First published in 2017 by Gloucester Publishers Limited, London.

Copyright © 2017 Nigel Davies

The right of Nigel Davies to be identified as the author of this work has been asserted in
accordance with the Copyrights, Designs and Patents Act 1988.

All rights reserved. No part of this publication may be reproduced, stored in a

retrieval system or transmitted in any form or by any means, electronic,
electrostatic, magnetic tape, photocopying, recording or otherwise,
without prior permission of the publisher.

British Library Cataloguing-in-Publication Data

A catalogue record for this book is available from the British Library.

ISBN: 978 1 78194 408 0

ISBN: 978 1 78194 409 7

Distributed in North America by National Book Network,

15200 NBN Way, Blue Ridge Summit, PA 17214. Ph: 717.794.3800.

Distributed in Europe by Central Books Ltd.,

Central Books Ltd, 50 Freshwater Road, Chadwell Heath, London , RM8 1RX.

All other sales enquiries should be directed to Everyman Chess.

email: info@everymanchess.com; website: www.everymanchess.com

Everyman is the registered trade mark of Random House Inc. and is used in this work
under licence from Random House Inc.

Everyman Chess Series

Chief advisor: Byron Jacobs
Commissioning editor: John Emms

Typeset and edited by First Rank Publishing, Brighton.

Cover design by Horatio Monteverde.
Printed by TJ International Limited, Padstow, Cornwall.
About the Author
Nigel Davies is an International Grandmaster and respected coach. He’s the author of
numerous books and DVDs on the game and is known for the clarity of his explanations.

Also by the Author

10 Great Ways to Get Better at Chess
Alekhine’s Defence
Gambiteer I
Gambiteer II
Play 1 e4 e5!
Play the Catalan
Starting Out: The Modern
Taming the Sicilian
The Dynamic Réti
The Grünfeld Defence
The Pirc: Move by Move
The Rules of Winning Chess
The Trompowsky
The Veresov
Contents
About the Author
Bibliography
Introduction

1 Exchange Variation with Nf3

2 Exchange Variation with Nge2
3 Main Line with Bg5
4 Main Line with Bf4
5 The Catalan and the Réti

Index of Complete Games

Bibliography
Books
Queen’s Gambit Declined, Matthew Sadler (Everyman Chess 2000)
Queen’s Gambit Declined, Sergiu Samarian (Batsford 1974)
Queen’s Gambit Declined: Bg5 Systems, Bogdan Lalic (Everyman Chess 2000)
Starting Out: Queen’s Gambit Declined, Neil McDonald (Everyman Chess 2006)
The Queen’s Gambit & Catalan for Black, Lasha Janjgava (Gambit Publications 2000)

Databases
Chess Tempo Online Database
Introduction
The Queen’s Gambit Declined is normally introduced via the moves 1 d4 d5 2 c4 e6,
but it can also be reached via 1 Nf3 and 1 c4. In fact this is the first great advantage of
playing the QGD as Black: it provides the basis of a defence to flank openings as well
as 1 d4. It is one of the oldest and most respected defences and has been played by
every World Chess Champion in history.
Although I have used the Queen’s Gambit Declined in my own
games, it has been as a coach that I really learned to appreciate
its value. Besides its well-deserved reputation for soundness,
which has made it a firm favourite with so many top players,
the QGD is relatively immune to both the changing tides of
opening theory and computer analysis. Meanwhile, the
greatest advantage for improving players is in its didactic
value: it is a superb vehicle for teaching positional play
because of the clarity of plans and pawn play. For this reason
it has become a major part of my teaching syllabus at my Tiger
Chess (http://tigerchess.com) website.
In writing this book my main focus has been on teaching the reader how to play the
black side, starting out in the simplest and most economical way by aiming to develop
with ... Nf6, ... Be7 and (usually) ... Nbd7 against all White’s possible systems. Besides
that, I look at how particular finesses can be used to improve Black’s chances against
particular systems. This is by far the best approach to learning an opening; having a vast
array of different set-ups and plans is impractical for people with busy lives – it makes
more sense for them to master a few set-ups and then expand these gradually as they
gain experience.
Some readers may also be interested in playing White and here I recommend a simple
and economical approach in capturing on d5 at the earliest opportunity. This is covered
in Chapter One.

Acknowledgements
My thanks goes to Byron Jacobs for his support and encouragement throughout this
project.
Nigel Davies, Southport
May 2017
Chapter One
Exchange Variation with Nf3
Introduction

1 d4 d5 2 c4

Question: What if I want to play a Torre Attack with 1 d4 Nf6 2 Nf3, followed by 3
Bg5, but my opponent replies 2 ... d5 so as to meet 3 Bg5 with 3 ... Ne4 - ?Can I still
play an Exchange Variation with 3 c4 and 4 cxd5 - ?

Answer: Yes, you can, and the Exchange Variation with 1 d4 Nf6 2 Nf3 d5 3 c4 e6 4
cxd5 exd5 5 Nc3 represents a simple and economical repertoire choice. Note, however,
that Black has important extra options besides going for the traditional set-up with ...
Nbd7. The most important of these is 5 ... Be7 6 Bg5 c6 7 Qc2 g6, intending 8 ... Bf5,
which is covered in Games 1-3. As for 5 ... c6 6 Qc2 Na6, see Game 4.
Having said that I’ve found that club players usually go back into the main line with
5 ... Be7 6 Bg5 0-0 7 Qc2 Nbd7 8 e3 c6 9 Bd3 Re8 10 0-0 Nf8, when White gets to
choose between a number of different 11th moves. Positional players will probably like
Anatoly Karpov’s 11 h3, while for those who are tactically inclined I suggest 11 Rae1
which plays for e3-e4. These moves are discussed in more detail below.
2 ... e6 3 Nc3 Nf6
Black can also play Alatortsev’s 3 ... Be7, though I don’t recommend this route for
newbies to the Queen’s Gambit Declined. You still have to face a form of Exchange
Variation with 4 cxd5 exd5 5 Bf4 and at the same time lose the important ... Nbd7 and ...
d5xc4 option against the 4 Nf3 and 5 Bf4 plan because Black has already committed the
bishop to e7.
If you meet 3 ... Be7 as White I suggest just going into a standard Exchange Variation
with 4 Nf3 Nf6 5 cxd5 exd5 (or 5 ... Nxd5 6 e4!) 6 Bg5, when 6 ... c6 7 Qc2 g6 requires
special treatment. White has three different choices on his eighth move: 8 e3, 8 Bxf6,
and 8 e4 (see Games 1-3).
4 cxd5 exd5 5 Bg5 Nbd7
 Question: Hang on a second, doesn’t this allow 6 Nxd5 winning an important
pawn?

Answer: No, because Black can meet that with 6 ... Nxd5! 7 Bxd8 Bb4+, when
White has nothing else than 8 Qd2, which gives back the queen and leads to the loss of a
piece. Black can also play 5 ... Be7, but as White is still falling into this trap (as a
database search will confirm!) it’s worth giving him some rope.
6 e3 Be7 7 Bd3 c6 8 Nf3
This is the most common destination for this knight. Part of the reason is that White
frequently puts the knight here before capturing on d5 and entering the Exchange
Variation.
White does have another important option here in 8 Qc2, which delays the decision
on where to put the king’s knight for another move.
8 ... 0-0

Question: Can Black play 8 ... Nh5 as he often seems to do in other variations?

Answer: Here it’s not such a good idea because of 9 Bxe7 Qxe7 10 0-0, and now if
Black gets his king out of the centre with 10 ... 0-0 White can play 11 Qb1,
simultaneously attacking h7 and preparing to play b2-b4. This is a far more efficient
version of the minority attack than many we will see.
9 Qc2 Re8 10 0-0
This looks obvious but it’s not the only move. Besides castling on opposite sides
with the aggressive 10 0-0-0 (as in Game 5), White can stay flexible with 10 h3. This
often transposes into the 10 0-0 Nf8 11 h3 line, though occasionally White gets the idea
to castle queenside (Game 6 is an example of this).
10 ... Nf8

11 h3
This is now the main line at grandmaster level, having taken over from more
traditional moves.

Question: What’s the idea?

Answer: It’s a subtle semi-waiting move that was brought into the limelight by
Anatoly Karpov. There are a number of effects in that 11 ... Ne4 is well met by 12 Bf4
and 11 ... Ng6 can be answered by 12 Bxf6, followed by 13 b4. Meanwhile the g4-
square gets taken away from Black’s minor pieces, which is important because he often
plays ... Ng4 in answer to Ne5 and will sometimes drive a white knight from f3 with ...
Bg4.
Here’s a round-up of the alternatives:
a) 11 Rab1 aims for the traditional minority attack with b2-b4-b5, creating some
sort of pawn weakness on the queenside which White then hopes to attack. All the
games I’ve featured have Black replying with 11 ... a5, to hold up White’s planned
advance, which for the most part is then met by 12 a3. In turn I like 12 ... Ng6 for Black,
preventing White’s bishop from going back to f4, when Black plays ... Ne4 on his next
move. White then has a choice: the obvious 13 b4 often sees White sleepwalk to defeat
as he does in Games 7 and 8, so 13 Bxf6 is better (as in Game 9), though here too Black
is not without his chances.
Apart from 12 ... Ng6, there is another interesting move in 12 ... g6, which intends a
regrouping with ... Ne6-g7 and then a possible exchange of light-squared bishops with
... Bf5. Sune Berg Hansen specializes in this plan and we can see a good example in
Game 10.
Note that, after 11 ... a5, White changed plans with 12 Rbe1 in Game 11. This is a
valid thing to do, though it’s difficult to see much advantage for White in having Black’s
a-pawn on a5 instead of a7.
b) 11 Bxf6 is another way of going for a minority attack, as after 11 ... Bxf6 White
can play an immediate 12 b4. The cost is that it gives away the bishop pair, which Black
used very effectively in Game 12. Particular attention should be given to the way Black
met 15 b5 with 15 ... a5 in this game, holding up White’s operations on the queenside
due to Black’s grip on the b4-square.
c) 11 Ne5 is a move which needs to be met accurately as a Stonewall attack plan
with f2-f4 could be very dangerous against passive play. The key move for Black is 11
... Ne4, immediately exchanging this knight off. Game 13 provides an excellent example
of how Black should play these positions.
d) 11 Rae1 is quite a good move to play at club level. White has two dangerous
plans in mind – e3-e4, or Ne5 followed by f2-f4 – and Black needs to play accurately to
obtain a decent game. The right way is 11 ... Ne4, after which 12 Bxe7 Qxe7 13 Bxe4
dxe4 14 Nd2 f5 15 f3 is covered within Games 14 and 15.
11 ... Be6
 As with 11 h3, this is a very modern treatment.

Question: What’s Black’s plan?

Answer: Besides developing his pieces he’s also preparing to bring his queen’s
rook to c8. This will mean that if White plays a minority attack with b2-b4-b5, he can
just push past with ... c6-c5.

Question: Won’t that leave Black with a weak d5-pawn?

Answer: The d-pawn tends to be quite easy to protect; the issue is more the d4-
square which White will gain as a possible outpost for a knight. But Black has
compensation here in that he gets squares along the c-file such as c5 and c4.
A more traditional response to 11 h3 is 11 ... Ng6, which prevents the retreat of
White’s bishop to f4 in preparation for 12 ... Ne4. In Game 16 White then chose 12
Bxf6, attempting to show that the insertion of 11 h3 and 11 ... Ng6 provide an improved
version of the 11 Bxf6 plan. They do, but Black’s position remains quite playable.
One more interesting option for Black is to meet 11 h3 with 11 ... g6, which is what
happened in Game 17. It looks as if there might have been some very serious opening
preparation behind Black’s 23 ... Bxd4! in this game.
 12 Rab1
The standard move, aiming for a minority attack with b2-b4-b5. But White has tried
various alternatives here:
a) 12 a3 looks like an inferior way of preparing b2-b4 as White may later want to
force b4-b5 through with a3-a4, apparently losing a tempo. But there is a subtle point to
this in that 12 ... Ne4 is no longer possible because of 13 Bxe7 Qxe7 14 Nxe4 dxe4 15
Bxe4 (whereas after 12 Rab1, Black would have 15 ... Bxa2). With 12 ... Ne4 out of the
question Black should play either 12 ... N6d7 (as in Game 18) or 12 ... Rc8 (as in Game
19).
b) 12 Rfc1 is Karpov’s original treatment, just improving the position of his pieces
before adopting a particular plan. This is illustrated in Game 20.
12 ... Rc8
Preparing to meet White’s b2-b4-b5 with ... c6-c5.
Black has a good alternative in 12 ... N6d7 which is covered in Game 21. As noted
above, Black can also play the surprising 12 ... Ne4 here (as in Game 22) because 13
Bxe7 Qxe7 14 Nxe4 dxe4 15 Bxe4 is met by 15 ... Bxa2!.
13 b4 Nh5
Ensuring the exchange of dark-squared bishops by stopping White’s bishop from
going back to f4.
14 Bxe7 Qxe7 15 Rfc1 Ng6 16 Ne2 a6
 We are following I.Sokolov-K.Asrian, Stepanakert 2005 (Game 23), in which White
soon gave up on the minority attack; but when he turned his attention to the kingside he
could make little progress there either.

Game 1
Bu Xiangzhi-J.Rowson
Turin Olympiad 2006

1 Nf3 d5 2 d4 Nf6 3 c4 e6 4 cxd5

Capturing on d5 at this stage is a sensible and practical option for White, not least
because it avoids sharp lines such as 4 Nc3 dxc4, not to mention the Semi-Tarrasch with
4 ... c5 and others. Black has some extra options because of this early capture, which
we will explore here and in the next couple of games. Yet at club level it has been very
noticeable to me that Black usually carries on his usual plan of development with ... Be7
and ... Nbd7.
4 ... exd5 5 Nc3 c6 6 Bg5
There are actually some subtle differences between this move and 6 Qc2!, and in the
position after 5 ... c6 I do think the queen move is more accurate. On the other hand, I
suggest that White meets 5 ... Be7 with 6 Bg5 rather than 6 Qc2.
6 ... Be7
Black should play 6 ... Bf5! in this situation, as he doesn’t even need to prepare the
bishop move with ... g7-g6. This is why White should prefer 6 Qc2 over 6 Bg5.

Question: Can’t White play 7 Qb3, attacking the b7-pawn?

Answer: Indeed he can, though Black then has 7 ... Qb6 (the dynamic 7 ... Nbd7!? 8
Qxb7 Rb8 9 Qxc6 Rxb2 is also interesting and was successful in A.Karpov-
G.Kasparov, blitz match, Valencia 2009) 8 Bxf6 gxf6 9 e3 Na6 10 Qxb6 axb6 11 Kd2
Nc7 12 Bd3 Be6 13 a3 b5 14 Nh4 Kd7 15 Kc2 Ne8 16 Nf5 Nd6 was fine for Black in
F.Berkes-Ki.Georgiev, Serbian Team Championship 2008, his two bishops and potential
pressure on the a-file compensating for the weakness of his kingside pawns.
7 Qc2 g6
 Now this is necessary if Black wants to put his bishop on f5.

Question: Doesn’t this weaken Black’s kingside?

Answer: Yes, it does, though White’s main plan to exploit this weakness is to castle
queenside and shove his h-pawn up the board. So Black might be careful not to castle
kingside before White is also committed to doing so.
8 e3
This looks obvious, but it’s not White’s only move. In the next two games we will
examine 8 Bxf6!? and 8 e4, respectively.
8 ... Bf5 9 Bd3 Bxd3 10 Qxd3 Nbd7!
Carefully delaying castling short in case White lays siege to the g6-pawn with h2-
h4-h5. Indeed, after 10 ... 0-0 11 h4 Nbd7 12 Bxf6 Nxf6 13 Ne5 Bb4 14 g4 c5 15 h5
Qe7 16 hxg6 fxg6 17 f3 cxd4 18 Qxd4 Rfe8 19 f4 Rad8 20 g5 Nh5 21 Ng4, White
developed a strong attack in B.Lajthajm-B.Ivanovic, Montenegrin Team Championship
2010.
11 0-0 0-0 12 h3!?
This and White’s next move constitute an interesting plan to play in the centre and
possibly activate his central pawn majority with a later f2-f3 and e3-e4.
The traditional plan is to play for a minority attack with 12 Rab1, trying for b2-b4-
b5, but this doesn’t seem very effective here because the c4-square can become weak (
... Nb6-c4 often happens in such positions). Black can also make b2-b4 very difficult
with 12 ... a5; for example, 13 h3 Re8 14 Bf4 Bf8 15 Qc2 Qb6 16 Rfc1 Qa7 17 Ne5
Nxe5 18 Bxe5 Nd7 19 Bh2 a4 20 Ne2 Nf6 21 Nf4 Ne4 saw Black gradually
outmanoeuvre his opponent in V.Burmakin-A.Dreev, Russian Team Championship 1999.
12 ... Re8 13 Bf4

13 ... a5
This certainly inhibits b2-b4 ideas, but White has other plans.
Black has an interesting alternative here in 13 ... Nh5 14 Bh2 Ng7, aiming to get this
knight to the nice d6-square after ... Nf5 and a subsequent exchange of White’s h2-
bishop with ... Bd6. Bu Xiangzhi-J.Lautier, Internet 2004, continued 15 Rac1 Nf5 16
Nd2 a5 17 Na4 Bd6 18 Bxd6 Nxd6 19 Nb3 Qg5 20 Qe2 Re7 21 Nbc5 Nf6 22 Nc3 Nf5,
when Black was successfully inhibiting White’s plans of a central expansion with f2-f3
and e3-e4. The exchange of dark-squared bishops is a useful way to do this in fact.
14 Qc2 Nb6 15 Ne5
This makes sense, putting the knight on a good square and freeing the way for his f-
pawn to nudge forward. White has also played a preliminary 15 Rfe1 here. D.Flores-
J.Granda Zuniga, Linares (rapid) 2008, continued 15 ... a4 16 Ne5 Bd6 17 Rad1 Qe7 18
Nd3 Ne4 19 f3 Nxc3 20 bxc3 with balanced play at this point.
15 ... Nfd7 16 Nd3 Qc8
 I’m not sure why Black played this rather than the immediate 16 ... Nf8. In a couple
of moves time the queen goes back to d8 with an apparent loss of time.
17 f3 Nf8 18 Qf2 Ne6 19 Bh2 Qd8 20 Rad1 a4 21 Kh1
All these moves can be perceived as preparation for e3-e4, but White has another
idea too as we shall see. Black meanwhile seems confined to preventing White’s ideas,
at least at the moment. His next provokes another pawn move by White in the hope that
this will later prove to be a weakness.
21 ... Bh4 22 g3 Bf6

23 f4!?

Question: Isn’t that just a horrible move, turning the bishop on h2 into a big pawn
and giving Black an outpost on e4?

Answer: That’s a good question! From a structural point of view this move does
create weaknesses and the bishop on h2 is shut in. But if we look more closely at the
position it’s possible to see some positive sides. First of all White is taking space and
may be able to put a knight on e5. The e4-square is currently defended by the knight on
c3 and there are ideas such as g3-g4 and f4-f5 in the position, which might open up
Black’s king while liberating the bishop on h2. So let’s say there’s “dynamic
compensation” for the weaknesses created.
23 ... Nc4 24 Ne5 Nd6 25 g4 Nc7 26 Qf3 a3 27 b3 Qe7 28 g5

Question: I thought you said White was playing for f4-f5 in order to liberate the h2-
bishop. Doesn’t this make that impossible?

Answer: This last move is another interesting strategic conception that often
happens in Stonewall structures. The main idea is to prevent White’s knight being
driven away with ... f7-f6 while, at the same time, staking out more space on the
kingside. White also has a new way to attack using the advance of his h-pawn.
Meanwhile it should be noted that 28 f5 can be answered by 28 ... g5, keeping the
kingside closed.
28 ... Bg7 29 h4 Qe6 30 Ne2 Nf5 31 h5!?

31 ... Nb5
My engine prefers the immediate 31 ... gxh5, after which 32 Qxh5 can be answered
by the cold-blooded 32 ... Bxe5 33 fxe5 Nxe3 34 Rf6 Qg4. It turns out that 35 Qxf7+? is
then a really bad move because of 35 ... Kh8 36 Rf2 Rf8, which isn’t easy to see when
playing 31 h5. Had White done so there might have been a case for the solid 31 Bg3.
32 Rd2 Nbd6 33 Nc3 gxh5?!
 This pawn grab is now very risky with the engine preferring 33 ... Nh4.
34 Qxh5 Nxe3
The logical follow-up to Black’s previous move, but storm clouds now gather
around his king.
35 Rf3 Nef5 36 Rh3 Ne4
Perhaps 36 ... h6 was marginally better, though it’s still very scary for Black after 37
Rg2. This kind of position is especially difficult to play near the time control and under
time pressure.
37 Qxh7+ Kf8 38 Nxe4 dxe4 39 Bg1 Ra5
At this point Black should perhaps head for the hills with 39 ... Ke7.
40 Re2
40 g6 was also possible; White evidently preferred to keep this in reserve.

40 ... Nd6
Understandably wanting to support the e4-pawn. It is equally understandable that
Black wished to avoid 40 ... Ke7, putting the king opposite White’s newly posted rook
on e2. Yet it turns out that this would have been Black’s best move, resulting in
approximate equality after 41 Rxe4 Rxe5 42 Rxe5 Bxe5 43 fxe5 (43 dxe5? Qd5+ 44
Kh2 Qd2+ 45 Kh1 Qxf4 is good for Black) 43 ... Qd5+ 44 Kh2 Nxd4 45 Bxd4 Qxd4 46
Qf5.
 41 Reh2
According to the engine, 41 Bf2 was objectively stronger.
41 ... Qf5??
Wanting to trade queens but allowing White to fan the flames of his attack. 41 ... Nf5
was the right way, keeping firm control over the position.
42 g6! Rd5
Losing immediately, but there was no good defence to threats like 43 Rh5 and 43
gxf7 in any case.
43 Rh5 f6
After 43 ... Rexe5 44 dxe5 Qxg6 45 exd6, Black has nothing for his lost rook.
44 Rxf5 Nxf5 45 Qh3 1-0

Game 2
S.Atalik-J.Wawrzaszek
Stillwater 2011

1 d4 d5 2 c4 c6 3 Nf3 Nf6 4 Nc3 e6 5 cxd5

A noteworthy and common way to get an Exchange Variation of the Queen’s Gambit
Declined. After 5 Bg5, Black can go for the famous and difficult Botvinnik Variation
with 5 ... dxc4 6 e4 b5 7 e5 h6 8 Bh4 g5, which has been very heavily analysed.
5 ... exd5 6 Qc2!
Here 6 Bg5 is less accurate because of 6 ... Bf5, as I pointed out in the notes to
Game 1.
6 ... g6
For 6 ... Na6 see Game 4. Meanwhile, 6 ... Be7 7 Bg5 g6 is just a transposition.
7 Bg5 Be7 8 Bxf6!?

Question: That looks odd. Why is White giving away the bishop for nothing?

Answer: There are several reasons. First of all, White will be committing his
pawns to dark squares after 9 e3, which means that this is not a particularly ‘good’
bishop. And secondly, by capturing on f6 straight away White forces Black to put his
bishop on f6, when he’d really prefer to have his b8-knight on that square and the bishop
on d6.
Instead, 8 e3 was covered in Game 1 and 8 e4 will be examined in Game 3.
8 ... Bxf6 9 e3 Bf5 10 Bd3 Bxd3 11 Qxd3
 11 ... Nd7
Black can delay White’s b2-b4 advance with 11 ... Be7, but then Atalik suggested 12
h4!?, looking to create play on the kingside.
12 b4
Black can now answer 12 h4 by evacuating his king to the queenside with 12 ... Qe7
13 h5 0-0-0. This leaves White’s main plan as the traditional minority attack.
12 ... a6 13 0-0 0-0 14 a4 Qe7
This doesn’t look right, as in many positions Black should prefer to have his bishop
on d6.
A typical counter to White’s minority attack was 14 ... Nb6, when 15 b5 can be
answered by 15 ... cxb5 16 axb5 a5, creating a passed a-pawn as compensation for the
weakness of d5.
Another plan is 14 ... b5, intending to bring the d7-knight to b6 and c4. White
probably then does best to open the position with 15 e4, after which the backward pawn
on c6 may become exposed and in need of protection.
 15 Rab1

Question: Can’t White push on with his minority attack with an immediate 15 b5 - ?

Answer: That is certainly something to consider, especially as Black can’t push past
with 15 ... c5. But in this case Black can cause some annoyance with 15 ... axb5 16 axb5
Ra3, making it difficult for White to unravel.
15 ... Nb6 16 Nd2 Bg7 17 Qc2 f5 18 Ne2
It’s important to prevent Black’s own minority attack based on levering open the e-
file with ... f5-f4.
18 ... Rac8 19 Nf4 Qf7 20 Nd3
The dark side of Black’s ... f7-f5 is underlined by this move: the e5-square is a
weakness and a white knight is poised to land there.
20 ... g5?!
Evidently feeling uncomfortable with the position, Black lashes out. Instead, 20 ...
Nc4 was preferable, after which 21 Rbc1 (or 21 Nxc4 dxc4 22 Nc5 f4) 21 ... b5 22 Nf3
Rfe8 23 Ra1 is only marginally better for White.
21 Nc5
21 f4 was worth considering, to get a grip on the e5-square; White probably
rejected it because of the weakness of e3 after 21 ... gxf4 22 Nxf4 Rce8.

21 ... Rc7
White can meet 21 ... f4 with 22 a5, after which 22 ... Nc4 (22 ... Na8 23 Qd3
would see White playing for e3-e4 while the black knight is badly misplaced) 23 Nxc4
dxc4 24 Rfe1 leaves Black in serious trouble because of the weakness of the c4-pawn
and the strength of White’s knight on c5.
22 a5 Nc8 23 Nf3 h6
Here 23 ... Qe7 was a better way to defend the g5-pawn. White is definitely for
choice after 24 Nd3 Nd6 25 Nfe5, but the problem remains as to how he could break
through given that b4-b5 is no longer possible (because of a4-a5) and e3-e4 will be
difficult to achieve.
24 Ne5 Bxe5?!
This is bad for Black, though it’s getting hard to give good advice. After 24 ... Qe8
25 f4 g4, trying to create a blockade, White continues with 26 h3 h5 27 hxg4 hxg4 28
Kf2 and can then transfer a rook (or two) to the h-file.
25 dxe5 Qe7 26 f4 Na7
Aiming for the outpost on b5, but meanwhile there are more pressing needs on the
other side of the board.
27 g4!
 27 ... gxf4
After 27 ... Nb5 White can merrily sacrifice the exchange with 28 gxf5 Na3 29 Qd3
Nxb1 30 f6, as his passed pawns constitute an overwhelming advantage.
28 exf4 Kh8 29 gxf5 Qh7 30 Ne6!
Looking to give up the queen for overwhelming compensation.
30 ... Rxf5 31 Nxc7 Rg5+ 32 fxg5 Qxc2
 33 Rbe1
Here 33 e6! was stronger, when 33 ... Nc8 34 Ne8 Qd3 35 Rf8+ Kh7 36 Nf6+ Kg6
37 Rg8+ Kf5 38 Rf1+ Kxe6 39 Re8+ Ne7 40 Re1+ Kf5 41 R8xe7 leaves White with a
decisive material advantage. Another solution is 33 Rf8+! Kh7 34 Rbf1, intending 35
Ne6, R1f7+ and mates, while 34 ... Qe2 35 Ne6 Qg4+ 36 Kh1 Qe4+ (36 ... Qxe6 loses
the queen after 37 R1f7+ etc) 37 R1f3 only slows White down temporarily.
33 ... Qg6?
Going down without much of a fight. 33 ... Qd2! was the most tenacious move, after
which White should play 34 Ne8 Qxg5+ 35 Kh1 with what should be a winning attack.
34 e6
Again not quite the best; White should have played 34 Rf8+! Kg7 (or 34 ... Kh7 35
Ne8 Qxg5+ 36 Kh1 with too many threats) 35 Ref1 with a winning attack.
34 ... Nc8 35 Rf8+ Kh7 36 Rf7+ Kg8
Now 36 ... Kh8 37 e7 will force Black to give up his knight as well.
37 Ne8 Ne7 38 Nf6+ Kh8 39 Rh7+
The simplest route to victory.
39 ... Qxh7 40 Nxh7 Kxh7 41 Rf1 Kg6 42 gxh6 Kxh6 43 Rf7 1-0

Game 3
A.Naiditsch-D.Pavasovic
European Championship, Rijeka 2010

1 d4 d5 2 c4 e6 3 Nf3 c6 4 Nc3 Nf6 5 cxd5 exd5 6 Qc2 g6

The alternative 6 ... Na6 will be covered in Game 4.
7 Bg5 Be7 8 e4
A favourite line of Anatoly Karpov which often leads to the exchange of queens. For
8 Bxf6 see Game 2, while 8 e3 was seen in Game 1.
8 ... dxe4 9 Bxf6

Question: Why is White giving up a bishop like this, especially in a position that
has just opened up?

Answer: He’s doing so to gain time and keep his pieces active. 9 Nxe4 Nxe4 10
Qxe4 Be6 would already be at least equal for Black.
9 ... Bxf6 10 Qxe4+

10 ... Qe7
In practice this has been the most popular choice but it is not Black’s only option:
a) 10 ... Kf8 also seems reasonable for Black; for example, 11 Bc4 Kg7 12 0-0 Re8
13 Qf4 Be6 14 Bxe6 Rxe6 15 Rfe1 Qd6! 16 Qxd6 Rxd6 17 Ne4 Rd8 18 Nxf6 Kxf6 19
g4 Nd7 20 g5+ Kg7 21 Rad1 Re8 22 d5 Rxe1+ 23 Nxe1 c5 24 f3 Re8 25 Kf2 f6 saw
Black neutralize White’s pressure in D.Neelotpal-P.Negi, Indian Championship,
Aurangabad 2011.
b) 10 ... Be6, on the other hand, seems to leave White with an edge after 11 Bc4 Qe7
12 Bxe6 Qxe6 13 Qxe6+ fxe6 14 Ne4, when 14 ... Ke7 15 0-0-0 Na6 16 h4 Bg7 17 h5
gxh5 18 Rxh5 Rad8 19 Ne5 was unpleasant for Black in S.Lputian-Y.Dokhoian,
Blagoveshchensk 1988.
11 Bc4 Bf5 12 Qxe7+
White has also tried 12 Qe3, inviting Black to exchange queens on e3, when f2xe3
lends the d4-pawn some protection. But after 12 ... Nd7 13 0-0 0-0 (13 ... Qxe3 14 fxe3
0-0 15 e4 caused Black some problems in D.Orzech-K.Miton, Polish Team
Championship 2009) 14 Qf4 (thus far D.Orzech-J.Parramon Guillaumet, Balaguer 2009)
and now 14 ... Qb4!, Black gets excellent counterplay; e.g. 15 Bb3 a5 16 g4 a4 17 Nxa4
Bd3 18 Rfd1 Be2.
Certainly White doesn’t seem to get much with 12 Qe3, though it might be useful as
a surprise weapon.
12 ... Kxe7 13 0-0-0 Rd8 14 h3

This move, intending g2-g4, is White’s most dangerous option here. The natural-
looking 14 Rhe1+ is relatively harmless; for example, 14 ... Kf8 15 h3 h5 16 Ne2 Be4
17 Ne5 Bd5 18 Bxd5 cxd5 19 Kd2 Nd7 20 f4 Be7! 21 Rc1? and now Black missed a
trick in 21 ... Nxe5!, winning material after 22 dxe5 Bb4+ or 22 fxe5 Bg5+, though 21 ...
Bb4+ 22 Nc3 Nf6 23 Kd3 Rac8 24 Rc2 Ne4 saw him obtain excellent counterplay
anyway in A.Riazantsev-S.Sjugirov, Novokuznetsk 2008.
14 ... Nd7
14 ... Bg7 has done well in practice but looks difficult for Black after 15 Ng5 Bh6
16 f4 with ideas such as Rhe1+, g2-g4 and Nxf7.

Question: What happens if Black stops White’s g2-g4 plan with 14 ... h5 - ?

Answer: White can actually play 15 g4! anyway, when 15 ... hxg4 is too dangerous
because White’s h1-rook gets into play along the h-file: 16 hxg4 Bxg4 17 Rh7! Rf8 18
Re1+ Kd8 19 Ne5 gives him a powerful initiative which more than compensates for the
sacrificed pawn. So Black does best not to take the g-pawn, but 15 ... Be6 16 Rhe1 hxg4
17 hxg4 Na6 18 Bxe6 fxe6 19 Ne4 still left her struggling in Shen Yang-M.Sebag, FIDE
Grand Prix, Nanjing 2009, because of the weakness of the pawns on e6 and g6.
15 Rhe1+ Kf8 16 g4 Nb6 17 Bb3

17 ... Bc8

Question: Does Black really have to pull the bishop all the way back like that?
Can’t he move it to d7 instead?

Answer: 17 ... Bd7 may be even more difficult for Black; for example, 18 Ne4 Be7
(18 ... Kg7 can still be met by 19 Nxf6 Kxf6 20 g5+ Kg7 21 Re7 Nd5 22 Bxd5, when
Black has lost the option to recapture on d5 with his rook and 23 Ne5 is coming after 22
... cxd5) 19 Neg5 Bxg5+ (19 ... Bb4 20 Nxf7 Rdc8 21 Re4 left Black with nothing for
the pawn in J.Maiwald-D.Baramidze, Austrian League 2008) 20 Nxg5 Be8 21 Nxh7+
Kg7 22 Ng5 a5 23 a3 and White was just a pawn up in N.Khurtsidze-S.Melia, Georgian
Women’s Championship, Tbilisi 2010.
18 Ne4 Kg7 19 Rd3
It does make sense to put the rook on the third rank like this, but it might be
somewhat on the slow side. The direct 19 Nxf6!? seems like the critical line; for
example, 19 ... Kxf6 20 g5+ Kg7 21 Re7 Nd5 22 Bxd5 Rxd5 (22 ... cxd5 23 Ne5 is very
unpleasant) 23 Re8 b6 (White also has pressure after 23 ... Rf5; e.g. 24 Rd3 h6 25 h4 b6
26 gxh6+ Kxh6 27 Ne5 Kg7 28 Nxc6 Rxf2 29 d5 leaves Black having to deal with the
powerful passed d-pawn) 24 Ne5 c5 25 Re1 cxd4 26 Nxf7! d3 27 Kd2 Rf5 28 Nd6
Rxf2+ 29 Kc3 Re2 30 R1xe2 dxe2 31 Kd2 Bb7 32 Re7+ Kf8 33 Rxb7 Rd8 34 Rf7+ 1-0
was G.Estevez Morales-J.Diaz, Cuban Championship, Camaguey 1988.
19 ... Nd5?
 This leads to serious difficulties. Black should have played 19 ... Be7, when 20
Neg5 Nd5 (20 ... Bb4 can be met by 21 Nxf7 Rd7 22 Re4 Rxf7 23 Bxf7 Kxf7 24 d5
with a strong initiative, and if 24 ... Nxd5 then 25 Rxd5 cxd5 26 Rxb4 gives White a
pull because of the nice blockade square in front of Black’s isolated d-pawn) 21 Bxd5
Bxg5+ 22 Nxg5 Rxd5 reaches what looks like an equal position.
20 Bxd5 Rxd5
After 20 ... cxd5 21 Nxf6 Kxf6 22 g5+ Kg7, White gets his rook in with 23 Re7 Kf8
24 Rc7 Bxh3 25 Rxb7 Rd7 26 Rdb3 and Black is struggling.
21 Nxf6 Kxf6 22 Re8

W
This awkward pin on the c8-bishop causes Black to try and shoot his way out of
trouble, but in doing so he drops the pawn on a7.
22 ... Bf5 23 Rxa8 Bxd3 24 Rxa7 Ba6
Protecting b7 and trying to shut White’s rook in, at least on a temporary basis.
Nevertheless, the more active 24 ... Be4 would have been a better chance; for example,
after 25 Ne1 Rxd4 26 Rxb7 Bd5 27 b3 Rf4 28 f3 Bxf3 29 a4 Bd5 30 a5 c5 31 Rb6+
Ke7, Black probably has enough to draw, despite the strength of White’s passed a-
pawn.
25 b4 h5 26 Kd2 hxg4 27 hxg4 Rd8 28 Kc3
The immediate 28 a4 was good, but there’s nothing wrong with Naiditsch’s choice
either.
28 ... Ke6 29 Ne5 g5
Black’s last chance was to get his rook active with 29 ... Rh8 30 f3 Rh1, though after
31 a4 Ra1 32 Nxc6 Rc1+ 33 Kb3 Rxc6 34 b5 Bxb5 35 axb5 Rc7 36 b6 Rd7 37 Kc4,
his survival prospects look rather remote.
30 a4 Rc8 31 Nc4

Preparing to bring the knight back to the nice e3-square. Meanwhile White is
confident about his chances in the rook endgame following 31 ... Bxc4 32 Kxc4.
31 ... Rb8 32 Ne3 Kd6
Going down without much of a fight. Black should still probably try to get active
with 32 ... Be2, though 33 Nc4!? f6 (33 ... Bxg4 34 Na5 wins either the b7- or c6-pawn)
34 Na5 b5 35 Nxc6 Rc8 36 Kd2 Rxc6 37 Kxe2 Rc4 38 Ra6+ Ke7 39 Kd3 Rxb4 40 a5
leaves White with all the chances in the rook endgame.
33 b5! cxb5 34 axb5 Bxb5 35 Ra5 Rc8+ 36 Kb4 Bd7 37 Nc4+ Ke7 38 Re5+ Kf8
On 38 ... Kf6, there follows 39 Nb6 Rd8 40 Rd5 Ke6 41 Kc5 Ke7 42 Rxg5, which
is winning for White.
39 Nb6 Rd8 40 Rxg5 Be8
After 40 ... Be6 41 Kc5 f6 42 Rh5 Kg7 (or 42 ... Bxg4 43 Rh8+ Ke7 44 Nd5+ etc)
43 Rh1, intending Rg1 and d4-d5, Black is also completely lost.
 41 Kc5 f6 42 Rf5 Kg7 43 d5 1-0

Game 4
R.Vaganian-D.Rogozenco
German League 1996

1 Nf3 Nf6 2 c4 c6 3 Nc3 d5 4 d4 e6 5 cxd5 exd5 6 Qc2 Na6

This is a major alternative to 6 ... g6 7 Bg5 Be7 (or 6 ... Be7 7 Bg5 g6), which was
covered in the first three games.
7 Bg5
White can also rule ... Nb4 ideas out with 7 a3, but this is arguably a suboptimal use
of time, even if it does prepare b2-b4. Garry Kasparov played a couple of games as
Black in this line which serve as a good model: 7 ... Nc7 8 Bg5 g6 (preparing the
exchange of the light-squared bishop, as with 6 ... g6) 9 e3 Bf5 10 Bd3 Bxd3 11 Qxd3
Be7 12 0-0 0-0 13 b4 Ne4 14 Bf4 Nxc3 15 Qxc3 (White should eliminate Black’s
knight with 15 Bxc7! Qxc7 16 Qxc3, when a draw was agreed in R.Vaganian-
G.Kasparov, Horgen 1995) 15 ... Bd6! 16 Bxd6 Nb5! 17 Qb3 Nxd6 left Black
somewhat better in L.Portisch-G.Kasparov, GMA World Cup, Skelleftea 1989, as the
beautifully placed knight on d6 effectively prevents White’s minority attack; meanwhile
Black can build pressure along the e-file and against White’s queenside. Kasparov went
on to win in classic style: 18 a4 a6 19 Ne5 Re8 20 Rfe1 Qg5 21 h3 Kg7 22 Qc2 Re6 23
Rac1 Rae8 24 Qb1 Qh5 25 Qb3 f6 26 Nd3 g5 27 Qd1 Qg6 28 Qc2 R6e7 29 Red1 h5 30
Qb1 h4 31 Qc2 g4 32 Nf4 Qxc2 33 Rxc2 g3 34 Rd3 Kh6 35 Kf1 Kg5 36 Ne2 Nc4 37
Rcc3 Nb2 38 Rd2 Nxa4 39 Rb3 Nb6 40 Ng1 Nc4 41 Nf3+ Kh5 42 Rdd3 a5 43 bxa5
Ra8 44 Rd1 Rxa5 45 Re1 b5 46 Re2 Ra1+ 47 Re1 Rea7 48 fxg3 Rxe1+ 49 Kxe1 Ra1+
50 Ke2 hxg3 51 Ne1 Ra2+ 52 Kd1 Rd2+ 53 Kc1 Re2 54 Kd1 Rxe3 55 Rxe3 Nxe3+ 56
Ke2 Nf5 57 Nc2 Nh4 58 Nb4 Nxg2 59 Kf3 Nh4+ 60 Kxg3 Nf5+ 61 Kf4 Nxd4 62 Ke3
Nf5+ 0-1. This game is well worth studying as it provides a wonderful model for
playing Black.
7 ... Be7

Question: What happens if Black plays 7 ... Nb4, followed by ... Bf5 - ? Doesn’t
that cause White problems?

Answer: That’s a good question, as this has in fact been played. I.Novikov-Z.Franco
Ocampos, Saint Vincent 1999, saw 8 Qb1 g6 9 Qd1! Bf5 10 Rc1, when White was
ready to begin ejecting the gatecrashers with a2-a3. The game continued 10 ... a5 11 a3
Na6 12 Qb3 Bc8 13 e4! Be7, and now 14 e5 Ne4 15 Bh6 would have put Black in
serious trouble.
8 e3 Nb4 9 Qb1 g6 10 Qd1
 Question: Hasn’t White just lost a tempo by making those two queen moves?
Shouldn’t he have played 9 Qd1 straight away?

Answer: By throwing in 9 Qb1 first White provoked his opponent into playing 9 ...
g6, which is far from being a useful move. For one thing, these positions sometimes
require Black to lift a rook to g6 or h6 via e6 in order to create counterplay against
White’s king; clearly a pawn on g6 gets in the way of this plan.
10 ... Bf5
In a later game, V.Gavrikov-J.Ehlvest, Kalev (rapid) 1997, Black preferred to avoid
this obvious bishop move and opted for 10 ... a5 instead. The problem is that White can
now play 11 a3 Na6 12 Bd3, developing his bishop on d3 without having it exchanged.
After 12 ... Nc7 13 0-0 Bg4 14 h3 Bxf3 15 Qxf3, White stood slightly better.
11 Rc1 a5
Black creates the possibility of retreating his knight back to a6 (after White’s a2-a3)
without contracting doubled pawns (after Bxa6), but he now gets driven back, with
White obtaining a very promising position. The critical move was 11 ... Qa5, after
which it seems likely that Vaganian would have kept the game going with 12 Bxf6 (here
12 Qd2 Nxa2! 13 Bxf6 Nxc1 14 Bxe7 Qa1 15 Ba3 Nd3+ 16 Ke2 Nc1+ 17 Ke1 is a
known drawing variation) 12 ... Bxf6 13 Qd2, intending a2-a3, and if 13 ... Nxa2? then
14 Ra1 wins a piece.
12 a3 Na6 13 Qb3 Qb8
Not the best way to defend b7, since a5 becomes weak. 13 ... Rb8 was a better idea.
14 Na4 0-0
Rainer Knaak suggested that Black should consider getting the dark-squared bishops
off here with 14 ... Nd7!? 15 Bxe7 Kxe7, implying that White would not then be able to
exploit the central position of Black’s king. This may well be the case as it is difficult to
open the centre. Even so, I think we have to prefer White after, say, 16 Nh4 Be6 17 Bd3.
15 Ne5
The pin on the f6-knight is now very unpleasant for Black, so much so that it allows
White to play quite expansively with a kingside pawn advance.
15 ... Rd8 16 Be2 Kg7 17 0-0 Nc7 18 g4!
 Question: I thought it was supposed to be unwise to move pawns in front of your
king like this?

Answer: Most of the time that is correct, but here the passive disposition of Black’s
forces allows White to play very ambitiously.
18 ... Be6 19 f4
The threat of f4-f5 is a serious problem for Black, who now tries for an exchange of
pieces rather late in the day.
19 ... Ng8 20 Bxe7 Nxe7 21 Nb6
Taking the opportunity to displace the rook on a8 before proceeding with his
kingside attack.
21 ... Ra7 22 f5
Opening up Black’s king at the cost of a mere pawn.
22 ... gxf5 23 gxf5 Nxf5 24 Bd3
 24 ... Nd6
My engine likes the materialistic 24 ... Nxe3, at least at first. After running for a
while it finds a brilliant forced win for White: 25 Bxh7!! Nxf1 (25 ... Kxh7 26 Qxe3
gives White a decisive attack without any material investment) 26 Rxf1 Kxh7 27 Rxf7+
Bxf7 28 Qh3+ Kg7 29 Qg3+ Kh7 (29 ... Kf8 30 Qf4 is immediately decisive) 30 Qh4+
Kg7 31 Qg5+ Kh7 32 Qf5+ Kh6 33 Qf4+ Kh5 34 Qxf7+ Kg5 35 Qg6+ Kf4 36 Kf2,
followed by 37 Qg4 mate or 37 Nd3 mate.
25 Qd1
Rerouting the queen towards Black’s vulnerable kingside.
25 ... Rg8 26 Kh1 Ne4
If 26 ... Kh8 then 27 Qh5 wins; e.g. 27 ... Rg7 28 Rg1 Qg8 29 Rxg7 Qxg7 30 Rg1
etc.
27 Bxe4 dxe4 28 Nbd7 Qd8
28 ... Bxd7 allows 29 Rxf7+ etc.
29 Qh5
 29 ... f5 30 Rg1+ Kh8 31 Rxg8+ Qxg8 32 Rg1 Qe8 33 Qg5 1-0
Black is no longer able to stave off mate.

Game 5
E.Bacrot-A.Karpov
Cannes (1st matchgame) 2000

1 d4 Nf6 2 c4 e6 3 Nc3 d5 4 cxd5 exd5 5 Bg5 Be7 6 e3 c6 7 Qc2 Nbd7 8 Bd3 0-0 9
Nf3 Re8 10 0-0-0
 Question: That’s different. What’s White’s idea with this move?

Answer: It’s a very sharp choice by Bacrot, presumably designed to unsettle his
mighty opponent. Positions with the kings castled on opposite sides often feature a
violent race of attacks.
10 ... Nf8 11 h4!? Be6!
This and Black’s next move aim for counterplay along the c-file with ... c6-c5.

Question: Can Black delay ... c7-c6 earlier in order to save a tempo?

Answer: Yes indeed, and we’ll be looking at that in two Yusupov games later on.
Karpov probably chose 6 ... c6 in this game because he wanted to meet White’s Bd3 and
Nge2 plan with an early ... Nh5.
12 Kb1 Rc8 13 Ka1 a6

Question: Why didn’t Black play 13 ... c5 straight away?

Answer: Partly because he’s concerned about a disruptive Bb5, and partly because
... b7-b5 will be a useful addition to Black’s arsenal of attacking ideas.

14 Qb1
Moving the queen away from the gaze of Black’s c8-rook in anticipation of the ...
c6-c5 break.

Question: I can understand White moving the queen from the c-file, but why choose
such a passive-looking square?

Answer: The only other reasonable squares for the queen are e2 and d2, but after 14
Qe2 Black can play 14 ... Bg4, and after 14 Qd2 the release of pressure against h7
allows Black to activate his f8-knight with 14 ... N8d7.
14 ... b5 15 Rc1
Once again preparing for Black’s ... c6-c5 break.

Question: I thought that in positions with opposite side castling you have to develop
your attack as quickly as possible. So aren’t White’s last two moves a bit slow?

Answer: The big thing is to develop your attack more quickly than your opponent,
which can mean a combination of punching and ducking. White’s last two moves can be
viewed as ducking, in a way that it makes it difficult for Black to get his attack going.
15 ... h6 16 Bf4 c5 17 dxc5 Bxc5 18 Nd4 Bd7 19 f3

Keeping the f6-knight out of e4 and preparing a possible g2-g4-g5.

19 ... Qb6
Karpov finds the most testing line, as one might expect. After 19 ... Nh5 20 Bh2
Rxe3 White can get his pawn back with 21 Nb3 Ba7 22 Nxd5.
20 Rhd1 Bxd4
Karpov goes for a clean and clear approach. At first the engines like 20 ... Ne6 21
Be5 b4, but their enthusiasm cools considerably after 22 Nce2 Nxd4 23 exd4.
21 exd4 b4 22 Ne2 Bb5 23 Rxc8 Rxc8 24 g4 Re8 25 Nc1
Black can meet with 25 g5 with 25 ... Rxe2; for example, 26 gxf6 Qxf6 27 Bxe2
Bxe2 28 Be5 Qxf3 picks up a piece and two pawns for the sacrificed exchange.
25 ... h5
 Karpov thought this was his best move of the game as it neutralizes the danger on the
kingside.
26 g5 N6d7 27 Bf5 Ng6 28 Bxg6 Qxg6
Keeping the queens on with 28 ... fxg6 was worth considering; for example, 29 Qc2
Rf8 30 Qd2 Qe6 would also leave White struggling. Karpov’s decision to exchange
queens was consistent with his love of simple positions and endgames.
29 Qxg6 fxg6 30 Bd2 a5 31 Nb3
Black could meet 31 Re1 with 31 ... Rxe1 32 Bxe1 Nb8, followed by 33 ... Nc6,
when White is still struggling in the minor piece endgame.
31 ... a4
Here 31 ... Ba4 was possible, though White seems to be okay after 32 Rc1 Re2 33
Bf4.
32 Nc5 b3
 33 Re1
And not 33 axb3?! axb3 34 Nxb3?? because of 34 ... Ba4!, winning a piece.
33 ... Rxe1+ 34 Bxe1 Nxc5 35 dxc5 Bd3 36 axb3 axb3 37 c6 Kf7 38 c7 Bf5 39
Bc3 Ke7 40 c8Q Bxc8 41 Kb1 Ke6 42 Kc1 Kf5 43 Kd2 Kf4 44 Ke2 Kg3 45 Be1+
Kg2 46 Ke3 ½-½

Game 6
D.Bocharov-S.Azarov
European Championship (play-off), Warsaw 2005

1 d4 Nf6 2 c4 e6 3 Nf3 d5 4 Nc3 Be7 5 Bg5 0-0 6 cxd5 exd5 7 e3 Nbd7 8 Bd3 Re8 9
Qc2 Nf8 10 h3 c6
 11 Bf4
The immediate 11 0-0-0 is well met by 11 ... Ne4!; for example, 12 Bxe7 Qxe7 13
Bxe4 dxe4 14 Nd2 f5 15 Kb1 Be6 16 Ka1 (16 g4 is the critical move, but Black seems
to be okay after 16 ... Nd7 17 gxf5 Bxf5 18 Qb3+ Be6 19 Qxb7 Rab8 20 Qa6 Qb4 21
Nb3 Bxb3 22 axb3 Qxb3) 16 ... Rad8 17 g4 c5 18 gxf5 cxd4 19 fxe6 dxc3 was roughly
even in D.Berczes-Z.Ilincic, Budapest 2005.

Question: Doesn’t 11 0-0 transpose into the 10 0-0 c6 11 h3 line?

Answer: It does indeed, and that was well spotted. We’ll examine this line later in
the chapter (Games 16-23).
11 ... Ng6 12 Bh2 Bd6 13 Bxd6 Qxd6
 14 0-0-0

Question: Can White still castle short and try for a minority attack?

Answer: Yes he can, though Black is quite well placed to attack on the kingside, and
White’s h2-h3 may well prove to be a weakness. G.Timoscenko-L.Ftacnik, Czech
League 2008, went 14 0-0 Qe7 15 Rfe1 Ne4 16 Nd2 Nxd2 17 Qxd2 Nh4 (threatening 18
... Bxh3) 18 Qe2 g6 19 Kh1 Qg5 20 Qf1 Bf5 21 Rad1 Rad8 22 Bxf5 Nxf5 23 g3 Re6 24
h4 Qe7 25 Kg2 Re8 26 Qd3 h5 27 Re2 Nd6 28 Rde1 Kg7 29 Rc2 g5! 30 Rh1 Rg6 31
Kf1 Qf6 32 hxg5 Qf3 33 Rh3 Nf5 34 Nd1 h4 with a decisive attack.
14 ... Be6 15 Kb1 Rac8
Simple and logical play; Black will get counterplay against White’s king position by
proceeding with ... c6-c5.
16 g4
White can also try to inhibit Black’s queenside play with 16 Rc1 before developing
his own attack. Black usually meets this with 16 ... b6, when 17 g4 (17 Ba6 Rb8 18 Bd3
was a quick draw in Al.David-V.Inkiov, Paris 2004) 17 ... c5 18 Ba6 Rb8 19 Nb5 Qd8!
(19 ... Qd7 20 Ne5 Nxe5 21 dxe5 Ne4 22 f3 was good for White in I.Khenkin-T.Halasz,
Recklinghausen 2001) 20 dxc5 bxc5 21 Qxc5 Ne4 leads to intricate play where both
sides have chances.
 16 ... c5 17 dxc5 Qxc5

18 Nd4
Instead, 18 Rc1 Qa5 19 Nd4 was seen in S.Plischki-J.Zeberski, Olomouc 2005, and
now 19 ... Bd7 (rather than 19 ... a6, as played) 20 Bf5 Ne5 21 g5 Ne4 22 Bxe4 dxe4 23
Qxe4 Nc6 gives Black a strong initiative for the sacrificed pawn.
18 ... Ne5?!
A natural but flawed idea, probably played quickly seeing as this was a rapid play-
off game. The right move is 18 ... Bd7, after which 19 Bf5 Bxf5 20 Qxf5 Nh4! (20 ...
Ne7 21 Qd3 Nc6 22 Nb3 was a bit better for White in V.Lazarev-D.Tyomkin, North
Bay, Ontario 1999) 21 Qd3 Ne4 leaves Black with fully equal play.
19 f4?!
Missing an opportunity in 19 g5! Nxd3 20 gxf6! Nb4 21 Qe2, when Black’s king
position is looking distinctly shaky. After the mistaken 19 f4?! Black gets excellent play.
19 ... Nxd3 20 Qxd3 Ne4! 21 Nxe4 dxe4 22 Qd2
Not 22 Qxe4?? because of 22 ... Bxa2+, winning the queen; but White might have
been better off with 22 Qb5, forcing Black to lose time if he wants to avoid the
exchange of queens.
22 ... Bc4 23 Rc1 Qd5 24 b3 Bd3+
 Both sides have a great minor piece here.
25 Kb2 h5 26 g5 a5 27 Rc3
Giving Black time to open the a-file. 27 Rxc8 Rxc8 28 Rc1 was safer, with a
position that can only be assessed as equal.
27 ... a4 28 Rhc1 Rxc3
At a fast time limit we can’t expect perfection. 28 ... Ra8! would have been more
testing.
29 Qxc3 Qd7 30 f5 axb3 31 Qxb3 b5 32 Rc6 Bc4
32 ... Ra8! was better here too.
33 Qb4 Qd8 34 h4 Qb8 35 Qd6 Qa7 36 a3 b4! 37 axb4 Qa2+ 38 Kc3 Ra8 39
Nc2
 39 ... Qb3+
Objectively, 39 ... Bb3 was correct, when it’s still about equal. But it looks like the
players were in a time scramble in which White’s flag went down.
40 Kd2 Bd3 0-1
White is now better, so presumably this was a loss on time.

Game 7
P.Trajkovic-B.Maksimovic
Serbian Championship, Kragujevac 2000

1 c4 e6

Question: Can I also get into a Queen’s Gambit Declined via 1 ... Nf6 2 Nc3 e6,
followed by 3 ... d5 - ?

Answer: Theoretically, yes, but this way allows White to play a Flohr-Mikenas
Attack with 3 e4, which adds a lot of complexity to what Black needs to know.
2 Nc3 d5 3 d4 Nf6 4 cxd5 exd5 5 Bg5 Be7 6 e3 0-0 7 Bd3 c6 8 Qc2 Nbd7 9 Nf3
Giving up the last chance to develop the knight on e2.
 9 ... Re8 10 0-0 Nf8

This is a key position for the QGD Exchange Variation, in which White has a
number of plans and even more possible moves.
11 Rab1
There are two other ways to go about preparing the minority attack: 11 Bxf6 Bxf6 12
b4 (see Game 12) and 11 a3. The latter attempt was met by 11 ... a5 in A.Beliavsky-
J.Pinter, Hungarian Team Championship 2002, after which 12 Rae1 (12 Rab1
transposes back into the main game) 12 ... Be6 13 Ne5 N6d7 14 Bxe7 Rxe7 15 Nxd7
Bxd7 16 Na4 Be8 17 Rc1 h6 18 Rfd1 b6 19 e4 Ne6 saw Black hold the balance.
11 ... a5

Question: Can’t Black play 11 ... Ne4 immediately?

Answer: He can and indeed has, but then 12 Bf4 avoids the immediate exchange of
bishops and leaves Black having to do something about the knight on e4, which is now
attacked. His best may be 12 ... Ng5, after which 13 Bxg5 Bxg5 14 b4 a6 15 a4 Be7 16
b5 axb5 (trying to block White’s queenside initiative with 16 ... a5 here is strongly met
by 17 bxc6 bxc6 18 Ne5!) 17 axb5 Bd6 18 bxc6 bxc6 19 Na4 Bg4 20 Nd2 left Black
with an unpleasant position due to his weakness on c6 in E.Gausel-M.Ehrke, Gausdal
1998.
12 a3 Ng6 13 b4

Question: I understand that 11 ... a5 was to try and stop White’s minority attack with
b2-b4, but as White has pushed through with this anyway, what was the point in playing
it?

Answer: Essentially there are two major arguments for 11 ... a5. Firstly, Black gets
the open a-file after an exchange of pawns on b4, and secondly he will not be left with
an isolated a-pawn if White plays b4-b5 and then takes on c6. There’s also a chance that
White can be left with a weak b4-pawn, especially if Black blocks it with ... b7-b5.
Alternatively, White can prevent ... Ne4 by playing 13 Bxf6, a move that we’ll be
looking at Game 9.
13 ... axb4 14 axb4 Ne4 15 Bxe7 Qxe7 16 b5
This all makes perfect sense; White’s minority attack is reaping concrete benefits in
terms of weakening Black’s queenside and opening files there. Yet at the same time he
needs to be quite wary about Black’s attacking chances on the kingside, as we shall see.
16 ... Bg4 17 Nd2?
This natural move is a surprisingly serious mistake. The fact that all White’s moves
have been so logical means that there’s a reasonable chance of this position coming
about if you play the Queen’s Gambit Declined as Black. White needs to insert 17 Bxe4
here, as in the next game.
17 ... Nxd2 18 Qxd2 Nh4

A really horrid move for White to have to meet, and in fact there’s no adequate way
to do so. The main threat is 19 ... Bh3, attacking g2, and White won’t be able to capture
the bishop because of either ... Qg5+, followed by mate on g2, or the knight fork on f3.
19 Be2
Looking at White’s alternatives, it seems that nothing can save him:
a) 19 Ne2 Bh3! 20 Nf4 Bxg2! 21 Nxg2 Nf3+ wins White’s queen.
b) 19 Rfc1 Bh3 20 Bf1 Qg5 threatens both the g2-pawn and ... Nf3+, and after 21 f4
Nf3+ 22 Kh1 Nxd2 23 fxg5 Nxb1 White loses the exchange.
c) 19 bxc6 bxc6 20 Qc2 Bh3 21 g3 Qf6! 22 Be2 Nf3+ 23 Kh1 Bxf1 also wins the
exchange.
d) 19 f3 Qxe3+ 20 Qxe3 Rxe3 wins a pawn because of the threat against the bishop
on d3. Nevertheless, this last line might be White’s best chance as he could conceivably
avoid defeat in the endgame.
19 ... Bh3 20 g3
The databases are full of different death agonies by White. 20 e4 Bxg2 21 Rfe1 Bf3
22 Qf4 dxe4 23 bxc6 Ng2 24 Qg3 Nxe1 25 Rxb7 Qd8 soon won for Black in
L.William-Art.Kovacs, Budapest 2012.
 20 ... Bxf1
Black can also play a preliminary 20 ... Qf6, threatening 21 ... Nf3+, before
capturing the rook on f1. For example, 21 Qd1 Bxf1 22 Bxf1 Nf3+ 23 Kg2 Ng5 24 bxc6
bxc6 25 Rb6 g6 26 Nxd5? Qd8 0-1 was M.Grabarczyk-K.Jakubowski, Polish
Championship, Warsaw 2003.
21 Kxf1 Nf5 22 Bd3 Nd6

Black is just the exchange up now, so we’re into the realm of technique. Black will
win by gradually improving his position and finding new targets.
23 Qc2 g6 24 Na4 Ne4 25 Bxe4 Qxe4 26 Qxe4 Rxe4 27 Nc5 Re7 28 b6
Here 28 bxc6 bxc6 29 Rb6 may be White’s best, hoping to tie Black down to the c6-
pawn, which is what White often tries to do in such positions when he has even
material. With Black having an extra exchange, he’s the one with all the chances.
28 ... Ra2 29 g4 f5 30 gxf5 gxf5 31 Rd1 Ra5
Preparing an exchange of rooks with a later ... Rb5, after which the win becomes a
lot easier.
32 Rb1 Kf7 33 Kg2 Kf6 34 Rb2 Rb5 0-1
 If 35 Rxb5 cxb5 36 Kf3 b4 then 37 Ke2 f4 is hopeless for White, and otherwise
Black will follow with 37 ... Re6, since the pawn endgame after 38 Nxe6 Kxe6 is a
simple win.

Game 8
D.Radovanovic-A.Savanovic
Bosnian Team Championship 2012

1 d4 d5 2 c4 e6 3 Nc3 Nf6 4 cxd5 exd5 5 Bg5 c6 6 e3 Nbd7 7 Bd3 Be7 8 Nf3 0-0 9 0-
0 Re8 10 Qc2 Nf8 11 Rab1 a5 12 a3 Ng6 13 b4 axb4 14 axb4 Ne4 15 Bxe7 Qxe7
 16 b5
Slightly different positions come about if White plays 16 Bxe4 at once. For
example, 16 ... dxe4 17 Nd2 f5 18 b5 Be6 19 bxc6 (it seems that this is Black’s most
accurate move) 19 ... bxc6 20 Rb6 c5! (and here this seems to be the most precise; 20 ...
Bd5 and 20 ... Rec8 have also been tried) 21 Nb5 (21 d5 Bf7 leaves Black with the nice
e5-square for his knight) 21 ... Ra2 22 Qc3 Nh4 23 Kh1 Rd8 24 g3 Nf3 25 Nxf3 exf3 26
Rc6 f4 27 Kg1 Bh3 28 Rc7 Qf6 29 Re1 was D.Magalashvili-A.Adly, World Junior
Championships, Heraklio 2004, where 29 ... Re2! would have given Black a winning
attack; e.g. 30 Rxc5 fxg3 31 hxg3 Qh6, intending to replace the bishop on h3 with the
queen.
16 ... Bg4 17 Bxe4!
Undoubtedly the right way for White to play this position. For the mistaken 17 Nd2?,
see the previous game.
17 ... dxe4 18 Nd2 f5 19 bxc6 bxc6
 20 Rb6
Another way to attack c6 is via 20 Qb3+ Kh8 21 Qb7, after which 21 ... Qg5 saw a
draw offered and accepted in P.Kiss-T.Polak, Hungarian Team Championship 2008. The
position is clearly a very complex one where we can have a lot of fun working out
different attacking possibilities for Black and defences for White. I offer the following
sample line as a starting point for the reader’s own investigations: 22 Kh1 Nh4 23 Rg1
Bh5 24 h3 Qf6 25 Nc4 Qe6 26 Ne5 c5 27 Nd5 cxd4 28 exd4 Qd6 29 Nc7 Qxd4 30
Nxa8 Qxe5 31 Nc7 Rg8 32 Nb5 Nf3 33 gxf3 Bxf3+ 34 Rg2 Bxg2+ 35 Kxg2 e3 36 fxe3
Qxe3 37 Qf3 Qxf3+ 38 Kxf3 g5 and the endgame should be drawn.
20 ... Qg5 21 g3 Qh5 22 Rfb1
White has to be very careful here. The text vacates f1 for the queen so as to defend
the threats against his king. Instead, 22 Rxc6 is strongly met by 22 ... Qh3, threatening 24
... Bf3, when 23 f3 exf3 24 Nxf3 Bxf3 25 Rxf3 f4 would see Black develop a strong
initiative.
22 ... Rac8 23 Rb7 Bf3
 24 Nxf3
It’s not easy for White to defend here. 24 Qa2+ Kh8 25 Qc4 Ne7 26 h4 Qg4 27 Kh2
Ng6 leaves him facing the brutal threat of 28 ... Nxh4.
24 ... exf3 25 Kh1
Now if 25 Qa2+ Kh8 26 Qf7 Rg8 (threatening 27 ... Qh3) 27 Qc4, then 27 ... f4! 28
exf4 Nxf4 gives Black dangerous threats on the kingside.
25 ... Kh8
Anticipating checks on the a2-g8 diagonal. 25 ... c5 was also worth considering,
eliminating the weak pawn on c6 and trying to get e5 for his knight.
26 Qa2 c5 27 Nb5?
This looks natural enough, but precise calculation is what is really required. 27 Qd5
Qh3 28 Qxf3?? fails to 28 ... Nh4! 29 Qd5 Rcd8 30 Qc6 Re4 31 Rg1 Nf3 32 Rg2 Nxh2!
etc. White’s best defence seems to be 27 Qf7 Rg8 28 Nd5, when 28 ... cxd4 29 exd4
Qh6 is adequately answered by 30 Ne3.
27 ... Qh3 28 Rg1
 28 ... cxd4
Not bad but not the best. Black can win by playing the immediate 28 ... Rc6!,
intending to route the rook to h6. For instance, after 29 Qf7 Rg8 30 Qc4, Black has 30 ...
Ne5! 31 dxe5 Rh6, forcing the losing 32 Qh4.
29 Nxd4 Ne5 30 Rbb1?
Underestimating the power of Black’s reply. White had to play 30 Qb2, when 30 ...
Rc6?? 31 Nxc6 Ng4 is very wrong because of 32 Qxg7 mate. Instead, Black is still
doing well after 30 ... h6, but the result would not be a foregone conclusion.
30 ... Rc6 0-1
Now after 31 Nxc6 Ng4, White is faced with inevitable mate.

Game 9
P.Peelen-P.Van der Sterren
OHRA-B, Amsterdam 1989

1 c4 e6 2 Nc3 d5 3 d4 Nf6 4 cxd5 exd5 5 Bg5 Be7 6 e3 0-0 7 Qc2 Nbd7 8 Bd3 Re8 9
Nf3 c6 10 0-0 Nf8 11 Rab1 a5 12 a3
 12 ... Ng6

Question: Can Black forestall White’s next move by playing 12 ... Ne4 at once?

Answer: He has tried this, but 12 ... Ng6 performs the important function of
preventing the bishop’s retreat to f4. This in fact happened in the game Z.Gyimesi-
P.Horvath, Hungarian Team Championship 1998, which continued 13 Bf4 Nd6 14 Ne5
g6 15 b4 axb4 16 axb4 Bf5 17 b5 f6 18 Nf3 Bxd3 19 Qxd3 Ra3 20 Bxd6 Bxd6 21 Ra1
Qa8 22 Rxa3 Qxa3 23 Qc2 Bb4 24 Nb1 Qa8 25 Rc1 Rc8 26 Qb3 and White had a
promising queenside initiative.
13 Bxf6

Question: Why is White giving up the bishop pair like that?

Answer: It’s a radical way of preventing Black’s freeing 13 ... Ne4 which did so
well in the previous two games.

Question: Is it a good idea?

 Answer: It leads to a somewhat different kind of position where Black’s attack on
the kingside is delayed, though the two bishops have value. For example, White’s
minority attack with b4-b5 can often be answered by ... c6-c5, as in this game.
13 ... Bxf6

14 b4 Bg4 15 Nd2 axb4

Van der Sterren adopts the simplest and most direct approach, opening the a-file for
his queen’s rook. He also prevents White opening the b-file with 16 bxa5, which should
be played after other Black tries such as 15 ... Be7 and 15 ... Qd6.
In both cases White will capture on b7 in return for the loss of the two a-pawns and
will therefore obtain a superior structure because of the weak pawn on c6. A sample
line is 15 ... Qd6 16 bxa5 Rxa5 17 Rxb7 Rxa3 18 Rfb1 with a small but persistent
advantage for White.
16 axb4 Be7

Question: Why does Black retreat his bishop when it has already been developed
on f6?

Answer: There are two major reasons. First of all, the bishop is not very effective
on the long diagonal and does little apart from bite on the granite d4-pawn; whereas it
will stand very well on d6, from where it menaces White’s kingside.
The second reason is in its specific relevance to coming pawn levers. Black is
ready to meet White’s b4-b5 push with ... c6-c5, which in fact is what now happens.
17 b5 c5!

Question: Doesn’t that leave Black a weak pawn on d5?

Answer: The timing of this excellent move is very important. It does indeed leave
Black with an isolated d-pawn but there are certain compensations. First of all, the
position is starting to open up for Black’s two bishops. Secondly, he may well gain the
c5-square because White’s b-pawn has advanced further than it would like to, and this
pawn also takes the b5-square away from White’s minor pieces.
18 Bf5
18 dxc5 Bxc5 19 Nb3 Bb6 would be nice for Black, whose dark-squared bishop is
well entrenched on the b6-square.
18 ... cxd4
Black has another good line in 18 ... Bxf5 19 Qxf5 cxd4 20 exd4 Qa5. Van der
Sterren prefers a move order which allows White to obtain opposite-coloured bishops,
but Black has the initiative here too.
 19 Bxg4
Otherwise 19 exd4 Bxf5 20 Qxf5 Qa5 transposes to the previous note.
19 ... dxc3 20 Qxc3 Bf6 21 Qb4 d4
Now White gets something close to equality. 21 ... Ne5 was probably a better idea.
22 exd4 Bxd4 23 Rbd1 Qf6 24 Bf3
This seems good at first but the bishop proves to be awkwardly placed here. 24 Nf3
was better.
24 ... Rad8 25 Ne4
In playing 24 Bf3 White might have missed the strength of 24 ... Rad8 and the fact
that 25 Bxb7 would be met by 25 ... Bc3!.
25 ... Qb6 26 Rd2
Instead, 26 Nc3 Nh4 doesn’t look pleasant for White, though he seems to be hanging
on after 27 Bd5.
26 ... Nh4 27 Rfd1 Nxf3+ 28 gxf3 Be5

29 Rxd8
It’s easy to understand White’s enthusiasm for exchanges, which often bring a draw
that much closer; but in this situation he would have been better off keeping more pieces
on with 29 Rd7.
29 ... Rxd8 30 Rxd8+ Qxd8 31 Qc5 Bd4 32 Qd6?
 The minor piece endgame is bad for White. Rather than continue his policy of
exchanges, White should have kept queens on, say with 32 Qc4.
32 ... Qxd6 33 Nxd6 b6 34 f4 Kf8 35 Nf5
Centralizing the king with an immediate 35 Kf1 seems better. Black, on the other
hand, plays in a very purposeful manner and his king soon races to the middle of the
board.
35 ... Bc5 36 Ng3 Ke7 37 Kg2 Ke6 38 Ne4 Be7 39 Nc3 f5

This is a sign that things have gone horribly wrong for White. He really needed his
pawns on light squares and have his knight protecting the dark ones.
40 Kf3 Bf6 41 Nd1 Bd4 42 h4 h5 43 Ne3
Going into a lost pawn endgame, though it isn’t easy to give White good advice now.
43 Ke2 Kd5 will soon see the b5-pawn fall.
43 ... Bxe3 44 fxe3 Kd5 45 Ke2 Kc4
White could have resigned with a clear conscience here. Perhaps he wished to play
it out for the benefit of the spectators.
46 Kf3 Kxb5 47 e4 fxe4+ 48 Kxe4 Kc4 49 Kf5 b5 50 Kg6 b4 51 f5 b3 52 Kxg7 b2
53 f6 b1Q 54 f7 Qg1+ 55 Kh7 Qf2 56 Kg7 Kd5 57 f8Q Qxf8+ 58 Kxf8 Ke6! 0-1
A final accurate move, shouldering off the white king which cannot now reach f4
when Black takes the h-pawn, so Peelen resigned.
Game 10
T.Kaasen-S.B.Hansen
Lund 2016

1 d4 Nf6 2 c4 e6 3 Nc3 d5 4 cxd5 exd5 5 Bg5 c6 6 Qc2 Be7 7 e3 Nbd7 8 Bd3 0-0 9
Nf3 Re8 10 0-0 Nf8 11 Rab1 a5 12 a3 g6

Question: That’s an odd-looking move. What’s the idea?

Answer: It’s a quite different treatment to 12 ... Ng6. Black is shelving any kingside
ambitions in order to carry out a sophisticated plan to exchange light-squared bishops.
He wants to bring his f8-knight to e6 and then g7 in order to prepare ... Bf5. The move
12 ... g6 is a vital part of this: first to prevent White grabbing the h7-pawn after Bxf6
when Black moves the f8-knight; secondly, to provide a square for this knight on g7; and
thirdly, to support the bishop coming to f5.
This is clearly a very beautiful idea and Sune Berg Hansen has revealed a great
fondness for it, though it does take a lot of time.
13 b4 Ne6 14 Bh4
 14 Bxf6 is also possible, but after 14 ... Bxf6 15 b5 Black can play 15 ... c5! with a
very good position already.
14 ... Ng7
Black tried 14 ... axb4 15 axb4 b5 in M.M.Pavlovic-V.Malakhatko, Bar 2005, but 16
Ne5 (16 Rfd1 Bb7 17 e4 dxe4 18 Nxe4 Nd5 was good for Black in the game) 16 ... Bb7
17 f4 makes this look quite dubious, as the f4-f5 thrust poses serious danger to Black’s
king.

15 Ne5

Question: Why did White not continue his minority attack with 15 b5 at this point?

Answer: Actually, I don’t know, because it looks very promising here. For example,
15 ... Bf5 (15 ... Bxa3 16 bxc6 bxc6 17 Ne5 is very difficult to meet) 16 bxc6 bxc6 17
Ne5 Rc8 (17 ... Qd6 is strongly met by 18 Rb6) 18 Rb7 Be6 19 Qa4 gave White a
powerful initiative in Noa.Omar-L.Guerra Tulcan, Istanbul Olympiad 2012.
15 ... axb4 16 axb4 Bf5
The logical continuation of Black’s plan, exchanging the light-squared bishops.
17 b5 c5
And this in turn is a typical counter to White’s b4-b5.
 18 dxc5 Bxc5 19 Bxf5 Nxf5??
This could have been proved costly had White now found some hidden tactics.
Black should have taken the knight with 19 ... Rxe5, when 20 Bd3 Ngh5 is roughly
equal.
20 Bxf6 Qxf6

21 Nd3??
Missing his chance for 21 Nd7!, when 21 ... Qd6 is met by 22 Nxd5! and wins.
Neither of the knights can be taken because of 23 Nf6+, and meanwhile Black’s bishop
on c5 is hanging.
21 ... Qd6 22 Nxd5 Qxd5 23 Nxc5 Rac8 24 Rbc1 Nh4 25 e4?
Another mistake, after which Black is better. This time White had to find the
ingenious 25 Ne4!, when 25 ... Qe6 (25 ... Rxc2 26 Nf6+ Kg7 27 Nxd5 is good for
White) 26 Nc3 Qg4 27 f3 Qg5 28 Qd2 Qxe3+ 29 Qxe3 Rxe3 30 Nd5 Rxc1 31 Nxe3
fizzles out to a likely draw.
25 ... Qh5 26 b6
Evidently designed against the threat of 26 ... b6, but White would have been better
off playing 26 g3, when 26 ... b6 27 gxh4 Rxc5 28 Qd3 at least offers him more drawing
chances.
26 ... Re5
 Not just threatening the knight on c5, as we shall see.
27 Nd7?

Missing Black’s lethal riposte. Here White had to try 27 Qd3, when 27 ... Rexc5 28
Rxc5 Qxc5 29 g3 Qd4! 30 Qe2 traps the errant knight on h4. Even so, this would still be
very good for Black after 30 ... Re8 31 gxh4 Rxe4 32 Qf3 Re7, winning either b6 or h4.
27 ... Nf3+! 0-1
Instantly decisive. White resigned in view of 28 gxf3 (or 28 Kh1 Qxh2 mate) 28 ...
Rg5+ 29 Kh1 Qxf3 mate.

Game 11
B.Esen-A.Korobov
Aeroflot Open, Moscow 2015

1 Nf3 Nf6 2 c4 e6 3 Nc3 d5 4 d4 Nbd7 5 cxd5 exd5 6 Bg5 c6 7 e3 Be7 8 Qc2 0-0 9
Bd3 Re8 10 0-0 Nf8 11 Rab1
For the immediate 11 Rae1 see Games 14 and 15.
11 ... a5 12 Rbe1

Question: That looks odd, why didn’t White play 11 Rae1 in the first place if he
wants to push the e-pawn?

Answer: White evidently thinks that Black’s 11 ... a5 is worse than useless against
the Rae1 and e3-e4 plan. Actually, he does have a point, as in some lines a white knight
may come to c4 or d5 and then be able to hop into b6.
12 ... Be6 13 Ne5 N6d7 14 Bxe7 Rxe7 15 f4

Question: That looks like a horrible move by White. Doesn’t he know he’s creating
weaknesses along the e-file with this wild advance?

Answer: I’m sure he knew what he was doing. This is in fact a standard plan for
White in such positions; the e-file weaknesses can’t usually be exploited (at least not
yet) and meanwhile White is gaining some useful space. You’ll see the same plan in
Game 13, but starting with 11 Ne5 in that case.
Instead, in J.Pinter-R.Kasimdzhanov, Mainz (rapid) 2003, White tried to exploit the
sensitivity of the b6-square with 15 Nxd7 Bxd7 16 Na4, but Black defended fairly
comfortably: 16 ... b6 17 Rc1 Re6 18 Bf5 Rh6 19 Bxd7 Nxd7 20 Qf5 Rf6 21 Qg4 Rg6
22 Qf5 Qe7 (playing to win!) 23 Rc2 Rf6 24 Qg4 Rg6 25 Qf5 Rf6 ½-½.
15 ... f6 16 Nf3 Rc8 17 f5
The advance of the white f-pawn has left Black rather cramped and short of squares
for his minor pieces.
17 ... Bf7 18 g4

Announcing his intention to go for a direct attack along the g-file, so Black needs to
generate rapid counterplay. White can also consider 18 e4 at this point, though Black
then gains the d5-square for his minor pieces after 18 ... dxe4 19 Nxe4 Nb6.
18 ... c5!
Not only good, absolutely necessary. Black needs to find some counterplay and this
is the only way to get it.
19 Qf2 Qc7 20 h4 Rce8 21 Kh1 cxd4 22 exd4 Qf4
Throwing a spanner in the works. White can get rid of the annoying queen by
exchanging it off, but then his attacking chances dissolve.
23 Rxe7 Rxe7 24 Nh2 Qxf2 25 Rxf2 Nb8
 Although Black remains cramped, the previous exchanges have reduced the
seriousness of this issue.
26 Re2 Rxe2 27 Bxe2
White is still better in the minor piece ending, but Black should have enough to
draw. The game only reaches an unexpected conclusion after a huge lapse by White.
27 ... Nc6 28 Nf3 g6 29 Bb5 Ne7 30 fxg6 Bxg6 31 h5
This looks like a first sign of excessive ambition coupled with shortage of time on
the clock. In theory this move gains space, but it also puts the pawn on the same colour
as his remaining bishop.
31 ... Bf7 32 Kg2 f5 33 gxf5 Nxf5 34 Bd3 Ne7 35 h6 Ne6
 36 Nb5??
An outright blunder, presumably as a result of time trouble. 36 Kf2 looks drawish.
36 ... Nf4+ 0-1
Simply winning the bishop on d3.

Game 12
M.Mozny-D.Dumitrache
Odorheiu Secuiesc 1995

1 Nf3 d5 2 c4 e6 3 d4 Nf6 4 Nc3 Nbd7 5 cxd5 exd5 6 Bg5 c6 7 Qc2 Be7 8 e3 0-0 9
Bd3 Re8 10 0-0 Nf8 11 Bxf6
This is a simple and direct way of getting b2-b4 in, reminiscent of the 11 Rab1 a5
12 a3 Ng6 13 Bxf6 sequence we saw in Game 9. The similar drawback is that it hands
Black the bishop pair and, as a result, greater influence on the dark squares.
11 ... Bxf6 12 b4 a6 13 a4
Renewing the idea of b4-b5 after Black prevented it with his previous move.
13 ... Bg4 14 Nd2 Be7 15 b5
 15 ... a5!

Question: So Black doesn’t open the a-file this time? Why is that?

Answer: This is actually a very interesting and instructive move which throws a
spanner in the works of White’s queenside ambitions. Black introduces the possibility
of ... Bb4 to block the b-file at an opportune moment.
Note that 15 ... axb5 16 axb5 Bd6 17 bxc6 bxc6 leads to a pawn structure which is
typical for the QGD Exchange. After 18 Bf5 Qg5 19 Bxg4 Qxg4 20 Nf3 Rxa1 21 Rxa1
Re6 22 Ne2 Rg6 23 Ng3, Black was struggling in A.Omelja-S.Nelyubin, Yuzhny 2009.
16 bxc6
After 16 h3 Bh5 17 bxc6 bxc6 18 Ne2 Bd6 19 Rfe1 (19 Qxc6 Rc8 20 Qb5 Rb8 21
Qc6 Rc8 is a draw by repetition) 19 ... Rc8, White was finding it hard to make any
headway in J.Fedorowicz-L.Christiansen, US Championship, Seattle 2000; and
following 20 Ng3?! Bxg3 21 fxg3 Bg6 22 Bxg6 hxg6 23 Nf3 Qd6 24 g4 c5, Black could
even claim to have slightly the better of it.
16 ... bxc6
 17 Na2
Evidently trying to stop ... Bb4 ideas while hitting the pawn on c6 with the queen;
but the knight is poorly placed on a2, and c6 is easily defended. Having said that, White
hasn’t had much joy with other moves either:
a) 17 Ne2 Bb4!? 18 Nb3 was played in A.Yuneev-V.Akopian, Daugavpils 1989, and
now 18 ... g6 followed by 19 ... Ne6 looks rock solid.
b) 17 Rfc1 Bd6 18 Bf5 (18 h3 Bd7 is fine for Black) 18 ... Bxf5 19 Qxf5 Rb8 20
Rab1 Rb4 (that b4-square again!) 21 Qc2 Ne6 22 Na2 Rxb1 23 Rxb1 c5 24 dxc5 Bxc5
was at least equal for Black in M.Skalski-K.Jakubowski, Polish Team Championship
2008.
17 ... Bd6 18 Rfc1 Qf6 19 Nf1 Rac8 20 Rab1
 20 ... Ng6
Here 20 ... c5! was already good. Instead, Black chooses to marshal his forces on
the kingside.
21 Rb6 Ba3 22 Rcb1 Nh4 23 R1b3??
Probably thinking that the bishop has to move, when White can attend to the dangers
on the other flank. Unfortunately for him, it doesn’t.
23 ... Nxg2!!
 A real bolt from the blue, after which Black wins in all variations.
24 Rxa3
There is no saving clause; for example:
a) 24 Kxg2 Qf3+ 25 Kg1 Bh3 is the end.
b) 24 Be2 Ne1 25 Qd2 Bxe2 26 Qxe1 Bxf1 wins.
c) 24 f3 Ne1 gets f3 with check. Then Black can, if needed, attend to the threat
against his other bishop.
24 ... Ne1!
The black knight makes room for his queen with gain of tempo.
25 Qc1 Qf3 26 Qxe1 0-1
White must have belatedly realized that 26 ... Bh3 would lead to mate.

Game 13
A.Gershon-K.Asrian
World Junior Championships, Yerevan 1999

1 d4 d5 2 c4 e6 3 Nc3 Nf6 4 cxd5 exd5 5 Bg5 Be7 6 e3 0-0 7 Qc2 Nbd7 8 Bd3 c6 9
Nf3 Re8 10 0-0 Nf8 11 Ne5
 An interesting and somewhat unusual move that requires accurate play by Black.

Question: Is the idea to support the knight with f2-f4 in Stonewall style?

Answer: That is one possibility, although Black effectively rules this out with his
reply. In fact White often follows up with f2-f4 ideas without a knight being on e5; in
this case the idea is to cramp Black’s kingside.
11 ... Ng4!
Immediately forcing an exchange of the e5-knight. White can’t support it with f2-f4
because the e3-pawn would hang.
12 Bxe7 Qxe7 13 Nxg4 Bxg4 14 Rae1
Protecting e3 in preparation for 15 f4.
White has also played 14 Bf5, exchanging light-squared bishops and perhaps hoping
to proceed with a queenside minority attack. But at the same time Black gets quite
serious counter chances on the kingside; for example, Lu.Popov-S.Smagin, Yugoslav
Team Championship 1994, continued 14 ... Bxf5 15 Qxf5 Rad8 16 Ne2 Rd6 17 Ng3 Rh6
18 Rfe1 Qh4 19 h3 Qe7 20 Rab1 a5 21 a3 Rg6 22 Re2 Ne6 23 Rbe1 Qd8 24 Kh1 h6 25
Kg1 Nc7 26 Nf1 Nb5 27 f3 Rge6 28 Qd3 a4 29 Nd2 Nd6 30 Nb1 f5 31 Nc3 b5 32 Kh1
Kh8 33 Kg1 Qh4 with very strong pressure for Black.
14 ... Bh5 15 f4 f6
 16 f5
A thematic follow-up, restricting both of Black’s minor pieces. On the other hand,
White has contracted certain structural weaknesses which Black can now aim to exploit.
White has tried other moves in this position but without notable success; for
example:
a) 16 Qf2 b6 17 f5 Bf7 18 g4 c5 19 Bb5 Red8 20 Be2 Re8 21 Bf3 Rad8 22 Re2 a6
23 Rd1 c4 24 e4 dxe4 25 Nxe4 Qc7 26 d5 Nd7 led to a double-edged game in
A.Goldin-J.Dorfman, France 1993.
b) Gershon himself subsequently opted for 16 a3, aiming to restrict Black’s
queenside counterplay before proceeding on the other flank. A.Gershon-B.Antilov,
Israeli Championship, Tel Aviv 2001, continued 16 ... Bf7 17 b4 a5 18 Qb3 axb4 19
axb4 Ra7 20 Rf2 and now, rather than 20 ... Ne6, Black should probably play 20 ...
Rea8, and if 21 Qb2 then 21 ... Nd7, aiming to bring the knight to c4 and gain
counterplay on the queenside.
16 ... Bf7 17 Qf2 Rad8
In I.Ibragimov-A.Botsari, Ano Liosia 1995, Black went for an immediate 17 ... c5,
but then 18 e4! (instead of 18 a3 as played) 18 ... cxd4 19 Qxd4 dxe4 20 Bxe4 gives
White some initiative that will need careful defusing. Asrian’s move makes it harder for
White to get e3-e4 in effectively.
18 a3 a6 19 Qf4 c5 20 Re2
 20 ... b5
The immediate 20 ... c4 was also possible, but Black evidently wants to keep his
options open.
21 Rf3 c4 22 Bc2 Qd6 23 Qh4 h6 24 Rg3 Nh7 25 Rf2
Not 25 Qxh6? because of 25 ... Qxg3!.
25 ... Ng5 26 Ne2?!
White tries to strengthen his play on the kingside but this allows Black to take the
initiative on the other flank. Instead, it might have been time for 26 Qf4, hoping for a
peaceful outcome after the exchange of queens.
26 ... b4
 27 a4
After 27 axb4 Qxb4, White faces the threat of 28 ... Qe1+ and will at the very least
have to give up his b-pawn.
27 ... Re7 28 Nf4
Black could meet 28 Qg4 with the calm 28 ... Qc6, after which 29 h4 h5 30 Qf4 Ne4
puts White in a desperate plight.
28 ... Rde8 29 Re2 b3 30 Bb1
30 Bd1 Qb4 doesn’t help White much.
30 ... c3 31 bxc3 Qa3 32 Nd3?
32 Re1 would have been marginally more tenacious, but 32 ... Qa1 is still winning
for Black.
32 ... Qa1 33 Re1 Re4 0-1
After 34 Nf4 Qxc3, White’s position falls apart.

Game 14
C.Bürger-J.Klovans
Werfen 1996

1 d4 Nf6 2 c4 e6 3 Nc3 d5 4 cxd5 exd5 5 Bg5 c6 6 Qc2 Be7 7 e3 Nbd7 8 Bd3 0-0 9
Nf3 Re8 10 0-0 Nf8 11 Rae1

An old treatment favoured by the many times US Champion Frank Marshall. Mikhail
Botvinnik became an adherent of this move as well.

Question: What’s the idea?

Answer: White has two very simple and direct ideas: opening the position with e3-
e4 or putting his knight on e5 and supporting it with f2-f4.
11 ... Ne4
Understandably cutting across White’s plans.
12 Bxe7
By far the most popular move, though not the only one. A number of strong players
have tried 12 Bf4, although 12 ... Bf5! then seems very comfortable for Black; for
example, 13 h3 Bd6 14 Bxd6 Qxd6 15 Nh4 Bd7 16 Nf3 (16 Nxe4 dxe4 17 Bxe4 and 16
Bxe4 dxe4 17 Nxe4 both loses a piece to 17 ... Qe7!) 16 ... Bf5 17 Nh4 Bd7 18 Nf3 ½-
½ was B.Gulko-V.Bagirov, USSR Championship, Leningrad 1977.
12 ... Qxe7 13 Bxe4 dxe4 14 Nd2 f5
 15 f3
Opening the centre like this certainly fits in with the positioning of White’s rooks.
The pawn structure, on the other hand, lends itself to 15 d5, which would probably be
quite good for White with his rooks on c1 and d1. As it is Black can defend very
comfortably with 15 ... Bd7, after which 16 f3 (16 Rd1 can be met by 16 ... Ng6, trying
to come to e5) 16 ... exf3 17 Nxf3 cxd5 18 Nd4 Qc5 19 Nxf5 Ng6 20 Kh1 Be6 was
comfortable for Black in N.Murshed-G.Milos, FIDE World Championship, Groningen
1997.
15 ... exf3 16 Nxf3 Be6 17 b3
Keeping his options open while covering the c4-square, though this is not a
particularly productive use of time. 17 e4 is the main line and will be covered in the
next game.
17 ... Rad8 18 Na4 Nd7
Aiming to bring the knight to f6 where it touches the two holes in White’s pawn
structure on d5 and e4.
19 Rf2 Nf6 20 Qc5
Perhaps hoping that a queen exchange will bring him closer to a draw against his
illustrious opponent.
20 ... Ne4
A nice outpost for the knight.
 21 Qxe7 Rxe7 22 Rc2 Bd5 23 Ne5

Now White has his knight on a good post too, which should allow him to hold the
balance.
23 ... g6 24 Nc5 Kg7 25 Nxe4 Bxe4 26 Rc5 f4!
This opening of the position favours Black as he has a long-range bishop against a
short-stepping knight.
27 exf4 Rxd4 28 Rc4 Rxc4 29 Nxc4 Re6?
An unfortunate slip that White fails to exploit. Black should have defended his rook
with either 29 ... Kf6 or 29 ... Kf8, after which it’s definitely more pleasant to have a
bishop rather than a knight.
30 Kf2?
Missing his chance. 30 Na5! wins a pawn, as after 30 ... b6 (or 30 ... b5 for that
matter) White has 31 Nxc6! Rxc6 32 Rxe4.
30 ... Kf6 31 g3 Bf5 32 Rxe6+ Bxe6
 Question: Why didn’t they agree a draw at this point? Isn’t it just equal?

Answer: The bishop is the superior minor piece in this endgame, especially with
pawns on both sides of the board. It can strike out on both the kingside and queenside,
protecting its own passed pawns while controlling those of the opponent.
33 Ne3
33 Na5 is answered by 33 ... Bc8.
33 ... a5 34 Ke2 Ke7 35 Kd3 b5 36 Kc3 Kd6 37 a3 c5 38 b4?!
Handing Black a supported passed a-pawn, which certainly doesn’t hurt his winning
prospects. 38 Kb2 seems like a better move.
38 ... cxb4+ 39 axb4 a4
This newly emergent passed pawn will tie White’s pieces down. The question now
is whether Black can get his king in.
40 Kb2 Bh3 41 Kc3 h6 42 g4 Ke6 43 Kd4 g5 44 fxg5
White might have tried 44 f5+, though 44 ... Kd6 45 Kd3 Ke5 46 Kc3 h5 47 gxh5
Bxf5 still leaves Black with all the chances.
44 ... hxg5 45 Kc5 a3 46 Kxb5 Bxg4! 47 Kc4
The bishop can’t be taken as the a-pawn promotes after 47 Nxg4 a2.
 47 ... a2 48 Nc2 Bd1 49 Na1 Kf5

Heading towards White’s h-pawn.

50 Kd3?!
Here 50 Kc3! would have been more accurate; for example, 50 ... Kg4 (or if 50 ...
Ba4 then 51 Kb2 Kg4 52 Kxa2 Kh3 53 Nb3 g4 54 Nd4 Kxh2 55 b5 Bxb5 56 Nxb5 g3
57 Nd4 etc) 51 b5 Kh3 52 b6 Bf3 53 Kb2 Kxh2 54 Nc2 Kg3 55 Kxa2 Bb7 56 Nd4 and
White holds the draw.
50 ... Ba4 0-1
Did Bürger resign here or was this a loss on time? I’d guess the latter is what
happened. The former would have been especially unfortunate as after 51 Kc3 Kg4 52
Kb2 Kh3 53 Kxa2 Kxh2 54 Ka3 Bd7 (not 54 ... g4?! 55 Kxa4 g3 56 Nc2 g2?? 57 Ne1
and White even wins due to the fork on f3) 55 Kb3 Kg1 56 Nc2 Kf2 57 Kc4 g4 58 Kd4
g3 59 Ne3, White saves himself.

Game 15
J.Hjartarson-N.Short
Dubai Olympiad 1986

1 d4 Nf6 2 c4 e6 3 Nc3 d5 4 Bg5 Be7 5 cxd5 exd5 6 e3 0-0 7 Bd3 Nbd7 8 Nf3 Re8 9
0-0 c6 10 Qc2 Nf8 11 Rae1 Ne4 12 Bxe7 Qxe7 13 Bxe4 dxe4 14 Nd2 f5 15 f3 exf3
16 Nxf3 Be6 17 e4

Question: Doesn’t that leave White with an isolated queen’s pawn?

Answer: Well, yes, but White hopes his active pieces will more than compensate,
which is why it has become the main line. 17 b3 featured in the previous game.
17 ... fxe4
Black has played 17 ... Qd7 in a few games, but then 18 d5! is a strong reply; for
example, 18 ... cxd5 19 exd5 Bf7 (19 ... Bxd5? 20 Rd1 Rac8 21 Qf2 wins material) 20
Ne5 Qc7 21 Nxf7 Qxf7 22 Rxe8 Rxe8 23 Qxf5 left White a good pawn up in R.Garcia-
D.Roselli, Mercedes 1969.
18 Rxe4
White has tried 18 Nxe4 as well but, as the engine points out, Black has an effective
response in 18 ... Bd5!. My silicon-aided analysis runs 19 Nc5 (19 Ne5 Nd7 20 Nc3
Nxe5 21 Nxd5? Nf3+ 22 gxf3 Qg5+ is good for Black) 19 ... Qd6! 20 Nxb7 (20 Ne5
Ng6 21 Nc4 Qc7 is level) 20 ... Qb4 21 Rxe8 Rxe8 22 Nc5 Bxf3 23 Rxf3 Qxd4+ 24
Kf1 Qh4 25 Nd3 Qxh2 26 Qc4+ Ne6 27 Qxc6 Qh1+ 28 Kf2 Qh4+ 29 Kg1 Qd4+ 30
Kh2 Re7 31 Qc8+ Qd8 32 Qxd8+ Nxd8 with stone cold equality in the endgame.
 18 ... h6 19 Rfe1
Far and away White’s most popular move. The most important of White’s
alternatives is 19 Ne2, though it doesn’t particularly trouble Black. J.Timman-
A.Yusupov, Tilburg (8th matchgame) 1986, went 19 ... Qb4 20 a3 Qb3 21 Qd2 Bd5 22
Rxe8 Rxe8 23 Ne5 Ne6 24 Nc3 Qb6 25 Nxd5 cxd5 26 Nf3 Nf8 and Black had
comfortable equality.
19 ... Rad8

20 Re5
White has also played 20 R1e3, which my engine confirms is best answered by 20
... Qf7 21 Ne5 Qc7! (better than 21 ... Qf5 22 Rf3 Qh7 23 Qe2 Bd5 24 Nxd5 cxd5, as in
V.Karpman-A.Panchenko, Belgorod 1991, when 25 Ree3 would have left Black
struggling) 22 Qe2 Bd5 23 Rf4, as in Bu Xiangzhi-S.Azarov, World Junior
Championships, Athens 2001, and now 23 ... Nd7! 24 Nxd5 cxd5 25 Nxd7 Qxd7 gives
Black full equality.
Another possibility is 20 Qa4, but then 20 ... a6 21 Qa5 Qd6 22 R4e3 Re7 and 23 ...
Rde8 is fine for Black.
20 ... Qf7 21 b4?!

Question: Is that some kind of minority attack by White?

 Answer: It seems possible that b4-b5 was the idea, but the position has changed
completely from a minority attack scenario with White’s opening of the centre. In the
current situation it looks more like a simple loss of time.
Here 21 Qe4 is a better move, when M.Tal-R.Vaganian, USSR Team Championship
1975, continued 21 ... Rd6 22 Qe3 Nd7 23 Ra5, and now Black’s most solid line is 23
... a6 (in the game 23 ... Nf6 24 Qf4 Red8 25 Rxa7 left Black skating on very thin ice)
24 Ne5 Qf8 25 Ne4 Rd5 26 Rxd5 cxd5 27 Nc5 Nxe5 28 dxe5 Qe7 29 Qd4 Rc8 and
White’s initiative is fizzling out.
21 ... Nd7

22 Ra5
Hjartarson goes badly astray with this decentralization. 22 R5e3 was preferable,
when 22 ... Bg4 is still about equal.
22 ... Bg4 23 Ne5
After 23 Rf1 Black has a powerful reply in 23 ... Qc4!.
23 ... Nxe5 24 dxe5
24 Raxe5 Rxe5 25 dxe5 Qc4 is also good for Black.
24 ... b6 25 Ra3 Rd4 26 Qf2?!
White should insert 26 h3, though 26 ... Bf5 27 Qb2 Red8 still leaves him in a very
unpleasant position.
26 ... Rf4 27 Qg3 Rxe5

28 h3
Unfortunately, 28 Rxe5 allows mate on f1.
28 ... Rxe1+ 29 Qxe1 Be6 30 Qe5?
Missing Black’s next, but White was losing anyway.
30 ... Bxh3! 31 Ne4
31 gxh3 also leads to mate after 31 ... Rf1+ 32 Kh2 Qf2.
31 ... Rf1+ 32 Kh2 Rh1+! 0-1
White resigned in view of 33 Kg3 Qg6+ 34 Ng5 Bc8 and if 35 Rxa7 hxg5 36 Ra8
then 36 ... Qd3+ 37 Kf2 Rf1 mate.

Game 16
M.Grabarczyk-Y.Balashov
German League 1997

1 d4 Nf6 2 c4 e6 3 Nf3 d5 4 Nc3 Nbd7 5 cxd5 exd5 6 Bg5 c6 7 e3 Be7 8 Qc2 0-0 9
Bd3 Re8 10 0-0 Nf8 11 h3
 Question: That looks like a strange move; doesn’t it just weaken the kingside?

Answer: Actually it’s quite a subtle move, popularized by Anatoly Karpov. White
intends to meet 11 ... Ne4 with 12 Bf4 and 11 ... Ng6 with 12 Bxf6, followed by 13 b4.
11 ... Ng6

Question: Does Black have any moves apart from this and 11 ... Ne4 - ?

Answer: Yes, he can also play 11 ... Be6, which we’ll look at thoroughly in due
course (see Games 18-23).
The text move leads to positions similar to Games 9 and 12. Having this sort of
consistency in your opening repertoire means that you can develop a better
understanding of the lines you play.
Meanwhile, I’ll give just one example of 11 ... Ne4. E.Romanov-M.Gurevich,
European Championship, Plovdiv 2008, continued 12 Bf4 Ng5 13 Bxg5 Bxg5 14 b4
Be7 15 b5 Bd6 16 bxc6 bxc6 17 Rab1 Qa5 18 Qa4 Qxa4 19 Nxa4 Bd7 20 Rb7 Reb8 21
Rfb1 f6, when Black had more or less equalized.
12 Bxf6
 Question: So White is giving up the bishop pair again? Wouldn’t it have been better
to do so on the previous move?

Answer: That’s a good question which goes to the heart of White’s reasons for
playing 11 h3. White is arguing that having the moves h2-h3 and ... Ng6 included
actually favours him, since h2-h3 prevents ... Bg4 and, rather than ... Ng6, Black would
have preferred to play ... g7-g6 and ... Ne6. Having said that, the argument is not
particularly clear-cut. There are good things about having the knight on g6 and potential
drawbacks to h2-h3 too.
12 ... Bxf6 13 b4 a6 14 a4
Renewing the idea of b4-b5.
14 ... Be7 15 b5

15 ... Bd6

Question: Why doesn’t Black take the opportunity to get the a-pawns off the board,
and make sure he isn’t left with a second weakness besides the c6-pawn?

Answer: That’s another excellent question, as Black often does eliminate the a-
pawns for this reason and has also done so in this position. In this particular game
Balashov decides he wants to keep the option of playing a later ... a6-a5 and planting
his dark-squared bishop on b4

Question: Well, can’t Black play an immediate 15 ... a5 with this idea?

Answer: Yes, this has been tried too, but it appears to be slightly better for White.
Here are examples of those two moves:
a) 15 ... axb5 16 axb5 Bd7 17 bxc6 bxc6 18 Na4 Ra5 19 Nd2 Qc7 20 Nb3 Ra7 21
Nac5 Bc8 22 Rxa7 Qxa7 23 Ra1 was clearly good for White in Pr.Nikolic-
L.Ljubojevic, OHRA Tournament, Amsterdam 1988.
b) 15 ... a5 16 bxc6 bxc6 17 Ne2 Ba6?! (17 ... Bd7 would have been better, with
only the most marginal edge for White) 18 Nc1 Qb6 19 Rb1 Bb4 20 Bxa6 Qxa6 21 Nd3
Rab8 22 Nfe5 Nxe5 23 Nxe5 left Black struggling in A.Barsov-M.Becker, Goch 1994.
16 bxc6 bxc6

17 Bf5
Understandably eliminating one of Black’s two bishops, even if the one on c8
doesn’t seem like a particularly active piece at the moment. The problem was that after
17 ... Qf6 Black threatens 18 ... Bxh3 because the knight on f3 will hang.
Interestingly, this position has been reached in a couple of other games. Each time
White played something else and something captured on h3 within the next few moves:
a) 17 Rfb1 Qf6 18 e4 Nf4!? 19 e5 Nxh3+ 20 Kf1 Qh6 21 exd6 Nf4 22 Ng1? (22
Kg1 was correct, intending 22 ... Nxg2 23 Bxh7+!, when it’s not clear what Black has
for his sacrifice) 22 ... Qh1 23 f3 Re3? (23 ... a5! looks okay here) 24 Qf2?? (after 24
Qd2! Rxd3 25 Qxf4 Rxc3 26 d7!, White turns the tables completely) 24 ... Rxd3 was
good for Black in J.Barkhagen-H.Cardon, Eeklo 1991.
b) 17 a5 Qf6 18 Rfe1 Bxh3 19 Bxg6 hxg6 20 e4 Bb4 left White scrambling to save
himself in G.Buckley-M.Taylor, British League 2008.
With 17 Bf5 White prevents such horrors but doesn’t particularly trouble Black.
Balashov’s play actually deserves careful study as his pieces float to excellent squares
and he equalizes with ease.
17 ... a5!
Getting an outpost on b4.
18 Rfc1 Qf6 19 Bxc8 Rexc8 20 Rab1 Ne7

The last link in the chain of Black’s fortress. The knight provides an impregnable
defence for the weak pawn on c6, after which White can’t hope to make further progress
on the queenside.
21 Qd3 Qg6 22 Qd1
Exchanging queens will lead to a draw; for example, 22 Qxg6 hxg6 23 Rb7 Rab8 24
Rcb1 Bb4 25 Rxb8 Rxb8 26 Na2 Bd6 27 Rxb8+ Bxb8 with a level endgame.
 22 ... Rcb8 23 Ne2 Qf5 24 Ne1 Rxb1 25 Rxb1 Rb8 26 Rxb8+ Bxb8 27 Nd3 Qc8
28 Qb3 Bc7 29 Nec1 g6 30 Nc5 Nf5 31 N1d3

White has established a knight on c5, but this is hardly enough to give him any
winning chances.
31 ... h5 32 Nb7 Nd6 33 Nxd6 Bxd6 34 Qb6 Qc7 35 Qxc7 Bxc7 36 f3
Activating the central pawn majority is White’s reserve plan, though it’s not
sufficient for anything more than stone cold equality here.
36 ... Kf8 37 Kf2 Ke7 38 e4 Bb6 39 Ke3 Ke6 40 Nf4+ ½-½
Not a very exciting game, but an excellent model from Black’s perspective.

Game 17
D.Navara-S.B.Hansen
European Team Championship, Reykjavik 2015

1 d4 Nf6 2 c4 e6 3 Nf3 d5 4 Nc3 Nbd7 5 cxd5 exd5 6 Bg5 Be7 7 e3 c6 8 Qc2 0-0 9
Bd3 Re8 10 0-0 Nf8 11 h3 g6
 Sune Berg Hansen is very fond of this move and it’s really quite interesting. In fact,
a number of very strong players have used it in their games, including Garry Kasparov.

Question: That may be so but what’s the point? Black certainly won’t be able to
play 12 ... Bf5 because White will simply take it.

Answer: He certainly won’t be able to play his bishop to f5 straight away, but this
isn’t the full extent of Black’s idea. He intends first to bring the f8-knight to e6 and then
to g7; after that ... Bf5 is a real possibility.

Question: But isn’t that a bit on the slow side?

Answer: Not necessarily. First of all, Black will gain a tempo on the g5-bishop, and
one can argue that White’s 11 h3 wasn’t a great use of a move anyway. Most of all, the
strategic benefits of this manoeuvre are so attractive that an investment of time may well
be worth it.
12 Rab1
Standard preparation for a minority attack with b2-b4-b5.
12 ... Ne6 13 Bh4
 Question: What if White avoids losing time with 13 Bxf6 - ?

Answer: That’s a good question. The answer is that Black can use the two bishops
to support ... c6-c5 in response to White’s b4-b5; for example, 13 ... Bxf6 14 b4 a6 15
a4 Qd6 16 Rfc1 Bd7 17 b5?! axb5 18 axb5 c5 19 dxc5 Nxc5 suddenly left White in
serious trouble in V.Akopian-G.Kasparov, Internet (blitz) 1998.

Question: Is 13 Bh6 any better, trying to exchange the other black knight when it
comes to g7?

Answer: Well, that’s another move to have been tried. B.Gelfand-V.Ivanchuk,

Linares 1993, went 13 ... Ng7 14 b4 a6 15 a4 Bf5 16 Ne5 Rc8 17 Bxg7 Bxd3 18 Nxd3
Kxg7 19 Rb3 Bd6 20 b5 cxb5!? 21 axb5 a5, when Black’s passed a-pawn proved to be
surprisingly powerful.
13 ... Ng7 14 b4 a6 15 a4
Continuing to build for the thematic b4-b5. Meanwhile, Black proceeds with his
own plan undeterred.
15 ... Bf5 16 b5 axb5 17 axb5 Bxd3 18 Qxd3 Nf5
Besides attacking White’s bishop on h4 this knight has another goal: it is aiming for
the d6-square from where it might establish itself on the outpost on c4.
19 Bxf6 Bxf6 20 Qc2 Nd6
 21 Nd2
White has also inserted 21 bxc6 bxc6 before 22 Nd2, which may be nothing more
than a harmless transposition of moves if Black then plays as in the game with 22 ...
Nf5. Instead, 22 ... Ra3 23 Ne2 Qa8 24 Rfc1 Ra2 25 Rb2 Rxb2 26 Qxb2 Nf5 27 Nf4
Ne7 28 Nf3 Rb8 29 Ra1 Rxb2 30 Rxa8+ Kg7 31 g4 h6 32 h4 was promising for White
in I.Khenkin-K.Asrian, FIDE World Championship, Moscow 2001; while after 22 ...
Qd7 23 Na4 Rad8 24 Nb6 Qe6 25 Qxc6 Nf5 26 Qxe6 fxe6 27 Nf3 Re7 28 Rfc1, White
went on to win in I.Khenkin-B.Heberla, European Championship, Plovdiv 2012.
From these examples it seems reasonable to assume that Black’s position is going to
be unpleasant unless he finds a way to gain counterplay. Sune Berg Hansen now goes
about obtaining this counterplay in a highly creative way.
21 ... Nf5!? 22 bxc6 bxc6 23 Na4
White could certainly stop Black’s evil intentions by returning his d2-knight to f3,
after which Black might just put his own knight back on d6 with the tacit offer of a draw.
Another possibility is 23 Rb4, but then Black can eliminate his backward c-pawn
with 23 ... c5! 24 dxc5 d4! 25 Nce4 dxe3 26 fxe3 Bg5, when the engine tells me that the
position is likely to fizzle out to equality.
23 ... Bxd4!
 A real bolt from the blue, which may well have been computer preparation. Black
gets a very strong initiative for the sacrificed piece.
24 exd4 Nxd4 25 Qd1 Qa5 26 Ra1
Pretty much the only move. After 26 Nb2 Black wins the exchange with 26 ... Ne2+
27 Kh1 Nc3 and will have a rook and two connected passers for the two knights.
26 ... Qb4
 27 Nf3?
No doubt somewhat shell shocked by Black’s 23rd move, Navara doesn’t find the
best defence. 27 Ra2 is strongly met by 27 ... Re2, but White can hang on with 27 Nb1.
That’s not to say that his position is at all pleasant after 27 ... Ne2+ 28 Kh2 d4, because
none of his pieces are particularly well placed.
27 ... Nxf3+ 28 gxf3
28 Qxf3 Rxa4 just leaves White two pawns down for nothing.
28 ... Re5 29 f4
Not 29 Nc3? Rg5+ 30 Kh1 Rxa1 31 Qxa1 Qh4 32 Qa8+ Kg7 33 Qc8 Qh5 and White
is defenceless.
29 ... Qxf4 30 Kh1
After 30 Re1 Rxe1+ 31 Qxe1 Rxa4, Black regains the piece with three extra pawns.
 30 ... Qe4+
Missing an immediate win, though Black can get it back by repeating the position.
Here 30 ... Rh5! would have been decisive; for example, 31 Ra3 Rxa4 32 Qxa4 (or 32
Rxa4 Rxh3+ 33 Kg2 Qh2 mate) 32 ... Rxh3+ 33 Rxh3 (or 33 Kg2 Qh2 mate again) 33 ...
Qxa4 with an easy win in sight.
31 Kh2 Qf4+ 32 Kh1 d4
Making a harder job of it than it needs to be. 32 ... Rh5! returns to the previous note.
33 Rg1 Qxf2 34 Rf1 Qe3 35 Rf3 Qe4 36 Raa3
36 Kh2 c5 is also quite hopeless for White.
36 ... Rb8 37 Ra1 c5 38 Kh2 c4 39 Qf1 Qd5 40 Ra2 Rbe8 41 Rf6 d3 42 Nc3 Qd4
43 Rxf7 Qxc3 44 Raa7
If 44 Raf2 then 44 ... Re2 would snuff out White’s counterplay.
44 ... Re2+ 45 Kg3 d2+ 46 Qf3 Qxf3+
A sensible practical decision making the win ‘idiot proof’.
47 Kxf3 R2e7 0-1
Black wins easily after 48 Raxe7 d1Q+ 49 Kf2 Qd4+ 50 Ke2 Rxe7+ 51 Rxe7 Qb2+
52 Ke3 Qa3+ 53 Kd4 Qxe7 and suchlike.

Game 18
Y.Pelletier-M.Gurevich
Gibraltar 2006

1 d4 d5 2 c4 c6

Question: Isn’t this a Slav Defence?

Answer: At the moment it is, but we will soon transpose into a Queen’s Gambit
Declined.

Question: So why not just play 2 ... e6 instead?

Answer: By adopting this move order Gurevich gets White to commit to an early
Nf3, which cuts out particular set-ups such as those with Nge2 in the next chapter.
3 Nf3 Nf6 4 Nc3 e6 5 Bg5 Nbd7 6 cxd5 exd5 7 e3 Be7 8 Bd3 0-0 9 Qc2 Re8 10
0-0 Nf8 11 h3 Be6
A relatively recent idea pioneered largely by Artur Yusupov. At first it looks like a
simple developing move, but Black is also preparing to counter the minority attack by
putting a rook on c8. This will make it easier for him to meet White’s b4-b5 with ... c6-
c5.
 12 a3!?
One of several moves, the other main ones being 12 Rfc1 (Game 20) and 12 Rab1
(Games 21-23). It has been established that 12 Ne5 is fairly harmless after 12 ... N6d7;
for example, 13 Bxe7 Qxe7 14 f4 f6 15 Nf3 Qd6 16 Qf2 Re7 17 Rae1 Rae8 18 a3 ½-½
was K.Miton-A.Sokolov, Belfort 2012.

Question: Why is White preparing b2-b4 with a2-a3 rather than 12 Rab1 - ? Won’t
he lose a tempo if Black opts for ... a7-a6 at some point and White then decides to play
a3-a4 etc?

Answer: That’s an excellent question; the answer is neither simple nor clear cut. But
one problem with 12 Rab1 is that Black can reply 12 ... Ne4, as we’ll see in Game 22.
The point is that 13 Bxe7 Rxe7 14 Nxe4 dxe4 15 Bxe4 can be met by 15 ... Bxa2, which
is obviously not possible after 12 a3.
12 ... N6d7
The main alternative 12 ... Rc8 will be considered in the next game.
13 Bf4 Nb6
 Black can play 13 ... Rc8 here too, but then 14 b4 Nb6 15 Nd2 Bd6 16 Bxd6 Qxd6
17 Rfc1 Qe7 18 Ne2 Ng6 19 Nb3 Rc7 20 a4 Nc4 21 Nc5 saw him suffering a bit in
Z.Izoria-M.Gurevich, Izmir 2003.
14 Rfc1
One of several moves to have been tried:
a) 14 Ne5 f6 15 Nf3 Bf7 16 Nd2 Bg6 allowed Black the strategically desirable
exchange of light-squared bishops in R.Dautov-A.Graf, German Blitz Championship
2003. After 17 b4 Qd7 18 Rab1 a6, he could claim full equality at this stage.
b) 14 b4 Nc4!? 15 Na4 was played in M.Kantorik-S.Smagin, Bad Wiessee 1999,
when 15 ... Ng6 (15 ... b6 16 Ne5 was annoying in the game) 16 Bg3 b6 seems quite
comfortable for Black.
14 ... Bd6
Black opted for 14 ... a5 in A.Obodchuk-S.Smagin, Moscow 1995, though this
advance of the a-pawn creates certain obligations, such as the need to protect it. After
15 Na4 Nxa4 16 Qxa4 Ng6 17 Bh2 Bd6 18 Bxd6 Qxd6, White could have tried 19 Rc5
(rather than 19 Qc2), when 19 ... Qc7 20 Qc2 is not so easy for Black.
15 Bxd6 Qxd6 16 b4 a6
 A familiar idea, making it more difficult for White to push through with his minority
attack. At the same time Black guarantees that he won’t be left with a weak a-pawn if
White does get b4-b5 in.
17 Nd2

Question: Presumably White avoided 17 a4 because the b-pawn would drop?

Answer: Actually, White could have played this way as after 17 ... Qxb4 18 Rab1
Qa5 19 Nd2, Black’s queen would be in trouble.
17 ... Qe7
Vacating the d6-square so as to transfer the b6-knight there via c8.
18 Ne2
Pelletier seems to be drifting. Again he could have considered 18 a4, when 18 ...
Rac8 (18 ... Qxb4? now loses to 19 Rab1 Qa5 20 Nb3 Qb4 21 Nc5 etc) 19 Nb3 leaves
White with a pleasant position.
18 ... Nc8 19 Qb2
And here White might have tried 19 Nb3 Nd6 20 Nc5. It’s hard to see what his plan
is after 20 ... Ng6, but at least the exchange of light-squared bishops isn’t on the near
horizon.
 19 ... Nd6 20 a4 Bf5! 21 Bxf5 Nxf5

Suddenly Black’s prospects look rather rosy as he’s threatening 22 ... Nxe3. He can
also bring this knight back to d6 in order to stop b4-b5, though in the game Gurevich
chooses a more aggressive approach.
22 Nc3 Ng6
As we’ll see, threats are now quickly materializing against White’s king.
23 b5 Ngh4 24 Nf1
This, together with his next move, represents a standard way to reinforce White’s
castled king. The downside is that it leads to other weaknesses being created.
Nevertheless, it is often necessary; for instance, White might have preferred to play 24
Kh1? Qg5 25 Rg1 with a defensible if passive position, but 24 ... Nxg2! leads to a very
strong attack after 25 Kxg2 Qg5+ 26 Kh2 Rxe3! 27 fxe3 Qg3+, which the computer
insists is winning for Black despite his current huge material deficit.
24 ... Qg5 25 f4 Qd8
Preparing to regroup with ... Nd6 and ... Nhf5, latching onto the newly created
weakness on e3.
26 bxc6 bxc6 27 Qf2 Nd6 28 g3 Nhf5 29 g4 Ne7
So the knight has been driven from f5, but it can come round to the c4-square if it
wants to put pressure on e3. Otherwise 29 ... Nh4!? was an interesting alternative.
 30 f5 Nec8 31 Nd2 Nb6 32 Re1

Trying to get counterplay with e3-e4. Obviously this doesn’t do a whole lot for
White’s king position, even if it helps him activate his forces.
32 ... Nbc4 33 Nxc4 Nxc4 34 e4 dxe4 35 Rxe4 Qa5
35 ... Rxe4 36 Nxe4 Rb8 was a pretty good alternative, trying to bring the rook into
b2.
36 Rae1 Rf8
Now 36 ... Rxe4 37 Nxe4 Qxa4 would allow White dangerous counterplay after 38
f6.
37 Qg3 Rad8 38 Qd3
38 f6 gxf6 39 Qd3 was still worth considering, when 39 ... Ne5 40 Qg3 Nc4 could
be a draw by repetition.
38 ... Nb6 39 Re5 Qb4 40 Rb1 Qd6!
 Underlining the weakness of White’s king. After 40 ... Qc4 41 Qxc4 Nxc4 42 Rc5,
White would have had an easier time as the exchange of queens leaves his king secure.
41 Re2
And not 41 Rxb6? because of 41 ... Qxe5!.
41 ... Nd5 42 Nxd5 Qxd5 43 Rd1 Rb8!?
Trying to infiltrate White’s position via b3.
44 Re5 Qa2 45 Qxa6?!
This takes the queen far away from the defence of his king. Again 45 f6 was
probably the best chance, after which 45 ... Rb3 46 Qe2 Qxe2 47 Rxe2 Rxh3 is
probably only a draw with best play.
45 ... Rb2 46 Qxc6??
A blunder in time trouble. White had to play 46 Qf1, when 46 ... Qxa4 is good for
Black but not decisively so.
46 ... Qb3 0-1
The twin threats of ... Qxd1 and ... Qg3+ are fatal.

Game 19
A.Graf-A.Yusupov
German Championship, Altenkirchen 2005
1 d4 Nf6 2 c4 e6 3 Nc3 d5 4 cxd5 exd5 5 Bg5 Be7 6 e3 Nbd7 7 Bd3 0-0 8 Nf3 Re8 9
Qc2 Nf8

Question: Is Black delaying ... c7-c6 for a particular reason?

Answer: Yes he is, and that was well spotted. He wants to wait and see which side
White castles; if the white king goes to the queenside Black will be able to play ... c7-
c5 in one move.
10 h3
Slightly more flexible than 10 0-0, since White keeps the option of 0-0-0.
Consequently, Yusupov again defers moving his c-pawn, so as to maintain the
possibility of ... c7-c5.
10 ... Be6

Question: I guess from your answer to my previous question, Graf might have
answered 10 ... c6 with 11 0-0-0 - ?

Answer: Yes indeed, since playing for ... c6-c5 would then cost Black a tempo.
 11 0-0
As already implied, one of the points of Yusupov’s move order is to meet 11 0-0-0
with 11 ... c5, setting about opening the c-file and without wasting time on ... c7-c6.
11 ... c6
With White’s king committed to the kingside, Yusupov finally plays ... c7-c6. Note
that we have now transposed to the 11 ... Be6 line first examined in Game 18.
12 a3 Rc8

This time Black follows the plan outlined in the previous game, developing his a8-
rook to the c-file so that he can counter b2-b4-b5 with ... c6-c5.
13 Na4

Question: White did White avoid the standard 13 b4 here?

Answer: Moving the knight to a4 is a logical alternative, but a bigger factor might
have been Yusupov’s considerable experience against 13 b4. For example, 13 ... N6d7
14 Bf4 (or 14 Bxe7 Qxe7 15 Nd2 Qh4 16 Ne2 Nf6 17 Nf4 Bd7 18 Nf3 and a draw was
agreed in V.Beim-A.Yusupov, German League 2001) 14 ... Nb6 (instead, 14 ... Ng6 15
Bg3 Ndf8 16 Na4 Bd6 17 Bxd6 Qxd6 18 Nc5 Rc7 19 Rac1 Bc8 20 Rfe1 Qf6 21 Qd1
Ne6 22 Nxe6 Bxe6 23 b5 was better for White in V.Topalov-A.Yusupov, Frankfurt rapid
2000) 15 Nd2 Bd6 16 Bxd6 Qxd6 17 Ne2 Nbd7 18 Nb3 Rc7 19 Rfc1 a6 20 Nc5 Nxc5
21 Qxc5 Qxc5 22 bxc5 Bc8 23 Rab1 g6 24 Rc3 Ne6 25 Nc1 Kg7 26 Bc2 f6 27 f3 Rce7
28 Kf2 Nd8 made it very hard for White to make progress in A.Gershon-A.Yusupov,
European Championship, Ohrid 2001, though he’s certainly on the positive side of the
position because he can at least try to achieve something.
13 ... Ne4 14 Bxe7 Rxe7
Not 14 ... Qxe7? 15 Bxe4 dxe4 16 Qxe4 and White simply wins a pawn.
15 Nc5 Nxc5
Black can’t keep the knight on e4 since 15 ... f5?! gives the e5-square away, while
15 ... Bf5? is bad because of 16 Nh4.
16 dxc5!?

Radically changing the pawn structure. White abandons the standard minority attack
in order to gain the d4-square for his knight. 16 Qxc5 was certainly playable, but after
16 ... Nd7 17 Qc2 (17 Qxa7? Ra8 18 Qxb7 Ne5 is just good for Black) 17 ... g6, Black
looks okay, being prepared to meet a later b4-b5 with ... c6-c5.
16 ... Re8 17 Qc3 Bd7 18 Rfe1 Ne6 19 Rad1 Qe7 20 b4 Ng5
 21 Nd4
The critical move, putting the knight on its best square in order to prepare b4-b5. At
the same time this leaves Black’s g5-knight on the board which can menace White’s king
position, but 21 Nxg5 Qxg5 22 Kh2 Bf5 is nothing for White.
21 ... Qf6 22 Ne2
Concern about his king encourages Graf to bring the knight back for defence.
Otherwise, he might well have considered 22 a4, aiming to continue his queenside play
with b4-b5. Black can then try lifting his rook with 22 ... Re5, when 23 f4?! is wrong
because of 23 ... Ne4. This kind of position is worth further study in order to deepen
one’s understanding of the opening.
22 ... Qh6
Obviously Black doesn’t want to exchange queens when he has good attacking
chances on the kingside.
23 Ng3
White could also defend his king with 23 Nf4, which might have been a better
square for the knight.
23 ... Rcd8 24 Qb2 Qh4 25 Bb1 g6
 26 Qf6??
A very surprising blunder from Graf, even if his position is far from comfortable.
For example, after 26 Bd3 Black can sacrifice with 26 ... Bxh3! 27 gxh3 and follow up
with the powerful 27 ... d4. One of the points is that 28 exd4? loses at once to 28 ...
Nf3+ 29 Kf1 (or 29 Kg2 Nxe1+) 29 ... Qxh3 mate. So White probably has to play 28
Bf1, when 28 ... Nf3+ 29 Kg2 Nxe1+ 30 Rxe1 dxe3 31 Rxe3 Rxe3 32 fxe3 Qg5 gives
Black the edge in view of White’s exposed king.
26 ... Nxh3+ 0-1
Since 27 ... Qxf6 follows. Backwards captures are often overlooked.

Game 20
Li Wenliang-A.Yusupov
Minneapolis 2005

1 d4 Nf6 2 c4 e6 3 Nf3 d5 4 Nc3 Nbd7 5 cxd5 exd5 6 Bg5 Be7 7 e3 0-0

As in the previous game Yusupov delays ... c7-c6 so that he can meet a potential 0-
0-0 with ... c7-c5 in one move.
8 Bd3 Re8 9 0-0 c6 10 Qc2 Nf8 11 h3 Be6 12 Rfc1
 A move introduced by Anatoly Karpov back in 1988, which continues to represent a
good way for White to proceed.
12 ... Rc8
Once again, this standard follow-up to 11 ... Be6 allows Black to answer a later b2-
b4-b5 with ... c6-c5. In the stem game with 12 Rfc1, Black opted for 12 ... N6d7.
A.Karpov-A.Kharitonov, USSR Championship, Moscow 1988, continued 13 Bf4 Nb6
14 Rab1 Bd6?! (due to the strength of White’s 17th move Black should have delayed the
exchange of dark-squared bishops; accordingly he might have preferred 14 ... Rc8 at
this point) 15 Ne2 Ng6 16 Bxd6 Qxd6 17 a4! (a very strong and somewhat unusual idea;
White advances on the queenside without giving Black the c4-square) 17 ... Rac8 (if 17
... a5 White plays 18 b3, intending 19 Qd2 and 20 Rc5) 18 Qc5 Qb8 19 Qa3 a6 20 Rc3
Qc7 21 Rbc1 Ra8 22 Nd2 a5 23 Rb1 Nc8 and now Karpov finally got his 24 b4 in with
a very good position for White.
V.Malaniuk-G.Feher, Budapest 1989, varied from the above game with 13 ... Ng6 14
Bh2 Nb6, when White should probably play 15 Rab1. Instead, Malaniuk chose the
immediate 15 a4?!, which is less effective with the dark-squared bishops still on the
board due to the difficulty White now has in forcing b2-b4 through. The game went on
15 ... a5 16 Ne2 Nd7 17 Qb3 Bb4 18 Nc3 Ra7 19 Na2 Bf8 20 Qd1 Qf6 21 Ne1 Nh4 22
Nc3 Bf5 23 Bxf5 Nxf5 and Black had a perfectly good position, even if he later lost.
13 Rab1 N6d7 14 Bf4
 14 ... Ng6

Question: When should Black play ... Ng6 in such positions? Is it good here?

Answer: That’s a very pertinent and quite complex question. Sometimes a knight on
g6 can be used for kingside action, at others it is just rooted to the spot because of the
need to cover h7. In this particular case it is part of an interesting regrouping which will
be revealed on the next move.
A couple of alternatives have been tried:
a) 14 ... c5?! is an attempt to change the nature of the position to one in which
White’s rooks are misplaced. Even so, the isolated d-pawn is the biggest factor after 15
dxc5 Nxc5 16 Nd4 (16 Nb5?! d4! was good for Black in B.Nickoloff-R.Ziatdinov,
Toronto 1998) 16 ... Nxd3 17 Qxd3 Ng6 18 Bg3 with a pleasant edge for White.
b) 14 ... f6!? is a noteworthy concept, planning to exchange light-squared bishops
via ... Bf7-g6. I.Khenkin-A.Mirzoev, Saint Vincent 2002, continued 15 Na4 Bf7 16 Bg3
Bg6 17 Bxg6 hxg6 18 Qb3! Qa5 (transferring the queen to a6 is the only good way to
defend the b7-pawn) 19 Ne1 Qa6 20 Nd3 Ne6 21 Qc2 g5 22 b4 b5 23 Nac5 Ndxc5 24
bxc5 Qb7 25 a4 a6, and now 26 axb5 axb5 27 Nb4 Ra8 28 Ra1 would have left White
with some pressure.
15 Bh2 Ndf8!?
 Question: What’s the point of this move? Doesn’t that knight usually go to b6 in
order to try and get in on c4?

Answer: Yes, that’s often the case. In this instance Yusupov has seen another way to
approach the position: he wants to exchange the knight on f3 with ... Nh4 and defends h7
first in order to prepare this.
16 Na4!?
The point behind Yusupov’s play can be seen after 16 b4 a6 17 a4 Nh4! 18 Nxh4
Bxh4 19 b5 axb5 20 axb5 c5 with a comfortable game for Black. Accordingly White
tries to cut across this idea.
16 ... Bd6!
Now 16 ... Nh4?! is answered by 17 Nxh4 Bxh4 18 Nc5, so Yusupov exchanges
dark-squared bishops in order to defend b7 with ... Rc7.
17 Bxd6 Qxd6

This position has arisen before, albeit via a slightly different move order. White’s
next move was new at the time of the game.
18 b4
In S.Kümin-F.Begovac, Swiss Team Championship 2001, White played 18 Nc5,
when 18 ... Rc7 was correct but Black should probably have answered 19 b4 with 19 ...
Nd7! (in the game 19 ... a6 20 a4 was a bit better for White); for example, 20 Rb3
Nxc5! 21 bxc5 Qe7 22 Rcb1 Bc8 and with the queenside securely held, Black can think
about being a nuisance on the other side of the board.
18 ... Nd7 19 Nc5 Nxc5 20 dxc5
After 20 bxc5 Qe7, followed by ... Rc7 and ... Bc8 again, it would be impossible
for White to break through on the queenside. Instead, 20 Qxc5 Qxc5 21 Rxc5 looks
slightly more positive for White, though he is still unlikely to be able to make progress
if Black centralizes his king to d6.
20 ... Qc7 21 Nd4

The white knight is very strong, so Black hurries to exchange it off.

21 ... Bd7 22 a4 Nf8 23 b5 g6 24 Qb2
Similarly, 24 bxc6 bxc6 followed by ... Ne6 will be fine for Black.
24 ... Ne6 25 Rc2 Qa5 26 Nxe6 Rxe6 27 Qd4 Rce8
 Black stands no worse here. White can’t make further progress on the queenside,
and Black’s rooks have potential activity on the e-file.
28 Rcb2
Now after 28 bxc6 Black can just recapture with the bishop.
28 ... Re5 29 Rb4 Qd8 30 Qc3 Qa5 31 Qd4 Qd8 32 Qc3 ½-½

Game 21
A.Lytchak-U.Bönsch
German League 2005

1 d4 Nf6 2 c4 e6 3 Nc3 d5 4 cxd5 exd5 5 Bg5 c6 6 Qc2 Be7 7 e3 Nbd7 8 Bd3 0-0 9
Nf3 Re8 10 0-0 Nf8 11 h3 Be6 12 Rab1
 The most natural move for White, preparing the usual minority attack with b2-b4-b5.
12 ... N6d7
A logical response, though not the only one. 12 ... Rc8 will be covered in Game 23,
while the interesting 12 ... Ne4 is dealt with in the next game.
13 Bf4

Question: Was there a reason White avoided the exchange of bishops here?

Answer: 13 Bxe7 Qxe7 14 b4 Rac8 prepares, as usual, to answer b4-b5 with ... c6-
c5. A.Bekker Jensen-M.Rivas Pastor, Stockholm 1999, continued 15 Rfc1 Qf6 16 Qd1
Ng6 17 b5 c5 18 dxc5 Rxc5 19 Na4 Rcc8 with approximate equality, though Black may
wonder where his winning chances are in such positions.
13 ... Nb6
Black has also tried 13 ... a5, with the clear aim of delaying or preventing White’s
minority attack. The drawback is that the a-pawn may become weak, or even the b6-
square should Black be forced to take an invading knight on c5. After 14 Rbd1 Nb6 15
Nd2 Bd6?! (exchanging the dark-squared bishops only helps White; the immediate 15 ...
a4!? would have been better) 16 Bxd6 Qxd6 17 Rde1 f6 18 f4!? Bf7 19 Nf3 a4 20 a3
Qb8 21 g4, Black was struggling to cope with White’s kingside initiative due to his lack
of counterplay in A.Kharlov-J.Meister, Sochi 2005.

14 Na4
Exchanging the knight rather than allow it land on c4 looks like a good idea if White
wants to play a minority attack. On the other hand, it loses time and leaves White’s
queen on a funny square.
Instead, 14 b4 continues the minority attack in the most direct way, though Black can
then get counterplay with 14 ... Nc4! (14 ... Bd6 15 Bxd6 Qxd6 16 a4 a6 17 Nd2 Re7 18
Nb3! Rc7 19 Nc5 Nbd7 20 f4 f6 21 f5 was good for White in A.Beliavsky-V.Ivanchuk,
Linares 1993); for example, 15 Bxc4 dxc4 (intending 16 ... b5 if given the chance) 16
b5 Ng6 17 Bg3 Bd6 18 bxc6 bxc6 19 Bxd6 Qxd6 20 Rb7 c5 21 d5! Bxd5 22 Qd2 Rad8
23 Qxd5 Qxd5 24 Nxd5 Rxd5 soon fizzled out to a draw in S.Safin-J.Sriram, Lucknow
2004.
On the other hand, the flexible 14 Rfc1 seems like quite a good idea here; for
example, 14 ... Bd6 (after 14 ... a6 15 Na4 Nxa4 16 Qxa4 f6 17 b4 Qd7 18 Qc2 Bf7 19
a4 Rac8 20 Bg3 Bg6 21 Ne1 Bxd3 22 Nxd3 Ne6 23 Nc5 Bxc5 24 bxc5, White had a
clear advantage in I.Khenkin-K.Bischoff, Belgian Team Championship 2004) 15 Ne2
Ng6 16 Bxd6 Qxd6 17 a4 Rac8 18 Qc5 Qb8 19 Qa3 a6 20 Rc3 Qc7 21 Rbc1 Ra8 22
Nd2 a5 23 Rb1 Nc8 24 b4 axb4 25 Qxb4 gave White serious pressure in A.Karpov-
A.Kharitonov, USSR Championship, Moscow 1988.
14 ... Nxa4 15 Qxa4 Bd6
 Question: Why doesn’t Black play 15 ... Ng6 here, gaining a tempo on the bishop
before swapping it off?

Answer: The problem is that the knight may not be well placed on g6. I.Sokolov-
A.Yusupov, Amsterdam 1994, continued 16 Bh2 Bd6 17 Ne5 (17 Bxd6 Qxd6 18 b4 a6
19 Qc2 was another way to play this for White, as it’s not clear that the knight belongs
on g6, but if it moves anywhere apart from back to f8 White can snaffle the h7-pawn) 17
... Qf6 18 Qc2 (18 f4 is another interesting plan when Black has played ... Ng6, as the
knight is vulnerable to f4-f5 possibilities) 18 ... Nxe5 19 dxe5 Bxe5 20 Bxh7+ Kh8 21
Bxe5 Qxe5 22 Bd3 Rac8 (22 ... d4?! 23 e4! Bxa2 24 Rbe1 Qe6 25 f4 would give White
a dangerous attack) 23 Qa4 c5 and by this stage Black’s centralized pieces and pawns
compensated him for any weaknesses elsewhere.
16 Bxd6
In this position 16 Ne5 is met simply by 16 ... f6.
16 ... Qxd6 17 b4 a6 18 Qc2 g6!

After this the knight is no longer tied to the defence of the h7-pawn.

Question: Doesn’t that make Black’s bishop really bad. since all his pawns are
now on light squares?

Answer: Not really as the bishop isn’t blocked in by the pawns. Instead, Black has a
certain weakness on the dark squares, but he can use his newly freed knight to neutralize
this.
19 a4 Nd7
The knight helps defend against White’s threatened minority attack in two different
ways. Firstly, it can support ... c6-c5 in response to b4-b5 and, secondly, Black can take
with the c-pawn, play ... a6-a5, and then post the knight on b6.
20 Rfc1
The previous note is illustrated by the fact that Black can meet 20 b5?! with either
20 ... axb5 21 axb5 c5 or 20 ... cxb5! 21 axb5 Rec8 22 Qb2 a5!, intending 23 ... Nb6,
and if White tries to prevent that with 23 b6? he loses a pawn after 23 ... Rc6!.
20 ... Rac8
A few months later Bönsch switched to 20 ... Rec8, perhaps to sidestep preparation
or possibly because he thought it was better; but if he foresaw variations where leaving
the rook on a8 is useful, they didn’t arise. D.Jacimovic-U.Bönsch, European Cup, Saint
Vincent 2005, continued 21 Qc3 Qe7 22 Nd2 Qg5 23 f4!? Qg3 (23 ... Qh4!? makes it
easier for the queen to retreat to e7 or d8 while keeping the option of sacrificing on h3)
24 Nf1 Qh4 25 Qd2 Nb6 26 a5 (committing himself to playing on the kingside or in the
centre) 26 ... Nd7 27 Qf2 Qe7 28 g4 f6 29 Ng3 Bf7 30 Re1 Re8 31 e4 with a strong
initiative for White.
21 Qc3
As before, Black is well prepared to meet b4-b5 by taking with the a-pawn and
playing ... c6-c5. In this specific position he has an even stronger line in 21 b5?! cxb5
22 Qxc8 Rxc8 23 Rxc8+ Nf8!, when the bishop on e6 hits the rook on c8, denying White
the time he needs to recapture on b5.
21 ... Qe7 22 Nd2
 Rerouting the knight to the queenside like this improves the preconditions for a
successful minority attack, but at the same time it takes a piece away from the king’s
protection team and thus allows Black counterplay there.
22 ... Qg5
Threatening 23 ... Bxh3, which White sidesteps with a king move.
23 Kf1 Qf6 24 Kg1
A tacit offer of a draw, since Black could repeat the position with 24 ... Qg5.
Bönsch’s next move shows that he wants to win!
White perhaps didn’t like 24 b5 because of 24 ... axb5 25 axb5 Bxh3!? 26 gxh3
Rxe3, though this looks like no more than a draw for Black after 27 Kg2 Qg5+ 28 Kh2
Qf4+ 29 Kg2 Qg5+ etc.
24 ... Kg7 25 Nb3 Qg5 26 Kf1 Bf5
Even at this stage, the trade of light-squared bishops is a common goal for Black in
the QGD Exchange.
27 Na5 Rc7 28 Bxf5 Qxf5 29 b5 axb5 30 axb5 Nf6
Suddenly Black’s pieces are becoming active. White may well have been short of
time at this point which would compound his issues on the board.
 31 Kg1 Ra8 32 bxc6 Ne4 33 Qe1 bxc6 34 Rb2 c5 35 dxc5 Nxc5 36 Qb4?! Qf6
37 Qc3?? 0-1
Suddenly White finds a horrible blunder which drops the knight. On seeing what
he’d done Lytchak immediately resigned. He should have played 36 Qc3+ on the
previous move with equality, or here 37 Qb5, when Black’s slight initiative can be held
in check; e.g. 37 ... Rca7 38 Ra2 Ne4 39 Qb4 seems okay.

Game 22
A.Mastrovasilis-M.Gurevich
European Championship, Dresden 2007

1 d4 d5 2 c4 c6
Gurevich adopts the same move order as in Game 18.
3 Nf3 Nf6 4 Nc3 e6 5 Bg5 Nbd7 6 cxd5 exd5 7 e3 Be7 8 Bd3 0-0 9 Qc2 Re8 10
0-0 Nf8 11 h3 Be6 12 Rab1 Ne4
 This interesting possibility for Black is perhaps the reason some players prefer
either 12 a3 or 12 Rfc1 as White.
13 Bf4

Question: Can’t White simply win a pawn with 13 Bxe7 Rxe7 14 Nxe4 dxe4 15
Bxe4 here?

Answer: No, because White’s 12 Rab1 means the a2-pawn is now hanging. If you
remember, this was one of the points behind 12 a3!? in Game 18.
However, I should mention that White has a very reasonable continuation in 13 Bxe7
Rxe7 14 b4. For example, 14 ... Bf5 15 Bxe4 (15 Nd2 Nxc3 16 Qxc3 Bxd3 17 Qxd3
Ne6 18 b5 c5 was equal in E.Bacrot-M.Stojanovic, Basel 2011) 15 ... Bxe4 16 Nxe4
dxe4 17 Nd2 Qd5 18 Rfc1 Re6 19 a4 Rae8 20 Qc5 Qa2 21 Qc2 Qd5 22 b5 saw White
secure an edge in F.Bindrich-M.Stojanovic, Swiss Team Championship 2012.
13 ... Bd6 14 Bxd6 Nxd6 15 b4 Rc8
 By now this plan should be quite familiar to the reader. Black intends to meet b4-b5
with ... c6-c5, and even though he contracts an isolated d-pawn, the unfortunate position
of White’s advanced b-pawn gives Black long-term control of valuable squares such as
c5 and c4.
16 Rfc1 Qf6
Threatening 17 ... Bxh3.
17 Qd1 a6
Another familiar idea. Black does not intend to halt the white b-pawn altogether, he
simply wants to have the a-pawns exchanged when b4-b5 occurs. This means he won’t
be left with a weak a-pawn and may also be able to make use of the newly opened a-
file.
18 a4 Nd7 19 b5
There was a case for delaying this consistent move, since Black gets his knights
established on excellent squares, but White may have been concerned about how Black
might improve his position otherwise.
19 ... axb5 20 axb5 c5
 21 dxc5
Similarly, there was a case for leaving the c5-pawn alone and playing something
like 21 Rb2.
21 ... Nxc5 22 Bf1 Nce4 23 Nxe4
The d5-pawn was immune to capture because of 23 Nxd5 Bxd5 24 Rxc8 Rxc8 25
Qxd5 Nc3, winning the exchange.
23 ... Nxe4 24 Qd4 Qxd4 25 Nxd4 Bd7 26 Bd3
My engine likes 26 b6 at first, but then 26 ... Nd2 27 Ra1 Rxc1 28 Rxc1 Nc4 29
Bxc4 Rc8 gives Black chances to play for a win, which Gurevich was probably looking
to do at this stage.
26 ... Nd6 27 Kf1 Kf8 28 Ke2
Obviously White cannot play 28 Bxh7 because 28 ... g6 closes a trap on the bishop.
28 ... Ke7 29 Kd2 Ne4+ 30 Bxe4 dxe4
 Black’s isolated d-pawn has been converted into kingside space, which in turn
provides an opportunity to attack White’s kingside pawns via a rook lift.
31 Rc3 Kd6 32 Rb4 Re5 33 Rxc8 Bxc8 34 Kc3
Here 34 Ra4 may have been better, though after 34 ... Rg5 35 Ra8 Bd7 White is left
with the problem of how he should defend g2.
34 ... Rc5+ 35 Rc4?
Under increasing pressure White makes a serious mistake which leads to the loss of
his b-pawn. He had to keep the rooks on in order to maximize his drawing prospects.
35 ... Kd5 36 Rxc5+ Kxc5 37 h4 b6 38 Kb3 Bd7 39 Kc3 Bxb5 40 Nb3+ Kd5 41
Kb4 Bf1
Creating further weaknesses in White’s kingside.
42 g3 f6 43 Nd2 Bd3 44 Nb3 Ke5 45 Nd4 Ba6 46 Kc3 g5 47 f4+?!
And now White creates some more unprovoked.
47 ... exf3 48 Nxf3+ Kf5 49 e4+ Kg6 50 hxg5 fxg5 51 Ne5+ Kg7
 52 Nd7?
White’s resistance starts to crumble. 52 Kd4 would have been a tougher defence.
52 ... h5 53 Nxb6?
After this Black’s h-pawn heads towards promotion. 53 Kd2 was somewhat better,
though White is losing in any case at this point.
53 ... h4 54 gxh4 gxh4 55 Nd5 h3 56 Nf4 h2 57 Nh5+ Kg6 58 Ng3 Kg5 59 Kd2
Kf4 60 Nh1 Kxe4 61 Ke1 Kf3 62 Nf2 Kg2 0-1
White would like the right not to make a move. Unfortunately, he has to.

Game 23
I.Sokolov-K.Asrian
Stepanakert 2005

1 d4 d5 2 c4 e6 3 Nc3 Nf6 4 cxd5 exd5 5 Bg5 Be7 6 e3 Nbd7 7 Bd3 0-0 8 Nf3 c6 9
Qc2 Re8 10 h3
Again White plays this flexible move, keeping open the option of castling long. In
the event we end up with a harmless transposition.
10 ... Nf8 11 0-0 Be6 12 Rab1 Rc8
In this final game of the chapter, we’ll look once more at this standard plan, with
which Black prepares to answer b4-b5 with ... c6-c5.
13 b4

13 ... Nh5

Question: If Black wants to exchange the dark-squared bishops why does he go

here with the knight rather than to d7?

Answer: Moving the knight to h5 ensures the exchange, since if the bishop goes back
to f4 the knight will take it. On the other hand, the knight may not be as well placed here.
An example of 13 ... N6d7 was M.Sorokin-G.Soppe, Villa Gesell 1996, which
continued 14 Bf4 Nb6 15 a4 Ng6 16 Bg3 Bd6 17 Ne2 Nc4 18 Bxc4 dxc4 19 Nd2 b5
with complex play.
14 Bxe7 Qxe7 15 Rfc1 Ng6
Black has also tried 15 ... g6, presumably with the idea of ... Ng7 and ... Bf5 (as in
Game 10), but it looks too slow here. After 16 b5 c5 17 dxc5 Rxc5 18 Qa4 Rec8 19
Ne2 b6 20 Nfd4, White had a clear edge in Jav.Garcia-M.Nina, Peruvian
Championship, Lima 2000.
16 Ne2 a6
For now the battle rages over White’s intended b4-b5.
 17 a4 Bd7 18 Re1 Qf6 19 Nh2 Qd8 20 Nc3 Nf6 21 a5

Question: Doesn’t that rule out White’s minority attack?

Answer: Well spotted; indeed it does. White is essentially announcing a change of

plan with this move, abandoning the minority attack (b4-b5) in favour of central or
kingside action. And part of this new plan will be to bring his knight on c3 to a4 and c5.
21 ... Rc7

Question: That looks like an odd move; what’s Black up to now?

Answer: He’s regrouping his pieces so as to inhibit White’s e3-e4. His bishop will
go back to c8 and then the c7-rook can swing across to e7.
22 f3
A standard way of preparing for e3-e4, though this can also be used as part of a
kingside build-up with g2-g4, followed by massing White’s forces behind the kingside
pawns.
22 ... Bc8 23 Na4 Rce7 24 Qf2 Nf8 25 Nf1 Ne6 26 Qg3
 Question: What does that move do? Is it some kind of attacking gesture?

Answer: Not really. It looks more like quiet probing, asking Black what he intends
to do next. Indeed, despite the solidity of his position, it’s not easy to find an active plan
for Black.
26 ... g6 27 Qh4 Rc7 28 g4

Question: That looks like some kind of pawn storm. Is it really a good idea in front
of your own castled king?

Answer: Advancing pawns in front of your own king does have its risks, but if you
can pack pieces in front of the king and keep the centre closed, it can be a good plan. In
this position Sokolov has achieved these preconditions.
28 ... Ng7 29 Rb2 Qd6 30 Qg3
Evidently deciding that his kingside pawn advance will be safer without queens on
the board. White is certainly on the positive side of the game after this, but whether or
not he can make progress remains to be seen.
30 ... Qxg3+ 31 Nxg3 h5 32 Nc5 Rce7 33 Rbe2 hxg4 34 hxg4 Ne6 35 Kg2 Kg7
Getting ready to bring a rook to the h-file if necessary. Black’s position appears
more or less impregnable now, and White’s attempt to make progress from here pushes
the boundaries of correctness.
36 Nb3?!
Sokolov foresees the possibility of 36 ... Ng5 followed by 37 ... Nxf3 38 Kxf3
Bxg4+ and so brings his knight back to cover f3. But it’s hard to believe that moving this
knight from such a beautiful square can be good. It would be better to use the other one
with 36 Nf1-d2 or 36 Nh1-f2.
36 ... Ng5 37 Nd2?!
So White has covered f3, but he should probably just put the knight back on c5 and
allow Black his trick: 37 Nc5 Nxf3! 38 Kxf3 Bxg4+ 39 Kf4 Bxe2 40 Rxe2 and while
White is slightly worse, it is hard to imagine him losing from this position.

37 ... Nxg4!
A bolt from the blue! It seems likely that Sokolov had missed this coup.
38 fxg4 Bxg4 39 Ndf1
If White saves the rook with 39 Rf2 then 39 ... Rxe3 40 Rxe3 Rxe3 41 Be2 Bxe2,
followed by 42 ... Rd3, leaves him struggling in a rather simple endgame. So it’s
understandable that he prefers to give up the exchange so as to complicated matters.
39 ... Bf3+ 40 Kf2 Bxe2 41 Kxe2?
Losing another pawn, after which White is on the brink. 41 Nxe2! was better, even if
Black now has decent winning chances after 41 ... Rh8.
 41 ... Ne6!
Threatening ... Nxd4, and there is no way for White to avoid this.
42 Kd2 Nxd4! 43 Kc3 Nf3 44 Re2 Rh8 45 Rg2 Kf8 46 Rf2 Ne5 47 Be2 Kg7 48
Rg2 Kf6 49 e4 Rd8 50 Nd2 d4+ 51 Kb2 d3 52 Bf1 b6!

Activating his queenside pawns.

53 axb6 Rb7 54 Kc3 Rxb6 55 Nb3 Rdb8 56 Nc5 Rxb4 57 Bxd3 a5 58 Nf1 Rh8 59
Rf2+ Ke7 60 Ra2 Rh3 61 Rxa5 Rc4+ 62 Kd2 Nxd3
Changing the material balance once again. Rook and three pawns should be enough
to overcome two knights, though it doesn’t prove so easy with the clock running.
63 Nxd3 Rxe4 64 Ra7+ Kf8 65 Nf2 Rd4+ 66 Ke2 Rb3 67 Ne3
67 Nd2 was probably a better way to keep the knights coordinated.
67 ... f5 68 Rc7
 68 ... f4?!
The start of an incorrect plan. By pushing the f-pawn forward by itself, Black
renders it vulnerable to attack, and if it falls White’s task becomes considerably less
difficult. Instead, 68 ... Rb2+ was a better way to restrict the white king; for example,
69 Kf1 (if 69 Kf3 then 69 ... g5! threatening ... Rf4+, or 69 Ke1 f4 70 Ned1 Rc2,
followed by ... g6-g5) 69 ... Rf4 70 Ned1 Rc2 71 Kg1 c5 and Black should win by
advancing all three pawns.
69 Ned1 f3+? 70 Kf1 Rd6 71 Kg1 Rb5
Taking the knights leads to a drawn rook endgame; for example, 71 ... Rb1 72 Kh2
Rbxd1 73 Nxd1 Rxd1 74 Rxc6 Kg7 (or 74 ... f2 75 Rf6+) 75 Kg3 etc.
72 Ne3 Re6 73 Neg4 Rh5 74 Kf1 c5 75 Rc8+?!
Forcing the king to come forwards only helps Black. The immediate 75 Ra7,
followed by Ra3, targeting the f- and c-pawns looks more accurate.
75 ... Ke7 76 Rc7+ Kd6 77 Ra7 c4 78 Ra6+ Ke7 79 Ra7+ Kd8 80 Ra8+ Kc7 81
Ra3 Rc5 82 Nd1
Not yet 82 Rxf3?? since 82 ... c3 wins at once.
82 ... Rd6 83 Ngf2 Rcd5 84 Nc3 Rf5 85 Nfe4 Rd4 86 Ra7+?
Another useless check; 86 Ra6 g5 87 Rg6 should draw now.
86 ... Kb6 87 Rg7
 87 ... g5?
Trying to save all his pawns is no good, not least because White can just take the g-
pawn. Here 87 ... Re5 88 Rxg6+ Ka5 and ... Kb4, activating the king, would have
renewed Black’s winning prospects.
88 Kf2
Simply 88 Rxg5 Rxg5 89 Nxg5, followed by Nce4 and Kf2, would draw. White will
pick up the f3-pawn and sacrifice a knight for the c-pawn if and when it advances to c3.
But White is sufficiently active to hold the position anyway.
88 ... Rd3 89 Rg8 Re5
89 ... Kc7!? might have kept a few chances.
90 Rc8 g4 91 Rxc4 Rh5 92 Kg3 Rh3+ 93 Kxg4 f2 94 Nxf2 Rhg3+ 95 Kh4 Rxc3
96 Rxc3 Rxc3 97 Kg4 Kc6 98 Ne4 Rc1 99 Kf5 Kd5 100 Nf6+ Kd4 101 Ke6 Rg1 102
Kf5 Rf1+ 103 Ke6 Rf2 104 Ng4 Ra2 105 Kf5 Ra5+ 106 Kf4 ½-½
A narrow escape for Sokolov. Of course if he hadn’t moved his knight from c5 in
the first place, it would have been difficult for Black to do much.
Chapter Two
Exchange Variation with Nge2
Introduction

1 d4 d5 2 c4
Obviously White has to delay bringing his knight out to f3 if he wants to play the lines in
this chapter. This in turn gives Black a few ways to avoid them; for example, he can
make a feint towards a Nimzo-Indian with 1 d4 Nf6 2 c4 e6 and then meet 3 Nc3 with 3
... Bb4 and only play 3 ... d5 against 3 Nf3.

Question: Can White also adopt an English Opening move order with 1 c4 - ?

Answer: Indeed he can, though in this case he needs to be prepared for various
other systems starting with 1 ... c5 and 1 ... e5.
2 ... e6 3 Nc3 Nf6

Question: In the previous chapter you mentioned Alatortsev’s 3 ... Be7. Would this
prevent the lines in this chapter?

Answer: Yes it would, though White has 4 cxd5 exd5 5 Bf4, when he still has plans
which delay Nf3 and can potentially put the knight on e2.
4 cxd5 exd5 5 Bg5 Nbd7 6 e3 Be7 7 Bd3 c6
 8 Qc2

Question: Does it matter whether White plays 8 Qc2 0-0 9 Nge2 or 8 Nge2 at once?

Answer: They can in fact transpose after 8 Nge2 0-0 9 Qc2, but there are a number
of differences. The main one is that after the immediate 8 Nge2 White gets an extra
option against the simplifying 8 ... Nh5 (which is quite a good answer to 8 Qc2); this
line runs 9 Bxe7 Qxe7 10 g4 Nhf6 11 Ng3, and now Ulf Andersson played 11 ... h6 in
Game 24.
Black probably does better to play 8 ... h6 first, when 9 Bh4 Nh5 10 Bxe7 Qxe7
saves a tempo on the Andersson game.
Another possibility is 8 ... Nf8, which can be used against both 8 Nge2 and 8 Qc2.
After 8 Nge2 Nf8 9 Qc2 Ne6, White played 10 Bxf6 in Game 25 and 10 Bh4 in Game
26.
8 ... 0-0
Going into a line which is recommended for White in a number of repertoire books.
To avoid confusion over move orders I think this is the easiest way to start declining
the Queen’s Gambit. As one’s understanding and appreciation of the different nuances
grows, there’s a case for looking at Black’s alternatives here:
a) 8 ... Nf8 9 Nge2 transposes to 8 Nge2 Nf8 9 Qc2 above.
 b) 8 ... Nh5 has proven to be very popular among top players and with quite good
reason. White doesn’t seem to get much with 9 Bxe7 Qxe7 10 0-0-0 (as in Game 27),
while after 10 Nge2 Black seems fine with either 10 ... Nb6 (as in Game 28) or Max
Euwe’s 10 ... g6 (see Game 29).
9 Nge2 Re8 10 0-0 Nf8

Black is playing all the same moves he uses against the Exchange Variation with
Nf3, which we covered in the previous chapter. The advantage of this plan is simplicity;
we only need one set-up for these two flavours of the Exchange. The disadvantage here
is that White has a new plan based on creating a central pawn duo with f2-f3 and then
e3-e4, and Black needs to adjust in order to combat it.
11 f3 Be6 12 Rad1
A logical way to prepare e3-e4; White sees that the main problem will be the
weakness of his d4-pawn and looks to reinforce it.
Another move to have been played here is 12 Rae1, after which 12 ... Rc8 13 Kh1
N6d7 14 Bxe7 is the main line. Having struggled with 14 ... Rxe7 (see Game 30), Black
switched to the superior 14 ... Qxe7, echoing the main line with 12 Rad1. I’ve presented
two games from this position: 15 Qd2 is featured in Game 31, and 15 Nf4 was played in
Game 32.
12 ... Rc8 13 Kh1
Biding his time with another preparatory move. The immediate 13 e4!? seems a bit
premature (see Game 33). White can jockey for position in other ways besides 13 Kh1;
for example, 13 a3 was played in Game 34. This also leads to a complex struggle in
which White’s chances seem somewhat preferable.
13 ... N6d7
This typical manoeuvre again, offering the exchange of dark-squared bishops. It’s
not the only move here, and in Game 35 we’ll look at 13 ... a6!?.
14 Bf4
Exchanging bishops with 14 Bxe7 Qxe7 would lead to similar play to Games 30 and
31.
14 ... b5!?

We are following Lei.Williams-E.Dervishi, Bled Olympiad 2002 (Game 36). The

text is a very interesting move which commences counterplay on the queenside.

Game 24
J.Speelman-U.Andersson
GMA World Cup, Reykjavik 1991

1 d4 Nf6 2 c4 e6 3 Nc3 d5 4 cxd5 exd5 5 Bg5 c6 6 e3 Be7 7 Bd3 Nbd7

 Question: Is Black’s move order a big deal at this stage? Different GMs seem to
have different ways of reaching the same position.

Answer: Particular move orders can matter a lot. For example, Black often just
plays 5 ... Be7 and 6 ... 0-0. By delaying castling, he maintains the option of going long,
as we’ll see in the notes below and again in Game 27. Similarly, White can choose
whether to give priority to Nge2 or Qc2. We’ll begin our coverage with the former.
8 Nge2

8 ... Nh5
Rather risky in the current situation because of White’s 10th move. If Black wants to
play the ... Nh5 plan against 8 Nge2, it is better to start with 8 ... h6. For example, 9 Bf4
(note that 9 Bh4 Nh5 10 Bxe7 Qxe7 11 g4 Nhf6 gains a tempo on Andersson’s play in
the main game) 9 ... Nh5 10 Bg3 Nxg3 11 hxg3 Nf6 12 Qc2 Qa5 13 a3 Bd7 14 Nc1 Bd6
15 Nb3 Qc7 16 Ne2 0-0 17 Nc5 Bc8 18 Rc1 Re8 19 0-0 Qe7 gave Black a comfortable
position in A.Tolush-G.Lisitsin, Leningrad 1956.
Another plan worth considering is the immediate 8 ... Nf8, which will be examined
in the next two games.

Question: Can’t Black just play 8 ... 0-0 here?

 Answer: Indeed he can; then 9 Qc2 transposes into the main line considered in
Games 30-36.
9 Bxe7 Qxe7 10 g4 Nhf6 11 Ng3

11 ... h6
This move is the argument in favour of 8 ... h6. Black understandably wants to
prevent g4-g5.

Question: I don’t like the look of White’s knight coming to f5. Does Black have
another option here?

Answer: Yes, he does; 11 ... Nb6 is a possibility. Even so Black’s position is not a
bed of roses; for example, 12 g5 Ng8 13 h4 g6 14 Qe2 Be6 15 0-0-0 0-0-0 16 Kb1 h6
17 f4 Qd7 18 Qc2 Ne7 19 Rdf1 Kb8 20 h5 hxg5 21 fxg5 Qc7 22 Qf2 gxh5 23 Nxh5
Rdg8 24 Qf6 was pleasant for White in M.Prusikin-P.Dittmar, Schwäbisch Gmünd
2006.
Ulf Andersson subsequently tried 11 ... g6, but then 12 g5 Ng8 13 h4 h6 14 h5!?
looks very dangerous for Black.
12 Nf5
In G.Kasparov-B.Spassky, GMA World Cup, Barcelona 1989, White resisted this
tempting knight move and opted for 12 h3. After 12 ... Nb6 13 Qd2 Bd7 14 b3 g6 15 a4
a5 16 f3, White had a pleasant space advantage quite similar to our main game. Of
course the issue in this line is not in getting a nice-looking position for White but in then
breaking through Black’s rock-like defences.
12 ... Qf8

13 h4
White has tried other moves at this point without noticeably improving his game
from being somewhat better than Black’s. For example:
a) 13 e4 g6 14 Ne3 dxe4 15 Bxe4 Nxe4 16 Nxe4 Qb4+ 17 Qd2 Qe7 18 f3 Nf6 was
fine for Black in A.Vaisser-R.Fontaine, French Championship, Narbonne 1997.
b) 13 h3 was Speelman’s later effort in J.Speelman-P.Peelen, Leeuwarden 1997. But
despite his much higher rating, the Englishman failed to make much progress after 13 ...
g6 14 Ng3 Qe7 15 Qe2 Nb6 16 0-0-0 Bd7 17 Kb1 0-0-0 18 Ka1 Kb8 19 Rb1 Ne8,
when Black’s position proved a very tough nut to crack.
13 ... g6 14 Ng3 Nb6 15 f3
Protecting both g4 and e4, which leaves White with a nice space advantage.
15 ... Qd6 16 Nce2 h5
 This seals the kingside but allows White to increase his space further. Instead, 16 ...
Bd7 was worth considering, after which 17 h5 0-0-0 18 hxg6 fxg6 19 Bxg6? is bad
because of 19 ... Nxg4!, discovering an attack on the g6-bishop. Naturally White doesn’t
need to play 19 Bxg6 or even 17 h5, and one would think he’d maintain a pleasant
position with a number of normal moves.
17 g5 Ng8 18 Kf2
The king stands well here, regardless of the loss of castling rights.
18 ... Ne7 19 a4 Be6

Question: Why didn’t Black play 19 ... a5, sealing the ‘hole’ on b4?

Answer: Probably because it wouldn’t prevent White from opening a file on the
queenside with Qd2, Rhb1 and b2-b4. Meanwhile, with his dark-squared bishop long
gone, Black will struggle to establish a piece on b4.
20 Qc2 Nd7 21 Nf4 Nf8 22 b4 Bd7
If 22 ... Qxb4 then 23 Rhb1 will pick up the pawn on b7.
23 Qc5 Qc7! 24 b5
 White is certainly making the running at this point, but he has yet to achieve
something concrete.
24 ... Ne6 25 Nxe6 Bxe6 26 bxc6 bxc6 27 Rab1 0-0 28 Rb4 Rab8 29 Rhb1 Rxb4
30 Rxb4
If 30 Qxb4, Black can obtain counterplay with 30 ... Bf5!. As the game goes on
White’s king position becomes something of an issue after all.
30 ... Nf5 31 Bxf5 Bxf5 32 Rb3 Rb8 33 Rxb8+ Qxb8 34 Nxf5
Essentially agreeing to a draw.
 34 ... Qh2+ 35 Kf1 gxf5 36 Qxc6
36 Qxa7 would have led to the same result after 36 ... Qh3+ 37 Ke2 Qg2+ 38 Kd3
Qxf3 39 Qb8+ Kg7 40 Qf4 Qd1+ 41 Kc3 Qe1+ and so on.
36 ... Qh1+ 37 Ke2 Qg2+ 38 Ke1 Qg1+ 39 Ke2 Qg2+ 40 Ke1 Qg1+ ½-½

Game 25
Quan Zhe-K.Spraggett
Canadian Championship, Toronto 2004

1 d4 d5 2 c4 e6 3 Nc3 Nf6 4 cxd5 exd5 5 Bg5 Be7 6 e3 c6 7 Bd3 Nbd7 8 Nge2 Nf8
 An idea worth noting, especially after White’s 8 Nge2.

Question: What is the idea exactly?

Answer: Black brings his knight to e6 without first castling and playing ... Rfe8, so
it saves a tempo in getting there.

Question: What’s the big deal about having the knight on e6 if White moves his
bishop away from g5?

Answer: The attack on the bishop gains a tempo to then play ... g7-g6 and ... Ng7,
after which Black can bring either his bishop or knight to f5. If Black’s knight lands on
f5 it will gain further time on White’s bishop, while after ... Bf5 Black will carry out the
invaluable exchange of light-squared bishops. A further idea is to exchange dark-
squared bishops with ... Nh5 and keep the ... Ng7 idea for later.
9 Qc2 Ne6 10 Bxf6
Giving up the bishop pair in this particular position is of dubious merit as the plan
that’s usually associated with Nge2 is to play f2-f3 and e3-e4. This will be very
difficult with Black’s dark-squared bishop bearing down on d4. The usual move here is
10 Bh4 which we’ll see in the next game.
10 ... Bxf6 11 0-0 g6 12 b4

As central action is more or less ruled out, a minority attack is the only show in
town for White, but Black is quite well placed to meet this as well.
12 ... a6 13 Na4 0-0 14 Nc5 Qd6 15 Rac1 Bd8!?
This, together with Black’s previous move, is very rare at GM level. Lining his
queen and bishop up against h2 prompts White to create weaknesses in his king
position.
16 a4 Bc7 17 Ng3 h5 18 Rfe1 h4 19 Nxe6 Qxe6
Black could have considered an exchange sacrifice with 19 ... hxg3 20 Nxf8 gxf2+;
for example, 21 Kxf2 Qxh2 22 Rh1 Qg3+ 23 Kf1 Kxf8 gives him a pawn and some
positional compensation. But the move played is also good.
20 Nf1 Bd6 21 b5 Bb4 22 Red1 Bd7 23 bxa6 bxa6
 White has succeeded in inflicting pawn weaknesses on the queenside, but he lacks
the piece support to tie Black down. In fact Spraggett soon takes the initiative there
himself as he forces ... c6-c5 through.
24 e4
White evidently doesn’t like the way things are going so tries to change the nature of
the game, but opening the position up can only be good for Black’s bishop pair.
24 ... Rfe8 25 e5 c5 26 dxc5 Rac8 27 Qa2 Bxc5 28 h3 Ba7 29 a5?!
This makes matters worse. 29 Rxc8 Rxc8 30 Be2 is superior, although 30 ... Qxe5
31 Rxd5 Qf6 32 Bxa6 Rc7 then leaves White facing threats of 33 ... Qxa6 and 33 ...
Be6.
29 ... Rxc1 30 Rxc1 Qxe5 31 Bxa6?
In grabbing this pawn, White sets himself up for tactical shot on the kingside.
Instead, 31 Qd2 was relatively best, though the position after 31 ... Rb8 is still
unattractive for White.
31 ... Bxh3!
 32 Rd1?
Not 32 gxh3? either, since 32 ... Qg5+ picks up the rook on c1. The best chance was
still 32 Qd2, when the win is rather more complicated: 32 ... Bxg2! (simply 32 ... Bf5
looks good too) 33 Kxg2 Qe4+ 34 f3 h3+ 35 Kh2 (or 35 Kxh3 Qxf3+ 36 Ng3 Re3 37
Qg2 Qh5 mate) 35 ... Qxf3 36 Rc3 Bb8+! 37 Ng3 Bxg3+ 38 Kxh3 Qh1+! 39 Kxg3 Qg1+
40 Kf3 Re4! and White has no satisfactory defence according to the engine, if you care
to check.
32 ... Bxg2! 33 Kxg2 h3+ 34 Kxh3 Qh5+ 35 Kg2 Qxd1
White could have saved himself the rest.
36 Qb2 Qa4 37 Qd2 Qe4+ 38 f3 Qh4 39 Qxd5 Qf2+ 40 Kh3 Kg7 41 Ng3 Rh8+
42 Kg4 f5+ 0-1
Mate follows after 43 Kg5 Qxg3 or 43 Kf4 Qe3, while 43 Nxf5+ gxf5+ 44 Kxf5
Rh5+ won’t last much longer.

Game 26
Zsu.Polgar-B.Spassky
Women vs. Veterans, Prague 1995

1 d4 Nf6 2 c4 e6 3 Nc3 d5 4 cxd5 exd5 5 Bg5 Be7 6 e3 Nbd7 7 Bd3 c6 8 Qc2

 Our first official example with 8 Qc2, though 8 Nge2 Nf8 9 Qc2 just transposed in
the previous game.
8 ... Nf8 9 Nge2

Question: What if White puts his knight on f3 here?

Answer: This does look like a better option for White, but Black still seems to have
a decent game. P.Harikrishna-B.Grachev, Russian Team Championship 2008, went 9
Nf3 Ne6 10 Bh4 g6 11 0-0 Ng7 12 b4 a6 13 Bxf6 Bxf6 14 a4 0-0 15 b5 a5 16 bxc6
bxc6 17 Ne2 Bd7 18 Rab1 with only a tiny plus for White.
9 ... Ne6 10 Bh4 g6 11 f3
The most thematic plan for White after her Nge2 development, but Black is not
without resources. Note that there are several ways to reach the position after White’s
13th move, such as 11 0-0 0-0 12 f3 (or 12 Rad1 b6 13 f3) 12 ... b6 13 Rad1. These
different routes seem to be equally valid, though 11 f3 allows some additional options.
11 ... 0-0

12 0-0
White has tried other moves too:
c) 12 0-0-0 a5 13 g4 b5 gives Black good attacking chances on the queenside.
 a) 12 g4 invites the immediate 12 ... c5, after which 13 Qd2 cxd4 14 exd4 Bd7 15
Bg3 Re8 16 0-0-0 b5 gave Black excellent play in J.L.Chabanon-B.Ivkov, Cannes 1995.
c) 12 Rd1 b6 13 Bf2 Bb7 14 h4 c5 15 dxc5 Bxc5 16 Kf1 Rc8 17 Qa4 a6 18 b4 b5
19 Qb3 Bd6 was also nice for Black at this stage in Z.Azmaiparashvili-A.Beliavsky,
Groningen 1994.
12 ... b6
An effective set-up for Black, who develops his queen’s bishop while at the same
time preparing the thematic ... c6-c5.
13 Rad1 Bb7 14 Kh1
N.Brunner-B.Spassky, French League 2002, varied at this point with 14 Bf2, after
which 14 ... Rc8 15 h3 c5 16 dxc5 Nxc5 17 Nf4 Bd6 18 Qd2 d4 19 exd4? Bxf4 20 Qxf4
Nxd3 21 Rxd3 Ba6 22 Bh4 Nh5! 23 Qg4 f5 won the exchange for Black and later the
game.
14 ... Rc8
It’s too soon for Black to play 14 ... c5? because of 15 dxc5 bxc5 16 Bc4!,
exploiting the pin along the d-file.

15 a3
Hoping to restrain Black’s ... c6-c5 advance by playing b2-b4. A number of other
moves have been tried in this position, but Black’s chances seem quite adequate:
a) 15 e4?! is premature: 15 ... dxe4 16 fxe4 (after 16 Nxe4 Nxe4 17 Bxe7 Nf2+! 18
Rxf2 Qxe7, White's strategy has already failed, G.Grosse Frintrop-R.Kilian, Ruhrgebiet
1997) 16 ... Ng4 17 Bf2 Nxf2+ 18 Rxf2 b5! and Black seizes the initiative, eyeing up
the weak d4-pawn (not yet 18 ... Nxd4?! 19 Nxd4 Qxd4 because of 20 Ba6).
b) 15 Qc1 is best met by the energetic 15 ... c5! (15 ... Rc7 16 e4 dxe4 17 fxe4 Rd7
18 e5 Nd5 19 Bxe7 Qxe7 20 Nxd5 cxd5 21 Rf6 was nice for White in B.Kouatly-
J.Piket, France vs. Holland, Cannes 1990); for example 16 dxc5 Rxc5 17 Qd2 Ne4! 18
Nxe4 dxe4 19 Bxe7 Qxe7 20 Bxe4 Bxe4 21 fxe4 Ng5 is at least equal for Black.
c) 15 Bf2 should probably be answered by something like 15 ... Qc7 (15 ... c5 16
dxc5 bxc5 17 Bc4 was awkward for Black in A.Vaisser-E.Bricard, French League
2007), after which 16 Bg3 (16 e4 dxe4 17 fxe4 Ng4 18 Bg1 Bg5 19 Rf3 c5 gives Black
excellent play) 16 ... Bd6 17 Bh4 Be7 (17 ... Nh5 is also possible) 18 Bg3 Bd6 could
result in a surprising draw by repetition.
d) 15 Qa4 a6 (15 ... a5 16 a3 Nd7 17 Bxe7 Qxe7 18 e4 dxe4 19 fxe4 Nf6 20 h3 was
a bit better for White in A.Alvarez Pedraza-O.Rodriguez Vargas, Barcelona 2011) 16 a3
c5 17 dxc5 Nxc5 18 Qc2 Qc7 gives Black a very decent position.
15 ... Nh5 16 Bxe7 Qxe7 17 Qa4 a6
This time 17 ... a5 was possibly a better way to safeguard the a-pawn.

18 e4
Here 18 Qb4!? would have been interesting, when Black’s best line may be 18 ...
Qc7 19 Na4 b5 20 Nc5 Nxc5 21 Qxc5 Rfe8 22 e4 Nf6 23 e5 Nd7 24 Qc3 c5 with
counterplay. The possibility of 18 Qb4!? is why Black should consider 17 ... a5.
18 ... dxe4 19 fxe4 b5 20 Qa5
20 Qc2 would also be answered by the thematic 20 ... c5!, when 21 d5 c4 traps the
bishop on d3.
20 ... Rfd8 21 Bb1 Nf6 22 h3
Black can meet 22 Qb6 with 22 ... Ng4!, threatening to come into e3.
22 ... c5 23 d5 Ng5 24 Rxf6?
This looks good at first but White has missed something. She should have played the
mundane 24 Ng3, when the chances are still fairly balanced.
24 ... Qxf6 25 h4 Ne6!

Presumably Polgar overlooked this move, cleverly saving the apparently trapped
knight.
26 Qb6
After 26 dxe6, Black wins the piece back with 26 ... Rxd1+ 27 Nxd1 Qf1+ etc.
26 ... Qxh4+ 27 Kg1 Ba8 28 Rf1
Now 28 dxe6 Rxd1+ 29 Nxd1 Qe1+ 30 Kh2 Qxe2 is even worse for White.
28 ... Ng5 29 Qf6 Qg4 30 Bd3 c4 31 Bc2 a5 32 Nd4
Note that 32 Nxb5 is met by 32 ... Rxd5! (rather than 32 ... Qxe2?! 33 Qxg5 Qxc2 34
Qf6! Rf8 35 Nd6 Rc7 36 Nf5 gxf5 37 Qg5+ and White survives) 33 exd5 Bxd5 34 Rf2
Nh3+ 35 Kf2 Nxf2 26 Qxf2 Rd8 and Black is clearly better.
32 ... b4 33 Bd1?! Qg3 34 axb4 axb4 35 Ncb5
Or 35 Nce2 Qe3+ 36 Kh1 Nxe4 etc.
35 ... Rd7 36 Bf3 Qh4 37 Bd1
The threat was ... Nh3+ (with shades of Game 19) and 37 Qa6 Bb7 doesn’t help
White.
37 ... Qg3 38 Bf3 h6 39 Nc6
Tricks are not going to work against a former World Champion, but White’s position
is already hopeless; e.g. 39 Ne2 Qb8 or 39 Qa6 Bb7 40 Qa4 Nxe4 41 Nc3 Qe5! etc.
39 ... Bxc6 40 dxc6 Rxc6!

40 ... Rd2 would win as well, but Spassky’s choice is much more crisp.
41 Qxc6 Nxf3+ 42 Rxf3 Rd1+ 43 Rf1 Qe3+ 0-1
White loses the rook, and the engine declares mate in nine.

Game 27
J.Loxine-S.Atalik
Dresden 2005

1 c4 e6 2 Nc3 d5 3 d4 Nf6 4 cxd5 exd5 5 Bg5 Be7 6 e3 c6 7 Bd3 Nbd7 8 Qc2

 In his notes to the game Atalik said he was more afraid of 8 Nge2, as White can then
meet 8 ... Nh5 with 9 Bxe7 Qxe7 10 g4 Nhf6 11 Ng3. We saw this plan in Game 24.
8 ... Nh5

Question: Isn’t Black breaking opening principles by moving an already developed

piece again?

Answer: As an exchange of pieces is helpful to Black in this line, there’s a case for
doing this straight away. He can live with the loss of time because the position is not an
open one.
9 Bxe7 Qxe7 10 0-0-0

10 ... Nb6
Having seen that White intends to castle queenside Atalik makes plans to send his
king that way too. Going kingside with 10 ... g6 11 h3 Ng7 12 Nf3 Nf6 13 g4 0-0, as in
E.Lobron-B.Ostenstad, European Team Championship, Haifa 1989, is very dangerous
for Black.
11 h3
The ... Ng7 and ... Bf5 plan is a good reason for White to be ready to play g2-g4.
Instead, Black can meet the immediate 11 Nf3 with 11 ... Bg4, but it might be better if he
resists that temptation. M.Dlugy-G.Kasparov, Internet (blitz) 1998, saw 11 ... Be6 12
Kb1 g6 13 Rhe1 0-0-0 14 Ka1 Kb8 15 Rb1 Ng7 16 Nd2 Bf5 17 f3?! Bxd3 18 Qxd3
Ne6 19 b4?! Rhe8 20 Rb3?! f5 21 a4?! f4, when White’s position quickly imploded.
11 ... g6 12 Nf3
White can always play 12 g4 first. Either way f3 is the best square for White’s
knight. 12 Nge2 is less effective here, and in E.L’Ami-I.Sokolov, Dutch Championship,
Boxtel 2011, Black got good play after 12 ... Ng7 13 g4 Bd7 14 Nf4 g5! 15 Nfe2 h5 16
Ng3 0-0-0 17 Nf5 Bxf5 18 gxf5 Ne8 19 Rhe1 g4 20 hxg4 hxg4 21 Ne2 Nf6 22 Rh1
Rh3!.
12 ... Ng7 13 g4 Be6

14 Ne5?!
This knight sally doesn’t make a lot of sense because Black can just drive it away
with ... f7-f6. White’s usual move is 14 Kb1, after which A.Barsov-M.Gurevich, Bastia
(rapid) 2005, went 14 ... 0-0-0 15 Na4 Kb8 16 Nc5 Bc8 (16 ... Ne8 at once is also very
reasonable) 17 Rc1 Ne8 18 a4 Nd6 19 b3 Nd7 20 a5 a6 (instead, 20 ... Nxc5 21 Qxc5
Qf6?! 22 Ne5 Qxf2 looks very risky after 23 Rhf1 Qxe3 24 Rf3! Qd2 25 Rf6 Be6 26 a6)
21 Nxd7+ Bxd7 22 Qc5 and now 22 ... f6 appears critical and not bad for Black (rather
than 22 ... Nc8 as played).
14 ... 0-0-0
The immediate 14 ... f6!? is interesting too. If White tries 15 Bxg6+, then 15 ...
Kd8!? (15 ... Kf8 16 Bd3! fxe5 17 dxe5 would offer decent compensation for the
sacrificed piece) 16 Bd3 fxe5 17 dxe5 Kc7 makes it hard for him to justify his play.
15 Na4 f6 16 Nxb6+
White seems to be drifting here, as this just strengthens Black’s pawn structure.
White should accept that 14 Ne5 was inaccurate and just go back with 16 Nf3, when 16
... Kb8 17 Nc5 (or 17 Qc5 Qc7) 17 ... Bf7 is still fairly equal.
16 ... axb6 17 Nf3 Ne8 18 Rhg1
White starts a plan to advance on the kingside which only worsens his position. He
would have been better off playing the solid 18 Kb1, when 18 ... Nd6 is nice for Black
but it won’t be easy for him to win.
18 ... Nd6 19 h4 Qd7!

A very annoying move.

20 Nh2
White can’t have been happy playing this move, but 20 g5 is met by 20 ... Bg4!.
20 ... Kb8 21 Kb1 Rc8
The immediate 21 ... c5 was also possible because Black can meet 22 dxc5 with 22
... Rc8 23 c6! bxc6, but there’s certainly no harm in further preparation.
22 Qe2
22 Qb3 Qc7 23 Rc1 doesn’t stop Black playing 23 ... c5.
 22 ... c5 23 g5 c4 24 Bc2 c3!

Black’s attack has materialized with surprising speed.

25 b3
Taking the pawn loses quickly after 25 bxc3? Nb5, while 25 gxf6? runs into 25 ...
Nc4! 26 b3 Nd2+ 27 Rxd2 (or 27 Ka1 Qd6 28 Nf3 Qa3 29 Rb1 Kc7! etc) 27 ... cxd2 28
Qxd2 Rxc2! 29 Kxc2 (or 29 Qxc2? Bf5) 29 ... Qc7+, winning the knight on h2.
25 ... Nb5 26 Ka1 Na3 27 Rc1 fxg5 28 hxg5 Bf5
Playing 28 ... Nxc2+ first is probably even stronger.
29 Bxf5 Qxf5 30 Nf3
Here 30 f4 looks tempting in order to defend the g5-pawn and secure the e5-square
for the knight. The problem is White’s weakness on e3; for example, 30 ... Rhe8 31
Nf3? (31 Rg2 is better but then 31 ... Re7 32 Ng4 b5 33 Ne5 b4 looks like a decisive
bind) 31 ... Rxe3! 32 Qxe3 Qb1+! 33 Rxb1 Nc2 with a sort of smothered mate is a
spectacular exploitation.
30 ... Rhf8 31 Ne5
31 Rg3 b5 followed by 32 ... b4 would again leave White totally tied up.
31 ... Qxf2 32 Rg2 Qxe2 33 Rxe2 Rf5
 34 Rg2
After 34 Rh2 Rxg5 35 Rxh7, another nice finish is possible: 35 ... Rg2 36 Rg7 Rb2!,
followed by 37 ... Rb1+ and 38 ... Nc2 mate. Instead, 34 e4 is an attempt to get
counterplay but Black keeps control; for example, 34 ... dxe4 35 Rxe4 Rxg5 36 d5 Rg2
37 Nd7+ Kc7 38 Nf6 Rc2 39 Rxc2 Nxc2+ 40 Kb1 Na3+ 41 Kc1 Rf8 42 Re7+ Kd8 43
Rd7+ Kc8 44 Rd6 h5, rolling his passed pawns forwards.
34 ... Ka7 35 Nd3 Rf3?
Not the best. Black could gradually run White out of moves with 35 ... b5!; for
example, 36 Rg3 Kb6 37 Rh3 Rc7 38 Rg3 Ka5 39 b4+ Ka4 and if 40 Nc5+ then 40 ...
Rxc5! 41 bxc5 Rf2, threatening 42 ... Rb2 and 43 ... Rb1+ again, as well as 42 ... b4 and
43 ... b3, or 42 ... Nc2+ and 43 ... Nb4.
36 Nb4 Rxe3 37 Nxd5 Rd3 38 Rxc3 Rdxc3 39 Nxc3 Rxc3 40 Kb2 Nb5 41 a4
Regaining the piece, after which White has decent drawing chances, as long as he
can prevent the black king from becoming active.
41 ... Rd3 42 axb5 Rxd4 43 Kc3 Rd5 44 Kc4 Rc5+ 45 Kb4 Kb8 46 Rh2 Rc7 47
Rh4 Kc8 48 Rf4 Re7
 49 Rd4?
The immediate 49 Kc4 would have been more precise, when it looks hard for Black
to make progress.
49 ... Kc7 50 Kc4 Re5?
This rook move was based on an oversight and could have led to a draw. 50 ... Rd7
straight away was correct.
51 Rh4 Re7
Atalik originally intended 51 ... Rc5+ 52 Kb4 Rxg5, but then 53 Rxh7+ Kd6 54
Ka4! is a draw, since Black will lose both his b-pawns.
52 Rd4?
Giving Black renewed chances. Instead, 52 Kd5 Kd8 53 Rf4 Ke8 54 Rf6 should
hold.
52 ... Rd7 53 Rf4 Kd6 54 Rd4+?!
Aiming to try his luck in the king and pawn endgame, but he loses straightforwardly
by a tempo. It would have been better to stay in the rook endgame, where things aren’t
quite so easy for Black. For example, after 54 Rf6+ Ke5 55 Rf1 (55 Rxb6 Kf5 will get
two connected passed pawns on the kingside) 55 ... Rd4+ 56 Kc3 Rg4 57 Rf7 Kd5 58
Rxh7 (58 b4 Rg3+ 59 Kb2 Kc4 60 Rxh7 Kxb4 61 Rxb7 Kxb5 would see Black going
two pawns up) 58 ... Kc5 59 Rxb7 Rxg5 60 Rc7+ Kxb5 61 Re7, Black has a bit of work
to do yet, even if the tablebase does declare a win for him.
 54 ... Ke6 55 Rxd7 Kxd7 56 Kd5 Ke7 57 Ke5 Kf7 58 b4 Kg7 59 Kd6 h5 60 gxh6+
Kxh6 61 Kc7 g5 62 Kxb6 g4 63 Kxb7 g3 64 b6 g2 65 Kc8
After 65 Ka6, Black wins with 65 ... g1Q 66 b7 Qa1+ 67 Kb6 Qe5 68 Ka7 (68 b5
Qb8 is easy) 68 ... Qd4+ 69 Ka8 Qa1+ 70 Kb8 Kg5 and so forth.
65 ... g1Q 66 b7 Qg8+ 67 Kc7 Qc4+ 0-1

Game 28
E.Bareev-U.Bönsch
German League 2000

1 d4 Nf6 2 c4 e6 3 Nc3 d5 4 cxd5 exd5 5 Bg5 c6 6 Qc2

Question: Why didn’t White play the usual 6 e3, when he can meet Black’s set-up in
the game with 6 ... Be7 7 Bd3 Nbd7 8 Nge2 Nh5 9 Bxe7 Qxe7 10 g4 - ?

Answer: Bareev probably wanted to avoid the line 6 e3 Bf5, after which 7 Qf3 Be6
8 Bxf6 Qxf6 9 Qxf6 gxf6 produces a difficult queenless middlegame where White’s
superior pawn structure is offset by Black’s bishop pair. For instance, S.Conquest-
U.Andersson, Spanish Team Championship 2004, continued 10 Bd3 Nd7 11 Nge2 Nb6
12 Nf4 Bd6 13 g3 Bxf4 14 gxf4 f5 15 Rg1 Ke7 16 Ke2 Rag8 17 Rg3 Rg6 18 f3 Rh6 19
Rg2 Nc8 20 Kf2 Nd6 21 Ne2 Rg6 22 Rag1 Rxg2+ 23 Rxg2 Kf8 24 Rg3 h6 25 Rg1 Rg8
26 Ng3 Rg6 and White was unable to make progress.
There are quite a few move order subtleties in the QGD Exchange with which both
players can try to steer their opponents into something they may find less acceptable.
6 ... Be7 7 e3 Nbd7 8 Bd3 Nh5
As mentioned before, this is a far more acceptable idea for Black when the white
queen has already gone to c2.
9 Bxe7 Qxe7 10 Nge2
 Bareev opts for a much quieter treatment than the 10 0-0-0 Nb6 11 h3 g6 12 Nf3 of
the previous game.
10 ... Nb6 11 Rb1
It was still possible to play 11 0-0-0, but as we noted in Game 27, the knight is
better placed on f3 when White castles long.
11 ... g6
Covering the h7-pawn so as to be able to castle short, while preparing ... Ng7 and ...
Bf5. In the next game we’ll see Euwe play 10 ... g6 on the previous move.
Black has also tried a preliminary 11 ... a5 12 0-0 and only then 12 ... g6. For
example, T.Hellborg-F.Hellers, Swedish Team Championship 2007, went 13 Na4 Nxa4
14 Qxa4 Bd7 15 Qb3 Qb4 16 Qc3 0-0 17 a3 Qd6 18 Qc5 Qb8 19 a4?! b6 20 Qc3 Qa7
21 b4 axb4 22 Rxb4 c5 23 dxc5 bxc5 and Black’s position was already preferable.
12 b4 a6 13 a4 0-0
 14 a5
Seeing that Black is well prepared for the minority attack, Bareev chooses another
path.
After 14 b5 Black can capture with either pawn; for example, 14 ... cxb5 (the
standard response 14 ... axb5 15 axb5 c5 is okay too) 15 axb5 (15 a5 b4 16 axb6 bxc3
is also fine) 15 ... a5 with a passed pawn to compensate for the weakness on d5.
Alternatively, White could just castle with 14 0-0, when Ki.Georgiev-U.Andersson,
French League 2000, continued 14 ... Be6 15 Rfc1 Rfc8 16 a5 Nd7 17 Na4 Re8 18 Nc5
Nxc5 19 bxc5 Rad8 20 Rb3 Bc8 21 Qd2 Qg5 22 Kh1 Nf6 23 Ng1 Re7 24 Nf3 Qh5 25
Ne5 Ng4 26 Nxg4 Qxg4 27 Re1 Rde8 28 Rbb1 h5 29 Re2 Qg5 30 Qe1 Bf5 31 Bxf5
Qxf5 32 Reb2 Qc8 33 Qd1 Rc7 34 Rb3 Ree7 and White was nominally a bit better but
struggled to make further progress, since opening the kingside is fraught with risk.
14 ... Nd7
Going forward with 14 ... Nc4 looks suspicious as, after 15 Bxc4 dxc4, the c4-pawn
would have to be protected with pieces.
15 Na4 f5!
 Question: Doesn’t that make the c8-bishop really ‘bad’?

Answer: It looks ugly but Black wants to prevent White’s plan of e3-e4. For this ...
f7-f5 is very useful.
16 0-0 Ndf6
16 ... f4?! is strategically desirable but premature, as it opens the position while
Black’s queenside is still undeveloped. Then 17 Nxf4 Nxf4 18 exf4 Rxf4 19 Rbe1 Qg7
(19 ... Qxb4? 20 Bxg6! is a decisive attack) 20 g3 Rf8 21 Qb2 is nice for White, who
has potential outposts on e5 and c5 and an active pawn majority on the kingside,
whereas Black’s queenside majority can be easily blockaded.
17 Qc1 Be6 18 Bc2 Ne4 19 Rb3 g5 20 f3 Nd6 21 Nc5 Rae8 22 Re1 Bc8
 This completes Black’s regrouping, after which he can claim full equality. In fact
White needs to be careful here due to the actively placed enemy forces on the kingside.
23 Ng3 Nxg3 24 hxg3 f4!
Unleashing his light-squared bishop.
25 g4
White is still absolutely fine after 25 gxf4 gxf4 26 e4; for example, 26 ... Nf5 27
Qxf4 Nxd4 28 Qg3+ Qg7 29 Qxg7+ Kxg7 30 Rc3 produces what will probably be a
drawn endgame.
25 ... Qg7 26 Qd1 Nb5
26 ... Nc4 would be answered by 27 e4, since 27 ... Ne3 allows White a powerful
exchange sacrifice with 28 Rbxe3!.
27 Bd3 h5 28 Bxb5 axb5 29 gxh5 g4 30 exf4 g3!?
 In his enthusiasm for developing his initiative, Black raises the stakes. Otherwise he
could have opted for 30 ... Rxe1+ 31 Qxe1 Qxd4+ 32 Qe3 Rxf4, when the position is
equal.
31 Rbe3
The cold-blooded 31 Re5! would have forced Black to find some accurate moves to
stay in the game: 31 ... Qf6 32 Qe1 Qh4 33 Qe3 (getting ready to meet 33 ... Qh2+ 34
Kf1 Qh1+ with 35 Qg1) 33 ... Rxe5! 34 fxe5 Rf4! and now, for example, 35 e6 Rxd4 36
Rb1 Qh2+ 37 Kf1 Qh1+ 38 Qg1 Qh4 39 e7 Kf7 with a probable draw.
Maybe Bareev was short of time at this point, which might explain why he avoided
this and why Bönsch played so flamboyantly for the attack.
31 ... Rxe3 32 Rxe3 Qf6 33 Kf1
Heading for the hills is the right idea here. Not 33 Qd3? Bf5 34 Qc3 Qh4 and
Black’s attack suddenly gets nasty; for example, 35 Nxb7 Qh2+ 36 Kf1 Qh1+ 37 Ke2
Qxg2+ 38 Kd1 Qf1+ 39 Re1 g2 etc.
33 ... Qh4 34 Ke2 Qh2 35 Qf1 Bh3 36 Kd2 Bxg2
 37 Qe1??
The decisive mistake. Amazingly, it seems that after 37 Qb1 again leaves Black
needing to play precisely to maintain equality: 37 ... Bxf3+! (not 37 ... Bh3+? 38 Re2 g2
due to 39 Qg6+ Kh8 40 Qh6+ Kg8 41 Ne6 and wins) 38 Kc3 Bxh5 39 Re7 Rf7 (or 39
... Bf7 40 Re5) 40 Re8+ Rf8 41 Re7 with a draw by repetition.
37 ... Bh3+ 38 Re2 g2 39 Qf2 Qxf4+ 0-1
After 40 Kc2 Qxf3 41 Qxf3 Rxf3, White has to give up his rook for the g-pawn.

Game 29
C.Kottnauer-M.Euwe
Amsterdam 1950

1 c4 e6 2 Nc3 d5 3 d4 Nf6 4 Bg5 Nbd7 5 e3 c6 6 cxd5 exd5 7 Bd3 Be7 8 Qc2 Nh5 9
Bxe7 Qxe7 10 Nge2 g6
 This time, rather than insert 10 ... Nb6, Black just plays ... g7-g6 straight away.

Question: Is the point to prepare a later ... Ng7 and ... Bf5 again?

Answer: This is one of the ideas which often comes into play. In addition, Black is
preparing to castle short by covering h7, while the knight on h5 is ready to exchange
White’s king’s knight should it come to g3 or f4.
11 0-0
Instead, 11 0-0-0 Nb6 12 h3 is a set-up considered within the notes to Game 27,
where we noted that White’s knight is better placed on f3 than e2.
11 ... 0-0 12 Ng3?!
Alternatively, 12 Rab1 is very similar to the previous game and could easily
transpose. For instance, 12 ... Nb6 13 b4 a6 14 a4 was the actual route taken in
Georgiev-Andersson (see the note with 14 0-0 there).
Another possibility is 12 Rae1, but the problem for White is that the f2-f3 and e3-e4
plan will be rather difficult to execute. A.Beliavsky-U.Bönsch, German League 2004,
continued 12 ... Nb6 13 Nc1 Be6 14 Nb3 Nd7 15 Na4 Rac8 16 Qd2 Nhf6 17 f3 b6 18
Rc1 c5 19 dxc5 Nxc5 20 Nbxc5 bxc5 with a comfortable game for Black whose hanging
pawns are balanced by the weakness on e3.
12 ... Nxg3 13 hxg3 Nf6
 With White’s remaining knight so far from e5, Black could well consider 13 ... f5
here.
14 Rfe1 Qd6 15 a3 Kg7

Question: What’s that move for? Shouldn’t Black be trying to drum up an attack on
the kingside?

Answer: This is in fact an attacking move! Black is getting ready to play ... h7-h5, ...
Rh8 and ... h5-h4.
16 b4 a6 17 Na4
Going for a minority attack with 17 Rab1 would now be too slow after 17 ... h5 18
a4 Rh8 and ... h5-h4. So White decides instead to try and interfere with Black’s efforts.
17 ... h5 18 Qc5 Qe6
Obviously Black doesn’t want to exchange his queen as that will be his best
attacking piece.
19 Nb6 Rb8 20 f3
Preparing to keep kingside files closed by meeting ... h5-h4 with g3-g4 and at the
same time keeping Black’s pieces out of e4 and g4.
20 ... Rd8
 21 e4?!
But this is a poor follow-up, after which Black gets a great outpost for his knight. 21
Re2 was better, when 21 ... Nd7 22 Nxd7 Bxd7 would leave the position about equal.
21 ... dxe4 22 fxe4 Ng4! 23 Bc4
It’s hard to believe that White allows the black knight to establish itself on g4. He
would be well advised to remove it as soon as possible with 23 Be2 and Bxg4. It seems
that Kottnauer must have missed something.
Perhaps he originally intended 23 e5, but now realized that after 23 ... Qb3 24 Rad1
Be6 he wouldn’t have the kind of dynamic play he’d need to compensate for his many
pawn weaknesses.
23 ... Qd6
Now it’s all about White’s weak d4-pawn.
24 Qxd6 Rxd6 25 Rad1 Be6 26 Rd3?!
Another inferior move. 26 e5 was White’s best here, after which 26 ... Rdd8 27
Bxe6 fxe6 28 Nc4 Nh6 29 Nd6 Rd7 30 Kf2 Nf5 31 Nxf5+ gxf5 gives Black the superior
rook endgame but nothing instantly decisive.
26 ... Rbd8 27 Red1 Bxc4 28 Nxc4 Re6 29 Nd2?!
Again 29 e5 was the best try, but then 29 ... f6 30 exf6+ Kxf6 intends ... Nh6-f5 with
nothing but pain for White to look forward to.
 29 ... Ne5!
A nice tactical blow that wrecks White’s position.
30 d5
Of course 30 dxe5 loses to 30 ... Rxd3.
30 ... Nxd3 31 dxe6 Ne5!?
31 ... Nb2 32 Rb1 Na4 was also good, but not 32 ... Rxd2?? 33 e7 and White wins.
32 exf7
Ftacnik suggests 32 e7 Re8 33 Nf3 Rxe7 34 Nxe5 Rxe5 35 Rd7, presumably
considering that White has more drawing chances in the rook endgame, though it is hard
to imagine him holding this one, especially against Euwe.
Perhaps White should have tried the tricky 32 Nf1!?, when 32 ... Rf8 (again not 32
... Rxd1?? because of 33 e7) 33 exf7 Kxf7 34 Ne3 Ke6 is good for Black, but he hasn’t
won any material yet.
32 ... Nxf7
Note that 32 ... Nc4?? is a blunder because of 33 f8Q+!.
33 Kf2 Rd3 0-1
Given all his pawn weaknesses this endgame is hopeless for White, so Kottnauer
resigned.
Game 30
E.Bareev-K.Asrian
New York Open 1998

1 d4 d5 2 c4 e6 3 Nc3 Nf6 4 cxd5 exd5 5 Bg5 Be7 6 e3 c6 7 Bd3 Nbd7 8 Qc2 0-0 9
Nge2
This development of the king’s knight became a popular weapon for White after
Mikhail Botvinnik’s patronage. White introduces the possibility of creating a pawn
centre with f2-f3 and e3-e4.

Question: Does White still have the minority attack plan?

Answer: Yes, he still has that option, but usually he goes for the central expansion.
9 ... Re8 10 0-0 Nf8 11 f3
Preparing to play e3-e4 while ruling out ... Ne4 by Black.
11 ... Be6

Black has tried virtually every move here that doesn’t lose material at once. The
most significant alternatives are 11 ... Nh5, 11 ... Ng6, and 11 ... g6, but we’ll stick with
our familiar 11 ... Be6 set-up from the previous chapter.
 12 Rae1

Question: That’s different. Doesn’t the rook normally go to d1?

Answer: This is one of several alternatives to 12 Rad1, which we’ll look at later in
the chapter. The rook may turn out to be well placed on e1 but there are questions over
the vulnerability of the d-pawn when White plays e3-e4.
White has also tried 12 Bh4, which aims to avoid the exchange of dark-squared
bishops and bring the bishop back to f2. There’s certainly some logic to this idea but it
is rather slow, and this in turn makes it possible for Black to develop counterplay. L.Van
Wely-J.Piket, Antwerp 1996, continued with the simple and direct plan of playing for ...
c6-c5: 12 ... Rc8 13 Rad1 a6 14 Kh1 Ng6 15 Bf2 c5 16 dxc5 Bxc5 17 Nd4 Bxd4 18
exd4 Nf4 19 Rfe1 Nxd3 20 Qxd3 Qb6 21 Rd2 Qc6 22 Rde2 b5 23 a3 h6 24 g4 Qc4 25
Qd2 Nd7 with approximately equal play.
12 ... Rc8 13 Kh1 N6d7 14 Bxe7 Rxe7
It does now seem that this is the worst of Black’s two recaptures, though the reasons
for this deserve exploring. For the superior 14 ... Qxe7 see Games 31 and 32.
15 Nf4

15 ... Nf6
 Asrian is looking to put pressure on White’s centre should he advance with e3-e4.
This and his next move would almost certainly have been thought up at home, though it
seems that something goes horribly wrong with his preparation, which might explain
why he later switched to 14 ... Qxe7.
Instead, 15 ... Rc7 was played in G.Kasparov-U.Andersson, GMA World Cup,
Belfort 1988, but seems a bit too passive. Kasparov took the centre with 16 Qf2 Nf6 17
e4, after which 17 ... dxe4 18 fxe4 Rcd7 19 d5! cxd5 20 Bb5 Rc7 21 exd5 Bd7 22 Be2
left Black in a difficult position.
16 Qd2 b5
Looking to destabilize the knight on c3, which in turn reinforces e4.
17 e4 b4 18 Na4 dxe4 19 fxe4 Qxd4?

If this was ‘prepared’, Black is in for a shock because there’s a big problem with it.
19 ... Bxa2 was a better try, though it’s still unpleasant after 20 Qxb4 Ne6 21 Nxe6
Bxe6 22 Nc5 with a dominating position for White.
20 e5!
Threatening both the knight on f6 and the win of Black’s queen by 21 Bxh7+. White
should be winning here, until he in turn slips up.
20 ... Bf5 21 exf6 Rxe1 22 Qxe1 Bxd3 23 Qd2 Rd8 24 Rd1 Qxf6 25 Nc5??
Apparently picking up the pinned bishop on d3 but missing a great tactic. Instead,
simply 25 Nxd3 Ne6 26 Qc2 leaves Black a piece down for inadequate compensation.
 25 ... Bc2!
Boom! After this one move White goes from an almost certain win to a position in
which he’s scrambling for a draw. As Frank Marshall once said, attention is more
important than concentration.
26 Qxd8
26 Qxc2? is answered by 26 ... Qxf4, when 27 Rxd8? Qf1 is mate.
26 ... Bxd1 27 Qxf6 gxf6 28 Kg1 Ne6 29 Nfxe6 fxe6 30 Nxe6

So White is just a pawn down in a minor piece endgame. He does at least have
some compensation because of Black’s ragged structure and a weakness on the dark
squares which gives the knight some outposts. All the same, a pawn is a pawn, or at
least it ought to be.
30 ... Bc2 31 Kf2 Be4 32 Nc5 Bd5?
Here 32 ... Bb1 would have given Black more chances. After the move played
White pulls close to equality.
33 b3 a5 34 g3 Kf7 35 Ke3 Ke7 36 Kd4 h5 37 Nb7 a4?!
And after this White is on top again! 37 ... Ke6 38 Nxa5 Kf5 39 Nc4 Kg4 would
draw.
38 bxa4 Bxa2 39 a5 Kd7 40 Nc5+ Kc7 41 Nd3 b3 42 Kc3 b2 43 Kxb2 Bf7 44 Kc3
Bg6 45 Nc5 Bf5 46 Kd4 Kb8 47 Ne4 Ka7 48 Nxf6 Bg6 49 Kc5 Ka6 50 Kxc6 Kxa5 51
h4 Kb4 52 Kd5 1-0
White is now winning since the h5-pawn is going to fall. Given the deterioration in
Black’s play towards the end, I would guess that he was in terrible time trouble. One
can only describe this as a “fluctuating struggle”!

Game 31
C.Lutz-A.Yusupov
Tilburg 1993

1 c4 e6 2 Nc3 d5 3 d4 Nf6 4 cxd5 exd5 5 Bg5 Be7 6 e3 0-0 7 Bd3 Nbd7 8 Nge2 Re8
9 0-0 c6 10 Qc2 Nf8 11 f3 Be6 12 Rae1 Rc8 13 Kh1 N6d7 14 Bxe7 Qxe7

This looks better than 14 ... Rxe7 from the previous game.
15 Qd2

Question: So why isn’t White getting on with it by playing 15 e4 - ?

Answer: Because Black can then break up the newly-formed centre with 15 ... dxe4
16 fxe4 c5, after which 17 d5 is effectively answered by 17 ... c4 18 dxe6 Qxe6 19 Nd4
cxd3 20 Qxd3 Ne5 with a good game for Black.
 Alternatively, 15 Ng3 c5 (15 ... Nb6 16 Qf2 c5 17 a3 a6 18 f4 f6 19 e4 dxe4 20
Ncxe4 cxd4 21 Nh5 gave White a strong attack in E.Bareev-A.Yusupov, Frankfurt rapid
2000) 16 Qf2 Nf6 17 dxc5 Qxc5 18 Nb5 was G.Pap-M.Stojanovic, Serbian Team
Championship 2012, and now 18 ... Ng6 (rather than 18 ... Bd7 19 Nd4 Ne6 20 Ngf5 as
played) 19 Qd2 Ne5 would have given Black good counterplay.
White’s best may be 15 Nf4 which we’ll see in the next game.
15 ... Nb6 16 b3

Question: How about 16 e4 now?

Answer: Here it’s definitely more testing, though White’s hanging pawn centre has
an element of vulnerability after, for example, 16 ... dxe4 17 fxe4 Rcd8 18 Rd1 f6. Both
sides would have chances in this kind of position.
16 ... Rcd8 17 a4
Looking to destabilize the b6-knight but giving Black a nice post for his queen on
b4.
17 ... Qb4 18 Ne4
The exchange of queens ensures the long-term security of Black’s king, which in turn
lessens the effect of a central pawn advance by White. 18 Bc2 seems more in tune with
the features of the position.
18 ... Qxd2 19 Nxd2 Bd7!
 20 Ng3
Not 20 a5? due to 20 ... Rxe3, winning a pawn, while 20 Nf4 would also be
answered by 20 ... a5.
20 ... a5 21 Re2?!
It seems hereabouts as if White didn’t fully appreciate how Black could develop his
position. 21 Kg1 makes more sense, in order to bring the king to f2 and defend e3.
21 ... Ne6 22 Rc1 Nc8
The knight is en route to a better post on d6. White should still bring his king to f2
but instead starts drifting badly.
23 f4?! Nd6 24 Nf3 f6 25 Kg1 Re7 26 h4?!
A further weakening of White’s position, after which it becomes critical. 26 Rce1
was better.
26 ... Rde8 27 Kf2 Nd8!
 This knight is heading for the newly weakened g4-square via f7 and h6.
28 Ree1 N8f7 29 Nh2?!
He should have dispatched the other knight to h2 with 29 Nf1 as this one is needed
to help restrain the ... g7-g5 lever.
29 ... Nh6 30 Ngf1 Bf5 31 Be2
Yusupov considers that White should have exchanged bishops here, though 31 Bxf5
Nhxf5 32 Nf3 Ne4+ 33 Kg1 h5, followed by an eventual ... g7-g5 still looks pretty
unpleasant for White.
31 ... Ne4+ 32 Kg1 g5! 33 g4?
A time-trouble error which makes matters much worse by allowing a decisive
combination. White’s best line was 33 hxg5 fxg5 34 Nf3, after which 34 ... gxf4 35 exf4
Rf8 is good for Black but by no means winning for him.
 33 ... gxf4! 34 gxf5
Losing quickly, but there’s not much to be done now. On 34 exf4 Black wins with 34
... Nd2! 35 Nxd2 Rxe2 36 Rxe2 Rxe2 37 Nhf3 Bxg4 etc.
34 ... Rg7+ 35 Bg4
35 Kh1 Nf2 is mate.
35 ... Nxg4 36 Nxg4 Rxg4+ 37 Kh2 f3 38 Rc2 Rxh4+ 39 Kg1 Kf7 0-1

Game 32
K.Sakaev-K.Asrian
Dubai 2000

1 d4 d5 2 c4 e6 3 Nc3 Nf6 4 cxd5 exd5 5 Bg5 Be7 6 e3 0-0 7 Bd3 Nbd7 8 Nge2 Re8
9 0-0 c6 10 Qc2 Nf8 11 f3 Be6 12 Rae1 Rc8 13 Kh1 N6d7 14 Bxe7 Qxe7
Evidently rethinking the 14 ... Rxe7 of Game 30. In fact the queen recapture is now
widely accepted as Black’s best line.
15 Nf4
 When someone of Sakaev’s pedigree plays in a particular way it should be taken
seriously. The immediate 15 Qd2 was examined in the previous game.
15 ... Qd6 16 Qd2
A high-level encounter, Ma.Carlsen-V.Ivanchuk, Leon (rapid) 2009, varied here
with 16 Qf2, after which 16 ... f6!? 17 Nxe6 Nxe6 18 f4 Ndf8 19 e4 dxe4 20 Nxe4 Qd5
21 Rd1 f5 22 Nc3 Qxd4 23 Bxf5 Qxf2 24 Rxf2 Rcd8 led to a draw in the endgame.
16 ... Rcd8
Alternatively, 16 ... f6 is similar to Carlsen-Ivanchuk in the note above, and looks
quite playable for Black. Asrian’s move is met sharply by Sakaev.
17 e4!? dxe4 18 Nxe4 Qc7
The point behind White’s 17th move is that 18 ... Qxd4 is answered by 19 Nd6!,
though this only leads to a marginally favourable endgame for White after 19 ... Ng6 (19
... Qxd6?? 20 Bxh7+ wins the queen) 20 Nxe6 (20 Nxe8? Qxf4 loses material) 20 ...
Rxe6 21 Nxf7 Rxe1 22 Rxe1 Rf8 23 Rd1 Nc5 24 Bxg6 Qxd2 25 Rxd2 hxg6 26 Ng5 Rf5
27 h4 Rd5 28 Re2. I’d be impressed if either player had seen all that at the board.
19 Nh5
There’s no time to lose if White wants to squeeze something out of this position.
After 19 b3 Bf5, Black is very comfortable.
 19 ... f6
Fearing his opponent’s attack Asrian sensibly covers the g5-square. He refrains
from grabbing on a2 as this would allow White to seize the initiative; for example, 19 ...
Bxa2?! 20 Neg3! (White has nothing special after 20 Qg5 Ng6, while 20 Ra1?! Bd5 21
Rxa7 Ne6 leaves Black slightly better) 20 ... Rxe1 21 Rxe1 Ne6 22 Nf5 Ndf8 (not 22 ...
g6?? 23 Qh6 and wins) 23 b4! g6 (this is forced; 23 ... Bd5?? loses to 23 Nfxg7! Nxg7
24 Nf6+ Kh8 25 Qh6 etc) 24 Nh6+ Kh8 25 Qxa2 gxh5 26 Qa1 and White has good
compensation for the pawn.
20 Qb4! Bd5
Now 20 ... Bxa2?? is just bad because of 21 b3 followed by 22 Qa3.
21 Bc4 Re6 22 Bxd5 cxd5 23 Nc3
Instead, 23 Neg3 might have been slightly more challenging, though Black is okay
after 23 ... Rde8.
23 ... Nb6
Here 23 ... Nb8! was a good idea, as on c6 the knight will hit the d4-pawn.
24 Rxe6
Possibly 24 Qa5 was more annoying for Black.
24 ... Nxe6 25 Re1 Qf7 26 Ng3 Re8 27 Nf5 Nf4 28 Rxe8+ Qxe8 29 Qe7 Qxe7
30 Nxe7+
 It’s not clear who was playing for what at this point as the endgame looks equal to
me; and it continues to be more or less equal until the players agree a draw.

30 ... Kf8 31 Nf5 Ne6 32 Kg1 g6 33 Nd6 Nxd4 34 Nxb7 Nc6 35 Nc5 Ke7 36 Nb5
Nd7 37 Nd3 Nde5 38 Nc5 Nd7 39 Nb3 Nde5 40 N3d4 a6 41 Nxc6+ Nxc6 42 Nc3 ½-
½

Game 33
J.Timman-A.Yusupov
Tal Memorial, Riga 1995

1 c4 e6 2 Nc3 d5 3 d4 Nf6 4 cxd5 exd5 5 Bg5 Be7 6 e3 0-0 7 Bd3 Nbd7 8 Nge2 Re8
9 0-0 c6 10 Qc2 Nf8 11 f3 Be6 12 Rad1
This is the main line; White overprotects the d4-pawn in preparation for pushing in
the centre with e3-e4.
12 ... Rc8
 Question: I’ve seen these last two moves before, with Black aiming to meet a later
b2-b4-b5 with ... c6-c5. But I thought that White was playing for e3-e4 and not a
minority attack?

Answer: That’s a good point. Actually, the idea here is to play for ... c6-c5 anyway,
without waiting for b4-b5. The reason is that there will be compensation for the isolated
d-pawn Black contracts, and not least because e3 was weakened by White’s 11 f3.
13 e4!?
This is clearly a critical move. White’s plan is to mobilize his centre pawns so why
not do it straight away?
White also has a number of preparatory moves, aiming to go e3-e4 at a better
moment. 13 a3 will be covered in Game 34, while for 13 Kh1 see Games 35 and 36.
13 ... dxe4 14 fxe4 Ng4
 The critical reply, setting up some tactical possibilities and getting ready to
undermine White’s control of e5 with ... c6-c5. This square could become a great
outpost for the knight on g4.
15 Bf4
Other moves fail to trouble Black:
a) 15 Bxe7 Qxe7 16 Qd2 c5 17 d5 Bd7 18 h3 Ne5 saw Black achieve his strategic
aim with a very comfortable game in A.Vaisser-G.Flear, Fuerteventura 1992.
b) 15 Bd2 c5 16 d5 c4! 17 dxe6 Nxe6 wins the piece back with advantage.
c) 15 Bc1 c5 16 Bb5 is strongly met by 16 ... cxd4! 17 Bxe8 Bc5! (17 ... Qxe8 was
worse in K.Mekhitarian-G.Milos, Sao Paulo blitz 2011) 18 Kh1 Qh4 19 Bf4 dxc3 with
much the better game for Black.
15 ... Ng6 16 e5
Again the critical move, which makes this a game of great theoretical value.
Instead, 16 Bc1 is again strongly met by the thematic 16 ... c5!, after which 17 d5 c4
18 dxe6 Qb6+ 19 Kh1 Qxe6 recovers the piece with the better game for Black.
16 ... Bg5
At first the engines like 16 ... Nxf4, but they change their minds after 17 Bxh7+ Kh8
18 Nxf4, since 18 ... Ne3 19 Qd3 Nxf1 (19 ... Nxd1 20 Nxe6 fxe6 21 Rxd1 is even
worse) 20 Rxf1 Rf8 21 Rf3 gives White a strong attack for the sacrificed exchange,
although 21 ... Bg5 and 22 ... Bh6 may well defend.
 17 Bxg6
Not 17 Bxg5? Qxg5, threatening both 18 ... Ne3 and 18 ... Qe3+.
17 ... hxg6 18 Qd2 Bxf4 19 Nxf4 Bc4 20 Rfe1
After 20 Rf3, Black’s 20 ... c5! is even stronger than the game.
20 ... c5

21 Ne4
After 21 e6, Black can defend with 21 ... cxd4 (rather than 21 ... Bxe6?! 22 Nxe6
Rxe6 23 Rxe6 fxe6 23 Qf4 Nf6 24 dxc5 and White regains the pawn with advantage) 22
Qxd4 Qxd4+ 23 Rxd4 g5 24 exf7+ Bxf7, reaching a roughly equal endgame.
Instead, Timman suggested 21 d5!? as an improvement. This was tested almost 20
years later in S.Sjugirov-A.Yusupov, Doha 2014, which continued 21 ... Rxe5 22 b3
Ba6 23 d6 Qg5? 24 Ne4 Qh6 25 h3 Nf6 26 Nxf6+ gxf6 27 Qf2, when White was clearly
better and went on to win. Going back, Black had to find 23 ... Qh4! 24 g3 Qf6, after
which he should be okay; for example, 25 Ne4 (or 25 h3 Rxe1+ 26 Rxe1 Ne5 27 Qe3
Re8 28 Ne4 Qf5 29 g4 Nxg4 30 hxg4 Qxg4+ 31 Kf2 Bb5 with three pawns and
sufficient play for the piece) 25 ... Qf5 26 Nd5 (or 26 h3 Nf6 27 Nxf6+ gxf6 28 Rxe5
fxe5 29 Nd5 Qf3 30 Ne7+ Kg7 31 Nxc8 Be2! and draws) 26 ... Rxe4 27 Rxe4 Qxe4 28
Ne7+ Kh7 29 Re1 (not yet 29 Nxc8? Ne5 30 Qg2 Qg4 and White is in trouble) 29 ...
Qf3 (showing the point of encouraging g2-g3) 30 Nxc8 b6 31 Qf4 Qxf4 32 gxf4 Bc8 33
Re7 Nf6 with a level endgame.
 21 ... cxd4
This natural move sacrifices the exchange, but Black will get excellent
compensation.
22 Nd6 d3

23 Nh3?!
A slip which hands Black the advantage. 23 Nxc8?? is very bad because of 23 ...
Qd4+ 24 Kh1 Nf2+ 25 Kg1 Nxd1+ 26 Kf1 Ne3+ 27 Rxe3 Qxe3 28 Qxe3 d2+ and Black
wins. But White could have maintained equality after 23 Nxd3 Bxd3 24 Qxd3 Rxe5 25
Rxe5 Qb6+ 26 Qe3! Qxe3+ (not 26 ... Nxe3?? 27 Nxc8 and White gets too much for the
queen) 27 Rxe3 Nxe3 28 Re1 Rc6 29 Nxb7 and the endgame should be drawn.
23 ... Rxe5! 24 Nxc8
Forced. 24 Rxe5?? Qxd6 25 Re4 Qxh2+ 26 Kf1 f5 wins for Black.
24 ... Qxc8 25 Rxe5 Nxe5 26 Qe3 f6 27 Nf2
White’s other possibility was 27 Qxa7, though he’s certainly on the back foot after
27 ... Be6 28 Nf2 Qc2 (if 28 ... Bg4 White can hang on with 29 Ra1) 29 Qe3 Bf5 due to
the strength of Black’s d-pawn.
27 ... Bd5 28 Rc1
The d-pawn cannot be taken: 28 Nxd3? loses to 28 ... Qg4 29 Rd2 Nf3+ etc.
28 ... Qf5 29 b3
 White might have played 29 Rc3, after which 29 ... Bxa2 30 Nxd3 Bc4 31 Nf2 Bd5
32 Qd4 Bc6 is nice for Black in view of his beautifully entrenched minor pieces, but it’s
not clear that he’ll be able to make progress.
29 ... Ng4

30 Qd4
There were a couple of ways for White to lose on the spot:
a) 30 Qxa7?? d2 31 Rf1 Qf4 32 Nxg4 Qxf1+ 33 Kxf1 d1Q+, picking up the knight
on g4.
b) 30 Nxg4 Qxg4 31 Qd2 Bxg2! 32 Qxg2?? d2! and Black wins the rook back to
reach a winning queen endgame.
30 ... Nxf2 31 Qxf2 Qg5 32 Re1
32 Rd1 Be4 holds the monster pawn on d3.
32 ... f5 33 Qg3 Qxg3
If 33 ... Qd2 then 34 Qb8+ Kh7 35 Qg3 sets up 36 Qh4+ and draws.
34 hxg3 Kf7 35 Kf2 Kf6 36 Ke3 Be4 37 Rc1 a6 38 Rc7 Bxg2 ½-½
An interesting and well-played game by both players.

Game 34
B.Gelfand-J.Piket
Wijk aan Zee 1998

1 d4 d5 2 c4 e6 3 Nc3 Nf6 4 cxd5 exd5 5 Bg5 c6 6 Qc2 Be7 7 e3 0-0 8 Bd3 Nbd7 9
Nge2 Re8 10 0-0 Nf8 11 f3 Be6 12 Rad1 Rc8 13 a3

Rather than go straight in with 13 e4 as in the previous game, Gelfand tries first to
inhibit Black’s counterplay with ... c6-c5.

Question: If White’s going to prepare e3-e4 are there other ways to do it?

Answer: Yes indeed, and in fact several other moves have been tried here. The most
important of these is 13 Kh1, which we’ll examine in the next two games.
13 ... a6
This is a typical move, preventing White’s pieces from landing on b5 after Black’s
intended ... c6-c5. Several others have been tried too:
a) 13 ... N6d7 14 Bf4 Ng6 15 Bg3 Bh4 was played in G.Kamsky-P.Van der Sterren,
Wijk aan Zee (4th matchgame) 1994, but White kept an edge after 16 b4 Bxg3 17 Nxg3
Nf6 18 Qd2 a5 19 Bb1 axb4 20 axb4 and then took the initiative with 20 ... b6 21 f4
Bg4 22 Rc1 Bd7 23 f5 Ne7 24 e4.
b) 13 ... Ng6 14 Kh1 c5 was a very direct approach used in D.Sharavdorj-
G.Kacheishvili, Lubbock 2009. Probably rather too direct, as after 15 dxc5 Rxc5 16
Nd4 Qc8 17 Qb1 a6?! (17 ... Ne5!? 18 Bxf6 Bxf6 19 Bxh7+ Kh8 might have been a
better practical try) 18 Na4 Ra5 19 Rc1 Qd7 20 Bxf6 Bxf6 21 Nc5 Qe7 22 Ndb3, Black
got his rook trapped and soon had to resign.
14 Na4
Again inhibiting Black’s ... c6-c5. Another way to do this is with 14 b4, but then 14
... a5 15 Qb2 axb4 16 axb4 Ng6 17 Kh1 h6 18 Bxf6 Bxf6 19 Bxg6 fxg6 20 e4 dxe4 21
fxe4 Qb6 gave Black active play with his bishop pair in M.Cebalo-D.Brumen, Croatian
Team Championship 1999.
In either case, the issue with reverting to queenside play is that White needs to look
after the weakness on e3. On the other hand, in M.Illescas Cordoba-D.Campora,
Spanish Team Championship, White allowed Black to get his ... c6-c5 break in 14 Bh4
Ng6 15 Bf2, when 15 ... c5 16 dxc5 Bxc5 17 Bxg6 hxg6 18 Qd2 Qc7 19 Rc1 Qe5 20
Nd4 Bd7 left Black with a comfortable game.
14 ... N6d7 15 Bxe7 Qxe7 16 b4

16 ... a5
Counter-attacking b4. There was another interesting option here in 16 ... b5!?, when
17 Nc5 (17 Nac3 is met by 17 ... Nb6 and 18 ... Nc4) 17 ... Nxc5 18 dxc5 (18 Qxc5
Nd7!? would route the knight towards c4 again, after which the weakness of the
backward c6-pawn is largely cosmetic) 18 ... Bd7 19 e4 dxe4 20 Bxe4 f5 21 Bd3 Qe3+
22 Kh1 Rcd8 sees Black obtain a very satisfactory position.
 17 Rb1
After this Black emerges with a very reasonable game. The critical reply was 17
bxa5, when 17 ... Qxa3 (17 ... Ra8 18 Rb1 isn’t easy for Black either, whereas 18 Qc3
Qd8 19 a6 b5 20 Nc5 Nxc5 21 Qxc5 Rxa6 seems okay) 18 Rb1 Ra8 19 Nec3 Rxa5 20
Rxb7 is going to be a bit better for White because the long-term weakness of c6 is more
significant than the one on e3.
17 ... axb4 18 axb4 b6
Preventing White from coming in on c5, while renewing the idea of playing ... c6-c5
himself.
19 Rfc1 g6
Preparing his next.
20 Nf4 Bf5! 21 Re1 Bxd3 22 Nxd3 Ne6 23 Nc3 Nf6 24 Qf2 c5!

Mission accomplished, after which Black gets full equality. In fact White is the one
who needs to be careful here because of the weaknesses in his king position caused by
f2-f3.
25 dxc5 bxc5 26 b5 d4 27 exd4 cxd4 28 Ne4 Nxe4 29 Rxe4 Qa3 30 Qd2 Rc3 31
Nf4 ½-½
A possible sequel would be 31 ... Rb3 32 Rxb3 Qxb3 33 Nxe6 Rxe6 34 Qxd4 Rxe4
35 Qxe4 Qxb5 with a very drawn endgame.
Game 35
M.Breutigam-R.Rabiega
German Championship, Altenkirchen 2005

1 d4 d5 2 c4 e6 3 Nc3 Nf6 4 cxd5 exd5 5 Bg5 Be7 6 e3 0-0 7 Qc2 Nbd7 8 Bd3 Re8 9
Nge2 c6 10 0-0 Nf8 11 Rad1 Be6 12 f3 Rc8 13 Kh1

The most popular and important continuation. White does not rush in advancing his
e-pawn (as in Game 33) but spends some time preparing it.
13 ... a6!?
Somewhat rare at this particular stage; 13 ... N6d7 is the most common move, which
we’ll look at in the next game. With the text Black has the clear and direct idea of
playing ... c6-c5 and by first advancing his a-pawn one square he prevents a white
bishop or knight coming into b5.
14 a3
White adopts a similar strategy. The modest advance of the a-pawn is designed to
slow down Black’s queenside play at a later stage.
Naturally, 14 e4!? is a critical test of Black’s move order, since White is getting a
version of Game 33 with the useful Kh1 included (vs. ... a7-a6 for Black). D.Eggleston-
S.Jessel, British Championship, Douglas 2005, continued 14 ... dxe4 15 fxe4 Ng4 16
Bc1 Bg5 17 h3 Ne3!? 18 Bxe3 Bxe3 19 e5 and now 19 ... Ng6 (19 ... Qh4 should have
been answered by 20 Bf5 in the game) 20 Bxg6 hxg6 21 Ne4 Rc7 22 Nd6 Rf8 is
somewhat better for White due to his powerfully placed knight on d6, though the main
struggle still lies ahead.
14 ... N6d7 15 Bf4
Avoiding exchanges, which would help free Black’s position. After 15 Bxe7 Qxe7
16 e4 dxe4 17 fxe4, Black gets the e5-square for his knight with 17 ... c5 18 d5 Bg4.
15 ... c5

As planned. Otherwise Black might consider 15 ... b5!?, transposing into the next
game.
16 Qd2
If White has any chances of an objective advantage he should play 16 dxc5, giving
Black an isolated d-pawn, though this is not going to be an easy thing.
16 ... c4
Here 16 ... cxd4 17 Nxd4 Nb6 isn’t out of the question, but Black probably
welcomed the opportunity to get a mobile queenside pawn majority rather than an
isolated d-pawn.
17 Bb1 b5 18 e4 Nb6 19 Be3 Qd7 20 e5?!
Fixing the centre like this prepares for an advance of White’s kingside pawns, but at
the same time it becomes easier for Black to push forward on the queenside. Keeping
the central tension with 20 Nf4! would make it more difficult for Black.
20 ... a5
Preparing ... b5-b4.
21 f4 Bg4
Black has inhibited White’s kingside pawn advance for the time being as 22 f5 is
answered by 22 ... b4 23 axb4 axb4 24 Na2 c3! 25 bxc3 Nc4, leaving White unable to
avoid a loss of material.
22 Rde1 b4 23 axb4 axb4 24 Nd1

24 ... Bf5
There was a better move here in 24 ... Ra8!, when 25 f5 is well met by 25 ... Ra1.
Rabiega’s continuation is potentially very dangerous for Black.
25 Bxf5 Qxf5 26 Ng3 Qd7
A case can be made for exchanging queens with 26 ... Qd3, after which 27 Qxd3
cxd3 28 Nf2 d2 29 Bxd2 Rc2 30 Bc1 Ne6 gives Black reasonable play for the pawn.
27 f5 c3
With storm clouds gathering on the other flank, Black desperately needs counterplay.
28 bxc3 bxc3 29 Qe2
Black would meet 29 Nxc3 with 29 ... Bb4 30 Nge2 Na4 31 Rc1 Rxc3 32 Nxc3
Rc8, but this was still the best line for White.
 29 ... c2 30 Nf2 Nc4 31 Bf4 Bh4
Aiming to restrain White’s kingside ambitions. 31 Ng4 would have been answered
the same way.
32 Nd3

32 ... Nd6
Both sides are struggling to find the best moves, which is understandable in such a
complex position. Here Black should have played 32 ... Qb5! 33 Qg4 Bxg3 34 Qxg3
Kh8 or even 32 ... Qa7!? 33 Nc5 Nxe5! 34 Bxe5 Bxg3 35 hxg3 f6, when the game is
still in the balance.
33 Nc5
Now Black drops the c2-pawn and White is definitely better.
33 ... Qb5 34 Qxc2 Bxg3 35 Bxg3 Ne4 36 Qd1 Nxc5 37 dxc5 Rxc5?!
Taking with the queen was preferable, for the reason given in the next note.
 38 h3?!
This looks a precautionary move made when short on the clock. Consequently,
White missed a shot in 38 e6! fxe6 (or 38 ... f6 39 Qh5 Re7 40 Bd6) 39 f6! gxf6 40
Qg4+ Ng6 41 Rxf6, when Black is in serious trouble. (Note that after 37 ... Qxc5 etc,
Black could defend with 41 ... Qe7!, though 38 Qg4 is critical there.) Both players
missed the strength of e5-e6 over the next few moves, so it seems that the time trouble
was mutual.
38 ... Rc3 39 Kh2 Qc4 40 Rf4 Qd3 41 Rf3
Here 41 Qg4! was more testing. After the move played Black gets to bail out into a
pawn down endgame where he can set up a light square blockade.
41 ... Qxd1 42 Rxd1 Rxf3 43 gxf3 g6 44 fxg6 hxg6 45 Rxd5 Ne6 46 Ra5 Rc8 47
Ra2 Rc3 48 Kg2 g5 49 h4 ½-½

Game 36
Lei.Williams-E.Dervishi
Bled Olympiad 2002

1 d4 Nf6 2 c4 e6 3 Nc3 d5 4 Bg5 Be7 5 cxd5 exd5 6 e3 0-0 7 Bd3 Nbd7 8 Qc2 Re8 9
Nge2 Nf8 10 0-0 c6 11 f3 Be6 12 Rad1 Rc8 13 Kh1 N6d7
 Black’s main move here, hoping for the exchange of dark-squared bishops while
routing the knight over to the queenside.
14 Bf4 b5

This treatment is worth noting as GM Dervishi specializes in the Queen’s Gambit

Declined. He wants to bring the knight on d7 to the c4-square and use this as a
springboard for queenside counterplay.

Question: Doesn’t that weaken the c5-square and leave Black with a backward c-
pawn.

Answer: Strictly speaking it does, but White will find it difficult to exploit these
weaknesses because of the disposition of his forces. In fact it will be very difficult to
stop Black playing a later ... c6-c5, when any weaknesses on the c-file will disappear.

Question: But doesn’t that then give Black an isolated d-pawn.

Answer: That’s true, but he’ll have some compensation due to the weakness of
White’s e-pawn and the fact that he can bring his knight to c4.
15 a3
 This inhibits ... b5-b4 but makes the c4 outpost even more inviting. White could
consider 15 Bg3, intending 16 Nf4, but then Black can get counterplay with 15 ... b4 16
Na4 Qa5, followed by 17 ... c5.
15 ... a6
Now that the black b-pawn has been halted, it’s useful to defend it as preparation for
... c6-c5.
16 Qd2?!
This gives Black even more incitement to bring his knight to c4, since the knight will
now attack White’s queen. 16 Bg3 is a better move, but then 16 ... Nb6 17 Nf4 Nc4 18
Rfe1 Bh4 gives Black counterplay, and the immediate 16 ... c5 is possible too.
16 ... Nb6

17 e4
If White prevents ... Nc4 by playing 17 b3, intending 17 ... Bxa3 18 Ra1, Black can
answer with the consistent 17 ... Ng6 18 Bg3 c5.
17 ... Nc4 18 Qc1 Qb6 19 Bc2 Red8 20 e5
This was certainly encouraged by Black’s play against d4, but the reduction in
pressure against d5 makes it easier for Black to pursue his ambitions on the queenside.
20 ... c5! 21 Bg3 cxd4 22 Nxd4 Bc5 23 Nce2
23 Nxe6?! fxe6 would merely reinforce the d5-pawn and rid Black of his less active
bishop.
23 ... Ng6 24 h3

Preventing ... Bg4 in preparation for f3-f4.

24 ... Ncxe5?
A tempting but very risky pawn grab. Black radically prevents the creation of a
white pawn duo with f3-f4, but he had another way to handle this issue, via 24 ... a5 25
f4 Bd7, after which 26 f5 Ngxe5 27 b3 leads to great complications; for example, 27 ...
Nxa3 (27 ... Bxa3 28 Qf4 Nb2 is also possible) 28 Bxe5 Nxc2 29 Nxc2 Re8 30 Bxg7!
Kxg7 31 Nf4 d4 with a very unclear position.
25 Bxe5??
The decisive mistake. The immediate 25 f4 was critical, when 25 ... Nc6 26 Nxc6
Be3 27 Bf2! (presumably both players overlooked this computer resource) 27 ... Qxc6
(or 27 ... Bxf2? 28 Nxd8) 28 Bxe3 Qxc2 29 Qxc2 Rxc2 30 Nd4 Rxb2 31 f5 leaves
White very much on top, as after 31 ... Nf8 32 fxe6 Nxe6 33 Nc6! Rd7 34 Rxd5, Black
does not have enough for the piece.
25 ... Nxe5 26 f4 Bxd4 27 Nxd4 Nc6 28 Nxe6 fxe6
 Now Black is just a good pawn up.
29 Qb1 h6 30 Bh7+ Kh8 31 Rde1 Ne7 32 Bd3 Rf8 33 Qd1 Rf6 34 f5 Nxf5 35
Bxf5 Rxf5 36 Rxf5 exf5 37 Qxd5 Qf2
Instructive play. Not satisfied with his extra pawn, Black also wants the initiative.
38 Rd1 Re8 39 b3 Kh7 40 a4 b4 41 a5
Or similarly 41 Qd4 Re1+ 42 Kh2 Re3 and so on.
41 ... Re2 42 Rg1 Re3 43 Qc4 Re1 44 Rxe1 Qxe1+ 45 Kh2 Qe5+ 46 g3 Qxa5
With the fall of a second pawn White’s position becomes hopeless.
47 h4 Qb6 48 Qf4 Qc5 49 h5 Qc2+ 50 Kh3 Qe4 51 Qf1 a5 0-1
Since 52 ... a4 53 bxa4 b3 follows.
Chapter Three
Main Line with Bg5
Introduction

1 d4 d5 2 c4 e6 3 Nc3 Nf6 4 Bg5

Or 4 Nf3 Nbd7 5 Bg5, which transposes into the main line.
4 ... Nbd7

Question: Is that the only move?

Answer: Several different move orders are possible here, though as pointed out
earlier this one lures White into playing 5 cxd5 exd5 6 Nxd5??, after which 6 ... Nxd5 7
Bxd8 Bb4+ leads to the win of a piece. In this book almost all the lines I give have
Black playing ... Nbd7, ... c7-c6 and ... Be7, and when White plays his bishop to g5 it
doesn’t matter which order Black plays them in.
5 Nf3
 5 cxd5 would take us back to the Exchange Variation, which was covered in the first
two chapters. White could still play this on either of the next two moves, but as soon as
Black has played the unpinning ... Be7 White needs to bear in mind that Black might
also recapture on d5 with the knight.
5 ... Be7 6 e3 c6 7 Rc1

Question: That looks a bit odd. Shouldn’t White be getting his pieces out?

Answer: Essentially he’s playing a useful move while waiting for Black to develop,
since after the ‘natural’ 7 Bd3 Black can gain time to mobilize his queenside with ...
d5xc4, followed by ... b7-b5, ... a7-a6 and ... c7-c5. We have a couple of examples of
this plan in Games 37 and 38 with White then retreating his bishop to b3 and d3
respectively. Note too the approach that Mikhail Gurevich took in Game 48.

Question: Then couldn’t White find some more useful move, such as 7 Qc2 - ?

Answer: This has been tried as well, and quite extensively. After 7 ... 0-0 White can
play the position in two different ways. With 8 0-0-0 (Game 39) he intends some ultra-
violence on the kingside, whereas 8 Rd1 (Game 40) is a sophisticated move which tries
to inhibit Black’s freeing ... c6-c5.

Question: What about 8 cxd5 or waiting some more with 8 Rc1 - ?

Answer: 8 cxd5 goes back into an Exchange Variation after 8 ... exd5 (see Chapter
One), but as noted above Black can also consider 8 ... Nxd5 here, freeing his game; and
8 Rc1 transposes into the next note in which White plays 7 Rc1 0-0 and then 8 Qc2 or 8
a3.
7 ... 0-0 8 Bd3
 Question: So now White develops, allowing Black to win the tempo with ... d5xc4.
Couldn’t he continue to wait?

Answer: That’s an excellent question, and indeed he could. The problem is that
Black also has useful waiting moves, such as ... a7-a6, preparing ... d5xc4 and ... b7-b5,
and White runs out of useful moves first. This jockeying for tempi is illustrated in Game
41.
8 ... dxc4
This has been the main line for a considerable length of time and I don’t see a
particular reason to depart from tradition.

Question: Doesn’t it give White a central pawn majority?

Answer: Yes it does, and Black must remember that he needs counterplay after
playing this move.

Question: How does he get this counterplay?

 Answer: There are essentially two methods, both of which aim to develop the c8-
bishop. One is to play for simplifications with ... Nd5, and then eventually try to get ...
e6-e5 in. The other is to develop the bishop on b7 after ... b7-b5, when Black will aim
to hit back at White’s centre with ... c6-c5.
Black can also play an interesting preparatory move here in 8 ... a6, after which 9 0-
0 is met by 9 ... dxc4, followed by 10 ... b5, as in Game 42.

Question: What if White cuts across this plan with 9 cxd5 or 9 c5 - ?

Answer: 9 cxd5 exd5 leads to an Exchange Variation in which Black’s ... a7-a6 is
probably more useful than White’s Rac1, because the rook belongs on b1 for a minority
attack whereas ... a7-a6 will help stop a later b4-b5. Instead, 9 c5 is a serious attempt to
interfere with Black’s plans but seems more or less neutralized by 9 ... e5! as in Game
43 (which reached this position via another move order).
9 Bxc4 Nd5
Capablanca’s manoeuvre, which aims at freeing Black’s position with exchanges
and later playing for either ... e6-e5 or ... c6-c5.

Question: Can’t Black play 9 ... b5 here?

Answer: He can indeed, and I’ve included two illustrations of this plan in Games
44 and 45.
10 Bxe7
White can also try protecting the bishop with 10 h4, but then 10 ... f6 11 Bf4 Nxf4
has been known to give Black good counterplay ever since Janowski-Capablanca, New
York 1924 (Game 46).
10 ... Qxe7 11 0-0
Here too White has an alternative in 11 Ne4, attempting to avoid the exchange of
knights. Black equalized in classic style in Game 47.
 11 ... Nxc3
This is far and away Black’s most popular move, though Mikhail Gurevich has
experimented with 11 ... N5f6, as in Game 48 (which again reached the key position via
a different route).
12 Rxc3

Question: Can’t White strengthen his centre with 12 bxc3 - ?

Answer: Some strong players have tried going this way, but it’s quite good for
Black after, for example, 12 ... b6 13 e4 Bb7 14 Re1 c5 15 d5 exd5 16 Bxd5 Bxd5 17
exd5 Qd6 18 Qa4 Nf6 as in W.Browne-H.Ree, Wijk aan Zee 1980.
12 ... e5 13 Bb3
One of several options that put the onus on Black to play another move, which
usually means either capturing on d4 or pushing his pawn on to e4. The same applies
after most of White’s other possibilities:
a) 13 Qc2 exd4 is examined in Game 50, while 13 ... e4 is showcased in Game 51.
b) 13 Qb1 exd4 features in Game 52, while 13 ... e4 is considered in Game 53.
c) 13 dxe5 is the old line that has now been shorn of its terrors. A contemporary
example is Game 49.
 13 ... e4
Just as with the two queen moves above, the question is whether Black should push
his e-pawn or capture on d4. Note that the latter did not work out well in Game 54.

Question: Does Black have a useful waiting move?

Answer: It’s not easy to see one in this position. Black he really needs to commit,
and he should probably be ready to push on with 13 ... e4, which creates a double-
edged game and scores much better in practice.
14 Nd2 Nf6

The position reached is strategically quite complex with both sides having chances.
Black’s e4-pawn certainly gives him space and can form the basis of a kingside pawn
storm after due preparation. On the other hand, White’s queenside chances are not to be
taken lightly. We will examine this line further in Game 55.

Game 37
A.Burn-J.Showalter
England vs. USA, cable match 1898
1 d4 d5 2 c4 e6 3 Nc3 Nf6 4 Bg5 Be7 5 Nf3 Nbd7 6 e3 0-0 7 Bd3

A favourite move of Botvinnik that continues to attract the attention of players who
dare to be different.

Question: Doesn’t Black just take on c4 and gain time on White’s bishop?

Answer: Yes, but the follow-up is what matters. Black can choose between the
Capablanca style ... Nd5 and expanding on the queenside with ... b7-b5.
7 ... c6
The theoretical line runs 7 ... dxc4 8 Bxc4 c5 9 0-0 a6 10 a4! (otherwise ... b7-b5
gives Black a tempo on the main game, having played ... c7-c5 in one go) 10 ... cxd4 11
exd4 Nb6, but I don’t like this because of the lack of pressure on the isolated d-pawn.
B.Spassky-Le.Lengyel, Tallinn 1975, then went 12 Bb3 Nbd5 13 Ne5 Nb4 14 Re1 Bd7
15 Qe2 b6? 16 d5! Bc8 17 Rad1 Qc7 18 dxe6 Bxe6? 19 Bxe6 fxe6 20 Nd7 and Black
resigned.
8 0-0 dxc4 9 Bxc4 b5
 Opting for queenside expansion. Alternatively, 9 ... Nd5 10 Bxe7 Qxe7 11 Rc1
transposes into the main line of this chapter, as outlined in Game 48; but Black should
bear in mind that putting the rook on c1 there is not obligatory for White, who might
decide to play something else, such as 11 Ne4 or 11 e4.
10 Bb3
White far more usually retreats the bishop the other way with 10 Bd3. We’ll look at
this in the next game.

Question: Which of these two moves is better?

Answer: Well, putting the bishop on d3 prepares e3-e4-e5, driving Black’s knight
on f6 away from the defence of his king. Accordingly, I think we should see 10 Bd3 as
the critical line, though it’s not necessarily better.
10 ... a6 11 Qe2
Here 11 Rc1 is often played but seems to cause Black no problems. M.Kaabi-
K.Bischoff, Tunis 1997, went 11 ... Bb7 12 Qe2 c5 13 Rfd1 c4!? 14 Bc2 Rc8 15 e4 b4
16 Nb1 Re8 17 Nbd2 c3 and Black was already better.
11 ... c5 12 dxc5
If White doesn’t take on c5 there’s the danger that Black might set up a queenside
pawn majority with ... c5-c4, as in the previous note.
 12 ... Nxc5 13 Rfd1 Qb6 14 Bc2 b4

15 Na4
A later game varied at this point: 15 Bxf6 Bxf6 16 Ne4 Nxe4 17 Bxe4 Bb7 18 Bxb7
Qxb7 was very comfortable for Black in Mavel-J.Dobos, Hungarian Championship,
Eger 1987.
15 ... Nxa4 16 Bxa4 Bb7 17 Bb3 Qa5
Threatening ... Bxf3 and ... Qxg5.
18 Bf4 Qh5 19 Nd4
Leading to an endgame which should be equal and indeed rather drawish.
19 ... Qxe2 20 Nxe2 Rfd8 21 Bc7 Rxd1+ 22 Rxd1 a5
Looking to dislodge the bishop on b3 with ... a5-a4 while moving the queenside
pawns nearer potential promotion.
23 Bd6 Kf8
And this takes the opportunity to bring the king nearer the centre.
24 f3 a4 25 Bc2 Nd5 26 Bxe7+ Kxe7 27 Kf2 a3!?
Showalter is getting ambitious hereabouts, though there’s a danger that his queenside
pawn advance is creating weaknesses.
28 e4 Nf6
 29 Rd4
Probably missing Showalter’s clever response. He should have played 29 bxa3,
when 29 ... Rxa3 30 Bb3 leaves Black having to worry about the weakness of his b4-
pawn. This would have brought out the dark side of his 27th move.
29 ... b3! 30 Bxb3 axb2 31 Rd1?
Once again missing Black’s reply. Here White should play 31 Bc2 Rxa2 32 Rb4,
followed by 33 Bb1 and Rxb2, after which the game would have ended in a draw.
31 ... Nxe4+!
This brilliant sacrifice turns the b2-pawn into a monster.
32 fxe4 Bxe4 33 Nc3 Bg6 34 Nb1
After 34 Ke3 there follows 34 ... Rc8 35 Kd2 Rd8+ 36 Ke3 Rxd1 37 Bxd1 b1Q 38
Nxb1 Bxb1, transposing to the game.
34 ... Rc8 35 Ke3 Rc1 36 Na3 Rxd1 37 Bxd1 b1Q 38 Nxb1 Bxb1 39 a4 Kd6 40
a5 Kc5
Now it’s just a matter of picking up the a5-pawn and winning with the 4-2 majority
on the kingside.
41 Be2 Bc2 42 h4 Ba4 43 a6 Kb6 44 g4 Bd7 45 Ke4 Bc8 46 a7 Kxa7 47 Ke5
 47 ... Kb6??
A careless move, which allows White to save himself. Black should have played 47
... f6+! 48 Kd6 e5, when going round the back with the king is too slow; for example, 49
Ke7 (or 49 h5 Kb6 50 Ke7 e4 51 Kf7 f5 52 gxf5 Bxf5 53 Kxg7 e3! 54 Kf6 Bc2 55 Ke5
Kc5 etc) 49 ... e4 50 Kf7 Ba6 51 Bd1 g6! 52 Kxf6 e3 53 Kg7 Bd3! and wins.
48 Kd6
Now 48 g5 Kc6 49 Bh5! draws, as after 49 ... g6 50 Be2 Kd7 51 Kf6 Ke8, White
has the dark squares under full control and can just pass with his bishop.
48 ... Ba6 49 Bd1 Bc4 50 Ke7?!
Here 50 h5 was simpler, intending 50 ... f6 (or 50 ... f5 51 Ke5!) 51 Ke7 e5 52 Kf8
g6 53 hxg6 hxg6 54 Kg7 and so on.
50 ... e5! 51 Kf8??
Missing his final chance to draw: 51 Bc2 Kc5 (or 51 ... Bd5 52 Kd6!) 52 h5 Be2 53
Kxf7 Bxg4 54 Kxg7 Bxh5 55 Bxh7 and Black won’t get his last pawn past the white
bishop.
51 ... g6 52 Kg7 f5 53 gxf5
Or 53 Kxh7 e4 54 Kxg6 f4 55 Kf5 f3 56 Kxe4 f2 etc.
53 ... gxf5 54 Kxh7
After 54 Kf6 e4 55 Kxf5 e3 56 Kg5 e2 57 Bxe2 Bxe2 58 Kh6 Bd3, Black is in time
to protect the h7-pawn and has the right bishop to prevent the white king from sitting
tight in the h1-corner.
54 ... f4 55 Kg6 e4 56 Bg4 f3 57 Kg5 f2 58 Bh3 e3 0-1

Game 38
D.Ayupov-B.Savchenko
Russian Team Championship 2004

1 d4 Nf6 2 c4 e6 3 Nf3 d5 4 Nc3 c6 5 Bg5 Nbd7 6 e3 Be7 7 Bd3 0-0 8 0-0 dxc4 9
Bxc4 b5 10 Bd3 a6
Preparing for ... c6-c5 which would otherwise leave the b5-pawn hanging.
11 e4 c5 12 e5

12 ... Ng4!?
12 ... Nd5?! is better for White after 13 Bxe7 Qxe7 (not 13 ... Nxe7?, which allows
a winning Greek bishop sacrifice with 14 Bxh7+ Kxh7 15 Ng5+ Kg6 16 Qg4, as can be
verified by consulting any engine) 14 Nxd5 exd5 15 Qd2. For example, 15 ... h6 16
Rae1 c4 17 Bb1 Re8 18 h3 Nf8 19 Nh2, followed by f2-f4-f5, saw White build up a
decisive kingside attack in A.Mirzoev-D.Agaragimov, Baku 1999.
13 Bxe7
 E.Zhurina-E.Mukhortova, Moscow 1997, varied at this point with 13 Be4 Rb8 14
Bf4 (14 Bxe7 Qxe7 leads back into the main game) 14 ... cxd4 15 Nxd4 Ngxe5 16 Qe2
Bf6 17 Rad1 Qb6 18 Nf3 and now 18 ... b4 (rather than 18 ... Nxf3+) 19 Na4 Qa5
would have left White with very little compensation for her pawn.
13 ... Qxe7 14 Be4 Rb8 15 h3 cxd4 16 Qxd4
Leaving White with a space advantage, while forcing the black knight back to h6
where it looks quite out of play. But these advantages are quite difficult to exploit and it
seems that Black’s pieces will eventually re-emerge. On the other hand, 16 hxg4?! dxc3
17 bxc3 Nc5 is just nice for Black.

16 ... Nh6 17 Rac1 Nc5

The knight isn’t very stable on this square. Black should probably play 17 ... Nb6,
followed by ... Nf5.
18 Rfd1
Here 18 Bc6 is quite awkward for Black. His best may well be 18 ... Nf5 19 Nd5
exd5 20 Qxc5 Qxc5 21 Rxc5 Ne7 22 Bxd5 Nxd5 23 Rxd5 Bb7 with some compensation
for the pawn in the endgame.
18 ... b4 19 Ne2 Nxe4 20 Qxe4 Bb7
The bishop is very strong now and can be anchored on the d5-square.
 21 Qf4 Nf5 22 Ned4 Nxd4 23 Nxd4 Rfc8 24 Nb3 Bd5! 25 Rxc8+ Rxc8 26 Rc1
Rxc1+ 27 Qxc1 Qb7

If anything Black’s position is preferable here and White has to tread with care. He
does a good job in keeping his opponent at bay.
28 f3 h6 29 Qc5 Qb8 30 Qd4 Qb5 31 Nc1 Qc4
Instead, 31 ... a5 looks more testing. The minor piece endgame should favour the
bishop, with pawns on both sides of the board, but White’s e5-pawn gives him some
compensation in terms of space.
32 Qxc4 Bxc4 33 Kf2 f6 34 exf6 gxf6 35 Ke3 Kf7 36 Nd3 Bxa2 37 Nxb4 Bc4 38
f4 h5 39 Nd3 Bd5 40 g3 a5 41 Kd4 Bg2 42 h4 Kg6 43 Nf2 Kf5
Only White’s knight is stopping the black king coming in to g4, but this is enough.
44 Ke3 Bd5 45 Kd4 Bc6 46 Kc5 Bf3 47 Kd4 a4 48 Ke3 Bd5 49 Kd4 Bc6 50 Ke3
Bd5 51 Kd4 Bb3 52 Ke3 e5 53 fxe5 fxe5 54 Kd3 Ba2 55 Ke3 Bd5 56 Kd3 Bf3 57
Ke3 e4 58 Kd4 Bg2 59 Ke3 Ke5 ½-½

Game 39
F.Jaeger-M.Gurevich
Tromsø 2007
1 d4 d5 2 c4 c6 3 Nc3 Nf6 4 Nf3 e6 5 Bg5 Be7 6 e3 Nbd7 7 Qc2

Question: That looks like a natural move. Is White planning to castle queenside and
launch an all-out attack on Black’s king?

Answer: White does have the option of castling queenside, as we will see in this
game; but the modern interpretation is just to put the rook on d1, inhibiting ... c7-c5, and
then aim to castle kingside.
7 ... 0-0 8 0-0-0
White takes the aggressive approach. We will look at 8 Rd1 in the next game.
8 ... Nb6!?

Question: That’s an odd-looking move. What happens if White just drives

the knight back with 9 c5 - ?

Answer: This is exactly what Gurevich wanted to provoke as after bringing the
knight back to d7, he could open up the queenside with ... b7-b6.
Instead, 8 ... h6 looks natural, but White can answer this with the thematic 9 h4!. For
example, 9 ... b6 (9 ... dxc4 10 Bxc4 Qa5 11 g4 was also promising for White in
N.Kelecevic-P.Cladouras, Budapest 1992) 10 Bxf6 Nxf6 11 Ne5 Qc7 12 g4 Nd7 13 f4
Nxe5 14 dxe5 f6? 15 exf6 Bxf6 16 g5 Be7 17 gxh6 Bf6 18 Rg1 Qf7 19 Bd3 Rd8 20
hxg7 d4? 21 Bh7 mate was a demolition in A.Karpov-D.Marty, Zürich (simul) 1988.
9 Ne5
As just indicated, 9 c5 Nbd7 10 e4 b6 leads to a quick opening of lines on the
queenside.
9 ... dxc4

10 Nxc4
White has tried two other moves here:
a) 10 e4 a5 11 h4 a4 12 a3 Qc7 13 Rh3 Ra5 14 Nxc4 Nxc4 15 Bxc4 was D.Sanchez
Repullo-A.Ahmed Harjour, Asturias 2001, and now 15 ... e5! would have given Black
very good play.
b) 10 h4 was played in D.Vass-E.Ongai, Hungarian Team Championship 2011,
where 10 ... a5 11 Nxc4 Nbd5 would leave Black ready to push his queenside pawns
forward and create threats against White’s king.
10 ... Nbd5 11 h4 Nxc3 12 Qxc3 b6
This is certainly logical but it seems that Black had a better move. 12 ... Nd5 looks
stronger; for example, 13 Qd3 Nb4 14 Qb3 f6 15 Bf4 a5, followed by ... b7-b5, would
give Black a dangerous queenside attack.
13 Ne5 Bb7 14 Kb1
 Of course 14 Nxc6?? Bxc6 15 Qxc6 would lose the queen to 15 ... Rc8.
14 ... c5
Not bad, but again not the best. Black had a very strong line here in 14 ... Nd5! 15
Qc4 f6 16 Nxc6 Qd6 17 Nxe7+ Qxe7 18 Bf4 Nb4!, threatening both ... Bd5 and ...
Be4+.
15 Bb5
White should in turn get his queen out of the way with 15 Qb3.
15 ... Nd5! 16 Qb3 Rc8

I get the impression that Gurevich was taking his lower rated opponent somewhat
lightly. 16 ... cxd4 looks correct here before White captures on c5.
17 Bxe7
The immediate 17 dxc5 was much better, when Black has to play with great
accuracy to hold the balance. He achieves this via 17 ... Rxc5 18 Nd7 (instead, 18 e4
Bxg5 19 Nd7 Bf6 20 Nxc5 bxc5 21 exd5 Bxd5 seems very good for Black because of
his potential play along the b-file) 18 ... Rxb5 19 Qxb5 a6 20 Bxe7 Qxe7 21 Qb3 Qxd7
22 e4 Qc7 23 exd5 Bxd5 with good compensation for the ‘sacrificed’ exchange.
17 ... Qxe7 18 dxc5 bxc5 19 e4 Nb6 20 f3 Rfd8 21 g3?!
Presumably White wanted to defend his h4-pawn, but this leaves his kingside very
weak. 21 Qa3 was preferable, when 21 ... Rxd1+?! 22 Rxd1 Qxh4? is bad because of
23 Qxa7.
 21 ... Qc7 22 Ng4?
Losing on the spot. White had to protect the knight with 22 Qc3.
22 ... c4! 0-1
White’s resignation could be described as “polite”, but he is quite lost here. For
example, after 23 Qb4 (or 23 Rxd8+ Rxd8 24 Qb4 a6 25 Ba4 h5 etc) 23 ... Rxd1+ 24
Rxd1 h5 25 Nh2 Qxg3, White’s kingside is falling apart.

Game 40
E.Ubilava-M.Gurevich
Spanish Team Championship 2007

1 d4 d5 2 c4 c6 3 Nf3 Nf6 4 Nc3 e6 5 Bg5 Be7 6 e3 Nbd7 7 Qc2 0-0 8 Rd1

Putting the rook opposite Black’s queen, which discourages any opening of the
centre with ... d5xc4 and ... c6-c5 or ... e6-e5. For 8 0-0-0 see the previous game.
8 ... b6
A logical response. With White’s rook having committed itself to d1 Black creates a
structure where it would be better placed on c1.
Gurevich experimented with 8 ... Kh8 9 a3 Ne8 in J.Werle-M.Gurevich, Dutch Team
Championship 2006, though I’m not sure he will repeat it. Here White should play 10
Bxe7 (10 Bf4 f5 11 Bd3 Nd6 set up a tough stonewall formation in the game, which
Werle found very frustrating) 10 ... Qxe7 11 Bd3, when 11 ... f5 would be distinctly
unpleasant for Black after 12 cxd5 cxd5 13 Rc1, since his dark square weaknesses and
the open c-file trump any sort of attacking hopes based on using the e4-square.
One other move worth considering is 8 ... a5. A.Shneider-P.Korzubov, Lvov 1985,
continued 9 Bd3 h6 10 Bh4 dxc4 11 Bxc4 Nd5 12 Bg3 and now, rather than 12 ... Nxc3
as played, 12 ... a4!? would have been interesting, when 13 Nxa4? is bad because of 13
... Nb4 and ... b7-b5.
9 Bd3
Another Gurevich game varied at this point with 9 Be2 Bb7 10 0-0 Rc8 11 Qa4 a6
12 Qb3, when 12 ... dxc4 13 Bxc4 Nd5 14 Bxe7 Qxe7 wasn’t bad in Z.Izoria-
M.Gurevich, Corsica (rapid) 2005, but 12 ... h6 13 Bh4 c5 14 cxd5 Nxd5 15 Bxe7 Qxe7
would have allowed Black to equalize quite comfortably.
9 ... Bb7 10 0-0 h6 11 Bh4
11 Bf4 is probably best met by 11 ... Rc8 as well. Instead, V.Tkachiev-A.Graf, FIDE
World Championship, Tripoli 2004, saw 11 ... dxc4 12 Bxc4 b5 13 Bd3 Qb6 and now
14 Ne5 (14 e4 c5 was okay for Black in the game) 14 ... Nxe5 15 Bxe5 Rac8 16 Bxf6!
Bxf6 17 Ne4 would give White a clear edge because of his control of c5.
11 ... Rc8 12 cxd5 cxd5 13 Qb3 Ne4 14 Bxe7 Qxe7 15 Rc1 Ndf6

Black has fully equalized, so the issue now is in how he might try for a win.
Gurevich provides one possible solution: he just keeps on playing.
 16 Rc2
Even in such a benign-looking position it’s important to be careful. For example, the
attempt to exchange the rooks with 16 Ne2 Rxc1 17 Rxc1?! is met by 17 ... Ng4 18 Bxe4
dxe4 19 Ne5 (19 Nd2?? Qh4 wins on the spot) 19 ... Nxe5 20 dxe5 Rd8 and White is
the one who will need to be careful.
16 ... Nxc3 17 bxc3 Rc7 18 a4 Rb8
18 ... Ne4!? was perhaps a better move, since 19 Bxe4?! dxe4 20 Ne5? Rfc8 sees
White suddenly in trouble in view of ideas like ... f7-f6, as well as ... Ba6 or ... Bd5,
taking control of the c4-square.
19 Ra1 Ne8
19 ... Ne4 was a possibility here too.
20 a5 Bc6 21 Ne5 Nd6 22 c4?!
A slip which leaves Black on the positive side of the game. 22 axb6 Rxb6 23 Qa3
was objectively better, with a rather equal position.
22 ... dxc4 23 Bxc4 Be4 24 Bd3 Rxc2 25 Bxc2 Bxc2 26 Qxc2 Rc8 27 Qb2 b5

Black has a passed pawn and his rook was the first to the c-file. White tries to
counter by bringing his knight to the c5-square.
28 Nd3 Qc7 29 Nc5 Qc6 30 a6 Qd5 31 h3 e5 32 Rb1?
After this move White’s position becomes critical. He should have played 32 Re1,
which seems to stay more or less afloat.
32 ... exd4 33 exd4 Nf5 34 Nb3 Nh4
The immediate 34 ... Rc6 looks even stronger, threatening both ... Rg6 and ... Rxa6.
35 f3 Rc6 36 Qf2 Nf5 37 g4 Nd6
At first it seems that 37 ... Rc3 is a powerful move, but after 38 gxf5 Rxf3 39 Qg2
Qe4 40 Rf1 Qe3+ 41 Kh2 Rxf1 42 Qxf1 Qxb3 43 Qg2, the queen endgame is far from
clear.
38 Nc5 Rc8 39 Re1 b4 40 Rb1 Rb8
The rook goes behind the passed pawn.
41 Rb3 Nc4 42 Rd3 Na5 43 Qh2 Re8
And now it switches to the open e-file, eyeing threats against the white king.
44 Qd2

44 ... Nc6?!
Passed pawns should be pushed. Here Black missed the chance for 44 ... b3!, when
45 Qxa5? loses at once to 45 ... b2, while 45 Qb2 leaves the white queen as a very poor
blockader. Indeed, 45 ... Rc8 already looks decisive, threatening ... Rxc5. Instead, the
game enters a manoeuvring phase as Gurevich probes for another way through.
45 Kg2 h5 46 Nb3 h4 47 Re3 Rf8 48 Qe1 g5 49 Rd3 Nd8 50 Nc5 Nc6 51 Ne4
Kg7 52 Qe3 Rb8 53 Nc5 Rd8 54 Nb3 Kf8 55 Qc1 Ne5 56 Re3 Ng6 57 Qc5+ Kg7?!
 Black should swap queens himself: 57 ... Qxc5! 58 Nxc5 Rxd4 leaves him a pawn
up for nothing, while 58 dxc5? Nf4+ 59 Kh2 Re8 forces the rooks off to reach an easily
winning knight endgame.
58 Qxd5
Not 58 Qxb4?! Rc8, when Black would have a very strong attack with his knight
coming to f4 and his rook getting in.

58 ... Nf4+?!
And here 58 ... Rxd5 was better. After the text White manages to activate his rook,
which gives him more drawing chances. To be fair, both players were probably in time
trouble at this stage.
59 Kf1 Nxd5 60 Re5 Kg6 61 Nc5 Nf4 62 Re7 Rxd4 63 Rxa7 Rd1+ 64 Kf2 Ra1
65 Rb7 Nxh3+ 66 Ke3?
The fatal mistake. After 66 Kg2 Nf4+ 67 Kh2, staying close to the passed h-pawn,
White could still put up strong resistance.
66 ... Nf4 67 a7 h3 68 Na6!?
Trying for a last trick: 68 ... Rxa6 69 Rb6+! Rxb6 70 a8Q and White can play on for
a while. Black is probably still winning after 70 ... Kg7, but he doesn’t need to allow
this.
68 ... Nd5+ 69 Kd4 Rxa6 70 Kxd5 Ra5+ 0-1
 The end.

Game 41
Z.Sturua-A.Kharitonov
USSR Young Masters, Tallinn 1981

1 d4 d5 2 c4 e6 3 Nf3 Nf6 4 Nc3 Be7 5 Bg5 0-0 6 e3 Nbd7 7 Rc1 c6 8 a3

Question: That looks odd. What’s the idea? Is White preparing to play b2-b4 - ?

Answer: Not really. He’s trying to find useful moves while waiting for Black to
play ... d5xc4 and in this way avoid losing time with his f1-bishop.

Question: Does this work?

Answer: The problem is that if Black waits as well, and with useful moves such as
... a7-a6 and ... Rf8-e8, White runs out of waiting moves that have much value. So Black
gets to play ... d5xc4 and ... b7-b5 even though White’s bishop goes to c4 in one
movement.
 White can also wait with 8 Qc2, but then 8 ... a6 9 a3 Re8 comes to the same thing,
while after 9 cxd5 exd5 the white rook does not belong on the c-file.
8 ... Re8 9 Qc2 a6 10 h3 h6
This h-pawn move can help Black save time if White lines up against h7 with Bd3.
Instead, 10 ... dxc4 11 Bxc4 c5 was played in K.Sasikiran-G.Kaidanov, Bled Olympiad
2002, but then 12 dxc5 Nxc5 13 Rd1 Ncd7 14 0-0 gave White an edge due to his lead in
development.
11 Bf4
Going back the other way with 11 Bh4 is also answered by 11 ... dxc4 12 Bxc4 b5.
V.Hort-Be.Toth, Biel 1982, continued 13 Ba2 c5 14 dxc5 Nxc5 (14 ... Bxc5 was
probably better, with full equality according to the engine) 15 0-0 Bb7 16 Rfd1 Qb6 17
Bxf6 Bxf6 18 b4 Bxf3 19 gxf3 Na4 20 Nxa4 bxa4 21 Bb1 g6 22 Qxa4 Qb5 and with the
help of the opposite-coloured bishops, Black held a draw against his illustrious
opponent.
11 ... dxc4 12 Bxc4

12 ... b5
Black must be careful to avoid 12 ... Nd5?, after which 13 Bxd5 exd5 14 Nxd5!
cxd5 15 Bc7 Bb4+ 16 axb4 would leave him a pawn down with a bad position.
13 Be2 Bb7
Probably Black’s safest move. The immediate 13 ... c5 risks opening the position up
while behind in development, though it looks okay too. For example, G.De Boer-
R.Cifuentes Parada, Wijk aan Zee 1993, went 14 dxc5 Bxc5 15 Ne4 Nxe4 16 Qxe4 Ra7
17 0-0 Bb7 18 Qc2 Qa8 19 Rfd1 Be4 20 Qc3 Be7 21 Bc7 and the players agreed a
draw.
14 0-0 c5 15 dxc5 Nxc5 16 Rfd1 Qb6 17 Ne5
The engine likes 17 b4 at first, but after 17 ... Nce4 18 Nxe4 Bxe4 19 Bd3 Qb7,
Black’s position is very comfortable.
17 ... Rac8 18 Bf3 Bxf3 19 Nxf3 Qb7 20 Be5 Ncd7 21 Bf4 Nb6

Black is doing quite well now. Although the position is still objectively equal there
are some potential weaknesses for him to exploit on the queenside.
22 Qe2 Nfd7 23 Bd6 Bf6 24 Bb4 Ne5 25 Nxe5 Bxe5 26 Nb1?!
After this White is clearly worse. 26 b3 would have held the balance.
26 ... Na4 27 b3
A further slip, though one can understand him underestimating Black’s reply. 27 Nc3
was correct, when 27 ... Nxc3 28 Bxc3 Bxc3 29 Rxc3 Rxc3 30 bxc3 Rc8 31 Rd3 offers
White serious drawing chances with just major pieces on the board.
27 ... Nb2 28 Rf1 Qd5 29 Nd2 Nd3 30 Rxc8 Rxc8 31 e4 Nf4 32 Qd1
After 32 exd5? Nxe2+ 33 Kh1 exd5 34 Re1 Rc1, White is left a pawn down.
32 ... Qd3
 33 Nf3?
A blunder which loses material. 33 Qf3 was the best chance for White, though still
very difficult with Black’s pieces crawling into his position.
33 ... Ne2+ 34 Kh1 Rc1 35 Qxc1
White can’t take the queen since 35 Qxd3? Rxf1+ 36 Ng1 Rxg1 is mate.
35 ... Nxc1 36 Rxc1 Bf4 37 Re1 Qxb3 38 e5 a5! 39 Bd6
39 Bxa5 Qxa3 creates a decisive passed pawn.
39 ... Qd3 40 Kg1 Kh7 41 Be7 g5 0-1

Game 42
G.Munkhgal-J.Ehlvest
Jakarta 2013

1 d4 Nf6 2 c4 e6 3 Nc3 d5 4 Nf3 Nbd7 5 Bg5 Be7 6 e3 0-0 7 Rc1 c6 8 Bd3 a6 9 0-0
The alternative 9 c5 will be examined in the next game.

Question: Can’t White exchange on d5 and show that 8 ... a6 is a waste of time?

Answer: 9 cxd5 exd5 actually gives Black a favourable version of the Exchange
Variation, since ... a7-a6 is useful against the minority attack (b2-b4-b5), whereas the
rook on c1 wants to be behind the b-pawn.
9 ... h6 10 Bh4 dxc4 11 Bxc4 b5 12 Bd3 c5

Showing the beauty of having played ... a7-a6 already. Black can proceed with ...
c6-c5 at once.
13 dxc5 Nxc5 14 Bb1 Qxd1
An interesting decision by Ehlvest; evidently he’s confident about outplaying his
lower rated opponent in the endgame. Otherwise 14 ... Bb7 gives Black a fully equal
game.
15 Rfxd1 g5 16 Bg3 Bb7 17 b4
Interestingly, this had all been played before in C.Nyberg-N.Gaprindashvili,
Lucerne Olympiad 1982. On that occasion White chose 17 Bd6 but found herself
struggling after 17 ... Bxd6 18 Rxd6 Rfc8 19 Rcd1 b4 20 Ne2 Bd5 21 Ne5 a5.
17 ... Nce4 18 Nxe4 Bxe4 19 Be5 Bxb1 20 Rxb1 Rfc8
 21 Bxf6
Snatching at simplification, perhaps in the hope that this might lead to a draw.
Objectively, 21 a3 would have been preferable, notwithstanding the fact that Black is
better after 21 ... Nd5.
21 ... Bxf6 22 Rdc1 Rxc1+ 23 Rxc1 Bb2 24 Rc7 Ba3!?
Simply going after the b-pawn and disregarding the idea of Ne5. Nevertheless,
playing 24 ... Rd8 first and then meeting 25 Kf1 with 25 ... Ba3 might have been more
precise.
25 Ne5 Bxb4
Here too Black could consider 25 ... Rd8, but Ehlvest wants to keep his rook behind
any passed pawn that emerges on the queenside.
26 Nxf7 Bf8 27 f4 gxf4 28 exf4 b4 29 Ne5 a5
 30 Rb7??
The wrong way of fighting against the queenside pawn majority; White needs to be
able to get his rook back to the first rank. He should play 30 Kf1 a4 31 Ke2 b3 32 axb3
a3 33 Rc1, when Black doesn’t have a good way of getting his a-pawn through. After
the move played White is in fact utterly lost.
30 ... a4! 31 Nd7
Threatening 32 Nf6+ Kh8 33 Rh7 mate, but this is easily dealt with. Similarly, 31
Ng4 can be met by 31 ... Bc5+ 32 Kf1 b3!, since the checks come to nothing; for
example, 33 Nf6+ Kf8 34 Nd7+ Ke8 35 Nf6+ (35 Nxc5 loses at once to 35 ... bxa2) 35
... Kd8 36 Rd7+ Kc8 37 Rd2 a3 will win soon enough.
31 ... Be7 32 Kf2
Or 32 Rc7 b3 33 axb3 a3 34 Rc1 a2 35 Ra1 Kg7!, followed by ... Bb4 and ... Bc3
etc.
32 ... b3! 33 a3
Black was threatening to capture on a2, while after 33 axb3 a3!, White would have
to give up his rook for the promoting pawn.
 33 ... Rc8
Simply 33 ... Bxa3 looks more straightforward. After the text White might play 34
Nb6, followed by Nxa4 and Rxb3, though 34 ... Rc7! 35 Rb8+ Kf7 36 Nxa4 Rc4 should
win easily enough, given that 37 Nb6 fails to 37 ... Rxf4+ 38 Ke3 Rb4!.
34 Kf3 Rc1
Note that 34 ... Bxa3?? would be dead wrong because of 35 Nf6+ Kf8 36 Nh7+ Ke8
37 Nf6+ with a draw by repetition, since 37 ... Kd8?? 38 Rd7 is mate.
35 g4 Bxa3 36 g5
Now 36 Nf6+ no longer helps because Black’s king can run across to c8 again.
36 ... hxg5 37 fxg5 Be7 38 Ne5 Bxg5 39 Kg4 Rg1+ 40 Kh3 Bf6 41 Nd7 Bg7 42
Nf6+ Bxf6 43 Rg7+!
A nice joke, hoping for stalemate should Black take with the king or rook.
43 ... Bxg7! 0-1

Game 43
Val.Popov-R.Ziatdinov
Voronezh 2004

1 d4 d5 2 c4 e6 3 Nc3 Nf6 4 Bg5 Be7 5 e3 0-0 6 Nf3 Nbd7 7 Rc1 a6

 Question: That’s different. I guess the idea is to follow up with ... d5xc4 and ... b7-
b5, but I’m surprised it is played here before ... c7-c6.

Answer: 7 ... a6 is known as the Swiss Variation and represents a playable option
against 7 Rc1. But as it is only good against this move, and not against White’s other 7th
move alternatives, I consider it confusing as a recommendation for Queen’s Gambit
Declined newbies. From a repertoire perspective the move order that is relevant to us is
7 ... c6 8 Bd3 a6 9 c5, reaching the position after Popov’s 9th move.
8 c5 c6 9 Bd3

9 ... e5!
The most logical and thematic reply to White’s c4-c5 plan. Having said that, 9 ... b6
is also playable; for example 10 cxb6 c5 (10 ... Qxb6 11 0-0 Qxb2 12 Na4 Qb7 13 Ne5
gave Black problems in V.Hort-L.Portisch, Madrid 1973) 11 b3 (preventing ... c5-c4, as
11 0-0 c4! 12 Bc2 Qxb6 is okay for Black) 11 ... cxd4 12 exd4 Qxb6 13 Na4 Bb4+ 14
Bd2 Qd6 15 Bxb4 Qxb4+ 16 Qd2 a5 17 Qxb4 axb4 18 0-0 gave White a favourable
endgame in G.Sargissian-Y.Seirawan, World Team Championship, Ningbo 2011, though
Black managed to hold the draw.
10 dxe5 Ne8 11 h4
 The critical line, which was thought to be good for White, but Ziatdinov’s play in
this game fully rehabilitates it for Black. Then again, White has not achieved much with
other moves either:
a) 11 Bf4 Nxc5 12 Bb1 Ne6 13 Qc2 g6 14 Bh6 N8g7 15 Rd1 Re8 16 Bxg7 Kxg7 17
0-0 Qb6 18 a3 Bd7 19 Rd2 Rad8 and Black had no problems in V.Tukmakov-
Y.Balashov, Dortmund 1987.
b) 11 Bxe7 Qxe7 12 Qc2 h6 13 0-0 Nxe5 14 Nxe5 Qxe5 15 Ne2 Nf6 16 Qc3 Qc7
17 Qd4 Bg4 18 Ng3 Rae8 19 f3 Bc8 20 Rfe1 Re5 21 Re2 Rfe8 was fine for Black in
E.El Gindy-Y.Seirawan, World Team Championship, Ningbo 2011 (played the day after
the game in the previous note).
11 ... Nxc5 12 Bb1 f6 13 Qc2 g6 14 Bh6 Bf5 15 Qe2 Bxb1 16 Rxb1
After 16 Bxf8 Nd3+ 17 Kd1 Nxc1 18 Kxc1 Bf5, Black has a very comfortable
position.
16 ... Ng7 17 h5

17 ... g5
This looks like best option, preventing the opening of the h-file. Alternatives are as
follows:
a) 17 ... gxh5?! is strongly met by 18 Bxg7 Kxg7 19 b4 Ne6 20 Nd4 Nxd4 21 Qxh5!.
b) 17 ... f5?! 18 hxg6 hxg6 19 Nd4 Qd7 20 f4 Rf7 21 g4 fxg4 22 Bxg7 Rxg7 23 b4
Ne6 24 Qxg4 was very good for White in the earlier game V.Epishin-R.Ziatdinov,
Philadelphia 1997.
c) 17 ... Qe8 18 hxg6 (18 Rh3 fxe5 was messy in G.Ageichenko-A.Gavrilov,
Moscow 1989) 18 ... hxg6 (18 ... Qxg6 19 Bxg7 Qxg7 20 Rh3 is unpleasant for Black)
19 b4 Nce6 20 Qc2 fxe5 21 Nxe5 is somewhat more comfortable for White.
18 0-0 fxe5 19 Nxe5 Qd6
Another set-up was 19 ... Bd6, looking to put the queen on e7.
20 Ng4 Rae8 21 b4 Nd7 22 e4 Bd8
Unveiling the rook on e8 against White’s queen and possibly preparing to bring the
bishop to c7.
23 Qb2
23 Rfd1 is well met by 23 ... Ne5, though this might have been White’s best option.
23 ... d4 24 Bxg7 Kxg7

25 Rfd1??
A fatal mistake, since Black can just sidestep the pin by the rook on the d-file.
Instead, 25 Qd2 was correct; or else 25 Ne2, when 25 ... Rxe4 26 f3 Rxg4 27 fxg4
Rxf1+ (or 27 ... Nf6) 28 Rxf1 Bf6 gives Black more than enough compensation for the
sacrificed exchange, but White would certainly be able to fight on.
25 ... Qe6
Now both his knights are en prise, so White is just losing.
 26 h6+ Kg6 27 Rxd4 Qxg4 28 Rbd1 Ne5 29 Rd6+ Bf6 30 Qc2 Rd8
Black is providing an object lesson in winning a won game by exchanging off his
opponent’s active pieces.
31 Na4 Rxd6 32 Rxd6 Rd8 33 Rxd8 Bxd8 34 Nc5 Nf3+ 35 Kf1 Qxg2+ 0-1
In view of 36 Kxg2 Ne1+ and 37 ... Nxc2.

Game 44
A.Yusupov-A.Goloshchapov
German League 2008

1 d4 d5 2 c4 e6 3 Nf3 Nf6 4 Nc3 Be7 5 Bg5 0-0 6 Rc1 c6 7 e3 Nbd7 8 Bd3 dxc4 9
Bxc4 b5

This is a good alternative to the main line 9 ... Nd5 (Games 46-55), especially if
Black is interested in playing for the full point.

Question: Doesn’t 9 ... b5 create a backward c-pawn on c6?

Answer: It does, but Black expects to achieve the ... c6-c5 advance in order to
solve this problem.
 10 Bd3
White usually returns the bishop to d3, where it lines up towards the black king. For
10 Bb3 Bb7 11 0-0 a6, see the note to White’s 11th move in the notes to Game 37.
10 ... Bb7 11 0-0 a6

Question: How about 11 ... b4, chasing that white knight away?

Answer: Unfortunately it gets chased to a better square; for example, after 12 Ne4
Nxe4 13 Bxe7 Qxe7 14 Bxe4, Black’s weaknesses on the c-file are becoming a real
problem. 12 Na4 also looks pretty good, again with pressure down the c-file.
Black does have a good alternative here in 11 ... Rc8, which we’ll see in the next
game.
12 Ne4 Nxe4 13 Bxe7 Qxe7 14 Bxe4
Here too Black’s c6-pawn is attacked, but this time it’s harder to exploit it. For one
thing Black is geared up for ... c6-c5 as soon as it doesn’t lose the bishop on b7.
14 ... Rac8

15 a4
This looks like it should discourage ... c6-c5 but doesn’t actually achieve its aim.
On the other hand, Black frees himself after other moves as well; for example:
 a) 15 Nd2 f5! 16 Bf3 e5 17 Nb3 e4 18 Bh5 c5 was already slightly better for Black
in P.Saering-M.Malachowski, German League 2008.
b) 15 Qc2 f5! 16 Bd3 c5 17 dxc5 Bxf3 18 gxf3 Rxc5 19 Qb1 Rf6 20 Kh1? Rh6 gave
a decisive attack in L.Jorgensen-Duc Tri Pham, World Junior Championships, Vung Tau
2008.
15 ... Nf6
Both threatening White’s bishop and protecting his own, while after any retreat (to
d3 or b1, say), Black frees his position with ... c6-c5.
16 Nd2 e5!
Obviously Black wants to avoid 16 ... Nxe4 17 Nxe4, since the white knight would
make itself at home on the c5-square.
17 dxe5 Qxe5

18 Bf3
Black would also be fully prepared for ... c6-c5 after 18 Qc2 Rfe8 19 Bf3 Re7.
18 ... Qxb2 19 Nb3 Ba8 ½-½
White probably accepted the draw offer with a degree of relief, as he may be a little
short of full compensation here.

Game 45
Im.Balogh-Vl.Georgiev
European Championship, Plovdiv 2008

1 d4 Nf6 2 c4 e6 3 Nf3 d5 4 Nc3 c6 5 Bg5 Be7 6 e3 Nbd7 7 Rc1 0-0 8 Bd3 dxc4 9
Bxc4 b5 10 Bd3 Bb7 11 0-0 Rc8

Question: Is that the most logical preparation for ... c6-c5 given that Black will
have to protect his b-pawn with ... a7-a6 - ?

Answer: Actually, it’s possible that Black will omit ... a7-a6 altogether and instead
play ... b5-b4, followed by ... c6-c5. In this sense 11 ... Rc8 can be considered a more
dynamic treatment than the 11 ... a6 featured in the previous game.
12 Qe2
A logical-looking move, vacating the d-file for the f1-rook and further supporting the
advance of White’s e-pawn. After the immediate 12 e4 Black can free himself with 12
... b4 13 Na4 c5, obtaining equal play.
Another possibility is the subtle 12 a3 which intends to meet 12 ... a6 with 13 b4,
preventing ... c6-c5. But then Black can get counterplay with 13 ... a5 14 Qb3 h6 15 Bh4
c5! 16 Bxf6 Bxf6 17 bxc5 Nxc5 18 dxc5 Qxd3 19 Qxb5 Ba6, which is a line I worked
out together with the engine and offers Black full equality.
12 ... b4 13 Bxf6 gxf6!
A very instructive recapture that most club players would avoid.

Question: So what’s the idea behind this horrid move?

Answer: Black needs both his bishop on e7 and knight on d7 to force through the
critical ... c6-c5 advance. At the same time he is not worried about his king since White
is unable to bring enough forces over to the attack.
14 Ne4

14 ... c5!
The key follow-up. In F.Deak-M.Ezsol, Hungarian Team Championship 2003, Black
went on a fishing expedition with his queen via 14 ... Qa5?!, but after 15 Nfd2 f5 16
Nb3 Qc7 17 Nec5, he found himself in a disgusting position. It’s difficult to overstate
the importance of getting ... c6-c5 in, and Black should therefore make it his priority.
15 dxc5 Bxe4?!
It looks safer just to recapture the pawn with 15 ... Nxc5 16 Nxc5 Rxc5 17 Rxc5
Bxc5, since 18 Qc2 Bxf3 19 Bxh7+ Kg7 20 gxf3 Bd6 21 Bd3 Rh8 22 f4 f5, intending ...
Qh4, is fine for Black
 16 Bxe4 Nxc5 17 Bb1 f5 18 Rc4 Na4!?
Black is trying to create chances by targeting the b2-pawn. At the same time he is
taking some risks by moving the knight away from its task of covering e4.

19 Rd4
The immediate 19 Rxc8 Qxc8 20 e4 is critical, opening up the b1-h7 diagonal
towards Black’s king. This might have led Georgiev to regret his 18th move.
19 ... Qb6 20 e4 Bf6 21 e5 Rfd8 22 Rf4
Both 22 exf6 and 22 Rxd8+ Bxd8 23 Bc2 came into consideration, when White
seems to have slightly the better of it in either case.
22 ... Bg7 23 h4?!
Going astray. White should have chosen 23 Rh4, after which he might yet open the
kingside with g2-g4.
22 ... Bh6 24 Ng5 Bxg5 25 hxg5 Rd4!
Now White’s hoped for attack isn’t working and he is left with weaknesses on g5,
e5 and b2.
26 Rf3 Qd8 27 Rg3 Nb6 28 Qh5 Qf8 29 f4
 29 ... Qg7
Black in turn doesn’t find the best line. He should have played 29 ... Nd5! 30 Rh3
Qg7 and if 31 Qh4 then 31 ... Nxf4! 32 Rxf4 Rc1+ 33 Kh2 Rxf4 34 Qxf4 Rxb1, winning
a pawn.
30 Rd3 Rcd8 31 Qf3 Qf8 32 Rb3??
A losing move. White is okay after 32 Rc1.
32 ... Nc4 33 Qh5 Nd2 34 Rh3 Qg7 35 Re1 Nxb1 36 Rxb1 Rxf4 37 Re1 Rg4
The g5-pawn is falling as well and with it White’s position.
38 Rh4 Rxg5 39 Qe2 a5 40 Rh3 Rd5 41 Re3 Rd4 42 Qf3 h6 43 R1e2 Kh7 44
Qb7 Rd1+ 45 Kf2 Rd5 46 Kg1 f4 47 Rf3 Rdxe5 48 Rxe5 Rxg2+ 0-1
Black wins easily after 49 Kf1 Rg1+ 50 Ke2 Qxe5+; for example, 51 Kd3 Qf5+ 52
Kd4 Rd1+ 53 Kc4 Rc1+ and so on.

Game 46
D.Janowski-J.R.Capablanca
New York 1924

1 d4 Nf6 2 Nf3 d5 3 c4 e6 4 Nc3 Be7 5 Bg5 0-0 6 e3 Nbd7 7 Rc1 c6 8 Bd3 dxc4 9
Bxc4 Nd5
 This freeing manoeuvre is generally attributed to Capablanca, following his 1927
World Championship match against Alekhine (see the notes to the next game), even if he
was by no means the first player to use this idea.
10 h4!?

Question: Wow, that looks dangerous for Black. Does this count as a pawn storm?

Answer: Not really, it’s more that White wants to get Black to make concessions in
order to free his position with exchanges. Of course he doesn’t want to capture on g5,
since h4xg5 will open the h-file for White’s rook on h1.
Note that similar positions to the game arise after 10 Bf4 Nxf4 11 exf4 Nb6 12 Bb3
Nd5, when A.Pokorny-N.Bergqvist, Munich Olympiad 1936, continued 13 Qd2 Bd6 14
g3 Qe8 15 0-0 f6 16 Ne4 Bc7 17 f5 Qh5 18 Bd1? (a tactical error, although the more
sensible 18 Nh4 is fine for Black after, say, 18 ... Rd8) 18 ... Qxf5 19 Nh4 Qxe4 20 Bf3
Bf4! 21 Qd1 Ne3 and Black won material.
10 ... f6 11 Bf4 Nxf4 12 exf4 Nb6 13 Bb3 Nd5
 Black will find it difficult to achieve either of his main pawn levers ( ... e6-e5 or ...
c6-c5) now, but he has the bishop pair and a well-placed knight on d5. Over the
following moves Capablanca gets counterplay against White’s weakened kingside by
bringing his queen to that region.
14 g3 Qe8! 15 Qd3 Qh5 16 Bd1 Bb4 17 0-0 Bxc3
This, together with the combination that follows, is a logical way to play the
position. Evidently Capablanca did not wish to contravene that logic in pursuit of the
full point.
18 bxc3 Nxf4!
 19 gxf4 Qg4+ 20 Kh1
Not 20 Kh2?? due to 20 ... Qxf4+, winning the rook on c1.
20 ... Qh3+ 21 Kg1 Qg4+ ½-½

Game 47
I.Ivanisevic-M.Gurevich
European Team Championship, Heraklio 2007

1 d4 d5 2 c4 c6 3 Nf3 Nf6 4 Nc3 e6 5 Bg5 Nbd7 6 e3 Be7 7 Rc1 0-0 8 Bd3 dxc4 9
Bxc4 Nd5 10 Bxe7 Qxe7 11 Ne4
 Question: That’s different; where’s the knight heading?

Answer: It’s a move that was popularized by Alexander Alekhine, the main idea
being to make it hard for Black to free his position with a trade of knights, followed by
... e6-e5.
11 ... N5f6
Given that Gurevich also likes to play this way against 11 0-0 (see the next game), it
probably wasn’t too traumatic for him to do it here too.
Note that 11 ... e5 can be answered by 12 dxe5 Nxe5 13 Nxe5 Qxe5 14 Bxd5 cxd5
15 Nc3 Rd8 16 Qd4 with a small but persistent edge for White, who has racked up the
points in practice.
12 Ng3
White naturally wants to avoid the knight exchange; and 12 Nxf6+ would be
answered by 12 ... Nxf6! 13 0-0 b6!, intending ... Bb7, ... Rfd8, ... Rac8 and ... c7-c5.
12 ... Qb4+
This classic treatment stems from the great Capablanca, who used it seven times in
his World Championship match with Alekhine and drew every game. Eighty years later,
it still proves to be a sound line in which White struggles to make anything out of his
slight plus.
 13 Qd2
Otherwise Black will capture the pawn on b2.
13 ... Qxd2+ 14 Kxd2 b6

15 Rhd1
Not a new move, but one that sends us back decades in order to find other examples.
A.Alekhine-J.R.Capablanca, Buenos Aires (16th matchgame) 1927, went 15 e4 Rd8 16
e5 Ne8 17 Ke3 Bb7 18 Rhd1 c5 (18 ... Kf8 was probably more accurate) 19 d5 exd5 20
Bxd5 Bxd5 21 Rxd5 Nc7 22 Rd2 Nf8 23 Rcd1 Rxd2 24 Rxd2 ½-½. Although White has
a space advantage, he is unlikely to convert it into anything tangible as it’s difficult to
locate a realistic target in Black’s ranks.
15 ... Bb7 16 Ke2 Rfd8 17 Rd2 Kf8 18 Rcd1 Ke7 19 e4 Nf8
An innovation by Gurevich which does little to change the overall assessment of this
line. The 18th game of the Alekhine-Capablanca match went 19 ... h6 20 h3 g6 21 Rd3
c5 22 dxc5 Nxc5 23 Rxd8 Rxd8 24 Rxd8 Kxd8 25 Ne5 Ke7 26 f3 Nfd7 27 Nxd7 Nxd7
28 Kd3 with another draw. 19 ... c5 also seems to be eminently playable.
20 Ke3 Ng6 21 h3 h6 22 a3 c5 23 dxc5 Rxd2 24 Nxd2 bxc5
 The isolated c-pawn is arguably a weakness, but Gurevich shows that it doesn’t
matter very much. Among other factors he now has the half-open b-file for his rook.
25 Nb3 Rc8 26 Na5 Rb8 27 Rd2 Ne5 28 Be2 g5
Preventing his well-placed knight on e5 being driven away with f2-f4. Black is at
least equal here.
29 Rc2 Nfd7 30 b4
30 f4 gxf4+ 31 Kxf4 Bc6 is fine for Black, who might make use of the half-open g-
file as well.
30 ... cxb4 31 axb4 Rc8
There was a case for 31 ... f5!? here, when 32 exf5 Bxg2 33 fxe6 Kxe6 leaves White
with the problem of how to protect his h-pawn. Lines like this are a good example of
how victory can be sought even in apparently level endgames.
32 Rxc8 Bxc8 33 Kd4 Kd6 34 b5 Nc5 35 Nf1 a6 36 Ne3 axb5 37 Bxb5 Ncd7 38
Be2 Ng6 39 Nec4+ Kc7 40 Na3 Nf4 41 Bf1 e5+ ½-½

Game 48
S.Brynell-M.Gurevich
Cappelle la Grande 2011
1 d4 d5 2 c4 c6
The standard move order is 2 ... e6 3 Nc3 Nf6 4 Bg5 Be7 5 Nf3 Nbd7 6 e3 0-0 7
Rc1 c6 8 Bd3 dxc4 9 Bxc4 Nd5 10 Bxe7 Qxe7 11 0-0.
3 Nf3 Nf6 4 Nc3 e6 5 Bg5 Be7 6 e3 Nbd7 7 Bd3
This makes sense after Black’s early ... c7-c6, not least because 7 Rc1 can be
answered by 7 ... Ne4. For example, V.Smyslov-U.Andersson, Tilburg 1984, went 8
Bxe7 Qxe7 9 Qc2 Nxc3 10 Qxc3 0-0 11 Be2 dxc4 12 Qxc4 e5 13 0-0 exd4 14 Nxd4
Rd8 15 Rfd1 Nf8 16 a3 g6 17 b4 Bd7 18 Rd2 Rac8 19 Rcd1 Ne6 20 h3 and the players
agreed a draw.
7 ... dxc4 8 Bxc4 Nd5 9 Bxe7 Qxe7 10 0-0 0-0

11 Rc1
This is the most common move here, reaching the main line for this chapter, a
position that has been known and studied for well over a century.
As we mentioned in the notes to Game 37, White is not forced to put the rook on c1.
Instead, 11 Ne4 b6!? (11 ... N5f6 12 Ng3 is better for White than the previous game
since Capablanca’s 12 ... Qb4+ is not available) 12 Qe2 Bb7 13 Rac1 c5 14 dxc6 Nxc5
15 Nxc5 bxc5 was P.Wells-M.Gurevich, Gibraltar 2006; while 11 e4 Nxc3 12 bxc3
b6!? (similarly, 12 ... e5 13 Re1 improves for White on the main line with Rc1) 13 a4
Bb7 14 a4 c5 15 Qe2 Nf6 was N.Dzagnidze-M.Gurevich, Chalkida 2009. Black seems
slightly worse in these games, but Gurevich drew them both without much difficulty; and
in any case it’s not a concern unless you wish to adopt Gurevich’s move order.
11 ... N5f6!?

Question: Can moving the knight again really be a good idea?

Answer: It’s a rare but interesting continuation. Rather than exchange knights on c3
before playing ... e6-e5 (which we’ll examine in Games 49-55), Gurevich simply
retreats it.
Black can also consider 11 ... Rd8, which is an old favourite of Ulf Andersson and
seems quite solid. For example, V.Borovikov-R.Cifuentes Parada, Pamplona 2003,
continued 12 Qc2 Nxc3 13 Qxc3 b6 14 Be2 (or 14 b4 Bb7 15 Be2 Rdc8 16 Qb2 h6 17
Rfd1 Rc7 18 Rc3 Nf6 19 Rdc1 Rac8 20 h3 Nd5 21 Rb3 Ba8 22 Ne5 c5 saw Black
equalize in A.Karpov-U.Andersson, USSR vs. Rest of the World, London 1984) 14 ...
Bb7 15 Rfd1 a5 16 a3 c5 17 dxc5 Nxc5 18 Rxd8+ Qxd8 19 Rd1 Qe7 with equality.

12 e4
Taking space in the centre and more or less forcing Black’s reply, since he can’t
allow White to push the pawn on to e5. White has tried a couple of other moves here:
a) 12 Qc2 e5 13 Bb3 (13 Rfd1 can be answered by 13 ... e4 and if 14 Ng5 then 14
... Nb6; while G.Stahlberg-G.Puiggros, Buenos Aires 1941, saw 14 Nd2 Re8 15 d5,
when 15 ... Ne5! would have been fine for Black) 13 ... exd4 (here 13 ... e4? 14 Ng5! is
just bad for Black) 14 exd4 Nb6 15 Rfe1 Qd8 16 Ne4 Bf5 17 Nxf6+ Qxf6 18 Qc5 was
A.Jedlicka-P.Bazant, Czech League 2003, and now 18 ... Rfe8 would have been best,
preventing Qe7.
b) 12 Qe2 e5 13 dxe5 Nxe5 14 Nxe5 Qxe5 15 h3 was a game L.B.Silva-A.Nery,
Juazeiro do Norte 2011, where 15 ... Bf5 would have been the simplest way to get a
fully equal game.
12 ... e5 13 Re1
At first some engines like 13 d5, but after 13 ... Nb6 14 d6 Qe8 15 Bb3 Bg4 they
soon change their ‘minds’. The d6-pawn will actually be very difficult to defend.
13 ... h6
Preventing White’s knight from coming to g5 is an important prelude to moving the
rook from f8. Instead, 13 ... exd4?! 14 Qxd4 left Black in an awkward situation in
M.Kapp-S.Buchwald, Überlingen 2000.
14 a3
Here too the engines like 14 d5, but as before 14 ... Nb6 15 d6 Qe8 16 Bb3 Bg4
shows Black as being very comfortable.
14 ... Rd8 15 Bf1 exd4 16 Qxd4 Nf8 17 Qe3 Ng4 18 Qe2 Ng6

Black’s agile knights are doing a good job of restraining White’s kingside pawn
majority. Having achieved full equality out of the opening I imagine Gurevich was
already thinking in terms of the full point.
 19 Rcd1 Be6 20 Rxd8+ Rxd8 21 Rd1 N4e5 22 Nxe5 Nxe5 23 Rxd8+ Qxd8 24
Qe3 Bc4 25 h3 Bxf1 26 Kxf1 Nc4 27 Qxa7
Dropping back with 27 Qe2 would be unpleasant after 27 ... Qd4.
27 ... Nxb2 28 Kg1

28 ... Qd3
Missing his chance. 28 ... Qd2! seems stronger, when 29 Qb8+ Kh7 30 Qe5 can be
met by 30 ... Qe1+ 31 Kh2 Qxf2, so White would have to try his luck in the line 29
Qxb7 Qe1+ 30 Kh2 Qxc3 31 Qc8+ Kh7 32 Qf5+ g6 33 Qxf7+ Qg7 34 Qb3, which isn’t
a straightforward win despite Black’s extra piece. Nevertheless, even after the move
played, Black is better.
29 Qb8+ Kh7 30 Qe5 Qd7 31 a4 Nd3 32 Qf5+ Qxf5 33 exf5
White does well to save this ending in view of his inferior pawn structure.
33 ... Kg8 34 Kf1 Kf8 35 Ke2 Nf4+ 36 Kf3 Nd5 37 Ne4
Not 37 Nxd5? cxd5 38 Kf4 f6 and Black has a winning endgame since White’s
damaged kingside can’t produce a passed pawn, effectively leaving him a pawn down.
37 ... b6 38 g3 Ke7 39 Ke2 Nf6 40 Nd2 Kd6 41 Nc4+ Kc5 42 Ne5 Kb4 43 Nxc6+
Kxa4 44 Ne5 b5 45 Nxf7
 Now White can give up his knight for the black b-pawn if needs be and draw on the
kingside.
45 ... b4 46 Ne5 b3 47 Nd3 Ka3 48 Ke3 Nd5+ 49 Kd2 Nb6 50 Kc1 Ka2 51 Nb4+
Ka1 52 Nc2+ Ka2
Essentially agreeing to split the point. After 52 ... bxc2?! 53 Kxc2, Black would
have to play precisely to hold the game as White’s king races towards the kingside
pawns.
53 Nb4+ Ka3 54 Nd3 Ka2 55 Nb4+ Ka1 56 Nc2+ Ka2 ½-½

Game 49
H.Spangenberg-P.Ricardi
Trelew 1995

1 Nf3 d5 2 d4 Nf6 3 c4 e6 4 Nc3 Be7 5 Bg5 Nbd7 6 e3 0-0 7 Rc1 c6 8 Bd3 dxc4 9
Bxc4 Nd5 10 Bxe7 Qxe7 11 0-0 Nxc3 12 Rxc3 e5
 After trading a pair of knights, Black seeks to free his game further with the thematic
thrust of his e-pawn. The key question is whether White can, in the meantime, make
something of his slight lead in development. We will examine this position in the
remaining games of this chapter.
13 dxe5
The double exchange on e5, followed by f2-f4, is often attributed to Akiba
Rubinstein, though it was actually played a decade earlier by a New Zealander, William
Mason. Obviously 13 Nxe5 Nxe5 14 dxe5 Qxe5 comes to the same thing.
White doesn’t get anywhere by playing 13 e4 at once; for example, 13 ... exd4 14
Qxd4 (14 Nxd4 can be met by 14 ... Ne5 or just 14 ... Qxe4!?) 14 ... b5!? (alternatively,
14 ... Nb6 15 Bb3 Be6 is a solid equalizer) 15 Be2 c5 16 Qd5 Rb8 17 Rd1 c4 18 Re3
Nf6 19 Qg5 h6 20 Qh4 Qb4 gave Black good play in H.Pillsbury-J.Showalter, New
York 1897.
13 ... Nxe5 14 Nxe5 Qxe5 15 f4
 15 ... Qe4
The stem game went 15 ... Qf5 16 Bd3 Qg4 17 Qc2 f5 18 Bc4+ Kh8 19 Rd3 Qh4 20
Rd6 Qe7 21 Qd2, which left Black in a difficult position in W.Mason-A.Davies,
Auckland 1914. Ricardi’s 15 ... Qe4 eventually became established as Black’s most
accurate move and then the entire line went out of fashion for White.
16 Qe2
Instead, 16 Qd3 Qxd3 17 Bxd3 Re8 18 Kf2 Be6 was very comfortable for Black in
N.Cusnariov-R.Stern, Berlin 2010, who in fact went on to win this endgame. Another
possibility is 16 Bb3, but then 16 ... Bf5 17 Qh5 g6 18 Qh4 Rad8 19 Bc2 Qd5 20 Rd1
Qa5 was at least equal for Black in P.Keres-I.Nei, Tallinn 1975.
16 ... Bf5
This move, on the other hand, appears to be an inaccuracy. The engine-approved
line is the less usual 16 ... Rd8; for example, 17 Bb3 Be6 18 Bc2 Qb4 19 a3 Qb6 and
now 20 f5 Qxb2 21 fxe6 Qxc3 22 Bxh7+ Kxh7 23 Qh5+ Kg8 24 Qxf7+ Kh7 was a draw
by perpetual check in G.Garcia Gonzales-J.J.Hernandez, Cienfuegos 1985.
17 Bd3 Qe6 18 e4 Rfe8 19 Re1 Qd6 20 Qf2 Be6
 21 e5
Not bad, but possibly not the best. The immediate 21 f5 appears to be good for
White after 21 ... Bd7 22 Bc2, since Black is struggling to get his pieces out without
dropping the a7-pawn.
21 ... Qb4 22 f5!?
Sacrificing the a2-pawn in order to go for broke on the kingside. In a subsequent
game White chose to keep his a-pawn with 22 a3, and after 22 ... Qb6 23 f5 Qxf2+ 24
Kxf2 Bd5 25 g4 a5 26 Kg3 a4 27 Kf4, he could claim to have somewhat the better of it
in G.Garcia Gonzales-B.Toth, Thessaloniki Olympiad 1984.
22 ... Bxa2 23 Re4 Qb6 24 Re3 Rad8 25 Ra3 Bb1!
 A neat tactic that Spangenberg might have missed.
26 f6?!
Not 26 Bxb1?? because of 26 ... Rd1+ etc. White should have tried 26 h3 Bxd3 27
Raxd3 Rxd3 28 Rxd3, when he still has some counterplay.
26 ... gxf6 27 Rg3+ Kf8 28 Qxb6 axb6
Now the bishops come off as well to leave Black with a very favourable rook
endgame.
29 exf6?!
Here 29 Bxb1 Rd1+ 30 Kf2 Rxb1 31 Rgb3 fxe5 32 Rxb6 offers more drawing
chances.
29 ... Re6 30 Rf3
Now if 30 Bxb1 Rd1+ 31 Kf2 Rxf6+ 32 Ke2 Rxb1 33 Ra8+ Ke7 34 Rb3 then 34 ...
b5 35 Ra7 Kd6 36 Rxb7 h5 is winning for Black.
30 ... Bxd3 31 Raxd3 Rxd3 32 Rxd3 Rxf6 33 Rd7 Rf5 0-1

Game 50
S.Lputian-Y.Balashov
Petrosian Memorial, Yerevan 1986
1 d4 Nf6 2 c4 e6 3 Nf3 d5 4 Nc3 Be7 5 Bg5 0-0 6 e3 Nbd7 7 Rc1 c6 8 Bd3 dxc4 9
Bxc4 Nd5 10 Bxe7 Qxe7 11 0-0 Nxc3 12 Rxc3 e5 13 Qc2

Question: What’s the idea behind this move?

Answer: This (and other moves) became popular when Black started to neutralize
13 dxe5 (see the previous game). The ongoing pressure against the e5-pawn means that,
in order to move his knight from d7 and develop the c8-bishop, Black will have either
to capture on d4 or push his pawn on to e4. So White is looking for a waiting/improving
move that will be good against both those options.
13 ... exd4
This is the first of Black’s two approaches. We will examine 13 ... e4 in the next
game.
14 exd4

Question: Why is White voluntarily taking on an isolated d-pawn?

Answer: By recapturing with the e-pawn White obtains excellent piece play down
the e-file and on the kingside, simultaneously allowing the c3-rook to slide along the
third rank. Whereas 14 Nxd4 gives White nothing after 14 ... Nf6 15 Rd1 Bd7, followed
by 16 ... Rfd8.
14 ... Nf6
Black’s position isn’t easy, and this move may in fact be a mistake. His best line
seems to be 14 ... Nb6 15 Re1 Qf6, though this is also very dangerous for Black.
V.Chekhov-D.Campora, Moscow 1989, continued 16 Bb3 Bf5 17 Qd2 Rad8 18 Ne5
Qd6? (18 ... Nd7 was relatively best, but still good for White after 19 Qf4) 19 Nxf7!
Rxf7 20 Bxf7+ Kxf7 21 Rf3 Kg6 22 h4 Qf6 23 Re5 Nc4 24 Qf4 with a winning attack.
15 Re1

15 ... Qd8
Instead, 15 ... Qd6 was played in the old game, A.Alekhine-C.H.Carlsson, Örebro
1935, with White winning in dramatic style after 16 Ng5 h6 (16 ... Bg4 17 Qb3 Bh5 18
Qxb7 Ng4 19 Nf3 Rfb8 20 Qe7 gives White a favourable endgame) 17 Nxf7 Rxf7 18
Qg6 Qf8 19 Rf3 Nd5 20 Rxf7 Qxf7 21 Re8+ Qf8 22 Rxf8+ Kxf8 23 Bxd5 cxd5 24
Qd6+ 1-0.
16 h3
Stopping anything landing on g4 and asking Black what he wants to do next. The
engine confirms that Balashov played the best move; even so his position remains
unpleasant.
16 ... Nd5 17 Bxd5 Qxd5
 17 ... cxd5 18 Qb3 is also difficult for Black, who is struggling to get his pieces
working without dropping a pawn.
18 Re5 Qd6 19 Ng5 g6 20 Rf3

20 ... f6?
Losing on the spot. Black should have gone for greed with 20 ... Qxd4, though 21
Qe2 (21 Re4 is met by 21 ... Qd8!) 21 ... Qd6 22 Ne4 Qxe5 23 Nf6+ Qxf6 24 Rxf6
leaves him fighting for a draw at best.
21 Qb3+ Kg7 22 Re8! a5
Shortening the struggle. Black gets mated after 22 ... Rxe8 23 Qf7+ Kh6 24 Qxh7+
Kxg5 25 h4+ Kg4 26 Qxg6+ Kxh4 27 g3+ Kh3 28 Qh5 mate, but he could have put up
more resistance with 22 ... b6.
23 Rxc8! 1-0
Boom! Balashov resigned in view of 23 ... Raxc8 24 Qxb7+ Rc7 25 Qxc7+ Qxc7 26
Ne6+ and 27 Nxc7.

Game 51
J.Timman-R.Hübner
Teesside 1975
1 d4 Nf6 2 c4 e6 3 Nc3 d5 4 Bg5 Be7 5 e3 0-0 6 Rc1 Nbd7 7 Nf3 c6 8 Bd3 dxc4 9
Bxc4 Nd5 10 Bxe7 Qxe7 11 0-0 Nxc3 12 Rxc3 e5 13 Qc2 e4

We examined the simplifying 13 ... exd4 in the previous game. Hübner’s move gains
space on the kingside and hopes to get an attack there.

Question: Which of these two approaches is better in general?

Answer: That’s difficult to say categorically, as a lot depends on the nuances of the
position. Certainly pushing on to e4 is the more ambitious option.
14 Nd2 Nf6 15 Rb1
Aiming to play b2-b4-b5 in order to open lines on the queenside, but this is certainly
not the only move or the only plan:
a) 15 f3 is a radically different idea, undermining the e4-pawn so as to gain a
potentially powerful pawn centre and open the f-file. After 15 ... exf3 16 Rxf3, Black
should probably challenge White’s dangerous bishop with 16 ... Be6, and if 17 Bd3 then
prevent the advance of his e-pawn with 17 ... Rfd8. In this case Black’s position looks
very comfortable.
b) 15 Bb3 was well met by 15 ... Bg4 in M.Najdorf-G.Stahlberg, Mar del Plata
1945, which continued 16 Re1 Kh8 17 a3 Rae8 18 Nf1 Be6 19 Bxe6 Qxe6 20 b4 Nd5
21 Rc5 a6 22 Ng3 f5 23 Qd2 Qh6 24 a4 f4! 25 exf4 Nxf4 26 Re3?! (26 Rd1 keeps the
balance) 26 ... Nd5 27 Re2 Qxd2 28 Rxd2 e3 and Black was better.
c) 15 a3 Bf5 16 Rc1 Rad8 17 b4 h5 18 f4 g6 19 Qb3 Rd7 20 b5? c5! also saw
Black take the initiative in W.Winter-M.Vidmar, Nottingham 1936.
15 ... Bd7
This is not Black’s most popular move but, as Hübner’s choice, it deserves our
attention. 15 ... a5 is another logical move, cutting across White’s plan of b2-b4. Then
C.Van den Berg-H.Pfleger, European Team Championship, Hamburg 1965, went 16 a3
Bd7 17 Rc1 (17 b4 axb4 18 axb4 Rfc8 19 b5 Nd5 is perfectly okay for Black) 17 ...
Rfe8 18 Ba2 Be6 19 Bxe6 Qxe6 20 Rc5 Nd7 and the players agreed a draw.
16 b4 Rfc8
Evidently looking to neutralize White’s queenside play.
17 Bb3 Nd5 18 Bxd5 cxd5

Black is not worse now. His space advantage fully compensates for having a ‘bad’
bishop, which is hardly restricted by the pawns in any case.
19 Rc1 g6
The b-pawn is taboo, since 19 ... Qxb4?? 20 Rxc8+ Rxc8 21 Qxc8+! Bxc8 22
Rxc8+ Qf8 23 Rxf8+ wins a piece.
20 Qb3 Qd6 21 Rxc8+ Rxc8 22 Rxc8+ Bxc8 23 Nb1 Be6 24 Nc3
 If White had expected this position to be favourable for him he will prove to be
mistaken. With his next move Black starts taking more space on the kingside, whereas
White will struggle to find an effective plan.
24 ... g5! 25 Ne2 Qa6 26 b5 Qa5 27 Nc3 Kg7 28 f3
Perhaps sensing the danger of staying passive White seeks to remove the black e-
pawn, but with the position opening up the bishop will become more effective.
28 ... exf3 29 gxf3 h5 30 Qb2
And not 30 Nxd5?? because of 30 ... Qe1+ 31 Kg2 Qe2+ 32 Kg1 (or 32 Kg3 h4
mate) 32 ... Bh3 followed by mate on g2, f1, or e1.
30 ... f6 31 Qd2 Qb4 32 e4 dxe4 33 fxe4 Kg6
After 33 ... Bxa2 White can either push the d-pawn or make things messy with 34 e5.
34 Qe3 Bc4
 35 e5!?
Rather than try and sit tight, White seeks counterplay by exposing Black’s king.
Otherwise 35 a4 was possible, simply defending the b5-pawn.
35 ... fxe5 36 Qe4+?
This is inconsistent with his previous move. 36 dxe5?! Be6 is just good for Black,
whose king is now quite secure, while White has a lot of weak pawns. So he should
have followed up with 36 h4!, undermining the remainder of Black’s pawn shield, when
White looks to have sufficient play to draw.
36 ... Kh6 37 Qf3 Qe7 38 Qe3
Here 38 d5 might have been a better try, but there’s no doubt about Black’s
advantage after 38 ... b6.
38 ... Qf6
Winning because of the threats against d4 and f1.
39 h4 exd4 40 hxg5+ Qxg5+ 41 Qxg5+ Kxg5 0-1

Game 52
V.Inkiov-Vl.Antonov
Pernik 1978
1 d4 Nf6 2 c4 e6 3 Nc3 d5 4 Nf3 c6 5 Bg5 Be7 6 e3 0-0 7 Rc1 Nbd7 8 Bd3 dxc4 9
Bxc4 Nd5 10 Bxe7 Qxe7 11 0-0 Nxc3 12 Rxc3 e5 13 Qb1

Question: That looks like an odd square for the queen. What’s the point?

Answer: As with 13 Qc2 in the previous two games, the queen switches to the b1-
h7 diagonal. Meanwhile there’s one distinct advantage to having the queen on b1 in that
she can support the advance of the b-pawn. This is particularly relevant if Black plays
13 ... e4.
13 ... exd4
Again we’ll look at the exchange on d4 first, which is the respectable attempt to
give White an isolated d-pawn and get developed. 13 ... e4 is featured in the next game.
14 exd4 Nb6 15 Re3
Putting a rook on the e-file is the logical way to proceed. A more difficult question
is which rook should be placed there, since 15 Re1 also makes sense. L.Pachman-
K.Ojanen, Trencianske Teplice 1949, went 15 ... Qf6 16 Bb3 Bf5 17 Qc1 Rad8 18 Re5
Bg4 19 Qe1 h6?! 20 h3 Bc8 21 Rce3 Qd6 22 Qa5 and Black’s position was critical.
15 ... Qd6
P.Ricardi-A.Acebal Muniz, Spanish Team Championship 2000, varied with 15 ...
Qd8 16 Bb3 Nd5, when 17 Bxd5 (17 Re5 f6 18 Bxd5+ cxd5 19 Re3 was equal in the
game) 17 ... Qxd5 18 Re5 Qd6 19 Rfe1 keeps some pressure.
16 Bb3

16 ... g6
Playing 16 ... Nd5 first looks preferable, when 17 Ng5 g6 18 Ne4 Qd8 19 Bxd5
cxd5 20 Nc3 should be slightly better for White in view of Black’s dark square
weaknesses, but it might not be possible to capitalize on this.
17 Qe4 a5 18 Ng5 Nd5 19 Bxd5 cxd5 20 Qe7 Ra6
And now 20 ... Qb4 seems better, provoking 21 a3 before exchanging queens with
21 ... Qxe7 22 Rxe7. Then Black can target the white b-pawn with 22 ... h6 23 Nf3 Ra6,
followed by 24 ... Rb6, which looks okay for him.
21 Rc1 Rc6 22 Rxc6 Qxc6 23 h3 Qd7 24 Qc5 Qd8 25 Re7 Bf5 26 Qc7 Qxc7 27
Rxc7
 By comparison with the previous note, it’s now White who has the active rook. This
makes quite a difference.
27 ... h6?!
Black is suffering after 27 ... f6 28 Nf3 as well, but at least the knight can’t hop into
e5 straight away, and there are real chances to save the game after 28 ... Rb8 29 Rc5 a4
30 Rxd5 Be6 31 Ra5 Bxa2 32 Rxa4 Bd5.
28 Nf3 Rb8 29 Ne5 Be6 30 Nd7 Bxd7 31 Rxd7 b6 32 Rxd5 Kf8 33 Rd7 Rc8 34
Rb7 Rc6 35 d5 Rd6 36 Kf1 Kg7 37 Ke2 Kf6 38 f4 Kf5?!
Here 38 ... Rxd5 39 Rxb6+ Ke7 was a better try, though White obviously has all the
chances.
39 Ke3 Rxd5 40 Rxb6 h5 41 a4
 41 ... h4?
This allows White to transpose into a winning pawn endgame. 41 ... Rd1 was the
last chance to fight on.
42 Rb5 Ke6 43 Rxd5 Kxd5 44 Kd3 Kc5 45 Kc3 f6 46 b3 f5
Or if 46 ... g5 then 47 f5, followed similarly by b3-b4 and wins.
47 b4+ axb4+ 48 Kb3 Kd5 49 a5 1-0
After 49 ... Kc5 50 a6 Kb6 51 Kxb4 Kxa6 52 Kc5, White’s king is the first to the
kingside.

Game 53
O.Garibaldi-H.Pilnik
Argentine Championship, Buenos Aires 1946

1 c4 e6 2 Nc3 d5 3 d4 Nf6 4 Bg5 Be7 5 e3 0-0 6 Nf3 Nbd7 7 Rc1 c6 8 Bd3 dxc4 9
Bxc4 Nd5 10 Bxe7 Qxe7 11 0-0 Nxc3 12 Rxc3 e5 13 Qb1 e4
 Now we’ll look at the second approach for Black. 13 ... exd4 was the previous
game.

Question: Which of these two moves do you think is better?

Answer: Objectively speaking, there may not be much difference between the two;
but I’d say that 13 ... e4 is the more challenging and pugnacious response.
14 Nd2 Nf6 15 b4
A case can be made for 15 Rfc1 in this position, which sidesteps 15 ... Be6 as White
would definitely take on e4. R.Sanguinetti-J.Bolbochan, Mar del Plata 1948, continued
15 ... Bd7 16 b4 Rad8 17 b5 Kh8 18 bxc6 Bxc6 19 Bb5 Bd5 20 Rc5 and White had an
edge; so Black should have held up the b-pawn’s advance with 16 ... a6, in order to
avoid being left with a weak a-pawn when White gets b4-b5 in.
15 ... Be6
The exchange of light-squared bishops is a key idea in Black’s defensive arsenal.
The main issue lies in whether White can now capture on e4 or not.
Of the alternatives only 15 ... a6 offers a real challenge to White’s b4-b5 plan.
N.Stull-Hel.Schumacher, Buenos Aires Olympiad 1978, continued 16 a4 Bd7 17 b5
axb5 18 axb5 Ra5 19 Rb3 Rfa8 20 bxc6 Bxc6 21 Bb5 Bd5 22 Bc4 Bc6 with a perfectly
sound position for Black. White declined the repetition with 23 Bb5 here and went on to
lose.
16 Rfc1
The critical line seems to be 16 Nxe4, after which M.Sjoberg-Fra.Andersson,
Swedish Team Championship 2000, continued 16 ... Nxe4 17 Qxe4 Qxb4 18 Qc2 Bxc4
19 Rxc4 Qe7 20 a4 Rad8 21 a5, threatening a5-a6. Black felt obliged to play 21 ... a6,
when 22 Rc5 would have prevented the idea of ... Rd5-b5. White clearly has an edge in
this position but it’s not easy to do much with it.
16 ... Nd5 17 Bxd5 Bxd5

This bishop proves to be a tower of strength on this square, holding the queenside
together and allowing Black to pursue his counterplay on the other flank.
18 a4 f5! 19 b5 f4! 20 bxc6 bxc6 21 Nc4 Rab8 22 Qc2
After 22 Rb3 Rxb3 23 Qxb3 f3, Black sets up threats against White’s kingside
similarly to the game.
22 ... f3 23 g3 Qe6
 Threatening mate with ... Qh3 and ... Qg2. There’s also the possibility of a rook lift
with ... Rf6 and ... Rh6, all of which should keep White busy defending his king!
24 Qd1 Rfc8 25 Qf1 Rb4 26 a5 Ra4
With White now tied down on the kingside, Black looks for a target elsewhere. This
is also a classic military strategy.
27 Ne5 Rxa5 28 Ra1?
This just makes matters worse. 28 Rc5 would have been better, though 28 ... Rxc5
29 Rxc5 Rb8 still leaves Black well on top.
28 ... Rxa1 29 Qxa1 Qh3 30 Qf1 Qxf1+ 31 Kxf1 Rb8
 The problems for White’s king persist into the endgame.
32 Rc1 Rb2 33 Ke1 a5 34 Ra1 h5 35 Ng6
If White does nothing Black can just charge his a-pawn down the board.
35 ... Re2+ 36 Kf1 Rc2 37 Nf4 Bb3 38 Nxh5
This loses quickly, but White’s position is hopeless in any case.
38 ... a4 39 Ke1 Re2+ 40 Kf1 Bc4 41 Kg1 a3 0-1
Nothing can be done about ... Rb2, ... a3-a2 and ... Rb1+.

Game 54
A.Karpov-D.Campora
Villarrobledo (rapid) 1997

1 d4 Nf6 2 c4 e6 3 Nf3 d5 4 Nc3 Be7 5 Bg5 0-0 6 e3 c6 7 Rc1 Nbd7 8 Bd3 dxc4 9
Bxc4 Nd5 10 Bxe7 Qxe7 11 0-0 Nxc3 12 Rxc3 e5 13 Bb3
 Question: What’s the idea behind this move? What does the bishop do on b3 apart
from block the advance of White’s b-pawn?

Answer: It’s actually a useful move whether Black pushes on with his e-pawn or
captures on d4. The bishop was in the way on c4; for example, a route has now been
opened for White’s rook to come to c5, and d4-d5 will fully open the c-file.
13 ... exd4
Again we’ll look at the capture on d4 first, with 13 ... e4 examined in the next game.
14 exd4 Nf6 15 Re1 Qd6 16 Ne5
 16 ... Bf5?!
This looks like an error in view of White’s possibility in the next note. Black should
probably play 16 ... Be6 to eliminate the dangerous white bishop. After 17 Bxe6 fxe6,
White has to be objectively better because his knight outpost on e5 is more advanced
than Black’s on d5; but it may be difficult for White to find targets in the enemy position,
especially if Black brings his a8-rook to e7, defending b7, g7 and e6.
17 Rf3
White had a strong continuation here in 17 Nxf7! Rxf7 18 Bxf7+ Kxf7 19 Qb3+ Kf8
20 Qxb7 Rb8 21 Qxa7 Rxb2 22 Rxc6 with a clear advantage, since 22 ... Rxa2? 23 Qb6
Qe7 fails to 24 Rxf6+! and 25 Qb8+, as in N.Doric-Ir.Pejic, Hum na Sutli 2010.
17 ... Bg4?
This meets with a powerful reply too. Having been given the chance, Black should
have secured f7 with 17 ... Bg6, when he has a reasonable game.
18 Rxf6! Qxe5
The only way to keep the game going, since recapturing on f6 just allows 19 Qxg4.
But now White forces a rook endgame in which he has an extra pawn.
19 dxe5 Bxd1 20 Bxf7+ Rxf7 21 Rxf7 Kxf7 22 Rxd1 Ke6
 Campora might have hoped that his active king would give him some compensation,
but Karpov’s precise play snuffs out Black’s chances within a few moves.
23 f4 a5
After 23 ... g5 there follows 24 Rd6+ Kf5 25 Rf6+ Ke4 26 fxg5 Kxe5 27 Rf7 etc.
24 Kf2 Ra6 25 Rd6+ Ke7 26 Ke3 Rb6 27 b3 a4
Here 27 ... Rb5 was perhaps a better try, though White can just answer with 28 Ke4,
further improving his king, and if 28 ... Rc6 then 29 Rd2.
28 Rd3 Ke6 29 g4 c5 30 h4 Rb4 31 h5
 Strategically speaking Karpov’s move is very attractive, even if the engine prefers
the immediate 31 bxa4.
31 ... h6
Obviously 31 ... axb3 32 Rxb3! wins easily, while 31 ... b5 allows 32 h6! gxh6 (or
32 ... g6 33 Rd6+ Ke7 34 f5) 33 bxa4 Rxa4 34 Rd6+ Kf7 35 Rxh6 and White’s kingside
pawns are much the stronger.
32 bxa4 Rxa4 33 Rd6+ Ke7 34 Rb6 Rxa2 35 Rxb7+ Ke6 36 Rb6+ Ke7 37 Rc6
Rg2 38 Kf3 Rc2 39 Ke4 c4 40 Rc7+ Kd8
After 40 ... Kf8 White would just play 41 Kf5, followed by either Ke6 or Kg6.
41 Rxg7 Rd2 42 Ra7 c3 43 Ra3 Re2+ 44 Kf5 Re3 45 Ra4 c2 46 Rc4 Re2 47 g5
hxg5 48 fxg5 Ke7 49 h6 Rh2 50 Rc7+ Kd8 51 Rc5 Ke7 52 Kg6 1-0

Game 55
A.Lahiri-S.Irwanto
Kuala Lumpur 2005

1 d4 Nf6 2 c4 e6 3 Nc3 d5 4 Bg5 Be7 5 e3 0-0 6 Nf3 Nbd7 7 Rc1 c6 8 Bd3 dxc4 9
Bxc4 Nd5 10 Bxe7 Qxe7 11 0-0 Nxc3 12 Rxc3 e5 13 Bb3 e4
 Here too this is the fighting move. 13 ... exd4 featured in the previous game.

Question: Has Black ever played anything other than pushing or capturing?

Answer: Occasionally, yes, though Black will need to push or capture anyway
within a few moves. An example is P.San Segundo Carrillo-D.Campora, Spanish Team
Championship 1999, which went 13 ... Rd8 14 Re1 g6 15 Qc1 e4 16 Nd2 Nf6 17 Qb1
Bf5 18 Nf1 Be6 19 Bxe6 Qxe6 20 b4 a6 with a solid position for Black, similar to our
main game.
14 Nd2 Nf6 15 Qc2
This position has been reached quite frequently, with White having tried several
other moves here, such as:
a) 15 Rc5 Bg4 16 Qc2 (16 Qb1 is similar) 16 ... Nd7 17 Rc3 Be2 (17 ... Rae8 18
Qxe4 Qxe4 19 Nxe4 Rxe4 20 f3 Bxf3 21 gxf3 was a bit better for White in
A.Kharitonov-H.Pfleger, Royan 1988) 18 Rc1 Bd3 19 Qd1 Rad8 20 f4 Nf6 21 Rc5 Nd7
(21 ... Nd5 would also have been fully equal) 22 Ra5 a6 23 Nf1 b6 24 Rh5 c5 gave
Black good play in N.Sulava-S.Brunello, Montecatini Terme 2006.
b) 15 Qb1 Bg4 16 Re1 Kh8 17 h3 Be6 18 Bxe6 Qxe6 19 Rec1 Rae8 20 Ra3 a6 21
Nb3 Nd5 22 Nc5 Qe7 23 Rb3 b6 24 Na4 b5 25 Nc3 f5 26 Nxd5 cxd5 27 Rc5 f4! 28
exf4 Rxf4 29 Rxd5 Qf6 30 Re5 ½-½ was J.R.Capablanca-G.Stahlberg, Buenos Aires
Olympiad 1939. Stahlberg played this line very effectively for Black and his treatment
deserves close study.
15 ... Bg4 16 Re1 Rad8

17 Nf1
New, but not particularly good because the knight no longer ties Black down by
putting pressure on e4. Instead, 17 Rc5 Rfe8 18 Rc1 was tried in Ab.Gupta-
T.Purushothaman, Commonwealth Championship, Nagpur 2008, and now rather than 18
... Nd7 19 Ra5, Black should have played 18 ... Qc7; for example, 19 d5 (or 19 h3 Be6)
19 ... Bd7 20 dxc6 Bxc6 21 Ba4 Ng4! 22 Nf1 Ne5 with a perfectly good position.
17 ... Be6!
As already noted, the exchange of light-squared bishops is a major achievement for
Black in this line.
18 Rc1 Bxb3 19 Qxb3 Nd5 20 Rc5 Rd6
 Black’s powerful pawn wedge on e4 gives him the chances here, for which this
game provides an excellent illustration.
21 Qc4 g6
The immediate 21 ... f5 was also good.
22 a3 h5 23 b4 a6 24 a4 b5
Preventing White’s b4-b5 at the cost of a weakness on c6, but Black reasons that he
can defend this successfully while building up on the kingside.
25 Qb3 Qe6 26 Nd2 f5 27 g3 Kg7 28 axb5 axb5 29 Ra1 Rf7 30 Ra6
30 h4 would act against Black’s next move, but then White’s kingside would be
rather fragile.
30 ... g5
 31 f4?
White must have underestimated the danger on the g-file. He should have played
something like 31 Qb1, though 31 ... h4 still looks very unpleasant.
31 ... gxf4 32 gxf4 Kh7 33 Kf2 Rg7 34 Nf1
Trying to run with 34 Ke1 might have been slightly better, but White’s position
remains desperate after either 34 ... Qf6 or 34 ... Rg1+ 35 Nf1 Qe7.
34 ... h4
The immediate 34 ... Rg2+! would have won.
35 h3
Giving Black a second bite of the cherry. 35 Qa3 would have allowed White to fight
on.
35 ... Rg2+! 0-1
White loses his queen after 36 Kxg2 Nxf4+ or 36 Ke1 Re2+ 37 Kxe2 Nxf4+ etc.
Chapter Four
Main Line with Bf4
Introduction

1 d4 d5 2 c4 e6 3 Nc3 Nf6 4 Nf3

This move is usually associated with the idea of meeting 4 ... Be7 with 5 Bf4. This
system is quite important and useful; it has been played at the very highest level and can
be adopted despite White having committed a knight to f3 early on.
4 ... Nbd7
This clever reply is specifically designed against the 5 Bf4 line. It is made possible
by a willingness to play ... Nbd7 anyway, such as in the Bg5 lines covered in Chapter
Three. Many players who defend the Queen’s Gambit Declined as Black don’t have this
option; for instance, if they favour the Tartakower Variation with 4 ... Be7 5 Bg5 h6 6
Bh4 0-0 7 e3 b6.
For those who want to play lines without ... Nbd7, or would like an alternative to
my main recommendation, I’ve included the option of 4 ... Be7 5 Bf4 0-0 6 e3 b6, after
which 7 cxd5 Nxd5 8 Nxd5 Qxd5 9 Bd3 Qa5+ is the critical line. White has two main
moves here, 10 Ke2 and 10 Kf1, featured in Games 56 and 57 respectively. White can
also play more quietly with 7 Bd3 as in Game 58. Black seems to be doing just fine in
all of these.
 5 cxd5
Clearly this can transpose into Chapter One should White follow up with Bg5. Here
we will focus on the development of his bishop to f4.

Question: Can’t White just play 5 Bf4 anyway, without capturing on d5 first?

Answer: He can indeed, but then the point of our move order is revealed with 5 ...
dxc4!?, when Black obtains excellent counterplay after either 6 e4 Bb4! (see Games 59
and 60) and the extraordinary 6 e3 b5!? (Games 61 and 62). Note that he can also meet
6 e3 with 6 ... a6, which is summarized in the notes to Game 61.
5 ... exd5 6 Bf4 c6 7 h3
After 7 e3, Black forces the exchange of White’s dark-squared bishop with 7 ... Nh5
and obtains a good game after both 8 Bg3 (Game 63) and 8 Bg5 Be7 (Game 64). The
text move anticipates Black’s ... Nh5 by preparing a retreat square for the bishop on h2.
7 ... Bb4!?
Counter-attack. Black’s ideas of ... Ne4 and/or ... Qa5 need to be taken seriously.
8 Qb3 Qe7 9 a3 Bxc3+ 10 Qxc3 Ne4
 We are following Zhu Chen-B.Spassky, Women vs. Veterans, Marbella 1999 (Game
65). Black already had a good game and went on to win handsomely.

Game 56
T.Radjabov-Ma.Carlsen
Bazna Kings, Medias 2011

1 Nf3 Nf6 2 c4 e6 3 Nc3 d5 4 d4 Be7 5 Bf4 0-0 6 e3 b6

Black’s main moves here are 6 ... c5 and 6 ... Nbd7, but the first of these has too
much theory and the second doesn’t look very good! Accordingly, I decided to
recommend the unusual text move, which has attracted the attention of several top
players. Magnus Carlsen’s patronage is certainly worth noting.
7 cxd5 Nxd5 8 Nxd5 Qxd5 9 Bd3 Qa5+
 10 Ke2
We’ll examine 10 Kf1 in the next game. Instead, 10 Nd2 Ba6 can get interesting after
11 Be4 (11 Bxa6 Nxa6 12 0-0 c5 13 Qe2 b5 14 Nb3 Qb6 15 dxc5 Nxc5 16 Nxc5 Bxc5
gave Black comfortable equality in Z.Gyimesi-U.Bönsch, Austrian League 2006) 11 ...
Nd7! (11 ... c6 12 a3 Qb5 13 b4 made it difficult for Black to free his position in
M.M.Ivanov-S.Halkias, German League 2006, in spite of the peaceful outcome) 12 a3
(12 Bxa8 Rxa8 13 Bxc7 Rc8 14 Rc1 Qb5 is very awkward for White) 12 ... Rac8 13
Bc6 Bb5 14 b4 Qa6 15 Bxb5 Qxb5 16 Qe2 Qa4 17 0-0 c5 18 dxc5 bxc5 19 Nf3 Bf6
and Black had rather more than equalized in P.Harikrishna-F.Döttling, Mainz (rapid)
2007.
10 ... Ba6 11 a3

Question: Why can’t White just take the c7-pawn? I thought this book was about a
solid defence to 1 d4, not some risky gambit!

Answer: Don’t worry, in this case Black has very clear compensation after 11 Bxc7
Bxd3+ 12 Qxd3 Na6; for example, 13 Rhc1 Rac8 14 Bg3 Nb4 15 Qb3 Qa6+ 16 Kd2?
Nc2 17 Rxc2 Rxc2+ 18 Qxc2 (or 18 Kxc2 Qe2+ 19 Nd2 Rc8+ etc) 18 ... Bb4+ 19 Kc1
Qf1+ 20 Qd1 Rc8+ leads to mate.
 11 ... c5!
This thematic pawn lever makes especial sense with White’s king in the centre.
12 b4 Bxd3+ 13 Qxd3 Qa4
Black can’t take the pawn, as after 13 ... cxb4 14 axb4 Qxb4? (14 ... Qd5 15 b5
leaves White with a clear positional advantage) 15 Rhb1, he loses his queen.

14 bxc5
On 14 Rhc1 Black should play 14 ... Nd7 (again 14 ... cxb4? 15 axb4 Qxb4? 16
Rcb1 traps the queen), when 15 dxc5 (or 15 ... bxc5 bxc5 16 Rc4 Qa5) 15 ... bxc5 16
Bd6 Bxd6 17 Qxd6 Rac8 started to look quite equal in B.Esen-U.Bönsch, European
Cup, Saint Vincent 2005.
Apparently Radjabov spent time calculating the aggressive 14 h4 but couldn’t get it
to work. A possible line is 14 ... Nd7 15 Ng5 Nf6 16 Be5 c4 17 Qb1 (17 Qxc4 Rac8 18
Qd3 Qc2+ 19 Qxc2 Rxc2+ 20 Kf3 Rfc8 gives Black compensation for the pawn due to
the strength of his rooks) 17 ... g6 18 h5 Nd5 19 hxg6 f6 (19 ... Nc3+? is bad because of
20 Kf1, when 20 ... Nxb1? allows 21 gxh7 mate) 20 Kf1 (or 20 gxh7+ Kh8 21 Qg6 fxe5
22 Nxe6 Bf6 and it’s White’s king that starts to look the more vulnerable) 20 ... fxe5 21
Nf7 Qb3 22 Rxh7 (22 Qxb3 cxb3 23 Rxh7 Rxf7 24 Rxf7 exd4, when 25 e4?? Rc8 26
exd5 b2 even wins for Black!) 22 ... Qxb1+ 23 Rxb1 Rxf7 24 Rxf7 exd4 25 exd4 Rc8
and Black emerges from the chaos with much the better endgame.
14 ... bxc5 15 Rhc1 Nd7 16 Qc4 Qa5
 Swapping queens leads to dead equality; e.g. 16 ... Qxc4+ 17 Rxc4 Nb6 18 Rc2
Rfc8. So Carlsen keeps them on in the hope that he can exploit the slight exposure of
White’s king.
17 dxc5
Here 17 Qc3 Qa6+ 18 Qc4 Qa5 would be a way to bring an immediate halt to
hostilities, unless Black wants to play on with 18 ... Qc6!?. Alternatively, 17 a4 would
be equal too, though Black can keep the game going with 17 ... Rfc8 18 Qb5 Qd8.
17 ... Nxc5 18 Nd4 Rac8

19 Qb5
Another possibility was 19 Kf1, after which 19 ... Rfd8 20 Qb5 would force the
exchange of queens, though the position is looking equal in either case.
19 ... Qd8
Still trying to keep his queen. Otherwise 19 ... Qxb5+ 20 Nxb5 and now 20 ... g5 21
Be5 f6 22 Bd6 Nb3 23 Rxc8 Rxc8 24 Rd1 Bxd6 25 Rxd6 Rc2+ would be a draw.
20 Nc6 Qd7 21 Nxe7+ Qxe7 22 Be5 Ne4 23 f3
Now 23 Qb4 is a bit awkward for White after 23 ... Qh4; for example, 24 Bg3 Qh5+
25 f3 (25 Kf1 Nxg3+ 26 hxg3?? Qh1+ wins a rook) 25 ... Nxg3+ 26 hxg3 Rb8. Even so,
White should be able to draw this position.
23 ... Qg5 24 fxe4 Rcd8
 One of several moves that will lead to a draw, such as 24 ... Qxg2+, 24 ... f6, and 24
... Rxc1 25 Rxc1 f6. Carlsen chooses a continuation which at least leaves White with a
possibility of going wrong.
25 Qb2
Now Black has nothing better than perpetual check.
25 ... Qxg2+ 26 Ke1 Qg1+ 27 Ke2 Qg2+ 28 Ke1 Qg1+
Not tempted by 28 ... Qxe4, as it’s White who has whatever chances are going after
29 Kf2.
29 Ke2 ½-½

Game 57
A.Graf-U.Bönsch
German League 2005

1 d4 d5 2 c4 e6 3 Nc3 Nf6 4 Nf3 Be7 5 Bf4 0-0 6 e3 b6 7 cxd5 Nxd5 8 Nxd5 Qxd5 9
Bd3 Qa5+ 10 Kf1

Instead of 10 Ke2 (or 10 Nd2) as in the previous game, this time White safeguards
his king, albeit at the cost of hampering his own development.
10 ... Ba6 11 Bxc7 Bxd3+ 12 Qxd3 Na6 13 Bg3 Nb4 14 Qb3 Rfc8 15 h3
 Making room for the king on h2. Sooner or later it must go somewhere in order to
allow White’s rook on h1 to get into play.
15 ... Rc2 16 Ne1
Black can meet 16 Kg1 with the paradoxical 16 ... Qd5!, after which 17 Qxd5 exd5
18 a3 Nd3 gives him more than enough for the sacrificed pawn despite the absence of
queens. This is a good example of the strength of a rook on the seventh rank.
16 ... Rc6
It’s not clear which move is best in this position. Another possibility is 16 ... Rd2,
engine analysis suggesting that the position is equal after 17 a3 (both 17 Qc3 Qb5+ 18
Kg1 Qe2 and 17 Kg1 Qd5 18 Qxd5 Nxd5 seem rather good for Black) 17 ... Rc8 (17 ...
Qd5?? is bad because of 18 Qc3, winning material) 18 Kg1 h5 19 Rd1 Nd5 20 Rxd2
Qxd2 21 Nf3 Qc2 22 Qb5 Qc4 23 Qxc4 Rxc4 24 Kh2 Rc2. Black has adequate
compensation for the pawn in the endgame, but no more than that.
17 Kg1

17 ... h5?!
In view of the possibility in the next note, Black should probably have preferred 17
... Rac8!?, after which 18 Nf3 Rc2 (the immediate 18 ... Qd5 also looks okay as a black
rook will reach the seventh rank in the endgame) 19 Kh2 Qd5 20 Rhd1 h5 21 Qxd5
Nxd5 leaves him with compensation for the pawn once more.
18 Nf3
 Here 18 a3 seems like a better move, when Black is struggling to find sufficient play
for the pawn. For example, after 18 ... Rac8 (now 18 ... Qd5 is well met by 19 Qd1) 19
Rd1 Qd5 (after 19 ... Nd5 White has 20 e4 Nf6 21 d5, or if 19 ... Nc2 then simply 20
Nf3) 20 Qxd5 Nxd5 21 h4 leaves White a pawn up for nothing special.
18 ... Qf5
After 18 ... Rc2 White can activate his central pawn majority with 19 e4.
19 Kh2 Rac8 20 a3 Nd5 21 Rhe1 Nf6 22 Rad1 Qd5 23 Qxd5
Avoiding the exchange of queens with 23 Qd3 is well met by 23 ... Ne4.
23 ... exd5 24 Ne5 Rc2 25 Nd3 Ne4 26 h4 g5 27 hxg5 Bxg5 28 f3

28 ... Nd2?!
Again not the best. Black should play 28 ... h4!, after which 29 Bf4 (both 29 Be5
Nf2 and 29 fxe4 hxg3+ 30 Kxg3 dxe4 are fine for Black) 29 ... Bxf4+ 30 Nxf4 Ng5 31
Re2 Rxe2 32 Nxe2 Rc2 33 Nf4 Rxb2 34 Nxd5 Nxf3+ 35 Kh3 Ng5+ 36 Kxh4 Rxg2
starts to look drawish.
29 Kh3 Nc4 30 Bf4 Bxf4 31 Nxf4
Black is in big trouble now and it looks as if he was only saved by the ticking of the
clock.
31 ... Nxb2 32 Ra1 Kf8
Defending the d-pawn with 32 ... Rd8 is bad because of 33 Ra2.
 33 Nxd5 Nd3 34 Reb1 Rd8

35 Nf6
Letting Black off the hook. White should try 35 e4 or 35 Rb5 (and if 35 ... a6 then 36
Nb4) with good winning chances in either case.
35 ... Re2 36 Nxh5 Rxe3 37 Ng3 Nf4+ 38 Kh2 Ne6 39 Nf5 Re2 40 Rb5 Nxd4 ½-
½

Game 58
V.Anand-V.Kramnik
London Classic 2016

1 d4 Nf6 2 c4 e6 3 Nf3 d5 4 Nc3 Be7 5 Bf4 0-0 6 e3 b6

 7 Bd3

Question: That’s different from 7 cxd5 as in the previous two games. Is it connected
with any particular idea apart from getting the bits out?

Answer: White is looking for a more opportune moment to capture on d5, perhaps
leading to Black having ‘hanging’ d- and c-pawns after ... c7-c5.
White’s other major alternative is 7 Rc1, after which Black should still play 7 ... c5.
For example, 8 cxd5 Nxd5 9 Nxd5 exd5 10 Bd3 (or 10 Be2 c4 11 0-0 Nc6 12 Ne5
Nxe5 13 dxe5 Be6 14 e4 b5 15 exd5 Qxd5 16 Qxd5 Bxd5 17 Rfd1 Rfd8 was nice for
Black in R.Kempinski-V.Babula, Czech League 2010) 10 ... c4 11 Bb1 Bb4+ (another
possibility is 11 ... Nc6, when 12 h4 f5 13 h5 Be6 14 Ne5 Nxe5 15 Bxe5 Bb4+ 16 Kf1
Bd6 was fine for Black in Ki.Georgiev-J.Nogueiras Santiago, Terrassa 1990) 12 Kf1
Bd6 13 Ne5 Bxe5!? 14 dxe5 Nc6 15 h4 Re8 16 Bg5 Qc7 17 f4 h6 18 Qc2 hxg5 19
Qh7+ Kf8 20 Qh8+ Ke7 21 Qxg7 Be6 was messy in B.Predojevic-A.Dreev, Croatian
Team Championship 2009.
7 ... c5
Black in turn goes straight for a hanging pawns position.
8 dxc5 bxc5 9 0-0 Nc6 10 cxd5
 After 10 Rc1 Black could play 10 ... h6, when 11 cxd5 exd5 transposes back into the
game, though he has other possibilities too, such as 10 ... d4.
10 ... exd5

11 Rc1
Another option is 11 e4 to break up Black’s pawn centre. V.Bhat-D.Gurevich, US
Championship, St Louis 2010, continued 11 ... dxe4 (11 ... d4!? is worth a further look,
notwithstanding the fact that it creates an outpost for White on c4) 12 Nxe4 Nb4 13
Nxf6+ Bxf6 14 Bc4 and now 14 ... Bf5 (rather than 14 ... Bb7 as played) 15 Be5 Re8 16
Qxd8 Raxd8 17 Bxf6 gxf6 would give Black full equality because of his piece activity.
11 ... h6
Some may see this as a questionable use of time. In C.Von Bardeleben-Em.Lasker,
Hastings 1895, Black played 11 ... Be6 immediately, which might be his most precise
move. After 12 Ne2 (or 12 Ng5 c4 13 Be2 Qa5 with counterplay) 12 ... Qb6 13 Ne5
Nxe5 14 Bxe5 Ng4 15 Bc3 Qd6, Black had an excellent game, even if he shortly went
wrong and ended up losing.
12 h3
After this non-critical reply Black gets a comfortable game. 12 e4 was more
challenging, when 12 ... Be6 (or 12 ... d4 13 Na4) 13 exd5 Nxd5 14 Nxd5 Qxd5 15 Qe2
gives White a positional advantage in view of the split black pawns, though this is not
guaranteed to translate into anything more concrete. It would be interesting to know
what Kramnik had in mind here.
12 ... Be6 13 Bb5 Qb6 14 Qa4 Rfc8 15 Ne5 Nxe5 16 Bxe5 a6 17 Be2 Rd8

The hanging pawns don’t look in any way vulnerable, so Black can claim full
equality.
18 Bf3 Nd7 19 Bg3 Nf6 20 Rfd1 Rac8 21 Be5 Nd7 22 Bg3
And not 22 Rxd5?, hoping for 22 ... Bxd5? 23 Nxd5 Qe6 24 Bc7 Re8? 25 Nf4 and
wins, since 22 ... Nxe5! 23 Rxe5 Qxb2 is just good for Black.
22 ... Nf6 23 Be5 Nd7 24 Bg3 Nf6 ½-½
One gets the impression that neither side was that unhappy with a draw in this game.

Game 59
Br.Kostic-Sr.Cvetkovic
Belgrade 2009

1 d4 Nf6 2 c4 e6 3 Nf3 d5 4 Nc3 Nbd7

In the last three games we saw a 4 ... Be7 move order, allowing 5 Bf4 etc.
Cvetkovic’s choice is designed to sidestep this variation altogether.
5 Bf4
White can forestall Black’s next move by exchanging on d5 first, when 5 cxd5 exd5
6 Bg5 Be7 7 e3 c6 8 Bd3 is the main line from Chapter One, while 6 Bf4 will be
covered separately in Games 63-65.
5 ... dxc4!?

This unexpected capture is the main idea behind Black’s early ... Nbd7.
6 e4
Seizing the centre is very natural, but it’s not so easy to maintain it. The more solid
alternative 6 e3 is covered in Games 61 and 62.
6 ... Bb4 7 Qc2
It’s difficult to suggest another move. Neither 7 e5 Nd5 8 Bg5 f6 9 exf6 gxf6, nor 7
Nd2 c5 is very appetizing for White.
7 ... b5 8 a4
Seeking to undermine the black pawn phalanx. 8 e5 is addressed in the next game.
Instead, 8 Nd2 reinforces e4 and breaks the pin on the c3-knight, but it does leave
one wondering how White should regain the ‘sacrificed’pawn. L.Bass-H.Barez
Menendez, Madrid 2011, went 8 ... c6 9 Be2 Nb6 10 Be3 0-0 11 0-0 a5 12 Rad1 ½-½,
though the result might have more to do with the rating difference than White’s non-
existent compensation.
8 ... Bb7
 Threatening the e4-pawn due to the pin on the knight on c3.
9 Nd2
9 e5 Bxf3! 10 gxf3 Nd5 11 Bd2 a6 12 Rg1 g6 13 Bg5 Qb8 was very good for Black
in P.Almagro Llamas-H.Barez Menendez, Madrid 2012.
9 ... a6 10 Be2
Grabbing the b5-pawn with 10 axb5 axb5 11 Rxa8 Qxa8 12 Nxb5 Qa1+ is very
good for Black; for example, 13 Ke2 (13 Qd1 Qxb2 14 Nxc7+ Ke7 is even worse) 13
... 0-0 14 f3 and White has a really horrible position.
10 ... c5 11 Bd6
 Trying to get compensation by trapping Black’s king in the centre.
11 ... cxd4
Not bad but not the best. Black can more or less refute White’s play with 11 ... Qb6!;
for example, 12 e5 cxd4 13 Bxb4 d3 14 Bxd3 cxd3 15 Qxd3 Nxe5 16 Qg3 Qd4 17 0-0
Nd3 leaves Black a pawn up and with a strong initiative.
12 Bxb4 Ne5 13 Nf3?!
White in turn plays an inferior move. He should try 13 axb5, after which 13 ... d3 14
Qa4 dxe2 15 bxa6+ Bc6 16 Qa5 is equal (!) according to the engine.
13 ... Nd3+
Missing a much stronger move in 13 ... Nc6, when 14 Bc5 d3 wins a piece for
inadequate compensation.
14 Bxd3 cxd3 15 Qxd3 dxc3 16 Qxc3 Rc8 17 Qe3
The engine gives 17 Bc5 and then recommends 17 ... Rg8 as Black’s best reply.
Don’t expect me to explain why.
17 ... Rc4 18 Rd1 Qc8 19 Qb3
Here 19 Rd4 was relatively best, although 19 ... Rc1+ 20 Rd1 Rxd1+ 21 Kxd1 Bxe4
leaves Black a pawn up for nothing.
19 ... Rxe4+ 20 Kf1 Bd5 21 Qc3
 21 ... Qxc3?!
Swapping the queens off is tempting, but it actually makes his task more difficult.
Instead, 21 ... Rc4, and if 22 Qa3 then 22 ... Qb7, seems pretty much winning for Black.
22 bxc3 bxa4 23 Nd2 Rf4 24 f3 a5!
Giving a pawn back in order to get his king safe and the h8-rook in play.
25 Bxa5 0-0 26 Ra1 Bb3
Here 26 ... Ra8 would have been stronger, when 27 Bc7 Rh4 28 Bg3 Rh5 sees
Black a good pawn up.
27 Kf2 Rf5 28 Bb4 Rd8 29 Nxb3 axb3 30 Rhd1
White is finally able to develop his own rook and can then target the b3-pawn. This
should be a draw now.
30 ... Rfd5 31 Rxd5 Nxd5 32 Rb1 Ra8
 Trying to eke out winning chances by having his rook on the seventh and bringing the
knight to f4.
33 Rxb3 Ra2+ 34 Kg1 Ra1+ 35 Kf2 Ra2+ 36 Kg1 g5 37 Bc5?
Giving Black serious winning chances again. After 37 c4 Nxb4 (or 37 ... Nf4 38 g3)
38 Rxb4 Rc2, Black is at best on the positive side of a draw. The remainder of the game
looks as if it was played out in mutual time trouble, so we won’t dwell on it too much.
37 ... Nf4 38 Bd4 Rxg2+ 39 Kf1 h6 40 h4 Rg3?!
Now pushing the c-pawn with 41 c4 gxh4 42 c5 looks to offer White sufficient
counterplay. So 40 ... Rc2 might have been better, getting behind the passed pawn at
once.
41 Kf2?! gxh4 42 c4 Rg2+ 43 Ke3 Ng6 44 Rb8+ Kh7 45 Rb2 Rxb2 46 Bxb2
 46 ... h3?
There is no need to give up the front h-pawn, which might yet be useful as a decoy.
After the text, the game meanders its way to a draw.
47 Kf2 Ne7 48 Kg3 Kg6 49 Kxh3 Kf5 50 Bc1 h5 51 c5 e5 52 Kh4 Kg6 53 Bb2 f6
54 Bc3 Nf5+ 55 Kh3 Kf7 56 c6 Ke7 57 Bb4+ Kd8 58 Ba5+ Kc8 59 Bb6 Nd4 60 f4
Nxc6 61 fxe5 fxe5 ½-½
The database has the further moves 62 Kg2? e4? 63 Kf2? h4? 64 Kg2 e3 65 Kf3 h3
66 Bxe3, which look ridiculous even with seconds on the clock. So we’ll stop here and
just give the result.

Game 60
K.Robatsch-Sr.Cvetkovic
Trnava 1988

1 Nf3 d5 2 d4 Nf6 3 c4 e6 4 Nc3 Nbd7 5 Bf4 dxc4 6 e4 Bb4 7 Qc2 b5 8 e5

We examined 8 a4 in the previous game.
8 ... Nd5 9 Bg5
 9 ... Be7

Question: Doesn’t that give the b5-pawn back now that the c3-knight is no longer
pinned?

Answer: It does indeed, and that was well spotted! Cvetkovic is looking for a solid
way to play the position which keeps his kingside pawns intact as a safe spot for his
king.
A more critical line is 9 ... f6, when 10 exf6 N7xf6 (10 ... gxf6 11 Bd2 Bxc3 12
bxc3 N7b6 is also worth considering) 11 a4 Bb7 12 Be2 (12 axb5 is met by 12 ... Nxc3
13 bxc3 Nxf3 14 cxb4 Qxd4) 12 ... a6 13 0-0 c6 14 Ne5 was fine for Black but quite
sharp in L.Van Wely-I.Sokolov, Tilburg 1994.
10 Bxe7 Qxe7 11 Nxb5 Qb4+ 12 Nc3 Rb8 13 Nd2
After 13 0-0-0 Black can defend the c4-pawn with 13 ... Ba6.
13 ... Qxb2
Here too 13 ... Ba6 looks like a decent alternative.
14 Qxb2 Rxb2
 15 Nd1
A somewhat retrograde approach. There was a more interesting way to play it in 15
Nxc4 Rb4 16 a3 Rxc4 (tempting, as it wins two pieces for a rook; but White gets
serious compensation so Black might be better off playing 16 ... Rb3 with something
close to equality) 17 Bxc4 Nxc3 18 Rc1 Na4 19 Bb3 Nab6 20 Rxc7 a5 21 Kd2 0-0 22
Ra7 a4 23 Ba2 and Black is tied up.
15 ... Rb8 16 Bxc4 c5 17 dxc5 Nxe5 18 Bxd5 Nd3+
A good idea as Black keeps his pawns more unified. Otherwise 18 ... exd5 was fine
too; for example, 19 0-0 0-0 20 Ne3 d4 21 Nec4 Nd3 22 Nb3 Be6 23 Nca5 Bxb3 24
Nxb3 Rfc8 25 Rfd1 Nxc5 starts to look drawish.
19 Ke2
After 19 Kf1?! exd5 20 Nb3 0-0, White has a clearly inferior version of the
previous note.
19 ... Nf4+ 20 Kf3 Nxd5 21 Nc4 Ba6 22 Nd6+
Instead, 22 Nde3 would have been a simpler way to maintain equality. White’s
position is still playable for a while but its viability will soon depend on some intricate
pyrotechnics.
22 ... Ke7 23 Ne3 Nxe3 24 Kxe3
The horrid-looking 24 fxe3 would have given White’s king a little more security and
allow him to meet 24 ... Rb2 with 25 Rhc1.
 24 ... Rb2

25 Rhb1?
This is the official turning point. White should have tried 25 Rhc1, though he stands
somewhat worse anyway; for example, 25 ... Re2+ 26 Kf3 Rb8 27 Nc4 (I tried making
27 Rc3? Rbb2 28 Ra3 work, but 28 ... Rxf2+ 29 Kg3 Rxg2+ 30 Kf3 Rbf2+! 31 Ke3
Re2+ 32 Kf3 Rgf2+ 33 Kg3 f5! 34 Rxa6 g5 sees the white king in serious trouble) 27 ...
Bxc4 28 Rxc4 Rbb2 29 Rf1 a5! (after 29 ... Rxa2 30 c6! Rec2 31 Rxc2 Rxc2 32 Ra1
Rxc6 33 Rxa7+ Kf6 34 h4!, White should be able to draw) 30 a4 (or 30 c6 Rec2) 30 ...
Rec2 31 Rxc2 Rxc2 32 Rb1 Rxc5 33 Rb7+ Kf6 34 Ra7 g5! and White has to defend a
tricky rook endgame a pawn down.
25 ... Rhb8 26 Rxb2 Rxb2 27 a4
Now if 27 Rc1 then 27 ... Re2+ 28 Kf3 Rxa2 leaves Black a good pawn up.
27 ... Re2+ 28 Kf3 Rc2 29 Nb5 Bb7+ 30 Ke3 a6 31 Nd4 Rc3+ 32 Kd2 Rxc5
The beginning of the end; not only is White a pawn down, Black has much the
superior minor piece.
33 g3 Bd5 34 a5 g5 35 Nc2 Bc6 36 Nb4 Bb5 37 Na2 Rd5+
 38 Ke3?
This should lose quickly. 38 Kc1 would have allowed him to hobble on.
38 ... Rf5?!
Missing the chance to finish the game immediately with 38 ... Rd3+! 39 Ke2 (or 39
Ke4 f5+ 40 Ke5 Rd5 mate) 39 ... Ra3+, followed by 40 ... Bc4, winning a piece.
39 Kd4?
Dropping another pawn. 39 Nc3 was better.
39 ... Rxf2 40 Nc3 Bd7 41 Kc5 Rc2 42 Ra3 Rb2 0-1

Game 61
Liang Chong-Wang Chen
Chinese Team Championship 2010

1 d4 d5 2 c4 e6 3 Nf3 Nf6 4 Nc3 Nbd7 5 Bf4 dxc4 6 e3

 Rather than seize the centre straight away with 6 e4, as in the previous two games,
this time White seeks first of all simply to regain his pawn.
6 ... b5!?

Question: Black has just blundered a pawn away on b5, right?

Answer: It will no doubt cause some controversy that I’m giving 6 ... b5!? as my
recommended treatment for Black, but this seems to be the most active and interesting
move. From among the numerous alternatives here are the two which gain most high-
level support:
a) 6 ... Nb6 7 Bxc4 Nxc4 8 Qa4+ Bd7 9 Qxc4 Bd6 10 Bg5 h6 11 Bh4 Be7 (after 11
... Bc6 12 Ne5 Bxe5 13 dxe5 g5 14 exf6 gxh4 15 Qxh4 Qd6 16 Rd1 Qe5 17 e4, Black
was struggling in S.Mamedyarov-E.Miroshnichenko, Baku 2006) 12 0-0 0-0 13 Rfd1 a5
14 Rac1 c6 15 e4 Re8 16 e5 Nd5 17 Bxe7 Rxe7 18 Ne4 Be8 19 Re1 left Black with a
solid but miserable position in I.Naumkin-V.Epishin, Di Roseto 2010.
b) 6 ... a6 7 a4 (7 Bxc4 b5 8 Bb3 Bb7 9 0-0 Bd6 10 Ne5 0-0 11 Rc1 Qe7 12 Qe2 c5
equalized in R.Kempinski-A.Baburin, European Team Championship, Gothenburg 2005)
7 ... Bd6 8 Bg3 b6 9 Bxc4 Bb7 10 0-0 0-0 11 Qe2 Bxg3 12 hxg3 c5 13 Rfd1 Qe7 gave
Black a comfortable game in A.Wirig-B.Grachev, Biel 2011.
7 Nxb5 Nd5 8 Bxc4
 White can maintain his pawn structure by retreating with 8 Bg3, but then 8 ... Bb4+ 9
Nd2 N7b6 gives Black sufficient play; for example, 10 Qc2 (10 e4 is well met by 10 ...
a6!, while after 10 a3 Ba5 11 Rc1 0-0 12 b4 cxb3 13 Qxb3 Bb7 14 Bd3, as in
M.Valles-B.Heberla, Porto 2015, then 14 ... f5! looks better for Black) 10 ... 0-0 11 Be2
Ba6 12 Nc3 c5! 13 dxc5 Bxc5 14 a3 Rc8 15 Nde4 Be7 16 Be5 f5 17 Nd2 Bf6 18 Bxf6
Qxf6 and Black was fine in L.Lenic-R.Stern, German League 2014.
8 ... Bb4+ 9 Kf1

Question: Can’t White keep the right to castle with 9 Nd2 - ?

Answer: He can, but then 9 ... Nxf4 10 exf4 Nb6 11 Bd3 c6 12 Nc3 Qxd4 13 Qc2
Qxf4 would leave him a pawn down and fighting for his life.
9 ... Nxf4 10 exf4 0-0

11 a3
White has tried several moves here, but Black seems to have decent compensation
for the pawn whatever White does. He just needs to play calmly for ... c7-c5, opening
the position for his bishop pair while White is sorting his bad king position out. Here
are some sample lines:
a) 11 Bd3 is covered in the next game.
 b) 11 Qb3 Be7 12 Rd1 Rb8 13 Qc2 Ba6 14 a4 c6 15 Na3 Qb6 starts to look quite
attractive for Black.
c) 11 g3 Bb7 12 Kg2 a6 13 Nc3 c5! 14 a3 (or 14 dxc5 Bxc3 15 bxc3 Qc8 16 Be2
e5!) 14 ... Bxc3 15 bxc3 Qa5 16 Qd3 Rac8 and Black had good play in M.Kopisch-
R.Stern, Berlin 2015.
d) 11 Rc1 Bb7 12 Bd3 (or 12 Be2 Ba5!?, when 13 Kg1 Bb6 14 h3 Nf6 15 Kh2 Rb8
16 Qd2 a6 17 Nc3 Bxf3 18 Bxf3 Qxd4 is okay) 12 ... Ba5!? 13 Kg1 Nf6 14 h4 Bb6 15
Rh3 a6 16 Nc3 Qd6 17 Ne2 c5 and once again Black has sufficient compensation.
11 ... Be7
Here too Black might bring the bishop round to b6 with 11 ... Ba5; for example, 12
b4 Bb6 13 d5 exd5 14 Qxd5 Qf6 15 Nbd4 Rb8 16 Rd1 Bb7 17 Qg5 Bxf3, intending 18
Nxf3? Qb2 and Black even gets the better of it. It’s interesting to see how engines
misassess these variations because they count White’s extra pawn but don’t see any
immediate problem with his uncastled king.
12 Qc2 Bb7 13 Bd3

13 ... Nf6

Question: Can’t Black destroy White’s pawn structure with 13 ... Bxf3 - ?

Answer: He does this next move, but he could indeed have played that at once,
since after 14 Bxh7+? (14 gxf3 Nf6 15 Qxc7 a6 16 Qxd8 Rfxd8 leads back to the game)
14 ... Kh8 15 gxf3 a6 16 Nc3 f5 17 Bg6 Rf6 18 Rg1 Nf8 19 Bh5 Rh6 (even better than
19 ... g6) 20 Bf7 Qxd4, followed by 21 ... Qf6, Black wins a piece for inadequate
compensation.
14 Qxc7 Bxf3 15 Qxd8 Rfxd8 16 gxf3 a6 17 Nc7 Rac8 18 Nxa6 Rxd4 19 Rd1
Ra8
There was a case for just preventing White’s next move with 19 ... g6 and inviting
him to come up with something else. The engine confirms that Black has more than
enough compensation after, say, 20 Ke2 Nd5.
20 Bxh7+
Probably best. After 20 Nc7? Rad8 21 Ke2 Nh5, White’s position falls apart.
20 ... Kxh7 21 Rxd4 Rxa6 22 f5 Nd5

Black is giving priority to activity over material and he’s probably quite right.
23 fxe6 fxe6 24 Kg2 Bf6 25 Rd2 Rb6 26 Rb1 Rb3 27 a4
Here 27 Re1 looks better, after which 27 ... Rxb2 28 Rxb2 Bxb2 29 a4 Bc3 30 Re4
Kg8 31 f4 Kf7 would likely be a draw.
The rest doesn’t have great implications for the Queen’s Gambit Declined, but it is a
good example of how to seek out winning chances in an objectively drawn position. It’s
a question of continuing to fight against an opponent who just wants it over with.
 27 ... Nb6 28 Re2 Nxa4 29 Rxe6 Nc3 30 Rc1 Nd5 31 Kg3 Rxb2 32 Rd1 Ne7 33
Rd7 Nf5+ 34 Kg4 Kg6 35 Rd5 Ne7 36 Rd7 Kf7 37 Re4 Rxf2 38 Kg3 Rf1 39 Rb7
Rg1+ 40 Kf2 Ra1 41 Kg3 Ra8 42 h4 Kf8 43 Kg4 Nc6 44 f4 Ra1 45 Re3 Ra4 46 Rc7
Ne7

47 Rb3?
This drops the f-pawn and leaves White teetering on the brink, since the h-pawn
will fall behind it. He should have defended f4 with 47 Rf3, when 47 ... Ng6 48 Rf1
(avoiding the knight fork) 48 ... Nxh4 49 Rb1 Kg8 50 Rcb7 and 51 R7b4 should hold.
47 ... Be5 48 Rcb7 Rxf4+ 49 Kh3 Nf5 50 R3b4 Bd4 51 Rd7 Rxh4+ 52 Kg2
Now it’s ‘just’ a case of reorganizing his forces and pushing the g-pawn.
52 ... Kg8 53 Kf3 Kh7 54 Rd5 Kg6 55 Ra4 Kg5 56 Ke2 Rh2+ 57 Kf3 Rf2+ 58
Ke4 Bf6 59 Kd3 Kg6 60 Raa5 Ne7 61 Rd6 Kf7 62 Ra7 g5 63 Ke3 Rf5 64 Ke4 Rf1
65 Rd3 Rf4+ 66 Ke3 Kg6 67 Ke2 Nc6 68 Ra8 Nd4+ 69 Ke1 g4 70 Rg8+ Kf5 71 Rf8
Ke5 0-1

Game 62
R.Hundack-R.Stern
Bad Zwischenahn 2012
1 d4 Nf6 2 c4 e6 3 Nf3 d5 4 Nc3 Nbd7 5 Bf4 dxc4 6 e3 b5 7 Nxb5 Nd5 8 Bxc4 Bb4+
9 Kf1 Nxf4 10 exf4 0-0 11 Bd3
Threatening a double attack with 12 Qc2, which turns out to be a blank shot. We
looked at other moves in the previous game.
11 ... Nf6

12 Qc2
Attacking c7 a second time. Another way of doing this is with 12 Rc1, but Black has
no reason to fear that either since 12 ... a6 13 Nxc7? Ra7 just puts White’s knight in
serious trouble; e.g. 14 Qa4 Bd6 15 Nxa6 (or 15 Nb5 Bd7) 15 ... Bb7 and 16 ... Qa8
wins. Otherwise 13 Nc3 Bb7 14 Ne5 g6 15 a3 Bxc3 16 bxc3 Qd6 17 Qd2 c5 gives
Black excellent play for the pawn.
12 ... Bb7 13 Qxc7 Bxf3 14 gxf3
It might have been wiser to exchange queens here with 14 Qxd8, though after 14 ...
Raxd8 15 gxf3 a6 16 Nc3 Rxd4 17 Bxa6 Rd2, Black again gets excellent play for the
pawn.
14 ... Qd5
Here 14 ... a6 15 Qxd8 Raxd8 would transpose to the previous note, but with
White’s king lacking the possibility of long-term pawn cover it’s understandable that
Black wants to keep the queens on.
 15 Rg1
Giving up a couple of weak pawns in order to generate counterplay on the open g-
file. Alternatively, White might try and hang on to them with 15 Kg2 and develop the h1-
rook to the centre.
15 ... Qxf3 16 Rg3 Qh1+ 17 Rg1 Qxh2 18 Qe5 Rfd8

19 Be4

Question: Why can’t White snaffle that knight on f6 with 19 Qxf6 - ?

Answer: Because Black removes the rook that pins the g7-pawn with 19 ... Qxg1+,
winning material.
19 ... Rab8 20 Rh1 Rxb5!
Obviously Black had seen this in advance.
21 Qc7
Instead, 21 Qxb5 Qxf4 22 Bxh7+! Nxh7 23 Qxb4 is assessed by the engine as dead
equal due to the large number of perpetual check possibilities; for example, after 23 ...
Rxd4 24 Qc5 Rd2 25 b3, there follows 25 ... Qf3 26 Rh2 Rd1+ 27 Rxd1 Qxd1+ 28 Kg2
Qg4+ 29 Kh1 Qd1+ etc.
21 ... Ba5 22 Qc1
 Not 22 Qxd8+?? Bxd8 23 Rxh2 Nxe4, when Black has two pieces for the rook and a
lot of great targets (b2, d4, f2 and f4).
22 ... Qxh1+!
The point. Black gets only rook and knight for the queen but terrific positional
compensation because of the weakness of White’s pawns. Otherwise 22 ... Bd2!? 23
Qc7 (not 23 Qxd2?? Nxe4 etc) 23 ... Ba5 is a surprising draw by repetition.
23 Bxh1 Bb6

24 Qc2
White loses this game because he keeps playing ‘normally’ when something
exceptional was required. Here he should have pushed 24 a4 at once, after which the
attempt to win the f4-pawn with 24 ... Rf5?! can be met by 25 d5! Nxd5 26 a5 Bc7 (or
26 ... Bd4 27 Ra4, developing the rook) 27 Bxd5 Bxf4 28 Qc5 Rfxd5 29 Qxa7 and
suddenly White has good winning chances. Black should prefer 24 ... Rb4 in this line.
24 ... Rf5 25 a4 g6
Making luft and waiting to see where White commits his rook before deciding which
pawn to take. For instance, 25 ... Rxf4 might be met by 26 Ra3, whereas that move can
now be answered by 26 ... Rxd4.
26 Rd1?
The last chance was 26 a5!? Rxa5 (if 26 ... Bxd4 27 Qc7 or 26 ... Bxa5 27 d5) 27
Rxa5 Bxa5 28 Qc4, followed by 29 d5, when the queen can still make a nuisance of
herself, though it looks rather hopeless in the long run.
26 ... Rxf4 27 Rd3 Ng4
Time trouble might have been playing a role at this point. Simply 27 ... Rdxd4! is
very strong.
28 f3?

And this should lose at once to 28 ... Rfxd4!, because 29 Rxd4 Ne3+ wins the
queen. White had to play 28 Bf3, even if 28 ... h5 is still very nice for Black.
In fact, the computer throws up numerous improvements for both sides over the
remaining moves, but as they don’t alter the basic assessment of “winning for Black”,
and as we’re well beyond the domain of Queen’s Gambit Declined theory, I don’t want
to bore the reader with further analysis.
28 ... Nf6 29 b4 Nd5 30 Qd2 a6 31 Bg2 Rc8 32 a5 Ba7 33 Qb2 Rb8 34 Rb3 Bxd4
35 Qe2 Rb5 36 Qc4?
Losing the exchange, though White’s position was horrendous in any case.
36 ... Ne3+ 37 Rxe3 Bxe3 38 Qc8+ Kg7 39 Qxa6 Rfxb4 40 Ke2 Re5 41 Qd6
Rb2+ 42 Kf1 Rd5 0-1
White must surrender his queen to stave off mate.

Game 63
W.Browne-Y.Balashov
Taxco Interzonal 1985

1 d4 Nf6 2 c4 e6 3 Nf3 d5 4 Nc3 Nbd7 5 cxd5 exd5 6 Bf4

This move is peculiar to the 4 ... Nbd7 line. 6 Bg5 is a standard Exchange Variation,
after which 6 ... Be7 7 e3 c6 or 7 ... 0-0 leads back to Chapter One. Note that having
played his knight to f3 already, White does not have the Bd3 and Nge2 plan of Chapter
Two available.
6 ... c6 7 e3
Another possibility is 7 h3, which safeguards White’s dark-squared bishop from
exchange. We’ll be considering this move in Game 65.
7 ... Nh5 8 Bg3
Allowing Black to get the bishop pair, hoping the h-file might then be useful.
Instead, 8 Bg5 Be7 will be examined in the next game, whereas Balashov himself once
played 8 Be5 in this position. Y.Balashov-M.Taimanov, Leningrad 1977, continued 8 ...
Nxe5 9 dxe5 g6 10 h3 Bb4 11 Qb3 Qe7 12 0-0-0 0-0 13 g4 Ng7 14 e4 and now 14 ...
Bxc3 (rather than 14 ... Be6 as played) 15 Qxc3 Be6 would have give Black a
comfortable game.
8 ... Nxg3
 Question: Isn’t it dangerous to give White the open h-file?

Answer: Not especially, because Black can build his fortifications before castling
kingside or else wait for White to castle kingside himself.
9 hxg3 Bd6

10 Bd3
A.Anastasian-P.Kiriakov, St Petersburg 1994, saw 10 Qc2 Nf6 11 Ne5 (an example
of meeting White’s plan to castle long is 11 Bd3 g6 12 0-0-0 0-0 13 Rde1 Re8, when
Black’s king is not in any danger and can also be reinforced with, say, ... Bf8-g7) 11 ...
g6 12 f4 Ng4 13 Qd2 Bxe5 14 dxe5 Qb6 15 e4 Be6 16 exd5 Bxd5 17 Nxd5 cxd5 and
Black had a comfortable game.
10 ... Nf6 11 Qc2 Qe7 12 0-0
Seeing that Black is in no hurry to castle, Browne abandons the idea of giving mate
down the h-file. If White tries to keep these hopes alive with 12 0-0-0, Black can dash
them with 12 ... g6; for example, 13 Rde1 Be6 14 e4 (14 Ng5 0-0-0 15 Nxe6 Qxe6 was
fine for Black in E.Degtiarev-J.Stanke, German Championship, Höckendorf 2004) 14 ...
dxe4 15 Bxe4 Nxe4 16 Rxe4 was A.Veingold-Y.Kruppa, Sverdlovsk 1984, and now 16
... Qd7, intending ... 0-0-0, would have been fine for Black.
 In Y.Seirawan-U.Andersson, USA vs. Scandinavia, Reykjavik 1986, White
prevaricated for another move with 12 a3, but after 12 ... g6 both sides then castled
short: 13 0-0 0-0 14 b4 Re8 15 Na4 Ne4 16 Rfe1 Bg4 17 Nc5 h5 18 Nd2 Nxd2 19
Qxd2 b6 and Black had a very comfortable game. The two bishops are definitely worth
something in this kind of position.
Note that 12 Bxh7? is bad because of 12 ... g6, trapping the bishop.
12 ... 0-0 13 Rfe1 g6

14 a3
The immediate 14 e4 is also fine for Black after 14 ... dxe4 15 Nxe4 Nxe4 16 Rxe4
Qc7 17 Rae1 Be6.
14 ... Bd7 15 e4
Attempting to seize the initiative in the centre, notwithstanding the fact that Black’s
bishops will like the opening of the position. On the other hand, 15 b4 is strongly met by
15 ... a5, when White’s queenside pawns are more of a weakness than a strength.
15 ... dxe4 16 Nxe4 Nxe4 17 Bxe4
Here 17 Rxe4 was slightly better, as after doubling rooks on the e-file White will be
able to put his knight on e5 without having to play f2-f4.
17 ... Qf6 18 Ne5 Be6 19 Rad1 Rad8 20 f4 Bc7 21 Qc3 Bb6
 Applying pressure on the isolated d-pawn. The position isn’t easy for White as his
activity seems built on feet of clay.
22 a4 a6 23 Kh2 Rd6 24 a5 Ba7 25 Qb4 Bc8 26 Bb1
An almost imperceptible mistake, which allows Black to take the initiative. White
had to play 26 Bc2, defending the rook on d1. Then after 26 ... Qd8 27 Nf3 c5, he is
able to play 28 dxc5.
26 ... Qd8 27 Ba2
Now 27 Nf3 is strongly met by 27 ... c5! because the rook on d1 is insufficiently
protected.
27 ... Kg7 28 d5?
It seems likely that Browne was suffering from his habitual time trouble at this stage,
otherwise he might have played 28 Qc3. The text move makes it looks as if White is
pressing, but in fact it puts his king in great danger.
28 ... g5!
 This is the problem; Black is threatening checkmate with 29 ... Rh6.
29 fxg5 Qxg5 30 Nf3 Rh6+ 31 Nh4
Cobbling a defence together, at least for the time being.
31 ... Bf2!
A strong move, if followed up correctly. The engine suggests an alternative in 31 ...
Bb8!, threatening both ... Bxg3+ and ... Rxh4+, and if 32 d6 then 32 ... Qf6 picks up the
d-pawn to start with.
32 Qc3+ Kg8?
Here Black had to find 32 ... f6!, which wins since 33 Re4 can be met by 33 ... Re8!
34 Rxe8 Rxh4+ 35 gxh4 Qxh4+, forcing White to put his queen in the way.
33 Re4 Bg4
After 33 ... Bf5 Black would still be slightly better, but Balashov, perhaps short of
time himself, goes for a draw.
34 Rf1 Rxh4+ ½-½
Since 35 gxh4 Qxh4+ 36 Qh3 Bxh3 (or 36 ... Bg3+ 37 Kg1) 37 Rxh4 Bxh4 38 Kxh3
produces a drawn endgame.

Game 64
G.Hertneck-V.Ivanchuk
Munich 1994

1 d4 Nf6 2 c4 e6 3 Nf3 d5 4 Nc3 Nbd7 5 cxd5 exd5 6 Bf4 c6 7 e3 Nh5 8 Bg5 Be7 9
Bxe7 Qxe7

Question: Hasn’t this just transposed into an Exchange Variation where White plays
Bg5 and Black answers with ... Nh5 - ?

Answer: Not quite because White has lost a tempo. Instead of Bg5 and Bxe7, he has
played Bf4, Bg5 and only then exchanged on e7. This is clearly quite favourable for
Black.
10 Nd2
A move which requires some explanation. White would like to play 10 Bd3, but this
is well met by 10 ... Nf4!; for example, 11 0-0 Nxd3 12 Qxd3 0-0 13 Rfe1 Nf6 14 Ne5
Ne8 (intending ... Nd6 and ... f7-f6) 15 e4 dxe4 16 Qxe4 Be6 17 Rad1 Rd8 18 Nd3 Nf6
19 Qe5 Rd6 20 Nf4 Rfd8 gave Black a comfortable game in A.Kotov-M.Euwe,
Staunton Memorial, Groningen 1946. So Hertneck hopes to get the black knight on h5 to
retreat, after which he can develop his bishop optimally to d3.
Some players have opted for 10 Be2 instead, which isn’t the best square for the
bishop but at least prevents ... Nf4 ideas. B.Damljanovic-Pr.Nikolic, Yugoslav Team
Championship 1988, continued 10 ... Nhf6 11 Qc2 0-0 12 Nd2 Re8 13 0-0 Ne4 14
Ndxe4 dxe4 15 d5 c5 16 Rfd1 Nf6 17 a4 Qe5 18 h3 Bd7 19 Bb5 Red8 20 Rd2 h5 and
Black had fully equal play.
10 ... Nf4
This in turn is an interesting response. Unless he plays g2-g3, White can’t develop
his kingside until the knight goes away, and the simplest other way to drive it back is to
castle queenside. This might not have been Hertneck’s idea when he played 10 Nd2.
11 Qc2 Nb6 12 0-0-0 Ne6 13 Bd3 Bd7 14 Kb1
As usual, 14 Bxh7? is answered by 14 ... g6.
14 ... g6

Ruling out Bxh7 altogether and envisaging a possible exchange of light-squared

bishops with ... Ng7 and ... Bf5.
15 h4 0-0-0 16 Ne2 Kb8 17 Rc1 Bc8 18 a3 Nd7 19 b4
Starting a kind of minority attack but in front of White’s castled king! Actually, this
is not such a bad plan because White can pack this area of the board with pieces.
Nonetheless, he weakens his king position with this action, and that becomes more
significant as pieces are exchanged.
19 ... Nf6 20 Ka2 Ng4 21 Rcf1 Ng7 22 Nf4 Bf5!
Finally playing the thematic manoeuvre associated with ... g7-g6.
 23 Nf3 Rc8 24 Rb1 Rc7 25 Rb2 Rhc8 26 Qb3 Bxd3 27 Nxd3 Nf5!
Another thematic idea; the knight is heading for d6. Once it gets there it will control
a number of key squares such as e4, c4 and b5, and protect b7 at the same time.
28 h5 Nd6 29 hxg6 hxg6 30 Nfe5 Nxe5 31 Nxe5 g5 32 Re2 Qe6 33 Nd3 Re8 34
Rh2 Nc4
After the occupation of this weak square we can say that Black is clearly better. His
next step is to set his own minority attack in motion with ... f7-f5-f4, prising the half-
open e-file fully open using a pawn lever.

35 Nb2 f5! 36 Rc2

Exchanging knights himself with 36 Nxc4 dxc4 just gives Black a passed pawn.
36 ... Nxb2 37 Kxb2 f4! 38 exf4 gxf4 39 b5
Note that 39 Qf3 is well met by 39 ... Qf6.
39 ... Qf6
Actually, 39 ... cxb5 looks pretty good for Black too.
40 bxc6 Rxc6 41 Qxd5?
A losing move. White should have tried 41 Rxc6 Qxc6 42 Qd3.
41 ... Rxc2+ 42 Kxc2 Qg6+ 43 Kb3 a6
 Surprisingly, the white king has no defence against the attack from Black’s two
heavy pieces. For instance, 44 Rh3 Rc8 45 Qe5+ (or 45 Rc3 Qb1+) 45 ... Ka7 46 Qe2
Qb1+ 47 Qb2 Qd1+ 48 Kb4 (or 48 Ka2 Rc2) 48 ... Rc6 wins.
44 a4 Qb6+ 45 Kc2 Rc8+ 46 Kd3 Qb1+ 47 Ke2 Rc3 48 Qe5+ Ka7 49 Qxf4 Rd3
50 Qc1 Qxc1 51 Kxd3 Qg1 0-1

Game 65
Zhu Chen-B.Spassky
Women vs. Veterans, Marbella 1999

1 d4 d5 2 c4 e6 3 Nc3 Nf6 4 Nf3 Nbd7 5 cxd5 exd5 6 Bf4 c6 7 h3

 Question: What’s the point of that? Is White preparing a pawn storm with g2-g4?

Answer: In the last couple of games we saw how 7 e3 Nh5 forced the exchange of
White’s dark-squared bishop and looked rather nice for Black. Hence 7 h3 is designed
simply to preserve the bishop.
7 ... Bb4
Black’s top scoring response, adopting a set-up in which White’s h-pawn move is
irrelevant.
8 Qb3 Qe7
Black has also played 8 ... Qa5; for example, 9 Bd2 0-0 10 e3 Re8 11 Bd3 Qb6 12
Qc2 Bd6 13 0-0 Qd8 14 Rfe1 h6 15 Rab1 Nb6 16 b4 a6 17 a4 Be6 18 Ne5 was
M.Krasenkow-Y.Seirawan, Istanbul Olympiad 2000, and now 18 ... Nc4! seems to give
Black excellent play. Note that Black prepared well for the minority attack, as 18 b5 is
met by 18 ... cxb5 19 axb5 a5 with a good game.
9 a3 Bxc3+
Giving White the bishop pair but preparing to commence active operations in the
centre. Alternatively, 9 ... Bd6 isn’t a silly move either.
10 Qxc3
 Question: Is 10 bxc3 a good move here?

Answer: It’s certainly playable, but Black gets a pretty good game with 10 ... Ne4
11 e3 Nb6, intending ideas such as ... Bf5 and ... g7-g5.
10 ... Ne4 11 Qc2 0-0 12 e3 c5

Playing like this wouldn’t be a great idea if White had already castled, since Black
is giving himself a weak d-pawn. Here it’s an attempt to exploit his lead in development
which will disappear if Black plays slowly.
13 Bd3
After 13 dxc5 Ndxc5 14 Be2 Bd7 15 0-0 Ba4 16 Qb1 Rfc8, Black has very good
play.
13 ... b6
There’s also a case for 13 ... Re8 14 0-0 c4 15 Bxe4 Qxe4 16 Qxe4 Rxe4 with what
looks like an equal endgame.
 14 Ne5
Black was probably quite happy to see this move as he gets to exchange off the
inactive knight on d7. Castling short would have been a better choice, when 14 ... Bb7
just leaves it equal.
14 ... Nxe5 15 dxe5 Bf5 16 Qd1
Black was threatening 16 ... c4, while 16 b3 is well met by 16 ... b5! since 17
Bxb5? Ng3 wins material.
16 ... Rad8 17 0-0 Bg6 18 Qg4 Qe6
A good move for the older chess player, looking to get the queens off the board and
thus minimize White’s play.
 19 Bb5?!
I’m not sure what this move was for. Simply 19 Rfd1 made more sense.
19 ... Bf5 20 Qh4?
Probably overestimating her chances on the kingside or else underestimating the
strength of Black’s response. She should have retreated the queen to e2.
20 ... h6 21 Qh5 Nd2 22 Rfe1 Nb3 23 Rad1 c4!
The immediate 23 ... Bc2 would be met by 24 Bd3! Bxd1 25 Rxd1, when the two
bishops offer some compensation for the exchange. After the text Black threatens 24 ...
Bc2 anyway, so White’s reply is virtually forced.
24 e4 Bxe4 25 Bxc4 dxc4 26 Rxe4??
Losing on the spot. 26 Rxd8 Rxd8 27 Rxe4 was mandatory, though still very good
for Black after 27 ... f5 28 Re1 Nc5, intending 29 ... Nd3.
26 ... Qg6! 0-1
White’s queen is overloaded. If she retreats to f3 or e2 then 27 ... Rxd1+ 28 Qxd1 Qxe4
wins the rook on e4, while 27 Qxg6 sees her decoyed from the defence of the rook on
d1 and 27 ... Rxd1+ is check!
Chapter Five
The Catalan and the Réti
Introduction

1 d4
One of the joys of playing the Queen’s Gambit Declined as Black is that it provides
a ready-made answer to the English and Réti Opening as well. Black just adopts the
same set-up with pawns on d5 and e6 and then plays ... Nf6, ... Be7, ... Nbd7 and ... 0-0.
But he needs to make sure he answers 1 c4 with 1 ... e6 rather than 1 ... Nf6.

Question: Why is that?

Answer: Because 1 ... Nf6 can run into the Flohr-Mikenas Attack, 2 Nc3 e6 3 e4!?,
which would require way more study than a 1 ... e6 move order.

Question: What if White then delays or omits the move d2-d4 - ?

Answer: Just follow how Lputian plays the position in Game 66, which provides an
excellent model for Black.
1 ... d5 2 c4 e6 3 Nf3 Nf6 4 g3 Be7 5 Bg2 0-0 6 0-0 Nbd7
 Question: Can Black play 6 ... c6 here instead?

Answer: He can, but there are some positions in which Black might want to delay or
omit ... c7-c6 altogether. For example, after 6 ... Nbd7 7 Nbd2, Black’s most accurate
move is 7 ... b6! as in Game 67.
7 Qc2
A logical move, both protecting the c4-pawn and preparing to play for e2-e4.
Having said that, White can also put the queen on b3, when 7 ... c6 transposes into
Game 68.
7 ... c6 8 Nbd2
I’m giving this as the main line as it’s currently thought to be the most challenging
continuation for Black to face. White’s idea is simply to play e2-e4, gaining space in the
centre.
The older version of this plan involved playing 8 b3 first, after which 8 ... b6 9 Rd1
Ba6 10 Nbd2 Rc8 11 e4 achieves the same goal. But as I discovered in Game 69,
White’s queenside fianchetto does not particularly help him after he pushes on with e4-
e5. White has also tried 9 Nc3, though Rafael Vaganian showed a good way to play for
Black in Game 70.
One other possibility for White is to eschew the possibility of e2-e4 and instead
develop with 8 Bf4. This is certainly sound enough, but Black has few problems with
accurate play. Once again Vaganian showed the way in Game 71.
8 ... b6 9 e4 Bb7 10 e5 Ne8 11 cxd5 cxd5 12 Re1 Rc8 13 Qa4 a5

We are following D.Tratatovici-Y.Zilberman, Kiryat Ono 2016 (Game 72), in which

Yaacov Zilberman provided a textbook demonstration of how to play such positions as
Black. When White put his pieces on somewhat clumsy squares, Black hit out with ...
f7-f5.

Game 66
B.Badea-S.Lputian
Porto San Giorgio 1999

1 c4 e6 2 Nf3 d5 3 g3 Nf6 4 Bg2 Be7 5 0-0 0-0 6 b3

Here 6 d4 transposes to the Catalan, which is examined in the rest of the chapter.
6 ... c6
 This represents an interesting alternative to the standard 6 ... b6 and mirrors the
treatment covered against the Catalan.

Question: What’s the idea? If Black plays ... c6-c5 later, won’t it just lose a tempo?

Answer: In such a slow position time is not the most important factor. What Black
wants here is to have the option of meeting c4xd5 with ... c6xd5 and otherwise play ...
c6-c5 in his own good time. Meanwhile he keeps the option of ... e6-e5 as well, which
might be good under certain circumstances.
7 d3
Keeping the game in English territory. It is not too late for 7 d4, when 7 ... Nbd7 8
Qc2 reaches a line covered in Games 69 and 70.
White has a major alternative in 7 Bb2; for example, 7 ... Nbd7 8 e3 b6 9 Nc3 Bb7
10 Qe2, and now Bosko Abramovic demonstrated an interesting approach for Black: 10
... a5 11 d3 Nc5 12 Rad1 (12 d4 Nce4 13 Ne5 Nxc3 14 Bxc3 Ne4 is also fine) 12 ... a4
13 d4 Nce4 14 Nd2 Nxd2 15 Qxd2 a3 16 Ba1 Ba6 17 Rfe1 Bb4 with pretty good
counterplay in K.Thorsteins-B.Abramovic, Reykjavik 1982
7 ... Nbd7 8 Bb2 b6 9 Nbd2 Bb7
 10 Qc2
Ideally White would like to play e2-e4 and have his queen on e2 rather than c2, so
10 e4 makes sense here. E.Prokopchuk-V.Malakhov, Koszalin 1999, showed a good
way for Black to get counterplay with 10 ... a5 11 a3 b5 12 Qc2 (12 Qe2 Qb6 shadows
the bishop on b2, which is probably why White chose to protect it with the queen) 12 ...
Qb6 13 e5 Ne8 14 cxd5 cxd5, leading to a position where White’s space on the kingside
will not be of much value because his pieces aren’t well placed enough to capitalize on
it.
10 ... Rc8 11 e4 c5 12 e5 Ne8 13 Rfe1
After 13 Rae1, as in S.Berry-A.Pimenta, Leuven 2011, Black should play the same
way with 13 ... Nc7, intending ... b6-b5 and ... a7-a5.
13 ... Nc7
The knight is well placed here because it helps support a black pawn storm on the
queenside.
14 Rad1 b5 15 Nf1 a5
 These forthright pawn advances are worth noting. Black’s number one priority is to
prise open files on the queenside where he has more space.
16 h4 a4 17 Bc1
After 17 bxa4, Black would switch the emphasis of the struggle to the a-file with 17
... bxc4 18 dxc4 Ra8.
17 ... axb3 18 axb3 Na6 19 Bf4 Nb4
Jumping into the newly created outpost.
20 Qe2 Ra8 21 Ra1 bxc4 22 bxc4 Nb6 23 Ng5
 A standard attacking idea in such positions; White hopes to lure one of Black’s
kingside pawns forward (such as ... h7-h6) and then prepare a second wave of attack to
exploit this. He does succeed in creating a weakness but the second wave never
materializes.
23 ... h6 24 Nh3 Qd7 25 Ne3 dxc4 26 Bxb7 Qxb7 27 Nxc4 Rxa1 28 Rxa1 Rd8
With White’s attacking chances having disappeared, it’s all about Black now. There
are plenty of light square targets, plus that knight on h3 is a miserable beast.
29 Rd1 Nxc4 30 dxc4 Rxd1+ 31 Qxd1 Qe4 32 Qb3 Nd3 33 Qc2 Qf3 34 Qd2
 34 ... Bxh4!
A neat tactic which nets the first material gain.
35 gxh4 Qxh3 36 Bg3 Qf5 37 Qe3 h5 38 Kf1 Kh7 39 Ke2
This loses without much of a fight. 39 Kg2 would have been more tenacious.
39 ... Nb2 40 Kf1 Qd3+ 41 Qxd3+ Nxd3 42 f4
Otherwise 42 ... Nb2 just wins the pawn on c4. The text allows White to meet that
with 43 Bf2, attacking c5, though it doesn’t save him because Black just brings his king
up.
42 ... Kg6 43 Ke2 Nb2 44 Bf2 Kf5 45 Kf3 Nd3 0-1
White’s position is hopeless; for example, 46 Be3 f6 47 exf6 gxf6 48 Bd2 e5 49
fxe5 Nxe5+ 50 Ke2 Nxc4 wins a second pawn.

Game 67
W.Addison-A.Bisguier
US Championship, New York 1962

1 Nf3 Nf6 2 c4 e6 3 g3 d5 4 Bg2 Be7 5 0-0 0-0 6 d4 Nbd7 7 Nbd2

 7 ... b6!
A move order finesse that I think is worth noting. Black is willing to accept the
possibility of hanging pawns on the d- and c-files, but only when White’s queen’s knight
is already on d2.
On the other hand, the standard 7 ... c6 can be met effectively by 8 Re1!?, intending
e2-e4 without putting the queen on c2. For example, N.Pert-A.Trifunovic, Witley 1999,
continued 8 ... b6 9 e4 dxe4 (9 ... Bb7 10 e5 Ne8 11 cxd5 cxd5 12 Bf1 gives White a
favourable version of Game 72 because he will not need to move his queen off the c-
file) 10 Ng5 Bb7 11 Ngxe4 Rb8 12 Nc3 Rc8 13 Nf3 c5 14 d5 exd5 15 cxd5 Bd6 16
Nh4 with a clear advantage for White.
8 b3
V.Korchnoi-A.Karpov, Baguio City (19th matchgame) 1978, saw 8 cxd5 exd5 9 Ne5
Bb7 10 Ndf3 (after 10 Ndc4 Ne4 11 Qc2 f6 12 Bxe4 dxe4 13 Ng4 f5 14 Nge5 Nxe5 15
dxe5 Qd5, Black has a nice game as his queen is about to park herself on the e6-square)
10 ... c5! 11 b3 (11 dxc5 bxc5 creates a hanging pawn position in which there is no
pressure on the pawns; this is the problem with the development of White’s queen’s
knight to d2 rather than c3) 11 ... a5 12 Bb2 Ne4 13 Rc1 Re8 14 Nxd7 Qxd7 15 Ne5
Qe6 16 Nd3 Bd6 17 dxc5 bxc5 18 e3 a4 and Black was somewhat better.
8 ... Bb7 9 Bb2 c5 10 e3 Rc8 11 Ne5 cxd4 12 exd4 Rc7
 Question: That’s a strange-looking move; what’s the idea?

Answer: The main one is to bring the queen to a8 and follow up with ... d5xc4 in
order to exchange off White’s Catalan bishop. The rook meanwhile is in the process of
being doubled, either on the c- or d-file.
13 Qe2 Qa8 14 f4
This looks aggressive but its main effect is merely to weaken White’s position. 14
Rfe1 would have been better, though Black still stands well after 14 ... dxc4 15 Bxb7
Qxb7 16 bxc4 Nxe5 17 dxe5 Nd7, as in V.Stamenkov-J.Bogoevski, Macedonian Team
Championship 2005.
14 ... dxc4 15 Bxb7
Allowing the exchanges to take place on g2 also turns out to be good for Black; for
example, 15 bxc4 Bxg2 16 Qxg2 Qxg2+ 17 Kxg2 Nxe5 18 fxe5 Ng4 19 Rf3 Bg5 20 Ne4
Rxc4! 21 Nxg5 Rc2+ and ... Rxb2 leaves Black a pawn up, since 22 Kh3 is met by 22 ...
h5!, threatening mate on h2 as well.
15 ... Qxb7 16 bxc4 b5 17 c5?
 This runs into a very promising ‘sacrifice’ which ends up netting Black a rook and
two pawns for two minor pieces, and Black gets the ‘compensation’ as well.
Addison should have played 17 Rab1, after which 17 ... bxc4 18 Ba3 Nb6 19 Bxe7
Rxe7 20 Ndxc4 Qa6 looks nice for Black because of the weakness of the opposing king
position, but this would have been a lot better for White than the game.
17 ... Nxc5! 18 dxc5?!
Objectively speaking, there was a case for simply accepting the loss of a pawn and
playing 18 Rac1; but such moves are psychologically very difficult because they accept
both a loss and having made a mistake.
18 ... Bxc5+ 19 Rf2 Bxf2+ 20 Qxf2 Rc2 21 Qd4
After 21 Bd4 Ne4 22 Nxe4 Rxf2 23 Nxf2 Rd8, White is left with three minor pieces
for a queen and two pawns, but his position is hopeless because the bishop and knights
don’t have enough outposts.
21 ... Qb6
Threatening both ... Rxb2 and ... Rxd2.
22 Qxb6 axb6 23 Bc1 Rd8 24 Nf1?
Allowing Black’s rook to the eighth rank makes matters worse for White. 24 Nef3
was preferable; even so, it’s hard to see White surviving after 24 ... Rd3!. For example,
25 Ne5 Rd5 26 Nef3 Ne4 27 Kf1 (27 Nxe4 Rd1+ leads to mate) 27 ... Rdxd2 28 Bxd2
Nxd2+ 29 Nxd2 Rxd2 30 h4 Rh2 31 Kg1 Rb2 gives Black a winning rook endgame.
 24 ... Rd1 25 Bb2 Rxa1 26 Bxa1 Rxa2 27 Bd4 Nd5

There’s no coming back from this. In addition to his material inferiority (Black has
three pawns now), White has to cope with the powerful position of Black’s rook on the
seventh rank and his passed b-pawns.
28 Nf3 b4 29 N1d2 Kf8 30 Kf1 Rc2 31 Ke1 f6 32 Kd1 Rc8 33 Ke2 Ke7 34 Kd3
Kd7 35 Nc4 b5 36 Na5 Nc3 37 Nd2 Kd6 38 Nb7+
Swapping knights with 38 Ne4+ Nxe4 39 Kxe4 is also quite hopeless; for example,
39 ... Rc2 40 h4 f5+ 41 Ke3 Kd5 42 Bxg7 Ra2 43 Nb3 (or 43 Nb7 Ra7) 43 ... Ra3 etc.
38 ... Kc6 39 Nc5 Kd5 40 Nd7 Na4 0-1
White’s resignation coincided with the arrival of the time control. This is often the
first moment for calm reflection after some hectic final moves with the flag hanging.

Game 68
H.Grooten-M.Kuijf
Eindhoven 1981

1 d4 Nf6 2 c4 c6
Despite the unusual move order we’ll soon get back into a Catalan. The standard
route from here is 2 ... e6 3 Nf3 d5 4 g3 Be7 5 Bg2 0-0 6 0-0 Nbd7 7 Qb3 c6,
transposing to the position after Black’s 7th move.
3 Nf3 d5 4 Qb3 e6 5 g3 Nbd7 6 Bg2 Be7 7 0-0 0-0 8 Nc3 b6 9 Bf4 Ba6

Question: Why is Black playing this move instead of 9 ... Bb7 with a standard
fianchetto?

Answer: We’ll see him opt for ... Bb7 in a similar position in Game 71. Here Black
wants to force an exchange on d5, after which a later e2-e4 (trying to activate the
Catalan bishop on g2) would lead to White having an isolated d-pawn.
10 cxd5

Question: If it’s undesirable for White to capture on d5, can’t he just defend the c4-
pawn with 10 Nd2 - ?

Answer: That’s a good question, and White often does this kind of thing. The engine
then likes the tacit draw offer 10 ... Nh5 11 Be3 Nhf6, while the attempt to take the
initiative with 12 Qa4 is well met by 12 ... Bxc4! 13 Nxc4 b5! 14 Qa6 bxc4 15 Qxc6
Qb6 with a comfortable game for Black.
10 ... cxd5
 Alternatively, 10 ... exd5 has been played as well and is quite a good option if
Black wants to create strategic imbalance. D.Kryakvin-A.Dreev, Russian Team
Championship 2011, went 11 Qa4 Bb7 12 Rfd1 Re8 13 Ne5 Nxe5 14 Bxe5 (14 dxe5
Nd7 is also fine for Black) 14 ... Ng4 15 e4 b5 16 Qb3 Nxe5 17 dxe5 d4 18 Ne2 c5 19
Qxb5 Qb6 with more than enough compensation for the pawn.
11 Qa4
In K.Lahno-A.Muzychuk, Women’s World Rapid Championship, Batumi 2012,
White played 11 Nb5! at once, which led to a quick draw after 11 ... Bxb5 12 Qxb5 a6
13 Qd3 b5 14 Rfc1 Qb6 15 Rc2 Rfc8 16 Rac1 Rc4 17 b3 ½-½. This all seems solid
enough, but to fight for the full point Black would be well advised to follow the
treatment of Alexey Dreev in the previous note.
11 ... Bb7
Black can also defend the bishop with 11 ... Qc8 and then place the queen on b7; for
example, 12 Nb5 Bxb5 13 Qxb5 a6 14 Qb3 Qb7 15 Rfc1 Rac8 was quite equal in
A.Mastrovasilis-V.Belov, Kavala 2009.
12 Nb5 Ne4
A pugnacious approach. M.Pein-H.Elyoseph, Tel Aviv 1987, varied here with 12 ...
a6 13 Nd6 b5 14 Qb3 (14 Nxb7 Qb6 15 Qb3 Qxb7 transposes) 14 ... Qb6 15 Nxb7
Qxb7 16 Rfc1 Rfc8 17 a4 b4 18 Nd2 Rxc1+ 19 Rxc1 Rc8, when White’s bishop pair
gave him no more than a theoretical edge.
13 Rfc1 a6 14 Nc7!?
Clearly the critical move. It’s not very good, but after 14 Nc3 Nd6!, intending ...
Rc8 and ... b6-b5, Black is doing well anyway.
14 ... Rc8
 15 Nxa6?!
Winning a pawn temporarily but walking into a pin. We’re more in the domain of
computer analysis than understanding here; accordingly, the engine gives the variation
15 Bh3! (threatening Bxe6) 15 ... b5 16 Qb3 g5! 17 Nxe6 fxe6 18 Bxe6+ Kh8 19 Rxc8
Bxc8 20 Qxd5 gxf4 21 Qxe4 and White escapes with three pawns for the piece. All the
same, I prefer Black’s position after 21 ... Qc7.
15 ... Ra8 16 Rc7 Bc8!
Simply 16 ... Bxa6 17 Qxd7 (or 17 Rxd7 Qe8) 17 ... Qxd7 18 Rxd7 Bd8 is fine for
Black, but naturally he wants more than that.
17 Qc6?
Pretty much a losing move. The engine tells me that 17 Rac1! is the only show in
town, when 17 ... Rxa6 18 Qc2 (18 Qc6? would transpose to the game) 18 ... Ndc5 19
dxc5 bxc5 20 Nd4 leads to unclear complications; for example, 20 ... g5! 21 Be5 f6 22
Bxe4 dxe4 23 Nb5 fxe5 24 Rd1 Qe8 25 Qxe4 and so on. However, Black can avoid all
that with 17 ... Nef6! (17 ... Ndf6 18 Rxe7 Qxe7 19 Qb4 is not quite so bad) 18 Rxc8
Rxc8 19 Bc7 Qe8 and White is struggling to justify the loss of the exchange.
17 ... Rxa6 18 Rc1 Nb8
This is good for Black, but there was something stronger: 18 ... Ndc5!, when 19
dxc5 (19 Rxc8 Qxc8 20 Qxc8 Rxc8 leaves White with zero compensation for the
exchange) 19 ... bxc5 is just winning, as White now has to answer the threat to his
queen. After 20 Qb5 (20 Ne5 g5! puts even more white pieces en prise) 20 ... g5! 21
Nxg5 Bxg5 22 Bxg5 (or 22 Bxe4 Bxf4 23 Bxh7+ Kg7!) 22 ... Qxg5 23 Rc2 Qd8 24
Rxc8 Qxc8, White ends up a rook down.
19 Qc2 Bd7 20 Ne5 Bd6

Aiming for simplification into a favourable endgame, which is a very human

solution. 20 ... Nd6 gets the engine’s vote.
21 Bxe4?
This try for complications again just leaves White a whole rook down.
Instead, 21 Nxd7 Nxd7 22 Bxe4 Qxc7 23 Bxh7+ would limit White’s material
disadvantage, but after 23 ... Kh8 24 Bd3 (or 24 Qxc7 Bxc7 25 Rxc7 Kxh7 26 Rxd7
Rxa2) 24 ... Bxf4 25 Qxc7 Bxc7 26 Bxa6 Bd8, White would still be fighting desperately
for a draw.
His best chance was probably 21 Rxd7!? Nxd7 22 Bxe4 dxe4 23 Nc6, as suggested
by Van der Poel, when 23 ... Bxf4 24 Nxd8 Bxc1 25 Qc7 Nf6 26 Nxf7! still offers White
some counterplay, even if he doesn’t really have enough for the queen.
21 ... Bxc7! 22 Bxh7+ Kh8 23 Nxf7+
After 23 Qxc7 Kxh7, White has nothing for the rook.
23 ... Rxf7 24 Bxc7 Qf6
 25 Bxb8
Everything loses now. He could have defended f2 with 25 e3, but 25 ... b5! 26 Be5
(26 Bxb8 Rc6 is a killer) 26 ... Qf3 is still winning for Black.
25 ... Ra8
The immediate 25 ... Qxf2+ 26 Kh1 e5 was even stronger.
26 Bc7
White is merely aiming for confusion. After 26 Bg6 Qxf2+ 27 Kh1 Rxb8 28 Bxf7
Qxf7, Black would win with his extra piece.
26 ... Qxf2+ 27 Kh1 e5
Threatening ... Bh3 etc.
28 e3 Qf3+ 29 Kg1 Qxe3+ 30 Kg2 Raf8 0-1
Since mate is inevitable.

Game 69
N.Davies-S.Lputian
Isle of Man 2003

1 c4 e6 2 g3 d5 3 Bg2 Nf6 4 Nf3 Be7 5 0-0 0-0 6 d4 Nbd7 7 Qc2

Rather than b3 (as in the previous game) or later e2 (as in Game 67), in this game
and the remainder of the chapter we’ll examine the main line where White puts the
queen on c2, defending the c4-pawn and supporting the advance of his e-pawn.
7 ... c6 8 b3
The modern way to play the e2-e4 plan is with 8 Nbd2; we’ll look at this in Game
72. White has also tried just developing the queen’s bishop, either to f4 or g5, as we’ll
see in Game 71.
8 ... b6 9 Rd1

Question: What is White trying to do here?

Answer: This set-up was considered standard at the time of the game; the rook
shadows Black’s queen on the d-file, while White is prepared to meet ... Ba6 with Nbd2
and ... Bb7 with Nc3. In both cases the idea was to follow up with e2-e4, hopefully
gaining a space advantage in the centre whether Black takes on e4 or lets White push on
with e4-e5.

Question: Does it work?

Answer: Well, I went into this game with a set of ideas which would prove to be
too simplistic. First of all I believed that, once White has played e2-e4, pushing the
pawn on to e5 was some kind of positional threat, due to the cramping effect it would
have on Black’s kingside and the likelihood that White would then be able to build an
attack. This proved to be false, as we shall see.
9 ... Ba6 10 Nbd2
White has tried 10 Nc3 here too, offering the pawn on c4 as a gambit, though Black
doesn’t need to take it straight away. He can instead opt for 10 ... Rc8, after which 11 e4
dxc4 12 Bf4 (12 Ne5?! Nxe5 13 dxe5 Nd7 14 Be3 Qc7 15 f4 Nc5 16 Rd2 Rfd8 17
Rad1 Rxd2 18 Rxd2 Rd8 was good for Black in T.Grabuzova-S.Lputian, Internet blitz
2004) 12 ... b5 13 bxc4 bxc4 14 d5 cxd5 15 exd5 exd5 16 Nxd5 Nxd5 17 Rxd5 Bb7 18
Rdd1 Qe8 19 Re1 Bxf3 20 Bxf3 Nc5 21 Qxc4 Ne6 led to equality in L.Psakhis-
A.Huzman, Israeli Championship, Tel Aviv 1994.
10 ... Rc8 11 e4 Bb7

Question: Why doesn’t Black prevent the white e-pawn going to e5 with 11 ... dxe4
-?

Answer: The positions that arise after 12 Nxe4 Nxe4 13 Qxe4 are good for White
due to his space advantage. A possible continuation is 13 ... Nf6 14 Qc2 Qc7 15 Bf4
Bd6 16 Bxd6 Qxd6 17 Ne5 with a pleasant edge for White.
Black does have a major alternative here in 11 ... c5, but after 12 exd5 exd5 13 Nf1
cxd4 (13 ... dxc4 14 d5 gives White a passed pawn, which is at least a technical
advantage) 14 Nxd4 b5 15 Ne3 bxc4 (15 ... Nb6 can be met by 16 Ndf5 with dangerous
threats) 16 Nxd5 Nxd5 17 Bxd5 cxb3 18 Qxb3, White has a definite initiative as seen in
several games, such as I.Stohl-V.Dydyshko, Polish Championship, Zakopane 2000.
12 Bb2
As White ends up closing the position anyway, it’s better to do it immediately with
12 e5. The pawn structure indicates that White should play on the kingside, which
means that his queen’s bishop will probably be more useful on the c1-h6 diagonal.
12 ... Qc7 13 e5 Ne8
 14 Nf1
Lputian would reach this position again the following year. D.Fridman-S.Lputian,
Internet blitz 2004, went 14 cxd5 cxd5 15 Qd3 Qb8 16 Rac1 Nc7 17 h4 Ba6 18 Qb1
Nb5 19 a4 Nc7 20 Ng5 g6 21 Bf1 Qb7 22 Bd3 Bxd3 23 Qxd3 Na6 with about equal
play at this point.
From the other side, I faced 14 Rac1 in K.Arkell-N.Davies, Colwyn Bay 2014. The
game continued 14 ... Qb8 15 cxd5 (after 15 Nf1 Nc7 16 Ne3 Rfd8 17 cxd5 cxd5 18
Qb1 Ba6 19 h4 Qb7 20 a3 Bb5 21 Ng5 Nf8 22 Nh3 a5 23 a4 Be8, Black eventually got
... f7-f6 in and activated his light-squared bishop in V.Bogdanovski-S.Lputian, World
Team Championship, Yerevan 2001) 15 ... cxd5 16 Qb1 Rxc1 17 Rxc1 Nc7 18 Bf1 Rc8
19 h4 Nf8 20 Bd3 Na6 21 Ng5 Bxg5 22 hxg5 Nb4 23 Bf1 Ba6 and was soon drawn.
14 ... Qb8 15 cxd5 cxd5 16 Qd2 a5 17 Ne1 Nc7
This might have been time for the ... f7-f6 break. The main issue is usually the
weakness of e6, but in this position Black can meet 18 Bh3 with 18 ... Nc7.
18 Nd3 Qa7
In the post mortem Lputian indicated a certain dissatisfaction with his forthcoming
queen journey. 18 ... a4 was perhaps a better move.
19 a3 Qa6 20 Ne3 Qb5
 21 Nc1
I didn’t see it at the time but White can just play 21 a4! here, when the queen has to
go back again, since she gets trapped after 21 ... Qxb3? 22 Nc1 Qb4 23 Bc3.
21 ... Ba6 22 Bf1 Qc6 23 Nd3 Qb7 24 f4 f6 25 Bh3 fxe5 26 fxe5 Bg5
This pin is a bit awkward, but I still liked my position at this point.
27 Bg2?!
I was attracted by the idea of shadowing Black’s queen, but it makes no sense to
take the pressure off e6. White should play 27 Rac1.
27 ... Qc6 28 Nf4
Again 28 Rac1 is correct, even if not as good as on the previous move.
28 ... Qb5
Counter-attack. I was hoping for 28 ... Bxf4 29 gxf4 Rxf4 30 Rac1, when 30 ... Qb7
31 Nc4 Rcf8 32 Nd6 would have given me very active pieces.
29 Rac1 Qxb3 30 Rc3 Qa4 31 Rdc1 Nc5 32 dxc5 d4 33 Qc2
I hadn’t even considered Lputian’s tremendous reply. Had I done so I might have
opted for 33 cxb6!?.
 33 ... b5!
I was expecting 33 ... Qxc2, after which 34 R3xc2 dxe3 35 cxb6 is good for White.
The calculating machine, on the other hand, prefers the laconic 33 ... Qe8! and shows
that Black comes out on top in the complications.
34 Qxa4 bxa4 35 Rc4 dxe3 36 Rxa4 e2 37 Bf3
Here 37 Kf2!? was worth considering, but I didn’t want to walk into another pin.
37 ... Nd5 38 Bxd5 exd5 39 Rxa5
Even now 39 Kf2 might have been a better try; for example, 39 ... Bc4 40 Raxc4
dxc4 41 Rxc4 Be7 42 Kxe2 Rxc5 43 Rxc5 Bxc5 would at least leave White with more
practical chances than the game.
39 ... Bc4 40 e6 Rxf4?!
Lputian judges that the two bishops offer better support for his passed pawns, which
is positionally correct, apart from the fact that 40 ... Bxf4 41 gxf4 Rxf4 and 42 ... Rf1+
wins straight away.
41 gxf4 Bxf4 42 Ra1 Re8 43 Bc3
Or 43 Kf2 Bd2 44 Ra7 Rxe6 45 Rxg7+ Kf8 etc.
43 ... Be3+ 44 Kg2 d4 45 Kf3 Rxe6 46 Be1 Bd3 47 Ra8+ Kf7 48 Ra7+ Kg6 49
Rd7 Rf6+ 50 Kg3 Be4 51 Re7 Bf4+ 52 Kh3 Kf5 53 Rxg7 d3 54 Rxh7 d2 0-1
Game 70
R.Sukharisingh-R.Vaganian
German League 1996

1 d4 e6 2 Nf3 Nf6 3 g3 d5 4 Bg2 Be7 5 0-0 0-0 6 c4

6 ... c6
Personally, I prefer to play 6 ... Nbd7 first.

Question: Is there any particular reason for this?

Answer: Well, after 6 ... c6 7 Nbd2 Nbd7 White can play 8 Re1, intending 9 e4;
whereas 6 ... Nbd7 7 Nbd2 can be met by 7 ... b6, as seen in Game 67. But presumably
Vaganian, who plays this position from both sides, has his own reasons for preferring
the text (or else 6 ... dxc4, which leads to a completely different type of game).
7 Qc2 Nbd7 8 b3 b6 9 Nc3
Rather than playing 9 Rd1 as in Game 69, White immediately develops his queen’s
knight to c3 and plans to defend the c4-pawn with his king’s knight.
9 ... Ba6 10 Nd2
Here 10 Rd1 transposes to 10 Nc3 in the notes to the previous game.
 Question: Why can’t White secure his c4-pawn by just exchanging it off with 10
cxd5 - ?

Answer: Because after 10 ... cxd5 a black rook would come to the c-file and annoy
White’s queen. In addition, his bishop on g2 would then be badly placed because a later
e2-e4 would leave him with an isolated d-pawn.
10 ... b5

Intensifying the pressure on c4, when White must decide if he wants to sacrifice a
pawn.
11 Re1
Alternatives are as follows:
a) 11 cxb5 cxb5 12 a3 Qb6 13 Bb2 Rfc8 14 Qd1 b4 15 Na4 Qb5 16 Re1 bxa3 17
Bxa3 Bxa3 18 Rxa3 Qb4 19 Nb1 Bb5 gave Black a good game in M.Quinteros-
U.Andersson, Manila 1974.
b) 11 e4 is strongly met by 11 ... bxc4 12 bxc4 c5; for example, 13 exd5 cxd4 14
Nb5 (14 dxe6? dxc3 wins material) 14 ... Nxd5! 15 Nxd4 Rc8 16 Qa4 Nb4 17 a3 Nb6
18 Qa5 Nxc4 19 Nxc4 Bxc4 is good for Black according to the engine. Of course it’s
difficult for human players to go into such highly tactical lines with much confidence.
11 ... bxc4 12 bxc4 dxc4
 A challenging way to play the position; the pawn on c4 may be isolated but it is also
passed. The engine prefers 12 ... Rb8 here, intending 13 Qa4 Bxc4 14 Nxc4 Rb4.
13 Qa4
After 13 Bxc6 Rb8, Black starts the process of activating his forces in support of his
passed c-pawn.
13 ... Qc8 14 Qxc6
Black can meet 14 Nxc4 with 14 ... Nd5; for example, 15 Nxd5 cxd5 16 Ba3 Bxa3
17 Nxa3 Nb6 18 Qa5 Nc4 19 Nxc4 Bxc4 gives him comfortable equality.
14 ... Nb6 15 Rb1 Qd8 16 e3 Nfd5

17 Nxd5??
A serious mistake, after which White has trouble getting his queen out of danger. 17
Bxd5 was a better way to capture on d5, when 17 ... Rc8 18 Rxb6 Rxc6 19 Rxc6 Bb7
20 Bg2 Bxc6 21 Bxc6 gives White three good pieces for the queen, while 17 ... exd5 18
Nxd5 Qxd5 19 Rxb6 Qxc6 20 Rxc6 leaves him a pawn up, although Black’s passed
pawn and two bishops should provide sufficient compensation after 20 ... Bb5 21 Rc7
Bd6.
17 ... exd5
Now White faces the instant threat of 18 ... Rc8, winning his queen.
18 Bh3 f5
 Threatening 19 ... Rf6, trapping the queen another way; and 19 Qe6+ Kh8 doesn’t
help since 20 Bxf5 loses to 20 ... Bc8. White has to do something dramatic.
19 Rxb6 axb6 20 Bg2 Kh8 21 Qxd5
So White is just the exchange down, and Black still has his strong passed c-pawn.
21 ... c3 22 Nb1
After 22 Qxa8 cxd2 23 Bxd2 Qxa8 24 Bxa8 Rxa8, White’s two pawns are nowhere
near enough for his missing bishop.
22 ... Qxd5 23 Bxd5 Bb4!

Fine technique; Vaganian trades his extra exchange for the white a-pawn, and then
wins easily with his active pieces and connected passed pawns.
24 Bxa8
Or 24 Ba3 Bxa3 25 Nxa3 Bd3 etc.
24 ... Rxa8 25 Ba3 Bxa3 26 Nxa3 Bd3 27 Nb1 c2 28 Nd2
After 28 Nc3 Rc8, Black wins on the spot.
28 ... Rxa2 29 Nb3 Be4 30 Nc1
Or 30 Kf1 Rb2 31 Nd2 (or 31 Nc1 Bf3) 31 ... Bd3+ 32 Kg2 b5 and White is
helpless against the further advance of the queenside pawns.
30 ... Ra1 31 Kf1 Bf3 32 d5 Kg8 33 d6 Kf8 34 Nb3 Rd1 35 Nc1 Ke8 0-1
Since White would soon run out of moves.
Game 71
R.Buhmann-R.Vaganian
German League 2004

1 Nf3 Nf6 2 c4 e6 3 g3 d5 4 d4 Be7 5 Bg2 0-0 6 0-0 c6 7 Qc2 Nbd7

8 Bf4
An attempt to exploit Black’s coming queenside fianchetto by going for pressure on
the c-file. We saw a similar treatment (albeit with Qb3) in Game 68.
White has also played 8 Bg5 here, but that does little to disturb Black’s standard
method of development. S.Skembris-V.Epishin, Assisi 2003, continued 8 ... b6 9 Nbd2
Bb7 10 Rfd1 h6 11 Bxf6 Nxf6 12 e4 Rc8, after which 13 c5 a5 14 a3 bxc5 15 dxc5 a4
16 Rac1 (16 Qxa4 Bxc5 is good for Black, who is now targeting f2) 16 ... Qa5 17 Ne5
Rc7 18 Bf1 Rb8 left Black with the better game.
8 ... b6 9 cxd5 cxd5 10 Nc3
The immediate 10 Rc1 is also logical, though it doesn’t cause particular problems
for Black either. L.Gutman-K.Bischoff, Bremen 1998, continued 10 ... Bb7 11 Qa4 Rc8
12 Nbd2 (not 12 Qxa7?? Rxc1+ 13 Bxc1 Qc7 14 Bf4 Qc6 and White’s queen is unable
to escape) 12 ... a6 13 Qd1 Rxc1 14 Rxc1 Qa8 15 Ne5 Rc8 and the players agreed a
draw since the final position is dead equal.
 Meanwhile 10 Qc6 is well met by 10 ... Ba6, when 11 Nc3 Qc8 12 Qa4 Qb7 or 12
... Qc4 is fine for Black, who essentially has an extra tempo on 10 ... Ba6 lines.
10 ... Bb7
With the c-pawns already exchanged, 10 ... Ba6 makes slightly less sense, since
White can find a more useful move than transposing to Game 68 with 11 Qa4; for
example, 11 Rfc1 Nh5 (11 ... Qc8 might be met by 12 Qb3 Qb7 13 e4!) 12 Qa4 Nxf4 13
Qxa6 Nxg2 14 Kxg2 Qc8 15 Qxc8 Rfxc8 16 Nb5 gave White a little something in
V.Tkachiev-A.Pavlidis, Skopje 2014, although Black’s position still looks playable.

11 Nb5
Playing a rook to the c-file first gives Black time to oppose it with ... Rc8; for
example, 11 Rfc1 Rc8 12 Qd1 (or 12 Qd3 a6 13 a4 Ne4 14 Nd2 f5 and Black was fine
in U.Schulze-R.Vaganian, German League 2005) 12 ... a6 13 Rc2 Nh5 14 Bd2 Bd6 15
Rac1 Nhf6 16 Bf4 Bxf4 17 gxf4 Qe7 18 e3 g6 was equal in I.Salgado Lopez-
R.Wojtaszek, Paks 2011. In none of these lines does White look close to getting in on the
open file.
11 ... Ne8
Not now 11 ... Rc8? due to 12 Bc7! Qe8 13 Nxa7 and White wins a pawn.
12 Rfc1 a6 13 Nc3
Here 13 Bc7 looks scary at first, but White’s foray lacks sufficient backing.
L.Rindlisbacher-Z.Ribli, Geneva 2012, continued 13 ... Qc8 14 Na3 Nxc7 15 Qxc7 Rb8
16 Ne5 Qxc7 17 Rxc7 Nxe5 18 dxe5 Bc5 19 e3 Rfc8 20 Rxc8+ Rxc8, when Black had
successfully repelled boarders and gained the advantage of the two bishops.
13 ... Nd6
The knight is very well placed on d6, as it touches the e4- and c4-squares and
interferes with the f4-bishop’s action towards c7.
Instead, in E.Gereben-P.Benko, Bad Gastein 1948, Black opted for 13 ... Rc8 again,
after which 14 a4 Nb8 15 Bxb8 Rxb8 16 e4?! dxe4 17 Nxe4 Nf6 18 Nxf6+ Bxf6 19 Rd1
Qd6 gave him an excellent game.
14 Qb3
There was a case for 14 Bxd6, giving up the two bishops in order to prevent Black’s
knight coming to c4.
14 ... Nc4 15 e4

White senses that his game is becoming difficult and tries to get counterplay. If he
plays 15 Qc2, followed by b2-b3 to evict the knight, he weakens his position on the c-
file, so that 15 ... Rc8 is at least equal for Black.
15 ... b5
Simply 15 ... dxe4 16 Nxe4 Bxe4 17 Qxc4 Nf6 is another way for Black to obtain an
edge, in view of White’s isolated d-pawn.
16 exd5 exd5 17 Qc2
White decides to put the queen back on c2 again. With hindsight he might have
preferred something like 17 Re1, when he is still more or less okay.
17 ... g5!
Vaganian plays with great energy, showing that White’s minor pieces don’t have
enough squares.
18 Bd2?!
After this, White loses the exchange for inadequate compensation. 18 Be3 was
preferable here, when 18 ... g4 19 Ne5 Ndxe5 20 dxe5 Nxe3 21 fxe3 Bg5 is not at all
pleasant for White but looks better than the game.
18 ... g4 19 Nh4
19 Ne5 Ndxe5 20 dxe5 Nxd2 21 Qxd2 Bg5 is much the same.
19 ... Nxd2 20 Qxd2 Bg5

21 Qd3?!
Now White doesn’t get any compensation at all. He would have been better off with
21 Qd1 Bxc1 (even declining the material with 21 ... Nf6 22 Nf5 Qd7 23 Ne3 Rac8
looks good for Black) 22 Qxc1 Qf6 23 Bxd5 Bxd5 24 Nxd5 Qxd4 25 Qg5+ Kh8 26 Nf5
Qe5 27 Qxg4, when White still has a bit of play, though it shouldn’t really be enough.
21 ... Bxc1 22 Rxc1 Qg5
White is just the exchange down for nothing in particular. This is not a good
situation to be in against a grandmaster of Vaganian’s class.
 23 Rf1 b4 24 Nd1 Bc6 25 Re1 Rae8 26 Rxe8
Here 26 Ne3 would have been a better idea, keeping as many pieces on as possible.
26 ... Rxe8 27 Ne3 Bb5 28 Qd2 Nf6 29 Nhf5
Not 29 Qxb4? because of 29 ... Rxe3! 30 fxe3 (30 Qd2 fails to 30 ... Re5) 30 ...
Qxe3+ 31 Kh1 Qc1+, leading to mate.
29 ... Bd7 30 Nd6 Re6 31 Nb7
The b4-pawn is still immune, as after 31 Qxb4? Rxe3! 32 fxe3 Qxe3+ 33 Kf1, Black
has the powerful 33 ... Ne8!. Your engine will confirm that Black is winning in all
variations!
31 ... Bb5 32 Nc5 Re8 33 Kh1
With ideas of taking on b4 after all, but he never gets the chance.

33 ... Qg6 34 Qd1 h5 35 h4 gxh3 36 Bxh3 Qg5 37 Qf3 h4 38 Kh2 hxg3+ 39 Qxg3
Qxg3+ 40 Kxg3 Ne4+
This endgame is of course winning for Black.
41 Kf4 Nxc5 42 dxc5 d4 43 Nc2
After 43 Nd5 there follows 43 ... Re2 44 Nxb4 d3 and so on.
43 ... d3 44 Ne3 Re7 45 Bf1 d2 46 Bxb5 axb5 0-1
In view of 47 ... Rxe3 and 48 ... d1Q, or 47 Nd1 Re1 etc.
Game 72
D.Tratatovici-Y.Zilberman
Kiryat Ono 2016

1 d4 Nf6 2 c4 e6 3 Nf3 d5 4 g3 Be7 5 Bg2 0-0 6 0-0 Nbd7 7 Qc2 c6 8 Nbd2

Question: What’s White’s idea with this move?

Answer: He prepares an improved version of the e2-e4 plan. If White is going to

push on with e4-e5, he doesn’t really want his bishop on b2 and it’s better not to have
his b-pawn on b3 either.
8 ... b6
Continuing our standard scheme of development for this chapter.

Question: Besides getting his pieces out, what’s Black’s plan in this position?

Answer: He will be looking to bring a rook to c8 and then play ... c6-c5. This could
be annoying for White’s queen.
 9 e4 Bb7 10 e5 Ne8
It’s worth comparing this position to the one I obtained in Game 69. Black has saved
some time here by moving his bishop to b7 in one go. From the other side, White
reasons that this will be more than offset by the time he has saved himself in not playing
b2-b3 and Bb2. How much these factors favour White, if at all, we’ll have to see.
11 cxd5 cxd5 12 Re1 Rc8 13 Qa4

13 ... a5
The most popular of several moves. Here are a couple of alternatives:
a) 13 ... a6 was Zilberman’s earlier try, but after 14 Nf1 (or 14 Bf1 and 15 Bd3 at
once, as in A.Huzman-Y.Zilberman, Israeli Championship, Haifa 2008) 14 ... Nc7 15
Bf4 Nb5 16 h4 h6 17 Qd1 Re8 18 N1h2 Bf8 19 Bf1 f6 20 Bd3 fxe5 21 dxe5, White had
very serious chances on the kingside in D.Zifroni-Y.Zilberman, Herzliya 2009.
b) 13 ... Bc6 is quite a bright idea for those who are worried about the g2-bishop
coming round to the b1-h7 diagonal. White can’t take on a7 because his queen would be
trapped, so he should play 14 Qb3 (after 14 Qd1 Bb5 Black forestalls Bf1, which
would just be swapped off) 14 ... b5 15 Bf1 b4 16 Ba6 (16 Qe3 Qb6 17 Bd3 Bb5 18
Bb1 Nc7 19 Ng5 h6 20 Nh3 f6 was good for Black in I.Almasi-L.Portisch, Hungarian
Team Championship 1994) 16 ... Rb8 17 Bd3 Bb5 18 Bb1 Rc8 19 Nf1 Nb6 20 h4 gave
White some attacking chances on the kingside in B.Avrukh-H.Al Tamimi, Santa Cruz de
la Palma 2005.
 These examples show the care with which Black needs to conduct the defence, but
Zilberman does a good job in the main game.
14 Bf1 Nc7
Preparing to exchange light-squared bishops with ... Ba6, while at the same time
defending the e6-pawn in preparation for ... f7-f5.
In Va.Filippov-L.Portisch, European Cup, Kallithea 2002, Black opted for 14 ...
Nb8 which also supports ... Ba6. The game continued 15 Bd3 Ba6 16 Bb1 Nc6 17 Nf1
Bb4 (17 ... Qd7 was also worth considering, threatening ... Nxe5, or if 18 Qb3 then 18
... f5) 18 Bd2 f5 and Black was fine.
The Bf1-d3-b1 manoeuvre is obviously directed towards Black’s kingside, but it
does take time and gets in the way of the development of White’s queenside pieces. A
timely ... f7-f5 will often bring out the negative side of White’s idea.
15 Bd3 Ba6 16 Bb1 f5!

Again the key defensive idea, which is similar to many lines of the French Defence.
If White fails to capture en passant, Black may take the initiative on the kingside himself
with ... g7-g5.
17 exf6 Bxf6
Thus far the players had followed G.Halldorsson-Y.Kuzubov, Reykjavik 2010. On
that occasion Black played 17 ... Rxf6, after which 18 Qc2 g6 19 Qd1 Qf8 20 Re3 Ne8
gave him a very satisfactory game, again because of the poor state of White’s queenside
development. I don’t know whether Zilberman’s choice was intended as an
improvement or not; it’s entirely possible that neither player knew about the earlier
game.
18 Nf1 Re8
The immediate 18 ... Nb5! was more to the point; White can’t capture the e6-pawn
because the bishop on c1 needs defending.
19 h4
In turn White should develop his pieces with 19 Bf4.
19 ... Nb5 20 Qd1?!
White doesn’t put up much of a struggle from here on in, and the text allows a
powerful exchange sacrifice. 20 Ne3 or 20 Be3 was better, though Black has a pretty
good game in any case.

20 ... Rxc1! 21 Qxc1 Nxd4 22 N1h2 Rf8

Possibly 22 ... Nxf3+! was a stronger follow-up, after which 23 Nxf3 Nc5 leaves
White struggling against the powerful central pawns.
23 Kg2
Here 23 Qd2! was preferable, when White is still on the board.
23 ... Nc5 24 Nxd4 Bxd4 25 f4 Qa8 26 a4 Rc8 27 Qd1?
Giving the b-pawn away certainly doesn’t help White’s cause. Perhaps he meant to
play 27 Nf3, intending 27 ... Nb3? 28 Qd1, but then made the queen move first by
mistake.
27 ... Bxb2 28 Ra2 Bc3 29 Ree2 d4+ 30 Kh3 d3 31 Re3 Qd5

32 Rf2?
And now White loses the exchange back, after which he is reduced to futile hopes of
a swindle. 32 Nf3 was the best chance, albeit not a very good one.
32 ... Bd4 33 Ba2 Bc4 34 Bxc4 Qxc4 35 Rff3 Bxe3 36 Rxe3 Rd8 37 Nf3 Qxa4 38
Qb1 Qb3 39 Qa1 Qd5 40 Nd2 a4 41 Re5 Qc6 42 Rg5 Rd7 43 Qe5 h6 44 Qb8+ Kh7
45 Qe8 Rc7 46 Qg6+ Kh8 47 Nf3 Ne4 48 Ne5
Or 48 Re5 d2 49 Rxe6 Qxe6+ 50 Qxe6 Nf2+ 51 Kg2 d1Q 52 Kxf2 Rc2+ etc.
48 ... Nf2+ 0-1
White resigned just before 49 Kh2 Qh1 mate.
Index of Complete Games
Addison.W-Bisguier.A, US Championship, New York 1962
Anand.V-Kramnik.V, London Classic 2016
Atalik.S-Wawrzaszek.J, Stillwater 2011
Ayupov.D-Savchenko.B, Russian Team Championship 2004
Bacrot.E-Karpov.A, Cannes (1st matchgame) 2000
Badea.B-Lputian.S, Porto San Giorgio 1999
Balogh.Im-Georgiev.Vl., European Championship, Plovdiv 2008
Bareev.E-Asrian.K, New York Open 1998
Bareev.E-Bönsch.U, German League 2000
Bocharov.D-Azarov.S, European Championship (play-off), Warsaw 2005
Breutigam.M-Rabiega.R, German Championship, Altenkirchen 2005
Browne.W-Balashov.Y, Taxco Interzonal 1985
Brynell.S-Gurevich.M, Cappelle la Grande 2011
Bu Xiangzhi-Rowson.J, Turin Olympiad 2006
Buhmann.R-Vaganian.R, German League 2004
Bürger.C-Klovans.J, Werfen 1996
Burn.A-Showalter.J, England vs. USA, cable match 1898
Davies.N-Lputian.S, Isle of Man 2003
Esen.B-Korobov.A, Aeroflot Open, Moscow 2015
Garibaldi.O-Pilnik.H, Argentine Championship, Buenos Aires 1946
Gelfand.B-Piket.J, Wijk aan Zee 1998
Gershon.A-Asrian.K, World Junior Championships, Yerevan 1999
Grabarczyk.M-Balashov.Y, German League 1997
Graf.A-Bönsch.U, German League 2005
Graf.A-Yusupov.A, German Championship, Altenkirchen 2005
Grooten.H-Kuijf.M, Eindhoven 1981
Hertneck.G-Ivanchuk.V, Munich 1994
Hjartarson.J-Short.N, Dubai Olympiad 1986
Hundack.R-Stern.R, Bad Zwischenahn 2012
Inkiov.V-Antonov.Vl, Pernik 1978
Ivanisevic.I-Gurevich.M, European Team Championship, Heraklio 2007
Jaeger.F-Gurevich.M, Tromsø 2007
Janowski.D-Capablanca.J, New York 1924
Kaasen.T-Hansen.S, Lund 2016
Karpov.A-Campora.D, Villarrobledo (rapid) 1997
Kostic.Br-Cvetkovic.Sr, Belgrade 2009
Kottnauer.C-Euwe.M, Amsterdam 1950
Lahiri.A-Irwanto.S, Kuala Lumpur 2005
Li Wenliang-Yusupov.A, Minneapolis 2005
Liang Chong-Wang Chen, Chinese Team Championship 2010
Loxine.J-Atalik.S, Dresden 2005
Lputian.S-Balashov.Y, Petrosian Memorial, Yerevan 1986
Lutz.C-Yusupov.A, Tilburg 1993
Lytchak.A-Bönsch.U, German League 2005
Mastrovasilis.A-Gurevich.M, European Championship, Dresden 2007
Mozny.M-Dumitrache.D, Odorheiu Secuiesc 1995
Munkhgal.G-Ehlvest.J, Jakarta 2013
Naiditsch.A-Pavasovic.D, European Championship, Rijeka 2010
Navara.D-Hansen.S, European Team Championship, Reykjavik 2015
Peelen.P-Van der Sterren.P, OHRA-B, Amsterdam 1989
Pelletier.Y-Gurevich.M, Gibraltar 2006
Polgar.Zsu-Spassky.B, Women vs. Veterans, Prague 1995
Popov.Val-Ziatdinov.R, Voronezh 2004
Quan Zhe-Spraggett.K, Canadian Championship, Toronto 2004
Radjabov.T-Carlsen.Ma, Bazna Kings, Medias 2011
Radovanovic.D-Savanovic.A, Bosnian Team Championship 2012
Robatsch.K-Cvetkovic.Sr, Trnava 1988
Sakaev.K-Asrian.K, Dubai 2000
Sokolov.I-Asrian.K, Stepanakert 2005
Spangenberg.H-Ricardi.P, Trelew 1995
Speelman.J-Andersson.U, GMA World Cup, Reykjavik 1991
Sturua.Z-Kharitonov.A, USSR Young Masters, Tallinn 1981
Sukharisingh.R-Vaganian.R, German League 1996
Timman.J-Hübner.R, Teesside 1975
Timman.J-Yusupov.A, Tal Memorial, Riga 1995
Trajkovic.P-Maksimovic.B, Serbian Championship, Kragujevac 2000
Tratatovici.D-Zilberman.Y, Kiryat Ono 2016
Ubilava.E-Gurevich.M, Spanish Team Championship 2007
Vaganian.R-Rogozenco.D, German League 1996
Williams.Lei-Dervishi.E, Bled Olympiad 2002
Yusupov.A-Goloshchapov.A, German League 2008
Zhu Chen-Spassky.B, Women vs. Veterans, Marbella 1999