. One of the most sensational games of the last 50 years, Short – Timman, Tilburg 1991, featured a remarkable finish. But it was the pawn play that sets the stage – and teaches us the most: 1 e4 f6 2 e5 d5 3 d4 d6 4 f3 g6 5 c4 b6 6 b3 g7 7 e2 c6 8 0-0 0-0 9 h3 a5 10 a4 dxe5 11 dxe5 d4 12 xd4 xd4 13 e1 e6?! 14 d2 d5 15 f3 c5 16 e4 b4!.

White to play White can exploit the Black kingside if he can get his queen to h4,

where it supports h6 and can threaten mate on h7 with g5. But 16 … b4 interrupts that plan. Black allows 17 xd5 exd5 18 xd5 because 18 … e6 would offer him excellent play for a pawn. If, instead, White retreats, 17 e2, Black can complete development and be close to equality with 17 … b6 and 18 … a6. White realized this and found 17 c4! and then 17 … b6 18 b3!!. The point is that 18 … xc4 19 bxc4 is actually a very favorable pawn structure. It allows White to begin his kingside attack undisturbed. Black cannot trade queens and he doesn’t have time to threaten the c4-pawn. After Black met the a3 threat with 19 … e8, play went 20 d1 c5 21 h4 b6 22 e3 c6 23 h6 h8 24 d8! b7 25 ad1 g7 26 8d7! f8 27 xg7 xg7 28 1d4 ae8 29 f6+ g8 30 h4! h5.

White to play Thanks to the pawn structure, White’s pieces dominate. Thanks to the dominating pieces, Black has no counterplay. All it takes for White to win is to add one more piece to the attack. He did it with the dazzling 31 h2 c8 32 g3! ce8 33 f4! c8 34 g5!. Black resigned because h6/ g7 mate is threatened and 34 … h7 35 xg6+ is also a quick mate. An unforgettable finish – made possible by 18 b3!!. Most amateurs don’t think about which pawn structures they will play. The opening chooses for them. The variations they adopt in a 1 e4 game will determine whether they end up in a Caro formation, an Open or Closed Sicilian structure, or a Lopez or Panov formation, for example.

But in some cases the structure is in flux past move 10 and you get to choose it, as in Tomashevsky – Ganguly, Moscow 2007: 1 f3 d5 2 c4 dxc4 3 e3 f6 4 xc4 e6 5 xc4 e6 6 e2 c5 7 0-0 a6 8 d1 b5 9 b3 b7 10 a4 b4 11 d3 c6 12 bd2 c7 13 c4 g4 14 g3 ge5 15 fxe5 xe5 16 xe5 xe5.

White to play The position seems at least equal for Black because his pieces have greater mobility and White’s kingside seems weak (17 d4 e4!). But it isn’t equal – after 17 e4!. The structure gives White a strong plan of f2-f4-f5, directed at the weak points e6 and f7. After Black’s bid for play with … c4 was foiled by 17 … c8 18 c4! he was clearly worse, 18 … e7 19 f4! c7 20 f1! f6?! 21 f5!. He didn’t like 21 … exf5 22 xf5 and e3/ af1 so he chose 21 … e5.

White to play But this is a losing position – losing because the structure allows White to overload the kingside with pieces. The trend continued with 22 e3 e7 23 h4 0-0 24 f2 h8 25 af1 fd8 26 g2. Seeing how hopeless his chances would be after h5 and g3-g4-g5, Black became desperate, 26 … d4 27 xd4 exd4 and lost after 28 h5 f8 29 g4 g6 (29 … xh4 30 h2) 30 h6 g5 31 hxg5 (31 … xg5 32 f6!). Most pawn structures are more familiar than these first three examples. Masters handle them better than amateurs do because they know the typical ways to exploit the structures and to favorably change them. A structure may change quickly – and violently – in the course of a few moves. Take the case of San Segundo – Pigusov, Menorca 1994: 1 d4 f6 2 c4 e6 3 f3 d5 4 c3 e7 5 g5 h6 6 h4 0-0 7 e3 b6 8 e2 b7 9 xf6 xf6 10 cxd5 exd5 11 b4 c6 12 0-0 a5 13 a3 d7 14 b3 e8 15 d3.

Black to play The trade of pawns on move 10 created what we call the Orthodox Exchange formation (d4+e3 versus d5+c6). White is carrying out a well-known strategy called the Minority Attack. We’ll examine it further in Chapter Seven. What you need to appreciate here is that he intends b4-b5 and bxc6 to create targets at b6 and d5.

Black responded with 15 … b5. This is a counter-strategy well known to masters. Black not only stops b4-b5 but sets up White’s b-pawn as a target that can be attacked by … e7-d6 and … e7. He also prepares to bring his knight into play on an excellent outpost square, … b6-c4. But there’s a counter-counter-strategy. White knew that the way to exploit … b5 in very similar situations is to break through in the center, that is with 16 e4. Then he threatened to win a pawn with 17 exd5 or gain space with 17 e5. The natural 16 … dxe4 17 xe4 favors White after, say, 17 … e7 18 ad1 axb4 19 axb4 f6 20 e5!. However, this is a rare case when Black has a strong response to the e3-e4 plan and it lies in the stunning 16 … c5!.

White to play And you thought pawn play was dull? If it were Black’s move he could choose from among four pawn-takes-pawn captures. Instead, it is White’s turn and he has the choice of four captures with his pawns. But it is the threats of 17 … c4! and 17 … dxe4, winning a piece, that take precedence. White can’t afford 17 xd5 c4, 17 exd5 c4, or 17 dxc5 dxe4 and doesn’t like 17 xb5 dxe4. He opted for 17 e5 and 17 … c4 18 b2 to keep material equal. But it wasn’t positionally equal after 18 … xe5! 19 f5 f6 20 xb5 f8. Black has a strong passer, the two bishops and targets at d4 and b4. His edge grew after 21 fe1 xe1+ 22 xe1 b6 23 a4 c6 and he won after 24 bxa5 xa5 25 b4 xb5 26 e8 e7! 27 xe7 xa4 (or

27 xe7 xe8). Pawn play becomes trickier when the structure is so fluid that it can evolve into different formations over the course of several moves, and each potential change requires evaluation. An illustration of that is Tal – Anand, Cannes 1989, which began: 1 c4 c5 2 f3 c6 3 c3 d4 4 e3 xf3+ 5 xf3 g6 6 b3 g7 7 b2 d6 8 g3 b8 9 g2 h6 10 d1 0-0 11 0-0 d7 12 a4 c6 13 d4 xg2 14 xg2 c8.

White to play White’s center is about to come under attack from … f5. He can obtain a slight edge with 15 d5!, particularly since 15 … f5 16 e4! d4 17 b5! trades pieces favorably (17 … xb5 18 xg7 and 19 axb5 make a7 a target for rooks.). But he erred with 15 d3? and Black made a good structural change, 15 … cxd4! 16 exd4 f5. This created another fork in the road: If White had gone on the defensive, 17 e2?, Black would have obtained the edge with another change, 17 … d5!. Then 18 c5 a5! stops White from obtaining a favorable Panov Formation (with b3-b4!), as we’ll see in Chapter Eight. In this case Black would continue … b6 since cxb6/… xb6 turns the White b-pawn and d-pawn into chronic weaknesses.

After 18 … a5 But this didn’t happen. White sought something more ambitious than 17 e2. His best chance for equality was 17 d5! and then 17 … e6 18 e3. But he made another natural move, 17 d5, and there followed 17 … b6 18 d1 xb2 19 xb2. Black could secure a solid position, with no serious losing chances, after 19 … a5, since a4-a5 is stopped and the b3 pawn can become a target. But White could try for counter-balancing pressure against e7.

Black to play Black found the superior 19 … e5!, yet another change in the structure. If White allows that to stand, he is definitely worse. Black would have a wonderful outpost on d4 and there would no longer be a target at e7. From a positional point of view, 20 dxe6! was forced. Then came 20 … fxe6 21 ad1 f6!. Black will play … e5 and his

pressure on the f-file trumps the weakness of his center pawns or the hole at d5 (and 22 g4? h4+ is suicidal). Black’s advantage became apparent after 22 d2 e5! 23 d5+ g7 24 b5 c7 25 c5 xc5 26 xb7+ c7 27 d5 b4 28 fd1 c5 29 a8 xb3 and he won shortly. As these examples show, it is often hard to recognize what a good pawn structure looks like and sometimes harder to understand why it’s good. In this book we’ll explore 12 basic structures and several related formations to seek answers. But before we start, a few warnings: First, many pawn structures that you encounter are messy. They won’t fit neatly into the classifications you’ll find in these pages. You’ll have to deal with them the way you would any strange middlegame situation: Figure it out. 1 e4 g6 2 d4 g7 3 c3 d6 4 g5 h6 5 h4 c6 6 d5 d4 7 d2 c5 8 d1 d7 9 c3 b5 10 f4 f6 11 d3 c7 12 f3 b5 13 0-0 00 14 c2 h5.

White to play This resembles the more familiar formations which arise when White’s c-pawn is on c4 or c2. Then the thematic plan for White is e4-e5. But here 15 e5 g4 leads to obscure complications. In Vallejo Pons – Korobov, Aix-les-Bains 2011, White noticed something else that make this position different from typical Benonis. Black’s knights can’t get to e5 easily. That indicates 15 f5! – which is

often dubious when it surrenders e5 – is strong here. It threatens 16 fxg6 fxg6 17 e5! and xg6. White would sacrifice soundly on g5 after 15 … g5?!. The game went 15 … gxf5 16 exf5 f6 17 e3 c8. Thanks to the kingside pawn structure, Black desperately needs counterplay. He wants to target d5 with … b7. But White foiled that 18 a4 a6 19 b3! b7 20 c4!. This effectively took Black’s queenside pieces out of the game. And after 20 … d7 the time was right for 21 xf6!. Black would have only one piece to defend the kingside after 21 … xf6 22 ae1 followed by g4. Recognizing that as lost, he preferred 21 … exf6.

White to play This is a very one-sided structure. The pawns make it easy for White to attack, hard for Black to defend and impossible to create counterplay. The outcome was becoming clear after 22 g4 h5 23 f2! fe8 24 e4 and then 24 … bxc4 25 bxc4 a5 26 g3 a6 27 xh5. Black resigned shortly after 27 … b4 28 f4! f8 29 e4 a6 30 ac1 h8 31 h4 e7 32 g6+! fxg6 33 fxg6 d8 34 f5 e7 35 ce1 f8 36 g7. A second caveat is: There are some recurring pawn structures that have been omitted to keep this book to a reasonable size. It’s a textbook, not an encyclopedia, and there is no examination of, for example, the doubled pawns that arise after 1 e4 c5 2 f3 c6 3 b5 and xc6 or

1 d4 f6 2 c4 e6 3 c3 b4 4 a3 xc3+5 bxc3 or 1 e4 e5 2 f3 c6 3 c3 f6 4 b5 b4 5 0-0 0-0 6 d3 d6 7 g5 xc3 8 bxc3. And third, often the only way to favorably change a structure requires tactics. Here’s a case, Kramnik – Lputian, Debrecen 1992:

White to play White’s e-pawn has captured on f4, so this is a cross between the Slav formation you’ll find in Chapter Two and the Isolani of Chapter Seven. One of White’s goals in both those formations is to push his d-pawn. Here he can play 17 d5!! based on tactics: Black loses outright after 17 … xd5? 18 h7 mate and less quickly after 17 … cxd5 18 xd5! xc2 19 xe7+! or 17 … exd5 18 xd5!. Instead, Black played 17 … d8 and White was able to make a favorable change, 18 fe1 h8 19 dxe6! and then 19 … xd1 20 xd1 fxe6. He finished off by going after the e6-target: 21 e4 g6 22 c5! xc5 23 xc5 g8 24 a2! g7 25 xe6 f8 26 d7! Resigns. Enough caveats. Let’s start mastering pawn structures.

Chapter One: The Caro-Slav Family

By ‘family’ we mean a group of closely related pawn structures that share some features with or evolve from a common opening. In the CaroSlav family only one player has a center pawn on the fourth rank and it is a d-pawn. His opponent has traded off his own d-pawn for either an epawn – resulting in what we’ll call the Caro formation – or for a c-pawn – the Slav formation. These are the only center pawn exchanges made in the basic cases, as shown by the diagrams.

The Caro formation

The Slav formation

Besides the facial resemblance, the two structures share a basic solidity. White cannot open lines further except by means of a sacrifice (d4-d5!?) or with the help of another pawn: For example, c2-c4 followed by d4-d5 in the Caro, or e3-e4 and d4-d5 in the Slav. If Black doesn’t try to change the pawn structure, such as with … c5 or … e5, the middlegame is often slow-paced. But Black usually wants to change it because the d4-pawn gives White advantages such as greater control of the center and good outposts for pieces at e5 and c5. If Black competes for those squares, with … f6 or … b6, he creates weakness, as Supplemental Game # 1, at the end of this chapter, shows. Black has a natural outpost too, at d5. But the structure tends to limit his pieces to his first three ranks – while White can more easily put his pieces on his first four ranks. And four is simply better than three. One thing to remember, here and throughout this book: These formations can come about in reversed form. That is, Black could be the one with a pawn on the fourth rank, at d5, facing White pawns at c3 and e3. For the sake of convenience, and because these formations come about more often in the other way, we’ll look at the Caro and Slav with White having the d-pawn.

The Caro Formation The Caro formation is by no means limited to the Caro-Kann Defense (1 e4 c6 2 d4 d5 and … dxe4). It also arises out of French Defense when Black gives up the center (1 e4 e6 2 d4 d5 and … dxe4). It can come about from the Scandinavian Defense (1 e4 d5 2 exd5 xd5 or 2 … f6/3 … xd5). And it is familiar to 1 d4 players as well because it arises in Queen’s Gambit Declineds and Catalan Openings when White plays e2-e4 and Black replies … dxe4. What’s more, Black does not have to initiate the Caro pawn trade. White can do it: 1 e4 f6 2 e5 d5 3 d4 d6 4 f3 g6 5 c4 c6 6 0-0 g7 and now 7 exd6 and 7 … xd6. Even after 1 f3 we can find the Caro, such as after 1 … d5 2 c4 d4 3 e3 c6 4 exd4 xd4 5 xd4 xd4 6 c3 f6 and 7 d3 followed by 8 e3, 9 d4 and … e6/… c6. But it’s the qualities of the Caro, not how it comes about, that are our main concern. The first lesson is that Black’s spatial inferiority – the four-versus-three factor – can be fatal if he does not compete in the center. A model example of this is Lasker – Capablanca, Moscow 1935: 1 e4 e6 2 d4 d5 3 c3 b4 4 e2 dxe4 5 a3 e7 6 xe4 f6 7 2c3 bd7? 8 f4! xe4 9 xe4 f6 10 d3 0-0 11 xf6+ xf6 12 c3 d5.

White to play

Black tries to use the threat of … xg2 to allow him to challenge the center with … c5 or … e5. But he missed earlier freeing moves. For example, 7 … c6! and then 8 e3 xe4 9 xe4 e5! (10 dxe5 xd1+ with equality, or 10 d5 d4! 11 xd4 xd5! with advantage). In the diagram White played 13 e2! because preventing a liberating break is what matters most. He stops 13 … e5 and would meet 13 … c5 with 14 e4! (14 … b3 15 d6), winning a pawn. True, Black can carry out his threat with 13 … xg2. But White retains an edge after 14 e4 h3 15 xc7 and he has an attractive alternative in the gambit line 15 0-0-0. Instead, Black solidified his formation with 13 … c6 and natural moves followed, 14 0-0 e8 15 ad1 d7.

White to play Black still cannot play … c5 in view of e4, and he is not ready for … e5 (e.g. 16 … e5 17 c4 or 17 dxe5 xe5 18 xh7+). White could increase his edge with 16 e5!, to inflict damage with xf6 or trap the queen with b2-b4 and c3-c4 (16 … xe5? 17 xh7+! xh7 18 dxe5 moves 19 xd7). Instead, he played simply: 16 fe1 a5 17 c2 g6 (or 17 … h6 18 e2 followed by e4, intending h7+) 18 e5 g7 19 h4! d8 20 h5 g5 21 xg7 xg7 22 e5 e7 23 de1 g8 24 c1! ad8 25 1e3 c8 26 h3.

Black to play White’s attack seems effortless only because the favorable pawn structure makes it easier to find strong moves. He threatens 27 hxg6 and 28 h6+, e.g. 26 … h8 27 h6 g7 28 hxg6 fxg6 29 xg6! or 26 … f6 27 hxg6 hxg6 28 h6+ f7 29 g3 (29 … fxe5 30 xg6+ and mates). Black played 26 … f8 but succumbed to 27 h6+ g7 28 hxg6 hxg6 29 xg6!. He couldn’t capture the bishop in view of 30 h8+ and 31 f3+, so he surrendered his queen after 29 … f6 30 g5! e7 31 f3 and lost.

White’s d4-d5 Plan As favorable as the basic Caro formation tends to be, White can change it with the ambitious c2-c4 followed by d4-d5. This holds great rewards if he is better developed and can exploit the further opening of the center. An ideal example is Spassky – O’Kelly, San Juan 1969: 1 e4 e6 2 d4 d5 3 c3 dxe4 4 xe4 d7 5 f3 c6 6 d3 xe4 7 xe4 c6 8 0-0 f6 9 d3 bd7 10 c4 d6 11 b3 0-0 12 b2 c7 13 c2 fe8 14 fe1 f8 15 ad1 g6.

White to play Unlike the previous example, Black has a solid kingside and a target at d4 to attack by … g7 and doubling rooks on the d-file. But White has the two bishops and that encourages him to try to open the center: Step I – White repositioned his light-squared bishop so that it aims at d5 – 16 f1! g7 17 g3 ad8 18 g2. Step II – He placed his rooks so that they would gain the utmost mobility from d4-d5 – 18 … h5 19 e2! c8 20 h4 cd8 21 de1. Step III – He shifted the queen to the vulnerable a1-g7 diagonal – 21 … hf6 22 c1! h5 23 h3 f8 24 a1 g4 25 c3 h6 26 b2. Black missed his opportunities for counterplay, such as … c5 at moves

18-20. He could continue to play passively (26 … h7, for example). That would have forced White to take additional steps to make d4-d5 work, such as e5-d3-f4. Alternatively, White could shift to another strategy such as c4-c5 followed by e5-c4-d6. But after Black played 26 … f5?:

White to play Black has two pawns and a rook defending d5. But White made the push as a sound pawn sacrifice, 27 xf5! gxf5 28 d5! xc3 29 xc3 cxd5 30 d4 (threat of xf5) 30 … d7 31 c5!. The prospect of White pushing his c-pawn tied Black’s pieces up – 31 … h7 32 b4 a6 33 a4 c8 (not 33 … xa4 34 c6! and a1) 34 b5 axb5 35 axb5 f8 36 c6 bxc6 37 bxc6 d8 38 c1 f6 39 c7! d7 40 e3! and wins, e.g. 40 … e4 41 f3 d6 42 g5+ h7 43 xh5+ g7 44 g4!. But White’s d-pawn can turn from asset to liability after piece trades:

Black to play This is a typical position when that occurs. White’s attacking chances are slim and that makes the d-pawn the most obvious target on the board. This became apparent, in a game from an international tournament in 1968, after 1 … e7! and 2 … f6. White should have recognized the danger and played 2 c3 and 3 b2. Then Black can increase his pressure with … g6, … e7-f8-g7 and … c5, or by tripling his heavy pieces on the d-file as he does in the game. Instead of 2 c3 White fumbled about with 2 c2? f6 3 c3 e7 4 a4? and Black replied 4 … d7! 5 c1? cd8 6 b2 g6!. White’s loss of time meant that Black can add decisive pressure against d4 with his queen before White can double rooks on the d-file. For example, 7 d1 f8! 8 e2 g7 9 ed2 is too slow because 9 … c5! costs White a pawn. So White resorted to a desperation attack, 7 h4 e8 8 h5 d6 9 hxg6 hxg6 and was lost soon after 10 g2 d7 11 h1 xd4 12 xd4 xd4 13 d2! f6! 14 h6 g7!.

Black’s … c5 Black’s most natural counterplay is to trade his c-pawn for White’s dpawn. It is easier to engineer than the other break, … e5, because Black usually has much greater control of c5 – and White has less of it – compared with e5. A goal of … c5 is to free Black’s minor pieces. It opens up part of the a8-h1 diagonal for his bishop. A trade of pawns may give him excellent outposts for knights at e5 or c5. If White reinforces his d-pawn with c2c3 and then … cxd4/cxd4, Black gets an ideal blockading outpost at d5. We’ll look further at this structure, the Isolani, in Chapter Seven. A striking example of the penetrating power of Black’s pieces after … c5 is Gligoric – Smyslov, Moscow 1963: 1 e4 c6 2 d4 d5 3 c3 dxe4 4 xe4 d7 5 f3 gf6 6 xf6+ xf6 7 c4 f5 8 e5 e6 9 0-0 e7 10 b3 a5 11 c4? 0-0 12 f4.

Black to play “Distrust a pawn move,” Emanuel Lasker warned. “Examine carefully its balance sheet.” Here White has underestimated the cost of c2-c4. His d-pawn may not be vulnerable – because none of Black’s minor pieces can attack it. But the d4 square is weak after 12 … c5!. Had he not moved his c-pawn he could reinforce d4 with 13 c3.

The natural response, 13 e3 cxd4 14 xd4, is bad after 14 … a4!, e.g. 15 c2 xc2 16 xc2 xd4 and 15 xa4 xa4 16 xa4 xd4. (White apparently rejected 13 d5 because of 13 … exd5 14 cxd5 e4 – overlooking the tactical shot 15 d6! xd6 16 xf7 or 15 … xd6 16 g6.) Instead he chose 13 dxc5 xc5 14 e2, hoping to contest all of the d-file squares after 15 ad1. But the reply 14 … d4! threatened the bishop on f4 and led to 15 g3 a4!. That wins material whether White opts for 16 d1 a3! or 16 c2 xb2 or 16 ad1 e4!.

White’s Anti- …c5 Strategies White can meet … c5 in four ways. He can: (a) reinforce d4 with a piece, such as 13 e3 in the last example. Or (b), do this with a pawn (c2-c3). That usually leads to an Isolani following … cxd4/cxd4. The alternatives are (c) the trade, dxc5, and (d) the push, d4-d5. The trade is the most common and it has obvious benefits. White may be the first to exploit the open d-file or the now-longer diagonal leading to c6. He also clears d4 for use by minor pieces, typically a knight, and creates a queenside majority and a semi-passed c-pawn. White can dominate the center quickly as in Steinitz – Marco, Nuremberg 1896: 1 d4 d5 2 c4 e6 3 c3 c6 4 e4 dxe4 5 xe4 f6? 6 xf6+ xf6 7 f3 b4+ 8 d2 xd2+ 9 xd2 d7 10 0-0-0! 0-0 11 e3! c5 12 dxc5 f5 13 d3 xc5 14 he1.

Black to play A game from 1896? Yes, this book includes many ‘classic’ games. (That’s another way of saying ‘old’.) The reason they serve better than more recent games is that modern masters know a lot more than the players of the ‘classic’ era. They don’t make the instructive mistakes that are worth studying. Today’s players know, for example, that 5 …

f6 is bad compared

with 5 … b4+! 6 d2 xd4 and 6 c3 c5!. They also know 11 e3! was needed to stop … e5. And they appreciate that the position in the diagram favors White. Exchanging minor pieces eased Black’s space problem but left him weak on the dark squares and behind in development. Fleeing into the endgame, 14 … xe3+ 15 xe3, is bad. For example, 15 … f6 16 e5 stifles Black’s pieces (16 … d7 loses a pawn to 17 xd7 xd7 18 xh7+ and xd7.). Also unpromising is 15 … c5 16 c2 a5 – to stop 17 b4 – in view of 17 g5 h6 18 e4, when Black still has to solve the problem of developing his bishop. In the game, he chose 14 … c7 but after 15 e5! e8 16 b1 his difficulties persisted. If he had tried 16 … f6 he would have invited a ready-made attack with g2-g4-g5!. Instead, he chose the safer 16 … f8 so he could continue … f6 and … e5. But 17 c5! secured a protected outpost at d6 for White’s own knight. The possibility of pushing the pawn to c5 is another benefit of dxc5. The game went 17 … f6 18 c4 e5 19 d6 e7 20 f4 d7. White converted his spatial edge to a mating attack: 21 f5! c6 22 c4+ h8 23 g4! b6 24 g5 fxg5 (else 25 gxf6 gxf6 26 g1 g7 27 xg7 xg7 28 g1+) 25 xg5 h6 26 h5 bxc5 27 f7+ h7 28 g5+ h8 29 f6! gxf6 30 xh6+ h7 31 xh7 xh7 32 xf6+ g7 33 xe5 Resigns. The queenside majority is a long-term benefit of dxc5. White is usually able to convert his 3-to-2 pawn advantage there into a passed pawn. For example, Teschner – Golombek, Hamburg 1955: 1 e4 c6 2 d4 d5 3 c3 dxe4 4 xe4 f5 5 g3 g6 6 h4 h6 7 f3 d7 8 d3 xd3 9 xd3 c7 10 d2 gf6 11 0-0-0 e6 12 c4 0-0-0 13 c3 d6 14 e4 xe4? 15 xe4 f6 16 e2 he8 17 he1 c5 18 dxc5! xc5 19 e5 xd1+ 20 xd1 d8 21 xd8+ xd8.

White to play Black missed a chance to equalize with 14 … f4+ 15 b1 e5! since 16 xe5 xe5 17 e3 xe4! 18 dxe5 xc3+ 19 xc3 xd1+ 20 xd1 d8 trades down to nothingness. And if White avoids that with 18 xe4 he is a bit worse after 18 … f6 because his d-pawn is the board’s chief target, as we saw a few pages ago. Before activating his majority in the diagrammed position, he took aim at the other wing with 22 g4! and then 22 … e8 23 g5 hxg5 24 hxg5 d6 25 f3, threatening to ruin Black’s pawns with 26 g6!. After 25 … g6 Black’s majority was damaged and White changed direction with 26 b4! e7 27 b2 d6 28 c5 f5 29 e4 e8 30 e5!. A trade of queens allowed him to safely advance his king, 30 … xe5 31 xe5 a5 32 a3 axb4 33 axb4 d8 34 b3 e7 35 d2! d5 36 e4 e7.

White to play

White made progress with 37 f6! since the king-and-pawn endgame, 37 … xf6 38 xf6+ xf6 39 gxf6, is won (39 … d7 40 c4 c7 41 b5 followed by, for example, d4-e5-f4-g5-h6-g7). Black retreated, 37 … f8, and was soon limited to two ranks, 38 c4 d7 39 b5 c8 40 d2 d7 41 f3 e7 42 e5+ e8. White set up a winning plan of c5-c6 with 43 xe7 xe7 44 g4 f4 45 f6 d8 46 b6! (46 … e7 47 c6!). Black resigned after 46 … e5 47 e4 d7 48 d6 h3 49 xb7 xg5 50 d5. If pawns are equal, the presence of a majority on one wing means there must be a rival majority on the other. White’s queenside majority in the Caro formation rarely becomes a factor before the late middlegame. But Black’s kingside majority often makes itself felt much earlier. This is illustrated by Maroczy – Charousek, Nuremberg 1896: 1 d4 d5 2 c4 e6 3 c3 c6 4 e4 dxe4 5 xe4 f6? 6 c3? e7 7 f3 0-0 8 d3 c5! 9 0-0 cxd4 10 xd4 bd7 11 f3 a5 12 d2 e5 13 e4 xf3+ 14 xf3 c7 15 c3 xe4 16 xe4.

Black to play White’s bishops look murderous. But by mobilizing his pawns, Black renders them harmless with: 16 … f5! 17 e2 d6! 18 h3 e5. Black has seized the initiative. White sought to neutralize the pawn majority with 19 b4 e4 20 c2 b6 21 f3!. But instead of the faulty 21 … exf3?, which led to an equal game, Black could increase his edge with 21 … b7! 22 fxe4 ae8 and … xe4.

White’s d4-d5 When White meets … c5 by pushing his d-pawn, it is usually captured by Black’s e-pawn. If White recaptures with his c-pawn, the new d5pawn can become a strong passer – or a highly vulnerable target. On the other hand, if he recaptures on d5 with a piece, the resulting structure is symmetrical and, in theory, dead even. But in practice, one side is usually better developed and he profits from opening lines. For example, Quinteros – Henley, New York 1985: 1 f3 f6 2 c4 c6 3 d4 d5 4 c3 e6 5 g5 h6 6 xf6 xf6 7 c2 d7 8 e4 dxe4 9 xe4 d8 10 d3 c5.

White to play Now 11 d5! assures White of superiority because of his control of the two central files. This became manifest after 11 … f6 12 e2 d6 13 0-0-0! and then 13 … 0-0 14 c2!. White wanted to occupy d5 with pieces. After 14 … e8 15 d3! he threatened he1 followed by dxe6 and prepared h2-h3/g2-g4-g5. Black became so desperate in view of 15 … exd5 16 xd5 xd5 17 h7+ that he tried the desperate 15 … b5 and lost the endgame after 16 cxb5 exd5 17 xd5 xd5 18 xd5 f4+ 19 b1 xd5 20 xd5 b7 21 xc5.

A similar situation arises in the Catalan Opening: 1 d4 d5 2 c4 e6 3 f3 f6 4 g3 e7 5 g2 0-0 6 0-0 c6 7 b3 bd7 8 b2 b6 9 c2 b7 10 bd2 c8 11 ad1 c7 12 e4! dxe4 13 xe4 xe4 14 xe4 c5.

White to play If White retreats his queen and allows … cxd4, the position approaches equality. The ambitious alternative is 15 d5 (15 … exd5?? 16 xe7). Black has a choice of two good answers, 15 … f6, which seems to win a pawn, and 15 … f6, which neutralizes White’s pieces. The safer 15 … f6 works well after 16 xf6? xf6, because White loses the d-pawn without compensation. Better is 16 c2 exd5 17 cxd5 when the pawn will be highly vulnerable following 17 … d6 or 17 … xb2 18 xb2 f6 19 d6 b8. The double-edged 15 … f6 16 c2 exd5 makes 15 d5 into a temporary gambit. White stands well after 17 e5! d8 18 g5, which threatens 19 xf6 and xh7 mate and forces 18 … g6. Because of the double pin on the d-pawn, White will regain his pawn. But he wants to retake on d5 with pieces rather than with a vulnerable pawn. Yefim Geller won two games, in similar fashion, from this position. One went 19 a1 d7 20 h4 f6 and now 21 xd5 xd5 22 xd5 xa1 23 xa1 f6 24 ad1 e5 25 e4 with great centralization (and a favorable endgame that he won after 25 … c6 26 d6 e5 27 xe5).

White to play Geller did better in a later game with the immediate 19 h4 and play continued 19 … h5 20 xd5 xd5 21 xd5! e8 22 e1 c6 23 b2 fe8. White unleashed a farsighted combination with 24 xh7! xh4 (not 24 … xh7 25 xh5+) 25 ed1! e6 26 c3! f6 27 d3 g4 28 g5!. The main point is that 28 … xg5 29 xg6+ g7 loses to 30 xf6+ xf6 31 xg4 and White is better after 29 … h8 30 xg5! xd1+ 31 h2. Black tried 28 … e4 29 xf6+ xf6 30 gxh4 ed8 but lost the endgame after 31 xd8+! xd8 32 xd8+ f7 33 xf6 xf6 34 d6+ f7 35 dxg6 xh4 36 g7+. The takeaway is: When the structure becomes symmetrical after … c5/d4-d5/… exd5, it is piece play, not the pawns that usually matters most.

Preventing … c5 White’s most enterprising strategy is to stop … c5 mechanically by putting his own pawn on that square. This advance has a balance sheet filled with pluses and minuses, to use Lasker’s term. It cedes d5 to Black pieces, virtually rules out d4-d5, makes White’s dpawn a backward target on an open file, and enables Black to open lines with … b6. Yet the benefits of c4-c5 often outweigh these factors, as in Pillsbury – Winawer, Budapest 1896: 1 d4 d5 2 c4 e6 3 c3 c6 4 e3 f6 5 f3 bd7 6 d3 d6 7 0-0 0-0 8 e4 dxe4? 9 xe4 xe4 10 xe4.

Black to play In the next chapter, we’ll see why 8 … dxc4!, creating the Slav formation, is superior. Nevertheless, the slight inferiority of the Caro structure could have been repaired by 10 … c5 (or, as we’ll see in a few pages, by 10 … h6 and 11 … e5). When this game was played, it was believed that White could get the advantage in any case with 11 c2 and d3, threatening mate on h7. That’s true after 10 … c5 11 c2! c7 12 d3 f6 13 g5! or 12 … g6 13 h6 e8 14 ad1. But later it was found that 12 … f5!, activating Black’s majority, may equalize.

In the game Black made a tempting but faulty gain of time, 10 … f6?! 11 c2. Then he realized how bad 11 … c5? 12 g5 followed by d3 would be, e.g. 12 … cxd4 13 xd4 e7 14 h4! h6 15 xh6! gxh6 16 xh6 and g5 with an overwhelming attack. He missed another chance for … c5 after 11 … h6 12 e3! e8? and then came 13 d3 c7.

White to play White exploited Black’s failure to change the pawn structure by playing 14 c5! and then 14 … f8 15 e5!. Plus side: he has stopped both of Black’s liberating ideas, … c5 and … e5. Minus side: he allows … b6. But he can maintain his pawn on c5 by answering that with b2-b4. Black’s only counterplay is a slow-to-develop attack on the backward dpawn. White’s threats, such as 16 g4! and 17 xf6+ are more important (16 … xg4 17 xh7 mate). Black’s best may be 15 … g6 and then 16 d2 h7. But he would be clearly worse after 17 f4 d8 18 c4 followed by occupying d6 (and lost after 17 … d5? 18 xg6!). Instead he tried to achieve with tactics what he couldn’t with proper pawn play: 15 … xc5 (hoping for 16 dxc5 xe5) 16 xh6! xd4 (or 16 … gxh6 17 g3+ followed by g6+ and xc7) 17 xd4 gxh6. White finished off with 18 f4 (again threatening 19 g3+ and 20 g6+) 18 … d5 19 xh6 f6 (19 … xe5 20 h7+ h8 21 g6+

leads to mate) 20 f4! e7 21 g6 Resigns. See also Supplemental Game # 3 for another good c4-c5.

Black’s … e5 Strategy If Black has a choice between … c5 and his other liberating move, a good case can be made for … e5. Unlike … c5, this push quickly solves the problem of his queen bishop, which had been hemmed in by the pawn at e6. Also, a trade of the e-pawn eliminates the possibility of a White queenside majority. But … e5 occurs less often because White usually has superior control of e5, thanks to a knight on f3 and a heavy piece on the e-file. Let’s go back to a position we examined on page 33.

Black to play Black would love to play … e5 rather than … c5. But there is a tactical problem with 10 … e5? 11 dxe5 xe5 12 xe5 xe5. The position is virtually symmetrical but having the move is meaningful – 13 xh7+! xh7 14 h5+ and 15 xe5. You might think that Black’s best policy, therefore, is to shift towards … c5. However, the experts in this variation prefer 10 … h6. This costs a tempo but prepares … e5 by ruling out the xh7+ tactic. For example, 11 e1 e5! and now the attempt to pressure Black’s undeveloped pieces with 12 f5?! exd4 13 xd4:

Black to play Black turns the tactical tables with 13 … e5!. He would be winning after 14 xc8? xf3+ 15 gxf3 xh2+ and 16 … xd4. Or after 14 xe5 xe5 15 xe5 e8!. Best is 14 xe5 xf5, when Black enjoys the two-bishop edge. A better way of handling the position after 10 … h6 is 11 c2! so that White can exploit the d3-h7 diagonal with 12 d3. But Black should have equalizing play after 11 … e5! 12 d3 f5, e.g. 13 c5 c7 14 dxe5 xe5 15 b3+ h8 16 e1 xf3+ 17 xf3 and now 17 … f4! frees his c8-bishop. In view of the energizing effect of … e5, White may take extreme measures to stop it. An example is Bronstein – Nei, Leningrad 1962: 1 e4 e6 2 d4 d5 3 c3 f6 4 g5 dxe4 5 xe4 e7 6 xf6 xf6 7 c3 d7 8 d3 c6 9 e2 d5?.

White to play Black abandoned hope of an early … c5 when he posted his bishop on c6. He bet his middlegame on achieving … e5 or exerting enough pressure on the c6-h1 diagonal, particularly after 10 xf6+ gxf6 and … g8, to equalize. But White put … e5 out of question with 10 f4!. Suddenly Black has no way to change the unfavorable pawn structure. The game continued 11 … h4+ 12 g3 e7 13 f3 d7 14 f1! (so that his knight can leave e4 without allowing … xf3) 14 … 0-0? 15 h4! b5 16 eg5!. Then the defense of h7 with 15 … f6 invites a strong 16 e5 e8 17 f5. Black tried 15 … h6 16 h7+ h8 instead but he was lost after 17 e4 d6 18 e5!, since 18 … hxg5 allows mate in two. Black resigned after 18 … xe5 19 fxe5 d7 20 xf7+ g8 21 g6 in view of 22 c2 and h7 mate. Before we say goodbye to the Caro, we should notice that White can also stop … e5 by planting a piece on e5. If Black trades there he can find himself restricted to his first two ranks, conceding the other six to White. And six almost always beats two. See Supplemental Game # 4.

Supplemental Games # 1 Passive play and … f6 dooms Black. Spassky – Donner, Leiden 1970 – 1 e4 e6 2 d4 d5 3 c3 f6 4 g5 dxe4 5 xe4 e7 6 xf6 xf6 7 f3 d7 8 d2 c6 9 xf6+ xf6? (9 … gxf6 gives Black pawn control of e5 at less cost than in the game.) 10 e5! 0-0 11 0-0-0 d8 12 e3 e8 13 g3 d7 14 g2 c6 15 f4 e7 16 h4! f6? 17 f3 h5 18 h3 xf3 19 xf3 f8 20 he1 f7

White to play 21 f1! (The plan is c4/f4-f5.) 21 … d6 22 c4 ad8 23 f5! xd4 24 fxe6 xd1+ 25 xd1 xd1+ 26 xd1 Resigns (26 … xe6 27 d6 or 26 … e7 27 d8! xd8 28 e7+.). # 2 White’s pieces dominate after … c5xd4: Vaganian – Pogonina, Marianske Lazne 2011 – 1 d4 f6 2 f3 d5 3 c4 e6 4 c3 e7 5 g5 0-0 6 c2 bd7 7 0-0-0 c6 8 e4! xe4 9 xe7 xe7 10 xe4 dxe4 11 xe4 f6 12 h4 c5 13 d3 c7 14 he1 cxd4

White to play 15 e5! b5 16 b1 (Not 16 g4 f4+.) 16 …. bxc4 17 g4! d8 18 xf6+ gxf6 19 h6! f5 20 g4! f4 (Or 20 … cxd3 21 gxf5 c2+ 22 a1 and g1+.) 21 xh7+ h8 22 f5+ g8 23 h7+ f8 24 xe6 Resigns. # 3 White’s c4-c5 grants him a free hand to attack:Wilder – Lein, Lone Pine 1981 – 1 d4 f6 2 f3 d6 3 c3 f5 4 g3 c8 5 g2 h3 6 00 xg2 7 xg2 d5 8 d3 bd7 9 e1 e6 10 e4 dxe4 11 xe4 e7 12 g5 0-0 13 ad1 h6 14 d2 d8 15 c4 c6 (Black passes up the double-edged 15 … c5, e.g. 16 d5 b6 17 c3 exd5 18 xf6+ xf6 19 xf6 gxf6.) 16 c2 b5 17 b3 a6?

White to play 18 c5! xe4 19 xe4 f6 20 h4! d5 21 e1 ad8? 22 xh6! gxh6 23 xe6! (Based on 23 … fxe6 24 g6+ f8 25 xh6 h5 26 e5! and wins.) 23 … f8 24 xh6! g8 25 h8! f5 26 e2 g7 27

e5! Resigns (27 … xe5 28 h5! xe6 29 h7+ or 27 … xh8 28 h5+ and mates.). # 4 White’s e5 stifles … e5: Maroczy – Mieses, Vienna 1908 – 1 e4 d5 2 exd5 xd5 3 c3 a5 4 d4 f6 5 d2 c6 6 c4 f5 7 f3 c7 8 0-0 e6 9 e2 d6 10 g3 g6 11 e1 bd7 12 c3 d8 13 e2 0-0 14 ad1 f4

White to play 15 e5! xd2 16 xd2 xe5 17 dxe5 xd2 18 xd2 d8 19 e3 (Black equalizes if he can play … c5 and maneuver his knight to c6. Best is 19 … d5.) 19 … d7? 20 h4! h6 21 h5 h7 22 d1 b6 23 xd8+ xd8 24 e2 a6 25 c4! c8 26 a3 e7 27 a7! c7 28 b4 c8 29 d4 f8 30 f3 e8 31 e4 xe4 32 xe4 d7

White to play 33 d6+! xd6 34 exd6 f6 35 f4 b5 36 c5 d8 37 g4 f7 38 e4

d7 39 e2 g8 40 a2! (Threatening to win with a3-a4xb5 or f4f5.) 40 … g6 41 hxg6 e8 (Or 41 … xg6 42 xe6+! xe6 43 f5+.) 42 f5 Resigns.

Chapter Two: The Slav Formation

There are as many roads to the Slav formation as to the Caro. There is, first of all, the Slav Defense to the Queen’s Gambit and its cousins (Queen’s Gambit Accepted and Queen’s Gambit Declined with … dxc4). Slav pawn structures also arise in more distant family members, such as the Catalan Opening and Nimzo-Indian and Gruenfeld Defenses. There are also colors-reversed versions, such as in the Colle System, e.g. 1 d4 d5 2 f3 f6 3 e3 e6 4 d3 c5 5 c3 e7 6 0-0 c6 7 bd2 c7 and now 8 dxc5 xc5 9 e4 is a reversed version of what we’ll call Tchigorin’s plan. The Slav formation lacks some distinctive Caro features, like White’s queenside majority, and makes some strategies, like … c5, somewhat less appealing. But once again, passivity can be fatal to Black, as in Rubinstein – Schlechter, Berlin 1918: 1 d4 d5 2 f3 f6 3 c4 c6 4 c3 dxc4 5 e3 g4 6 xc4 e6 7 0-0 bd7 8 h3 xf3 9 xf3 e7?! 10 d1 0-0? 11 e4 e8? 12 f4! f8.

White to play Black missed chances to free his game with … e5 but the basic solidity

of the Slav should withstand anything except a d4-d5 break. Unfortunately, that’s just what White had prepared, 13 d5! and 13 … exd5 14 exd5. Black saw that both 15 d6 and 15 dxc6 were threatened. He also realized that he would be lost after 14 … cxd5 15 xd5 xd5 16 xd5, which threatens xb7 as well as xf7+/ xd8. He tried 14 … b6 15 d6! d8. But 16 g4! e6 (Else 17 g5 6d7 18 e3 -moves 19 xf7+.) 17 d7! e7 18 d6 xd7 19 xe6 fxe6 20 g5 was fatal.

Tchigorin’s …e5 Plan The Tsarist era master Mikhail Tchigorin often found himself in a position similar to one we considered in the last chapter, after, say, 1 d4 d5 2 c4 e6 3 c3 f6 4 e3 c6 5 d3 bd7 6 f3 d6 7 0-0 0-0 8 e4.

Black to play The addition of … 0-0 is significant because Black need not play 8 … dxe4, which reaches a Caro formation after 9 xe4. Instead, Tchigorin showed that 8 … dxc4! 9 xc4 e5! is good. (The significance of castling is that if Black’s king were still on e8 White would have a strong 10 dxe5 xe5 11 xe5 xe5 12 xd8+ and xf7.) Tchigorin demonstrated that with … dxc4 and … e5 Black could compete in the center by playing … exd4 and then make use of e5 with his minor pieces. If White foils that by pushing d4-d5, the closed center invites a kingside attack with … f8-g6-f4 and … g4. A modern example of how Black’s piece play accelerates after 9 … e5 is Lautier – Anand, London 1995. It went 10 g5 e7 11 e1 and then 11 … exd4 12 xd4 e5 13 f1 c5.

White to play Black is preparing a pin, 14 … d8. White played 14 a4? and was surprised by 14 … g4!. The key point is that 15 f3 allows 15 … xf3! 16 gxf3 xd4+ (17 xd4? xf3+). White replied 15 d2 and 15 … b4 16 c3. But 16 … ad8! hiked the pressure and White soon lost a piece and the game (17 e3 c5). An alternative situation arises when Black passes up 11 … exd4 in favor of 11 … d8 and White replies 12 d5. It is tempting for Black to exchange pawns on d5 but after 12 … cxd5? 13 xd5 White’s pieces swarm. And pushing his c-pawn to c5 in such positions is double-edged as Supplemental Game # 1 shows. The counterintuitive lesson: Black can allow dxc6/… bxc6 because keeping White minor pieces off d5 is much more important than the weakness of an isolated c-pawn. After 12 d5 Black should begin Tchigorin’s attacking plan, 12 … h6 13 h4 f8!. For example, Fridman – Fressinet, Ajaccio 2007 continued 14 dxc6 bxc6 15 c2 g4! 16 d2 g6 17 g3 h5 18 f1 hf4.

White to play It may seem that White stood well after 19 e3 e6 20 f5 f6 21 f1 b4 22 ed1. But then came 22 … h5! 23 f3 c5+ 24 h1 h4 25 e1 d4 26 e2 xe2 27 xe2 c5! and Black had the initiative. His edge grew after 28 ac1 xf5 29 exf5 f4 30 f1 g5 31 b3 h5! 32 e4 g3+!, since 33 hxg3 hxg3 would set up a mate on the hfile. Instead White tried to keep the kingside closed with 33 xg3 hxg3 34 h3 but lost after 34 … a5 35 c4 a4 36 f1 axb3 37 xb3 a6! 38 ce1 h6 39 f4? exf4 40 xf4 xh3+! 41 gxh3 g2+ 42 h2 gxf1( )+! 43 xf1 xf4+ 44 xf4 e5.

Consequences of dxe5 The biggest danger to the Tchigorin plan is a pawn trade on e5 followed by an advance of White’s e-pawn and f-pawn, blowing Black pieces backward. An example was Taimanov – Matulovic, Palma de Mallorca 1970: 1 d4 d5 2 c4 dxc4 3 f3 f6 4 e3 g4 5 xc4 e6 6 h3 h5 8 0-0 d6 9 e4 e5.

c3 bd7 7

White to play White can proceed patiently with 10 e3 or more violently with 10 g4 g6 11 dxe5 xe5 12 xe5 xe5 13 f4!. But he chose a middle course, 10 dxe5 xe5 11 e2 and was rewarded when Black naively traded pieces, 11 … xf3 12 xf3 xf3+? 13 xf3 e7 14 f4 e5? 15 xe5 xe5.

White to play This allowed White to mobilize his majority, 16 e3! 0-0 17 f4! e7 18 e5!. It became a positional rout – 18 … c6 19 fe1 fe8 20 f3 c5+ 21 f2 xf2+ 22 xf2 d5 (22 … d7 23 e4 and d6) 23 xd5 cxd5 24 ed1 ed8 25 ac1 d7 26 e3! ad8 27 d4 f8 28 f5 and wins. How does Black avoid this? One method is to pressure the White epawn before it’s ready to march. Davidson – Alekhine, Semmering 1926: 1 d4 d5 2 c4 c6 3 f3 f6 4 e3 e6 5 c3 bd7 6 c2 d6 7 d3 0-0 8 0-0 e7 9 e4 dxc4! 10 xc4 e5 11 d1 exd4 12 xd4 b6 13 f1.

Black to play White is a move or so (h2-h3/ e3) away from a favorable f2-f4!. Black stopped him with 13 … d8!, which threatens 14 … xh2+ 15

xh2 xd4 16 xd4 e5+ and 17 … xd4, winning a pawn. White can’t reply 14 f4 in view of 14 … c5 (15 e3 g4). Alternatives such as 14 g3 also run into tactics, e.g. 14 g3 g4! 15 f3? c5 or 15 de2 f3. White relied on 14 h3 and was a move closer to f2-f4!. But Black found 14 … c7!. This threatens 15 … xd4 16 xd4 e5, followed by … xd4 or … h2 mate. In addition, with the bishop removed from d6, there is no pawn fork on e5. That means Black has time to meet 15 f4 with 15 … h5! 16 e5 f6!, when White’s center is overextended. So White played one more preparatory move, 15 e3, putting him on the verge of 16 f4.

Black to play But Black crossed him up with 15 … e8!. Even though this is inconsistent with 13 … d8, it keeps an eye on the crucial e-pawn (16 f3 fd5! 17 xd5 xd5 18 exd5? xe3+ or 18 f2 f4). The game continued 16 d3 h5 17 ce2 g6 18 e1 d7!. With his bishop on e3, White is still reluctant to play 19 f4, because of 19 … df6 20 e5 d5 21 d2 b6!. Instead, White fought for control of the center, 19 f3 b6 20 g5 c5 21 c3?!, but Black’s pieces proved superior and he created a mating attack:

21 … e5 22 xe5 xe5 23 e3 c7 24 e2 h2+ 25 f1 xh3! 26 gxh3 xh3+ 27 g1 h2+ 28 h1 f4 29 xf4 xf4+ 30 g1 h2+ 31 h1 f3+ 32 xh2 e5 33 c5!? xc5 34 xc5 h5+ and … xc5 won. See also Supplemental Game # 2, a rare case of when … c5 is played along with … e5.

White meets … e5 with d4-d5 If White is not willing to exchange his d-pawn for Black’s e-pawn, he will push to d5. That leaves Black three options: (a) He can exchange pawns on d5. (b) He can allow a trade of pawns on c6. (c) He can push … c5, granting White a passed d-pawn. Let’s consider the choices in Gligoric – Gheorghiu, Tel Aviv 1966: 1 d4 f6 2 c4 g6 3 c3 d5 4 f3 g7 5 f4 c6 6 e3 0-0 7 c1 dxc4? 8 xc4 g4 9 h3 xf3 10 xf3 bd7 11 0-0 e6 12 fd1 e7 13 g5! h6 14 h4 ad8 15 e4 e5 16 d5 b6 17 b3.

Black to play The first option, 17 … cxd5 looks attractive if White has to retake with a pawn. For example, 18 exd5 g5 19 g3 e8 and … d6 sets up an effective blockade. Similarly, 18 xd5 xd5 19 xd5 g5! and 20 … xd5. But 17 … cxd5 is poor because White can insert 18 xf6! xf6 and then 19 xd5 or 19 xd5 xd5 20 xd5, with a powerful outpost on d5.

Black also rejected allowing dxc6/… bxc6, because he lacks the kingside play that compensated Tchigorin for the isolated c-pawn. Black chose door number three, with 17 … g5 18 g3 c5. He wants to blockade the d-pawn with 19 … c8 and 20 … d6. But a better way was using the other knight, … e8-d6. The difference is White can act quickly with 19 h4 c8 20 hxg5 hxg5 21 e3 h5 22 d6!.

Black to play White appreciated the value of clearing d5 for his knight, e.g. 22 … xd6 23 d5 e6 24 xg5 followed by a strong e7+. Black preferred to keep his g-pawn defended with 22 … xd6 23 d5 d8 and surrendered the c-pawn instead, 24 xc5 b6. But 25 f3! exposed his kingside to 25 … f4 26 xf4 gxf4 27 g4! with the plan of d3-h3. The rest was: 27 … h8 28 d3 g6 29 f5 xd5 30 h3+ h6 31 xd5 b6 32 dd3! d8 33 xh6+ xh6 34 xe5+ g8 35 e7! f8 36 d6 a5 37 xh6 Resigns.

White plays e4-e5 When Black delays … e5 White has good reason to stop the break with either e4-e5 or f2-f4. But even when there is no chance for … e5, White should be on the lookout for a well-timed e4-e5 because he gains space and attacking chances. A classic case was Pillsbury – Burn, Hastings 1895: 1 d4 d5 2 c4 e6 3 c3 f6 4 g5 e7 5 e3 0-0 6 f3 b6 7 c1 b7 8 cxd5 xd5 9 xe7 xe7 10 xd5 xd5 11 d3 c8 12 e4 b7 13 0-0 d7 14 e2 a6? 15 c3 c6? 16 fc1 b5 17 e3! c7 18 f4 ac8.

White to play Black took a hypermodern approach with 10 … xd5 rather than 10 … exd5. But that virtually requires him to play for … e5 or … c5. He missed those opportunities earlier but now is poised for 19 … c5!. White anticipated him with 19 e5!. He has two threats and Black could have stopped both with 19 … h6! 20 e4 f8 (although 21 c5 and 22 d2 still favors White). Black saw only the positional threat of 20 g5, which attacks h7 and heads the knight to e4 and then c5 or d6. He felt it was time for 19 … c5?.

He was surprised by the ancient 20 xh7+! xh7 21 g5+ sacrifice. Then came 21 … g8 22 h3 (threatening 23 h4 as well 23 h8+! xh8 24 h4+ and mates) 22 … e8 23 h4 f8 24 h7+ g8 25 f6+! and wins. See also Supplemental Game # 3 for a modern version. We saw how f2-f4 can stop … e5 effectively in a Caro formation – provided that Black cannot exploit White’s weakening of the g1-a7 diagonal with … c5. A similar situation arises in a Slav when Black has blocked his c-pawn. Consider Taimanov – Novak, Harrachov 1966: 1 d4 f6 2 c4 e6 3 c3 b4 4 e3 c6 5 e2 d5 6 a3 f8!? 7 g1 dxc4 8 xc4 d6.

White to play Black could have retreated his bishop to d6 at move six but was concerned about the space-gaining 7 c5. Now he’s ready for 9 … e5. But White made his own retreat at move seven so he could stop him with 9 f4!. If Black doesn’t change the pawn structure he will be clearly inferior after e3-e4-e5. White understood that Black needs several moves to play the desirable … c5. For example, 9 … a5 10 a2 c5? loses to 11 dxc5 and 12 b4. Black didn’t appreciate how quickly his position would worsen and he was lost soon after 9 … 0-0 10 f3 b6? 11 e4 e7 12 e5! d5 13 xd5 exd5 14 d3 and then 14 … g4 15 e3 d7 16 c1 a5? 17

h3 f5 18 xf5 xf5 19 xc7. Not much better was 17 … xf3 18 xf3 c6 19 f5!.

White plays e2-e3 The Slav formation we’ve focused on so far features White pawns on d4 and e4. When White’s e-pawn stops at e3, a new set of options arises: 1 d4 d5 2 c4 c6 3 c3 f6 4 f3 e6 5 e3 bd7 6 c2 d6 7 e2 0-0 8 0-0 dxc4 9 xc4 e5 10 d1 e7:

White to play By keeping his pawn at e3 White can meet … exd4 with exd4!?. That creates an isolated d-pawn which keeps Black’s pieces off of e5 and c5. Black has an alternative to … exd4, pushing his pawn to e4. Then he can reinforce it and enjoy chances of kingside attack (see the Wedge formation in Chapter Five). For example, 11 a3 e4 and now 12 g5 allows 12 … xh2+ 13 xh2 g4+ and … xg5. Suppose White plays 11 h3, which safeguards against the … xh2+ trick. Then Black can set up the Wedge with 11 … e4 12 g5 b6 13 b3 f5 with a threat of 14 … h6, embarrassing the knight at g5.

White to play White should liquidate with 14 f3! and then 14 … ae8 15 gxe4 xe4 16 xe4 xe4 17 xe4 xe4 18 fxe4 xe4. Then he has the only center pawns. They can become a target after 19 f1 c5!. But they will be a strength after 19 c2 e7 20 e4!, which favors White slightly.

Black plays … c5 While … e5 is the most unbalancing stroke in the Slav formation, Black can also attack the center with his c-pawn. This is a standard device of the Queen’s Gambit Accepted and several lines of the Gambit Declined. Once again White has three options. He can allow an exchange of pawns on d4. If he can recapture exd4, he reaches isolated d-pawn positions we’ll examine in Chapter Seven. The second strategy is to trade pawns on c5 and create a symmetrical structure. White often tries to exploit it by pushing his e-pawn. For example, Karpov – Anand, Brussels 1991 went: 1 d4 d5 2 c4 c6 3 f3 f6 4 c3 e6 5 e3 bd7 6 c2 d6 7 e2 0-0 8 0-0 dxc4 9 xc4 e7 10 h3 c5!? 11 dxc5 xc5 12 e4.

Black to play White is preparing 13 e5! followed by moves such as He can meet 12 … e5 with 13 g5 and d5.

e4 and g5.

But 12 … d6! was a good defense that White should have answered by 13 b5 followed by quickly mobilizing his QB and QR. Instead he preferred 13 d4? and accurate play gave Black the edge, 13 … e5! 14 b3 d7 15 e3 g6! 16 ad1 (16 f4 e5!) 16 … fd8

17 f3 c6 and … b4. This illustrates the dangers to both sides in symmetrical positions. The most aggressive way for White to handle … c5 is to push his dpawn. He tries to exploit his superiority in the center before Black’s queenside majority matters, e.g. Alekhine – H. Steiner, Bradley Beach 1929: 1 d4 d5 2 c4 dxc4 3 f3 f6 4 e3 e6 5 xc4 c5 6 0-0 a6 7 e2 bd7? 8 c3 c7.

White to play Black is mixing two ideas that don’t jell. If he isn’t going to expand immediately with 7 … b5 – and fight for control over d5 with … b7 – then he should develop his QN on c6, where it watches d4 and can go to b4 to attack a bishop if it retreats to d3 or c2. You can see the … xc2 idea work in Supplemental Game # 4. Black’s error allowed a strong 9 d5!. Keeping the center closed with 9 … e5 10 g5 d6 allows White too much play with 11 a4 b6 12 ge4. So the game went 9 … exd5 10 xd5! d6 11 e4! 0-0 12 g5 and it was clear White’s pieces were better placed (12 … e5 13 xf6). After one more mistake, 12 … g4? 13 h3 ge5, White established a winning edge by freeing his f-pawn to advance, 14 h4!. The rest was easy: 14 … b6 15 f4 c6 16 f5! e5 17 h5 e8 18

f4 e7 19 f6 f8 20 fxg7 xg7 21 af1 e6 22 xg7! g6 24 xe8 xe8 25 xd5 Resigns.

f5 xd5 23

But d4-d5 is not dangerous if Black can compete for control of d5. For example, suppose in the last example that he had played 7 … b5 8 b3 b7 and then 9 c3 bd7 10 d1 c7, reaching a book position.

White to play Now 11 d5 initiates a harmless trade of pieces and pawns (11 … xd5 12 xd5 xd5 13 xd5 exd5 14 xd5 and 14 … e7 15 e4 0-0 16 g5 fe8 followed by … b6). White should always be looking to push his e-pawn in this kind of structure. The reason Black meets 11 d5 with 11 … xd5 is that 11 … exd5 gives White a dangerous extra option – 12 e4!? dxe4 13 g5, with promising chances. Of course, White could just play 11 e4 in the diagram. But this allows Black good play by means of 11 … exd4! 12 xd4 c5. There is another option and it belongs to Black after a d4-d5 push. He can ignore that advance and expand on the queenside with … c4. For example, 11 d5 in the diagram leads to sharp play if Black chooses 11 … c4!? and then 12 dxe6 fxe6. Yes, Black’s pawns are loosened. But his piece activity usually makes up for that (13 c2 d6 14 h3 0-0 and … c5). This is a highly doubleedged situation and appears in slightly different form in both the Queen’s Gambit Accepted and the Meran Variation of the Slav (1 d4 d5 2

c4 c6 3 f3 f6 4 c3 e6 5 e3 bd7 6 d3 dxc4 7 xc4 b5 8 b3 a6 9 e4 c5 10 d5!? and now 10 … c4 12 dxe6 fxe6). This has been popular at the master level for more than a decade and it’s easy to find instructive games to learn more.

Supplemental Games # 1 Black fails to blockade after … c5/d4-d5: Polugayevsky – Biyiasas, Petropolis 1973 – 1 d4 f6 2 c4 e6 3 f3 d5 4 c3 c6 5 e3 bd7 6 d3 dxc4 7 xc4 b5 8 d3 a6 9 e4 c5 10 d5 e5 11 b3 d6 12 0-0 00 13 e1 b8 14 f1! e8? (Black should defend b5 with 14 … e8 15 a4 c7.)

White to play 15 a4! b4? (Because he can’t blockade on d6 after this, Black should sacrifice a pawn with 15 … c4 16 axb5 axb5 17 bxc4 b4! and … c5.) 16 b1 b6 17 bd2 e7 18 b2 e8 19 c1 f6 20 a5 a8 21 c4 c7 22 fd2! d7 23 f4! e7 24 fxe5 fxe5 25 xd6 xd6 26 c4 xc4 27 xc4 Resigns (28 d6 was threatened and 27 … d6 walks into 28 xe5 xe5 29 d6+.). # 2 Black mixes … c5 with … e5: Beliavsky – Berkes, Paks 2004 – 1 d4 d5 2 c4 c6 3 f3 f6 4 c3 e6 5 e3 bd7 6 c2 d6 7 d3 0-0 8 0-0 dxc4 9 xc4 a6 10 e4 e5 11 d1 c7 12 g3 b5 13 f1 c5!

White to play 14 dxe5 (Black has an excellent game after 14 d5 c4! because of his queenside pawns, e.g. 15 a4 b4 16 b1 c5.) 14 … xe5 15 xe5 xe5 16 g5 b7 17 g2 c4 (When White exchanges on e5 but can’t mobilize his kingside pawns, he has no edge. Here 18 f4? b6+ 19 h1 g4 is bad.) 18 ac1 h6 19 xf6 xf6 20 a4 b6 21 d5! xd5 22 xd5 ad8 23 cd1 xd5 24 xd5 d8 25 xd8+ xd8 26 axb5 axb5 27 f1 a5 28 b3 c3 29 d3 d4 30 e2 a3 Draw. # 3 White punishes passivity with e4-e5!: Kasparov – Short, London 1993 – 1 f3 d5 2 c4 e6 3 d4 f6 4 c3 dxc4 5 a4+ bd7 (5 … c6 6 xc4 b5 7 d3 a6 and 8 … c5! is a good alternative.) 6 e4 a6 7 xc4 c6 8 d1 e7?! 9 0-0 0-0 10 a4 b6 11 d3 b7 12 e5! d5 13 xd5 cxd5 14 d2 a5 15 c1

Black to play (White has a very favorable version of the e5 chain discussed in

Chapter Five.) 15 … b8 16 b5 a6 17 b3 c8 18 c3 c7 19 d3 b4? 20 xh7+! xh7 21 g5+ g8 22 h3 e8 23 f3 d7 24 h5 f8 25 h7+ e7 26 g5+ f6 27 xf6 Resigns. # 4 Black expands with … c4 and wins control of e4: Evans – Bisguier, U.S. Open 1950 – 1 d4 d5 2 c4 dxc4 3 f3 f6 4 e3 e6 5 xc4 c5 6 0-0 a6 7 e2 c6 8 d1 b5 9 b3 c4!? 10 c2 b4!

White to play (Now 11 e5 b7 12 c3 xc2 is excellent for Black, e.g. 13 xc2 e7 14 a4 b4 15 e2 c8.) 11 e4 xc2 12 xc2 b7 13 d5! c7! (Not 13 … exd5 14 c3 e7 15 e5! with a dangerous initiative.) 14 g5 d7 15 dxe6 fxe6 16 a4! b4! (Otherwise axb5/ c3 pressures the queenside.) 17 axb5 axb5 18 xa8+ xa8 19 c3 xc3! 20 xc3 00 21 e7 e8 22 d6 c6 23 e5 c5

White to play

(Black has won control of key light squares.) 24 e3 d3 25 b3 e4! (The endgame favors his queenside pawns.) 26 xe4 xe4 27 d4 c3! 28 f3 g6 29 a1 c8 30 f1 h5 31 e2 c2 32 c1 h7! 33 e2 a8! 34 xa8 xc1+ 35 e3 d3! (Better than 35 … xb3 36 a3.) 36 a3 b4 37 b2 xb2 White resigns.

Chapter Three: The Open Sicilian/English

The Sicilian Defense is the most popular way of beginning a 1 e4 game and the most popular treatment of it is 2 f3 and 3 d4. The Sicilian’s reversed form, the 1 c4 e5 English Opening, has grown in popularity – particularly after White realized he should not try to force matters in the center with d2-d4 and should wait for Black to play … d5. There are four major branches of the Open Sicilian and we’ll devote a subchapter to each of them.

A. The Scheveningen Formation Almost no non-Dutch person can pronounce ‘Scheveningen’ correctly and for that reason it was used as a password by Resistance fighters in the Netherlands during World War II. The pawn structure that bears its name is almost as difficult to deal with flawlessly.

The Scheveningen formation This is the most common Open Sicilian formation and it is by no means limited to the Scheveningen Variation, which can come about via 1 e4 c5 2 f3 d6 3 d4 cxd4 4 xd4 f6 5 c3 e6, for example. The sharp Sozin and Rauzer lines rely on the same pawn structure and so does the Najdorf when Black plays … e6. Variations with an early … e6, such as the Kan and Taimanov, become Scheveningen formations after … d6. If White doesn’t change the pawn structure his middlegame options are largely limited to kingside attack or pressure against the d6-pawn. More promising are these four plans: (a) White can play e4-e5, a prelude to kingside or central attack. (b) He can try f2-f4-f5, pressuring e6 in an effort to win control of d5. (c) He can launch a kingside pawn storm with g2-g4-g5. (d) He can try the queenside option of xc6 followed by c2-c4.

Timing is much more important than in the structurally solid CaroSlav positions. The price of error is also higher. A bad move in a Caro or Slav position may mean slight inferiority. A bad move in an Open Sicilian may be enough to lose the game. The most explosive option is e4-e5 supported by a pawn on f4. This is double-edged because it may create a weak e-pawn after … dxe5/fxe5. But the opening of lines – the d-file and half of the f-file as well as part of the b1-h7 and h1-a8 diagonals – can make this push worth the risk: 1 e4 c5 2 f3 d6 3 d4 cxd4 4 xd4 f6 5 c3 e6 6 e2 c6 7 e3 e7 8 0-0 a6 9 a4 0-0 10 f4 d7 11 b3 a5.

White to play White’s 9 a4 stopped … b5-b4, Black’s main counterplay source, and prepared a cramping a4-a5. Black averted a4-a5 but allowed 12 e5!?. It’s almost always a good sign for White when Black cannot answer e4e5 with … dxe5. Here 12 … dxe5 is punished by 13 fxe5 d5 14 xd5 exd5 15 xd5 or 13 … e8 14 xa5 xa5 15 xd7. After 12 … e8 play may run 13 xa5 xa5. White has a choice of ideas, such as 14 e4, threatening 15 exd6. But the most common response in master chess is 14 d2 and 14 … c7 15 d4.

Black to play When the first edition of this book appeared, Black’s position was considered risky. If he closes the center, 15 … d5?!, White gets a good version of the Wedge pawn structure (Chapter Five) and can attack the kingside with 16 e3 c8 17 d3! c5 18 e2! and f3-h3. Black’s natural alternative, opening the center with 15 … dxe5 16 fxe5 in order to trade bishops with 16 … c5, leads to a dominating White presence in the center after 17 e4! xd4+ 18 xd4. In Parma – Larsen, Teesside 1972, White won fairly easily after 18 … c6 19 ad1 a5 20 c5! c7 21 a5 g6 (or 21 … xa5 22 xe6! fxe6 23 xf8+ with a winning attack, e.g. 23 … xf8 24 f4+ g8 25 f1 or 24 … e7 25 g5+ f8 26 f1+) 22 b4! g7 23 d6! xd6 24 exd6. But the downside of e4-e5 was apparent in later games. After 15 … c6 (instead of 15 … d5 or 15 … dxe5) Black controls a key diagonal. His knight can emerge at f5, after … g6 and … g7. And a pawn trade on e5 leaves the e-pawn more of a weakness than strength. For example, 16 e3 dxe5 17 fxe5 g6 and 18 e4 xe4! 19 xe4 c5! 20 c3 g7. Black eventually won in Anand – Ponomariev, World Blitz Championship 2009. Note that if Black doesn’t meet e4-e5 with … dxe5, White can change the pawn structure further. His choice is between exd6 – which often liquidates whatever edge he had – and the explosive f4-f5!?, as in Supplemental Game # 1 at the end of this section.

White’s f4-f5 Plan The strategic goal of f4-f5 is to force the e6-pawn to advance or be traded off. White hopes to win control of d5 and occupy it with a minor piece, ideally a knight. Also, f4-f5 gains kingside space for an attack on the king, as in Short – Kasparov, World Championship match 1993: 1 e4 c5 2 f3 d6 3 d4 cxd4 4 xd4 f6 5 c3 a6 6 c4 e6 7 b3 bd7 8 f4 c5 9 f5 e7 10 f3 0-0 11 e3 e5 12 de2 b5.

White to play Black’s b-pawn and the likely … xb3 offers him counterchances, such as after 13 g4 b7 14 g3 b4! and 15 d5 xb3. When White ends up with a pawn, rather than a piece, on d5, he can usually claim an edge only if he can occupy e4 with a knight. White wanted a piece on d5 and chose 13 d5!. He would stand well after 13 … xd5 14 xd5. Black preferred 13 … b8 and White ended the … b4 threat with 14 b4!? cd7 and then 15 0-0 xd5 16 xd5 b7 17 ec3. He keeps control of d5, but made a big concession on the c-file. This was clear after 17 … f6 18 ad1 xd5 19 xd5 xd5 20 xd5 c8.

White to play If his pawn were still on b2, White could neutralize Black pressure on the file with 21 c3. That doesn’t work now and if he defends c2 with a piece, Black gets the upper hand with 22 … c4! and … c7/… c8. However, the benefits of f4-f5 enabled White to ignore the threat to c2. He chose 21 g4! with ideas of h6 and f5-f6. Black defended, 21 … f6, but 22 f3! xc2 23 h3 f7! 24 h5 h6 25 g6 threatened xh6. Black’s 25 … f8 was forced and White took a perpetual check draw following 26 xh6. But he would have had good winning chances after 26 h7! and then 26 … e8 27 g8+ f8 28 c5! or 27 … f8 28 xg7. The importance of controlling d5 after f4-f5/… e5 is further shown by Supplemental Game # 3. White loses the fight for it and that proves fatal.

White’s g4-g5 Plan Siegbert Tarrasch called g2-g4 the ‘hara-kiri move’ because it seemed suicidally loosening. But in the Sicilian the push to g5 has a positional punch. It makes … d5 difficult to achieve (except as a sacrifice). And by driving a knight off f6, it makes Black think twice about … e5. But the main purpose of g2-g4-g5 is to attack the king. There are two basic cases of the push: When it is accompanied by f2-f4 and when accompanied by f2-f3. The latter has become increasingly common in opening systems such as the English and Velimirovic attacks: 1 e4 c5 2 f3 e6 3 d4 cxd4 4 xd4 c6 5 c3 c7 6 e3 a6 7 d2 f6 8 f3 e7 9 g4 d6 10 0-0-0 0-0 11 g5 d7.

White to play By keeping his pawn at f3 White supported g2-g4 and protects the epawn, one of Black’s natural targets. White does not need f2-f4-f5 to open lines because he can do it with h2-h4-h5 and g5-g6. In fact, if White feels the need to attack is urgent, he can push the g-pawn as a sacrifice: After 12 h4 b5 there is a good maneuver, 13 xc6 xc6 14 e2! and d4, as in similar positions. But White can attack immediately with 13 g6!? and 13 … hxg6 14 h5 (not 13 … fxg6? 14 xe6 or 13 … xd4 14 gxh7+.).

In Adams-Sheldon, Hove 1997 Black was in trouble after 13 g6 f6 14 gxh7+ xh7 15 xc6 xc6 16 d3 h8 17 dg1. The game ended with 17 … b4 18 h6! g8 19 e5! (threatening 20 xg7+ xg7 21 h6+ and mates) 19 … g6 20 exf6 bxc3 21 g5 Resigns. See also Supplemental Game # 4, which shows a White pawn sacrifice to keep the center closed after … d5.

White’s xc6/c2-c4 Option When the Scheveningen Variation was young, even strong players felt … cxd4 and … a6 weakened queenside squares and pawns that could be exploited by xc6. Spielmann – Euwe, Bad Kissingen 1928 began: 1 e4 c5 2 f3 c6 3 d4 cxd4 4 xd4 f6 5 c3 d6 6 e2 e6 7 0-0 e7 8 e3 0-0 9 d2 a6.

White to play Black’s last move rules out an attack on the d6-pawn by means of ad1 and 4b5. White played 10 xc6?! bxc6 and then 11 fd1 c7 12 f3. He wants to pressure the queenside with f1 and f2 or aim at the isolated a-pawn with e1-f1. However, Black stood well after 12 … d5!, with … b7, … fd8 and … e5 in mind. White’s queenside pressure ran out of steam while Black’s chances grew following 13 exd5 cxd5 14 e1 b8 15 ab1 d6 16 g3 b7 17 f2 e5 18 d2 d4. But xc6 can be good if White adds c2-c4!. Psakhis – Ftacnik, Sochi 1982: 1 e4 c5 2 f3 d6 3 d4 cxd4 4 xd4 f6 5 c3 a6 6 e2 e6 7 f4 e7 8 f3 0-0 9 0-0 c6 10 xc6 bxc6 11 e3.

Black to play White’s plan is 12 a4 and 13 c4. He may then push c4-c5 if he can trade the c-pawn for the d-pawn and create an isolated target pawn at c6. Black replied with the traditional answer to the xc6 plan – 11 … d5 – and play continued 12 e5 d7 13 a4 a5 14 c4! a6 15 b3. The tactics would favor White’s bishops after 15 … bxc4 15 xc6. So the game went 15 … c8 16 f2 a3 (16 … dxc4 17 d2!) 17 d2 b4 18 c1 e7 19 c2.

Black to play Black realized the enemy position will keep getting better if he can play b2, e2, ac1 and a well-timed cxd5. He felt obligated to change the pawn structure. But 19 … c5? would just drop a pawn to 20 cxd5. He chose 19 … f6 instead. But 20 g4! made it hard to defend against xe6+. And since 20 … f5 21 e2 would leave Black with zero

counterplay, he desperately replied 20 … f7. But he was losing after 21 f1!, threatening 22 xe6+ xe6 23 f5+ or 23 cxd5+. The game ended with 21 … g6 22 xe6+ g7 23 xd7 xd7 24 b6 f5 25 xc8 xc8 26 f2 Resigns. See also Supplemental Game # 5.

Black’s Counterplay: … b5/… d5 Black’s best chances in the Scheveningen formation come from threats to the e4-pawn. The most common method is … b5-b4, driving away the best defender of the e-pawn, a White knight on c3. The popularity of the Najdorf Variation stems, to a considerable degree, from that simple plan. White’s methods of safeguarding e4 include a2-a4, b2-b4 and the simple f2-f3. But there is a big downside to f2-f3. Unless it is followed by g2-g4-g5, it surrenders White’s best middlegame options and makes … d5! stronger. For example, Wotulo – Larsen, Manila 1973: 1 e4 c5 2 f3 d6 3 d4 cxd4 4 xd4 f6 5 c3 e6 6 e2 a6 7 e3 c7 8 0-0 b5! 9 a3 b7 10 f3 e7 11 e1 bd7 12 g3 0-0 13 ad1 h8 14 h1 ac8

White to play White’s pieces are developed on reasonable squares. He has no glaring weaknesses. But he has no clear way to improve. Black, on the other hand, can make progress. He can maneuver a knight to c4. He can push his b-pawn with … a5/… b4, and take aim at c2. He can wait for the right moment to push his d-pawn. White’s lack of good options became clearer after 15 b3 a8! 16

d2 e5 17 f2 b8 18 d1 c4!. Then on 19 d3 d5!, Black forced the position open. White cannot play 20 e5 or allow 20 … dxe4.

White to play White was in full retreat after 20 exd5 xd5 21 c1 e5 22 d2 and Black took his time to mount a winning kingside attack: 22 … d6 23 c3 bd8 24 c2 d7 25 g1 h6 26 f2 f5! 27 d1 e7 28 e1 c4 29 d3 e5! 30 b4 a5 31 xd5 xd5 32 ce2 ef7 33 f2 f6 34 h4 c6 35 h3 a4 36 a1

Black to play 36 … e4! 37 f4 c5 38 c2 b6 39 b4 b7 40 g3 g6 41 h4 g4 42 h5 d8 and the threat of 43 … h4 (44 g6 xh2+!), trapping the queen, led to 43 d2 xd2 44 xg4 xb4 and White resigned. Punishing examples like this were so common in the 20th century that

a conventional wisdom arose that White should almost always prefer f2f4 to f2-f3 so that he could meet … d5 with e4-e5!. But this view was overthrown around the turn of the century in situations like: 1 e4 c5 2 f3 d6 3 d4 cxd4 4 xd4 f6 5 c3 a6 6 e3 e6 7 e2 c7 8 d2 b5 9 a3 bd7 10 0-0-0 b7 11 f3 d5 12 exd5 xd5 13 xd5 xd5 14 he1.

Black to play This seems to be a another case of an equalizing … d5. But in this, and similar positions, White has excellent tactical chances simply because of his superior development, e.g., 14 … c8 15 f4 b6 invites 16 xb5! axb5 17 xe6! fxe6 18 xd5 with a terrific attack. Better is 15 … b7 but the prospect of f5 gives White excellent prospects.

Black’s Counterplay: … e5 When White has played f2-f4, Black has an intriguing option in … e5!?. It is similar to situations that arise in the Boleslavsky Hole, seen later in this chapter. Mazzoni – Gligoric, Monte Carlo 1967 began: 1 e4 c5 2 f3 d6 3 d4 cxd4 4 xd4 f6 5 c3 e6 6 e2 e7 7 00 0-0 8 e3 c6 9 f4 d7 10 h1 a6 11 f3 c7 12 e1 ac8 13 d1?

Black to play White’s moves appear logical. But he allowed a good trade, 13 … xd4 14 xd4 e5!. Black would stand well after 15 e3 c6 (not 15 … e6 16 f5! c4 17 f2 and the bishop is misplaced) with … b5 and … b7 in mind. White compounded his problems with 15 fxe5?. He got little benefit from the half-open f-file and had given up on f4-f5. Black’s pieces, particularly the dark-squared bishop and KR, came alive: 15 … dxe5 16 e3 e6! 17 g3 h8 18 e2 b5 19 d3 fd8 20 a3 b8 21 e2 d7! 22 g1 c5 with the idea of … xd3 and … c2. White made another concession, 23 xc5, and found his queenside and e-pawn under fire following 23 … xc5 24 f3 f6 25 g5 xg5! 26 xg5 e8 27 d2 c7 28 fd1 h6 29 h4 e7! 30 xe7 xe7.

He was losing after 31 e2 h7 32 g1 a5! 33 c3 b7 34 d6 b3 35 1d2 b4! 36 cxb4 axb4 37 a6 bxa3 38 bxa3 c2 39 f3 c4 and resigned after 40 e2 b1+ 41 f2 d3! 42 d2 xe4 43 a7 xf3.

Supplemental Games # 1 White plays e4-e5 and f4-f5: Shamkovich – Damjanovic, Sochi 1967 – 1 e4 c5 2 f3 e6 3 c3 c6 4 d4 cxd4 5 xd4 a6 6 e2 c7 7 0-0 f6 8 h1 e7 9 f4 d6 10 e3 0-0 11 e1 d7 12 g3 h8 (Black passes up the … e5 plan, 12 … xd4 13 xd4 c6 14 d3 e5! 15 fxe5 h5 16 -moves dxe5.) 13 a3 ac8 14 d3 b5 15 ae1 b8 16 e5! g8? 17 xc6 xc6 18 d4! b4

White to play 19 f5!! (Now 19 … exf5 20 xf5 bxc3 loses to 21 exd6 f6 22 h3! h6 23 xf6!.) 19 … bxc3 20 f6 xf6 21 exf6 g6 22 xg6! e5 (22 … fxg6 23 f7+ e5 24 xe5!) 23 xe5! xg2+ 24 xg2 dxe5 25 e4 xf6 26 xe5 Resigns. # 2 Black’s bishops exploit e4-e5: Zinn – Monti, Budva 1963 – 1 e4 c5 2 f3 d6 3 d4 cxd4 4 xd4 f6 5 c3 a6 6 c4 e6 7 0-0 e7 8 b3 0-0 9 e3 c6 10 f4 d7 11 h1 xd4 12 xd4 c6 13 e2 c7 14 e5? (The position calls for f4-f5. White may have expected 14 … dxe5 15 xe5.) 14 … d7! 15 ae1 dxe5 16 fxe5 c5 xc5? (This was White’s better bishop.) 17 … xc5 18 f4 ad8 19 h4 d4! 20 h3 fd8 21 h5

Black to play 21 … d2! 22 xh7+ f8 23 h8+ e7 24 h4+ e8 25 e4 xe5! 26 xd2 xd2 (Since 27 xe5 xg2 is mate.) 27 f3 xf3 28 gxf3 xh2+ 29 xh2 xe1+ and Black won. # 3 Black meets f4-f5 with … e5 and … d5: Bena – Pavlov, Bucharest 1969 – 1 e4 c5 2 f3 c6 3 d4 cxd4 4 xd4 f6 5 c3 d6 6 c4 e6 7 0-0 a6 8 e3 c7 9 b3 a5 10 f4 b5 11 f5

Black to play 11 … e5! (Keeping the center fluid, 11 … xb3 12 cxb3! e7, allows a promising 13 c1 d7 14 f3 b7 15 h3 or 14 … 0-0 15 e5! b7 16 exf6! xf3 17 fxe7.) 12 de2 b7 13 g3 (If White occupies d5, 13 d5 xd5 14 xd5, Black gets a fine game with 14 … c4! 15 c1 xd5 16 exd5 c8 or 15 c1 xd5 16 xd5 c8.) 13 … c4 14 xc4 xc4 15 f3 h5! 16 h4

Black to play 16 … d5! 17 xd5 xd5 18 b3 c6 19 exd5 xd5 20 e2 e7 21 ad1 c6 22 d2 d8! 23 xd8+ xd8 24 g5 xg5 25 hxg5 h4! 26 d1+ c8 27 f1 h3 28 e3 b6 29 f2 hxg2 White resigns. # 4 White sacrifices to keep the center closed after … d5: Bologan – Ye Jiangchuan, Beijing 2000 – 1 e4 c5 2 f3 d6 3 d4 cxd4 4 xd4 f6 5 c3 c6 6 g5 e6 7 d2 e7 8 0-0-0 0-0 9 f3 a6 10 h4 xd4 11 xd4 b5 12 b1 b7 13 d2 c7 14 e2 ac8 15 d4 fd8 16 d3 e8 17 g4 f6 18 e3! (On 18 xf6 xf6 Black is ready for 19 … d5!.) 18 … d5

White to play 19 e5! (Now 19 … xe5? 20 g5 e7 loses to 21 f4 xd4 22 xh7+.) 19 … xe5 20 f4 d6 21 g5 e5 22 fxe5 xe5 23 g6! (White’s 19th move bought time for this, e.g. 23 … hxg6 24 h5 gxh5 25 dg1

and g2.) 23 … d6 24 gxf7+ xf7 25 hf1 d7? (25 … f5 is unclear.) 26 f3! f6 27 g5

e8! 26

Black to play 27 … c4 (Or 27 … h6 28 xf6 gxf6 29 h7 and d4/ xh6.) 28 xh7+ h8 29 e2 e8 30 xf6! gxf6 31 d4! c6 32 f2 f8 33 d3! d7 34 f5 e7 35 h7! c8 36 h5 f7 37 g6 Resigns. # 5 White wins with xc6/c2-c4: Topalov – Kramnik, Novgorod 1996 – 1 e4 c5 2 f3 c6 3 d4 cxd4 4 xd4 f6 5 c3 d6 6 c4 b6 (By attacking the d4-knight Black stops b3/ e3/f2-f4-f5) 7 xc6 bxc6 8 0-0 e6 9 e2 d7 (Black averts 10 e5, which would isolate his c-pawn.) 10 b3 e7 11 b2 0-0 12 a4! c7 13 f4 b7 14 ad1 ae8 15 d3 c5 16 b5! c6 17 xc6 xc6 18 c4!

Black to play (Black has no counterplay so White can test the defense of d6, g7 and

h7.) 18 … d8 19 fd1 fe8 20 f3 c7 21 g3 f6 22 c3 a6 23 h3 c6 24 g3 h8 25 e1 c8 26 ee3 g8 27 d1 d7 28 f5! e5 29 c3 d8

White to play 30 d5! h6 (Or 30 … xd5 31 xh7+! xh7 32 h3+ and mates.) 31 c1 xd5 32 exd5 f6 33 e4 f8 34 eg4 and Black resigned before 35 xh6!. # 6 White stops … b4 with b2-b4: Larsen – Portisch, London 1986 – 1 e4 c5 2 f3 d6 3 d4 cxd4 4 xd4 f6 5 c3 a6 6 d3 e5 7 de2 e7 8 0-0 bd7 9 e3 b5 10 g3 g6 10 b4 (Another promising idea is 10 a4 b4 11 a2 followed by c2-c3.) 10 … b7 11 a4! bxa4 12 xa4 b6 14 a5

Black to play (White piles up on the a6-pawn, e.g. 14 …

c7 15 a1.) 14 …

c8

15 xb6 xc3 16 a1! c6 17 e3 0-0 18 h6 fb8 19 h3 d8 20 a3 d7? 21 c3! b6 22 a4! (Threatening e3 as well as xd7) 22 … c8 23 xc8 xc8 24 xd7 h4 25 g4 e7 26 d2 f8 27 a1 h5 28 e2 Resigns.

B. The Dragon Formation

The Dragon formation You don’t see a Dragon? Don’t be surprised. The name comes from a constellation, not a monster. A master named Fyodor Duz-Khotimirsky was studying astronomy a century ago when it struck him that Black’s pawns from d6 to h7 resembled Draco the Dragon. The name stuck. The Dragon formation comes about after several Sicilian move orders, including 1 e4 c5 2 f3 d6 3 d4 cxd4 4 xd4 f6 5 c3 g6 and 1 e4 c5 2 f3 c6 3 d4 cxd4 4 xd4 g6. It also arises from Pirc and Modern Defenses when Black initiates a pawn exchange with … c5. What distinguishes all members of the Dragon family is, of course, Black’s kingside fianchetto. (We’ll call him Black for simplicity. In various English lines, such as 1 c4 e5 2 c3 f6 3 g3 d5 4 cxd5 and 3 f3 c6 4 g3 d5 5 cxd5, it is White whose pawns will form the Dragon.) The bishop at g7 fires at an enemy center loosened by … cxd4 and beyond that at c3 and b2. Unlike in the Scheveningen, Black rarely plays an early … e6 because without support from his bishop at g7, his d6pawn would come under heavy fire. White has three basic anti-Dragon plans: (a) He can castle kingside and attack with f2-f4.

(b) He can castle queenside and attack with f2-f3, g2-g4 and/or h2-h4. (c) He can occupy d5 with a knight.

White’s f2-f4 Attack The Dragon has relatively little to fear from e4-e5 because Black’s kingside is not usually vulnerable to the d3/ f3-h3 or g3 attacks seen in the Scheveningen. More solid is f4-f5. With neither a bishop nor a pawn on e6, Black has to be concerned about d5!:

Black to play This comes about via 1 e4 c5 2 f3 d6 3 d4 cxd4 4 xd4 f6 5 c3 g6 6 e3 g7 7 e2 c6 8 0-0 0-0 9 d2 and then 9 … g4 10 xg4 xg4 11 f4. Black’s ninth move wins the two-bishop advantage by threatening … xe3 (and punishing 10 g5? with 10 … xd4). White’s 11th move threatens to trap the bishop with 12 f5 and 13 h3. But Black has 11 … xd4 12 xd4 e5!. This is another form of the … e5 strategy we saw a few pages ago. Then 13 fxe5? dxe5 leads to an excellent position for Black. And 13 e3 exf4! grants him good piece play. Black’s d6-pawn seems weak then but after 14 xf4 e6 and … e5 he has few problems (and after 14 xf4 b6+ and 15 … xb2 he has the edge). But suppose back at the diagram Black retreats his bishop immediately with 11 … e6? instead of 11 … d7. This turns out to be faulty – but not because of 12 xe6 fxe6, when the doubled pawns can help control the center. Rather it is that White can play f4-f5!, e.g. 12 ad1 a5 13

f5!.

Black to play This works tactically because 13 … xd4 14 xd4 xd4+ 15 xd4 gxf5 invites 16 exf5 xf5? 17 d5! with threats of xe7 mate and xf5. But it succeeds strategically because it allows White to occupy d5 with an unchallenged knight. After 13 … d7 14 f2 he prepares d5, h4 and either g5 or h6. In Rodl – Herman, Bad Pyrmont 1949, White’s superiority was manifest following 14 … xd4 15 xd4 xd4 16 xd4 c6 17 f6! and he won after 17 … ae8 18 b4 and b4-b5, e.g. 18 … b6 19 xb6 axb6 20 b5 d7 21 fxe7 and d5 or 18 … g5 19 b5! xb5 20 h4 h5 21 fxe7 xe7 22 d5! and xe7+ or f6+. The chief drawbacks to the f2-f4-f5 plan are: (a) White’s e-pawn can become a target, since f2-f3 is impossible. (b) He gives up pawn control of e5, a great square for a Black knight. (c) There can be tactical problems on the b6-g1 diagonal. Let’s return to the position after 1 e4 c5 2 f3 d6 3 d4 cxd4 4 xd4 f6 5 c3 g6 6 e2 g7 7 0-0 0-0 8 e3 c6. Now 9 f4 – instead of 9 d2 – is premature because 9 … b6! threatens both 10 … xb2 and 10 … xe4 11 xe4 xd4.

White usually plays 9 b3, which safeguards against tactics as well as deterring … d5. Then 9 … e6 10 f4 a5 is common. Black will not be able to play … e5 easily but he controls c4. Therefore 11 f5 makes sense – and so does 11 … c4!, which is based on 12 xa5 xe2 and then 13 xe2 xa5. Trading pieces eases Black’s game a bit. But 14 g4! threatens to obtain a strong position with 15 g5 -moves 16 d5!.

Black to play When Tigran Petrosian was 15 he played 14 … b4?, threatening 15 … xe4 and 15 … xb2. But his opponent’s 15 ad1! was strong in view of 15 … xe4 16 d5!. Petrosian made matters worse with 15 … xb2 since 16 d4! prepared 17 g5 -moves 18 d5!. He relied on the tactics of 16 … xg4, e.g. 17 xg7? xg7 18 xg4 xc3 or 18 d5 f6 defends. But he missed 17 f6! and lost (17 … xf6 18 d5 a3 19 xe7+ or 17 … xf6 18 xf6! xf6 19 d5 xa2 20 xe7+ g7 21 f1 and xf6+). Well then, suppose that in the diagram Black treats the position like a Scheveningen and meets 14 g4 with 14 … d5 ? That’s better than Petrosian’s idea but 15 e5! d7 16 d4 favors White’s greater space and pieces. A third defensive idea is to occupy the e5 square that White relinquished with f4-f5. But here 14 … d7 isn’t quick enough in view

of 15 d5 fe8 16 g5 e5 17 f6! with an edge. It turns out that the situation is so urgent after 14 g4! that Black’s best policy is to sacrifice the Exchange with 14 … ac8 and 15 g5 and …

Black to play … then 15 … xc3!. Black has more than enough compensation after 16 bxc3? xe4. Better is 16 gxf6 but 16 … xe3 17 xe3 xf6 should equalize at least. His king is much safer than White’s and he has ample counterplay (18 c3 c8 19 a3 c4 and … e5 or … b5/… a5).

White’s f2-f3 Attack After f2-f3 White typically throws his kingside pawns forward and tries try to trade off Black’s KB with e3/ d2/ h6xg7. When Bobby Fischer popularized another wrinkle, posting the king bishop on c4, White’s attack won so often that it seemed like all he needed to do was “Sac, sac and mate!” as Fischer put it. Black’s difficulty is underlined by what happens if he takes the logical step of eliminating one of the attacking bishops: 1 e4 c5 2 f3 d6 3 d4 cxd4 4 xd4 f6 5 c3 g6 6 e3 g7 7 f3 c6 8 d2 0-0 9 c4 a5 10 b3! xb3 11 axb3!.

Black to play White retook with a pawn because it strengthens his intended king position (after 0-0-0). Black can’t occupy c4 now and won’t get much benefit from a … xc3 sacrifice. So where will Black’s counterplay come from? He can’t attack the e4 pawn, as in f2-f4 Dragons. And … d5 is dubious, e.g. 11 … d5 12 e5! e8 13 f4 since White has holes to exploit and pawns to attack after 13 … f6 14 exf6 xf6 15 0-0 and ad1. As a result, White is free in the diagram to attack, e.g. 11 … d7 12 h4! a6 13 h5!. He can win without his QR. For instance, if Black grabs

the pawn, 13 … xh5 14 g4 f6, White’s attack rolls on with 15 h6 followed by 16 xg7 xg7 17 h6+. The bleakness of Black’s chances was shown when Yefim Geller tried this line in a 1965 Candidates match and was crushed by Boris Spassky after 13 … c8 14 h6 e5 15 de2 e6 16 g4 c7? (Better is 16 … xh6 17 xh6 e7.) 17 g3 b5 18 b4!. For example, 18 … c6 19 xg7 xg7 20 hxg6 fxg6 21 h6+ and 22 g5. For another example of how White’s attack is more dangerous than Black’s, see Supplemental Game # 1. The chief liability of White’s plan is that his advanced pawns can turn out to be overextended in an endgame. Suppose: 1 e4 c5 2 f3 d6 3 d4 cxd4 4 xd4 f6 5 c3 g6 6 e3 g7 7 f3 0-0 8 d2 c6 9 0-0-0 xd4 10 xd4 e6 11 b1 c7 12 g4 fc8 13 h4 a5 14 d5?! xd2 15 xf6+ xf6 16 xd2 xd4 17 xd4.

Black to play With 13 … a5 Black was threatening 14 … xc3 and … xa2+ (which works even after 14 xf6 xc3!). White’s 15th move avoids the promising piece sacrifice of 15 xd2 xd5 16 exd5 xd5! 17 xg7 xf3 18 h3 xg4 19 g3 xg7 20 xg4. Therefore the position in the diagram seems both logical and favoring White’s space edge. But Black can raid the kingside with … g7-f6-e5f4.

For example, a 1981 game, Heinola – Haikola, went 17 … c5 18 h5 g7 19 hxg6 hxg6 20 g1? g5! and Black soon had an edge, 21 d1 h8! 22 d3 h3 23 df1 c8 and … ch8/… f6-e5. And Burgalata – Trifunovic, Mar del Plata 1953 went 17 … g7 18 g5 h6! 19 h3? xh3 20 xh3? (20 gxh6+!) 20 … hxg5 21 hxg5 h8! and Black won after 22 xh8 xh8 in view of 23 d1 h3 24 f1 g3. Nevertheless there are Dragon endgames that favor White as in Supplemental Game # 2.

Positional Plan: d5 One of the most favorable Dragon pawn structures for White appears below. Black’s e-pawn is backward and can be attacked by heavy pieces. White’s protected d-pawn has a cramping effect and can take en passant if Black pushes his e-pawn.

After the ‘Marco Hop’ The basic structure arises in several openings, such as the English Opening, King’s Indian Defense and Alekhine’s Defense, e.g. 1 e4 f6 2 e5 d5 3 d4 d6 4 c4 b6 5 exd6 cxd6 6 e3 g6 7 d5. But it arises most often in the Sicilian Defense, when White plants a knight on d5, Black captures it and White retakes with his e-pawn. The idea goes back at least to 1895, when Georg Marco of Rumania used it. The Marco Hop was born. Some favorable pawn structures deteriorate if minor pieces are traded. But even without minor pieces this formation favors White because he enjoys such an edge in space, as in Suchting – Chajes, Karlsbad 1911: 1 e4 c5 2 f3 c6 3 d4 cxd4 4 xd4 f6 5 c3 d6 6 e2 g6 7 e3 g7 8 0-0 0-0 9 h3 d7 10 d2 a6 11 a4 c8 12 xc6 xc6 13 f3 d7 14 d4 e5 15 e2 d7 16 d5 e6 17 f4 xd5? 18 exd5 c4 19 xc4 xc4 20 xg7 xg7 21 b3! b6+22 h2 d4

23 e2 c7?.

White to play After 24 c4! White can pile up against e7 or open the kingside with f4f5. Black helped him with 24 … f6 25 ae1 b8 26 f5 gxf5?. Once White recaptured on f5, 27 d3 g8 28 xf5, he targeted f7 and h7. Play went 28 … g6 29 ef1 h8 30 c3+ f6 31 1f2 b5 32 5f4. Black resigned soon after 32 … bxc4 33 g4! d3 34 xg8+ xg8 35 a5 a7 36 f3 e4 37 g3+ f7 38 d8!, threatening mate on g8. White’s position is so favorable after 24 c4! that Black should have avoided it with 23 … c3!. Complications ensue from 24 xe7 xd5. But at least Black is in the game.

Counter-strategies to d5 The best way to combat the Marco Hop is simply to keep enough pieces trained on d5 to prevent d5. Failing that, Black’s best options include (a) anticipating the hop with … d5 and (b) allowing an exchange on d5 but creating counterchances from … e6 or … b5. Anticipation is a familiar theme in Dragon lines such as 1 e4 c5 2 f3 c6 3 d4 cxd4 4 xd4 g6 5 c3 g7 6 e3 f6 7 e2 0-0 8 0-0? d5!. Neither 9 exd5 b4! and … bxd5 nor 9 xc6 bxc6 10 e5 d7 11 f4 e6 followed by … c5 or … f6 offers White much. This is why he typically stops … d5 with 8 b3 – or earlier with 5 c4, the Maroczy Bind. The same idea has been used in reversed form since the English Dragon was young: 1 c4 e5 2 c3 f6 3 g3 d5 4 cxd5 xd5 5 g2 e6 6 f3 f6 7 0-0 c6 and now 8 d4! exd4 9 b5! and fxd4, with better pieces and, thanks to … f6, better pawns, in Alekhine – DuzKhotimirsky, Karlsbad 1911. As for the second counterstrategy, we can see how Black’s game goes into a decline when he cannot play either … b5 or … e6, as in Alekhine – Golombek, Montevideo 1939: 1 e4 c5 2 f3 d6 3 d4 cxd4 4 xd4 f6 5 c3 g6 6 e2 g7 7 b3 c6 8 0-0 0-0 9 h1 a5 10 a4! e6 11 f4 c8 12 e3 g4? 13 g1 d8.

White to play Black’s 9 … a5 is a good idea, threatening to loosen White’s queenside with 10 … a4 and 11 … a3. But it gave up on … b5. Black’s 12th move reduced control of d5 and made d5 more likely, compared with 12 … b4! 13 f3 d8 since 14 d5? xd5 15 exd5 allows 15 … xc2!. The difference is that in the diagram 14 d5! is strong and play went 14 … xe2 15 xe2 xd5 16 exd5 b4 17 c4!. White threatens b6 followed by xd8 or xa5. But Black’s real problems are his e-pawn and a lack of counterplay. He should try to solve both with 17 … e6! 18 dxe6 xe6!. Then White’s cand b-pawns can become targets. Instead, he played 17 … c7 and White was freed to play on both wings – 18 d4! dc8 19 b3 a6 20 ae1 e8 21 f5! c5 22 f3 f8 23 b5 d7 24 xc5!. Black resigned soon after 24 … dxc5 25 e3!. The Marco Hop and an exchange on d5 can be followed by c2-c3, rather than c2-c4. This has the benefit of denying Black the possibility of opening lines with … b5, as in Geller – Kan, Moscow 1952: 1 e4 c5 2 f3 d6 3 d4 cxd4 4 xd4 f6 5 c3 a6 6 e2 c7 7 00 g6? 8 g5 bd7 9 d5! xd5 10 exd5 g7 11 c3! f6 12 f3 0-0 13 e1 e8 14 d2 d7 15 h3 ac8 16 e2!

Black to play Black should have shifted into a Scheveningen (7 … e6!). Once White carried out the Marco Hop, he chose 11 c3. If he had pushed that pawn to c4, Black could have pressured d4 and tried for counterplay with … b5. In the diagram White is preparing 17 ae1 which would tie Black to the defense of e7. Black’s solution was 16 … e5 17 dxe6 fxe6. Positions similar to this are occasionally defendable – as 17 … e6! in the Alekhine – Golombek game. But here it allowed 18 d1 cd8 19 f4! e5 20 h4. Black couldn’t play 20 … exd4 21 xf6 xf6 22 xf6 xe2 in view of 23 d5+!. He tried 20 … f8 21 c2! e6 but lost the battle for d5 after 22 b4!. With 23 d5 looming, Black made a desperate bid for air, 22 … d5? 23 g3! f7, but lost after 24 xe5 d4 25 cxd4 h8 26 d5 f5 27 d6.

Supplemental Games # 1 White’s queenside is stronger than Black’s kingside: Gipslis – Bilek, Sousse 1967 – 1 e4 c5 2 f3 d6 3 d4 cxd4 4 xd4 f6 5 c3 bd7 6 c4 g6 7 f3! g7 8 e3 e5 9 b3 d7 10 h4 c8 11 d2 a5? 12 a4! c4 13 xc4 xc4 14 b3! c8 15 0-0-0! 0-0 16 db5 (Black cannot open the queenside further without sacrifices.) 16 … h5 17 d4

Black to play 17 … e5? (A desperate bid to stop e3/ b6 or he1/e4-e5.) 18 e3 d5 19 xd5 xb5 20 axb5 xd5 21 xd5 c7 22 c4 a4 23 b6 e7 24 d6! e8 (The 24 … xd6 25 xd6 axb3 26 c5 endgame is also lost.) 25 b2 axb3 26 xb3 xc4!? 27 xc4 a4+ 28 d3 b5+ 29 d2 a8 30 d3 b2+ 31 e1 xg2 32 f1 c2 33 e2 b3 34 f2 f8 35 a1 c8 36 hc1 e8 37 c4 b2+ 38 g3 Resigns. # 2 The Marco Hop gets a favorable endgame: Mednis – Fedorowicz, New York 1977 – 1 e4 c5 2 f3 d6 3 d4 cxd4 4 xd4 f6 5 c3 g6 6 e3 g7 7 f3 0-0 8 d2 c6 9 g4 xd4 10 xd4 e6 11 0-0-0 a5 12 b1 fc8 13 a3 ab8 14 g5! (Black’s attack is competitive after 14 h4? b5!.) 14 … h5 15 d5! xd2 16 xd2 xd5 (Otherwise xe7+ or xa7.) 17 exd5 xd4 18 xd4 c5 19 h3

Black to play (White is better because his bishop controls the key squares c8, e6 and f5.) 19 … b5 20 e1 f8 21 e3 a5 22 c3 c7 (Black should hold up the queenside pawns, 22 … a4!.) 23 b4! a7 24 b2 ba8 25 h4 g8 26 he4 f8 27 f1 b7 28 h4 g7 29 d3 g8 30 f4 (White’s threat of 31 eh3 h5 32 e2 and xh5 forces Black’s hand.) 30 … axb4 31 cxb4! ba7 32 xb5! f5 33 hh3 xe3 34 xe3 e6 35 c4 exd5 36 xd5 b8 37 b3 f8 38 a4 (The a-pawn wins.) 38 … e8 39 xe8+ xe8 40 a5 d8 41 a4 c8 42 b5! b8 43 b6 e7 44 a6 c7 45 b7! xb7+46 axb7 d5 47 c5 Resigns. # 3 White exploits … b5 in a Marco Hop formation: Botvinnik – Kholodkevich, Moscow 1927 – 1 d4 f6 2 c4 g6 3 c3 g7 4 e4 d6 5 g3 0-0 6 g2 bd7 7 ge2 c5?! (With … bd7, a better plan is … e5.) 8 0-0 cxd4 9 xd4 e5 10 b3 d7 11 b2 c6 12 d5! xd5 13 exd5! xd4 14 xd4 b5

White to play

15 xg7 xg7 16 d4+ g8 17 cxb5! xb5 18 fe1 d7 19 e3 f5? (Black will defend e7 with … f7. But 19 … a5 and … a7 was better.) 20 a4! a6 21 ae1 f7 22 b4! b7 23 b5 a6? 24 b6 c8 25 a5 c5

White to play 26 c3! (Now 26 … xa5 27 c7 or 26 … b5 27 xc5 dxc5 28 e5 is lost.) 26 … xc3 27 xc3 f8 28 e3 e8 29 c1 c8 30 xc8+ xc8 31 e6+! wins a piece.

C: Maroczy Unbound

The Maroczy Bind In 20th Century manuals beginners were warned: ‘Do not take the queen knight’s pawn with your queen.’ But when they became stronger, they encountered more sophisticated advice such as ‘Do not permit the Maroczy Bind.’ The Bind is no longer the terror it once was but still requires great care from Black since his two best sources of counterplay, … d5 and … b5, are discouraged by White’s c-pawn. In addition, a pawn on c4 shortens the reach of Black’s heavy pieces on his half-open file. The Bind arises in two basic forms depending on whether Black plays … g6 (the Maroczy Dragon) or … e6, best known as the Hedgehog. The Maroczy Dragon comes about in the Sicilian Defense (e.g. 1 e4 c5 2 f3 c6 3 d4 cxd4 4 xd6 g6 5 c4) as well as in a variety of King’s Indian and English lines such as 1 c4 c5 2 f3 f6 3 d4 cxd4 4 xd4 g6 5 c3 g7 6 e4. The reversed form occurs in Englishes, such as 1 c4 c5 2 c3 f6 3 g3 d5 4 cxd5 xd5 followed by … e5. The benefit of the Bind, as in other favorable pawn structures, is that enemy passivity is fatal, as in Botvinnik – Golombek, Moscow 1956: 1 c4 f6 2 f3 c5 3 g3 g6 4 b3 g7 5 b2 0-0 6 g2 c6 7 0-0 b6 8 d4! cxd4 9 xd4 b7 10 c3 c8 11 c2! d6 12 e4!.

Black to play We owe Akiba Rubinstein for showing us the strength of d4-c2. That retreat avoids a trade of knights which might ease Black’s constricted position. From c2 the knight watches key squares such as b4, d4 and e3. It’s not obvious that this pawn structure favors White until you consider what happens if Black fails to change it. Instead, Black decided to occupy d4 with his own knight, 12 … d7 13 d2 c5 14 f4 e6?! 15 ad1 ed4? 16 xd4 xd4? but then came 17 d5!. Black had to meet the threat of 18 xe7+ and 18 xd4. After 17 … c6 18 xg7 White would have his choice of preparing c4-c5, e4-e5 and f2-f4-f5, all good plans as we’ll see in pages to come. But Black made matters worse with 17 … xd5? 18 cxd5!. Then 18 … e5 19 dxe6 would cost his d-pawn. He preferred 18 … b5 19 xg7 xg7 but resigned after 20 c1! in view of 20 … b7 21 a4! c7 22 c3+ and xc7. For a more tactical crush, see Supplemental Game # 1. Game # 2 offers an example of a disastrous Bind when minor pieces are misplaced.

White’s Bind Plans White usually has a choice of four good strategies: (a) He can play f2-f4-f5 to win control of d5 and launch a wing attack. (b) He can open the queenside with c4-c5. (c) He can open the center with e4-e5. (d) He can play the Marco Hop, d5. The kingside attack can seem effortless as in Smyslov – Timman, Moscow 1981, which arose from an un-Sicilian-like opening: 1 d4 f6 2 f3 g6 3 g3 c5 4 g2 cxd4 5 xd4 g7 6 c4 c6 7 c3 xd4 8 xd4 0-0 9 0-0 d6 10 d3 f5 11 e4 e6 12 b3 a6 13 b2 d7 14 d2 c5.

White to play Black eased his game with a trade of knights (before White could play c2). He provoked e2-e4 so he could develop his QB (rather than 10 … e6 11 xb7). He repositioned his knight to pressure the queenside. If he can’t engineer … b5 he might do well with … b6 and … a5/… a4!. For example, on 15 d5 xb2 16 xb2 he can try 16 … xd5!? 17 exd5 b6 and then … a5/… b4/… a4 or the liquidation of his main weakness with … fe8 and … e5. For both ideas see Supplemental Game

# 3. But in this game White exploited the Black knight’s absence from the kingside with 15 f4!. He would stand better after 15 … d7 16 f5! gxf5 17 exf5 xf5 18 d5, threatening b6 and xf5/ xe7+. Black thought he could allow f4-f5 and played 15 … c8? 16 f5 d7. But this invited 17 f6!. Black would have no pieces to defend his king after 17 … xf6 18 xf6! exf6 19 d5. Instead, he tried 17 … exf6 18 d5 f5 but the long diagonal proved fatal after 19 exf5 xf5 20 xg7 xg7 21 d4+ f6 22 g4! e6 23 xf6! xf6 24 g5. Another White plan, c4-c5, leads to a somewhat symmetrical pawn structure after … dxc5. This benefits whoever has better placed pieces, particularly rooks. Pachman – Gunnarson, Vrnjacka Banja 1967 began: 1 c4 f6 2 c3 c5 3 f3 g6 4 d4 cxd4 5 xd4 g7 6 e4 d6 7 e2 0-0 8 0-0 c6 9 e3 d7 10 d2 g4 11 xg4 xg4 12 ac1 xd4 13 xd4 e6 14 f4 xd4+ 15 xd4 c8 16 b3 f6.

White to play Now 17 f5 becomes unclear after 17 … gxf5 18 exf5 xf5 19 d5 f7 or just 17 … f7 and … c5. Better is 17 d5 and then 17 … xd5 18 xd5+! (rather than 18 exd5 c5!). The reason is that White favorably opened the c-file, 18 … g7 19 c5! dxc5 20 xc5.

His heavy pieces would dominate after 20 … b8? 21 d7 and c7. Black managed to confuse things with 20 … e6 21 c4 e8 but then 22 e5! was strong since 22 … fxe5 23 xe5 would doom the e-pawn. Black settled for 22 … f5 and was close to paralysis following 23 d1 f7 24 d6!. He set a trap with 24 … d8, hoping for 25 xe6 xe6 26 xe6 d2 or 25 xe6 d1+ 26 f2 d7, when he’s alive. But White ended the game with a pin, 25 c7! fd7 26 dxd7+ xd7 27 b5! Resigns. The third Bind plan, e4-e5, is particularly effective when Black’s pieces are constricted, as in Capablanca – Yates, Bad Kissingen 1928: 1 e4 c5 2 f3 d6 3 d4 cxd4 4 xd4 g6 5 c4! g7 6 c3 f6 7 e2 0-0 8 0-0 bd7? 9 e3 c5 10 f3! d7 11 d2 c8 12 fd1 a6 13 ac1 e6 14 b3 h5? 15 xe6 xe6 16 a4! c6 17 f4 f6 18 f3.

Black to play White threatens to push either of his center pawns, e.g. 18 … g4? 19 e5! xf3 20 exf6! xd1 21 fxg7 with advantage. Black tried 18 … g4 but was lost after 19 e5! xe3 20 then 20 … c7 21 c5!, e.g. 21 … d5 22 b6 and xd5.

xe3 and

The last of White’s major options is the d5 plan. If Black captures the knight, White has a choice between three recaptures – and often at least two of them are favorable. Supplemental Game # 4 shows how

xd5 can win in a tactical game. Here’s how it works in a more positional way, Bologan – Motylev, Russian Team Championship 2003: 1 e4 c5 2 f3 g6 3 d4 cxd4 4 xd4 c6 5 c4 f6 6 c3 d6 7 c2 g7 8 e2 0-0 9 0-0 d7 10 d2 a5 11 a3 c5 12 ab5 d4 13 xd4 xd4 14 h6 g7 15 xg7 xg7 16 g4 xg4 17 xg4 a4! 18 e2 a5 19 ac1 d7 20 d5 f6 21 fd1.

Black to play The Bind is another pawn structure that keeps some of its favorable qualities even after minor pieces are traded. That’s why White resisted the temptation to force matters (19 e5 a3!) and allowed 21 … xd5. He would have had a slight edge after 22 exd5 fe8 but Black can defend with a timely … e5. For example, 23 h3 e5!? 24 dxe6 xe6 25 f3 b4 followed by … ae8 and … e1+, … e2 or … f6. Instead, after 21 … xd5 White chose 22 xd5 a6 23 h4 because of his kingside chances after 23 … fc8 24 h5. Black liquidated pawns with 24 … e6 25 d4 c5 and then 26 d2! xh5 27 xd6. But the 27 … a5 28 xa5 endgame favored White after 28 … hxa5 29 d7 and he won soon following 29 … 5a7? 30 c5. Even after 29 … b8 30 c7 a6 White has good winning chances (31 d1 c6 32 dd7).

Counter-strategies: … b5 White is the one who benefits most from a stable Bind formation. Black is the one who wants to blow it up, and he can do that with … b5, … f5 or … xc3+. The most attractive idea is … b5, particularly if White’s e-pawn is vulnerable: 1 e4 c5 2 f3 c6 3 d4 cxd4 4 xd4 g6 5 c4 f6 6 c3 xd4 7 xd4 d6 8 e2 g7 9 e3 0-0 10 d2 e6 11 c1 a5.

White to play We’ve seen how White can obtain a sizeable advantage from f2-f4-f5 or e4-e5. Therefore it makes some sense to set the stage with 12 0-0 and then 12 … a6 13 f4. But this is faulty because it allows Black to break the Bind with 13 … b5! 14 cxb5 axb5. This is tactically based on 15 xb5? xd2 16 xd2 xe4 or 15 xb5 xe4! 16 xe4 xb5 with excellent chances. And it is strategically based on the impact of … b4. For example, 15 f5 d7 16 fxg6 looks promising because of 16 … hxg6 17 e5 dxe5 18 xf6 and xd7. But Black has a fine riposte in 17 … b4! (instead of 17 … dxe5). This reveals one of the plot lines of the Bind: If the game becomes tactical early on, Black comes out ahead more often than White.

Suppose, for example that instead of 15 f5, White forces matters and stops … b4 with 15 b4 xb4 16 b1 a3 17 xb5.

Black to play An example of what can happen was Cardoso – Adorjan, Lanzarote 1975 – 17 … fc8 18 d4 xc3! 19 xc3 xe4 (20 xa3 xd4+! or 20 a1 xd4+ 21 xd4 c3 and wins). White won the queen with 20 b8+ xb8 21 xa3 but eventually lost after 21 … xd4+ 22 h1 f2+ 23 xf2 xf2 24 h3 h5 25 f3 b1+ 26 h2 g1+ 27 g3. See also Supplemental Game # 5. Another Bind theme is: White should only open the position in the middlegame. Here this means after pawn-protecting moves such as f2-f3 and b2-b3. Going back to the last diagram, let’s look at 12 f3 fc8 13 b3. If Black continues 13 … a6 White might enter the 14 a4 xd2+15 xd2 endgame, which tries to exploit the new hole at b6. More ambitious is 14 0-0, waiting for the right time to play f3-f4.

Black’s Counter-strategies: … f5 White can often stop or delay … b5. But … f5 is another matter. This is a very double-edged idea popularized by the Russian master Vladimir Simagin in the 1950s. Black’s aim is to obtain play on the f-file and use f5 as a minor piece outpost. An example from an English line: 1 c4 f6 2 c3 c5 3 f3 c6 4 g3 g6 5 g2 g7 6 0-0 0-0 7 d4 cxd4 8 xd4 xd4 9 xd4 d6 10 d3 a6 11 d2 f5 12 e4 e6 13 b3 d7.

White to play White’s light squared bishop at g2 doesn’t watch b5 but he can still slow the queenside push with a2-a4. He played 14 fd1? and Black switched gears with 14 … g4! 15 a4 f5! in Saidy – Reshevsky, Netanya 1969. Black already has f-file threats (16 … xf2 17 xf2? fxe4+). White defended with 16 f1 but after 16 … e5 17 e3 fxe4 18 xe4 h3! he had to weaken his king position (19 fe1 g4). The result was 19 g2 xg2 20 xg2 f3! 21 e4 af8 with a powerful buildup that won quickly (22 ae1 d3! 23 e2 c5 24 c2 d4! 25 d5 e6 26 e3 c6! 27 g1 e4 28 g4 xg3! White resigns). But … f5 creates problems of Black king safety and pawn vulnerability:

1 e4 c5 2 f3 c6 3 d4 cxd4 4 xd4 g6 5 c4 g7 6 e3 h6 7 c3 0-0 8 e2 d6 9 0-0 f5.

White to play Now 10 exf5 poses a choice. After 10 … xf5 11 xf5 xf5, the most obvious reply, White has an edge with 12 d2 followed by 13 f4 and f3. White also gets a slight pull from 10 … xd4 11 xd4 xd4 12 xd4 xf5 13 d2 because of the hole at e6 and the d5-g8 diagonal, e.g. 13 … d7 14 f3 c6 and now 15 d5+ g7 16 fe1. The most unbalancing continuation is 10 … gxf5 after which 11 f4! d7 12 d2 favors White a bit, particularly if he can play 13 h3 and keep the knight off g4 (Even after 12 … g4 13 xg4 fxg4 14 d5 he’s better.). Black’s best answer in this and similar exf5 situations is to compensate for his looseness with piece play. Here that means 10 … xd4! (because 11 xd4 xf5 eliminates White’s two-bishop edge). Chances are mixed after 11 xh6! xf5.

The Bishop Capture on c3 or c6 Anyone who has played games against a computer knows a machine’s fondness for inflicting doubled pawns, even if it means trading a good bishop for a knight. Humans tend to distrust such a swap, particularly by Black in a traditional Dragon formation because of the weaknesses created on dark squares. But in the Maroczy Dragon, … xc3 is stronger: 1 c4 c5 2 c3 g6 3 f3 g7 4 d4 cxd4 5 xc3+!? 7 bxc3 d6 8 e4 f6.

xd4

c6 6

c2

White to play It used to be thought that Black’s goal is to win the c3-pawn, after … a5. Since White should be able to defend that pawn he must stand better, the conventional wisdom said. But in the 1980s it was realized that Black’s real target is at c4 and White doesn’t have enough wood to defend it. Black can attack c4 with … e5, … e6 and … c8. If White protects it with his knight, bishop and queen, Black can increase the pressure with … a4 and a shift of his king knight to b6 or c5. White must find play elsewhere. For example, 9 f3 a5 10 d2 e6 11 e3 c8 and then 12 d5 d7 is a typical scenario. The c4-pawn is doomed after … ce5, e.g. 13 b3 c5 14 b5 a4 or 14 … a3 followed by … d7 and … e5.

White may get sufficient play in tactical lines such as 13 b1 ce5 14 xb7 xc4. Or he can take aim at the kingside with 13 e2 ce5 14 b3 c5 15 b1 (better than 15 b5+ xb5 16 cxb5 f5) 15 … a4 16 0-0 xc4 17 h6 (Bareev – Khuzman, Haifa 2000). But in any case, Black has escaped the Bind stranglehold. Note that if Black delays the capture on c3, White gets an opportunity to avoid the doubled pawns. That occurs regularly with colors reversed: 1 c4 c5 2 c3 f6 3 g3 d5 4 cxd5 xd5 5 g2 c7 6 f3 d3 e5 8 d2.

c6 7

Black to play White’s last move has two ideas. He can attack the center with f2-f4, a reversed form of Simagin’s … f5. Or he can inflict damage on the queenside after 8 … e7?! 9 xc6+! and then 9 … bxc6 10 c4 f6 11 a4 d7 12 a5! with advantage. The best reply to 8 d2 is 8 … d7. Then 9 0-0 e7 10 c4 threatens the e-pawn. Black has to choose between defending it with 10 … f6 or gambiting it with 10 … 0-0. The problem with 10 … f6 is that it strengthens Simagin’s idea, 11 f4!. That’s why some players prefer 10 … 0-0 11 xc6 xc6 12 xe5 e8. As compensation for a pawn, Black gets the two bishops and juicy kingside light squares to exploit. For example, 13 b3 b6 14 e3 h8 15 ad1 f6! 16 f3 f7 and 17 … d5 offers fairly balanced chances.

Hedgehog

The Hedgehog When White gets his pawns to e4 and c4 and Black has a Scheveningen formation, they’ve created the Hedgehog. There is no consensus on how the name came about. But the compact Black setup – which aims to flare out sharply – may have something to do with it. The Hog arises in Sicilian variations in which White is allowed to push his c-pawn, such as in 1 e4 c5 2 f3 e6 3 d4 cxd4 4 xd4 a6 5 c4 or 4 … c6 5 b5 d6 6 c4. In other variations, White cannot set up the Hog because he defends his attacked e-pawn with c3 before he gets a chance to play c2-c4, as in 1 e4 c5 2 f3 d6 3 d4 cxd4 4 xd4 f6! 5 c3. Today the Hog appears more often in English and Reti Opening variations and the first exchange of pawns may not occur until after move seven. For example, 1 c4 c5 2 f3 e6 3 c3 a6 4 e4 c7 5 g3 d6 6 g2 f6 7 0-0 e7 8 b3 b6 9 d4 cxd4 10 xd4. White’s space advantage often allows him to force matters on the kingside, such as with e4-e5, as in Polugayevsky – Petrosian, Kislovodsk 1982: 1 c4 e6 2 f3 f6 3 c3 b6 4 e4 b7 5 d3!? d6 6 c2 c5 7 d4 cxd4 8 xd4 e7 9 0-0 0-0 10 b3 a6 11 b2 c6 12 xc6 xc6 13

d3 g6 14 a4 c7 15 f4 ad8.

White to play The pawn on c4 stops or at least discourages both … d5 and … b5. But as usual White needs to change the structure to create a middlegame target. The position seems to call for 16 f5 and perhaps fxe6/ h3. But Black would be able to meet that with 16 … exf5 17 exf5 d5!, when his d8-rook comes alive and he can try to exploit the c6-g2 diagonal. White prepared instead to push his other pawn: 16 e2 fe8 17 ad1 b7 18 h1. Black should take further precautions with 18 … d7, to stop e4-e5 and continue … f6. But he erred with 18 … c5?, allowing 19 e5!.

Black to play Driving away the knight from f6 yields White a significant advantage, e.g. 19 … h5 20 e4! xe4 21 xe4 c7 22 g4! g7 23 exd6.

Black chose 19 … d7?, not appreciating the strength of 20 e4!. Black would lose material after 20 … c7? 21 exd6 xd6 22 xb7 or 20 … c8 21 exd6 xd6 22 xg6! and e4/ xd6. In the end, he played 20 … c8 21 exd6 f8 but resigned soon after 22 f3 f5 (else e4) 23 b4! xb4 24 d5! threatened the queen and c7 (24 … exd5 25 xd5+ mates). Games like that did nothing to improve the Hedgehog’s reputation. But Black has an ample choice of resources. Three pawn pushes stand out. There is (a) … f5, similar what we saw in the Maroczy Dragon. This is difficult to carry out without accepting serious Black weaknesses. But Supplemental Game # 6 shows how this plan can work. More desirable is (b) … b5. This can open lines for Black’s heavy pieces, after cxb5/… axb5, and may weaken White’s pawns as in Supplemental Game # 7. The most attractive idea is (c) … d5 although this is typically the hardest to execute because White controls d5 with two pawns and usually by at least one knight and one rook. But the push can be effective as a sacrifice, temporary or permanent, if it reveals that White is overextended, as in Yermolinsky – Salov, Wijk aan Zee 1997: 1 f3 f6 2 c4 c5 3 c3 e6 4 g3 b6 5 g2 b7 6 0-0 e7 7 d4 cxd4 8 xd4 0-0 9 d1 c6 10 f4 b8 11 e4 d6 12 b3 e5 13 e3 ed7 14 e2 a6 15 d4 c7 16 d2 fe8 17 ac1 ac8 18 e3 b8 19 f3.

Black to play White seems to have everything in order because 19 … d5! 20 cxd5 exd5 21 exd5 favors him after 21 … xd5 22 xd5 xd5 23 f5!. But he underestimated 21 … a3!, which allows Black to simplify tactically, 22 c2 xc3! 23 xc3 xd5 24 cd3 c5. Black reaches a later middlegame in which his opponent has the weaknesses. For example, 25 c2 xd3 26 xd3 c5 or 26 xd3 xe3 27 xe3 c5. White preferred 25 f4 xd3 26 xd3 but after 26 … c5:

White to play This is what a favorable dissolution of the Bind center looks like. White’s e3 is exposed and the threat of 27 … xe3 forced new concessions. Black won after 27 xd5 xd5 28 f5 e4!. No better was 28 c2 e4 29 c3 xc2.

Supplemental Games # 1 Passivity is punished by a kingside attack: Darga – Yanofsky, Winnipeg 1967 – 1 c4 f6 2 c3 g6 3 e4 d6 4 d4 g7 5 f3 0-0 6 e3 b6 7 d3 bd7 8 ge2 c5 9 0-0 b7 10 d2 e8 11 ad1 c8 12 b3 cxd4 13 xd4 c5 14 b1 e6 15 h1 h5 16 de2! f6 17 fe1 c6 18 a4 c5 19 c2 a5 (Black gives up on … b5 in order to stop b2-b4.) 20 d4 b7 21 d5! fd7 22 h6 e6

White to play 23 f5! f8 (Or 23 … gxf5 24 exf5 xh6 25 xh6 ef8 26 xe7+ xe7 27 xe7 with a winning attack.) 24 xf8 dxf8 25 h6+ g7 26 g4 f6 27 h6+ f7 28 f4 g7 29 h4 h5 30 e5! xd5 31 xd5 dxe5 32 fxe5 g7 33 e6 ad8 34 xh5! gxh5 35 xh5 c6 36 f7+ h8 37 h6 g6 38 xg6 hxg6 39 xg6 xe6!? 40 f7+ Resigns. # 2 Black sets up a bad Bind: Dorfman – Nogueiras, Moscow 1990 – 1 c4 e6 2 f3 d5 3 g3 dxc4 4 a4+ d7 5 g2 c5 6 xc4 gf6 7 d3 d6 (Intending … e5, a reversed-color Bind.) 8 a4 0-0 9 0-0 b6 10 c2 d7 11 e4! (But now his minor pieces are misplaced.) 11 … c6? 12 bd2 e5 13 b3 e8 14 b2 fd7 15 a5 c8 16 c4 c7 17 fd1

Black to play (Threatening 18 b4 cxb4 19 d4! with advantage.) 17 … b5 18 axb6 axb6 19 xa8 xa8 20 h3! f6 21 b4! (With a4 and xd7 or xa8 in mind.) 21 … e7 22 a1 f8 23 bxc5 bxc5 24 a3 c6 25 xc5 e6? 26 xe6+ xe6 27 a2! Resigns (27 … b7 28 d6 wins.). # 3 Exploiting the Marco Hop: Tukmakov – Velimirovic, Odessa 1975 – 1 f3 c5 2 c4 g6 3 d4 cxd4 4 xd4 c6 5 e4 g7 6 e3 f6 7 c3 0-0 8 e2 d6 9 0-0 d7 10 c1 xd4 11 xd4 c6 12 f3 d7! 13 h1 xd4 14 xd4 b6! (Black competes on dark squares.) 15 d2 c5 16 fe1 ad8 17 d5? (A better plan is 17 b1 followed by pushing the b-pawn.) 17 … xd5 18 exd5 a5 19 f1 fe8 20 e3

Black to play 20 … e5! (Black seeks a good N-vs.-bad B endgame.) 21 dxe6 xe6 22 ce1 b4 23 f2 de8 24 xe6 xe6 25 d1 a4! 26 b3 b4! (To weaken White’s pawns with … a4.) 27 g1 a4 28 bxa4 xa4 29

d2 b4 30 b2 c3 31 d2 f6 32 b1 h5 33 e1 xe1 34 xe1 e5! 35 f2 f8 36 g3 h4

White to play 37 gxh4? e6 38 d2 d4 39 g2 f5 40 f2 f4 41 h3 xh4! 42 g3 f6 43 f4 f5 44 e1 d4 45 g2. White resigned before 45 … xf4. # 4 White plays c4-c5 and e4-e5: Korchnoi – Huebner, Leningrad 1973 – 1 c4 f6 2 c3 c5 3 f3 g6 4 e4 g7 5 d4 cxd4 6 xd4 c6 7 c2 d6 8 e2 d7 9 d2 (Avoiding doubled pawns from … xc3.) 9 … c5 10 b4 e6 11 c1 0-0 12 d5

Black to play 12 … ed4 13 xd4 xd4 14 g5 e8 15 0-0 e6 16 e1 xe2+ 17 xe2 d7 18 d2 xd5 19 xd5!? e6 20 d3 ac8 21 e3 a6 22 h3 f8 23 g4! (To stop … f5.) 23 … f6 24 g5 b2 25

a3 c7 26 c5! fc8 27 g2! f8

White to play 28 cxd6! (Based on 28 … xc1 29 d7! d8 30 xc1 xc1 31 c5 f4 32 c8 xd7 33 xd7 and wins.) 28 … exd6 29 xc7 xc7 30 e5! c2 31 d2! dxe5 32 d8 (Threat of d6.) 32 … g7 33 e3! e7 34 h6+ f6 35 h4+ e6 36 e8 xd2 37 xe7+ d5 38 d8+ Resigns (38 … c4 39 c5+ b3 40 e3+). # 5 An ideal version of … b5 via tactics: Bobekov – Joppen, Amsterdam 1954 – 1 d4 f6 2 c4 g6 3 c3 g7 4 e4 d6 5 f3 0-0 6 e3 bd7 7 d2 c5 8 ge2 a6 9 g3?! cxd4 10 xd4 e5! (After 11 xe5 dxe5 Black might occupy d4 with a knight.) 11 e2 a5 12 c1 e6 13 b3 fc8 14 0-0

Black to play 14 … b5! (Based on 15 cxb5 axb5 16 xb5 xc3 or 15 c5

c6!) 15

f4 c6! 16 e3 bxc4 17 f5 d7 18 xc4 e5 19 d5? xd2 20 xd2 xc4 21 xe7+? (A miscalculation but Black is better after 21 bxc4 xd5.) 21 … f8 22 xc8 xd2 23 fd1 xc8! 24 xd2 h6 25 xc8+ xc8 26 xd6 g4! White resigns. # 6 The … f5 plan in the Hedgehog: Dely – Suetin, Kecskemet 1972 – 1 e4 c5 2 f3 c6 3 d4 cxd4 4 xd4 e6 5 b5 d6 6 c4 f6 7 5c3 e7 8 e2 0-0 9 0-0 b6 10 b3 b7 11 b2 d7! 12 a3 c5 13 d2 f6 14 fe1 e5 (Black wants to provoke f2-f4 and prepare … f5.) 15 f1

Black to play 15 … f5! 16 exf5 xf5 17 d1 h4 18 g3 f6 19 g2 f8 20 b1 xb2 21 xb2 e5! 22 e3 xg2 23 xg2 g4 24 e1 h5! and White resigned in view of 25 h3 f3+ 26 g1 xh3 or 25 h4 f3+ 26 g1 d3. # 7 The … b5 plan in the Hedeghog: Larsen – Quinteros, Mar del Plata 1982 – 1 c4 f6 2 c3 c5 3 g3 e6 4 f3 b6 5 g2 b7 6 0-0 e7 7 e1 0-0 8 e4 d6 9 d4 cxd4 10 xd4 c7 11 b3 a6 12 b2 bd7 13 f4 fe8 14 f5? e5 15 c2

Black to play 15 … b5! (Based on 16 cxb5 axb5 17 xb5? b6+.) 16 e3 b4! 17 cd5 xd5 18 xd5 xd5 19 cxd5?! g5! (Black can rule the c-file.) 20 h4 h6 21 f1! ec8 22 c4 a5! 23 g2 a4 24 e2 a5 25 e1 b6 26 f3 ac5 27 c1 xc1 28 xc1 axb3 29 axb3

Black to play 29 … a5! 30 g4 ca8 31 g5 c5 32 h5 a3 33 e3 a2 34 g3 b2 35 f6 aa2 36 e2 c2 37 g6 fxg6 38 hxg6 h6 39 f7+ f8 40 xc2 xc2 41 g4 d8 42 g3 c3! 43 xc3 bxc3 44 d1 d3 45 f3 g5+ 45 h2 d2+ 47 g3 c2 48 xc2 xc2 49 b4 f4 White resigns.

D. The Boleslavsky Hole The arresting idea that … e5 generates enough counterplay to compensate Black for a backward d-pawn may seem fairly new. Actually, it’s one of the oldest in the Sicilian Defense. Louis Charles Mahe de LaBourdonnais tried it in his marathon match with Alexander MacDonnell – who immediately erred by erasing the d5-hole. See Supplemental Game # 1. The modern handling of … e5 can be credited to Isaac Boleslavsky. He showed how Black gets more space than in the Scheveningen and keeps White pieces off d4 and f4. And by itself, the hole at d5 isn’t enough to favor White significantly. (We call this the Boleslavsky Hole to distinguish it from the Boleslavsky Wall of Chapter Six.) The chief variations that feature this formation are Boleslavsky’s own 1 e4 c5 2 f3 d6 3 d4 cxd4 4 xd4 f6 5 c3 c6 6 e2 e5,

The Boleslavsky Hole on d5 the Najdorf Variation, 1 e4 c5 2 f3 d6 3 d4 cxd4 4 xd4 f6 5 c3 a6 and … e5, and the Sveshnikov, 1 e4 c5 2 f3 c6 3 d4 cxd4 4 xd4 f6 5 c3 e5 and … d6. An early Najdorf was Novotelnov – Petrosian, Moscow 1951: 1 e4 c5 2 f3 d6 3 d4 cxd4 4 xd4 f6 5 c3 a6 6 e2 e5 7 b3 e7 8 0-0 0-0 9 f4 bd7 10 e1? b5 11 a3 b7 12 f3 c8 13 h1

e8 14 f2 f8 15 d2.

Black to play White failed to restrain Black earlier with a2-a4. The table is set for 15 … d5!, based on 16 exd5 e4! and then 17 … xd5. Black would also be better after 16 fxe5 dxe4!. White chose 16 xd5 and was lost after 16 … xc2 17 fxe5 xd5 18 exd5 xe5 19 d4? d3!. See Supplemental Game # 2 for another … d5!.

The Battle for d5 The best way for White to exploit the Boleslavsky Hole is to occupy it with a knight. But since Black has at least two minor pieces that can easily eye d5, White needs help, as in Predojevic – Fogarasi, Budapest 2004: 1 e4 c5 2 f3 e6 3 c3 a6 4 d4 cxd4 5 xd4 c7 6 d3 f6 7 00 c5 8 b3 a7 9 h1 d6 10 f4 bd7 11 d2 0-0 12 e2 e8 13 ae1.

Black to play White threatens a kingside initiative with 14 e5!. Black stopped that with 13 … e5, creating a delayed Boleslavsky Hole. A tempting reply is 14 f5, because it prepares f3-g3 or -h3 or a pawn storm, g2-g4-g5. But f4-f5 works best when … d5 is ruled out. Here 14 … b5 15 a3 (to stop … b4) 15 … b7 prepares Black for pushing his dpawn. White chose 14 g4! instead. Black cannot stop a favorable 15 g5 (and 14 … h6 15 g5 hxg5 16 fxg5 h7 would be suicidal after 17 c4!). Instead, he tried to shift pieces to the kingside but 14 … f8? 15 g5! 6d7 16 d5 began a positional rout, 16 … d8 17 f5! b5 18 a5!. White won following 18 … b8 19 c4 b6 20 xb6 xb6 21 b4.

When White has a choice of how to retake on d5, a piece is generally better than a pawn. Let’s consider: 1 e4 c5 2 f3 d6 3 d4 cxd4 4 xd4 f6 5 c3 a6 6 e2 e5 7 b3 e7 8 0-0 c7.

White to play Here 9 d5?! xd5 is premature, since the desirable 10 xd5 costs a pawn (10 … xc2). Of course, is 10 exd5 playable. But unlike the Marco Hop case in the Dragon, Black’s e-pawn is strong on e5, not backward on e7. In such formations, Black usually stands well after … f5. Therefore White should try to win control of d5 first, with 9 a4, which stops … b5-b4. If Black tries to occupy the weakened b4 with 9 … c6, he makes 10 d5! xd5 11 exd5! favorable. Black’s natural continuation is 9 … e6. But suppose he chooses 9 … b6?!. White can win the fight for d5 with 10 g5! bd7 and then 11 d2! b7 12 c4 0-0 13 e3!. After the impending xf6, three White minor pieces can aim at d5 compared with two for Black. The upshot is 13 … fc8 14 c4! c5 15 xf6! xf6 16 cd5.

Black to play This kind of position was considered virtually lost in the 1960s. But now similar positions are regarded as just inferior. After 16 … xd5 17 xd5! a7 White must try to improve his position and Bogdanovic – Buljovcic, Kraljevo 1967 showed how: 18 c3 g6 19 g3 f8 20 f3 g7 21 fd1 b5 22 axb5 axb5 23 xa7 xa7 24 e2 b8 25 c2! xd5 26 xd5 d7 27 d3 b6 28 g2 f5 29 f3 with a winning plan of b2-b4 and a3xb5.

The exd5 Option Let’s examine the exceptions in which exd5 is preferable to retaking on d5 with a piece. In the Sveshnikov Variation, 1 e4 c5 2 f3 c6 3 d4 cxd4 4 xd4 f6 5 c3 e5 6 db5 d6 there is a positional approach, 7 d5 xd5 8 exd5, with queenside chances. This is plainly superior to 8 xd5?, which hands Black an initiative, 8 … a6 9 c3 d4 and a later … e6. One favorable scenario arises when White can dominate light squares: 1 e4 c5 2 f3 d6 3 d4 cxd4 4 xd4 f6 5 c3 a6 6 e3 c6 7 e2 e5 8 b3 e6?! 9 0-0 e7 10 d5! xd5? 11 exd5.

Black to play White would have the upper hand after 11 … a5 12 c4 xb3 13 xb3, regardless of whether he plays for c4-c5 or on the kingside with f2f4. Black’s best policy is the quiet 11 … b8 12 a4 bd7 so he can use this knight to watch c5 and e5. He is only slightly worse after 13 a5 0-0 14 c4. However, Tseshkovsky-Staniszewski, Lubniewice 1995 saw 12 … a5?, severely compromising his already-weak light squares. Then came 13

b5+! bd7 14 d2! 0-0 (or 14 … xd5 15 c4 xe3 16 fxe3 and 17 xd6+ or 16 … 0-0 17 xd7 xd7 18 b6). White had a big positional plus after 15 c4 c7 16 c3 e4 17 b4! axb4 18 cxb4 because of queenside pressure and his ability to create a passed pawn. He won after 18 … e5 19 b6 ad8 20 c1 b8 21 d4 g6 22 c2 e8 23 xe4. See also Supplemental Game # 3. But if Black can play … f5, the middlegame becomes a battle between White’s queenside expansion and Black’s kingside and center play. White wins the battle in games like Anand – Topalov, Monaco 2003: 1 e4 c5 f3 c6 3 d4 cxd4 4 xd4 f6 5 c3 e5 6 db5 d6 7 d5 xd5 8 exd5 b8 9 c4 e7 10 d3 0-0 11 0-0 d7 12 a4 f5.

White to play White struck first with 13 c5!, based on 13 … dxc5 14 d6 and 15 dxe7 or 15 c7. It wouldn’t have helped Black much to insert 13 … xb5 14 axb5 because 14 … dxc5 15 d6 xd6 16 c4+ h8 17 d5 or 15 … xd6 16 c4+ h8 17 xd6 and d5/ xb7 still favors White. So, Black chose 13 … xb5 14 axb5 e4 after which 15 c4 d7 16 c6 would have been bad for him. On the other hand, there are examples like Pilnik – Geller, Goteborg 1955: 1 e4 c5 2 f3 c6 3 d4 cxd4 4 xd4 f6 5 c3 d6 6 e2 e5 7

b3 e7 8 0-0 0-0 9 e3 e6 10 f3 a5 11 13 c4.

d5 xd5 12 exd5

b8

Black to play White wants to push to c5, with the support of a2-a3 and b2-b4 if necessary, and create a passed pawn. But Black stopped him with a darksquare blockade, 13 … a6! 14 d2 b6 15 c3 c5! 16 xc5 bxc5. After White repositioned his bishop, 17 e1 d7 18 d1 a4! 19 c2, Black’s 19 … f5! forced him to choose between confronting Black’s center with f2-f4, restraining it with f2-f3 or undermining it with g2g4!?. For example, 20 f4 and then 20 … e4 21 g4 collapses the support for e4. But this is risky in view of 21 … f6 22 gxf5 e8. Instead White chose 20 d1 g6 21 e2 f6 22 f3.

Black to play White is ready for 23 g4!? since a trade of f-pawns would benefit his

c2-bishop. But Black acted first, 22 … e4!! and 23 xf6 f4!.

xf6 24 fxe4

He secured the e5-outpost and kept White’s bishop bottled up, while he prepared to push his g-pawn. The rest: 25 f2 e5 26 df1 h4 27 d1 f7 28 c2 g5! 29 c3 af8 30 h3 h5 31 e2 g4! 32 xf4 xf4 33 xf4 xf4 34 g3 f3+ 35 f2 xh3 36 gxf4 g3+! 37 xf3 g2+ 38 f2 h2 White resigns. This exd5 formation gained new life in the 21st century in supersharp positions arising out of English Attack such as 1 e4 c5 2 f3 d6 3 d4 cxd4 4 xd4 f6 5 c3 a6 6 e3 e5 7 b3 e6 8 f3 bd7 9 d2 b5 10 0-0-0 e7. White can profit from d5/exd5 if he can also play g2-g4-g5. That’s because Black cannot maintain a pawn on f5. For example, 11 g4! b6 12 f2 fd7 13 b1 c7 14 d5! xd5 15 exd5.

Black to play Once again White was able to time d5 so that … xd5? would allow a piece-winning exd5. The exchange on d5 means he has access to several nice light squares including c6, e4 and f5. He also has choices: After g4-g5 and h2-h4 he will decide whether his bishop belongs on the h3-c8 or d3-h7 diagonal. Then he can choose between playing positionally ( a5-c6 or d2-e4) or just going for mate (h4-h5 and g5g6). Black’s counterplay comes from …

c4 and/or pushing queenside

pawns. But White can defend his king much more easily than Black, e.g. 15 … c4 16 c1 db6 17 f4!. In a typical game, Korneev – Draoui, Cergy Pontoise 2004 Black tried 15 … b8 16 g5 c4, hoping for 17 xc4 bxc4! or 17 c1 a5 and … a4 with a counterattack directed at b2. But White assured himself of an edge with 17 h4 xe3 18 xe3 d8 19 d3 b6 20 e4!. Preparing to castle with 20 … g6 is bad after 21 f4 0-0 22 f5. So is 20 … c5 21 xc5 xc5 22 f4. Black tried 20 … a5 21 he1 a4 22 d4 g6 but was soon lost, 23 c6 b7 24 f4 0-0 25 f5 b8 26 h5! xc6 27 dxc6 c7 28 fxg6, and resigned after 28 … fxg6 29 hxg6 xc6 30 g7!. See also Supplemental Game # 4.

The f2-f4 Problem If White pushes his f-pawn to f4 and f5 he gets a formation that was somewhat favorable in the Scheveningen. Black can avert that with … exf4. Then White will be better if he makes more out of the outposts at d4, f4 and f5 than Black can of e5. For example, in Geller – Polugayevsky, Moscow 1961, Black traded on f4 because White threatened 12 f5 c4 13 g4! followed by a strong 14 g5: 1 e4 c5 2 f3 d6 3 d4 cxd4 4 xd4 f6 5 c3 a6 6 e2 e5 7 b3 e7 8 0-0 0-0 9 e3 c7 10 d2 e6 11 f4 exf4 12 xf4 bd7.

White to play Black’s last move was an error because his knight belongs on c6 so he can take on d4 when White repositions his knight. The difference was 13 d4! e5 14 f5!. Then trades like 14 … xf5 15 xf5 c4 16 xc4 xc4 17 d4 would just benefit White’s remaining pieces. Instead he was allowed to concentrate his forces in the center and on the kingside, 14 … ac8 15 af1 fe8 16 d4! f8 17 d3 fd7 18 h4! g6 19 h3 de5 20 d1! b5 21 h5 h6. He threatened sacrifices on h6 and g7 after 22 g3 h7 23 xd5 24 exd5, e.g. 25 xh6 gxh6 26 xe5 dxe5 27 xf7+.

d5

The game continued 24 … xd3 25 cxd3 b7 26 h1 c2 27 e3 d2 28 g4 d7 29 h4 e7 and White finished off with the neat 30 e3! d7 31 xh6+! g8 32 xe8 Resigns (31 … gxh6 32 f6+ costs a rook).

Supplemental Games # 1 The anti-positional xc6: MacDonnell – LaBourdonnais, match 1834 – 1 e4 c5 2 f3 c6 3 d4 cxd4 4 xd4 e5 5 xc6? (This allows Black to regain control of d5. Better was 5 b5!.) 5 … bxc6 6 c4 f6 7 g5 e7 8 e2 d5 9 xf6 xf6 10 b3 0-0 11 0-0 a5 12 exd5 cxd5 13 d1 d4 (Not 13 … e6? 14 xd5 xd5 15 c3.) 14 c4? b6 15 c2 b7 16 d2 ae8 17 e4 d8 18 c5 c6 19 f3 e7 20 ac1 f5!

White to play 21 c4+ h8 22 a4 h6! 23 xe8 fxe4 24 c6 exf3 25 c2 (Or 25 cxb7 e3+ 26 h1 fxg2+ and … f2+ wins.) 25 … e3+26 h1 c8 27 d7 f2 28 f1 d3 29 c3 xd7 30 cxd7 e4! 31 c8 d8 32 c4 e1 33 c1 d2! 34 c5 g8 35 d1 e3 36 c3

Black to play 36 … xd1! 37 xd1 e2! White resigns. # 2 White’s slow play allows …d5!: Unzicker – Bronstein, Goteborg 1955 – 1 e4 c5 2 f3 d6 3 d4 cxd4 4 xd4 f6 5 c3 a6 6 e2 e5 7 b3 e7 8 0-0 0-0 9 e3 c7 10 a4 b6 11 d2 e6 12 fd1 c8 13 e1? (He should fight for d5 with c1-a2-b4!.) 13 … b7! 14 d2 bd7 (Not 14 … xe4 15 xe4 xe4 16 f3. The … xc3 sacrifice would have been good but the riskless … d5 can’t be stopped now.) 15 f3

Black to play 15 … d5! 16 exd5 xd5 17 xd5 xd5 18 ad1 f6 19 c1 e4! 20 f2 c5 21 xc5 bxc5 22 e3 e8 23 f4 c4 24 b3 ac8 25 h3 e6 26 h2 c7 27 d6 a5 28 bxc4 xc4 29 b3 xe2 30 xe2 e3! 31 6d4 e4 32 f3 g5 33 g4 e6 34 e4

Black to play 34 … h5! 35 f3 g5 36 xe8+ xe8 37 g3 xc2 38 d5 e2!? (Simpler was 38 … xb3 39 xg5 b2.) 39 xg5 e1( ) 40 xg7+ h8 41 g5 xg2+! 42 xg2 e2 White resigns. # 3 Model passed pawn creation: Korneev – Kovalev, Hamburg 1993 – 1 e4 c5 2 f3 d6 3 d4 cxd4 4 xd4 f6 5 c3 a6 6 e3 c6 7 f3 xd4 8 xd4 e5 9 d2 e6 10 0-0-0 c8 11 b1 a5 12 d5!? xd2 13 xd2 xd5 14 exd5 d7 (Or 14 … f5 15 d3! xd3 16 xd3 and c1/c2-c4.) 15 d3 e7 16 c4 h6 17 c2 f5 18 e1 h4 19 g3 e7 20 b4! 0-0 21 c5 a4 22 cc1

Black to play (White prepares cxd6 and c5.) 22 … b5 23 xb5 axb5 24 b2 fd8 25 ed1 f7 26 b3 d7 27 c6! bxc6 28 xc6 b8 29 dc1 g5 30 c7 b7 31 xd7+ xd7 32 xg5 hxg5 33 a4 bxa4+34 xa4 a7+ 35 b5 a3

White to play 36 c6! e7 37 b5 xf3 38 b6 b3 39 b7 e4 40 c7 e3 41 b8( ) xb8 42 xb8 f6 43 e1 f4 44 h4! gxh4 45 gxf4 f5 46 c7 xf4 47 xd6 f3 46 e5 f2 49 a1 Resigns. # 4 Kingside attack after exd5: Glek – Stefansson, Linares 1996 – 1 e4 c5 2 f3 d6 3 d4 cxd4 4 xd4 f6 5 c3 c6 6 e2 e5 7 f3 h6 8 0-0 e7 9 e1 0-0 10 h3 a6 11 f1 c7 12 h2! (To trade a knight for a Black defender of d5.) 12 … b5 13 g4 xg4 14 hxg4 b4 15 d5 xd5 16 exd5 d4 (No better is 16 … d8 or 16 … b8 in view of 17 a3! bxa3 18 xa3 with a queenside edge.) 17 d3 (Black cannot stop c2-c3, followed by getting the queen to e4 or f5.) 17 … c5 18 e3 f6

White to play 19 c1! a5 20 c3 bxc3 21 bxc3 b5 22 f3! d8 23 e4 e8 24 c4 c7 (24 … d4 25 xd4 exd4 26 h7+ and mates.) 25 h7+ f8 26 c5! xd5 27 cxd6 g5 (27 … xd6 28 c5) 28 c4 Resigns.

Chapter Four: Chain Reactions

The d5 chain Pawns are born free but are everywhere in chains. They form the central spine of French, King’s Indian and Benoni Defenses and numerous other openings. Chains divide the board and limit mobility. Whether they are long (White pawns at c3, d4, e5, f6 vs. Black pawns at c4, d5, e6, f7) or small (d4, e5 vs. d5, e6) they force us to focus on the wings. If one player can attack on ‘his’ wing while his opponent cannot on ‘his,’ the game usually becomes one-sided. This is kindergarten strategy for humans. But not for computers. Garry Kasparov constructed an enormous chain of White pawns from f2 to b6 against X3D Fritz in their 2003 match. Spectators in the playing hall began to laugh at the machine’s pointless maneuvers. Why, they wondered, wasn’t it attacking the base of White’s pawn chain? Kasparov methodically attacked the base of Black’s chain, created a passed b-pawn and won. Chains may require odd-looking strategies, as in Rivera – Suttles, San Juan 1965:

1 d4 g6 2 e4 g7 3 c4 d6 4 c3 c6 5 e3 e5 6 d5 ce7 7 f3 f5 8 c5! f6 9 b5+ f7!.

White to play Both players attacked the base of the enemy chain – White with c4-c5 and Black with … f5. Black’s last move avoids an exchange (9 … d7?), which would have rendered e6 highly vulnerable to enemy pieces. White continued 10 h3?. This stops the annoying 10 … fxe4 11 fxe4 g4 and prepares to fight for kingside space with g2-g4. But it’s a grave error because of 10 … f4! and 11 f2 g5. To use Aron Nimzovich’s phrase, Black transferred his attack from the base at e4 to a new base at f3. On 12 g4, Black would take en passant and bombard a terminally weak pawn at f3. The game proceeded with 12 ge2 h5 13 cxd6 cxd6 14 b3 g4 15 0-0-0 g3!. White had to avoid 16 g1?, which would entomb his king rook. He chose 16 e1 after which 16 … g6 prepared … h4!.

White to play To evaluate chains, compare wings: White correctly played c4-c5xd6 but can’t attack the d6-pawn or exploit the open c-file. Black correctly decided against … gxf3 in favor of more successful pressure on g2. If White defends it with g1 Black has a winning plan of sacrificing his knight on g2, followed by … xh3 and pushing the h- and g-pawns (For a queenside example of this plan see Supplemental Game # 1.). White covered g2 with 17 g1!? h4 18 f1. But Black seized the initiative on the other wing, 18 … a6 19 b1 b5 20 ge2 d7! 21 c1 c5 22 c2 d7. He soon ruled much of the board, 23 b4 a4 24 b3 xc3+ 25 xc3 b6 26 a5 hc8 27 d2 a7 28 b2 f6 29 c1 xc1+ 30 xc1 d8!. White’s king tried to reach safety at f1 after d3 and e1-e2. But it was futile: 31 c2 c7+ 32 d1 c8 33 b3 a7 34 d3 b6 35 e1 f2! 36 f1 xe1 37 xe1 g1 38 e2 xh3! White resigns. Appropriately, the farthest pawn in Black’s chain won the game (39 gxh3 g2!). The longer the chain, the more power exerted by advanced pawns. They can be as valuable as minor pieces, as in Anand – Shirov, Leon 2002:

White to play Black hoped to show that White is overextended on the queenside by attacking the a6-pawn with … b8 or … b8-b6 (after … e8-d7). But White began to exploit the chain with 25 g4! and f4-f5. That leaves Black a choice between restraint, blockade or neglect. The restraining 25 … g6 fails tactically (26 f5! exf5 27 gxf5 gxf5 28 f4 and 29 e6!, threatening xc7). But the blocking 25 … f5! would have been unclear after 26 exf6 gxf6 (27 h4 and possibly 28 g5). However, Black chose neglect, 25 … e8? 26 f5! d7 27 f4 g5 (based on 28 xg5 xg5 29 xg5 xd4+) and then 28 e3 h6. But White lengthened the chain with 29 f6! and 29 … f8 30 d3.

Black to play White has a choice of winning tries. The first is to attack the kingside with h2-h4. The pawn at h6 cannot be defended for long after … gxh4.

The second plan is to sacrifice a piece on g5 for two pawns, e.g. 30 … b8, e.g. 31 h4! gxh4 32 xh4 or 31 xg5! hxg5 32 xg5 and h2-h4h5 etc. Black helped his opponent choose by playing 30 … a5 and then 31 xc7+ xc7 32 xc7+ xc7. His threat to the b3-pawn made the h2h4 plan too slow (33 d2 b6). But after 33 xg5! hxg5 34 xg5 Black can’t stop the h-pawn – 34 … xb3 35 h4 a1 36 c1 b3 37 e3 a5 38 g5 c4 39 c1 Black resigned.

The d5 Chain Every experienced player knows that a pawn chain with White links on d5 and e4 has a distinctly different personality from one with links at e5 and d4. White usually plays on the queenside in a d5 chain and his modest goal is to win a pawn there. But the e5 chain, considered in the next chapter, almost forces White to attack the king. The d5 chain comes about most often in 1 d4 openings. But it also arises, somewhat surprisingly, in 1 e4 e5 lines. For example, in the Ruy Lopez, 1 e4 e5 2 f3 c6 3 b5 a6 4 a4 d6 5 0-0 d7 he may go in for 6 c4 g6 7 d4 g7 8 e3 ge7 9 d5! b8.

White to play White should attack the base of the enemy chain, 10 c5!. He is better after 10 … 0-0 11 c3 and could meet 11 … f5? with 12 xd7 and g5-e6. Black can spend a tempo on 11 … h6 to stop g5. That gives White time for 12 d2 so he can meet … f5 with f2-f3, reinforcing e4 and avoiding the bishop-trapping … f4. If Black defends d6 with 12 … c8, play may go 13 b4 xa4 14 xa4 f5 15 f3 d7. Then White can attack a new base with 16 c6! bxc6 17 dxc6. In Keres – Shiyanovsky, Baku 1961 he took his time with 16 ac1

and Black prepared to attack a new base after 16 … f4 with the intention of … g5-g4. White’s advantage became apparent after 17 f2 h7?! 18 c6!. Now 18 … bxc6 19 dxc6 db6 20 b3 and a2-a4-a5 would be a big edge. Instead, Black kept matters closed with 18 … b5 and 19 a5 db6.

White to play But this fails to 20 a4!. One of the queenside pawns is lost. After 20 … bxa4 21 xa4 xa4 22 xa4 White’s simplest winning plan is to double or triple heavy pieces against a6. The chain prevents Black’s kingside pieces from coming to the defense of his collapsing queenside. But Black could have defended better – and attacked better – such as with 17 … f6, taking the sting out of 18 c6. Let’s examine a similar variation. Anand – Carlsen, Monaco 2011 went: 1 e4 e5 2 f3 c6 3 b5 f6 4 e2 a6 5 a4 e7 6 c3 d6 7 0-0 0-0 8 d4 d7 9 d5 b8 10 xd7 bxd7 11 c4.

Black to play The good news for Black is he has more control of c5 this time. The bad news is that it takes more effort to achieve … f5. He chose 11 … g6 12 c3 h5 and 13 b1 g7. In similar King’s Indian Defense situations, White promotes c4-c5 with e1-d3. But here 14 e1 allows 14 … g5!, when Black rids himself of his bad bishop just as White did with 10 xd7. Instead, White looked for an opportunity to break through on the queenside, 14 b4 f5 15 e3 b8 16 fd1 e8 17 a4 f4! 18 d2 g5.

White to play White cannot stop … g4 and that counts as a minor victory for Black, e.g. if 19 e1 g6 20 f3 then 20 … h5 21 d3 g4 with even chances. Bear in mind: Black does not have to mate to win. But it helps if he can open the kingside before White opens the queenside. That’s what

happened here, 19 h3 h5 20 h2 g6 21 f3 e8!. Note how Black needs one knight at f6 to promote … g4 and the other at d7 to stop c4c5. Then came 22 dc1 ef6 23 f1 – better is 23 h1 g4 24 g1 to safeguard the g-file – 23 … f7 24 e1 g4! 25 d1 gxh3 26 gxh3 g8 27 c2 g2 and Black won the endgame by penetrating the kingside. The basic themes of the d5-chain can be so simple – trade off your pawn-bound bishop and attack the enemy base – that the real battle begins when the players try to foil their opponent’s plans. In the last two examples we saw how Black’s failure to stop c4-c5-c6 – and then White’s failure to counter … f5 in the later example proved decisive.

White Counter-strategies: exf5 and g2-g4 For White, the choices begin with trading off his base pawn, that is meeting … f5 with exf5. This typically leads to one of three reactions: (a) Black recaptures on f5 with a piece, allowing White to occupy e4 with his own pieces. (b) Black recaptures on f5 with a pawn. “As every Russian schoolboy knows,” Mikhail Botvinnik used to say, “one always captures with a pawn in such positions.” But … gxf5 may make f5 a target that can be attacked by d3, c2, g3 and/or g2-g4. (c) Black retakes on f5 with a pawn and White tries to change the center further with f2-f4. The first option is tempting because Black’s pawns remain solid and he can occupy f5. If he can play … f5 and … d4 – while White is playing his own knight to e4 – Black may come out ahead. But there are positional risks, as in Toth – Gasztonyi, Budapest Championship 1966: 1 d4 c5 2 d5 e5 3 c4 d6 4 e4 f5? 5 exf5 xf5 6 e2 f6 7 bc3 e7 8 g3 g6.

White to play Black’s temporary control of the light squares e4 and f5 disappeared

after 9 d3! and 9 … xd3 10 xd3. Moreover, after 10 … 0-0 11 0-0 bd7 12 ce4 he had to stop g5-e6. White made further progress after 12 … e8 13 d2 h4 14 ae1 a6 15 f4! e7 16 g5!, threatening mate. Since 16 … ef6 17 f5 would lose outright and Black didn’t like the looks of 16 … g6 17 e6 and f4-f5, he played 16 … xg5 17 fxg5 g6. But then the dark squares, principally f6, could be exploited, and the game ended with 18 xf8+ xf8 19 e4 d7 20 f1 b5 21 b3 b4 22 h4 g7 23 f6+ h8 24 h3! f5 25 h5 f8 (or 25 … xf6 26 gxf6 and 27 g4) 26 hxg6 d4 27 xh7! xh7 28 f7 Resigns. When Black retakes on f5 with a pawn he may find himself with greater control of the center than White. He temporarily owns d4, e4 and f4. But if he advances either his e- or f-pawn, both become weaker: 1 e4 d6 2 d4 f6 3 f3 e5 4 d5 c6 5 c4 exf5! gxf5 9 d3.

h5 6 e3 g6 7

c3 f5 8

Black to play This position, from Psakhis – Steffens, Ostende 1993, illustrates Black’s difficulties. How will he defend his f-pawn after c2 ? If he pushes it to f4, he hands e4 to White’s pieces and opens the d3-h7 diagonal. He could try the pawn sacrifice of … e4 and then fxe4/… f4 that we saw in the discussion of the Boleslavsky Hole. But here Black doesn’t

have the pieces to exact enough compensation. Black could also defend the f-pawn with his queen, perhaps at f6, or his knight, clumsily at g7. But both ideas enable White to seize a substantial positional edge with ge2, 0-0-0 and a well-timed g2-g4!. In the game, Black played 9 … g8 10 g3 f6? and had to meet 11 c2! with 11 … d7. White could have increased his superiority in various ways and chose 12 0-0-0 a6 13 a3 c7 14 b3 f7 15 ge2. Black has no safe home for his king and opening the position would backfire (15 … cxd5? 16 cxd5 fxd5? 17 c4). Meanwhile, White has two attractive plans, g3-g4 and f3-f4. The first may allow him to win control of light squares. The second seeks a trade for the e5-pawn that would sign the f5-pawn’s death warrant. But what happens if Black meets f3-f4 with … e4 ? Then Black’s pawns are vulnerable to a now-stronger g3-g4. This strategy is deceptive because it seems as if Black’s game is improving: 15 … h5 16 df1 h6 17 f4! e4 18 b1 g4 19 d2 a6 20 h3 c5 21 d1 f6.

White to play Despite the retreats, White gets to carry out his strategic stroke, 22 g4! – and Black resigned. He would lose a piece after 22 … g6 23 g5 or 22 … fxg4 23 hxg4 xg4 24 xh6. But even if this tactic wasn’t possible – because Black’s queen or king bishop were on other squares – White would achieve a substantial edge with simple moves such as e3, b2-b4, d2 and d4 or eventually g3-g4.

The … e4 advance is so committal after f2-f4, that anyone who finds themselves dealing regularly with a d5 chain, such as a King’s Indian Defense player, has to weigh alternatives. That means meeting f2-f4 with … exf4!? or allowing an exchange on e5. For example: 1 d4 f6 2 c4 g6 3 c3 g7 4 e4 d6 5 f3 0-0 6 e2 e5 7 0-0 bd7 8 d5 c5 9 c2 a5 10 e1 fd7 11 e3 f5 12 exf5 gxf5 13 f4.

Black to play Nowadays White is more likely to meet 11 … f5 in the traditional way, 12 f3. But when the King’s Indian was fairly young, the position in the diagram was popular. White’s ideas include 14 fxe5 dxe5? 15 xf5. The natural 13 … e4? is wrong because White can use the fine d4 and e3 squares while Black had little counterplay. For example, 14 d2! f6 15 c2 e8 16 b5 f7 17 d4! e8 18 e3. In Flohr – Suetin, Soviet Championship 1950, White combined a kingside attack ( ac1-c3-g3) with the positional threat of g2-g4 and won after 18 … h8 19 ac1 d7 20 c3 d8 21 h1 a6 22 c2 xd4 23 bxd4 f6 24 g3 g8 25 xg8+ xg8 26 e3!. Either the f-pawn or a-pawn falls. The lesson to be learned from the previous diagram was that Black needs counterplay and that led to 13 … exf4! 14 xf4.

Black to play Black gets nice centralized piece play after 14 … e5 15 f3 d7 16 ae1 f6. He can also try the more forcing 14 … e4 15 xe4 fxe4 in view of 16 xe4 xb2 17 b1 c5 or 16 g3 f6 17 b1 d4+ and … c5. There is another way for White to deal with the attack on his e4-base. He can answer … f5 with f2-f3 and g2-g4!?. This seems to force a wild, if not suicidal, opening of the kingside. In fact, it can mean keeping control of or even closing kingside lines, as in Supplemental Game # 3 and: 1 d4 f6 2 c4 g6 3 c3 g7 4 e4 d6 5 f3 0-0 6 e2 e5 7 0-0 c6 8 d5 e7 9 e1 d7 10 f3 f5 11 g4.

Black to play On 11 … f4, White replies 12 h4! and can keep matters closed by answering … h5 with g4-g5!.

The same goes for the immediate 11 … h5. It is best met by 12 g5! – not by 12 gxh5? f4!, when Black opens the kingside and enjoys more space there at the cost of a pawn. But what about … fxg4? Doesn’t that benefit Black? No, he is not ready to control the file. For example, 11 … h8 12 d3 g8 13 h1 and now 13 … fxg4 14 fxg4 h6 15 xh6 xf1+ 16 xf1 xh6. Black is rid of his bad bishop. But White had the better of the open lines, 17 f2 g5 18 f1 a6 19 c5! (Wang Ho – Ding Liren, Jin Zhou 2009). Recent experience indicates Black’s best policy after 11 g4 is to seek play on the other wing, with … c6 followed perhaps by … b5!?. And that leads us to consider Black’s basic choices in the d5 chain.

Black Counter-strategies: … c5 and … c6 Black can mechanically stop the attack on the base of the chain at d6 by planting his own pawn on c5. Then if White leaves the queenside unchanged, attention turns to the kingside and … f5. The benefits of … c5 were underlined in Bobby Fischer’s heyday, such as in games that began: 1 d4 f6 2 c4 g6 3 c3 g7 4 e4 d6 5 f3 0-0 6 e2 e5 7 0-0 c6 8 d5 e7 9 d2 c5.

White to play If White wants to open the queenside after … c5 he has to choose between b2-b4 and the en passant capture on c6. He usually gets more out of b2-b4 because this usually assures him of an open file. In contrast, 10 dxc6 bxc6 allows Black play in the center after, say, 11 b4 d5. A common continuation is 10 a3 e8 11 b4. White intends 12 bxc5 dxc5 followed by attacking c5 with b3/ e3 and meeting … b6 with a3-a4-a5xb6. Black can avoid this with 11 … cxb4? 12 axb4 but then White’s initiative, with c4-c5 and c4, is stronger. Fischer’s solution, 11 … b6!, minimizes the queenside danger and allows Black to turn to the other wing with 12 b1 f5 13 f3 f4! 14 a4 g5. One of his games (against Korchnoi, Herceg Novi 1970) saw Black’s

attack prevail after 15 a5 f6 16 bxc5 bxc5 17 b3 g6 18 d2 f6 19 h1 g4!. But if White has time, he will invade with force on the b-file. So, let’s replace 9 d2 c5 in the last diagram with 9 e1 d7 10 d3 f5 11 d2 f6 12 f3 h8 13 c1 and now 13 … c5 14 b1. Who is winning the race now?

Black to play White’s rook became misplaced on c1 once c4-c5 was out of the question. But now b2-b4 is coming and the loss of a tempo is minor. White has the edge after, say, 14 … a5 15 a3 eg8 16 b4 axb4 17 axb4 b6 18 c1 f4 19 b2 and a1. Neverov – Belov, Voskresensk 1990 went 19 … g5 20 a1 xa1 21 xa1 g4 22 bxc5 bxc5 with b8 and a7 coming up. Black needs to act faster, such as with 14 … f4! in the diagram. Then 15 b4 b6 16 bxc5 bxc5 offers chances for both sides (If White plays the blocking 17 g4 Black can open a good file with 17 … g5 and … h5.). A major alternative to … c5 is … c6 and … cxd5. If White recaptures exd5 we have something we’ve already looked at: the Boleslavsky Hole structure when White recaptures with e-pawn after a piece trade on d5. The more common recapture after … cxd5 is cxd5. At first glance, this seems to help White by exposing the base at d6 to attack. After all, the pawn structure that results from … cxd5/cxd5 is the same one that comes about if White plays c4-c5xd6 and Black retakes … cxd6.

However, the opening of the c-file may benefit the first player to occupy it with heavy pieces, as we saw in the Rivera – Suttles example at the beginning of this chapter. The risk to Black is illustrated by: 1 d4 f6 2 c4 g6 3 c3 g7 4 e4 d6 5 f3 0-0 6 e3 e5 7 d5 c6 8 d3 cxd5 9 cxd5 bd7 10 ge2 c5 11 0-0 h5 12 b4! xd3 13 xd3 f5.

White to play Black seems to be doing well. He eased his game by a trade of pieces and achieved … f5. After 14 exf5 he might try 14 … xf5 because it gives him play on light squares (15 e4 f6 16 2g3 d7 and … a6/ … ac8). But White doesn’t need to look at the kingside. Thanks to … cxd5, he should have good prospects on the c-file. This explains 14 ac1 f7 15 c2! in Bagirov – Bednarski, World Student Team Championship 1962. White prepared fc1 and b5! followed by a capture on a7 or invasion at c7. If Black prevents this with … a6 he invites a4-b6. So Black followed a traditional script with 15 … f4 16 f2 g5 17 h3 f6 18 fc1 e8. Black made kingside progress but White’s king can evacuate: 19 b5 a6 20 c7! xc7 21 xc7 xc7 22 b6 f8 23 xc7 and 23 … d7 24 a3 f6 25 a5 e7 26 b6 c8 27 f1! h5 28 e1! g4 29 hxg4 hxg4 30 d2.

Black to play Black couldn’t win on the kingside with just his queen. But if he tried to pass (… f6) White would win on the queenside after c2, c1-b2 and c1-b3-a5 (or c3-a4/ a5/ b6). So Black tried 30 … h6 and was lost after 31 d8! f8 32 c7 and e7. There is another possible pawn break to consider. After … c6 White might play c4-c5!?. That would liquidate four pawns and create one or two passed pawns. Supplemental Game # 4 is a colors-reversed example.

Complications on the f-file White is also entitled to attack the front, rather than the base, of the enemy pawn chain. This means f2-f4. It is promising when: (a) Black has kingside weaknesses, (b) White can push the pawn to f5, or (c) An attack on Black’s queenside pawn base is difficult to achieve. But f2-f4 is double-edged because it may grant Black a great outpost on e5. Najdorf – Ivkov, Bled 1961 began: 1 d4 f6 2 c4 c5 3 d5 g6 4 c3 g7 5 e4 d6 6 f3 0-0 7 e2 e5 8 0-0 g4 9 e1 h6 10 d3 d7 11 f4.

Black to play White looks to his right because c4-c5 is out of the question and the a2-a3 and b2-b4 plan seems slow. If he gets to advance f4-f5 he will own the lion’s share of the kingside. Alternately, if 11 … exf4 12 xf4 he may embarrass Black’s h6-knight with 13 d2. (And 12 … e5 13 xe5 dxe5 14 e3 must also favor him.) But 11 … f5! is a fine response in the diagram. It limits White to a small edge after, say, 12 exf5! xf5 13 e4 or 12 … gxf5 13 d2. White chose 12 fxe5? instead, overlooking 12 … xe5 13 xe5 xe5!.

Then 14 xh6 h4! would threaten 15 … xh2+ and mates, as well as … xh6. Black would have a nice lead in development 14 exf5 xf5. But instead White erred again with 14 f4?. That was punished by 14 … d4+ 15 h1 g5! 16 d2 f4! 17 b1 e5.

White to play King’s Indian players dream about middlegames like this. The pawn structure gives Black a winning plan of … g4/… h4 followed by pushing the g- or f-pawn. White temporarily stopped that with 18 c1 (since 18 … g4? 19 xf4). But he was clearly worse following 18 … g4! 19 xg4 xg4. And he was losing after 20 c2 e8 21 b5 e2! 22 f2 f3 23 c3 xc3 24 xc3 fxg2+ 25 xg2 xf2+ 26 xf2 xc4 followed by … h5/… f8+-f4.

Supplemental Games # 1 White targets the b7 base of a chain: Sokolsky – Livschitz, Minsk 1956 – 1 b4 e5 2 b2 f6 3 b5 d5 4 e3 e6 5 d4 e4 6 d2 c6 7 a4 a6? (This error is similar to Rivera – Suttles. Better is 7 … d6 8 c4 e7.) 8 e2 d6 9 c4 d7 10 c5! (Now 10 … e7 offers White good prospects with 11 f4 f7 12 g4!.) 10 … b8 11 b6! e7 12 b3 0-0 13 a5! c8 14 d2 f5 15 g3 g5 16 h4!

Black to play 16 … f4! (16 … g4? is positional surrender, e.g. 17 c1! f6 18 cb3 g6 19 xb7! xb7 20 a5 d7 21 c3 followed by b1, b6-b7, b6 and winning the a-pawn.) 17 hxg5 fxg3 18 fxg3 f3 19 0-0-0 f8 (Or 19 … xg3 20 xg3 xg3 21 h2!.) 20 f4 f5 21 xe6 xe6 22 h3! xe3 23 h2! h6 24 dg1 e8

White to play 25 xe6+ xe6 26 xh6 xh6 27 gxh6 xg3 28 xb7 h8 29 a5 c8 30 b1 f5 31 h5 e7 32 g5 f4 33 g7 f5 34 b7! e8 35 f7 xh6 36 xf4! xf4 37 xc6 e3 38 b8( ) xb8 39 xb8 e2 40 e1 Resigns. # 2 White plays b2-b4 and bxa5 after …a5: Ivanchuk – Ye, Moscow 2001 – 1 d4 f6 2 c4 g6 3 c3 g7 4 f3 0-0 5 e4 d6 6 e2 e5 7 00 c6 8 d5 e7 9 b4 a5 (Since White cannot play 10 a3? axb4 11 axb4?? xa1 this was once thought to favor Black.) 10 a3 b6 11 bxa5 xa5 12 b4 a8 13 a4 e8 14 d3 d7 15 b1 h6 16 a5

Black to play 16 … c5? (Better was 16 … c5! 17 dxc6 xc6 18 xd6 xa5.) 17 axb6 xa1 18 xa1 cxb6 19 b1 c7 20 b5 d8 21 xc5! bxc5 22 a7 g4?! (Last chance for 22 … f5!.) 23 h3 c8 24 b8 xf3 25 xf3 g5 26 a1

Black to play (White’s winning penetration is a matter of time.) 26 … h5 27 a6 e7 28 c6 f8 29 c7! g5 30 b7 d2 31 h4 a5 32 c6 e1 33 g3 a5 34 f1 d2?! 35 e2 e7 36 xe7 xf2 37 xf2 Resigns. # 3 White kills Black’s kingside attack with g2-g4!: Gurgenidze – Soos, Tbilisi 1965 – 1 e4 c5 2 f3 c6 3 b5 g6 4 c3 g7 5 0-0 e5 6 d3 ge7 7 e3 d6 8 d4 cxd4 9 cxd4 0-0 10 d5 b8 11 fd2! f5 12 f3! d7 13 g4! f6 14 h3 h5 15 g5! h7 16 h4 f7 17 g2

Black to play 17 … fxe4? (Black should switch to the queenside with … a6/… b5 and … d7/… c8.) 18 fxe4 g4 19 e2 xf1 20 xf1 xe2 21 xe2 d7 22 c3 a6 23 d2 f8 24 f1 c8 25 c1 f7 26 d1! f8 27 f2 (The kingside is secure.) 27 … d8 28 c4 b6 29 c2 g7 30 a4! e7? 31 a5 d7

White to play 32 c7! f8 (Desperation.) 33 xd7 xg5 34 h3 xe3 35 xe3 d8 36 f3 xa5 37 h3 b4 38 hg5 xb2+ 39 g3 xg5 40 xg5 e7 41 e6+ h7 42 c8 Resigns. # 4 In a reversed d5 chain, Black plays … c4: Milic – Spassky, Lyons 1955 – 1 e4 e5 2 f3 c6 3 b5 a6 4 a4 f6 5 0-0 e7 6 e1 b5 7 b3 0-0 8 c3 d6 9 h3 b8 10 a4 b7 11 d3 bd7 12 c2 e8 13 bd2 f8 14 f1 c5 15 g3 g6 16 g5 c7 17 h2 d5! 18 f3 g7 19 h4 d4! 20 h5

Black to play 20 … c4! 21 ac1 b6 22 axb5 axb5 23 b1 ac8 24 a2 a6 25 hxg6 hxg6 26 ed1 h7 27 d2 c5 (This completes the strategy begun at move 19.) 28 dxc4 bxc4 29 cxd4 d3! 30 b1 exd4 31 g4 e5 32 xe5 xe5 33 dc1 g5 34 g4

Black to play 34 … xe4! 35 xe4 f5 36 e2 xe4 37 f1 e6 38 b4 b5 39 d1 f8 40 f3 d3 41 c3 a6 42 a1 e2 43 f4 f6 44 d1 xd2! 45 xd2 xc3 46 d6+ g8 47 d1 h7 48 h2+ h6 49 d6 e8 50 c7+ g7 51 c5 d2! White resigns. # 5 Black gets a bad version of … c5: Averbakh – Panno, Buenos Aires 1954 – 1 d4 f6 2 c4 g6 3 c3 g7 4 e4 d6 5 e2 0-0 6 g5 c5 7 d5 a6 8 a4 a5 9 d2 e5?

White to play (This is the same structure as a d5 chain with … c5 but White has a strong initiative. Better was 9 … e6! and … exd5.) 10 g4! e8 11 h4! f5 12 h5 (White threatens 13 hxg6 hxg6 14 gxf5 gxf5 15 h5 with a strong attack.) 12 … f4 13 g5! f7 14 g4 d8 15 xc8 xc8 16 f3 f8 17 e2! (With his king safe in the center. White will win on the h-file.) 17 … g7 18 h4 d7 19 hxg6 hxg6 20 h1 e7 21 h8+

f7 22 h6 f8 23 h1! b8

White to play 24 xf4! c7 (Or 24 … exf4 25 h4 and xf4+.) 25 h2 d7 26 h3 f8 27 xf8+! xf8 28 e6 g8 29 h4! d8 30 xg6+ g7 31 xe5! Resigns. # 6 Black exploits an e5 outpost: Alatortsev – Levenfish, Tbilisi 1937 – 1 d4 f6 2 c4 e6 3 g3 b4+ 4 d2 xd2+ 5 xd2 c6 6 gf3 d6 7 g2 e5 9 d5 e7 9 0-0 0-0 10 e4 d7 11 e1 f5

White to play 12 d3? (Better was 12 exf5! xf5! 13 e4 c5 14 d3 although 14 … b6! and … d4 is OK for Black.) 12 … f4! 13 gxf4?! exf4 14 f3? (White should sacrifice 14 c5! xc5 15 xc5 dxc5 16 b3 with reasonable chances.) 14 … g6! 15 c1 e7 16 e1 de5 17 fxe5 xe5

White to play 18 f3 (Else 18 … f3!) 18 … b6 19 xe5 xe5 20 d2 d7 21 c3 fe8 22 xe5 xe5 23 a3 a5 24 b3 f7 25 f2 f6 26 e2 h5 27 h1 e5! 28 d3 h6 29 h3 g5! 30 h2 g3 31 h4 g8 32 e2

Black to play 32 … g5 33 hxg5 hxg5 34 f2 g4! 35 h5+ d4! 36 d1+ c3 37 h7 gxf3 38 f1 (38 f3 xf3+!) 38 … c2 39 d3 h3! and wins.

Chapter Five: The e5 Chain

The e5 chain If there were no such thing as checkmate, middlegame play would be very similar in the e5 and d5 chains. But the presence of kings means White can quickly develop an attack after e4-e5 thanks to his kingside space edge. Supplemental Game # 1 shows what can happen if Black makes no significant change in the structure. We can identify five basic changes in the e5 chain: (a) Black opens the c-file with … c5 and … cxd4. (b) Black advances … c4 and transfers the attack to the c3 base. (c) Black plays … f6 and trades for White’s e-pawn. (d) White exchanges his d-pawn for Black’s c-pawn. (e) White swaps off both his e- and d-pawns. The first case is the easiest to understand. If Black can use the c-file to draw White’s attention from the kingside, Black doesn’t get mated. Model play was shown in Ornstein – Friedgood, Nice 1974:

1 e4 e6 2 d4 d5 3 d2 f6 4 e5 fd7 5 d3 c5 6 c3 cxd4 8 cxd4 b6 9 f4 d7 10 f3 c8 11 0-0 e7.

c6 7

e2

White to play White aims for f4-f5 while Black hopes to mine the open file with … c7/… b4 and … c2 (or … xd3, if allowed). A good precaution is 12 a3 because stopping … b4 outweighs the weakening of b3. Then on 12 … 0-0 he can prepare f4-f5 with 13 g3. But White erred with 12 f2?, overlooking that 12 … b4! 13 b1 c7 14 c3 can be answered by 14 … a4! (15 xa4 xc1 or 15 d2 xb2). So, after 13 … c7, White fell back on 14 a3 c2 15 a2. But this allowed Black to run riot on the weakened light squares with 15 … a4! (threatening 16 … b3 or 16 … b4) 16 d3 c4! 17 xc4 xc4.

White to play

Black has doomed the d4-pawn: 18 d2 c7 19 b3 walks into 19 … xd4! 20 xd4 xc1+ or 20 bxa4 xe2+. He was on his way to a victory after 18 c3 xd4 19 xd4 xd4 20 e3 c5! and 21 b4 c4! 22 bxc5 xc3 23 cxb6 xe3 (24 bxa7 e1+ 25 f1 xf1+ 26 xf1 0-0 and … a8xa7). But opening, controlling and even penetrating along the c-file is often insufficient, as Supplemental Game # 2 shows. Black has to appreciate when the file is significant. He also needs to weigh other issues: Black may delay … cxd4 because he wants to retain the option of … c4 and because swapping allows White to regain the use of c3. Or he may time … cxd4 so that White cannot retake on d4 with pieces. And most of all Black should not open ‘his’ file if it will end up in White’s hands, as in Steinitz – Halpern, New York 1894: 1 e4 e6 2 d4 d5 3 d2 f6 4 e5 fd7 5 f4 c5 6 c3 c6 7 cxd4 8 cxd4 b4? 9 a3 a5 10 d3 f5 11 b4 b6 12 b3 e7.

gf3

White to play If Black had delayed … cxd4, such as with 7 … b6 8 b3 e7? he would be slightly worse after 9 dxc5 xc5 10 e3. But he made a different error, by failing to follow 7 … cxd4 8 cxd4 with 8 … b6, which threatens … xd4, e.g. 9 b3 a5 10 b1 a4 or 10 a4 b4+. Instead, Black decided on the bishop maneuver to b6, to exert a different kind of pressure on d4. But that was disastrous for his pieces:

13 d2! f8 14 b5 d8 15 b4 f7 16 a4! h6? 17 a5 c7 18 c1 b6 19 a6 g6 20 g3.

Black to play Black is about to lose a piece on the very file he wanted open. He lost soon after 20 … d7 21 c2!.

The … c4 Option If Black has time, transferring his attack from d4 to c3 can pay off handsomely. The reason is that it is usually harder for White to defend c3 than d4. A devastating illustration was Kasparov – Ivanchuk, Horgen 1995: 1 e4 e6 2 d4 d5 3 c3 b4 4 e5 b6 5 a3 f8!? 6 f3 e7 7 h4 h6 8 h5 a5 9 b5+ c6 10 a4 d7 11 e2 b5 12 b3 c5 13 c3 c6 14 0-0 c7 15 e1.

Black to play Black’s slight weakening – 7 … h6, to stop h5-h6 – means White has better than usual kingside prospects. With that in mind, he made a crucial decision, 15 … c4!? and then 16 c2 b6. He cannot be stopped from playing … b4, which opens up more useful lines than … cxd4 would. Chances are equal. But after 17 f4 e7 18 g3 b8 19 h2! d8 White rejected the thematic 20 f4! and 21 f5. Instead, he tried to exploit the kingside with pieces, 20 g4? Black struck first, 20 … b4! 21 axb4 axb4.

White to play Now Black might aim at b2 with 22 … b3 followed by … d7, … a8 and eventually … a4. Or he could target c3 with 22 … bxc3 23 bxc3 d7 followed by … a8/… a7-b5 and … a3-b2. White tried to compete on the queenside, 22 cxb4 xb4 23 b1 d7 24 b3, but couldn’t defend the d-pawn after 24 … a8! 25 xa8 xa8 26 bxc4 xc4 27 c1 a4. The positional massacre ended with 28 e2 a7! 29 e3 xd4 30 xc4 dxc4 31 f1 0-0! White resigns. The … c4 option also allows Black to capture en passant if White tries b2-b4!?. In the French Defense line that runs 1 e4 e6 2 d4 d5 3 e5 c5 4 c3 c6 5 f3 b6 6 a3, White’s plan of 7 b4 is often discouraged by 6 … c4. Similarly in a colors reversed case, 1 d4 d5 2 c4 c6 3 f3 f6 4 c3 e6 5 e3 a6, White often stops … b5 by means of 6 c5. But Supplemental Game # 3 shows that discouraging your opponent from pushing his b-pawn is not enough and you often need to open the position elsewhere.

Attack on the Chain’s Front: … f6 “A pawn chain should only be attacked at its base,” Aron Nimzovich wrote when he denounced … f6 in a French Defense as “strategically unsound.” But if Black cannot attack d4 or c3 quickly, the attack on the chain’s front is the only alternative to living with an inferior pawn structure. Case in point was Westerinen – Korchnoi, Palma de Mallorca 1968: 1 e4 e6 2 d4 d5 3 c3 b4 4 e5 e7 5 d2 b6 6 f3 d7 7 b5 xd2+ 8 xd2 a6 9 a3 a5 10 b5 a6 11 a4 bc6 12 d3 h6 13 c3 0-0 14 b1.

Black to play After White avoided a trade of light-squared bishops (… a6), he virtually announced his plan of d3 and h7 mate. If Black anticipates that with 14 … f5, White must decide whether to leave the structure as it is – and turn his attention to the queenside – or change it. And if he changes it, his choice is 15 exf6 en passant or perhaps a well-prepared g1/g2-g4. Black decided not to give White the later option and played 14 … f6! instead. He didn’t fear 15 d3? fxe5! 16 h7+ f7. The game went 15 exf6 xf6 16 c2 g6 17 0-0 af8. When Black recaptures on f6 with a piece, he loses his last chance for pawn control of e5 and makes his e-

pawn backward. But the opening of the kingside makes the White king the board’s most vulnerable target. Black is preparing a standard Exchange sacrifice on f3 and White did nothing to deter him: 18 e1 xb5 19 axb5 ce7 20 d3

Black to play and now 20 … xf3! 21 gxf3 h4 22 h7+ f7. Besides the fork on f3, Black’s winning ideas include … e5 followed by … h3-g2 mate, and … eg6 followed by the queen-trapping … h8. White found nothing better than 23 h1 xf3 24 g1?! and resigned after 24 … xg1 25 g6+ xg6 26 xg1 e7 27 xg7+ e8 28 g2 e5. But see Supplemental Game # 4 for a case, with colors reversed, of the backward e-pawn becoming significant. If c-pawns are also traded we get a center with a White pawn at d4 and Black ones at d5 and e6, e.g., 1 e4 e6 2 d4 d5 3 d2 f6 4 e5 fd7 5 d3 c5 6 c3 c6 7 e2 and now 7 … cxd4 8 cxd4 f6 9 exf6 xf6 is common. In theory, that structure should favor White, particularly if he can trade off dark-squared bishops. In practice, however, Black often obtains excellent chances through pressure on d4 and against the kingside.

The Wedge Formation One of the most common versions of an e5 chain emerges after White’s d-pawn is exchanged for Black’s c-pawn. This comes about when … c5 is followed by dxc5 or by … cxd4 (when a White piece rather than a pawn retakes on d4). The result is the Wedge formation. Its main features are outposts for White at d4 and for Black at c4 and e4. When Black wins in a Wedge he often benefits from the half-open cfile and from White’s lack of a d-pawn to support the e5-pawn. When White wins he usually exerts kingside pressure, as in Maroczy – Singer, Rogaska Slatina 1929, a colors-reversed of Tchigorin’s plan in the Slav formation: 1 d4 f6 2 f3 e6 3 e3 e7 4 d3 d5 5 bd2 c5 6 c3 c6 7 0-0 a6 8 e2 0-0 9 dxc5! xc5 10 e4 e8? 11 e5! d7 12 b3 e7 13 f4 f8 14 ad1 b5 15 fd4 d7 16 g3

Black to play Black has a bad version of the Wedge because (a) he hasn’t been able to use his c4-outpost, (b) his bad bishop is really bad, and (c) his only major source of counterplay to distract White from f2-f4-f5 is … f6, which seriously weakens the e-pawn. Black blocked the attacking line b1-h7 with 16 … f5. But this only delayed the opening of the kingside. Play went: 17 f4! d8 18 f2 b8

19 g4!. Black replied 19 … g6 so he could meet gxf5 with … gxf5. But this invited h2-h4-h5 and play went 20 h4! a5 21 xa5 xa5 22 a3! (to rule out … b4) 22 … b6 23 g2 xd4? 24 xd4 e7 25 h5! and then 25 … g7 26 f2! e7 27 h1 f7 28 dg1 e8 29 e3 d8 30 d2 f7 31 c5.

Black to play Black’s position is so bad that he should sacrifice, 31 … d4, to activate his bishop at b7. He chose the immediate 31 … b7 and resigned after 32 gxf5 gxf5 33 xg7 xg7 34 g1 f7 35 f2! c6 36 g3 b7 37 g8! e8 38 xf8 in view of 38 … xf8 39 xe6+. But Supplemental Game # 5 shows how Black can create counterplay in a Wedge.

Liquidation of e5 and d4 Nimzovich’s fellow Hypermoderns held back their center pawns because they feared they would become targets too easily. Nimzovich took a different approach: He pushed those pawns in order to trade them off, such as with exf6 and dxc5, and then occupy d4 and e5 with pieces. The key to success with such a pawn structure is, not surprisingly, control of the two outpost squares, as in Gurgenidze – Fuchs, Tbilisi 1969: 1 e4 e6 2 f3 d5 3 c3 f6 4 e5 fd7 5 d4 c5 6 b5!? c6 7 0-0 a6? 8 xc6 bxc6 9 a4 e7 10 c3 0-0 11 b3 cxd4? 12 cxd4 b6 13 xb6 xb6 14 g5! xg5 15 xg5 f6 16 f3 fxe5? 17 xe5 c5 18 dxc5 xc5 19 c1 d6 20 d4!.

Black to play The queen and knight dominate the center and the inevitable trades of rooks only emphasized this: 20 … d7 21 fe1 ac8 22 h3 xc1 23 xc1 c8 24 xc8+ xc8 25 b4! c7 26 f4! c1+ 27 h2. Black sought counterplay with 27 … g5 28 fxg5 xg5 29 a4 h5 30 h4 f6. But the b-pawn and supporting pieces were too strong, 31 b5! axb5 32 axb5 b7 33 b6 h7 34 d3+ h8 35 g3 h7 36 f3! h6 37 c7!.

But if Black pressures d4 and c5 or finds other counterplay the chances fall into balance, as in Van der Sterren – Korchnoi, Wijk aan Zee 1984: 1 e4 e6 2 d4 d5 3 c3 f6 4 g5 e7 5 e5 fd7 6 xe7 xe7 7 f4 0-0 8 f3 c5 9 dxc5 xc5 10 d2 b6 11 b5? c6 12 c3 f6! 13 exf6 xf6.

White to play Given time, White would get the edge via bd4xc6 followed by e5, 0-0-0, g2-g3 and d4. But that’s a lot of time. After White dawdled with 14 d3? c4! and 15 xc4 xc4 16 bd4 d7 he found he couldn’t castle (17 0-0-0 xa2) or secure an outpost (17 e5? xe5 18 fxe5 f7 and … af8). His best chance was to head for an endgame with 17 g3 and 18 e2. But he chose 17 b3 a6 18 0-0-0? instead and Black’s initiative grew overwhelming following 18 … xd4 19 xd4 af8 20 g3 e5!.

White to play White’s dxc5/exf6 policy has failed. He cannot allow 21 fxe5 f2 or 21 f3 exf4. The rest was a crush: 21 e2 d4! 22 cxd4 g4 23 de1 c6+ 24 b2 fc8 25 c1 f5 26 b4 b6 27 e7 a5! 28 a4 xa4 White resigns.

Supplemental Games # 1 No change in the pawn structure? White’s attack rolls on: Kasparov – Short, Zurich 2001 – 1 e4 e6 2 d4 d5 3 d2 c5 4 gf3 f6 5 e5 fd7 6 c3 b6 7 b5 e7 8 0-0 0-0 9 e1 (White can target e6 in case of … f5/exf6.) 9 … a5 10 f1 a6 11 a4 xb5? 12 axb5 (Black’s QN has no good square now.) 12 … e8 13 g3

Black to play 13 … f8 (This invites an attack on g7. But if Black had played … f8, h7 would be vulnerable to c2/ g5.) 14 h5! bd7 15 g3 g6 16 h4 df8? (Now was a good time for 16 … cxd4 17 cxd4 c8.) 17 g2 d7? 18 h6! gxh6 19 d2 f5 (Otherwise mate follows 20 xh6.) 20 exf6 d8 21 xh6 a7 22 g5 xb5 23 f7+ xf7 24 xf7 Resigns. # 2 The open c-file isn’t enough: Volokitin – Bareev, Russia 2007 – 1 e4 c6 2 d4 d5 3 e5 f5 4 f3 e6 5 e2 d7 6 0-0 g6 7 bd2 h6 8 c3 e7 9 e1 c5 10 df3 f5 11 d3 cxd4 12 cxd4 h5 13 f4 xf3 14 xf3 g6 15 g4! g7 16 d3 0-0 17 g3

Black to play 17 … c7 (Black should challenge the g4-bishop, e.g. 17 … h5! 18 e2 b8 and 19 … c6.) 18 h4 ac8? 19 h5! c4 20 hxg6 hxg6 21 e3! (White’s winning plan is g2/ h1-h6.) 21 … fd8 22 g2 g5 23 h1 c2

White to play 24 f3 (He could have won with 24 xg6!. Now 24 … xf4 may hold.) 24 … f8? 25 xd5! xd5 26 xg5 xb2 27 f6 d7 28 xg7 xg7 29 h7+! Resigns (29 … xh7 30 xf7+ h6 31 h1+). # 3 White can open the queenside favorably despite … c4: Romanishin – Mueller, Altensteig 1992 – 1 e4 e6 2 d4 d5 3 e5 c5 4 c3 c6 5 f3 b6 6 a3 c4 7 e2 a5 8 bd2 d7 9 0-0 h6 10 a4! (White avoids … b3 and then xb3/… a4! when Black can invade on the light squares.) 10 … 0-0-0 11 e1 f5?

White to play 12 b4! cxb3? (Having given up on … f6, he had to keep the queenside closed.) 13 d3 b8 14 a3 xa3 15 xa3 c8 16 a1 e7 17 b1 c7 18 c5 a8 19 dxb3 ec6 20 b5 xb3 21 axb3 b8

White to play 22 xc6! xc6 (22 … bxc6 23 a6 and 22 … xc6 23 a5 and 24 b6! are also lost.) 23 xe6 e7 24 c5 f4 25 f3 g5 26 a5 h7 27 b6! (Not 27 a6? b6.) 27 … c8 28 a6! bxa6 29 xa6 e8 30 b7! xb7 31 xb7 Resigns. # 4 The downside of attacking the front of the chain: Bondarevsky – Botvinnik, Moscow 1941 – 1 d4 d5 2 f3 c6 3 e3 g4 4 c4 e6 5 c3 d7 6 d3 gf6 7 0-0 e7 8 b3 0-0 9 b2 e5! 10 e2 e4! 11 d2 xe2 12 xe2 b4! 13 a3 xc3 14 xc3 e8

White to play 15 f3 (This is what Black’s last two moves were designed to exploit. But otherwise Black makes progress with … f8-g6-h4/ … d7, etc.) 15 … f8 16 f2 d7 17 af1 exf3 18 xf3 e6! 19 d3 ae8 20 b1 g6 21 e1?

Black to play 21 … dxc4! 22 xc4 (Or 22 bxc4? e5!.) 22 … xe3 23 xe3 xe3 24 f2 d5! 25 d2 gf4 26 h3 c3 27 a4 e2+ 28 h2 xh3+! White resigns (29 gxh3 df4 and … xh3 mate.). # 5 Black finds anti-Wedge counterplay: Bondarevsky – Lilienthal, Stockholm 1948 – 1 e4 e6 2 d4 d5 3 c3 f6 4 g5 e7 5 e5 fd7 6 xe7 xe7 7 f4 0-0 8 f3 c5 9 d3 f5! 10 0-0 a6 11 dxc5 xc5 12 e2 c6 13 c3 d7 14 ed4 ac8 15 c2 e4 16 e2 h8! (Black’s e4 outpost is at least as good as White’s d4 and he has his choice of breaking with … g5 or … b5-b4.) 17 h1? xd4 18 xd4

Black to play 18 … g5! 19 e3 (On 19 g3 Black attacks a new base with 19 … g4 and … h5-h4.) 19 … gxf4 20 xf4 g8 21 d1 b5! 22 xb5? axb5 23 e2 (Stopping … b4 with 23 a3 invites 23 … c4 and 24 e2? g3+! or 24 e3 g7.) 23 … b4! 24 c4!? dxc4 25 g4 h4 26 ac1

Black to play 26 … c3 27 bxc3 xc3 28 xc3 bxc3 29 e3 c2 30 g2 and Black won with 30 … c8 31 c1 c3 but 30 … xg4+! 31 xg4 xg4+ 32 h1 d1! is faster. # 6 Passivity loses a Wedge endgame: Tal – Stahlberg, Stockholm 1961 – 1 e4 e6 2 d4 d5 3 c3 f6 4 e5 fd7 5 f4 c5 6 f3 c6 7 e3 cxd4 8 xd4 c5 9 d2 xd4 10 xd4 xd4 11 xd4 b6 12 b5! xd4 13 xd4 b6? (Better is 13 … e7 and … b8-c6 or … c5.) 14 d3 d7 15 0-0-0 0-0-0 16 hf1 g6 17 e2 b8 18 d3 h5 19 h3 c6 20 ff3 c8 21 b3 d7 (Black needs to reposition, …

d7!.) 22 d2 cd8? (22 … cg8 and … g5 was his last real chance.)

White to play 23 h4! c8 24 fh3 hg8 25 g4 hxg4 26 xg4 h8 27 gh4 xh4 28 xh4 f8 29 h7 e7 30 f3 c6 31 g5 d8 32 h4 c6 33 e3 c8 34 d3 e8 35 d4 b6 36 e3 g8 37 f3! b7 38 h2 c6 39 g4 c5 40 c3 a5 41 a3 b5 (Or 41 … b5 42 f6 f8 43 xg6! fxg6 44 c7+.) 42 b4+ b6 43 f6 f8 44 f3 axb4 45 axb4 d4 (Otherwise g4-g5-h6-g7.) 46 cxd4 c6 47 xe8 xe8 48 xf7 d4+ 49 g4 Resigns.

Chapter Six: The King’s Indian Complex

The basic King’s Indian formation ’Complex’ is right. Each member of the family of structures that stem from the diagram has such distinct and difficult qualities that any pawn capture or advance must be weighed with extreme care. Obvious moves can be positional blunders. For example, after dxe5 and … dxe5 White may temporarily control the d-file. But the trade eases Black’s game and often allows him to occupy d4 later on. On the other hand, … exd4 in the diagram may increase Black’s piece play, particularly if he has a bishop on the g7-b2 diagonal. But it also exposes his now-backward d6-pawn to pressure. You can recognize the basic formation in King’s Indian and Old Indian Defenses. The difference between the two is the placement of Black’s king bishop, on g7 or e7. This structure also arises in Bogo-Indians and even, with colors reversed, in a Ruy Lopez or Sicilian Defense. The key changes in the basic structure are: (a) White pushes his d-pawn to d5.

(b) He trades it for Black’s e-pawn, dxe5. (c) Black exchanges on d4, … exd4, or (d) White pushes c4-c5. The first change, d4-d5, creates a d5 chain like those we saw two chapters ago. Supplemental Game # 1 shows the danger to White of … cxd5. The last of the changes, c4-c5, is the most unbalancing and is illustrated by Supplemental Game # 2. The others require greater explanation.

White plays dxe5 When White exchanges on e5 he is expecting to profit from the opening of the d-file, the occupation of c5, the seizing of the g1-a7 diagonal or all three. Some computers love to play dxe5 because of how they evaluate control of files and diagonals. Let’s examine White’s choices after: 1 d4 f6 2 c4 g6 3 c3 g7 4 e4 d6 5 f3 0-0 6 e3 e5 7 ge2 c6 8 d2 bd7 9 0-0-0 a6.

White to play First, we can cross off c4-c5 because it eases Black’s game after 10 c5 exd4! 11 xd4 dxc5 12 xc5 e8. Next up is 10 d5. If Black keeps matters closed, 10 … c5, we have a wing battle brewing. White has the better chances after 11 b1 and 12 g4. Black may have to sacrifice with … b5!? to have any queenside play at all. But the problem with 10 d5 is 10 … cxd5! 11 cxd5 b5! and 12 … b6. Then Black’s wing attack is coming much faster than White’s. True, White can play 11 xd5 (instead of 11 cxd5). That dooms the d-

pawn after 11 … xd5 12 xd5. But 12 … b6! and 13 xd6? xd6 14 xd6 xc4 is strong for Black. And 13 xb6 xb6 14 xd6 a5 or 14 … xd6 15 xd6 e6 offers compensation on dark squares for the lost pawn. When there’s no good way to change the structure, the best strategy is to wait. A useful pass is 10 b1! because then the opening of the c-file (or of part of the a-file after … b5/cxb5/… axb5) holds fewer dangers. Black shouldn’t change the structure since 10 … d5? and 10 … c5? just lose pawns and 10 … exd4 11 xd4 makes his d-pawn a target. His chief alternatives are 10 … a5 and 10 … c7. Geller – Boleslavsky, Moscow 1952 (reached by transposition) went 10 … a5 11 c1! e8 12 b3 c7. Note that Black played … a6 earlier because he wanted the option of … b5. But that weakens the dark squares a bit and White exploited them with 13 dxe5! dxe5 14 c5!.

Black to play Black can fight for control of d6 with 14 … f8 but after 15 a4 his pieces are tied up and White has a free hand for g2-g4 and h2-h4-h5. Instead, the game went 14 … f8 15 d6!. After a queen trade, White can target the b7-pawn by getting knights to a5 and d6. Black can put up some defense with 15 … xd6 and … e6/ … ab8. He chose 15 … e6 16 c4 f8 17 xc7 xc7 and allowed 18

a5 b8 19 a4 e6 20 xe6 xe6 21 c4 c7 22 ab6. White is preparing to double rooks on the file and invade the seventh rank with d7. The rest was one-sided: 22 … e6 23 b4 f4 24 d2 bd8 25 hd1 xd2 26 xd2 g7 27 a5 b8 28 xf4 exf4 29 d7! d8 30 xf6+ xf6 31 xd8+ xd8 32 xb7 and wins. Let’s scroll back to the position after 10 b1. If Black needs a more active plan, what about 10 … b5 and 11 c1 b7 ? Then he is inching closer to playing … bxc4 followed by … d5!? or … exd4/ … e5. This is why 12 dxe5! makes sense again. White need not rush c4-c5 this time but can exploit the d-file and g1-a7 diagonal after 12 … dxe5 13 b3 c7 14 c1! b8 15 f2, as in Ivanchuk – Zapata, Novi Sad 1990. Then … c5 would lose a pawn. Other changes in the structure, … dxc4/ xc4 and … b4/ a4, help White’s pieces. Black left the pawns where they were and play went 15 … e8 16 g3 f8 17 e2 e6 18 hd1.

Black to play White has various ways of making progress, such as a5 and b6 or c5xe6. The game went 18 … f8 19 a3 d8. But that made 20 xd8 xd8 21 a5! strong (21 … xa5 22 b6). Black was clearly in trouble following 21 … e7 22 f4! since 22 … exf4 23 gxf4 will be followed by e4-e5 of f4-f5. Instead, Black retreated,

22 … d8 23 f5! knockout.

d7 24 c5! c8 25 g4

b8. This is a pawn play

White still needed a winning target and he found one at f7: 26 f1 d7 27 h4 e8 28 d1! b7 29 xb7 xb7 30 b3 e7 31 fxg6 hxg6 32 g5! d7 33 d1 Resigns in view of 33 … c7 34 d8.

Downside of dxe5 But after dxe5 White may not be able to occupy the key invasion square at d6. And if he does occupy it, Black might drive the invading piece back with … f8. In addition, the liquidation of d-pawns means that Black can end up with more control of d4 than White, as in Hernandez – Berliner, Southern Championship 1949: 1 d4 f6 2 f3 g6 3 c4 g7 4 c3 0-0 5 e4 d6 6 d3 bd7 7 0-0 e5 8 dxe5? dxe5 9 e2 c6 10 b3 c7 11 a3 d8 12 ad1.

Black to play White controls the newly opened a3-f8 diagonal. That might mean something if he could play c2 and d6. But Black foiled that with 12 … f8!. Then 13 xf8 xf8 would allow his knight to head for d4 or f4. If … e6-d4 and xd4/… exd4 is played, Black gets a nice passed dpawn. White preferred 13 b2. His lack of pawn control of d4 began to pinch after 13 … h5 14 g3 c5 15 c2 g4! with … d4! coming up. Black won after 16 g2 e6 17 e3 c5 18 c1 and 18 … hf4+! 19 gxf4 xf4+ 20 g3 xf3 21 xf3 xd1 22 xd1 d8 in view of 23 c1 c8 and … h3 mate. If White plays c4-c5 after dxe5 he can exploit the a2-f7 diagonal as we saw in Ivanchuk-vs.-Zapata. But there is a danger that the c5-pawn will

be over-extended, as in Gross – Kalme, U.S. Junior Championship 1955: 1 d4 f6 2 c4 g6 3 c3 g7 4 f3 d6 5 f4 0-0 6 h3 fd7 7 e3 e5 8 h2 c6 9 e2 e7 10 dxe5 dxe5 11 0-0 d8 12 c2 f8 13 ad1 f5! 14 e4? d7 15 d2 h6? 16 dd1? f6 17 b4 a6 18 c5? c7 19 c4+ e6 20 b3 xc4 21 xc4+ e6 22 xe6+ fxe6 23 xd8+ xd8 24 d1 xd1+ 25 xd1.

Black to play If White realized how bad his position was becoming he would have exploited Black’s slip at move 15 with 16 xe5! xd2 17 xd2. That would free his bishop and eliminate the powerful Black center pawn. Instead, he advanced his queenside pawns based on tactics like 18 c5 xb4? 19 b3+ But the pawns went too far, as 25 … a5! showed, e.g. 26 a3 axb4 27 axb4 a6 and … xb4. White tried to complicate with 26 bxa5 (instead of 26 a3) and then 26 … xc5 27 d4 because the e-pawn was doomed any way (27 c3 b5). Black won with 27 … exd4! 28 xc7 xe4 29 b6 c5 30 f3 c3! 31 b2 f8 followed by the king’s advance to support the c-pawn’s promotion.

The Boleslavsky Wall: … exd4 The formation created by an exchange of pawns on d4 was dubbed the Boleslavsky Wall by the pawn play authority Hans Kmoch, in honor of the Soviet grandmaster who pioneered the Kings’ Indian Defense in the 1930s, ‘40s and ‘50s. The formation was then viewed suspiciously because Black’s d-pawn can come under heavy fire on the half-open dfile. That shouldn’t be surprising. The same pessimistic view prevailed for the Hedgehog. The only difference between the two formations is that Black has a pawn at c6 and no e-pawn in the King’s Indian, whereas in the Hog he has a pawn at e6 and no c-pawn. The chief benefits of … exd4 are tactical. Black seeks threats along the h8-a1 diagonal and against the e4-pawn and tries to use the e5 outpost. A strikingly modern example was Paulsen – Anderssen, Leipzig 1877: 1 e4 e5 2 c3 c5 3 f3 c6? 4 xe5! xe5 5 d4 d6 6 dxe5 xe5 7 e2 c6 8 0-0 f6 9 f3 h6 10 e3 d6 11 d4 e6 12 b3 0-0 13 d3 e7 14 ae1? fd8 15 d1 d7 16 c4 f6! 17 c3 a5! 18 e2 g5 19 f3 f8 20 d2 g6 21 g3.

Black to play The fork trick 4 xe5 gave White a superior center but he showed no understanding of how to handle it. He gave up chances for e4-e5 or f3-

f4-f5 and misplaced his QR (Ironically, White was Louis Paulsen, who is regarded today as a strategist well ahead of his time, whereas Adolf Anderssen, who was Black, is considered a onedimensional Romantic.). When Boleslavsky and David Bronstein popularized the Wall after World War II they used all of the Black ideas that Anderssen did, including … a5-a4!, the … e5/… f6 battery and the bid to occupy f4 with a knight. They would most likely have continued 21 … h5 followed by … g7 and … h4 to create a target at g3 or secure f4 for the knight. Anderssen chose another idea, liquidating the center with 21 … d5!? and then 22 cxd5 cxd5 23 e3 dxe4 24 xe4 d5. Black won after 25 g4 g7 26 f2 e6 27 a4 d4 and … g4. White’s basic strategy against the Wall is to keep d6 under pressure. If Black has developed his bishop at g7, White wants to force him into … f8. Then the same White strategies that are good in the Maroczy Bind and Hedgehog – such as looking for a chance for e4-e5 or c4-c5 – are at the top of his wish list. More interesting are Black’s weapons.

Wall Weapons: … d5 The most rewarding Black break, as in the Maroczy Bind, is a welltimed … d5. This works best when White has been forced to defend the e-pawn with f2-f3 or has voluntarily made that move. Then the liquidation of pawns after … d5 can expose a hole at e3. 1 d4 f6 2 c4 g6 3 g3 g7 4 g2 0-0 5 f3 d6 6 c3 bd7 7 0-0 e5 8 e4 exd4 9 xd4 e8 10 f3?

Black to play White’s e-pawn is often attacked by … c5 so his last move is logical. But here f2-f3 is bad because he cannot stop a … d5 break after 10 … c6!. In similar positions the best White policies are (a) trying to use betterplaced pieces to exploit the opening of the center or (b) exchanging exd5 and then push c4-c5. But in this case, Black controls c5 so policy (b) doesn’t work. White can pursue the first strategy with 11 e3 and then 11 … d5 12 cxd5 cxd5 13 db5, with the idea of d6. But what if White ignores … d5 ? Then Black has fine play in lines such as 11 c2 b6 12 d3 e6 13 e3 fd7! 14 b3 d5! or 11 b3 b6 12 e2 e6 13 d2 d5 (14 e5

fd7 15 cxd5 cxd5 16 f4 c5 and … d4!. In Liebert – Kuzmin, Zinnowitz 1971 White chose 11 e1? d5 12 exd5 xe1+ 13 xe1 but underestimated 13 … b6!, which threatened 14 … xc4 or 14 … fxd5 (15 cxd5 xd4+). His position quickly deteriorated: 14 c2 xc4 15 dxc6 bxc6 16 h1 d5! 17 f1 e5 18 d1 b6! 19 g2 f5 20 xd5 cxd5 21 g4 d3 22 e3 c4 23 xd5? d8 24 f4 xd5 25 xd5 c6 26 f3 xd5! White resigned (27 fxe5 d1+ 28 g2 g1+ 29 f2 f1+). But if White can carry out policy (b) he may obtain clear superiority. Why? Because it isolates a Black pawn on d5, establishes d4 and possibly d6 as outposts and creates a passed c-pawn. For example: 1 d4 f6 2 c4 g6 3 g3 g7 4 g2 0-0 5 f3 d6 6 0-0 bd7 7 c3 e5 8 e4 c6 9 b1 e8 10 h3 exd4 11 xd4.

Black to play On 11 … c5 White can avoid 12 f3? in favor of 12 e1. Then Black has to find tactical ideas, such as … fd7-e5-d3, to prevent White from coordinating his forces and making the d6-pawn the main middlegame issue. But suppose Black insists on the … d5 plan. He doesn’t have support for the immediate 11 … d5?. But once again he can try 11 … b6 12 b3 d5.

However, White can use policy (b) – 13 exd5 cxd5 14 c5! – which throws Black into disarray, e.g. 14 … bd7 15 b4 and db5-d6. In Kavalek – Garcia-Orus, The Hague 1967, Black tried to mix it up with 14 … e4!? 15 xe4 dxe4, attacking the knight on d4. But 16 b5 was still strong and White prevailed after 16 … d5 17 b2 xb2 18 xb2 f6 19 d2! a6 (19 … c3? 20 a1! or 19 … e7 20 c7) 20 d6 c3 21 a1 e3 22 xc3 exd2 23 xe8 and wins. Black needs to play sharply and accurately in exd5/c4-c5 positions and Supplemental Game # 3 shows how.

Wall Weapons: … f5, … c5 When … d5 is not possible – which is often the case when White has developed his bishop at g2 – Black may have a good alternative in … f5. This is not as loosening as in the Hedgehog because Black has a bishop, not an e6-pawn, to control f5, e.g. Kan – Boleslavsky, Moscow 1952: 1 d4 f6 2 c4 g6 3 f3 g7 4 g3 0-0 5 g2 d6 6 0-0 bd7 7 c2 e5 8 d1 e8 9 c3 c6 10 e4 exd4 11 xd4 e7 12 h3 c5 13 f3.

Black to play Black made a good prophylactic move, 13 … a5!, to delay b2-b4. White has to avoid 14 e3 because this makes 14 … d5! possible, since exd5?? would hang the bishop. And 14 b3 (to trade off the good Black knight) can be met by 14 … fd7 since 15 e3 is met strongly by 15 … f5!, e.g. 16 f4 e5 17 xc5 dxc5 18 exf5 xf5 19 b3 d3!. A good plan after 13 … a5! is 14 b3 and 15 b2. Instead White chose 14 g4?! to rule out both … f5 and … d5. But the weakening of the kingside makes … f5 more attractive, not less. Black prepared it with 14 … fd7!. There followed 15 b1 e5 16 b3 f5! 17 exf5 gxf5 and now 18 xf5 xf5 19 xf5 f8 and … xf3+ would favor Black. So White tried

18 gxf5, overlooking 18 … ed3! 19 xd3 xd3 20 xd3 e1+ 21 f1 g3+ 22 h1 e1! (23 e3 xh3+ 24 g1 g3+ 25 h1 h4+ 26 g2 xe3! 27 xe3 xd4). Around the turn of the 21st century another Black strategy, previously dismissed as positional poison, became popular. The key to it is … c5 followed by … b5. The goal is to create a target on c4 or to disorient White with … b4. An illustration was Lingnau – Golubev, Berlin 1993: 1 d4 f6 2 c4 g6 3 f3 g7 4 g3 0-0 5 g2 d6 6 0-0 bd7 7 c3 e5 8 e4 exd4 9 xd4 e8 10 h3 a6 11 e1 b8 12 b1 e5 13 b3.

Black to play Both sides took good preliminary steps at moves 10-12 and now, before White gets to play 14 f4!, Black began his strategy with 13 … c5!?. Given time, the d6-pawn can prove horribly weak. But Black’s counterplay comes first, e.g. 14 c2 b5 15 e3 bxc4 or 15 cxb5 axb5 16 e3 b4 17 cd5 xd5 18 xd5 a6. White preferred 14 f3 and 14 … xf3+ 15 xf3. That gave Black an alternative plan, 15 … d7, followed by … e5 and perhaps … c6-d4. But he chose 15 … b5, hoping for 16 cxb5 axb5 and … b7 or … b4. White preferred 16 f4! h5 17 d2 and thanks to … h5 he wasn’t afraid of 17 … b4 18 d5. The game continued 17 … e6 18 d3

f6.

White to play This illustrates a basic principle of the KID formation: Black must be willing to lose a pawn in many middlegames or he will be slowly ground down. Here he’s willing to play 19 cxb5 axb5 20 xb5 d7, for example. The game took a different turn, 19 ed1 d7!. Now 20 f4! would be thematic and the position becomes unclear. But White chose 20 xd6 bxc4 21 bxc4 xb1 22 xb1 so that 22 … xc4 23 b7 offers good chances. Instead, Black chose 22 … a5 and White succumbed to tactics, 23 c1 d4! 24 c2 a3 25 e1 e5 26 e2 f3+ 27 xf3 xf3 28 xd4 cxd4 (threatening … xh3 and mates) 29 f4 d3! 30 d2? b1 and wins. Supplemental Game # 4 shows how the … c5 plan can backfire badly. Note that if White develops his KB on e2 rather than g2, his e-pawn can become a target after … c5 as in Supplemental Game # 5.

Supplemental Games # 1 Black dominates the c-file after d4-d5/… cxd5!: Baburin – Trent, Bunratty 2012 – 1 d4 f6 2 f3 g6 3 c4 g7 4 g3 0-0 5 g2 d6 6 0-0 c6 7 c3 b6 8 e4 g4 9 b3 a5 10 d2 bd7 11 h3 xf3 12 xf3 e5 13 d5 cxd5 14 cxd5 fc8 15 b2 h5 (White should either contest the file or play for f2-f4.) 16 ad1?! h7

White to play (Black readies … h6!.) 17 a4 xd2 18 xd2 h6 19 e2 b5! 20 c3 ab8 21 h4? c5 22 d1 a5 23 b1 b4 24 d2

Black to play 24 … xd2! (To stop 25 c4!.) 25 dxd2 a4 26 bxa4 xa4 27 c2 xc2 28 xc2 b3! White resigns.

# 2 White breaks with c4-c5!: Benko – Medina, Havana 1966 – 1 d4 f6 2 c4 g6 3 g3 g7 4 g2 0-0 5 c3 d6 6 f3 bd7 7 0-0 e5 8 h3 c6 9 e4 a6 10 e3 e7 11 c2 e8 12 fe1 b5

White to play 13 c5! (Strong even without … b5. Now 13 … dxc5 14 dxe5! h5? 15 g4 costs a piece and 13 … exd4 14 cxd6 xd6 15 xd4 and e4-e5 favors White.) 13 … dxc5 14 dxe5 xe5 15 xe5 xe5 16 f4 c7 (Or 16 … e7 17 e5 d7 18 e4.) 17 xc5 d7 18 f2 b7 19 e5! f5 20 d5 c8 21 b6 xb6 22 xb6 f8 23 ac1 e6

White to play 24 b3! ac8 25 xe6+ xe6 26 ed1 g5 27 d7 e7 28 cd1 a8 29 xe7 xe7 30 d7 e8 31 a7 gxf4 32 gxf4 b4 33 f2 b8 34 a3 d2 35 c7 c8 36 e6 b4

White to play 37 e5! bxa3 38 bxa3 e8 39 g7+ f8 40 xh7 Resigns. # 3 Black tactics defeat … d5/exd5 and c4-c5 in the Wall: Gelfand – Topalov, Linares 1997 – 1 d4 f6 2 f3 g6 3 c4 g7 4 c3 0-0 5 e4 d6 6 e2 e5 7 e3 exd4 8 xd4 e8 9 f3 c6 10 f2 d5 11 exd5 cxd5 12 0-0 c6 13 c5 h5 14 g3 h3 15 e1

Black to play 15 … g5! (The threat is 16 … xg3 17 hxg3 xd4 18 xd4? xg3+.) 16 db5 ad8 17 d6 d4! 18 c1 (Black is winning after 18 xe8? xg3! or 18 xd4 xg3. But 18 f4!? is critical.) 18 … e3! 19 xe3 xe3 20 f2 d4 21 d1? (Chances are unclear after 21 ce4! e7 22 c4.)

Black to play 21 … xe2+! 22 xe2 d5 23 e4 e5 24 g5 e8! 25 d2 c4 (25 … g4+! is faster.) 26 dc3!? h6 27 h4 dxc3 28 d8 cxb2! 29 xe8+ g7 30 d1 c2+ 31 d2 b1( ) 32 xh3 cc1 White resigns. # 4 Black’s … c5 backfires: Mikhalchishin – Vujadinovic, Sibenik 1990 – 1 d4 g6 2 c4 g7 3 e4 d6 4 c3 e5 5 ge2 d7 6 e3 h6 7 g3 g4 8 xg4 exd4 9 g5 f6 10 f3 dxc3 11 xc3 0-0 12 d2 e8 13 d3 c6 14 0-0 b6 15 e3 c7 16 d4! d7 17 ad1 ad8 18 b3

Black to play 18 … c5?! (Bad since … b5 or other structure changes can’t be carried out. Better is 18 … e6 and … e8 to trade bishops.) 19 c3 c6 20 b2 e6 21 f5! gxf5 22 exf5 e5 23 f4! g4 (Or 23 … ee8 24 xf6 with a winning extra pawn.) 24 fxe5 dxe5 25 e2 xh2 (Hoping

for 26 xh2? e4+.) 26 f6! e4 27 xe4 e8

White to play 28 xh7+! Resigns (28 … xh7 29 h5+ h6 30 d2! and wins). # 5 Black mixes … c5 and … f5: Miles – Glek, Biel 1996 – 1 d4 f6 2 f3 g6 3 c4 g7 4 c3 0-0 5 e4 d6 6 e2 e5 7 0-0 a5 8 e3 a6 9 h3 exd4 10 xd4 e8 11 d2 (Now 11 … c5 12 xc5! dxc5 13 f4 and e4-e5 favors White.) 11 … c6 12 e1 c5 (But here 13 xc5 dxc5 14 f4 d7! 15 e5 f6 or 15 … g5 is unclear.) 13 f1 d7 14 g3 e6 15 e3 c5! 16 f3 c6 17 d2 d7 18 a4 h5! (Since … b5 is out of the question, Black aims for dark square play with 19 … h4 20 g4 e5.) 19 h4 e5 20 f3

Black to play 20 … f5! 21 d5 f8 (Now f3 and g3 are targets.) 22 f4 e8 23 exf5 gxf5 24 f2? xf4 25 gxf4 g6+ (Now 26 h2 xf3! 27 xf3

g4+ wins.) 26 h1 g4 27 g1 f6 28 g2 ae8 29 h3 h6 30 a2 xf4 31 b3 g6 32 xe8 xe8 33 e4

Black to play 33 … xe4! 34 fxe4 xe4 (The threats include 35 … e3! 36 xe3 g4+! and mates.) 35 g2 xg2+ 36 xg2 e3 37 d5+ h7 38 f3 g4+ White resigns.

Chapter Seven: The Queen’s Gambit Family

The Queen’s Gambit Declined formation In the early 20th century, this pawn structure threatened to monopolize the middlegame. In the 1927 World Championship match, the Queen’s Gambit Declined occurred in 32 of 34 games. The QGD is not quite so popular today but the handling of this formation and its relatives – the Isolani, the Hanging Pawns, the Orthodox Exchange – remains essential knowledge for every player. Let’s begin with the isolated d-pawn:

Isolani

It arises both in this form and with colors reversed, as in the French Defense Tarrasch Variation, 1 e4 e6 2 d4 d5 3 d2 c5 4 exd5 exd5 followed by dxc5 or … cxd4. The same goes for the QGD Tarrasch Defense, 1 d4 d5 2 c4 e6 3 c3 c5 4 cxd5 exd5, after dxc5 or … cxd4. Aron Nimzovich said if it is allowed to advance (“the lust to expand”) the Isolani has “dynamic strength.” But blockaded, it is both a target for enemy pieces and a bottleneck for its own.

White to play Masters know the recipe for such positions: White should reposition minor pieces, e.g. on c3 goes to d4 via e2. But here it would be too clumsy to get the knight to d4 after, say, 15 xe5 xe5 16 f1 and 17 e2. In Aronian – Anand, Moscow 2011 White preferred 15 e1, preparing to attack the d5-pawn with c2-b3 or ac1/ b1-a2. He invited 15 … xd3 16 xd3 since ad1 and d4 would be an excellent blockade of the Isolani. But the lust to expand was illustrated by 15 … b8!.

xf3+ 16

xf3

White to play White has no good way of meeting the threat of 17 … d4! (18 exd4?? xf3). He ended up playing 17 f4 xf4 18 exf4 and lost the endgame after after 18 … d7! and … c5 and … d4. The energy released by pushing the Isolani is so great that the advance is often made as a sacrifice (Supplemental Game # 1). But when the Isolani is blockaded it’s a different animal. It’s not the pawn that is weak but the square in front of it, as in Bogolyubov – Rosenthal, St. Petersburg 1914: 1 d4 d5 2 f3 c5 3 c4 e6 4 cxd5 exd5 5 c3 c6 6 g3 f6 7 g2 e6 8 0-0 h6 9 b3 c8 10 b2 cxd4.

White to play Black exchanged on d4 because he expected to win control of d4 after 11 xd4 (perhaps by 11 … xd4 12 xd4 c5 – although 13 a4+

would be a problem). What he didn’t appreciate was the power of 11 b5! followed by occupying d4 with a knight. On d4 the knight will restrict Black’s bishop (no … f5, for example) and the Isolani inhibits his efforts to drive the knight away. After 11 … a5 12 bxd4 xd4 13 xd4 e7 White cashed in one positional advantage for another, 14 xe6 fxe6 15 d3 0-0 16 e4!.

Black to play Black will end up with either an isolated e6-pawn (after 16 … dxe4) or a passive position bottled up by the Isolani (after 16 … fd8 17 exd5 exd5 18 fe1). He chose the former and lost a textbook endgame: 16 … dxe4 17 xe4 xe4 18 xe4 d5 19 g4 f6 20 xf6 xf6 21 ad1 f5 22 d4! c5 23 xc5 xc5 24 c1 ff5 25 fd1 xc1 26 xc1 f7 27 c8+ f8 28 xf8+ xf8 29 g2 e7 30 f3 d6 31 e4 b6 32 f4 g6 33 h3 h5 34 d4 (zugzwang) 34 … d7 35 e5 e7 36 b4 b5 37 g4 hxg4 38 hxg4 f7 39 d4 Resigns. See also Supplemental Game # 2. It might seem that only when the d-pawn can be liquidated – by d4-d5 in the case of White – is the Isolani a good formation and when it can’t be liquidated it’s a bad one. Wrong. The Isolani often provides the foundation for powerful kingside attacks or queenside positional squeezes. It supports outposts at c5 and e5 which can be much more useful than Black’s use of d5. Here’s a typical case:

1 d4 d5 2 c4 e6 3 c3 f6 4 f3 c5 5 e3 c6 6 cxd5 xd5 7 d3 cxd4 8 exd4 e7 9 0-0 0-0 10 e1 f6 11 g5 b4 12 b1 b6 13 e5 b7 14 e3!.

Black to play Take the queens off the board and the d-pawn becomes the weakest target in an endgame that favors Black. But as Siegbert Tarrasch put it, “before the endgame the gods have placed the middlegame.” White has a serious attack and threatens 15 xf6 xf6 16 xh7+! xh7 17 h5+ and h3 etc. He can meet 14 … h6 15 xf6 xf6 with a promising 16 a3 and 17 c2. Black’s best may be 14 … g6 but that makes g6 a target. How quickly the attack intensifies after 15 g3 is shown by a few examples: One went 15 … e8 16 h4 bd5 17 h5! since 17 … xh5 allows 18 xh5! gxh5 19 xe7+ and mates. Instead, 17 … xc3 18 bxc3 g7 was played but Black resigned after 19 hxg6 hxg6 20 xg6!. Another game was very similar after 15 … bd5 16 h6 xc3 17 bxc3 e8 and then 18 h4 f8 19 g5 e7 20 h5!. White would have the edge after 20 … e4 21 hxg6 hxg6 22 xf7! xf7 23 f3+ or 21 … fxg6 22 xe4 xe4 23 xe7 xe7 24 g4 and h4/ g4. And a third game, Keene – Miles, Hastings 1975-76, ended dramatically with 15 … c8 16 h6 e8 17 a3 c6 18 xg6! hxg6 19 xg6! fxg6 20 b1! e5 (else 21 xg6+ mates) 21 dxe5 e4 22 xe4 h7 23 f6+ xf6 24 xg6+ h8 25 g7+ and mates.

Yes, Black can defend better with his own sacrifice, 15 … c6 16 h6 xd4!. But in general an Isolani-based attack remains dangerous.

The Hanging Pawns The hanging pawns hang because they are vulnerable on half-open files. Mikhail Botvinnik said that when both are on the fourth rank they are neither essentially good nor bad. But Nimzovich took a pragmatic tack: It is harder to be the player with the hangers, he said. They should be advanced eventually but in practice they are often pushed too soon.

The hanging pawns Nimzo concluded, “Never let yourself be drawn into a dead blockaded position.” For example, if White pawns are at c3 and d4, a Black blockade based on c4 is likely to become permanent. With colors reversed, an example of this was Rubinstein – Salwe, Lodz 1908: 1 d4 d5 2 c4 e6 3 c3 c5 4 cxd5 exd5 5 f3 f6 6 g3 c6 7 g2 cxd4? 8 xd4 b6 9 xc6! bxc6 10 0-0 e7 11 a4 b5 12 e3! 0-0 13 c1 g4 14 f3 e6.

White to play Black’s seventh and eighth moves were designed to break White’s control of d4 before it could be reinforced with other pieces. White correctly reacted by converting the Isolani into the hanging pawns. But Black is ready to strengthen them with 15 … d7 and 16 … c5. White stopped that with 15 c5! fe8 16 f2! d7 17 xe7 xe7 and was able to maintain the blockade on c5 with 18 d4! ee8 19 f1! ec8 20 e3 b7 21 c5 xc5 22 xc5 c7 23 fc2.

Black to play This is what bad hanging pawns look like. The only changes in the structure would come via e3-e4 or b3-b4-b5. Then a trade on c6 or d5 would make the remaining pawn an isolated weakie. To make matters worse, Black has only one minor piece and it’s a bad bishop. He could have begun restraint with 23 … a5!. But he lost a pawn after 23 … b6? 24 b4! a6 25 a5!, e.g. 25 … xd4 26 exd4 c8 27

xd5. But if the hangers are together on the fourth rank, they can often be turned into a single isolated pawn that is an offensive strength, not a blockaded weakness. In Karpov – Anand, Leon 1999,

White to play White had just retreated the knight from f3 because he intended h3f4. Black’s reply (… c6) forces him to make a decision about d4 and the one he chose was 17 dxc5. Black had a choice between hanging pawns or an Isolani. He chose 17 … d4! 18 d2 xc5!. This is better than 18 … bxc5 for two reasons: First, White could establish a nice position after 18 … bxc5? 19 h3 followed by e4/ f4. That stops the d4-pawn from advancing safely. He would jump at a chance to meet … b4? with xb4 because that would give him a good N-vs.-bad B matchup. But after 18 … xc5! 19 h3 fe8! White’s e-pawn is threatened. Then 20 f4 would kill his f4 plan and 20 f4 invites 20 … b4. The second reason for 18 … xc5! is that the d-pawn can advance and make d4 available for Black’s pieces. Anatoly Karpov himself, playing Black, had illustrated that in a previous game that went 19 e4 fe8 20 f4 ad8 21 f3 d3! 22 g5 d5 23 ac1 d4 with advantage. In this game, as White he played 19 f4.

Black to play But 19 … d3! was strong in view of 20 xd3? ad8 21 e2 22 f3 d4. White had to settle for 20 e4 and was outplayed:

d5+

20 … ad8 21 f3 d4! 22 xd4 xd4 23 f3 fd8 24 f5 d5 25 ad1 a5 26 xd5 4xd5 27 e4?! e8 28 c3 f6! 29 e6 xf5 30 xd3 f2+ 31 h3 xa2 and so on. One final point. Another way of looking at hanging pawns at c3 and d4 is that they are an Isolani plus a c-pawn. Since the Isolani can be the bedrock of a kingside attack, there is no urgency in trying to push the cpawn. Supplemental Game # 4 is an illustration of that kind of attack.

Fixing the Pair When the hanging pawns both reach the fourth rank the other player wants to ‘fix’ them in a stationary, blockaded position. That usually means attacking them with his own b-pawn or e-pawn. This was the case in two Gligoric – Keres games that occurred in a 1958 team match: 1 d4 f6 2 c4 e6 3 c3 b4 4 e3 0-0 5 d3 c5 6 f3 b6 7 0-0 b7 8 d2 cxd4 9 exd4 d5 10 cxd5 xc3 11 bxc3 xd5!? 12 c4 d6 13 c3 bd7 14 e1 ac8 15 h3 fd8 16 e3.

Black to play Can Black do anything active before White completes development and launches an attack? Well, he would like to play 16 … b5 to exploit the pin on the c-file (17 cxb5? d5 and a capture on e3 or c3). The positional point is that even though 17 c5 makes the c-pawn a protected passer, it also allows a firm blockade (17 … c6 and … d5). For a good illustration of the … b5 strategy see Supplemental Game # 3. But there’s a tactical problem here: 16 … b5? allows 17 a5!. That attacks a rook and frees the pinned c-pawn to capture on b5. So with nothing immediate to do, Black chose to pass with the quiet 16 … h6 instead. White erred with 17 b3 and that allowed 17 … xf3! (since 18

xf3 is no longer possible) and 18 xf3 e5!.

White to play This seems wrong: Black now has two knights facing two bishops and yet he opens up the center. But his threats of 19 … exd4 and 19 … e4 neutralized the pawns and forced White to give up one of the bishops. White doesn’t have time to create a protected passed pawn (19 d5? e4). He opted for 19 dxe5 xe5 20 xe5 xe5 and offered a draw after 21 b1. Curiously, Black played worse when the position after 16 e3 recurred in this match. Instead of 16 … h6 he tried to reposition his KN on f4. But after he played 16 … h5? White had a perfect opportunity for 17 d5!.

Black to play White unleashes his bishops and works with ideas such as d4 and

g5 – as well as 18 dxe6 fxe6?? 19 xh7+ and xd6. Black tried 17 … c5 18 g5! g6 but there was too much to defend after 19 e2 g7 20 d4! f8 21 h4 h5 22 g4! and he even allowed mate, 22 … f5 23 xe6! 2xe6 24 dxe6 e8 25 xh5 h6 26 f6 f4? 27 f7 mate. Pressure on the hangers is particularly effective when it can be coordinated with threats to other weaknesses and with the flanking e2e4 and b2-b4. An example is Larsen – Radulov, Leningrad 1973: 1 c4 f6 2 f3 e6 3 b3 e7 4 b2 0-0 5 c3 d5 6 e3 b6 7 d4 b7 8 d3 c5 9 0-0 bd7 10 e2 e4?! 11 cxd5 exd5 12 fd1 xc3 13 xc3 c8 14 ac1 c7 15 dxc5! bxc5 16 c2 g6?!.

White to play White’s bishops aim at the kingside but the flanking 17 e4! targets d5. The threat is 18 exd5 xd5 19 xg6! followed by xh7 mate or xd5. White also works with a pin on the c-file (17 … d4 18 xd4!). Black found 17 … a8, which defends d5 and threatened 18 … dxe4. But White combined the mating and positional ideas with 18 d2!, e.g. 18 … dxe4 19 h6 f6 (18 … f6 and 19 … f6 lose to 20 g5!) 20 g5 e8 21 xh7! and wins, as does 18 … fc8 19 exd5 xd5 20 xg6! xf3 21 h6!. Black tried to pacify the c3-g7 diagonal with 18 … f6 and 19 f4 d8 20 a5 b6. Everything seems under control … until you see 21 b4!.

Black to play Thanks to the multiple pins (21 … c4 22 xc4!) there was no defense. Black resigned after 21 … e7 22 exd5 d6 23 h6 cxb4 24 g5 f5 25 e6 d7 26 xc7. But see Supplemental Game # 5 for a case of the e3-e4 idea failing because of … c4!.

The Orthodox Exchange Formation This pawn structure arises out of the QGD, the Nimzo-Indian Defense and related openings when White plays cxd5 and Black recaptures with his e6-pawn. It also occurs with colors reversed in a Caro-Kann Defense line (1 e4 c6 2 d5 d5 3 exd5 cxd5) with the addition of c2-c3 and … e6.

The Orthodox Exchange formation This looks like a chain in which the White pawn at c5 and the Black one at e4 have been mysteriously removed. Unlike a chain, the center can be opened easily. But neither player is eager to do it. Why? Because if White opts for e3-e4/… dxe4, he is left with an isolated pawn under less favorable circumstances than usual. Similarly, Black gets an Isolani after … c5 and dxc5. So instead of changing the center, the two players usually focus on the wings. Black can work on the kingside with pieces (… d6/… g4/… e8-e6-g6 or -h6) rather than with pawns. But White usually needs pawn action, such as b2-b4-b5, to create a target at b7, c6 or d5. This strategy is known as the Minority Attack. Reshevsky – Myagmasuren, Sousse 1967: 1 d4 e6 2 c4 d5 3 c3 f6 4 cxd5 exd5 5 g5 e7 6 e3 0-0 7 d3 c6 8 c2 bd7 9 f3 e8 10 0-0 f8 11 xf6 xf6 12 b4! g4 13 d2 c8 14 f5 xf5 15 xf5 g6 16 d3 d6 17 fb1 g7 18 a4

d7 19 a2 e6 20 c2 ce8 21 b3 f6 22 h3.

Black to play If White can play b4-b5xc6 in the Orthodox Exchange formation, Black will end up with either a weak, backward c-pawn (after … bxc6) or a weak, isolated d-pawn (after he recaptures on c6 with a piece). In this case Black should avoid 22 … xb4? 23 c5. He would like active counterplay such as … f5-f4 but 22 … h5 23 b5! f5 24 a5 e7 25 bxc6 bxc6 26 b7 is much too slow. Instead, he chose 22 … b6!?. He hopes to escape into a good version of the hanging pawns (23 b5? c5! 24 dxc5 bxc5). But White prepared for the break with 23 c1 h6 24 1e2 h5 25 b5!. Then Black couldn’t play 25 … c5? because it costs the d-pawn after 26 dxc5. So he played 25 … d7 26 bxc6 xc6 27 b5.

Black to play

Now we can appreciate why the Minority Attack makes sense. The isolated d-pawn is fatally weak, e.g. 27 … f6 28 xd5! xc2 29 xf6+. Black desperately relied on tactics, a pin on the open file, 27 … ec8 28 bc1 8c7. But 29 g4! was a killer since a move of the Black knight allows 30 xd5! and f6+ or xc6. Black resigned soon after 29 … a6 30 xa6 f6 31 xd5! and xc6. Black does have defenses to the Minority Attack, as we’ll see. But a piece attack on the kingside is preferable. Even if White’s king is safely on the other wing, Black should try to exploit ‘his’ side of the board.

Black to play White castled queenside, in Kasparian – Aronin, Moscow 1952, to deny Black the usual attack. He can pressure b7 with c5 and d3-b3. Black should focus on the kingside, with 22 … g4!, threatening … xg2. If 23 f3, the e3-pawn is vulnerable (23 … f5). Play went 23 g3 f3! 24 c5 e4 and then 25 xe4 xe4 26 he1 ae8 27 d2.

Black to play This reveals a feature of the Exchange Orthodox often overlooked: There are simply more squares on Black’s side of the pawn spine. That counts for something, thanks to the weakness he incurred at f3. Black didn’t appreciate how well he stood and the game was soon drawn. But he had a strong plan – his own minority attack. With 27 … h5!! followed by … f6 and … g5 he would force White into hxg5. Then … fxg5 and … h4 would become a powerful passed pawn. If kingside prospects are slim, Black’s active play options are limited to: (a) looking for an opportunity for … c5, and (b) stopping the Minority Attack with … b5. The first option is risky because it may isolate the d5-pawn after dxc5. And (b) is risky since the c6-pawn becomes backward on a half open file. But it can pay off big, as in Fairhurst – Smyslov, Hastings 1954/55. 1 d4 f6 2 c4 e6 3 c3 b4 4 e3 0-0 5 e2 d5 6 a3 e7 7 g3 b6 8 cxd5 exd5 9 e2 b7 10 f5 e8 11 xe7+ xe7 12 0-0 bd7 13 b4 c6 14 d2

Black to play White is not ready for 15 b5 because that would be a splendid time for the (a) option. After 15 … c5! Black would get either a good version of an Isolani (16 dxc5 xc5 and … e4) or of hangers (16 … bxc5). Nevertheless, Black chose 14 … a6! Had White guessed his thinking he would have considered 15 a4. It is tactically justified by 15 … xb4? 16 xd5! with a discovered attack on the queen. And it is based positionally on answering 15 … b5 with 16 a5!. Why does that matter? The game revealed two reasons. After 15 b3? there followed 15 … b5! 16 ae1 b6!. Black’s knight is headed for c4. It would be harder to get there if White had played a4-a5!. The second reason White should regret the lack of a4-a5 was that after he played 17 c1, Black took aim at the b-pawn with 17 … a5!. Then White needed counterplay and that means e3-e4, supported by 18 d3. But he tried to build a pawn center instead with 18 f3?. The difference was 18 … axb4 19 axb4 c8! 20 d3 e6 21 b1 c4 22 e4 a7!.

White to play White is losing a pawn and resigned shortly after 23 fxe4 g4! 25 xc4 bxc4 26 g3 xd4+.

e2 dxe4 24

If the Minority Attack were White’s only weapon, the Exchange Orthodox formation would have lost its luster in the 1930s. But he has other plans. He can attack on the kingside with e5 and f2-f4, as Harry Pillsbury did so effectively. Or he can change the structure in a more radical way, with f2-f3 and e3-e4, as in Zsu. Polgar – Boensch, Dortmund 1990: 1 d4 f6 2 c4 e6 3 c3 d5 4 cxd5 exd5 5 g5 c6 6 e3 e7 7 c2 g4?! 8 ge2 xe2 9 xe2 bd7 10 0-0 0-0 11 d3 e8.

White to play In the Exchange Orthodox Black lacks a naturally good square for his QB. So he made the common sense decision to get rid of it at move

eight. But that gave White an incentive to look for something more ambitious than b2-b4-b5, something that would benefit her two bishops. She found it in 12 f3! with the idea of ae1 and e3-e4-e5!. When White plays f2-f3 in this structure – and in several others as well – Black should consider opening the diagonal leading to White’s king with … c5. But here that is outweighed by the easily-attacked Isolani it creates, 12 … c5 13 dxc5 xc5 (13 … xc5 14 xd5) 14 b5 and 15 ad1. Therefore Black had to rely on quiet moves: 12 … f8 13 h4 a6 14 ad1 g6 15 f2. There was still no chance for … c5 (15 … c5? 16 dxc5 xc5 17 xg6 and 18 xd5). Black played 15 … d6, seeking kingside play after 16 e4! dxe4 17 fxe4 g4. But the center pawns were simply too powerful – 18 e5! c7 19 c4 e7 20 h1 xf2+ 21 xf2 d7 22 e4.

Black to play Thanks to the pawn structure, each of White’s pieces is better than its Black counterpart. The final stage was an attack aimed at f7 and h7: 22 … f8 23 f3 e8 24 h5 h8 25 d3! h6 26 df3 and the threat of 27 xg6! led to 26 … xe5?! 27 dxe5 xe5 28 xf7! xf7 29 xf7 Resigns.

Supplemental Games # 1 The liberating force of d4-d5: Petrosian – Balashov, Moscow 1974 – 1 c4 f6 2 c3 e6 3 d4 b4 4 e3 c5 5 d3 d5 6 f3 0-0 7 0-0 dxc4 8 xc4 c6 9 d3 cxd4 10 exd4 e7 (… xd4 here or later loses to xd4 and xh7+.) 11 e1 b6 12 a3 b7 13 c2 c8 14 d3 e8? (Black needed 14 … g6 to anticipate xh7 mate.)

White to play 15 d5! exd5 16 g5 (It’s too late for 16 … g6 in view of 17 xe7! and 18 xf6 or 18 xd5.) 16 … e4 17 xe4 dxe4 18 xe4 g6 19 h4 c7 20 b3! h5 (Or 20 … f8 21 f4 and g5!.) 21 e4! g7

White to play

22 xf7! xf7 23 h6! d6 (Another GM game ended with 23 … d6 24 g5+ f6 25 h7+! and Black resigned before 25 … f7 26 d5+ or 25 … xh7 26 f3+.) 24 c4+ f6 25 ad1 d4 26 xd4+ xd4 27 xd4 Resigns. # 2 The Isolani as endgame liability: Regedzinsky – Rubinstein, Lodz 1917 – 1 d4 d5 2 f3 f6 3 c4 e6 4 g5 bd7 5 c3 e7 6 e3 0-0 7 d3 dxc4 8 xc4 a6 9 0-0 b5 10 d3 b7 11 e2 c5 12 ad1 cxd4 13 exd4 b6

White to play 14 e4? (Attack with b1/ e5/ d3 is called for, not exchanges.) 14 … xe4 15 xe7 xe7 16 xe4 fd8 17 d3 xe4 18 xe4 ac8 19 fd1 d5 20 3d2 f6 21 e3 b7 22 h3 h6 23 e2 d5 24 b3 d6 25 c1 d5 26 d2

Black to play

26 … f4! 27 c2 xd2 28 exd2 xc2 29 xc2 b4! 30 b2 c8 31 f1 c1+ 32 e2 a1 33 d2 xa2 34 xa2 xa2 35 e5 b4 36 d7 f6 37 g3 f7 38 b6? e7 39 c3 a5 White resigns (before … d8-c7). # 3 Black blockades hanging pawns well: Ojanen – Trifunovic, Mar del Plata 1953 – 1 d4 f6 2 c4 e6 3 c3 b4 4 e3 0-0 5 f3 d5 6 a3 xc3+ 7 bxc3 c5 8 cxd5 xd5 9 c4 f6 10 c2?! cxd4 11 exd4 d7 12 e2

Black to play 12 … b5! (Based on 13 cxb5 xb5 14 xb5 a5+ and … xb5, when White has a bad bishop and a bad Isolani.) 13 c5 c6 14 0-0 e4 15 c3 c6! (So that 16 xb5 xf3 and 17 … xd4 favors Black.) 16 e3 d5 17 c1 b8

White to play

(A graphic example of well-blockaded hanging pawns.) 18 e5 c7 19 xc6 xc6 20 d2? (Even after 20 f3! Black is still much better.) 20 … xg2! 21 fe1 (Or 21 xg2 xe3+ and mates.) 21 … e7 22 f4 b7 23 d6 d5 24 c3 c8 25 xe7 xe7 26 eb1

Black to play 26 … e5! (The queen heads to g6.) 27 xb5 g6+ 28 g3 e4! 29 f3 xd4+ 30 f2 xf2+ 31 xf2 xc5 32 d3 g6 White resigns. # 4 The hangers provide a basis for attack: Hansen – Ki. Georgiev, Kiljava 1984 – 1 d4 d5 2 c4 e6 3 c3 f6 4 f3 c5 5 cxd5 xd5 6 e3 c6 7 c4 cxd4 8 exd4 e7 9 0-0 0-0 10 e1 xc3 11 bxc3 b6 12 d3 b7 13 h4! (This is based on 13 … xh4 14 xh4 xh4 15 e3 with kingside chances.) 13 … f6 14 g5 g6 15 g4 h5 16 g3 e7 17 a3

Black to play

17 … c8? (Necessary was 17 … d7, although 18 ad1 ac8 19 d5! is dangerous, e.g. 19 … exd5 20 f4 or 19 … xd5 20 c4.) 18 xe6! fxe6 19 xe6 (The threat is 20 xe7 xe7 21 xg6+ and mates.) 19 … c7 20 ae1 f7 21 xg6 d7 (21 … xg6 22 e8+) 22 xf7+ xf7 23 xf6+! xf6 24 e5+ f7 25 e6+ f8 26 f6+ Resigns. # 5 The hanging pawns advance despite e3-e4: Donner – Darga, Krefeld 1969 – 1 d4 d5 2 c4 e6 3 c3 e7 4 f4 f6 5 e3 0-0 6 f3 b6 7 cxd5 xd5 8 xd5 exd5 9 d3 c5 10 dxc5 bxc5 11 0-0 c6 12 e5 xe5 13 xe5 e6 14 c2 g6 15 b3 c8 (White should hold a small edge after 16 b2.) 16 e4?

Black to play 16 … c4! 17 bxc4 dxc4 18 e2 c3! (No chance for the blockading c3! now, since 19 xc3 allows a winning pin, 19 … c7 20 ac1 f6.) 19 d3 d6 20 xd6 xd6 21 f4? (There were drawing chances after 21 fd1!.) 21 … fd8 22 e2? (Also lost was 22 ad1 c5+ 23 h1 g4.) c5+ and White resigned before 23 … d2.

Chapter Eight: The Panov Formation

The Panov formation This semi-closed center has been popping up in games since the QGD first became popular. But it was only with the introduction of Vasily Panov’s system in the Caro-Kann Defense (1 e4 c6 2 d4 d5 3 exd5 cxd5 4 c4 and a later c4-c5) that the formation became popular. It’s become a standard feature of openings as varied as the Alekhine’s and NimzoIndian Defenses, and in colors reversed versions like the Colle Opening. The Panov formation favors White slightly. He has a protected, semipassed c-pawn and controls valuable outposts at e5 and d6. He can create a queenside pawn roller with b2-b4-b5 and a passed pawn after c5-c6. He can exert pressure on the a4-e8 diagonal with b5/ e5/ a4. But Black has two ways to change the structure, with … e5 and/or … b6. He also has a more vulnerable target, at d4, than White does. Panov introduced his idea in Panov – Mudrov, Moscow 1929: 1 e4 c6 2 d4 d5 3 exd5 cxd5 4 c4 c5!.

f6 5

c3

c6 6

f3 f5? 7

Black to play After 7 … e6 8 b5! it was clear that Black’s bishop is misplaced. It cannot go to d7 to meet threats such as 9 e5 c7 10 a4 c8 11 xa7. Black withstood the pressure for a while, 8 … d7 9 f4 e7 10 h3 0-0 11 0-0 a6 12 a4 c8 13 e2 e8 14 fe1. But he couldn’t play natural moves such as 14 … f6 (15 xc6 xc6 16 xd5 or 15 … bxc6 16 xa6). The situation became critical after 14 … f8 15 ad1 g6 16 h2 h4.

White to play White’s threats on the a4-e8 diagonal were too great after 17 e5! and he won soon after 17 … f8 18 xc6 bxc6 19 xa6 g5 20 g3 e5 21 xc6.

Black cannot afford passive play in the Panov, as Supplemental Game # 1 shows. He should strike back in one of three methods – or a mixture of two or even all three: (a) A pawn break in the center, … e5, (b) A pawn break on the queenside, … b6, or (c) Piece play with … f6/… f5 or … e4. The first and third options are standard devices in a reversed form of the Panov. It arises in the Tarrasch Variation of the QGD, 1 d4 d5 2 c4 e6 3 c3 c5 4 cxd5 exd5 5 f3 c6 6 g3 c4. After 7 g2 b4 8 0-0 ge7 White’s two best ideas are 9 e4 and 9 e5. The aim is to destabilize Black’s center before he can complete development with … f5 and … xc3/… e4. White isn’t taking a risk with 9 e4 since 9 … xc3 10 bxc3 dxe4 favors him after 11 g5 or 11 d2. A simpler version is 1 e4 c6 2 d4 d5 3 exd5 cxd5 4 c4 f6 5 c5?! which allows 5 … e5!. White, who is behind in development, should keep some center control with 6 c3 exd4 7 xd4 and 7 … c6 8 b5. But suppose 6 dxe5 and 6 … e4 are played:

White to play White is temporarily up a pawn but the one at c5, which was intended to be his biggest middlegame asset, is doomed, e.g. 7 b4 a5!.

In Mariotti – Pfleger, Olot 1972 White tried to relieve the tension with 7 d3 and xe4. But following 7 … a5+ 8 d2 xc5 9 xe4 dxe4 he got a very poor game – 10 c3 e3! 11 fxe3 xe3+ 12 e2 c6 13 d2 e4 14 0-0 c5+ 15 h1 0-0. See also Supplemental Game # 2 for … e5 as a gambit. The … e5/dxe5 break is double-edged because it creates a wonderful outpost for White at d4. If he can reinforce the c5-pawn with b2-b4, he has a positional plus, e.g. Botvinnik – Konstantinopolsky, Sverdlovsk 1943: 1 e4 c6 2 d4 d5 3 exd5 cxd5 4 c4 f6 5 c3 e6 6 f3 e7 7 g5 0-0 8 c1 c6 9 c5! e4 10 xe7 xe7 11 e2.

Black to play Black needs to change the center. But he can’t play 11 … e5 12 xd5. And 11 … xc3 12 xc3 e5 favors White after 13 dxe5 xe5 14 xe5 and 15 0-0. But undermining d4 with piece play is a good option. For example, 11 … g5 12 0-0 xf3+ 13 xf3 f6! allows him time to strike at c5 from the flank, 14 e2 d7 15 d2 a5! 16 fd1 b6!. Instead, Black decided to adopt a Stonewall formation, 11 … d7 12 a3 f5. White used the respite to fight for e5 with 13 b5! g5 14 xc6 xf3+ 15 xf3 bxc6 16 f4!. Then Black should have tried 16 … e5! as a gambit. He would get

some compensation from 17 xe5 xe5 18 fxe5 f4! and … f3. But play went 16 … ae8? 17 0-0 e5 18 xe5 xe5 19 dxe5 xe5.

White to play The pawn structure and bad bishop make Black much worse. But White had to exert care with 20 f4! e7 21 fe1 fe8 22 xe7 xe7 23 f2 f7 24 d1!. If he had rushed into 24 e1? xe1 25 xe1, Black would free his d-pawn and be only slightly worse, 25 … d4! 26 e2 e6 27 xd4+ d5. White’s winning plan after 24 d1! could have included a e2-d4 maneuver and the building of a queenside passer with b2-b4 and a2-a4. Black lost after 24 … e8 25 d2 h6 26 e2 b8 27 e3 b3 28 d4 f6 29 a2 b8 30 b4 g5! 31 g3! gxf4 32 gxf4 a6 33 c3 g8 34 a4 g4 35 f2 e6? (35 … g1! and … c1) 36 b5!.

Black’s …b6 Option When Black challenges the head of the mini-chain with … b6 he wants White to capture on b6. That would isolate the d4-pawn and enable Black to exert pressure against the b-pawn. This is a common theme in the Nimzo-Indian Defense line that goes 1 d4 f6 2 c4 e6 3 c3 b4 4 e3 c5 5 e2 cxd4 6 exd4 d5 7 c5 and 7 … e4! 8 d2 xd2 9 xd2.

Black to play Black’s bishops are not enough to equalize if he can’t open lines further. For example, 9 … c6 10 a3 a5 11 b4 c7 12 f4! prevents a quick … e5. In Simutowe – Nsubuga, New York 2001, Black tried to catch up with 12 … 0-0 13 g3 b6. But 14 g2 bxc5 15 dxc5! b8 16 b1 put White firmly in control (16 … a5? 17 b5!). After the delayed break 16 … e7 17 d4 e5 18 fxe5 xe5 he was winning, 19 0-0 g6? 20 xd5! xd5 21 c6! c7 22 xd5. Black can do better in the last diagram with 9 … a5! to anticipate b2b4. Typical play runs 10 a3 xc3 11 xc3 a4! 12 d3.

Black to play Black will take en passant if White engineers b2-b4. Without further support of c5, White has no edge, e.g. 12 … d7 13 0-0 b6 14 cxb6 xb6 and Black is fine (15 fe1 0-0 and … c8). The most interesting battles occur when Black employs both the … b6 and … e5 plans. An example arises in the Caro-Kann after 1 e4 c6 2 d4 d5 3 exd5 cxd5 4 c4 f6 5 c3 e6 6 f3 e7 7 c5 0-0 8 d3 b6 9 b4! a5!.

White to play Black cannot delay. After, say, 9 … d7? 10 e3 a5 11 b5! or 10 … g4 11 0-0 a5 12 a4, White’s queenside is solid. After 9 … a5 it appears Black has won the queenside battle (10 a3? axb4 11 axb4?? xa1). But White can reply 10 a4!, threatening xb6. That can lead to a liquidation, 10 … bd7 11 a3 axb4 12 axb4 bxc5

13 bxc5. Then Black needs to act before White consolidates with 0-0, f4 and e2 into a typically favorable Panov middlegame. Best is 13 … e5!, based on 14 dxe5 e4 15 c2 b7 and … dxc5 with an edge. The crucial continuation is 14 xe5 xe5 15 dxe5 e4:

White to play In a 1959 Soviet game, Estrin – Liberzon, White’s king never made it to safety – 16 b2? a5+! 17 f1 d7 18 c2 b5+ 19 e2 xa4! 20 xa4 d2+ 21 e1 xb2 22 d1 xa4 and he resigned. Black stands well after 16 0-0 xc5 and 17 c2 a6 or 17 b5 d4 18 f4 a5. And 16 xe4 dxe4 17 b2 is also bad after 17 … xd1+ 18 xd1 d7. The bottom line remains: Black must change the structure or the Panov favors White.

Supplemental Games # 1 Black fatally delays … e5: Matulovic – Damjanovic, Ljubljana 1960 – 1 c4 e6 2 f3 f6 3 c3 c5 4 e3 e7 5 d4 cxd4 6 exd4 d5 7 c5 c6 8 d3 0-0 9 0-0 b6 10 b5! d7

White to play 11 xc6 xc6 12 b4 e4 13 b2 a5 14 a3 f6 15 c2 (Since Black hasn’t created serious counterplay White’s edge grows.) 15 … bxc5 16 bxc5 b8 17 fb1 h6 18 c1! c7 19 e3 xc3 20 xc3 a4 21 d2! (Threatening to seize the open file with f4.) 21 … xb1+ 22 xb1 b8 23 b4! xb4 24 xb4

Black to play

24 … e5 (Otherwise White wins by maneuvering the knight to c3 and then b5-d6 or xa4.) 25 dxe5 xe5 26 d4! f6 27 xe5 fxe5 28 b6! xb6 29 cxb6 (Triumph of the c5-pawn.) 29 … e4 30 d4 b7 31 e6 a6 32 c5 c8 33 f1 f8 34 e2 e7 35 b7 xb7 36 xb7 Resigns. # 2 Black plays … e5 as a sacrifice: Seirawan – Sunye Neto, Dubai 1986 – 1 d4 f6 2 c4 e6 3 c3 b4 4 c2 0-0 5 g5 c5 6 e3 cxd4 7 exd4 d5 8 f3 c6 9 a3 xc3+ 10 xc3 e8 11 c5 h6 12 xf6 xf6 13 b5

Black to play 13 … e5! (Otherwise 14 xc6 favors White. Now 14 xe5 xd4! is unclear, e.g. 15 xe8 xe5+ 16 f1 g4.) 14 dxe5 d4! 15 0-0-0? (Not 15 xd4 xe5+ but 15 0-0 was better.) 15 … xe5! 16 c2 (16 xd4 g5+! and … xc5) 16 … f5 17 d3 xd3 18 xd3 d5 19 b1 f5 20 c1 xc5 21 xc5 xd3+ 22 c2 a6 23 e4 d8 24 h4 b6 25 d1 b3 26 d2 e6 27 xe6? fxe6 28 c2 f7 29 d3 e5 30 e4 e6 31 b4 d5 32 e1 b8! 33 g4 d7 34 g5 hxg5 35 hxg5 d6? (35 … b6, headed to c3 or d6, should win.) 36 d3 a5 37 c2 axb4 38 axb4 b5 39 c8 e7 40 g8 f7 41 d8 e7 42 g8 f7 43 d8 e7 Draw. # 3 Piece play plus the e-pawn break wins: Anastasian – Caruso, Paris 1999 – 1 d4 d5 2 c4 e6 3 c3 c5 4 cxd5 exd5 5 f3 c6 6 g3 c4 7 g2 b4 8 0-0 f6? (This knight belongs on e7.) 9 g5! e6 10 e5! 0-0 11 xf6! gxf6 12 xc6 bxc6

White to play 13 e4! dxe4 14 xe4 (Black has bad pawns and a bad king, e.g. 14 … d5 15 xd5 cxd5 16 f5! and h5.) 14 … a5 15 f3 xc3 16 bxc3 d5? 17 g4+ h8 18 f5! Resigns. # 4 White delays b2-b4 and is punished: Toran – Korchnoi, Uppsala 1956 – 1 e4 f6 2 e5 d5 3 d4 d6 4 f3 g6 5 e2 g7 6 0-0 0-0 7 c4 b6 8 exd6 cxd6 9 c3 c6 10 e3 g4 11 b3 d5! 11 c5

Black to play 11 … c8! (This knight is headed to f5 to attack d4. That is so strong White should immediately begin queenside action with 12 b4!, based on 12 … xb4 13 b3.) 13 h3? xf3 14 xf3 e6 15 b4 a6 16 b5? (Better is 16 b1.) 16 … axb5 17 xb5 8e7 18 g5 a5 19 xe7 xe7 20 d3 c6 21 a4 b6! 22 cxb6 xb6 (Black must win either the d-pawn or a-pawn.) 23 fd1 a5 24 c3 fa8 25 ac1

Black to play 25 … xd4! and Black won (26 axb5 xd4.).

xd4 xa4 or 26 xd4 xb5 27

Chapter Nine: Stonewalls and Other Prisons

The full Kingside Stonewall Stonewall formations are popular with players who prefer a relatively simple game with a closed center, little opening theory to master and easy-to-understand middlegame plans. The center pawns of each player are committed to squares of one color, so each side has a bad bishop he wants to trade for the ‘good’ enemy bishop. Each side has a natural outpost – e5 for White and e4 for Black in the Kingside Stonewall. Piece trades on those squares and those near them also have a great impact, as in Sultan Khan – Mattison, Prague 1931: 1 d4 f6 2 f3 e6 3 e3 b6 4 d3 b7 5 bd2 d5 6 e5 d6 7 f4 0-0 8 f3 fd7 9 h3 f5 10 df3 f6 11 d2 a6 12 g1.

Black to play A good way to defend against a Kingside Stonewall is to shut out enemy pieces by erecting your own wall of stone. Black did that with 8 … fd7 and 9 … f5. He followed up by trying to exchange his bad bishop. But 12 … xd3? 13 cxd3 was a Stonewall felony since White regains control of e4 and Black fails to get light-square play to compensate. Black’s desperate bid for counterchances led to 13 … e8 14 e2! bd7 15 h4 c5 16 c3 cxd4? 17 xd4 c5 18 g4! xe5 19 xe5 fxg4 20 xg4 xg4 21 xg4 g6 22 ag1 Resigns. He can’t stop 23 xg6+!. A case of a Queenside Stonewall was Albin – Janowski, Nuremberg 1896: 1 e4 e5 2 f3 c6 3 b5 f6 4 d3 d6 5 xc6+ bxc6 6 h3 e7 7 e2 0-0 8 c4 d7 9 g4 e8 10 c3 c5 11 e3 e6 12 d2.

Black to play White tried to lock the center so he could attack the king by means of c3-e2-g3-f5. But his exchange on c6 came back to haunt him when Black replied 12 … c5! 13 e2 c6! 14 g3 b8 15 f5 f8. White’s attack shouldn’t succeed because there is no kingside weakness and … d5 is coming fast. Black had the edge after 16 h4 d4 17 xd4 cxd4 18 h5 d5! and the wall was chipped away, 19 f1 dxc4 20 dxc4 f6 21 3h4 e6 22 b3 a5! and … a4. Some openings establish at least a partial Stonewall in the first moves, such as the Dutch Defense (1 d4 f5 2 f3 e6 3 g3 f6 4 g2 d5) and Sicilian (1 e4 c5 2 f3 d6 3 c3 e5). But often a player can build it later, as White did with 8 c4 and Black did with 12 … c5 in the last example. By far the biggest reason to build a kingside version is the ability to attack the king. Sadler – Ernst, Oslo 2011 began 1 d4 d5 2 c4 c6 3 c3 f6 4 e3 f5 5 cxd5 cxd5 6 b3! c8 7 f3 c6 and White played 8 e5 e6 9 f4!.

Black to play Black’s loss of time with his QB justifies White’s aggressive plan of d3, d2, 0-0, f3 and af1. If Black captures on e5, then fxe5 leaves White with more space and ammunition on the kingside. That’s what happened: 9 … e7 10 d3 0-0 11 0-0 d7 12 d2 dxe5 13 fxe5. Now on 13 … f5, White could open lines with g2-g4. An even better plan is 14 exf6! xf6 15 f3 or 15 c2 g6 16 e2 and f4. Instead, play continued 13 … d7 14 f3 b4 15 e2 a5 16 a3 a4 17 d1 c6 but the queenside action didn’t distract White from 18 d3, which threatens 19 xh7+ xh7 20 h3+ g8 21 h5 with a strong attack. Black defended with 18 … g6 19 e2 f6 20 exf6 xf6 21 af1 g7.

White to play White needed one more piece to break through and it shouldn’t be surprising that it was the bad-bishop-turned-good – 22 e1! xf3 23

xf3 f6 24 g3 a5? 25 c7 e7 26 d6 d8 27 c7 e7 28 d6 d8 29 h4! c6 30 h5 e8 31 h6+! and Black resigned in view of 30 … xh6 31 f8+! g7 32 xg7+ xg7 33 f8 mate. See also Supplemental Game # 1.

The Kingside Stonewall The primary issues in the full Kingside Stonewall or the partial version with a White pawn at f2 are: (a) The status of good and bad bishops. (b) How to recapture when knights are taken on an outpost square, and (c) How each player can open a wall edge, e.g. c2-c4xd5 and g2-g4xf5. The traditional view of the Dutch Defense is that if White can trade the dark-squared bishops only, he must be better. This is borne out by examples like Tarasov – Matsukevich, Kharkov 1956: 1 d4 f5 2 c4 e6 3 g3 f6 4 g2 e7 5 f3 0-0 6 0-0 d5 7 b3 c6 8 a3! bd7 9 c1 e4 10 xe7 xe7 11 a3 b6 12 c2 b7 13 a3 f7?.

White to play It’s natural with the player who sets up a Kingside wall to insist on attack and resist a queen trade. But here Black was in a bad middlegame after 14 cxd5 cxd5 15 ce1! since a White knight can occupy e5. White has no invasion square on the c-file but made progress nonetheless with 15 … a5 16 d3 fc8 17 fe5 xe5 18 xe5 e8

19 fc1. He won after Black tried too hard for counterplay: 19 … d8 20 e3 c3?! 21 f1 c7 22 b4 axb4 23 b3 a6 24 xb4 xa2? 25 xb6! xc1 26 xc1 c4 (26 … xc1 allows mate after 27 xe6+) 27 xe6+ f8 28 xf5+ g8 29 e6+ f8 30 xc4 e7 (Or 30 … dxc4 31 xc4 and g8+) 31 f5+ f7 32 xh7 Resigns. For most of the 20th century it was regarded as a mistake for Black to play … d6 in the Dutch Defense because that made it easier for White to trade off the right bishops with f4. It was believed that the exchange, … xf4/gxf4, strengthened White’s center at no expense. But by opening half of the g-file, White hands his opponent an opportunity to open the other half, with … g5!. See Supplemental Game # 2. And while bishops are very important in stonewall positions, their significance can be exaggerated, as in Kerek – Cs. Horvath, Cseppko 1997: 1 d4 d5 2 f3 e6 3 g3 c6 4 g2 d6 5 0-0 f5 6 c4 f6 7 c3 0-0 8 c2 bd7 9 b1 e4 10 b4 b5!? 11 cxb5 xc3 12 xc3 cxb5 13 g5 b6 14 fc1 b7 15 e3 fe8 16 f4 xf4 17 gxf4!? d6 18 e5 ac8 19 xd7 xd7.

White to play Black’s attack never got going and White could establish superiority

with 20 c5!. But he thought he could increase his edge by trading all four rooks and played 20 xc8+? xc8 21 c1. He misjudged 21 … c4!. Then a trade of rooks will create a protected, passed pawn that is more important than a bad bishop: 22 a3 c6 23 xc4 bxc4 24 b5 b6 25 a4 f7 26 c3 a6! 27 bxa6 xa6. Black was much better following 28 f3 c6 29 a5 a4 30 e3 b3! because 31 xb3 cxb3 32 d1 c4 would be a won endgame. White played 31 c1 c3 32 d1 b2 33 xb2 instead but was lost after 33 … cxb2 34 c2 e7 35 f3 d6 36 f2 c4 37 b1 c6 and … b5. When there is a capture on one of the outpost squares, the natural recapture is towards the center, e.g … xe5/fxe5. But there are advantages to dxe5. It preserves the prospect of f4-f5 and allows White to use a half-open d-file. This was the theme of Rubinstein – Cohn, Karlsbad 1911: 1 d4 d5 2 f3 e6 3 c4 c6 4 e3 d7 5 c3 f5 6 e5 gf6 7 f4 d6 8 e2 0-0 9 0-0 e4 10 xe4 fxe4 11 d2 xe5 12 dxe5!.

Black to play White’s winning chances would be slim after 12 fxe5? and the inevitable trade of heavy pieces on the f-file. But 12 dxe5! allowed him to expand after 12 … c5?! 13 b4 b6 14 c2 d7 15 a4 c8 16 a5 c7 17 c3 h4 18 d4 a6 19 b3 d8 20 c5. White has made his bad bishop work harder than Black’s good bishop.

Black could not stop a steady buildup on the kingside – 20 … f7 21 f2 e7 22 g3 h6 23 d6! g6 24 c5 f8 25 af1 g7 26 g4 cf8 27 d1 h6 28 h1 h8 29 g1 g8 30 f1 gf8 31 h4! xd6 32 cxd6. In view of h2, d1 and then g3-g4-g5 and h4-h5, Black became desperate with 32 … g5 and resigned after 33 fxg5 xf2 34 xf2 xh4+ 35 gxh4 xf2 36 d1 g7 37 g1.

The Queenside Stonewall Why should the queenside version of the Stonewall be any different from the kingside version? One answer is the size of the wings. Now there are two files on the queenside and three on the other side, a reversal of the Kingside Stonewall. This means there is more room for a break on White’s right. He can do it with f2-f4 or Black can with … f5. This has greater effect, at less risk, than g2-g4 and … g5 in the Kingside Stonewall. Similarly, there is less room to exploit on the queenside. After b2-b4 Black can play … b6 and White has to choose between bxc5 or b4-b5 followed by a4-a5xb6. Black, meanwhile, usually seeks play on the other wing. A typical situation arises after 1 e4 c5 2 f3 d6 3 b5+ d7 4 xd7+ xd7 5 c4 – preparing to create a Maroczy Bind with d2-d4 – and now 5 … e5!? 6 c3 g6 7 d3 g7 8 a3 c6 9 b1 ge7 10 b4 00.

White to play At some point one side or both will occupy a d-file outpost. After 11 d5 xd5 12 cxd5 d4 13 xd4 cxd4, for example, what matters most is the remaining minor pieces. White’s bishop is much better than Black’s. Take all the heavy pieces off the board and White is close to a won endgame, as in Supplemental Game # 3.

And if Black plays 12 … e7 rather than 12 … d4 White will create a protected passer with 13 bxc5 dxc5 that must favor him, at least slightly. Another natural continuation is 11 0-0 f5 12 d5. Then on 12 … h6 13 bxc5 dxc5 the chances are in balance because when White prepares to open a queenside line, 14 b3 b6 15 d2 Black has 15 … g5!.

White to play Black is ready for 16 … g4 and 17 … f4. White would like to create a target at e5 with 16 c3 followed by fe1 and/or b2. But Black obtains attacking chances with 16 … fxe4 17 dxe4 g6. In Torre – Radulov, Leningrad 1973 he tried for more with 17 … xf3!? 18 gxf3 g6 and got it after 19 a1 h4 20 h1 h3 21 g1 xf3 22 g2 f8 23 a4 cd4 because White failed to eliminate the knight on d4 (24 xd4!). The game ended with 24 xa7? e2! 25 xb6 h4 26 bg1 f3! 27 e6+ h7 28 g4 xg1 29 xg1 b8! 30 c3 xc3! White resigns. The value of the d5 outpost has been the focus in recent years of closed Sicilian Defense variations such as 1 e4 c5 2 c3 d6 3 ge2 e5 and 4 d5 followed by ec3! and c4. See Supplemental Game # 4. When both players capture enemy pieces on the d-file outposts, there are opportunities to create protected passed pawns, normally a huge asset. A remarkable example of that was Bobotsov – Gligoric, Belgrade

1961: 1 c4 g6 2 c3 g7 3 g3 c5 4 g2 c6 5 e4 d6 6 ge2 f5 7 d3 f6 8 0-0 0-0 9 b1 e5 10 d5 fxe4 11 dxe4 xd5 12 cxd5? d4 13 xd4

Black to play 13 … exd4! The exchanges created two-pawn majorities on opposing wings. Which counts more? The answer is Black’s since he can advance his more easily. In contrast, after 13 … cxd4? Black would have a protected passed d-pawn but his chances of winning would be much less. In the game White bet on a kingside attack, 14 f4 b5! 15 d2 a5 16 h6 xh6 17 xh6. Black just pushed pawns, 17 … c4 18 be1 b4 19 e5 c3!. The deciding factor is that White couldn’t make progress after 20 bxc3 bxc3 21 e6 e7 22 e4. But Black could: 22 … d3 23 c4 c2 24 d2 a6 25 c6 b5! 26 b6 ab8 27 xb8 xb8 28 e4 f6! 29 e1 (29 xd3 d4) 29 … d4 30 e7 b4! and White resigned.

Supplemental Games # 1 A slow-moving b2-b4-b5 loses to a Kingside Stonewall attack: Andric – Ivkov, Sarajevo 1951 – 1 d4 e6 2 c4 f5 3 f3 f6 4 g3 d5 5 g2 c6 6 0-0 d6 7 b3 e7 8 b2 0-0 9 bd2! (This gets both knights trained on e5.) 9 … d7! 10 e5 e8 11 df3 h5 12 d3 bd7 13 fe5 ad8 14 c2 h8 15 e3 g5 16 c5 c7 17 b4 g4! 18 xd7 xd7 19 c1 f6 20 a4 a6

White to play 21 e5? (White needs b1 and b4-b5 before Black breaks on the other wing with … e5 or … f4.) 21 … g7 22 a3 d7! 23 xd7 xd7 (Now 24 b3 allows 24 … df7 25 d3 g8 26 b5 axb5 27 axb5 f4.) 24 f4 gxf4 25 gxf4 g8 26 h1 g4 27 a2 dg7 28 d3?

Black to play

28 … xg2+! 29 xg2 xg2 30 e4 e2 White resigns. # 2 The f4/… xf4/gxf4 trade allows the g-file to open: Lavrov – Chigvintsev, Tomsk 2001 – 1 d4 e6 2 f3 f5 3 c4 f6 4 g3 d5 5 g2 c6 6 0-0 d6 7 f4 xf4 8 gxf4 0-0 9 c3 bd7 10 e5 h8 (Black envisions … g8/… g5.) 11 e3 e7 12 b1 xe5 13 fxe5 g4 14 h3 h6 15 f4 d7 16 e2 g8 17 h2 dxc4 18 xc4 h4! 19 e2

Black to play 19 … g5 20 e1 h5 21 e2 g4! (White has no defense to doubled Black rooks.) 22 h1 g7 23 g1 ag8 24 b4?! e8 25 e4 h4 26 e1

Black to play 26 … g3 27 exf5 g4! 28 e4 f2+ White resigns. # 3 White wins a Queenside Stonewall ending: Karpov – Ribli, Bath

1973 – 1 e4 c5 2 f3 d6 3 b5+ d7 4 xd7+ xd7 5 c4 e5 6 c3 c6 7 0-0 ge7 8 d5 xd5 9 cxd5 d4?! (Since Black can only hope for a draw now, 9 … d8 or 9 … b4 are better.) 10 xd4 cxd4 11 d3 e7 12 b3 0-0 13 f4 ac8 14 d2 f6 15 h3 c7 16 f2 fc8 17 af1 c2 18 g4! a6 19 a4 8c5 20 b4! xf2 21 xf2 c7 22 d2 c8 23 e2 c5

White to play (White takes his time deciding between g4-g5 or b2-b4-b5.) 24 f5 d8 25 a3 f7 26 b1 e7 27 d1 d7 28 b4 c8 29 a2 b6 30 e1 f7 31 a1 d8 32 d2 b6 33 a5! d8 34 a4! xa4+ 35 xa4 e8 36 h4 h6 37 a2 d7 38 e1 e7 39 g2 d8 40 d2 e7 41 g3 h8 42 c2 c8+ 43 b2 h8 44 g2 c8 45 g1 h8 46 c2 c8+ 47 d1 h8 48 e1 c8 49 g2 h8 50 b2! g8 51 d2 d8 52 b5 axb5 53 xb5 c7 54 e2 b8 55 b4 c7?

White to play

56 a6! a7 57 axb7 b8 58 d2 xb7 59 xb7+ xb7 60 g5! hxg5 61 hxg5 d8 (Or 61 … fxg5 62 xg5 c8 63 e7 and f8.) 62 f3 c8 63 g4 d7 64 h5 e8 65 b4 fxg5 (Also lost is 65 … e7 66 gxf6 gxf6 67 g6 and e1-h4xf6.) 66 xd6 f6 67 b4 f7 68 d2 e7 69 xg5 a3 70 d8 d6 71 g5 Resigns. # 4 A partial Queenside Stonewall allows White a kingside attack: Kasparov – Piket, Zurich 2001 – 1 e4 c5 2 f3 c6 3 c3 e5 4 c4 e7 5 d3 d6 6 d2 f6 7 f1! a5? 8 e3 xc4 9 dxc4! (Stopping … d5 forever is the priority.) 9 … 0-0 10 f3 e6

White to play 11 g4! (With the center closed White has all the reason to attack with f5/ g1/g4-g5.) 11 … g6 12 h4 d7 13 g1 h8 14 a4 g8 15 b3 af8 16 b2 h5?! 17 g5 h7 18 g3 f6 19 f4 fxg5 20 fxg5 g7 21 0-0-0 a6 22 cd5 d8 23 d3 f7 24 f1 e8

White to play

25 f6 c6 (Not 25 … xf6 26 gxf6 xf6 27 xd6.) 26 fd1 c7 27 fd5 b8 28 a5 g8 29 c3 gf7 30 b2 f4? (Black is also losing after, say, 30 … h8 31 c7 xc7 32 xd6! and xe5+ but 30 … d7 puts up a better fight.) 31 e7+ Resigns.

Chapter Ten: The Nimzo-Gruenfeld Formation

The Nimzo-Gruenfeld formation Aron Nimzovich helped create two major openings, the Queen’s Indian and Nimzo-Indian Defenses, that can lead to this pawn formation (or the similar one with a White pawn at e3). In addition, the Gruenfeld Defense’s Exchange Variation can lead to this formation as does the Semi-Tarrasch line of the QGD. There’s even a Neo-Gruenfeld Defense line, 1 d4 f6 2 c4 g6 3 g3 g7 4 g2 d5 5 f3 0-0 6 0-0 dxc4 7 a3, in which it arises after 7 … c3! 8 bxc3 c5 and a push of the White epawn. The basic questions of this formation are: (a) Should Black trade pawns on d4 – and, if so, when? (b) Should White advance his e-pawn to e4 or just e3? (c) Can White profitably push his d-pawn? There are several reasons for Black to trade on d4. He creates a potential target there. He opens a file that he might control and occupy. And he increases his chances of creating a passed queenside pawn. All three factors were in play in Petrosian – Fischer, Candidates match

1971: 1 d4 f6 2 c4 e6 3 f3 d5 4 c3 c5 5 e3 c6 6 a3 e4 7 c2 xc3 8 bxc3 e7 9 b2 0-0 10 d3 h6 11 0-0 a5 12 d2 dxc4 13 xc4 xc4 14 xc4 b6 15 e4 b7 16 e2

Black to play It may appear odd that Black spends time (… e4xc3) to make a trade that strengthens White’s center. But Bobby Fischer said that in general the value of a pawn center declines as pieces are traded. Black often seeks piece trades in a Nimzo-Gruenfeld. See Supplemental Game # 1. In this game Black was in no rush to swap on d4. He inserted 16 … c8, with its threat of 17 … cxd4 18 cxd4 xe4! 19 xe4 xc4. After 17 b3 he replied 17 … b5!, based on 18 xb5 xe4, a favorable pawn swap. White concluded he had to attack the kingside and played 18 f4 b6 19 h1. This was a good time to decide about the c-pawn and the choice isn’t easy. Black would get a splendid position after 19 … c4 20 c2 a5 and … b4. But he chose 19 … cxd4! and then 20 cxd4 b4! 21 axb4 xb4 because his counterplay is much faster than White’s kingside initiative.

White to play Black has 22 … a5 and 23 … a6 among other ways to make progress. White tried to force matters with 22 d5?!. But 22 … c3! extinguished his attack. He lost after 23 xc3 xc3 24 c2?! exd5 25 e5 e3 26 d2 d4. When Black delays … cxd4 he hopes to make c3 a target, rather than d4. He takes the risk that White will push his c- and d-pawns and obtain a super-center. Korchnoi – Smirin, Biel 2002 shows us how this works: 1 d4 f6 2 c4 g6 3 c3 d5 4 f3 g7 5 cxd5 xd5 6 e4 xc3 7 bxc3 c5 8 e3 a5 9 d2 0-0 10 e2 g4 11 0-0 d8 12 g5? xe2 13 xe2 a6! 14 xa6 xa6 15 d5.

Black to play White is a move or two away from a favorable c3-c4!. That suggests Black should play 15 … e6, based on 16 dxe6? xd2 and 16 c4? xa1.

But White would get good play then from 16 ab1!. For example, 16 … b6? 17 c4! creates an excellent center. And 16 … exd5 17 xb7 dxe4 18 xe4 f5 allows a complex 19 h6!. Black chose 15 … c4! instead because stopping c3-c4 takes precedence. After 16 e3 ac8 White faced the prospect of being slowly outplayed (17 fc1 c5 18 xc5 xc5) or allowing a passed cpawn with 17 xa7 xc3. He chose the latter and Black’s passer proved decisive after 18 ab1 d7 19 e3 g7 20 b5 c3! 21 f4?! c2. See also Supplemental Game # 2 for another effectively delayed … c4.

The e4-e5 Option When White pushes his e-pawn he is looking to attack. This happens regardless of whether c-pawns have been traded. The critical question White asks himself is whether e4-e5 is worth ceding control of d5.

White to play Here 21 e5! is worth it because 21 … d5 22 e4! would threaten h7+ and leave Black with too many kingside weaknesses (22 … g6 23 h4). When 21 e5! was played in Azmaiparashvili – Schneider, Soviet Young Masters Tournament 1984, Black settled for 21 … d7. The best he could hope for after White replied 22 g4 f8 23 h4 cxd4 24 cxd4 was reaching an endgame. But he was slowly ground down, 24 … d7 25 c3 xc3 26 xc3 c8 27 b4! c2 28 a6! d8 29 e2 a4 30 d6 and so on. But White bets so much on an initiative after e4-e5 that Black may be able to equalize by exploiting d5 or through trades, as in Reshevsky – Fine, Hastings 1937: 1 d4 f6 2 c4 e6 3 f3 d5 4 c3 c5 5 cxd5 xd5 6 e4 xc3 7 bxc3 cxd4 8 cxd4 b4+ 9 d2 xd2+ 10 xd2 0-0 11 c4 c6 12 0-0 b6 13 fd1 b7 14 f4.

Black to play Black’s 14 … f6! is a common idea in similar situations. If White trades queens his attack is over, Black’s pawns are only slightly messed up and the endgame is at least equal. White recognized that and replied 15 e3 fd8 16 e5. This looks good in view of 16 … e7 17 h4 intending g5/ d3 or h4-h5-h6. But Black shot back 16 … h6!. Having committed himself to e4-e5 White didn’t like 17 xh6 gxh6 and … ac8/… a5 or … b4. Instead, White played to equalize, with 17 ac1 xe3 18 fxe3 and drew after 18 … ac8 19 d2 h6 20 b3 a5 21 c2 d5!. See Supplemental Game # 3 for another example of Black favorably provoking e4-e5.

White’s d4-d5 Option White usually posts a rook on the d-file in the Nimzo-Gruenfeld formation and may be able to push his d-pawn. There are two basic d4d5 scenarios. In the first White creates a passed d-pawn. In the second, he sacrifices a pawn, by meeting … exd5 with e4-e5!?. Let’s go back to the position, before 11 … example.

c6, in the previous

Black to play Black’s knight has a choice. On c6 it can go to a5 and drive off the well-posted White bishop. But then the knight is offside. On the other hand, … d7-f6 allows Black to watch both d5 and e4. But … d7 invites e4-e5, which leaves the knight without a good square if he’s played … b6. So if Black is going to post his bishop at b7 anyway, how about an immediate 11 … b6 ? The answer is 12 d5!, since 12 … exd5?? 13 xd5 loses a piece and 12 … a6 13 xa6 xa6 14 d6 is a strong passed pawn. That’s why the main line is 11 … c6 and 12 0-0 b6. Then White can get his rooks to optimum squares with 13 ad1! b7 14 fe1 e7.

White to play Black repositioned his knight but doesn’t have enough ammunition to fight in the center after 15 d5! and then 15 … exd5 16 exd5. Even when it is blockaded, that d-pawn makes White’s pieces superior to Black’s. The bankruptcy of Black’s strategy was revealed in a 1977 Candidates match game, Petrosian – Korchnoi, which went 16 … f5 17 e5 d6 18 c6!. This was tactically based on 18 … xc6 19 dxc6 xc4 20 f4!. White regains his piece and after 20 … d6 21 xd6 c7 his chances improved with 22 g3 h6 23 e5 ac8 24 d5 and he eventually won. The sacrificial version of d4-d5 arises in positions like this: 1 d4 f6 2 c4 e6 3 f3 d5 4 c3 c5 5 cxd5 xd5 6 e4 xc3 7 bxc3 cxd4 8 cxd4 b4+ 9 d2 xd2+ 10 xd2 0-0 11 c4 c6 12 0-0 b6 13 ad1 a5 14 d3 b7.

White to play Here 15 e5? xf3 16 gxf3 is bad because Black can plug attacking lines with 16 … h4 or 16 … f5/… d5, leaving White with bad pawns. But 15 d5 and 15 … exd5 16 e5! is dangerous and similar to a celebrated Polugayevsky – Tal game. White can use d4, e.g. 16 … c8 17 f4 a6 18 xa6 xa6 19 d4 and 20 f5. If Black doesn’t trade bishops he has problems such as 16 … e7 17 f4 c4 18 b1 a3? 19 xh7+! and then 19 … xh7 20 g5+ g6 (20 … g8 21 h4 wins) 21 h4! and d3-g3. Bear in mind that – once again – when you have doubts about changing the structure, the best decision is to keep your fingers off the pawns.

White to play White is better developed and might be tempted by 1 d5. But that’s premature, e. g. 1 … c5! and then 2 c3 exd5 3 exd5 d6 4 xd6 xd6 with even chances. But what about 1 e5 ? White underestimated the impact of 1 … c4! in Eljanov – Kulaots, Kharkov 2003. Then 2 d5 c5! would prompt a favorable trade of queens. White could have recovered with 2 e3 or 2 d2. But he went ahead with his attack – 2 d2? c2! 3 g3 b8! 4 e1 xd4 5 xh6 d7

– and was losing after 6 c1 f5 7 e3 d5 8 g5 d3 9 f3 10 d4 c5.

xa3

So what should White have done back at the diagram? A good preparatory move, such as 1 e3, was in order.

White plays e2-e3 When White’s e-pawn stops on the third rank, additional possibilities arise. One is that he can retake with his e-pawn after … cxd4. That creates hanging pawns on c3 and d4. But this is usually not advisable: 1 d4 f6 2 c4 e6 3 c3 b4 4 e3 c5 5 f3 0-0 6 d3 d5 7 a3 xc3+ 8 bxc3 dxc4 9 xc4 c7 10 a2 b6 11 0-0 b7 12 b2 bd7 13 e2.

Black to play White may be preparing c3-c4 to liberate his b2-bishop so this is a good time for 13 … cxd4. How should he retake? After 14 exd4 he can meet 14 … d5 with 15 c4! and mobilize the hangers. But they are a liability if Black inserts 14 … b5! before … d5. Black has play for the pawn after 15 xb5 xf3 16 gxf3 but better is 15 … ab8! 16 e2 d5. White may be able to defend with 17 d2! f4 18 c1. But in Golz – Averbakh, Dresden 1956 play went 17 c4? f4 18 d2 xf3 19 gxf3 e5! and 20 dxe5 h3+ 21 g2 xe5! 22 xe5 xe5 23 xh3 b6! wins. The right recapture is 14 cxd4!, which allows White to contest the c-

file. True, Black is OK after 14 … d5! and 15 xd5 exd5. But so is White after 16 e5 ac8 17 ac1 b7 18 b5 or 16 fc1 b7 17 a4. The significance of the e-pawn being at e3 grows if Black pushes his epawn. Then it may not be tactically possible or positionally desirable for White to push his d-pawn to the fifth rank. 1 d4 f6 2 c4 e6 3 c3 b4 4 e3 d5 5 d3 0-0 6 f3 c5 7 0-0 c6 8 a3 xc3 9 bxc3 dxc4 10 xc4 c7 11 a2 e5.

White to play Black’s last move prepares pressure on d4 (12 … g4 and … ad8). And 12 d5?, creating a passed pawn, is met strongly by 12 … e4! and 13 dxc6 g4! with a powerful attack. In similar positions after … e5 White can favorably capture twice on e5 and then push f2-f3/e3-e4. But here that runs into tactics, e.g. 12 dxe5 xe5 13 xe5 xe5 14 b2 and now 14 … g4! 15 g3 h5 16 h4 and perhaps 16 … e6!? 17 xe6 fxe6 followed by … c4 or … ad8/… f5. Experience indicates 12 h3 is best in the diagram although 12 … e4! is still good, e.g. 13 h2 a5 14 f3 and 14 … b6! 15 fxe4 xe4 and … b7. In general, if the double exchange on e5 doesn’t bring White an edge he either has to deal with the double-edged … e4 or allow exchanges on

d4 that create an Isolani.

Supplemental Games # 1 When trades help Black: Cebalo – M. Gurevich, Bern 1989 – 1 d4 f6 2 f3 e6 3 c4 b6 4 a3 b7 5 c3 d5 6 cxd5 xd5 7 c2 xc3 8 bxc3 d7 9 e4 c5 10 f4 e7 11 d3 c8 12 b1 (12 0-0 cxd4! 13 cxd4 xc2 is a fine endgame for Black.) 12 … 0-0 13 0-0 a6! 14 a4 cxd4 15 cxd4 xd3 16 xd3 b7 17 fe1 f6! (This anticipates both e4-e5 and d4-d5.) 18 g5? fd8 19 ad1 ac8 20 e5

Black to play 20 … d7! 21 f4 (Black is better after 21 xd7 xg5 22 e5 f6 23 f3 xd4! 24 xd4 e5 or 21 xe7 xe5 22 g3 xe7 23 xe5 b4!.) 21 … b4 22 e2 xe5 23 xe5 c6! 24 g3 f8 25 b3 c4 26 b5

Black to play

26 … f6! 27 f4 e5 28 e3 exd4 29 ed2 xb5 30 axb5 dxe3! and White resigned before 31 xd8 xd8 32 xd8 e2. # 2 Black uses threats of both … cxd4 and … c4: Bareev – M. Gurevich, Lvov 1987 – 1 d4 f6 2 c4 e6 3 f3 b6 4 a3 b7 5 c3 d5 6 cxd5 xd5 7 c2 e7 8 e4 xc3 9 bxc3 0-0 10 b2 c5 11 d1?

Black to play 11 … c7! (So that 12 d3 c4! 13 e2 f4 or 12 b1 cxd4 13 cxd4 xa3! 14 xa3 c3+ and … xa3 favors Black.) 12 d5 exd5 13 exd5 c4! (Black sacrifices a pawn to stop c3-c4.) 14 a4 c8 15 d4 d7 (Now 16 xc4 c5 would have been fine for Black.) 16 xc4 f6 17 d6?! e8+ 18 f1 d8 19 d1 c5 20 c2 d7 21 h4?

Black to play 21 … g4! 22 a2 e4 23 d2 d3+! and wins, e.g. 24 g1 e2 25 c1 d8.

# 3 Black provokes e4-e5: Gligoric – Smyslov, USSR-Yugoslavia match 1959 – 1 d4 f6 2 c4 g6 3 c3 d5 4 cxd5 xd5 5 e4 xc3 6 bxc3 g7 7 c4 c5 8 e2 0-0 9 0-0 c6 10 e3 c7 11 c1 (White tries to discourage … cxd4 because his rook would control the file after cxd4.) 11 … d8 12 h3 b6 13 f4 e6! 14 e1 b7 15 f2 (The delay in … cxd4 pays off in lines such as 15 f5 a5! 16 d3 exf5 17 exf5 e8 and then 18 f2 c4! 19 b1 e7.) 15 … a5 16 d3

Black to play 17 … f5! (This works best when f4 is not possible.) 17 e5 c4 18 c2 c6 19 g4 e7! 20 h2 c6 21 g3 b5 22 a4 a6 23 b1 ab8 24 d2 bxa4! 25 a1 a8 26 xa4 c7 27 a2 b6 28 gxf5 exf5 29 c1 d5 30 e2 a5 31 c2

Black to play 31 … b3! 32 xb3 cxb3 33 a4 f8 34 b2 e3! (So that 35

xe3 c6! threatens the rook and mate on g2.) 35 fa1 c4 36 g3 e7 37 f1 c6 38 xc4 (Or 38 g3 h4) 38 … h1+ 39 g3 h5! White resigns. # 4 After e4-e5 and … d5, White plays xd5!: Smyslov – Ernst, Subbotica 1987 – 1 d4 f6 2 f3 d5 3 c4 e6 4 c3 c5 5 cxd5 xd5 6 e4 xc3 7 bxc3 cxd4 8 cxd4 b4+ 9 d2 xd2+ 10 xd2 0-0 11 c4 d7 12 0-0 f6 13 fe1 d7?! 14 e5! d5 (Worse is 14 … e8 15 d5!.)

White to play 15 xd5! exd5 16 ab1 b6 17 h3! (Black faces lift to g3.) 17 … c8 18 h2 a4 19 bc1

h2-g4 and a rook

Black to play 19 … b5? (The bad endgame of 19 … d7 20 g4 xc1 21 xc1 c8 22 xc8+ and g5 is better than what happens.) 20 g4 c4 21

c3! b5 22 g3 h8 23 f4! a5 24 f5 b4 25 f6 g8 26 f4 xa2 27 e6! fxe6 28 e5 Resigns (29 f7 mate or 29 g6+/30 h4 mate).

Chapter Eleven: The Lopez Formation

The Lopez formation This structure is a rarity outside of games that begin with the Ruy Lopez (1 e4 e5 2 f3 c6 3 b5). But the Ruy is so popular and the nuances of pawn management in it are so great that it deserves its own chapter. White has three main options: trading d-pawns, pushing d4-d5 or leaving the center as it is. Black has a greater choice. He can push … c4, open up the c-file with … cxd4, create a pawn imbalance with … exd4 or exchange twice on d4. Each option can be good – or bad – depending on circumstances. Let’s start with a trade of d-pawns. It gained popularity after one influential game won by Vsevolod Rauzer, Rauzer – Riumin, Leningrad 1936: 1 e4 e5 2 f3 c6 3 b5 a6 4 a4 f6 5 0-0 e7 6 e1 b5 7 b3 d6 8 c3 a5 9 c2 c5 10 d4 c7 11 bd2 c6 12 a4 b8 13 axb5 axb5.

White to play The conventional wisdom had held that since an open c-file gives Black play, White’s best policy was to close the center, d4-d5, and attack with g2-g4 and f1-g3-f5. But Rauzer showed that it was beneficial to stop … cxd4 in another way, 14 dxc5! dxc5. White pawns then guard both d4 and d5 and he has a strong plan of f1-e3-d5!. Black needs preparation to change and exploit the pawn structure with … c4 and/or … b4. Instead, he tried to control the d-file, 15 f1 e6 16 e3 0-0 17 g5 fd8 18 f3 d6. Then 19 xe6 xe6 20 d5 was tempting – but Black can reply 19 … fxe6!?, to control d5 and f5 and use the f-file with … f8. White did better with 19 f5! xf5 20 exf5!. He cleared e4 for his minor pieces and could use the f-pawn with g2-g4-g5 as a kingside battering ram.

Black to play White’s advantage appeared after 20 … h6 21 e4 xe4 22 xe4

f6 23 e3 e7 24 b4 c4 25 g3 d7 and now 26 a7 d8 27 xd7 xd7 28 h4! h8 29 g4!. The pawns can’t be restrained (29 … xh4 30 h3 and g4-g5.) Black went downhill: 29 … g8 30 g5 e7 31 d1 c7 32 f6! xf6 (32 … gxf6 33 f5!) 33 gxf6 xf6 34 c2 d8 35 xh6 xd1+ 36 xd1 e4 37 f4 d8 38 e2 Resigns.

White’s d4-d5 Option But in some situations, closing the center is better. This is seen in what for many years was the main line of the Ruy Lopez: 1 e4 e5 2 f3 c6 3 b5 a6 4 a4 f6 5 0-0 e7 6 e1 b5 7 b3 d6 8 c3 0-0 9 h3 a5 10 c2 c5 11 d4 c7 12 bd2 c6.

White to play This time the Rauzer plan, 13 dxc5 dxc5, and the natural follow-up, 14 f1 e6 15 e3 ad8 16 e2, offers little after 16 … c4!. Black is under no obligation to capture on f5, e.g. 17 f5 fe8! and 18 xe7+ xe7 with equal chances. More promising is 13 d5!. Then 13 … d8 disconnects Black’s heavy pieces and allows White a queenside edge with 14 a4!, e.g. 14 … b4 15 c4 or 14 … b8 15 axb5 axb5 16 b4, perhaps intending b3-a5. On the other hand, 13 … a5 allows White to pursue a kingside attack with f1-g3-f5 and/or f2-f4. He might start with 14 b3 so that 14 … c4 can be met by 15 b4! b7 16 a4 with the upper hand on the queenside. Better is meeting 13 … a5 14 b3 with 14 … d7. Then 15 f1 b7 16 g3 g6 17 h6 fc8 18 d2 d8 was Velvart – Lengyel, Balatonbereny 1992. White found 19 f5!, based on 19 … gxf5? 20

g5+. Play went 19 … f8 20 xf8 xf8 21 h6+ g7 22 g5! (22 … xh6 23 e6+) and then 22 … e8 23 f4! exf4 24 xf4. Black resigned after 24 … d7 25 f1 e7? 26 h4 f6 (before 27 f5+ gxf5 28 xh7+ and mates). Black’s best source of counterplay in d4-d5 positions lies in robust queenside action, such as a well-timed … c4 followed by … b4. In the opening we just looked at, Black may play the immediate 11 … c6. Then on 12 d5 a5 Black has chances after 13 bd2 c4! and 14 f1 b7.

White to play Black’s plan is … c5, … a5 and … b4 with the goal of creating a queenside target. For example, 15 g3 c5 16 e3 c7 17 d2 a5 18 ad1 and now he anticipated f2-f4 with 18 … d8! 19 h2 a7 20 h1 b6 21 f4 cd7 in Tukmakov – Romanishin, Moscow 1974. A good trade of dark-squared bishops is coming and Black eventually won an endgame after 22 f5 b4! 23 gf1 b3! 24 b1 bxa2 25 xa2 xe4 26 xb6 xd2 27 xa7 xf1 28 xf1 xa7 29 xc4 b6. White can combat the Black plan with 15 b3 in the last diagram. But he does better by anticipating … c4 earlier with 13 b3! b7 (13 … c4? 14 b4! b7 15 a4) 14 e3 since 14 … a5 15 a4! allows him to win control of c4.

Black exchanges on d4 When Black opens the c-file he is usually the first to occupy it with heavy pieces. But that may be temporary and/or insufficient to equalize. It may enable White to obtain an edge with a well-timed d4-d5, exposing a hole at c6. For example: 1 e4 e5 2 f3 c6 3 b5 a6 4 a4 f6 5 0-0 e7 6 e1 b5 7 b3 d6 8 c3 0-0 9 h3 a5 10 c2 c5 11 d4 c7 12 bd2 cxd4 13 cxd4 c6 14 b3. Recent experience indicates Black should try to drive this knight back, with 14 … a5 and … a4. Then a spirited battle shapes up based on Black’s greater queenside space and White’s targeting of the b5-pawn. Why weaken b5? The answer is shown when Black relies on routine play, e.g. 14 … d8 15 d2!, which threatens 16 d5 and 17 a5. Black can’t play 15 … d5 (16 xe5! xe5 17 a5) and is not quite equal after 15 … exd4 16 bxd4 and c1. White is also clearly better after 16 … b8 17 d5!.

Black to play In Smyslov – Keres, Leningrad-Moscow 1941, White had a queenside edge after 17 … a7 18 a5! d7 19 d3 c8 20 b4!. When Black minimized it with 20 … d8 21 b3 b6, his lack of kingside

defenders invited a strong 22 meaningless.

h4! and

f3/

f5. The c-file proved

The chief alternative to … cxd4 is … exd4, which unbalances pawns on the wings. Black will create a queenside majority with … c5 and White usually pushes d4-d5 and attacks the kingside. These positions have a long and rich history, beginning with instructive Capablanca games through the Kasparov – Karpov battles of the 1980s. See Supplemental Game # 1 for how White’s attack wins. One of the fundamental goals of … exd4 is to put pressure on the e4pawn along the file. For example, Pilnik – Smyslov, Mar del Plata 1966: 1 e4 e5 2 f3 c6 3 b5 a6 4 a4 f6 5 0-0 e7 6 d3 b5 7 b3 d6 8 c3 0-0 9 e2 a5 10 c2 c5 11 d1 c7 12 bd2 e8 13 f1 h6 14 e3 f8 15 d2 e6 16 d4.

Black to play By delaying the push to d4 White avoided an early … cxd4. But White’s loss of time allows a strong 16 … exd4! and then 17 cxd4 c4. White’s queen is misplaced on the e-file, such as after 18 d5 xe3! 19 xe3 g4. Black also gets good play from 18 dxc5 dxc5 19 c3 xe3 20 xe3 and 20 … g4! 21 e2 d6 22 h3 e5. Black’s queenside majority becomes more important than White’s kingside majority. In the game White played 18 dxc5 dxc5 19 xc4 xc4 but he had no convenient way of defending against the coming … d5, e.g. 20 e3

d5! or as the game went 20 e1 d5! 21 a5 f4. White tried 22 xd5!? xd5 23 e5 but he had scant compensation after 23 … b4 and quickly lost.

Double Trade on d4 The third major scenario occurs when Black exchanges twice on d4. That leaves only two pawns in the center, a White one at e4 and a Black one at d6. The Black pawn is more vulnerable and the White outposts at d5 and f5 are more useful than Black’s at c5 and e5. Supplemental Game # 2 shows how quickly White can develop a winning kingside attack. But Black has assets in his ability to attack e4 and the prospect of … d5. Trading off the last center pawns should favor the player who is better developed or who has fewer exposed weaknesses. 1 e4 e5 2 f3 c6 3 b5 a6 4 a4 f6 5 0-0 e7 6 e1 b5 7 b3 d6 8 c3 0-0 9 h3 h6 10 d4 e8 11 bd2 f8 12 f1 d7 13 g3 a5 14 c2 c5 15 b3 c6 16 e3.

Black to play White delayed a decision with his d-pawn in favor of e3/ d2. This would work well after 16 … c8 because removing the rook from the afile invites 17 d5 e7 18 a4!. Then White has no queenside cares and can attack with h2/f2-f4. Note that White played 15 b3 to avoid … c4. That’s often a wise precaution. But the pawn move does create a potential hole at c3. Black exploited it with 16 … cxd4 17 cxd4 exd4 18 xd4 d5! and 19 exd5 xd5.

Black’s pieces turn out to be better placed than White’s after 20 xc6 xc6 (21 d4 xe1+ 22 xe1 b4!), for example. In Ki. Georgiev – Beliavsky, Cacak 1996, White traded pieces, 20 e4 xe3 21 xe3 c8 22 xc6 xc6 23 f3 xe4 24 xe4, to release tension. He hoped that the N-vs.-B matchup would be relatively even. But it wasn’t after 24 … e6! 25 c3 xc3 26 xc3 c5! and he eventually lost after 27 d5 a5 28 g3 d2. Finally, it’s worth noting that Black can avoid … c5 entirely: 1 e4 e5 2 f3 c6 3 b5 a6 4 a4 f6 5 0-0 e7 6 e1 b5 7 b3 d6 8 c3 0-0 9 h3 b8 10 d4 bd7 11 bd2 b7 12 c2 e8 13 f1 f8 14 g3 g6 15 b3.

Black to play Before we explore this position, let’s talk about the hidden battle over the e4-pawn. Black’s … e8/… f8 and … b7 were designed to win it after … exd4. White anticipated the threat by retreating his bishop from the wonderful b3-f7 diagonal and carrying out the f1-g3 maneuver that is so much a part of the Lopez DNA. True, White could also eliminate threats to his e-pawn by playing d4d5 at some point. But then … c6! is a good response since dxc6 creates quite a nice version for Black of the Boleslavsky Hole. The possibility of d4-d5 also explains 15 b3. Unlike previous examples, b2-b3 was not intended to stop … c4 or to discourage … c4.

This time White played it to have the opportunity to play d4-d5 and, after … c6, maintain the d5-chain with c3-c4!. Back to the diagram. White can meet 15 … c5 with 16 d5! after which 16 … c4! and … c5 is a typical, roughly equal Ruy middlegame. The chief alternative is 15 … d5!?. This leads to double-edged play after 16 dxe5 xe5 17 xe5 xe5 18 f4 xe4! 19 xe4 dxe4 – but not 18 … e8? 19 e5 with a favorable structure. Black can even get a slight edge after 16 exd5 xd5 17 dxe5 and then 17 … xc3 18 d4 xe5! since 19 xc3? xf3+ 20 gxf3 g7!. Instead, White can retain chances for advantage with 16 g5! h6 17 h4! based on 17 … g5 18 xg5! being sound, or 17 … dxe4 18 xe4. See also Supplemental Game # 3.

Supplemental Games # 1 White attacks in the … exd4/… c5 structure: Khalifman – Gavrilov, St. Petersburg 1994 – 1 e4 e5 2 f3 c6 3 b5 a6 4 a4 f6 5 0-0 e7 6 e1 b5 7 b3 d6 8 c3 0-0 9 h3 e8 10 d4 b7 11 bd2 f8 12 a4 h6 13 c2 (White repositions the bishop to attack b5, e.g. 13 … b8 14 d3.) 13 … exd4 14 cxd4 b4 15 b1 g6 16 a3! g7 17 h2

Black to play 17 … c5 18 d5! d7 19 df3 b6 20 axb5 axb5 21 g4 h7 22 h4! c8 23 fh2 xg4 24 xg4 e7 25 xa8 xa8 26 h5 c4 27 e5! (Opening the b1-g6 diagonal dooms Black.) 27 … dxe5 28 b3 b6 29 hxg6+ fxg6

White to play

30 xe5! xe5 31 h5 g7 32 xe5 4xd5 33 e6! (So that 33 … f6 34 xg6+! xg6 35 e7+ or 34 … h8 35 f5 a1 36 xf6 xc1+ 37 h2 d7 38 e6 and e8+ wins.) 33 … g8 34 xg6 Resigns. # 2 A double exchange on d4 fails to provide Black counterplay: Shirov – Fernandez, Spanish Championship 2002 – 1 e4 e5 2 f3 c6 3 b5 a6 4 a4 f6 5 0-0 e7 6 e1 b5 7 b3 d6 8 c3 0-0 9 h3 a5 10 c2 c5 11 d4 c7 12 bd2 cxd4 13 cxd4 d8 14 b3 c6 15 b2 exd4 16 xd4 xd4 17 xd4 b7 18 c1 a5 19 f1 ac8

White to play 20 e3 (White offers the a- or e-pawn, e.g. 20 … xe4 21 xe4 xe4 22 f5!.) 20 … g6 21 f3! (Black was preparing 21 … d5!.) 21 … d7 22 g4 g5 23 h4! xc1? (A better try is 23 … e5 24 xe5 xc1.) 24 xc1 f5 (No defense to h6+.) 25 h6+ f8 26 f4! e5 27 exf5 Resigns. # 3 Black plays … d5 without … c5: R. Byrne – Spassky, Candidates match 1974 – 1 e4 e5 2 f3 c6 3 b5 a6 4 a4 f6 5 0-0 e7 6 e1 b5 7 b3 d6 8 c3 0-0 9 h3 b8 10 d4 bd7 11 bd2 b7 12 c2 e8 13 f1 f8 14 g3 g6 15 b3 g7 16 a4

Black to play 16 … d5! (16 … exd4 17 cxd4 c5 18 f4! favors White.) 17 dxe5 xe4 (Now 18 xe4 dxe4 19 xe4 xe4 20 xe4 xe5 favors Black slightly.) 18 xe4 dxe4 19 g5 exf3!? (Enough for equality was 19 … c8.) 20 xd8 axd8 21 axb5! xe5 22 bxa6? (22 c2! fxg2 23 xe5! xe5 24 bxa6 was unclear.) 22 … xd1 23 exd1 a8

White to play 24 gxf3 xf3+ 25 f1 xc3 26 ac1 d2+ 27 g1 a5 28 b4 f3+ 29 f1 h2+ 30 g1 f3+ 31 f1 b6 32 c2 h2+ 33 g1 f3+ 34 f1 f8 35 e2 h2+ 36 g1 f3+ 37 f1 e4 38 a2 h2+ 39 g1 f3+ 40 f1 h4 41 f4 f3 42 d3

Black to play 42 … g5! 43 e2 g2+ 44 e1 f3+ 45 d1 e5! 46 c3 d5 47 d2 c4! 49 f4 gxf4 50 xf4 xa6 and wins.

Chapter Twelve: The Closed Sicilian/English

The Closed Sicilian formation When no pawns are exchanged in the early stages of an English Opening or Sicilian Defense, each player concentrates on ‘his’ wing. Black works on the queenside. He also has good pressure on dark squares and can force matters in the center with … d5. White looks to the kingside. Unlike a Queenside Stonewall, he can control d4 with c2-c3 and can also open the center with c2-c3/d3-d4. If the center is left unchanged, wing attacks can proceed with surprising speed, as in Blatny – Taimanov, Decin 1975: 1 e4 c5 2 c3 c6 3 g3 g6 4 g2 g7 5 d3 d6 6 f4 b8 7 a4? f6 8 f3 0-0 9 0-0 a6 10 h4 g4! 11 d2?! d7 12 e2 b5 13 axb5 axb5 14 h3 b6.

White to play White failed to discourage … b5 and Black is ready to exploit the opened a-file with 15 … a8. White’s second mistake was allowing his queen and minor pieces to be confused. When the rival pawn storms began, 15 f5 b4 16 g4, Black broke through with 16 … c4+ 17 h1 b3!. Black’s ultimate target is b2, which his queen, rook and g7-bishop are gunning for. White could have kept the position partially closed with 17 d4 but then 17 … e8 and 18 c3 a5! prepares to occupy the hole at b3. In the game, White didn’t want to defend his pawns after 18 dxc4 e5! 19 cxb3 xb3. He ignored the queenside with 18 g5 e8 19 f4.

Black to play But the storm broke with 19 … c3! and 20 bxc3 b2, winning a piece and the game. Better was 19 c3 but after 19 … cxd3 20 xd3 e5 Black

would be better thanks to his pieces (… b5, … c7, … fc8, … a8). See also Supplemental Game # 1.

Closed English Strategies With colors reversed, it is White who pushes his a- and b-pawn – to create holes and make b7 a target – and it is Black who often responds with a kingside attack. De Santis – Naumkin, Verona 2006 went: 1 f3 e6 2 g3 f5 3 g2 f6 4 0-0 e7 5 c4 0-0 6 c3 d6 7 d3 e5 8 b1 e8 9 b4 h5.

White to play Black intended a Dutch Defense but White’s refusal to push d2-d4 left a belated Closed English on the board. Before Black can carry out his plan of … f4, … h3/… g4, White exposed the queen’s absence from d8 with 10 d5! and 10 … xd5 11 cxd5. That sets up c7 as a target. Black tried to repair his queenside with 11 … c6 and 12 dxc6 xc6 (rather than 12 … bxc6 13 b5!, which would exploit the g2-a8 line). But White was able to pursue the basic policy of 13 b5! d8 14 a4 f7 15 a3 f4 16 a5!. With a5-a6 coming up, Black could have tried 16 … h3 and … xg2. But White would dominate the light squares after b3-d5!. Instead, Black chose 16 … fxg3 17 fxg3! a6 and 18 bxa6 xa6 19 b5 d8 20 b3.

Black to play Black is worse in all three areas of the board and one more error, 20 … e8?, weakened the f7 square and decided the game: 21 d2! xa5 22 d5 f8 23 c4 c7 24 xe5! xa3 25 xf7! xb3 26 h6+ Resigns. In other situations … f5-f4 can be dangerous. But White can try to stop it with e2-e3 and e2. This works when Black cannot reinforce control of f4, say with … g5. Or Black can make the … f4 advance as a sacrifice. That’s what White often does in the Grand Prix Attack of the Sicilian, 1 e4 c5 2 c3 c6 3 f4 g6 4 f3 g7 5 c4 e6 6 f5!?. Another anti- … f4 plan is simply to put a White pawn on f4. Consider: 1 c4 e5 2 c3 c6 3 g3 g6 4 g2 g7 5 e3 d6 6 ge2 ge7 7 d3 0-0 8 0-0 h6 9 b1 g5 10 b4 g6 11 b5 ce7 12 b3 h8 13 a4 b8 14 d5 f5.

White to play White has made queenside progress and can prepare for c4-c5. He invites … xd5 because cxd5! again exposes c7. But to have a free hand on the queenside, he needs to stop Black on the kingside with 15 f4!. In Szabo – Damjanovic, Beverwijk 1966, White clarified the center after 15 … e6 16 xe7 xe7 17 c2 d7 18 d4!. Then 18 … e4 would have allowed him to expand with d4-d5/ d4 or prepare c4-c5 or b5-b6. Black accepted the consequences of 18 … gxf4 19 exf4 exd4? (better 19 … exf4 20 gxf4 h4). But White had a strategically won game in view of Black’s weak pawns and the seventh rank after 20 b2 c6 21 bxc6 bxc6 22 xd4 c5 23 xg7+ xg7 24 fd1 e7 25 a5! xb1 26 xb1 c8 27 f2! a6 28 d3 f6 29 b7. Yet another way of responding to the advance of the f-pawn is to trade it off for the e-pawn. We see this in the Closed Sicilian after 1 e4 c5 2 c3 c6 3 g3 g6 4 g2 g7 5 d3 d6 6 f4 e5!? 7 h3 ge7 8 0-0.

Black to play White’s knight may look strange on h3 but it can come into action at g5 after 8 … 0-0? 9 f5!. Since Black has weakened f6 with his sixth move, he would be in trouble after 10 g4 and later g4-g5. But the attack is also strong after 9 … gxf5 10 exf5. e.g. 10 … xf5 11 h5 and 12 e4 or 10 … xf5 11 xf5! xf5 12 e4 fd4 13 h5. Going back to the diagram, a good idea is 8 … exf4!. Then 9 gxf4 f5

halts White’s pawns. Black would have the better of the structure after 10 exf5 xf5 and good center play after, say, 10 e3 0-0 11 d2 h8 12 ae1 e6 13 g5 g8! followed by … d7 and … ae8. The more popular answer to 8 … exf4 is recapturing with a piece. After 9 xf4 White can later play a knight to d5. And after 9 xf4 he can attack with d2/ h6 and f2/ af1. But Black has defensive resources in either case. For example, 9 xf4 0-0 10 d2 b8 11 h6 was met in Larino Nieto – Tatai, Vecindanio 2010 by 11 … xh3! 12 xh3 b5 13 xg7 xg7, intending 14 … b4 15 d5 xd5 16 exd5 e5 and … f5! to make the knight better than the bishop. Instead, the game went 14 e2 b4 15 f2 and Black had the edge after 15 … d5!.

Counter-strategies Although both players have their eyes on the wings, either one can push his d-pawn. In the English, White may follow b2-b4-b5 with d3-d4 now that Black has less control over d4 and e5. But Black can try to build a monster center with … c6 and … d5 as in Filep – Osnos, Debrecen 1969: 1 c4 e5 2 c3 c6 3 g3 f5 4 g2 g6 5 d3 g7 6 d2 d6 7 b1 a5 8 a3 f6 9 b4 axb4 10 axb4 0-0 11 c1 e7! 12 h3?!.

Black to play White’s 11 c1 discourages … h6/… g5. But his lack of pressure on squares such as e5 invited Black to play 12 … c6! 13 0-0 d5. Black has nothing to fear from 14 b5 d4 or 14 cxd5 cxd5 15 b5 c6. White tried 14 c5 instead but Black seized the upper hand with 14 … d4! 15 d1 fd5. And when White tried to change the structure further, 16 e4? dxe3 17 fxe3, he was ground down: 17 … c7! 18 b2 h6 19 f2 e6 20 c4 b5 21 h1 h7 22 b6 a2 23 a1 xa1 24 xa1 and now 24 … f4! 25 exf4 exf4 26 e1 fxg3 27 hxg3 f5! 28 xe6 bd4 29 e1 xg3+ 30 h2 h4+ White resigns.

A second way Black can fight for d5 is to meet d5 with … xd5 and … c6. We saw an unsuccessful version of this earlier in this chapter. Let’s see how it succeeds: 1 c4 e5 2 c3 c6 3 g3 g6 4 g2 g7 5 d3 d6 6 e3 ge7 7 ge2 0-0 8 d5?!.

Black to play Now 8 … xd5 9 cxd5 e7 equalizes in view of 10 0-0 c6! 11 dxc6 xc6 or 11 c3 cxd5 12 xd5 e6. In one GM game, White got the worst of 10 d4 c6 11 dxc6 xc6 12 00? because of 12 … exd4 13 xd4 xd4 14 exd4 b6! (or 13 exd4 g4). In another, Bilek – Smyslov, Polanica Zdroj 1968, he tried 10 0-0 c6 11 dxc6 and then 11 … bxc6! 12 d4?! a6 13 e1 b6! 14 dxe5 dxe5.

White to play Once again a theoretically inferior pawn structure – inferior for Black because of the isolated c6-pawn – is in reality superior because of what it does for pieces. White’s were confined and Black’s were lively after 15 c2 fd8 16 d2 d3! 17 c1 d5 and then 18 c3 b4 19 a4 b5 20 xb4 xb4 21 c3 e4!. Black won without incident: 22 d1 ab8 23 d2 c5 24 e1 xc3! 25 bxc3 a5 (Not 25 … xc3 26 xe4!) 26 c1 c4! 27 d1 d5 (Five ranks beat three.) 28 g4 db5 29 f4 a3 30 dd1 e8 31 h3 b2 32 d7 e7 33 g4 c5 34 f6 e5 35 a6 g7 36 a4 c7! 37 a5 b5 38 a1 bc5! and White resigned when he saw that 39 … 5c6 would trap the queen.

The Nimzo-Botvinnik Formation A close relative of what we’ve examined is an English Opening structure in which White plays e2-e4. Aron Nimzovich was the first master to demonstrate that surrendering a hole at d4 was a small price to pay for the positional benefits. A generation later Mikhail Botvinnik adopted a reversed form of the formation against the Closed Sicilian, 1 e4 c5 2 c3 c6 3 g3 g6 4 g2 g7 5 d3 d6 6 ge2 by playing 6 … e5!?. He won a 1954 World Championship game that went 7 d5 ge7 8 c3? (8 ec3!) 8 … xd5! 9 exd5 e7 10 0-0 0-0 11 f4 d7 12 h3 c7 13 e3 ae8 14 d2 f5 15 f2 h5! 16 ae1 d8 17 h2 h6 18 h4 f6!. When facing this structure in the English, Black has a big decision about which of his center pawns goes to the fourth rank: 1 c4 f6 2 c3 g6 3 g3 g7 4 g2 0-0 5 e4 d6 6 ge2.

Black to play Black might be thinking, ‘I’m playing the King’s Indian Defense. What are you doing?’ What White is doing is subtle. He wants Black to play 6 … e5, the traditional King’s Indian move. Then f2-f4 enjoys added strength. For instance, 6 … e5 7 0-0 c6 8 d3

and now, instead of 8 … e8! (9 f4 f5 or 9 … exf4 10 gxf4 f5) Black might decide on queenside play with 8 … a6 9 f4 b8 10 h3 b5. This makes some sense because he has stolen White’s Closed English plan of b1/b2-b4-b5. But it leaves White free to expand on the other wing, with 11 f5! as in D. Byrne – Myagmasuren, Varna 1962. Black gets little benefit from … bxc4 so play went 11 … b4 12 b1 d4. Then after 13 xd4 exd4 Black can occupy c5 or e5 with a knight, But White continued 13 g4! d7 14 d2 c5 15 f3.

Black to play Black has created a Queenside Stonewall. That’s a bad decision since there is more space on the kingside to exploit than on the queenside. White can slowly build up with g3, e3, d2, f2 and af1 along with g4-g5 and h3-h4-h5. Black tried to trade dark-squared bishops with 15 … xf3+ 16 xf3 h4 17 e3 h6!. But the pawn structure allowed White to make careful progress with 18 f2! e7 19 h4! gxf5 20 xf5 g7 21 g3 f6 22 g5 e8 23 e3! f6 24 e2 xf5 25 exf5!. White dominated the light squares: 25 … c7 26 h5 fd8 27 g6 f8 28 e4 d7 29 f1 e8 30 g5! fxg5 31 hxg5 g7 32 d5+ h8 33 e4 e7 34 h4! f6 (else 35 f6!) 35 gxf6 xf6 36 h5 e7 37 f3 f8 38 h3 a5 39 h6 fd8 40 d5 g8 (White threatened f7 and g7+.) 41 xg8 xg8 42 gxh7+ h8 43 g3 f7 44 g8+ Resigns (44 … xh7 45 g6! and mates).

There were several things that Black could have done better but the simplest improvement is 6 … c5!, to focus on the d4 hole with … e8c7-e6. For example, on 7 d3 c6 8 0-0 with plans of h2-h3, f2-f4, e3, d2 or a2-a3, b1 and b2-b4, Black begins his strategy with 8 … e8!.

White to play Black shouldn’t fear f2-f4-f5 because his e-pawn still controls f6 and there is less danger of g3-g4-g5 and f5-f6. If White realizes that f2-f4 would go nowhere he may turn to b2-b4, as in 9 a3 c7 10 b1. Then 10 … a5 is a popular move but also good is 10 … e6 11 b4 ed4. Trades of minor pieces ease Black’s game, as in 12 xd4 xd4 13 e2 b6 14 xd4 xd4 15 b2 xb2 16 xb2 b7 17 f4 e6 18 f5 exf5 19 exf5 xg2 20 xg2 f6! with the better game in Stolyar – Nezhmetdinov, Moscow 1957.

Supplemental Games # 1 White fatally ignores the queenside in a closed Sicilian: Fedorov – Kasparov, Wijk aan Zee 2001 – 1 e4 c5 2 d3 c6 3 g3 g6 4 g2 g7 5 f4 d6 6 f3 f6 7 0-0 0-0 8 h3 b5 9 g4 a5 10 f5 b4! 11 e1? (White’s main kingside idea, h4 followed by h6, g5 and fxg6, is too slow because of … c4.) 11 … a6! 12 h4?

Black to play 12 … c4! 13 h6 cxd3 14 cxd3 (On 14 fxg6 Black can win with 14 … hxg6 or the cute 14 … dxc2 15 gxh7+ xh7 16 xg7 xg7 17 g5 f6 18 xf6 c1( )+ 19 h2 h8!.) 14 … xd3 15 e1 xh6 16 xh6 b6+ 17 h1 e5! 18 bd2 ac8 19 g5 c2 20 f1 xf1 21 xf1 fc8 22 fxg6 hxg6 23 b3

Black to play

23 … xg2! 24 xg2 c2+ 25 g3? lost was 25 h1 e3.

e3+ White resigns. Also

# 2 White uses the b-file after b2-b5-b5xc6: Ribli – Mokry, Moscow 1994 – 1 c4 e5 2 g3 c6 3 g2 g6 4 c3 g7 5 e4 d6 6 ge2 ge7 7 d3 0-0 8 0-0 e6 9 d5 d7 10 e3 f5 11 d2 ae8 12 ac1 f7 13 b4 c8 14 b5 d8 15 exf5 xf5 (White is also better after 15 … gxf5 16 f4!.) 16 fe1 h3 17 h1 c6 18 bxc6 bxc6 19 dc3 c5? (Stopping 20 d4 isn’t worth this concession.) 20 b1 h8 21 e4! ff8 22 2c3 e6 23 d5 e7 24 xe7 xe7 25 b5 d4

White to play 26 xd4! exd4 (Black’s bishops are useless.) 27 eb1 f5 28 g2 xe4 29 xe4 h3? 30 b8 ee8 31 xe8 xe8 32 b7 h6 33 a5 Resigns. # 3 White’s e2-e3 setup in the Closed English: Goldin – Murey, Moscow 1966 – 1 c4 e5 2 c3 c6 3 g3 g6 4 g2 g7 5 e3 ge7 6 ge2 0-0 7 0-0 d6 8 b3 e6 9 d5 (Black was threatening 9 … d5! since 10 cxd5 xd5 is a good version for him of a reversed Dragon.) 9 … f5 10 xe7+? xe7 11 d4 (Now 12 d5 is threatened and 11 … exd4 12 exd4 favors White after b2/ f4.)

Black to play 11 … f4! 12 d5 (Or 12 gxf4 exd4! and 13 xc6 d3!.) 12 … g4 13 f3 e4! 14 fxg4 f3 15 b1 (Safer was 15 dxc6 xa1 16 cxb7.) fxg2 16 xf8+ xf8 17 xg2 e5 18 f4 g5 19 e6 f3! 20 xg7? xg4 21 e6 f6 22 g1 xg3+ White resigns. “Enough, there are still problems to be solved; the whole truth in chess is not by any means known yet – fortunately.” – Emanuel Lasker.