Sunteți pe pagina 1din 384


Title Page
Foreword by the Publisher 5
Introduction 6

Part 1. 1.d4 d5 2.c4 dxc4 3.e4 9

1. 3.e4 Nc6 10
Main Ideas Step by Step Annotated Games
2. The QGA la Chigorin 37
Main Ideas Step by Step Annotated Games
3. 3.e4 b5?! 51
Main Ideas Step by Step Annotated Games
Part 2. The Classical System 63
4. The Exchange Variation 64
Main Ideas Step by Step Annotated Games
5. The 7.Bb3 Variation 87
Main Ideas Step by Step Annotated Games
6. The 7.Bd3 Variation 97
Main Ideas Step by Step Annotated Games
7. The Old Main Line 107
Main Ideas Step by Step Annotated Games
8. The 7.a4 Variation 121
Main Ideas Step by Step Annotated Games
9. The 7.b3 Variation 143
Main Ideas Step by Step Annotated Games
10. The 7.e4 Variation 149
Main Ideas Step by Step Annotated Games
11. The Furman Variation 155
Main Ideas Step by Step Annotated Games
Part 3. Deviations From The Classical System 163
12. The Gambit Line 164
Main Ideas Step by Step Annotated Games
13. 4.Qa4+ 175
Main Ideas Step by Step Annotated Games
14. Rare Variations 181
Step by Step
Part 4. Alternative Repertoires 185
15. 3.e3 186
Main Ideas Step by Step Annotated Games
16. The 4...Bg4 Variation 201
Main Ideas Step by Step Annotated Games
17. 1.d4 d5 2.c4 dxc4 3.Nf3 a6 219
Main Ideas Step by Step Annotated Games
Ods and Ends 227
18. 1.d4 d5 without 2.c4 228
Main Ideas Step by Step Annotated Games

Index of Variations
Understanding the QGA
A Black Repertoire with 1.d4 d5 2.c4 dxc4

Alexander Delchev
Semko Semkov

Chess Stars Publishing

Copyright 2015 by Alexander Delchev and Semko Semkov

Cover by Kalojan Nachev, Rustam Taichinov

ISBN: 978-619-7188-05-9



Playing 1.d4 - The Queens Gambit, Lars Schandorff, Quality Chess 2012
Wojos Weapons, Jonathan Hilton and Dean Ippolito, Mongoose Press 2010
The Queens Gambit Accepted, Konstantin Sakaev and Semko Semkov, Chess Stars 2008
Grandmaster Repertoire 1 1.d4 Volume one, Boris Avrukh, Quality Chess 2008
The Chigorin Defence According to Morozevich, Alexander Morozevich and Vladimir Barsky, New in Chess


Chess Informant
New in Chess

Internet resources

The Week In Chess (
10 Days (
Internet Chess Club (
Chess Publishing (
Chess Today
Foreword by the Publisher

Is the Kings Indian better than the Nimzo? And what about the Benko?!
My mailbox is full with such questions and I always try to explain that the choice of opening repertoire should depend
on the players tastes and needs. For instance, I think that 1.Nf3 would be a bad move for a beginner. Open games with
empty or mobile pawn centres put the accent on piece play. They demand detailed calculation and help refining
endgame technique.

The QGA is in the same category. It is an excellent opening if you want to learn to play chess better. The overwhelming
majority of games is decided in deep middlegame or endgame. The complexity of theory is not even close to mammoth
openings as the KID or the Grnfeld where one need several volumes and the corresponding memory!
At the same time, modern QGA is positionally sound and reliable opening, played by world champions Kasparov,
Karpov, Anand, Topalov. Its only drawback is that White could choose ultra solid drawing lines.

I solved this problem by inviting GM Delchev to share his personal repertoire which includes the sharp and challenging
Alekhine Variation based on ...Bg4. That allowed us to present a multifaceted approach with a wide choice for Black
ranging from the blunt drawish line 1.d4 d5 2.c4 dxc4 3.Nf3 a6!? 4.e3 b5 to the pawn sacrifice 3...Nf6 4.e3 Bg4
without ...a6!

It so happens that I have been studying the QGA all my life. That made our collaboration with Delchev an easy task and
resulted in our third mutual work after The Most Flexible Sicilian and The Safest Sicilian.

Semko Semko
August 2015

In 2009 during a teamwork with GM Artur Kogan, he briefly showed me the line 1.d4 d5 2.c4 dxc4 3.Nf3 Nf6 4.e3
Bg4, together with his analysed games. I found it promising and attempted to build up a complete repertoire. It turned
out, however, that I faced a serious problem as early as on move 6. After 5.Bxc4 e6 6.Nc3!,

White was threatening to capture our bishop with h3, g4, Ne5. Artur claimed that best was 6...Nbd7, but then 7.0-0!
Bd6 8.h3 Bh5 9.e4 e5 10.g4! was well known and definitely unpleasant. I tried to improve with 9...Bxf3 against
Swiercz, but got an extremely murky position. It became clear that it was a dead end. 6...a6 7.h3 Bh5 8.g4 Bg6 9.Ne5
was not appealing either.

Thus by the natural method of exclusion I reached to the idea of 6...Nc6. I found in the database a game of Spassky of
1961 where he met the critical 7.Bb5 by the weak 7...Bb4?!. Only a quarter of century later did Black come with an

Kinsman-Brynell, 1988, saw 7.Bb5 Bd6 8.e4 Nd7 9.Be3 0-0 10.0-0 Nb6 11.Be2 f5 and Gausel-Brynell, 1999,
introduced the pawn sacrifice 7.h3 Bh5 8.Bb5 Bd6 9.Qa4 Bxf3 10.gxf3 0-0. On the ground of these games, I developed
the main directions and managed to build a detailed repertoire. My pupils were satisfied with the new weapon and soon
I also had a chance to test it against GM Edouard (an ardent QGA protagonist himself!) see the annotated game 34
Edouard-Delchev, Linares 2013.
My compensation for the pawn is obvious and it grew up in a rout: 18...Nxf4 19.Nc3 Nh3+ 20.Kg2 Qc4 21.f4 g5
22.Qd1 Rxb2+ 23.Bxb2 Rxb2+ 24.Kxh3 Qxc3 25.Qg4 Qxe3+ 26.Qg3 Bxf4 27.Qxe3 Rxh2+ 28.Kg4 h5+ 29.Kf3
Rh3+ 0-1
The variation got its first taste of gunpowder and the ball was in Whites court (it is still there!).

The Classical System with 4...e6 has never really attracted me due to Whites possibility to play with a draw in the
pocket in the Exchange Variation 7.dxc5. Here is what happened a couple of days ago in Kavala Open, August 9th
Rychagov 2552-Delchev 2604
1.d4 d5 2.c4 dxc4 3.Nf3 Nf6 4.e3 e6 5.Bxc4 c5 6.0-0 a6 7.dxc5 Bxc5 8.Qxd8+ Kxd8 9.Be2 draw.
When playing opens against 100-150 Elo points lower rated opponents, such a line should be out of question. On the
other hand, a solid endgame without risk could be a fair choice against a decent player.

I would not like to impose my tastes so I provide the following alternative repertoires:

The first part of the book is devoted on Whites most challenging option against the QGA 3.e4. You should learn

1. Chapter 1, which deals with 3...Nc6 4.Nf3 Bg4, or

2. Chapter 2, which is my recommendation. It covers 3...Nc6 4.Nf3 Nf6!? and offers extremely interesting play la
Chigorin Defence.

I also analyse 3...b5?!, mostly to show Blacks problems in that topical line.

If White plays 3.e3, we have the following choice:

1. 3...Nf6 followed up by the Classical System.

2. 3...e5!? my recommendation see Chapter 15.

After 3.Nf3, we can choose:

1. 3...Nf6 the Classical System.

2. 3...a6 with a draw in mind see Chapter 17.

3. 3...Nf6 4.e3 Bg4!? which is my weapon of choice in most games.

If you stop your attention on options 2 or 3, you can safely skip the whole part about the Classical System.

Chapter 18 offers advice against Queens Pawn Openings.

I have written this book in first person, but it is a collaborative work.

Every line has been checked and discussed by both authors.

Alexander Delchev
August 2015
Part 1
1.d4 d5 2.c4 dxc4 3.e4

This is the most challenging system against the QGA and the first thing you should learn. Its importance made me offer
several alternatives for Black. I recommend 3...Nc6 4.Nf3 Nf6!? see Chapter 2. This move order transposes to a very
interesting line from the Chigorin Defence. It is strategically unbalanced, but easy to play with Black as he does not
have any weaknesses while Whites pawn structure is compromised and often overextended. I do not see any
theoretical problems so you can use it even against well prepared opponents. Another plus is that the variations are not
forced and you will not be risking to face a groundbreaking novelty.

The more conventional 3...Nc6 4.Nf3 Bg4 is the subject of Chapter 1. I offer reliable defences against everything White
has tried so far. However, I also discuss new approaches for White which might give him a pleasant position without
much counterplay for Black. Also note that in one line White has a forced draw. On the positive side, we might trick
the opponent with the move order, and in the most popular line for White, based on the game Karpov-Milov, we can
even get a self-playing attack. So, if you hope to capitalise on your better home preparation, Chapter 1 will give you
plenty of advice.

Finally, I pay a tribute to fashion and devote Chapter 3 on the super-topical 3.e4 b5, but mostly in order to discourage
you from playing it at a higher level. Black has had his day or two of glory, but I expect murky days for him in future.
There are two many lines where his limit is a forced draw. Or a struggle for the draw.

Chapter 1. 3.e4 Nc6

Main Ideas

1.d4 d5 2.c4 dxc4 3.e4 Nc6!?

This development is alien to the spirit of the QGA, but it does lead to very complex and strategically unbalanced
positions with considerable winning chances... for both sides, to be sure. Instead of unloading the centre, Black hopes to
prove that it is overextended. His first goal is to double the f-pawns with ...Bc8-g4xf3 or provoke d4-d5. Then ...e7-e5
would block the centre while ...e7-e6 would try to undermine it.

Theoretical status

It strikes me that all the authors who write on this system seem clearly uncomfortable in Whites shoes. Lets consider
their recommendations.

Lars Schandorff gives in his White repertoire book Playing 1.d4 The Queens Gambit 4.Nf3 Bg4 5.Bxc4 e6
He offers as a backup line 6.d5 exd5 7.Bxd5 Qf6 8.Nc3 when 8...Nge7 9.Bxc6+ Qxc6 evens the game.
6...Bxf3 7.gxf3 Qf6 (Black can also transpose to the Chigorin with 7...Nf6!? 8.Nc3 Bb4 see Chapter 2). White is in
danger, but Schandorff points out that he can still make a draw with 8.Nc3! 0-0-0 9.Qa4 (9.d5 Bb4) 9...Qxf3 10.Rg1
Bb4 11.d5 exd5 12.exd5 Re8 13.dxc6 Rxe3+ 14.Kf1 Bc5 15.Rg3 (15.Rg2) 15...Qh1+ 16.Rg1.
As we see, there is nothing to worry about so far.

Many commentators rely on Karpovs erudition and claim a small advantage for White on the ground of his widely
cited game against Milov. It continued 4.Nf3 Bg4 5.d5 Ne5 6.Bf4 Ng6 7.Be3 e5 8.Bxc4 a6 9.Nc3 Nf6 10.0-0 Bd6
11.Be2 0-0 12.Nd2 Bd7 (Keep this bishop! We need it for an attack see game 1 Anand-Vallejo Pons,
Bilbao 2014.) 13.Rc1 Qe7 14.a3 b5 15.Nb3 Nf4 16.Bf3,

when instead of 16...Kh8? which allowed 17.Na2! (heading for c6), Black should have immediately launched the g-
pawn: 16...g5 17.Na5 Kh8! 18.b4 Rg8

Black is aiming to bring his heavy pieces to the h-file with ...R8-g6-h6 or ...Qe7-f8-h6 followed by ...Rg8-g5-h4. The
key point is that Black can ignore the threat g2-g3xf4 as his threats on the h-file will provide him with adequate
compensation. I consider this set-up in game 5 Miton-Berzinsh, Warsaw 2013 and in game 4 Olszewski-Mista,
Warsaw 2011.
On a final note, I would add that in the first of these two annotated games Black revealed his plan even earlier:
13...Nf4!? to take over the initiative after 14.Bf3? g5 15.Nc4 Kh8! 16.a3 Rg8.
The bottom line is that Karpovs stand offers Black a free attack. It suffices to put a knight on f4, bolster it with ...g5
and never budge from there (except for eating the white bishop on e2 if it is available).

Problems and critical lines

Lets return to the position after 12...Bd7.

White has the better centre. His pawn on d5 makes the difference on the queenside. The d2-knight can follow the route
Nc4-a5 or Nb3-c5(a5). The c3-knight could go to a2-b4 or d1-e3. Both bishops on e2 and e3 support the knights stabs.
In short, Whites game is playing itself. Blacks only counterplay is based on ...Nf4 followed up by ...g5. If White
overcame the spell of Karpovs name and thought logically, he could find the very unpleasant move 13.g3!. It
effectively dissipates our hopes for an attack and brings us down to prose. Although this line is practically nexplored, I
believe that 13...Qe7 14.Rc1 b5 15.a3! (a new move, which aims to block the queenside and take control of c5) retains
some initiative.

Of course, Black can enter a symmetrical position with 15...c5 16.dxc6 Bxc6 17.b4! Rfd8 and gradually exchange most
pieces to draw, but I do not like such static positions.

I use another approach to avoid the abovementioned line:

4.Nf3 Bg4 5.d5 Ne5 6.Bf4 Ng6 7.Be3 Nf6!?
Now 8.Nc3 e6 introduces a totally different approach to the centre. Instead of blocking it, Black destroys it. Critical
here is 9.Qa4+ Qd7 10.Qxd7+ Kxd7! 11.Ng5! (Karpov again!) 11...exd5 12.Nxf7 Rg8 13.f3 Be6 14.Ng5 Bb4!

Game 3 Sakaev-Yakovich, Moscow 2009 saw further 15.0-0-0 Bxc3 16.bxc3 b5 with a tangled and very interesting
position. A sterner test of Yakovichs idea, according to my analysis, might be the line 15.Nxe6 Kxe6 16.exd5+ Kd6
17.Bxc4 Rae8 18.Kd2 with a sharp nontrivial endgame.

The move order with 7...Nf6 is also effective against 8.Nbd2 when 8...e6! is strong.
However, Tomashevsky pinpointed against me in 2013 a downside of Blacks seventh move. He sacrificed a pawn by
8.Bxc4!? when 8...Nxe4 is practically the only decent answer since 8...a6?! 9.Nbd2!? e5? 10.dxe6 would be horrible.
After 9.h3 Bd7 10.0-0, I should have probably forced play with 10...e5! 11.dxe6 Bxe6 12.Qc2!? Bxc4 13.Qxe4+ Be6
This position also needs practical testing.
For completeness sake, I will also mention the retreat to g3:
4.Nf3 Bg4 5.d5 Ne5 6.Bf4 Ng6 7.Bg3. Then 7...e5 is the only move but this version of the static structure should not be
a problem for us since Whites bishop is badly missing from the g1-a7 diagonal. We can exploit that by advancing
pawns on the queenside. Here is an example:

Moscow 2002


In conclusion, 4.Nf3 Bg4 is playable, but the Chigorin with 4...Nf6 offers much sharper play. I examine it in Chapter 2.

Part 1. 3.e4 Nc6.

Step by Step

1.d4 d5 2.c4 dxc4 3.e4 Nc6!?

This development is alien to the spirit of the QGA, but it does lead to very complex and strategically unbalanced
positions with considerable winning chances... for both sides, to be sure. Instead of unloading the centre, Black hopes to
prove that it is overextended. The first goal is to double the f-pawns with ...Bc8-g4xf3 or provoke d4-d5. Then ...e7-
e6xd5 would devalue Whites pawn centre and would give us an easy and natural development.


Of course, White may refrain from 4.Nf3, but 4.Be3 and 4.d5 (an attempt to sidestep the Chigorin which arises after
4.Nf3 Nf6) do not pose substantial problems. Lets see:
a) After 4.d5 Ne5, we can get a good version of the main plan with ...e6. The key point is that we have not played ...Bg4
so we are not to fear Qa4+ which is the most challenging retort to ...e6 in line A. For instance:
Minor alternatives are:
5.Nf3 Bg4;
5.f4?! Nd3+ (5...e6 6.fxe5 Qh4+ 7.Kd2 Qf4+ is only a repetition) 6.Bxd3 cxd3 7.Qxd3 c6;
5.Nc3 e6!;
5.Qd4 Ng6! (do not develop the enemy by 5...Nd3+. He gets an initiative after 6.Bxd3 cxd3 7.Nc3 c6 8.Bf4 Nf6
9.0-0-0) 6.Nf3 e5 7.Qxc4 (7.Nxe5? drops a piece to 7...Bb4+ 8.Nc3 c5. I encountered once 7.dxe6? Bxe6 8.Nc3
Nf6 9.Bg5 when simplest was 9...c6.)
White counts on seizing space on the kingside with h2-h4-h5 which could be embarrassing after 7...Nf6. It is
better to prepare a retreat on e7 with 7...Bd6. Black took over the initiative in the game Goldin-Morozevich, St
Petersburg 1993, following 8.Nc3 Nf6 9.h4 h6 10.h5 Ne7 11.Be3 0-0 12.Be2 a6 13.Nd2 c6.
5...Ng6 6.Bg3
6.Be3 Nf6 7.Nc3 e6! opens the e-file in Blacks favour: 8.Bxc4 exd5 9.exd5 Bd6 10.h3 0-0 11.Nf3 a6 (11...Qe7
12.0-0 Ne5= Petkov-Godena, Warsaw 2013) 12.0-0 b5 13.Bb3 Re8
It is good to gain space on the kingside before switching to the thematic plan with ...e6.
6...e5 leads to the pawn structure of line A (see below): 7.Bxc4 a6! 8.Nc3 Bd6 9.Nf3 Nf6.
7.h3 h4 8.Bh2 and now 8...e6 9.Bxc4 Nf6 10.Nc3 exd5 11.Bxd5!? Bd6=.

b) 4.Be3 Nf6 5.Nc3

5.e5 Bg4 6.Nf3 Nd5 gives Black comfortable development.
5.f3 does not look inspiring as besides 5...e5 6.d5 Nd4 7.Bxd4 exd4 8.Qxd4 c6 which offers sufficient
compensation for the pawn, Black also has 6...Na5!? 7.Qa4+ c6 when 8.Bd2 b6 or 8.dxc6 Nxc6 9.Bxc4 Bd7 are
clearly harmless so White should take the piece and struggle after 8.b4 b5 9.Qxa5 Qxa5 10.bxa5 cxd5 11.Nc3 d4
12.Nxb5 dxe3 13.Nc7+ Kd7 14.Nxa8 Bb4+ 15.Kd1 Ba6 16.Rc1 Rc8 17.Nb6+ axb6 18.axb6 Bd2.
5...e5 6.d5 Na5
The insertion of 7.Nf3 Bd6 enables the nice piece sac 8.Qa4+ Bd7!? 9.Qxa5 a6!! 10.Na4 Qe7! with great
compensation, for instance: 11.a3 Nxe4 12.Bxc4 0-0.
7...Bd7 is also possible, but not so strong as in the previous example.
8.dxc6 Nxc6 9.Bxc4 Bb4 10.Nf3 0-0=.

Lets now return to the most topical line 4.Nf3:

We have three possible approaches here. The most popular one (line A) is to force d4-d5 and block the centre with ...e5.
More testing is the plan with ...e7-e6xd5. I consider it as line B. Finally, we can further unbalance the position with
4...Nf6!? which is the subject of Chapter 2.
Note that in the 4...Bg4 line Black cannot avoid the static structure with ...e7-e5 if White retreated his bishop to g3 (after
5.d5 Ne5 6.Bf4 Ng6) when 7.Bg3 e6 would be dubious. On the other hand, at g3 the bishop is clearly more passive
than on e3 so Black should not be too concerned about this rare line.

4...Bg4 5.d5

a) 5.Bxc4 e6 is comfortable for Black. Schandorf chooses it for his White repertoire (Playing 1.d4 - The Queens
Gambit), but he does not offer sensible arguments in Whites favour. Play continues:
6.Bb5 Bxf3! (It is always good to damage the enemys kingside pawn structure) 7.Bxc6+ (7.gxf3 Nge7!? 8.Bg5
Qd6) 7...bxc6 8.gxf3 Bb4+! (Semkov) 9.Nc3 Ne7 is double-edged. Whites king does not have a safe haven.
6...exd5. Simple and straightforward.
6...Bxf3 7.Qxf3 Ne5 8.Qc3! offers White an easier endgame. (8.Bb5+ c6 9.Qc3 cxb5 10.Qxe5 Qd6 11.Qxd6
Bxd6 12.Nc3 a6. White scores modest 25% in this balanced endgame. Epishin-Sengupta, Leiden 2013,
continued 13.dxe6 fxe6 14.Bd2 Nf6 15.f3 Rc8.) 8...Nxc4 9.Qxc4 exd5 10.Qb5+ c6 11.Qxb7 Qc8 12.Qxc8+
Rxc8 13.exd5 cxd5 14.0-0 Bb4. It is possible that this position could be analysed to a draw, e.g. 15.Be3 Rc7
16.Rd1 Ne7.
7.Bxd5 Qf6 8.Nc3 Nge7!

Black falls behind in development, but I did not find a way for White to exploit that: 9.Bxc6+ Qxc6 10.0-0 (10.h3 Rd8
11.Bd2 Bxf3 12.Qxf3 Ng6 13.Nd5 Bd6) 10...Rd8 11.Qb3 Bxf3 12.gxf3 Ng6 13.Nd5 Bd6.

b) 5.Be3 Bxf3 6.gxf3 e6 7.Bxc4 Qf6!

Threatening ...0-0-0. White is in danger, but he can still make a draw with 8.Nc3!
8.e5?! Qh4 9.Bb5 Bb4+ 10.Nc3 Nge7;
8.Bb5?! Bb4+ 9.Nc3 Nge7 10.a3 Ba5 and the bishop goes to b6.
8...0-0-0 9.Qa4 (9.d5 Bb4) 9...Qxf3 10.Rg1 Bb4 11.d5 exd5 12.exd5 Re8 13.dxc6 Rxe3+ 14.Kf1 Bc5 15.Rg3 (15.Rg2)
15...Qh1+ 16.Rg1 Qf3 17.Rg3 draw, Ding,Liren-Timofeev, Moscow 2011.
Black can escape the draw by transposing to the Chigorin with 7...Nf6!? 8.Nc3 Bb4 see Chapter 2.

5...Ne5 6.Bf4

6.Qd4?! is a tricky move which does not yield White enough compensation after 6...Nxf3+ 7.gxf3 Bxf3 8.Rg1
8.Bxc4 e5! is the simplest way to deflect the attack (8...a6!? 9.Rg1 is unexplored. Then 9...e5 10.Qxe5+ Qe7 is
balanced, but 9...c5!? 10.Qd3 Bh5 may be a sterner test of Whites pawn sacrifice.) 9.Bb5+ c6 10.Qxe5+ Be7
11.Rg1 cxb5 12.Nd2 f6 13.Qf5 Bxe4 14.Nxe4 g6 15.Qe6 Qd7.
8...e5 9.Qxc4 a6. White cannot regain the pawn by 10.Rg5 due to 10...Qd6 although 10...h6 11.Rxe5+ Be7 is not bad at

6.Nbd2 e6! 7.Qa4+ Qd7 8.Qxd7+ Nxd7 9.dxe6 Bxe6 10.Nd4 Ne5 11.Nxe6 fxe6 is equal.

Do not give free tempi with 6...Nd3+ or 6...Nxf3+. 6...Bxf3?! 7.gxf3 Qd6 looks strategically sound, but then 8.Bxe5
Qxe5 9.Qa4+ c6 10.Nc3! is awkward.

Main branches now are: A. 7.Bg3; B. 7.Be3

7.h3 Bxf3 8.Qxf3 e5! 9.dxe6 fxe6 10.Nc3 Qf6! 11.Bg3 Qxf3 leads to a typical endgame for this line. Blacks play is
natural and easy.

A. 7.Bg3

This seems to be coming into fashion. It is directed against ...Nh4. Another point in its favour is that Black does not
have the plan with 7...Nf6 8.Nc3 e6?! when 9.Qa4+ Qd7 10.Qxd7+ Kxd7 11.dxe6+ Bxe6 12.0-0-0+ favours White.
This leaves us with:

7...Nf6 8.Nc3 e5

Note that the inclusion of 7...Nf6 8.Nc3 is essential. Black needs to parry the check on the a4-e8 diagonal with ...Nf6-
d7. The immediate 7...e5 8.Bxc4 Bd6 allows 9.Bb5+! which disturbs Blacks harmonious development. Then 9...Kf8
10.Be2 Nf6 (10...h5 11.h3 Bd7 12.Nbd2) 11.Nbd2 h6 12.0-0 Kg8 13.Qb3 offers White the better game. Trading
light-squared bishops with 9...Bd7 10.Bxd7 is a more common reply, but practical experience has seen Black struggling
as he lacks counterplay. See the similar game 1 Anand-Vallejo Pons, Bilbao 2014 for an example. 8...a6 (instead of
8...Bd6) is not an improvement due to 9.h3 and the double attack on e5 forces Black to part with his light-squared
bishop. We see here the impact of the g3-bishop on the centre.

9.dxe6 Qxd1+ 10.Rxd1 fxe6! (10...Bxe6 11.Nd4 Bb4 12.f3) 11.Bxc4 Bb4 12.0-0 Ke7 is equal. Well trade our only bad
piece on f3 after 13.Rc1 (13.Rd3 Rad8=) 13...Bxf3 14.gxf3 c6=, remaining with active knights and a centralised king.

9...Bd6 10.0-0

After 10.h3 Bd7! (Of course we must preserve our good bishop. This should be a rule of thumb, but I was surprised to
see that both available games in my database featured 10...Bxf3?!), White will have sooner or later to play f3 thus
making an ugly hole on g3. Moreover, the h3-pawn might become a lever for an attack.
The checks are rather harmless:
10.Qa4+!? Nd7 11.Nd2 0-0 12.Qc2 Nf4 with mutual chances.
10.Bb5+ Nd7 11.h3 Bxf3 12.Qxf3 0-0!. It might appear that 13.Bxd7 Qxd7 14.0-0 accomplishes Whites strategic aim
of trading light-squared bishops. However, that has come at a price. He lost tempi which allows Black to take over the
initiative with 14...f5.

10...0-0 11.Be2 a6 12.Nd2 Bd7

Well see in line B that White is slightly better in this structure with his bishop standing on e3. From g3, it does not
control the important g1-a7 diagonal and is practically out of play.


If White keeps the tension with 13.Rc1 b5 14.f3, Black can turn his attention to the kingside with 14...Nf4 15.Bf2 N6h5.
In this position f3 undoubtedly weakens Whites kings pawn shelter.
After the text, the natural retort is:

13...b5 14.Nxd6

14.Nd2 c6 or 14...c5

14...cxd6 15.a3 a5

Blacks queen will find a perfect stand on b6 as if to underline the absence of the enemys dark-squared bishop from e3:

16.b4 Qb6 17.Qd2 Rfc8=

B. 7.Be3 e5

A difficult decision. It is very tempting to insert 7...Nf6!?. Then 8.Nc3 e5 would transpose to the main line having
dodged the set-up with Nbd2. Black has an interesting alternative though 8...e6!?
The pawn structure after 9.Bxc4 exd5 10.exd5 a6 11.Be2 Bd6 is excellent for Black, and 9.dxe6 Bxe6 10.Nd4 Ng4
11.Nxe6 Qxd1+ 12.Rxd1 fxe6 13.Bxc4 Nxe3 is completely equal. Therefore, White must seek concrete play with
9.Qa4+ Qd7 10.Qxd7+ Kxd7! when a tangled position with mutual chances arose in game 3 Sakaev-Yakovich,
Moscow 2009. However, I point out to an improvement of Whites play which could offer him a slightly better
The same plan is good after 7...Nf6 8.Nbd2 8...e6 9.Qa4+
Or 9.dxe6 Bxe6 10.Nd4 Bd7 11.Bxc4 Bd6 12.0-0 (12.f3 0-0 13.0-0 c5; 12.Nf5 Bxf5 13.exf5 Ne5) 12...Ng4=.
9...Qd7 10.Qxd7+ Kxd7 (10...Nxd7 11.Bxc4 exd5 12.Bxd5 c6 13.Bb3 Bc5=) 11.Bxc4 exd5 12.exd5, Fressinet-
Delchev, Porto Carras 2011, 12...Bd6=.

Critical for 7...Nf6 is the cunning pawn sac 8.Bxc4!

If we attempted to transpose by 8...a6?! 9.Nc3 e6?, then 10.Qb3 b5 11.dxe6 would be awful.
Thus we are left only with 9...e5 and even this may not be a simple transposition since 10.dxe6 Qxd1+ 11.Rxd1 fxe6
12.h3 Bxf3 13.gxf3 e5 is solid, but not easy to play with Black. At first glance White does not have any target, but if we
dug deeper, we could notice that the set-up Rg1, Ne2-g3-f5 or Be6-f5 maintains some pull. Black should be able to
defend this endgame, but he can hardly hope to win.
No one has ever played this as White, but the last nail in Blacks coffin was my discovery that after 8...a6?! the first
player has 9.Nbd2!? eliminating the queens trade from the above variation. Then 9...e5? 10.dxe6 would be just
Evidently, we must accept the gift 8...Nxe4 9.h3 Bd7 10.0-0

In Tomashevsky-Delchev, Struga 2013, I braced myself for a dogged defence and went on to struggle after
10...Nd6 (10...c6 11.Nc3 Nd6 12.Bb3 cxd5 13.Nxd5 e6 14.Re1) 11.Bb3 Nf5 12.Nc3 e5 13.dxe6 fxe6 14.Ne4
Be7 15.Nd4!. White regains the pawn remaining with the better pieces.
It is better to return the pawn at once:
10...e5! 11.dxe6 Bxe6 12.Qc2!?
It should not be a problem to neutralise Whites initiative without queens, e.g. 12.Bxe6 Qxd1 13.Bxf7+ Kxf7
14.Rxd1 Be7 15.Rd4 Nc5 16.Nc3 Ne6 17.Re4 Bf6 18.Nd5 Rad8 19.Nxf6 gxf6=.
12...Bxc4 13.Qxe4+ Be6
White will capture on b7 to remain with the better pawn structure on the queenside. On the other hand, Black succeeds
in completing development. In my opinion, Black has enough counterplay here. For instance:
a) 14.Qxb7 Bd6 15.Nd4 Rb8 16.Qe4 Qd7 17.Nxe6 Qxe6 18.Nd2 Qxe4 19.Nxe4 Rxb2;
b) 14.Nc3 Bd6 15.Nd4 Qd7 16.Qxb7 0-0.

Another topical alternative is:

7...a6 (or 7...e5 8.Bxc4 Nh4 9.Rg1 a6) 8.Bxc4 Nh4 9.Rg1 (9.0-0) 9...e5 10.Nbd2. It has been tested in 3 games. The fact
that White has lost the right of short castling does not change substantially the strategic canvas. Perhaps his best plan is
to leave his king on the kingside and play in the centre as in the game Goganov-Pridorozhni, Khanty-Mansiysk 2014:
10...Ng6 (10...Bd6 11.Qb3 b5 12.Be2 Bxf3 13.Nxf3 Nxf3+ 14.Bxf3) 11.Qb3 b5 12.Be2 Bd7 13.h4 h5 14.g3 Nf6
15.Ng5 Bd6 16.f3 Qe7 17.Kf2 0-0 18.Rac1 Rfc8 19.Rc2 c6 20.dxc6 Rxc6 21.Rxc6 Bxc6 22.Rc1.

8.Bxc4 a6

8...Bd6 9.Bb5+ Bd7 is quite solid and it is played often. However, trading the light-squared bishops is a serious strategic
achievement of White since it reduces the enemys attacking resources. Lets be realistic. What we have on the
kingside? A symmetrical pawn structure with e4-pawn vs. the e5-pawn. The extra-knight on g6 is neutralised by the e3-
bishop. On the opposite flank White has a clear space advantage (d5-pawn vs. c7) and a semi-open c-file. The maximum
Black could get is to achieve ...c7-c6, but even then his pieces will be passive. See Game 1 Anand-Vallejo Pons, Bilbao
2014, as an instructive example.

It is more challenging to keep the bishops with 9...Kf8, but the simple 10.g3 (well see that this prophylaxis effectively
neutralises Blacks counterplay in the main line, too) should be enough to scorch Blacks hopes for a kingside attack.
My game Delchev-Stevic, Zadar 2007, continued 10...Nf6 11.Nc3 (11.Nbd2! is even better) 11...a6 12.Be2 b5 13.a3
Bd7 14.b4 Qe7 15.h4 Kg8 16.h5.

8...Nf6!? 9.Nc3 a6 would transpose since 9.Bb5+ could be met by 9...c6.

Only 9.Nc3 Nh4 would have an independent significance, but I have discussed this idea before. In short, White
can answer 10.Rg1, then leave his king on the kingside. The computer likes 10.0-0, but then 10...Nxf3+ 11.gxf3
Bh3 12.Re1 Bd6 13.Kh1 a6 14.Rg1 g6 would be unclear.


9.Nc3 looks somehow unfocused. At best, White will transpose to the main line after 9...Bd6 10.Be2 Nf6, but he also
allows 9...Nh4.
9.Nbd2 is a natural, but rare alternative. See Game 2 M.Rodshtein-Raetsky, Biel 2012 for details.

9...Nf6 10.Nc3

Dreev played 10.Nfd2 to transpose after 10...Bd7! 11.Nc3.

Or 10.Nbd2 Bd6 as noted above.

10...Bd6 11.0-0

Or 11.Nfd2 Bd7

11...0-0 12.Nd2 Bd7

This set-up is strategically risky. It is too static and the positional factors favour White: he has the better centre, more
space, and a semi-open c-file. His long-term prospects on the queenside are excellent. The d2-knight follows the route
Nc4-a5 or Nb3-c5(a5). The c3-knight could go to a2-b4 or d1-e3. In short, Whites game is playing itself. This scenario
is discussed in line B3.

Still, many Black players happily enter this line, hoping for a kingside attack. It is tempting to have in store such a
threatening sequence of moves as ...Nf4, ...g5, ...Kh8, ...Rg8, ...g4. In rapid games, a clear plan for a kingside attack
often overweighs a small positional plus on the opposite flank. Black commonly aims for the following line:

B1. 13.Rc1 (Karpov)

If White started with 13.a3, we could try to transpose to the main line with 13...Qe7 14.Rc1 (14.b4 Nf4 15.Bf3 g5) since
it would be nice to trade the bad bishop after 14.Nc4 Bc5, e.g. 15.Rc1 Bxe3 16.Nxe3 Nf4 17.Re1 Ne8 18.Bf3 Nd6
19.Ne2 Nxe2+ 20.Rxe2 Rac8 21.a4 g6
There is no much sense in taking control over c4 with 13...b5 because the white knight would go to either c5 or a5
after 14.b4! [after 14.Nb3 Black achieves a dream set-up with 14...Qe7 (or 14...Nf4 15.Bf3 Qe7 16.Na5 g5)
15.Na5 Nf4 16.Bf3 g5 when 17.g3 is not a threat due to 17...Kh8! 18.b4 Rg8 19.Rc1 Rg6 20.Nb7 Rag8 21.Re1
Qf8] 14...Nf4 15.Nb3! Nxe2+ 16.Qxe2 Ng4 17.Bd2 f5 18.f3 (18.Nc5) 18...Nf6 19.exf5 Bxf5 20.Bg5.

13.Nc4 gives us time to undermine the centre with 13...Bb4 14.Qc2 b5 15.Nd2 c6. Black should gradually equalise
after 16.dxc6 Bxc6 17.Rfd1 Qe7 18.a3 Bd6.


This is a good flexible move. Black preserves plans with ...Bc5 or ...c6 while taking c5 under control. It is all the more
venomous because it may lure White into following in Karpovs footsteps to reach the position on the next diagram.
The straightforward 13...Nf4!? is also good. Then 14.Bf3? g5 15.Nc4 Kh8! 16.a3 Rg8 is game 5 Miton-Berzinsh, rapid,
Warsaw 2013.
White should probably continue 14.Nc4 when 14...Nxe2+ (14...g5 15.g3) 15.Qe2 b5 is balanced. Black also does not
have any problems with his bishop pair after 14.Re1 Nxe2 (14...g5 15.Bf1 Kh8 16.Nc4) 15.Qxe2 Ng4.

14.a3 b5 15.Nb3 Nf4 16.Bf3


Scherbakov finishes his survey in Chess Publishing before this move granting White an advantage. My assessment is
quite the opposite as Blacks attack is running by itself.
Note that 16...g5 is an important improvement over the stem-game Karpov-Milov, Biel 1997, which saw 16...Kh8?!
17.Na2! and both Whites knights found terrific stands on c5 and b4. The novelty 16...g5 was introduced in the game
Lugovoi-Kharlov, Rethymnon 2003, but Black did not connect it with the rook lift ...Rf8-g8-g6. Its point is that
Karpovs manoeuvre Na2 would drop the e4-pawn.
I would also like to mention that the other possible plan ...f7-f5 offers White a free hand in the centre, e.g. 13.a3 b5
14.b4 Nf4 15.Nb3 Nxe2+?! 16.Qxe2 Ng4 17.Bd2 f5, Wojtaszek-Mista, 2011, 18.f3 Nf6 19.exf5.

17.Na5 Kh8! 18.b4 Rg8

Black is aiming to bring his heavy pieces to the h-file with ...R8-g6-h6 or ...Qe7-f8-h6 followed by ...Rg8-g5-h4. The
arising positions may be dynamically balanced, but Whites mistakes have a higher cost. The key point is that Black
can ignore the threat g2-g3xf4 as his threats on the h-file will provide him with adequate compensation.
I consider this set-up in game 5 Miton-Berzinsh, Warsaw 2013. A similar course followed game 4 Olszewski-Mista,
Warsaw 2011, which saw another move order: 13.a3 Qe7 14.b4 Nf4 15.Bf3 g5.
If White chooses a waiting game, we have a number of ways to generate threats against his king. For example:
19.g3 Rg6
The manoeuvre 19...Qf8!? maintains more tension 20.Kh1 Qh6 21.Rg1 g4 22.Be2 Rg5.
20.Kh1 Rag8 21.Rg1 Rh6 22.gxf4 gxf4 23.Rxg8+ Nxg8 24.Ba7 Qh4 25.Qg1

Black can force a draw with 25...Rg6 or play on with 25...Nf6 26.Nb3 Ng4 27.Bxg4 Bxg4 28.Nd2 Rg6 29.f3 (29.Qg2?
f3! 30.Qg3 Qh5) 29...Bxf3+ 30.Nxf3 Rxg1+ 31.Kxg1 Qh3.

19.Qd2 Qf8 20.Nb7 Qh6 21.Bd1 Nxg2 22.Kxg2 Qh3+ 23.Kg1 Ng4 with compensation.

19.Nb7 Qf8 20.Qd2 Qh6 21.Bd1 Bh3 22.Nxd6 cxd6 23.gxh3 Qxh3 24.f3 Rg6

White is unable to disentangle his pieces after 25.Rf2 Rag8 threatening ...g4.

B2. Just for the record, I want to mention the solid, but rather boring approach when Black plays for equalisation with


Blacks task would be easier after 13.Nc4 Bb4 14.Qc2 b5 15.Nd2 c6.

13...b5 14.a3 c5 15.dxc6 Bxc6 16.g3

Or 16.Na4 Rc8 17.Nc5 Nf4.

16...Qe7 17.b4 Rfd8

Blacks bishops are less stable than the the e3-e2 pair so Black should aim to trade both rooks (for instance, ...Qe6,
...Be7) and hold the symmetrical position.
I consider similar positions with symmetrical pawn structures in Game 2 M.Rodshtein-Raetsky and Game 1 Anand-
Vallejo Pons.

B3. Unfortunately, in 2010 Bacrot introduced the novelty


It effectively hinders Blacks counterplay with ...Nf4 thus reducing our options down to the plan of line B2.


Evacuating the queen from the d-file. 13...b5 only offers White a lever on the queenside after 14.a4 Rb8 15.axb5 axb5


14.a3 might transpose after 14...b5, but Black could also swap his bad bishop with 14...Bc5.

Taking c4 under control. Black is better prepared to meet 15.a4 than he was one move ago since he can protect the b5-
pawn with his kings rook 15...Rfb8, intending ...bxa4.


Navara-Mista, Czechia 2011, saw 15.Bg5 h6 16.Bxf6 Qxf6 17.Bg4. White has a clear positional plan. He wants to
eventually remain with a knight vs. a bad dark-squared bishop. I would argue that a bishop is rarely weaker than a
knight except for totally closed symmetrical pawn structures. In our example, Black has enough counterplay in the
centre, but instead of 17...Bxg4 18.Qxg4 Qg5, he should have played 17...Qd8 18.Nb3 Ne7 19.Bxd7 Qxd7 20.Qd3

The d5-pawn is a target now. Black will swiftly cumulate 3 hits by Qf7, Rad8.

I think that blocking the queenside with b2-b4 is a sterner test of Blacks set-up. After 15.a3, his only sensible plan is to
reach the symmetrical positions from line B2. Im not too sure about the best shaping of this idea. He can either play
...c5 at once, or wait for a better timing.

a) 15...c5 16.dxc6 Bxc6 17.b4!

17.Bg5 h6 18.Bxf6 Qxf6 19.Nd5 Bxd5 20.exd5 Qd8 21.a4 bxa4 22.Nc4 Ne7 23.Qxa4 Nf5=.
17.a4 Qb7 18.Bg5 (18.Nb3 Bb4 19.axb5 axb5 20.f3 Rfd8) 18...b4 19.Ncb1 Be7=.

White has the better minor pieces, but nothing tangible. Black should aim to trade rooks and bring the g6-knight into
play, probably with ...Ng6-f8-e6.


Black might leave the other rook on a8 in order to support ...a6-a5, e.g. 17...Rfc8 18.Rc2 Nf8.
18.Qc2 Rac8 19.Rfd1 Nf8. The first line of Komodo now is 20.Bf3 Qe6 21.Qb1 Be7 22.Nb3 Rxd1+ 23.Rxd1 Rd8
24.Rxd8 Bxd8 25.Nc5 Qc8 26.Nd5 N8d7 27.Nxf6+ Nxf6. The only thing White has achieved is multiple exchanges
which favour the defending side, of course. Houdini is closer to the truth with an evaluation of 0.03!

b) 15...Rfb8

Aimed against 16.b4 which will be attacked with 16...a5. 16.Nb3 is also harmless due to 16...b4.

16.Qc2 h6 17.f3 a5 18.Rfe1 a4 19.Red1 Ra6

Black has covered everything and he can choose when to play ...c5.

Chapter 1. 3.e4 Nc6

Annotated Games

1. Anand 2785 Vallejo Pons 2712

Bilbao 18.09.2014

1.d4 d5 2.c4 dxc4 3.e4 Nc6 4.Nf3 Bg4 5.d5 Ne5 6.Bf4 Ng6 7.Be3 e5 8.Bxc4 Nf6 9.Nc3 a6 10.Be2 Bd6 11.Nd2

This exchange might be not worse than the retreat to d7 from a computers standpoint, but it dooms Black to a long
struggle in a slightly inferior position. Without light-squared bishops, Blacks chances for a kingside attack are none.
We should not easily accept such scenarios. At the same time, it is amazing in how many games Black avoids ...a6 and
covers the check from b5 by ...Bd7 which leads to the same type of positions.

12.Qxe2 0-0 13.0-0 Qe7 14.Rfd1

Anand obviously assumes that Blacks most promising plan is connected with ...c6 and he puts a rook on the future open
file. I think that ...c6 (especially while White has not weakened his kingside yet) does not solve any problem. Instead of
the d5-pawn White will get a nice outpost for his pieces. Blacks only real counterplay is on the kingside. It may be
based either on ...Nf4, followed up by ...g5, or on ...f5. Therefore, White should probably start with 14.g3! and if
14...Qd7, then 15.f3 Ne8 16.Rac1 Ne7
17.Nc4 f5. White preserves the better centre, but at least Black is kicking. He will concentrate all his pieces on the right
wing and wait for an opportunity to activate them.


Apparently Vallejo believes that his position is solid and he could hold it with calm defence. In my opinion, it would be
more consistent to try 14...Nf4 15.Qf3 g5 16.Nc4 Kh8, following the same scheme as with light-squared bishops.

15.g3! h6

Black is consistent in pursuing his plan of undermining the centre. Before ...c6, he prevents the possible pin Bg5.

16.Rac1 c6 17.Nc4! cxd5 18.Nxd5 Nxd5 19.Rxd5 Bc5 20.Rcd1 Bxe3 21.Nxe3

Black has fulfilled his plan, but only to discover that he is in straits. It turns out that he cannot unload further the
position with 21...Rfd8 due to 22.Nf5 Qf6 23.Qg4! when 23...Rxd5 fails to 24.Nxh6+!.
Instead, 21...Rcd8 does not lose material, but Blacks kings rook remains passive. That gives White time for building
pressure with h2-h4-h5.

21...Qb4 22.Nf5 Rc4 23.Nd6

Anand avoids forced variations like 23.a3! Rxe4 24.Qf3 Re1+ 25.Kg2 Qe4 (25...e4 26.Nxh6+!) 26.Rxe1 Qxe1 27.Rd1
Qa5 28.Qxb7 with the terrible threat Rd1-d6Xg6, but his move drops a considerable deal of his advantage.


23...Rc7 was essential. If White then continued as in the game, 24.a3 Qb3 25.R5d3 Qb6 26.Nf5, Black would have
26...Ne7. A trade of knights would be a great achievement.

24.a3 Qb3 25.R5d3 Qb6 26.Nf5 Re8? 27.Rd7

White has a decisive advantage. The game finished with:

27...Rf6 28.Qg4 Qc6 29.h4 h5 30.Qxh5 Qxe4 31.Rd8 Qc6 32.Qg5 Qe6 33.R1d6 1-0

The moral of this game is that Blacks light-squared bishop is an important piece and it is should be preserved. It not
only bolsters kingside activities, but also could control the d5-square in the event of ...c7-c6.

2. Rodshtein 2642 Raetsky 2449

Biel 03.08.2012

1.d4 d5 2.c4 dxc4 3.e4 Nc6 4.Nf3 Bg4 5.d5 Ne5 6.Bf4 Ng6 7.Be3 e5 8.Bxc4 a6 9.Nbd2
This set-up is more unpleasant against 7...Nf6 8.Bxc4! a6? 9.Nbd2!.

9...Bd6 10.Be2 Nf6 11.0-0 0-0

Lets compare this position with the standard set-up with Nc3.
Both side are deprived of their most common plans. White can not play Nfd2, nor can he jump to c4 since the e4-pawn
is hanging.
Black, for his part, lacks the option of ...Nf4 due to the fork on e5 after Bxf4 and e4-e5 (Whites knights are connected
so ...Bxf3 does not help!). On the other hand, his plan B, which is based on undermining the centre with ...c6, gains
in strength because of the passive stand of the d2-knight. That transpires from variations like 12.Ne1 Bd7 13.Nd3 c6
14.dxc6 Bxc6 15.f3 Bb5 (or 15...Bc7!? 16.g3 Bb6). Rodshtein tries to uncoordinate the opponents pieces by hitting the

12.Qb3 Qb8

This may look weird, but it is already obvious that the main battleground will be the queenside so Black sends there his
heavy pieces. Next, he can play ...Rc8, ...Bd7 , then choose the right timing for ...c6. Even if he plays ...b5, he should
not shun the idea of destroying the centre with ...c6. Another point in favour of Blacks move is that from b8 the queen
protects the d6-bishop. This could be important after ...c6.


White needs this move anyway. 13.Rac1 would not change Blacks plan. He can start preparing ...c6 with 13...Bd7 when
14.Bg5 hits the air due to 14...Nh5 (or even 14...c6) 15.g3 h6 16.Be3 Nf6, returning to the main track. Then Black
could wait with moves like ...Ne7, or rush with ...c6 as in the line 17.Rc2 Rc8 18.Rfc1 c6.

The knight is no longer useful on g6 so it shifts closer to the critical squares c6-d5. However, it was better to spend a
tempo on the more restraint 13...h6. White does not have a clear plan anyway.

14.Rac1 h6! 15.Rfe1 b5

Having defended the e2-bishop, White was threatening Nc4.


Enabling Nd2-b3-c5. Black is just in time with his thematic counterplay:

16...Rc8 17.Nb3 c6 18.dxc6 Nxc6

The opening stage is over. Black has achieved his main strategic goal to reach a balanced game. As it often happens, the
stronger player went on to win, but the opening was not to blame.

19.a3 Bf8 20.Nfd2 Bxe2

20...Rd8 21.Qc2 Bxe2 22.Rxe2 Rc8 was a solid alternative, but Black has a plan to occupy the light squares. If White
plants a knight on c5, he will trade his bishop for it.

21.Qxe2 a5 22.Rc2 Nd7 23.Rec1 Qb7 24.Nf3 a4 25.Nbd2

25.Nc5 Bxc5! 26.Bxc5 Na5 27.Bb4 Nc4 would be a strategic triumph. After the text Raetsky could have maintained
the tension with 25...Rc7. Instead, he charges down on the a-pawn, but misses the target. There was no need to hand the
opponent the c4-square and open the b-file in Whites favour.

25...b4 26.Qd3 bxa3?! 27.bxa3 Na5? (this loses a pawn) 28.Rb1 Nb3 29.Rxc8 Qxc8 30.Nxb3 axb3 31.Rxb3 Nc5
32.Bxc5 Qxc5 33.Rc3 Qa5

Of course, 33...Rxa3 should be a draw.

34.Nd2 Rd8 35.Qc2 Qb5 36.Nf3 Qa5 37.a4 g6 38.h4 Ra8 39.Rc4 h5 40.Kg2 Ra7 41.Qb2 Bd6 42.Rc6 Bf8 43.Rc8
Ra8 44.Rxa8 Qxa8

Q+N easily converts the extra pawn against Q+B in this symmetrical position.

45.Qc2 Qa5 46.Qb3 Be7 47.Qb7 Bd6 48.Qd7 Qc7 49.Qb5 Qb8 50.Qd5 Kg7 51.Ng5 1-0

3. Sakaev 2619 Yakovich 2556

Moscow 2009

1.d4 d5 2.c4 dxc4 3.e4 Nc6 4.Nf3 Bg4 5.d5 Ne5 6.Bf4 Ng6 7.Be3 Nf6 8.Nc3

This move is considered dubious in view of the check from a4 and the result of the current game seems to confirm such
evaluation. However, my analysis suggests that the queenless middlegame is rather unclear and perhaps White himself
should look for improvements.

9.Qa4+ Qd7 10.Qxd7+ Kxd7!

10...Nxd7? 11.Nd4 is embarrassing. Black does save the piece with 11...exd5 12.h3 c5, but at the price of Whites
strong initiative.


It seems that White has a simple solution: 11.dxe6 Bxe6 12.0-0-0+ Ke8, but then he will be unable to regain the pawn
profitably: 13.Ng5 (Dreevs attempt to prevent the threat of ...Ng4 is outright dubious: 13.h3?! Bb4! 14.Ng5 Bxc3
15.bxc3 b5.) 13...Ng4 14.Nxe6 fxe6 15.Bxc4 Nxe3 16.fxe3 Bc5 17.Bxe6 c6 with a draw due to the opposite coloured
bishops, e.g. 18.Kd2 Ke7 19.Bb3 Rad8+.

White risks to get 5(!) pawn islets in the event of the natural 11.Bxc4 exd5 12.exd5 (12.0-0-0 Bd6 13.Bxd5 Nxd5
14.Rxd5= Bxf3 15.gxf3 c6) 12...Bxf3 13.gxf3. White can only hope to maintain the balance thanks to his bishop pair,
e.g. 13...a6 (or 13...Nh4 14.Ke2 a6) 14.Ne4 Re8 15.Nxf6+ gxf6 16.0-0-0 Bd6, M.Gurevich-Ivanchuk, Reggio Emilia

Having rejected the above two options, Karpovs move 11.Ng5!? seems best.

11...exd5 12.Nxf7 Rg8 13.f3

White wants to castle long and keep the d-file open. 13.exd5 has no venom due to 13...b5! and the f7-knight might find
itself without reteat squares. For instance, 14.Nxb5?! Bb4+ 15.Nc3 Nxd5 16.a3 Bxc3+ 17.bxc3 h6. 14.f3 is not much
of a threat, too 14...Re8 15.Kf2 a6. Finally, 14.h3 Bf5 15.Nxb5 Bb4+ 16.Nc3 h6 also puts Whites knight in a

13...Be6 14.Ng5


This improves on 14...c6 15.0-0-0 h6 16.Nxe6 Kxe6 17.Bd4 Bd6 18.Bxf6 gxf6 19.exd5+ cxd5 20.Rxd5 Ne5 21.g3 Bb4
22.Rd4, Karpov-Lautier, Monte Carlo 1997.


Critical for Yakovichs idea is the line 15.Nxe6 Kxe6 16.exd5+ Kd6 17.Bxc4 Rae8 18.Kd2 Nxd5
Now 19.Bxd5 Bxc3+ 20.Kxc3 Kxd5 is marginally better for White, but Black does not have weaknesses. A sterner test
of Blacks idea is 19.Bf2 Bxc3+ 20.bxc3 Ne5 21.Bb3 b5. In theory, Whites bishop pair should give him an edge, but
the central knight pair is no less impressive, in my opinion.

15...Bxc3 16.bxc3 b5 17.g3

Preparing a sortie for the bishop on h3. 17.Bd4 c6 18.g3 transposes.

17...c6 (17...Rge8!?)

A very unusual position has arisen. It transpires that Blacks knights are quite dangerous, especially when supported by
a rook on the e-file. I would say that White must be very accurate in order to keep the balance.


Inserting 18.a4 a6 is fruitless 19.Bd4 Rge8 20.e5 c5=.

18...Rge8 19.Nxe6

Another funny position arises after 19.e5 c5 20.exf6 cxd4 21.fxg7 d3

Blacks pawns provide sufficient counterchances, for instance, 22.Re1 Bg8 23.Bh3+ Kd6 24.f4 b4 25.Kd2 bxc3+
26.Kxc3 Rab8.

19...Kxe6 20.Bh3+ Kd6 21.Bf5 Kc7 (22...a5!?) 22.exd5 Nxd5

Whites bishops are balanced by Blacks active rook. The computer also likes 22...cxd5 since 23.Bxf6 gxf6 24.Rxd5
Ne7 25.Rxb5? a6 is bad, but the human GM prefers to activate his pieces.

23.Bxg7 Re2?!

It was better to immediately attack c3 with 23...Ne5 24.f4 Ng4 25.Bg4 Re3 with counterplay.

24.Rd2 Rxd2 25.Kxd2 Rd8 26.Kc1 a5 27.h4 Rg8 28.Bd4 c5 29.Bxc5 Nxc3 30.h5

The decisive mistake. 31...Ne5 was still unclear.

31.Kc2 Nxg3 32.hxg6 hxg6 33.Be6 Rd8 34.Rg1 Ne2 35.Rxg6 Nf4 36.Rf6 Nxe6 37.Rxe6 Rd5 38.Bb6+ Kd7 39.Re4
Kc6 40.Bxa5 Rd3 41.f4 Ra3 42.Bb4 Rxa2+ 43.Kc3 Rf2 44.Kd4 Rf3 45.Re6+ Kd7 46.Rd6+ Ke8 47.Ke5 Rb3 48.Bc5
Kf7 49.Rd7+ Ke8 50.Ke6 Rb1 51.Bd4 Rd1 52.Rh7 Kd8 53.Be5 1-0

4. Olszewski 2532 Mista 2565

Warsaw 15.02.2011

1.d4 d5 2.c4 dxc4 3.e4 Nc6 4.Nf3 Bg4 5.d5 Ne5 6.Bf4 Ng6 7.Be3 Nf6 8.Nc3 e5 9.Bxc4 a6 10.Be2 Bd6 11.0-0 0-0
12.Nd2 Bd7 13.a3 Qe7 14.b4
The set-up with a3, b4 looks logical. White hinders a possible trade of bishops via c5 and supports the manoeuvre Nd2-
b3-c5. A minor drawback could be the option of 14...a5, but in fact White would be only happy to open the queenside
with bxa5 where his pieces are dominant.

14...Nf4 15.Bf3 g5 16.Re1 Kh8 17.Kh1 Rg8 18.g3

A cautious move which eliminates any ideas of ...Nxg2.


I do not understand the idea behind this move. Instead of attacking on the h-file, Mista voluntarily plugs it!
I can offer at least three more consistent plans:

a) 18...Qf8 19.Rg1 Rg6. White can hardly wait for Black to concentrate more forces against h2 so he should take the
piece and play a complex position after 20.gxf4 gxf4
21.Bc5 Bxc5 22.bxc5 Qxc5 23.Ne2 Qxf2;

b) 18...g4 19.Be2 Nxe2 20.Qxe2 h5

21.Rac1 (21.Nc4 Nh7) 21...h4;

c) 18...Rg6 19.Nc4 Rag8 20.Rg1 Rh6. Now 21.gxf4 gxf4 22.Rxg8+ Nxg8 23.Bxf4 exf4 24.Qd4+ f6 is unclear. Another
option is 21.Rc1 g4 22.Be2 Ne8 when 23.gxf4?? Rxh2+ mates.

19.Rg1 Rg6 20.Nc4 Rag8 21.b5?

A hasty breakthrough. 21.Bd2 would have threatened the f4-bishop since 21...Rh6 22.gxf4 gxf4 23.Qe2 is clearly better
for White while 21...Ne8 22.Bg4 neutralises Blacks attack. Nc4-e3-f5 would be looming.

21...axb5 22.Nxd6 cxd6 23.Rc1 Bd7

24.gxf4 gxf4 25.Rxg6 Rxg6 26.Bb6 Ng8 27.Ne2 Qh4 28.Rc7 Bg4

Black has full compensation for the piece. White can still draw, but he must find 29.Ng1 Nf6 30.Rc2 (30.Qe2 Rh6)
30...Rh6 31.Kg2 Bh3+ (31...Qxh2+ 32.Kf1 Bd7 33.Ke2 Rg6 34.Rc1 Ng4 35.Qe1 Ne3 36.Bxe3 fxe3 37.Kxe3 Qf4+
38.Ke2 b4=) 32.Nxh3 Qxh3+ 33.Kg1 Rg6+ 34.Kh1 Rh6 35.Kg1=. After his blunder, the game is over.

29.Nd4?? Qh3 30.Rc1 exd4 0-1

5. Miton 2591 Berzinsh 2421

rapid Warsaw 14.12.2013

1.d4 d5 2.c4 dxc4 3.e4 Nc6 4.Nf3 Bg4 5.d5 Ne5 6.Bf4 Ng6 7.Be3 e5 8.Bxc4 Nf6 9.Nc3 a6 10.Be2 Bd6 11.0-0 0-0
12.Nd2 Bd7 13.Rc1 Nf4
Black takes the bull by the horns and embarks upon his main plan without any preparation. Indeed, it is not clear how
White could exploit the fact that Black did not include neither ...b5, nor ...Qe7. For instance, 14.Bxf4 exf4 15.Nc4 Re8
16.Nxd6 cxd6 17.Qd4 h5 18.Rfe1 (18.f3 h4) 18...Ng4 19.Bxg4 hxg4 is unclear. Miton chooses to ignore the specific
move order and adopts Karpovs set-up. His next move aims to provoke ...g5-g4 which would take the g4-square away
from Blacks minor pieces. Instead, 14.Re1 Nxe2+ 15.Qxe2 Ng4 would be comfortable for Black.

14.Bf3 g5 15.Nc4 Kh8 16.a3 Rg8 17.b4 Qe7

Black must bring somehow his queen to the h-file. He commonly plays ...Rg6, ...Rag8, and then ...Qf8!?-h6. In this
game Berzinsh saved ...Qe7 at an earlier stage and now he could have lead his queen directly to f8-h6. Then he could
set up a heavy battery on the h-file with ...Rg5-h5.

18.Qc2 Rg6 19.Bc5 Rag8 20.Rfe1 g4 21.Bd1

Starting from move 14, White has made all the natural moves. There is one little nuance though. He refrained from
g2-g3. Perhaps he thought he could make it later, but 21...Nxg2!! in the diagrammed position would have disillusioned
him. After 22.Kxg2 Nh5 23.Be3 Qh4 Black gets a decisive attack: 24.Kf1 Nf4

25.Bxf4 exf4 26.Ke2 g3 27.Nxd6 cxd6 28.hxg3 fxg3. White is helpless against the march of the h-pawn! 29.Qd3 Qf6
30.f3 g2 31.Rg1 h5+. Do not get the impression, however, that White is victim of an accidental turn of events. In fact
his king is in a precarious situation and Black has many ways to lead the attack with more conventional means:
21...Bxc5 22.bxc5 Rh6 or 21...N6h5 22.g3 Nh3+ 23.Kf1 b5 24.Bxd6 cxd6 25.Nb6 f5 26.Nxd7 f4+.

21...Rh6? 22.g3 Ne8 23.Be3 Qg5 24.gxf4?

24.Qd2 threatening Nxe5 would have neutralised the attack. Whites subsequent play is cooperative. He obviously
missed something.

24...exf4 25.Bd4+ f6 26.Nxd6?? (26.e5) 26...cxd6 27.Qd2 Rxh2 28.Ne2 Qh4 0-1

Chapter 2. The QGA la Chigorin

Main Ideas

1.d4 d5 2.c4 dxc4 3.e4 Nc6 4.Nf3 Nf6!?

This transition to the Chigorin is the most interesting and promising Blacks weapon against 3.e4. Here are my
arguments in its favour:

1. It is somewhat underestimated.

2. There is little to learn. The databases contain only a few games at top level and your opponents will find it difficult to
prepare against you since he would not know what example to follow!

3. This variation is based not on long forced lines, but on understanding of the arising asymmetric pawn structures. The
positions are strategically unbalanced and general considerations rarely helps.
What is the difference in comparison to the lines from the previous chapter?
First of all, we avoid the static structures connected with ...e5. Instead we apply piece pressure on the centre and our
goal is to double the enemy pawns on the f-file.

The tactical background of our play is the line 5.d5 Nxe4! 6.Bxc4 e6! 7.0-0 Bc5!
The positions with 2 pieces for a rook and pawns are dynamically balanced.

Another tangled line is 5.Nc3 Bg4 6.d5 Ne5 7.Bf4

Now 7...Ng6 8.Be3 e6!? would transpose to Chapter 1, see game 3 Sakaev-Yakovich, Moscow 2009. However, the
insertion 4...Nf6 allows us to stay on the Chigorin wave with:

This should come as a bolt out of the blue for your opponents. Morozevich/Barsky briefly mention:
8.Qa4, but then 8...Nxf3+ 9.gxf3 e5! 10.dxe6 Bxe6 11.Bxc4 Bxc4 12.Qxc4 Qf6 equalises.
A critical position arises after: 8.Qd4 c5 9.Qe3 Bxf3 10.gxf3 g6 11.0-0-0 Bg7 12.Bh6 Bxh6! 13.Qxh6 Nxf3 14.Bxc4
Black will castle long with a good game (15.d6 e6).

The most natural and popular retort to 4...Nf6 is:

5.Nc3 Bg4 6.Be3 e6 7.Bxc4 Bb4 8.Qc2! 0-0 9.Rd1

9.0-0-0 Bxc3! 10.bxc3 Qe7 is double-edged, see game 6 Shirov-Morozevich, Amsterdam 1995.

9...Bxf3! 10.gxf3 Nh5

How to evaluate this position? Whites trumps are the mobile pawn centre and the bishop pair. However, the threat ...e6-
e5 followed up by ...Nd4 forces White to play e4-e5 himself. Then Whites centre will not be flexible anymore and
could easily become a target. Another point in Blacks favour is that he has nice stands for his pieces on the kingside. I
would prefer to be in Blacks shos here. Morozevich/Barsky give:

11.e5 Ne7 12.0-0

Black has two possible plans here.

1. To keep his dark-squared bishop and aim for ...f7-f5, see game 7 Magerramov-Al Modiahki, Dubai 2000.

2. To kill the enemy knight and prove that the bishop pair is not too helpful: 12...Bxc3!? 13.bxc3 c6 14.Qe4 Nd5.

Blacks knights have good stands, his pawn formation remains flexible. He possesses a wide scope of active plans,
starting from ...b5 or ...c5 on the kingside and finishing with ...f6 on the other side.
Although White looks active, Blacks play is more obvious. He has to move ...g6 and ponder over ...f5.

Chapter 2. The QGA la Chigorin

Step by Step

1.d4 d5 2.c4 dxc4 3.e4 Nc6!? 4.Nf3 Nf6


a) 5.e5 Nd5 6.Bxc4 is a sideline of the 3...Nf6 system. It has been established that 6...Nb6 7.Bb3 Bg4 8.Bxf7+ (8.Ng5
is a draw!) 8...Kxf7 9.Ng5+ Ke8 10.Qxg4 Qxd4 is equal, for instance 11.Qe2 Qc4 12.Nc3 Qxe2+.
Black can escape the draw with 6...Bg4 7.Nc3 e6.

b) 5.d5 Nxe4 6.Bxc4 (6.dxc6?! Qxd1+ 7.Kxd1 Nxf2+ 8.Ke1 Nxh1 seems to be in Blacks favour) 6...e6! gives Black
sufficient play. (6...Nb8 is dubious as 7.0-0 Nd6 8.Bd3 g6 is too passive.) 7.0-0 Bc5!
The arising positions with 2 pieces for a rook and pawns are dynamically balanced. Theoretically, White should aim to
keep the queens hoping to build up an attack with his minor piece:
8.dxc6 Bxf2+ 9.Kh1 Qxd1 10.Rxd1 Bb6 11.Rf1 Nf2+ 12.Rxf2 Bxf2 is safe for Black, or:
8.Qb3 Nd4 9.Nxd4 Bxd4 10.Nc3 Bxc3 11.bxc3 0-0 12.dxe6 Bxe6 13.Bxe6 Nc5 14.Bxf7+ Rxf7.
8...Nxf2 9.Rxf2 Bxf2+ 10.Kxf2 exd5. Scherbakov assesses this position as unclear. In my opinion, it is rather balanced
since White cannot keep the queens after 11.Bxd5 Ne7 12.Nc3 Be6.


A. 6.d5; B. 6.Be3

In his latest game Kiril Georgiev surprisingly opted for 6.Bxc4, but it does not look like a deep home preparation. Im
inclined to think that he just wanted to deviate from his game against Miladinovic which we are going to analyse
below. The forced variation 6.Bxc4 Bxf3 7.gxf3 (7.Qxf3 Nxd4 8.Qd3 e5) 7...Qxd4 8.Qb3 Ne5 (Morozevich/Barsky
suggest 8...0-0-0, but then 9.Be3 Na5 10.Be6+!! fxe6 11.Qb5 retains some pull.) 9.Qxb7 Qxc4 10.Qxa8+ Kd7 11.Be3
Qd3 12.Ne2 Nxf3+ 13.Kf1 e6 14.Rc1 Bc5= is quite convincing, for instance, 15.Qb7 could be met by 15...Rc8=
(Black can play on with 15...Bd6!?) .

A. 6.d5 Ne5 7.Bf4

7.Bxc4? Nxc4 8.Qa4+ c6! favours Black.

7.Qd4?! Nxf3+ 8.gxf3 Bxf3 9.Rg1 e5 (9...e6 is also playable) 10.Qxe5+ Qe7 effectively neutralises Whites activity.


7...Ng6 transposes to Chapter 1 with an important nuance White has already committed his knight to c3 so he does not
have the nasty pawn sacrifice Bxc4. However, the inclusion of Nf6 Nc3 offers us the alternative ...Nfd7. The big
question is: to take first on f3 or not.

Morozevich/Barsky recommend:
7...Bxf3 8.gxf3 Qd6,
when 9.Bg3! is the most testing.
Unfortunately, they do not mention this move at all. Alternatives are:
a) 9.Qa4+ Nfd7 10.Bxe5 (10.0-0-0 Nd3+ 11.Rxd3 Qxf4+ 12.Re3 a6) 10...Qxe5 11.Bh3 e6 12.dxe6 fxe6 13.Qxc4
0-0-0 (13...Nc5 14.0-0-0 a5) 14.Qxe6 Kb8 15.0-0-0 Bd6 (15...Qf4+ 16.Kb1 Bd6);
b) 9.Bxc4 0-0-0 10.Bg3 Nh5 11.Be2 Qf6 12.Rc1 (12.Nb5 a6 13.Nxc7 Kxc7 14.f4 Nxg3 15.fxe5 Qxe5 16.hxg3)
12...e6 13.Nb5!?

9...Nfd7 (9...0-0-0 10.f4 Ned7 11.e5 Qa6 12.exf6 exf6 13.Qf3) 10.Bxc4 It is unclear what Black can oppose to the
bishop pair and the pawn centre. I checked:
10...g5 11.h4 Bg7 12.hxg5 Qb4 13.Bb3 0-0-0 14.f4 Ng6 15.Qf3 h5 16.gxh6;
10...h5 11.h4 e6 12.Qe2 [12.Be2 exd5 13.f4 Ng4 14.Bxg4 hxg4 15.f5 Qa6 16.Qxd5 (16.Nxd5 0-0-0 17.Bxc7 Re8
18.Qc2 Nc5) 16...c6 17.Qd2 0-0-0] 12...Qb6 13.Bb3.


Another version of this idea is 8.Bxe5 Nxe5 9.Qd4

Karpov beat Morozevich in a blitz game with 9.Bxc4 Nxc4 10.Qa4+ Qd7 11.Qxc4 Bxf3 12.gxf3 when instead of
12...g6?!, Morozevich suggests 12...0-0-0!?. Black can follow up with ...Kb8 and ...e6 or ...g6.
9...Nxf3+ 10.gxf3 Bxf3 11.Qxc4 (11.Bxc4 a6) when:
11...e6! (11...Qd7 12.Rg1) leads to sharp interesting play: 12.Qa4+ c6 13.dxc6 b5 14.Bxb5 Qc7 15.Rg1 Bd6 16.Rxg7

Morozevich/Barsky claim that 8.Qa4 could cause considerable trouble, but without providing any further analysis. Lets
check it: 8...Nxf3+ 9.gxf3 e5 10.dxe6 (10.Bxe5? Bxf3 11.Rg1 a6) 10...Bxe6

11.Nb5 Rc8 12.0-0-0 Bc5 13.Rg1 0-0 14.Bh6 (14.Bg5 Be7) 14...g6 is even more tangled, but Blacks pieces are
better coordinated.
11...Qf6 12.Bxc7 Rc8 13.Bxe6 fxe6 14.Nb5 a6 15.Nd6+ Bxd6 16.Bxd6 Qxf3. Black does not risk much here.

My wife faced recently 8.Be2 Bxf3 9.gxf3 and she went here for 9...g6?! 10.Bxe5?! Nxe5 11.f4 Nd3+ 12.Bxd3 cxd3
13.Qa4+ Qd7 14.Qd4 d2+ 15.Kxd2 f6 16.Ke3 Bh6 Stopa-Djingarova, Kavala 2015. In the postmortem, we found that
White could launch an attack with 10.Be3! Bg7 11.f4 Nd3+ 12.Bxd3 cxd3 13.Qxd3 c6 14.dxc6 bxc6 15.h4.
Black should attack the centre immediately with 9...e6!

Now 10.dxe6? fxe6 11.Be3 would fail to 11...Nc5, and 10.Be3 Bc5 also gives Black an initiative.

8.Bg3 e6! 9.Bxe5 Bxf3 10.Qa4 a6 11.Bg3 is very interesting and chaotic after 11...b5 (11...Bh5 12.dxe6 fxe6 13.Qxc4
Nc5 is certainly safer.) 12.Qa5 Bxe4!


8...Nxf3+ 9.gxf3 e5 10.dxe6 Bxe6 11.Bxc4 Bxc4 12.Qxc4 c6 13.0-0-0 Qf6 (Morozevich/Barsky) is a playable
alternative, but White has some initiative after 14.Be3.

9.Qe3 Bxf3 10.gxf3 g6 11.0-0-0 Bg7 12.Bh6

We have been following the game Razuvaev-Morozevich, Moscow 1993. Morozevich/Barsky offer as an improvement
on 12...Bf6? 12...Bxh6! 13.Qxh6 Nxf3 14.Bxc4 Qb6 and assess the position as slightly better for White. Indeed Black
is behind in development, but he has a nice control over the dark squares. After castling, he could take over the
initiative. The only sensible try for White seems 15.d6 e6 when nothing can prevent ...0-0-0. For instance 16.Qg7 0-0-0
17.Qxf7? fails to 17...Nfe5 18.Qxe6 Rhe8 19.Nd5 Qxb2+!!.
Instead of 14...Qb6, Black has the even more tempting 14...a6!? 15.d6 e6 and White must be very ingenious to keep the

B. 6.Be3 e6

6...Bxf3 7.gxf3 e5 8.d5 Ne7 is dubious due to 9.Qa4+ Nd7 10.d6! Nc6 (10...cxd6 11.Bxc4) 11.dxc7 Qxc7 12.0-0-0.


7.Qa4 gives us tempi 7...Nd7 8.Qxc4 Nb6 9.Qd3 f5! and Whites centre is not so impressive anymore.


Black is strictly abiding to the spirit of the Chigorin Defence. He is freely developing his minor pieces, hoping to
provoke the opponent to advance one of his central pawns. Then White would remain with a bunch of weak pawns. For

a) 8.d5? exd5 9.exd5 Bxc3+ 10.bxc3 Ne5 11.Bb5+ Kf8!;

b) 8.e5? Nd5 9.Rc1 0-0 10.0-0 Bxf3 11.Qxf3 Nxe3! 12.fxe3 Nxd4 13.Qxb7 Rb8 14.Qe4 Nf5.

c) Therefore, White should defend the e4-pawn by his queen, conceeding doubled pawns on the f-file. He should also
aim to keep more pieces on the board in order to exploit more effectively his space advantage.
8.Qd3 would leave the c4-bishop without retreat squares. Black levels the game with 8...Bxf3! (8...0-0 9.Nd2! e5 10.d5
Ne7 11.f3 Bc8 12.a3 Bxc3 13.Qxc3 c6 14.dxc6 Nxc6 15.Nb3 is clearly better for White) 9.gxf3 Na5 10.a3 Bxc3+
11.Qxc3 Nxc4 12.Qxc4 0-0 intending 13...Nh5, ...Qf6. The c-pawn should remain on c7. Then the possible
breakthrough d5 is less dangerous since the position remains relatively closed, e.g. 13.0-0-0 Nh5 14.d5 Qf6 15.dxe6
Qxe6 16.Qxe6 fxe6 17.Rd7 Rac8!=.

8.Qc2! 0-0 9.Rd1

9.0-0-0 is a bit overambitious. After 9...Bxc3 10.bxc3 Qe7, intending ...Rfb8, all Blacks heavy pieces will be directed
toward the enemy king.

See game 6 Shirov-Morozevich, Amsterdam 1995.

With 9.Rd1, White has defended everything and now he is threatening Be2 followed up by 0-0, with a stable edge in the
centre. We must inflict some damage on his pawn structure quickly.


Innocuous moves like 9...Ne7 lead to difficulties: 10.Be2 Bxc3+ 11.bxc3 c5 12.0-0 Qc7 13.Qb1 Rfd8 14.h3 Bh5

The only reasonable alternative is a strike in the centre with:

9...Qe7?! 10.Be2 e5 11.d5 Nd4 12.Nxd4 exd4 13.Rxd4 Bxe2 14.Kxe2
This pawn sacrifice occurred for the first time in the game Kiril Georgiev-Morozevich, Tilburg 1994 which White won
convincingly. In 2008 Kiril also beat Miladinovic. The moral of his games is that Whites king feels fine in the middle
of the board and it supports the pawn centre. Black does not have enough counterplay. For example:
14...Bc5 15.Rd2 Bxe3 16.Kxe3 Ng4+ 17.Ke2 Qg5 18.Kd1.
15.Nd1 Bc5
Morozevich/Barsky suggest 15...f5 (!), but White is on top after 16.exf5 Nf6 17.g4;
15...Rae8 16.f4.
16.Rc4 Nxe3 17.Nxe3 Bxe3 18.Kxe3 Rae8 19.f4.

10.gxf3 Nh5


Other logical moves are:
a) 11.a3 Ba5 12.b4
12.e5 Ne7 (12...Bxc3+ 13.bxc3 Qh4) 13.b4 Bb6.
12...Bb6 13.e5 Nxd4 14.Qe4 f5 15.exf6 Nxf6 16.Rxd4 Nxe4 17.Rxd8 Nxc3 18.Bxe6+ Kh8=.

b) 11.Rg1 e5!? (11...Qh4) 12.dxe5 (12.d5?! Nd4!) 12...Qh4 with a strong initiative. For instance, 13.e6 would only
open new lines in Blacks favour.

11...Ne7 12.0-0 Bxc3!?

It is a good idea to kill the knight immediately, but 12...c6, intending ...Be7, (or 13.Kh1 Bxc3 14.bxc3 Nd5 15.Rg1 b5)
is also possible. See game 7 Magerramov-Al Modiahki, Dubai 2000.

13.bxc3 c6 14.Qe4 Nd5 15.Qg4

Possible continuations are 15...g6!? 16.Bd2 b5 or 15...Nxe3 16.fxe3 g6 17.Rd2 Ng7 18.Rg2 c5. It is easy to play such
positions since Whites mistakes cost higher.
Chapter 2. The QGA la Chigorin

Annotated Games

6. Shirov 2695 Morozevich 2630

Amsterdam 1995

1.d4 d5 2.c4 Nc6 3.Nc3 dxc4 4.Nf3 Nf6 5.e4 Bg4 6.Be3 e6 7.Bxc4 Bb4 8.Qc2 0-0 9.0-0-0

Since Black will open the g-file, it looks logical to castle left. The current game, however, brilliantly underlined the
shaky state of Whites king.

9...Bxc3! 10.bxc3 Qe7

Despite the opposite castles, Black cannot focus only on the queenside. He must also think of the centre. His last move
not only lets the kings rook pass, but it prepares ...e5.
The more direct approach 10...Rab8 11.Bd3 b5 is insufficient. The following line shows that Black lacks enough
striking power after 12.Kb1! (12.e5 Nd5 13.Bxh7+ Kh8 14.Be4 Na5) 12...b4 13.Ka1 bxc3 14.Qxc3 Nb4 15.Bb1.


It is a race game where every tempo could be fateful so Shirovs decision to open the g-file quickly is understandable.
Still, the course of the game shows that Blacks counterchances are considerable. It would be better to evacuate the
king to a1 first, but White cannot start with 11.Kb1?? due to 11...Nxe4, nor with 11.Kb2 e5 12.h3? Bxf3 13.gxf3 exd4
14.cxd4 Nxd4 15.Bxd4 Qb4+. It transpires that:
11.Bd3 may be the best option.
Then 11...b5 12.Kb1 b4 13.Ka1 suddenly gives White a strong initiative on the queenside along the c-file. ...b5 is
also a weakness in an endgame which could arise in the event of 12...e5 13.h3 Bxf3 14.gxf3 exd4 15.cxd4 Nb4
Therefore, it is better to play 11...e5 instead of ...b5. Play continues 12.h3 (or 12.d5 Na5 13.Kb2 c6 14.Qa4 b6 15.c4,
Bakic-Zakic Yugoslavia 1994, when 15...Bd7 would have been in Blacks favour.) 12...Bxf3 13.gxf3 exd4 14.cxd4

Here 15.Qc5 drops the a2-pawn while 15.Qb2 Nxd3+ 16.Rxd3 c6 would be with mutual chances.

11...Bxf3 12.gxf3 Rfb8

Blacks kings rook does not have a better place than b8 since on d8 it could be harassed by Bg5.

13.Bd3 b5 14.e5

Shirov picks up the gauntlet, but this is a very risky decision from a practical standpoint. Nobody followed in his
footsteps in later games.

It was possible to run away with:

14.Kb1 when 14...b4 allows 15.Qa4. So Black should first prevent this option with a6-a5-a4.

In two games White chose to return to the middle:

14.Kd2?! e5 15.Rb1 a6. Black then takes over the initiative with ...Qd7 (Qd6), ...Ne7, ...c6.

14...Nd5 15.Bxh7+ Kh8 16.Be4 b4


17.c4 fails to 17...b3 18.axb3 Na5.

17...exd5 18.Kd2 Na5 19.Ke2 Nc4 20.Bd2

White was not threatening to take on b4 as ...a5 would regain the pawn. That gives Black time to improve his queen
20...Qe6, eyeing a6 and taking the f5-square under control. Still the most consistent continuation would be to
undermine the e5-pawn with 20...f6!. Morozevich attacks its base:

20...c5 21.Rhg1

It transpires that 21...bxc3 22.Bxc3 cxd4 23.Rxd4 Nxe5 24.Kf1 turns the tables in Whites favour. Black must think up
another ways of developing its initiative. The previous line hints that Whites bishop may become a dangerous piece so
it would be wise to trade it with 21...Nxd2 22.Qxd2 c4. Black gets a passed pawn and enough play. Another argument
in favour of ...Nxd2 is that the bishop defends Whites king. The best move order is:
21...Qe6! 22.Rg4 Nxd2 23.Qxd2 bxc3 24.Qxc3 Qa6+ 25.Ke1 c4. Blacks answer is inaccurate since dxc5 will be
with a tempo.

21...Rb6?! 22.Rg5 (22.Rg4) 22...g6?! (22...Rh6) 23.dxc5 Nxd2 24.Qxd2 bxc3

24...Qxc5 25.cxb4 Rxb4 26.Kf1 is grim for Black he is a pawn down and under attack. The rest is irrelevant. Black
saved a half-point in a lost position after many Whites mistakes.

25.Qxc3 Ra6 26.f4 Rc8 27.Rxd5 Qe6 28.Rd2 (28.Qc4!) 28...Rxa2 29.Rxa2 Qxa2+ 30.Kf1 Kg8 31.f5 Rd8 32.Kg2
Qd5+ 33.Kh2 Qe4 34.e6 Qf4+ 35.Qg3 Qxg3+ 36.Rxg3 fxe6 37.fxe6 a5 38.Rxg6+ Kf8 39.Rg4 Ke7 40.Ra4 Ra8 41.f4
Kxe6 42.Kg3 Kd5 43.Kg4 Kxc5 44.Kg5 Kb5 45.Ra1 Rg8+ 46.Kf6 Rf8+ 47.Ke5 Re8+ 48.Kd5 Rd8+ 49.Ke5 Re8+
50.Kd5 Rd8+ 51.Ke4 Re8+ 52.Kf3 Rh8 53.Rh1 Kc5 54.f5 Kd6 55.Kf4 a4 56.Kg5 Ke7 57.f6+ Kf7 58.Rb1 Rxh3
59.Rb7+ Kf8 Draw.

7. Magerramov 2562 Al Modiahki 2538

Dubai 07.05.2000

1.d4 d5 2.c4 Nc6 3.Nc3 dxc4 4.Nf3 Nf6 5.e4 Bg4 6.Be3 e6 7.Bxc4 Bb4 8.Qc2 0-0 9.Rd1 Bxf3 10.gxf3 Nh5 11.e5
Ne7 12.0-0


I chose for the main line in the Step by Step section 12...Bxc3 because the knight is potentially a powerful attacking
pieces. Modiahkis idea of keeping the dark-squared bishop is also viable. If White trades it with:
13.Ne4 Nd5 14.Bg5 Be7 15.Bxe7?, he will completely lose control of the kingside dark squares. Instead he should try
14.a3 (to prevent Blacks queen from reaching h4) 14...Be7 15.Kh1 Kh8 16.Rg1 Nhf4 17.Rg4 Ng6 with a complex
Kopylov opted for 14.Ng3 Nhf4, but then he was unable to break the grip on f4 since 15.Qe4? faces 15...f5!. He
actually continued 15.Kh1 when Black could calmly prepare ...f7-f5 with 15...Kh8 (there is no need to allow
15...f5 16.Bxf4 Nxf4 17.a3 Be7 18.Nxf5).

13.Kh1 is also consistent. Black may wait to see the opponents next move with 13...Qd7. Then 14.Rg1 b5 15.Bd3
Bxc3 16.bxc3 g6 would be similar to 12...Bxc3. 14.Ne4 could be met by 14...f5 this advance is a constant threat

13.Bg5 h6

A brave, but not obligatory move. I do not see anything wrong with 13...Qd7, avoiding to make a lever for an attack on

14.Bc1 Nd5

I would feel more at ease with this knight on the kingside, e.g. 14...Nf5.

15.Ne2 Qh4

Blacks previous move enabled the thematic thrust 15...f5!.

It nips Whites attack in the bud. 16.Ng3 is easily tamed with 16...Qh4.

16.Qe4 Be7 17.Bd3?! f5! 18.exf6 Qxf6 19.Qg4 Qxf3 20.Qxe6+ Qf7

Black has an easy game due to his beautiful knight on d5 and connected pawns. Further manoeuvring did not tip the
scales in anybodys favour.

21.Qe4 Nhf6 22.Qg2 Kh8 23.Bg6 Qe6 24.Ng3 Nh7 25.Bb1 Ng5 26.Rde1 Qh3 27.f4 Qxg2+ 28.Kxg2 Nh7 29.a3 Bf6
30.Nf5 Rad8 31.Kg3 Rd7 32.Re2 Rfd8 33.Rfe1 Nf8 34.Bd2 Kg8 35.Kf3 Nc7 36.Bb4 Nd5 37.Bc5 b6 38.Bxf8 Kxf8
39.Re6 Ne7 40.Nxh6 Bxd4 41.Nf5 Bf6 42.b3 c5 43.a4 Kf7 44.h4 Nxf5 45.Bxf5 g6 46.Bb1 Rd1 47.R6e2 Bxh4
48.Rxd1 Rxd1 49.Bc2 Rd6 50.Re3 g5 Draw.
Chapter 3. 3.e4 b5?!

Main Ideas

1.d4 d5 2.c4 dxc4 3.e4 b5?! 4.a4 c6

5.axb5 cxb5 6.Nc3 a6 7.Nxb5 axb5 8.Rxa8 Bb7 9.Ra1

I do not believe in the future of this variation and I cannot recommend it. It has already turned into a battle between
computer analyses. Perhaps it will not be completely refuted, but I suppose that not before long White will find
relatively safe lines where he could be squeezing the opponent for many moves. Even now, White has established
forced draws in many branches, and I see at least one line where he should be clearly better. Thus White can learn only
one draw sequence for his repertoire, and build up on it when he has more time for preparation. For his part, Black must
study all Whites options and refresh his repertoire virtually after every new game. I believe that such an approach is
unrewarding. The Chigorin treatment of 3.e4 Nc6 offers more chances to play chess in very complex positions.

Meanwhile, it is still a big fun to plunge into the crazy complications, especially in rapid games.
In order to avoid tons of computer variations which you will forget anyway, I focus only on the best lines for White,
according to my current opinion.

9.Ra2 is another cause of concern for Black, but it is completely unexplored yet.

Whites safest choice here is 10.Be2 Nf6 11.Nf3, but 11...Bb4+! would leave him little chance to win. See game 8
Howell-Das, Leiden 2015. As a rule, if Black takes the e4-pawn without losing immediately, he should be able to hold
since the d4-pawn is weak and White cannot successfully defend it and attack Blacks queenside pawns at the same
time. Thus a position like this:

tends to transform into this pawn structure, which is a draw in the general case.
These considerations teach us that White should defend his e4-pawn with f3 with all the consequences it implies a
weak e3-square, and a weak g1-a7 diagonal. Im still not sure what the perfect move order should be. It looks like both:

10.f3 Nc6 11.Ne2 Nf6 12.Be3 and 10.Ne2 Nf6 11.f3 Nc6 12.Be3 should be good enough to reach the following
critical position:

Whites main threat is b3, so 12...Na5 is a logical try, but concrete analysis shows that any tempo is vital and White
takes over the initiative with countersacrifices: 13.Nc3 Nb3?! 14.Ra7 Qb6 15.Nxb5! Qxb5 16.Bxc4 or 13...Bb4
14.Be2 Nb3 15.0-0!? Nxa1 16.Qa1 with a clear positional advantage. So Black keeps on developing:

12...Bb4+ 13.Nc3 0-0 14.Ke2!

An amazing move, suggested by Morozevich in his excellent article in Chess Informant 123. It turns out that Whites
pawn centre is a good shelter to the king and the f1-bishop can go to h3. Black has not shown yet a convincing retort.

We could look for early deviations from this line. The most popular one is 10.f3 f5, but I cannot offer good advice
against 11.Nh3:

White is intending 11...fxe4 12.Ng5 or 11...Nf6 12.Ng5 Qb6 13.exf5. The play is very sharp, but Black is on the

Typical tactical ideas

I will show now some typical blows which are an essential part of the 3...b5 variation.

Black often destroys Whites shelter by hits on e4 or f3. Most often they lead to perpetual check, but here is a winning

Edmonton 2015

22...Rxa4 23.Bxa4 Nxf3+! 24.gxf3 Qe3+ 25.Kf1 Qxf3+ 26.Kg1 Qe3+ 27.Kf1 Qf3+ 28.Kg1 Qe3+ 29.Kf1 Nd3 30.h4
Qf4+ 31.Kg1 Qe3+ 32.Kf1 c3 33.Rhh2 Ba6 34.Rhe2 Qxe2+ 0-1


16...Bxe4! 17.fxe4 Nxe4+ 18.Kg1 Bd2! 19.Bxd2 Nbxd2 20.Bg2 Qg5 21.Nf4 Qf6!=

13...Nxe4!? 14.fxe4 Qh4+ 15.g3 Bb4+ 16.Kf2 Qxe4. White does retain a material advantage with 17.Bh3!, but Blacks
counterplay is not easy to tame: 17...c3 18.Nd3 Nc4 19.Qe2 cxb2 20.Nxb4 bxa1=Q 21.Rxa1 h5 22.Nd3 h4 23.g4 Qc6
24.Ne1 Qd6 25.Kg1 0-0 26.Bf2 Rc8 27.Bg2 Bxg2 28.Nxg2 b4.
For his part, White commonly targets the c4- and the b5-pawns:


16.Nxb5! Qxb5 17.Qc2 Na5 18.Qa4 and the rook is stronger than the uncoordinated minor pieces.

15.Nxb5! Qxb5 16.Bxc4.

Chapter 3. 3.e4 b5?!.

Step by Step

1.d4 d5 2.c4 dxc4 3.e4 b5?! 4.a4 c6


5.Nc3 has been neutralised in the last years. This position commonly arises after 2...c6 3.Nc3 dxc4 4.e4 b5 5.a4. Now
5...a6 transposes to the main line, but 5...b4! is objectively better. In response, White has tried all the retreats:

a) 6.Nce2 e6 7.Nf3 Ba6 8.Ng3 c5 with a fine game. Baldauf-Felgaer, Graz 2014, shows that the pawn advance is not
dangerous 9.d5 exd5 10.exd5 Nf6 11.Bg5 Qxd5 12.Qxd5 Nxd5 13.0-0-0 h6 14.Re1+ Be7 15.Bxe7 Nxe7 16.Nf5
Nbc6 17.Re4= (17.Nd6+ Kf8).

b) 6.Nb1 Ba6 7.Qc2 Nf6

White has good compensation for the pawn after 7...Qxd4 8.Nf3 Qb6 9.Be3.
8.Nd2 e6 9.Bxc4 Bxc4 10.Qxc4 c5!
The simplest solution. 11.dxc5 Nbd7 12.Nb3 Rc8 13.Be3 Nxc5 14.Nxc5 Bxc5 15.Bxc5 Nd7 16.Rd1 Rxc5 17.Qd4 b3=,
Jakovenko-Najer, Yerevan 2014.

c) 6.Na2 The main move. By attacking simultaneously two pawns, White evens the material, but at the cost of the d5-
square 6...Nf6 7.e5 Nd5 8.Bxc4 e6 9.Nf3 a5 10.Bg5 Qb6 11.Nc1 Ba6

Nothing can stop Black from reaching this position. His next step is to prepare ...c6-c5 and hit the a4-pawn. 12.Qe2
12.Bxa6 is even more easy as White has to trade queens 12...Qxa6 13.Nb3 Nd7 14.Qe2 (14.Rc1 h6 15.Be3 Be7
16.Qe2 N7b6 17.Qxa6 Rxa6 18.Rxc6 Kd7 19.Rc2 Nxa4) 14...c5 15.Qxa6 Rxa6 16.Rc1 h6 17.Be3 c4! 18.Rxc4
N7b6 19.Rc6 Kd7 20.Rc2 Nxa4 21.Ke2 Nab6 22.Ra1 a4=.
12...h6 13.Be3 Bxc4 14.Qxc4 Nd7 15.Nb3 Be7 16.0-0 0-0 17.Rfc1 Rfc8 18.Rc2 c5 19.Rac1
I have been following the game Hera-Balogh, Austria 2011. Black can always take on e3, but perhaps the simplest is
19...Qd8=, threatening N7b6.

5...cxb5 6.Nc3 a6 7.Nxb5 axb5 8.Rxa8 Bb7 9.Ra1

9.Ra2 also deserves attention. So far it has occurred only in one practical game, Hambleton-Ganguly, Edmonton 2015:
9...e6 10.f3 Nc6
10...Bb4+ 11.Bd2 (11.Kf2 Bc5 12.Be3 e5 13.Ne2 Nc6 14.b3 exd4 15.Nxd4 f5 16.exf5 Bxd4 17.Bxd4 Nxd4
18.Rd2 Ne7 19.Rxd4 Qb6 20.Ke1 0-0) 11...Nc6 12.d5! (12.Ne2 Nge7) 12...exd5 13.exd5 Qh4+ 14.g3 Qe7+

Or 11.b3!? Bb4+ (11...Nxd4 12.bxc4 Bb4+ 13.Bd2) 12.Kf2 Bc5 13.bxc4 Bxd4+ 14.Ke1 b4.
11...Bb4+ 12.Bd2 Nge7 13.b3 0-0 14.bxc4 bxc4, when 15.Qc1 Qb6 16.Qxc4 Rc8 17.Nf4 h6 18.Nd3 Bxd2+ 19.Kxd2
would be in Whites favour.


A very topical position lately. Whites most challenging continuations are: A. 10.f3; B. 10.Ne2.
They can both lead to the same position if Black answers 10.f3 with 10...Nf6.
10.Be2 is in no way a mistake, either. White practically plays with a draw in the pocket making the most natural moves:
10...Nf6 11.Nf3 Bb4+! 12.Bd2 Bxd2+ 13.Qxd2 Nxe4

His main problem is that he will be often getting positions with an extra exchange, but only 3:4 pawns on the same side
which should be a technical draw.
See the annotations to game 8 Howell-Das, Leiden 2015.

A. 10.f3 f5

Or 10...Nc6 11.Ne2! with the following branches:

a) 11...Nf6 transposes to line B.

b) 11...Bb4+ 12.Bd2 Nge7 13.b3 0-0 14.bxc4 bxc4

Black has castled and he is ready to strike at the centre with ...f5 or ...Qb6, ...Rd8, but his c4-pawn is very weak:
In Javakhishvili-Edouard, Caleta 2015, White blundered here with 15.Kf2? f5 16.Nf4? Qxd4+ 17.Be3 Qb2+
18.Be2 fxe4;
15.Bxb4 Nxb4 16.Nc3 Nec6 17.Bxc4 Nxd4 18.0-0 Nbc2 19.Nb5 Nxa1 20.Nxd4 Qb6 21.Qxa1 Ra8 22.Qc3 Rd8
23.Rd1 e5 24.Qb3 wins a pawn in a an endgame, but Black should be able to hold the draw;
15.Rb1 Qa5 16.Qc1 Rb8 17.Bc3 Bc8 is unclear.
15...Qb6 16.Rb1 Ba6 17.Kf2 Rb8 18.Ng3.

c) 11...Na5 12.Nc3!
Attacking b5. 12.Be3 Nb3 13.Ra2 Bb4+ 14.Nc3 Ne7! is less precise.
12...Qb6? stumbles into 13.Nxb5 Qxb5 14.Qa4 Qxa4 15.Rxa4 Nc6 16.Bxc4! with the better endgame, e.g.
16...Nxd4 (16...Nge7 17.Bb5) 17.Kf2 Bc5 18.Be3.
12...Nb3? 13.Nxb5 Nxa1 14.Bxc4.
13.Kf2 (Reviving the threat on b5.) 13...Qb6 (13...Nb3 14.Nxb5) 14.Be3 Nb3 15.Ra2 Ne7
15...f5 allows the same combination 16.Nxb5 Qxb5 17.Qc2 Na5 18.Qa4 Bc6 19.Qxb5 Bxb5 20.b3 with a
decisive advantage, e.g. 20...Nxb3 21.Ra8+ Kf7 22.Rb8.
16.Nxb5! Qxb5 17.Qc2 0-0
17...Ba6 18.Qxc4 Qxc4 19.Bxc4 Bxc4 20.Ra8+ Kd7 21.Rxh8.
18.Qxc4 Qb6 19.Qb5. White trades queens and remains with 2 pawns and active rooks against 2 helpless knights. It is
unlikely to see candidates to defend Blacks side.


White might be better after 11.exf5, but this capture is more risky than the text since Black can put a knight on f5 after
11...Ne7 or 11...Nc6 12.exf5 Nge7.


After 11...Nf6, we should probably attempt to transpose by 12.Ng5 (12.Be2 Nc6 is unclear) 12...fxe4 13.fxe4. The only
deviation I see is 12...Qb6 when we have no choice, but take on f5 13.exf5 exf5 14.Qe2+ Kd7 15.Nf7 Bb4+ 16.Bd2
Bxd2+ 17.Qxd2 Re8+ 18.Ne5+ Kc7 19.Be2 Nd5. White has a clear advantage. It becomes obvious after 20.Rc1
(20.f4!?) 20...f4 21.b3 c3 22.Rxc3+ Nxc3 23.Qxc3+.

12.fxe4 Nf6 13.Ng5 Qb6

13...Bb4+ 14.Bd2 Bxd2+ 15.Qxd2 occurred in Shirov-Sulskis, Suomi 2015. Black gambled with 15...0-0? 16.Nxe6
Qe7 17.Nxf8 Nxe4 18.Qc2 and it turned out that Whites material advantage was too big.
15...Nxe4 is a relatively better try and White is only slightly better after 16.Nxe4 Bxe4 17.Qf4 Bd5 18.Be2 Rf8 19.Qg3
g6 20.Rf1 Nc6 21.Rxf8+ Kxf8 22.Bf3 Kg7 23.Bxd5 exd5 (23...Qxd5 24.Qc7+ Kf6 25.Kf1) 24.Qf4.


Development has a higher priority than material in this variation. White fell under attack after 14.e5?! Ne4 15.Nxe4
Bxe4 16.Be3? Bb4+ 17.Kf2 0-0+ in Kovalenko-Heberla, Katowice 2015.

14...Nxe4 15.0-0.

B. 10.Ne2 Nf6

10...Bxe4 is the acid test of 10.Ne2, but it offers White tempi for development 11.Nc3 (11.b3 Nc6!)

11...Bc6 12.Be2 b4 (12...Bxg2 13.Rg1 Bc6 14.Bf3 turns the tables and it is already White who is attacking.)
13.Nb1 Bxg2 14.Rg1 Bc6 15.Nd2 c3 16.bxc3 bxc3 17.Nc4. The knight is heading for e5 when Black suddenly
finds it difficult to defend the a4-e8 diagonal. 17.Nf3 achieves the same effect.
12.Be2 Nf6
12...Bxg2 13.Rg1 Bc6 14.Rxg7 Ne7 15.Kf1 is in Whites favour according to Morozevich.
White has castled and he has a small material advantage.

11.f3 Nc6 12.Be3 Bb4+

After 12...Na5 13.Nc3, White may countersacrifice material 13...Bb4 (13...Nb3?! fails to 14.Ra7 Qb6 15.Nxb5!
Qxb5 16.Bxc4) 14.Be2 Nb3 when besides 15.Ra2, White has even 15.0-0!? Nxa1 16.Qxa1 with a clear positional
advantage since the b5-pawn is weak 16...Qb6 17.b3 0-0 18.bxc4 bxc4 19.Na4.


13.Kf2!? 0-0 14.g3 is safe, but rather passive.

14.b3 Na5 15.bxc4 Nxc4 is only a draw since White cannot prevent a destructive sacrifice on e4, for instance,
16.Rb1 Nxe4+ or 16.Qb3 Bxe4!.
14...Na5 15.Qc2
15.Bh3 Nb3 16.Ra2 Bxe4!.
15...Ne8 16.Nc3 f5 17.Bh3 Nf6 18.Rad1 Nb3 (18...Qe8 is clearly inferior) 19.Nxb5 Nd5 20.exd5 Qxd5 21.Bg2 Qxb5
22.Rhf1 Ra8 23.Kg1 Ra2 offers Black some compensation.



Morozevichs suggestion.
14.Be2 Qb6 15.Qd2 Rd8 16.Rd1 e5 underlines the weakness of d4. Morozevich analyses:
17.d5 Bc5 18.Bxc5 Qxc5 19.f4 (19.Ra1 Nd4 20.Bd1 b4 21.Ne2 Ne6 22.Bc2 Bxd5 23.exd5 Nxd5) 19...exf4! 20.Qxf4
Nd4! 21.Rxd4 Qxd4 22.Qc7 Ra8 23.Qxb7 Ra1+ 24.Nd1 g5 25.Rf1 Qxb2 and:
17.dxe5 Rxd2 18.Bxb6 Rxb2 19.exf6 Bxc3+ 20.Kf2 Bxf6 (20...gxf6 21.Rd7 Ba6 22.Rhd1 b4 23.Rc7 Ne5 24.f4 Nd3+
25.Kf1 Nxf4 26.Bxc4 Bxc4+ 27.Rxc4 Rxg2 28.Be3 Rg4 should be a draw.) 21.Rb1 Ra2 22.Rxb5 Ba6 23.Rbb1 c3
24.Rhe1 c2 25.Rbc1 Bc3 26.Bxa6 Rxa6 27.Re2 Rxb6 28.Rexc2 Bd4+ 29.Ke2 Nd8 30.Rd1 Bf6 31.Rc8 h5! and this is a
theoretical draw.
The surprising king move unpins the c3-knight. The f1-bishop hopes to pop up on h3. Now that White has defined the
place of his king, Black must seek counterplay with ...f5.


Another critical line is 14...Nd7 15.Nxb5 f5 16.exf5 Rxf5 17.Kf2

Black has an unpleasant choice:

a) 17...Rxb5 18.Bxc4 Rh5!? 19.Bxe6+ Kh8 20.d5 Nce5 21.f4 Nf6 22.fxe5 Rxe5 23.Rf1 Nxd5 24.Bxd5 Rxd5 25.Qe2
Re5 26.Rfc1! h6 27.Ra7 Qd5 28.Rxb7 Qxb7 29.Kg1 with an extra pawn.

b) 17...Nf6 18.Bxc4 Ng4+ 19.Ke2 Nxe3 20.Kxe3 Rxb5

Or 20...Rg5 21.Kf2 Qb6 22.Re1 Bxe1+ 23.Qxe1 Re5 24.Be2 Nxd4 25.Nxd4 Qxd4+ 26.Kf1.
21.Bxb5 Qg5+ 22.Kf2 Qxb5 23.Qe2! Qd5 24.Rhc1 White has the upper hand. 24...Qxd4+ 25.Qe3 Qd5 26.Rd1 Bc5
27.Rxd5 Bxe3+ 28.Kxe3 exd5

I fiddled with this endgame for a while and decided that a computer may be able to save it, but it is nevertheless quite
difficult for Black:
29.b4 Kf8 30.b5 Nd8 31.f4 Ke7 32.Rc1 Kd6 33.f5 Nf7 34.b6 Bc6 35.Rc3 Nd8 36.g4 Ke5 37.Rc5 h6.

15.g3!? f5 16.d5 Ne7 17.dxe6 Qb8 18.Ba7 Qe5 19.Qd4 Qxe6 20.Bh3
A funny position. Whites king feels safe in the middle while Black seems unable to create any threat. 20...Qf7 would
only cede more space after 21.Qe5 Bc8 22.Rhf1.

20...Nf6 21.Bc5 Nc6 22.Qe3 Bxc5 23.Qxc5 Nd7 24.Qe3 Qf6 25.Rad1.
Chapter 3. 3.e4 b5?!

Annotated Games

8. Howell 2698 Das 2457

Leiden 24.07.2015

1.d4 d5 2.c4 dxc4 3.e4 b5 4.a4 c6 5.Nc3 a6 6.axb5 cxb5 7.Nxb5 axb5 8.Rxa8 Bb7 9.Ra1 e6 10.Be2 Nf6 11.Nf3


This game puts to the question this capture. Critical is 11...Bb4+! 12.Bd2 Bxd2+ 13.Qxd2 Nxe4 14.Qe3
Mamedyarov-Kasimdzhanov, Baku 2014, saw 14.Qb4 Qd5 15.Ra5 (15.0-0 Nc6 16.Qa3 Nd6) 15...Nd6 and
White was unable to extract anything from his extra exchange.
14...Bd5 15.b3
White gets a symbolic advantage, but it should be unconvertible since sooner or later Black will reach a 4:3 pawn
formation without any weakness:
15...0-0 16.bxc4 bxc4 17.0-0 Nd6 or
15...cxb3 16.Bxb5+ Nd7 17.0-0 0-0 18.Ba4 b2 19.Rab1 Nb6 20.Bb3 Nd6 21.Nd2 Nf5 22.Qd3 Bc6 23.Rxb2 Nxd4.
On the ground of these variations, I suppose that the 10.Be2 line will remain the choice of players who want to ensure
the draw in the first place.

12.0-0 Nc6

Whits only hope to win the game is to make good use of his lead in development so he must rush to open files.
12...Qd5 only helps him in that respect due to 13.b3! Nc3 (13...cxb3 14.Qd3 Nd6 15.Ba3 Nc6 16.Bc5) 14.bxc4 bxc4
15.Qd2 Nxe2+ 16.Qxe2 Bd6 17.Ba3.


13.Be3 was played previously, but 13...Qd5 14.b3 cxb3 is roughly equal. I repeat, Whites advantage is not in his extra
exchange, but in the vulnerable stands of Blacks pieces. One mundane move, and Black will consolidate. The text
busts his queenside.
Perhaps the best defence is 13...Nc3 14.Qc2 Nxe2+ 15.Qxe2 Nxd4 16.Nxd4 Qxd4 17.Bb2 Qd5, but 18.Qg4 h5
19.Qg3 retains an initiative. It is important that 19...Bd6 20.bxc4 bxc4 21.Qxg7 Bf8 22.Qg3 Bd6 does not force a draw
in view of 23.Rfd1.

13...Na5 14.Bxc4?

It is typical for White to sacrifice a piece in order to take over the initiative. For instance, most endgames with a
rook+pawn vs. two minor pieces are much better for him due to the raging rook and unstable black pieces. However,
the simple 14.bxc4 Nc3 15.Bg5! Qb8 (15...f6 16.Qe1) 16.Qd3 Nxe2+ 17.Qxe2 Nxc4 18.d5 would have been much
more unpleasant than the text.

14...bxc4 15.bxc4


It is not easy to calculate that 15...Be7 16.Qa4+ Nc6 17.Re1 f5 18.Rb1 Qa5! (18...Qd7 19.Rxb7 Qxb7 20.Ne5) 19.Qxa5
Nxa5 does not lose a piece. For instance: 20.Ne5 0-0 21.f3 Nc3 22.Rb6 Bd8 23.Rb2 Bh4 24.g3 Bf6 25.Bd2 Bxe5
26.Bxc3 Bxf3 27.dxe5 Nxc4=.

16.Qd3 Bb4 17.d5 Nb3?

The decisive mistake. It was high time to give back a piece with 17...0-0 18.Ba3 Bxa3 19.Rxa3 Nxd5 20.cxd5 Bxd5
21.Ne5 and Black is rock-solid.

18.Ba3 Nc5 19.Qe3 Qa5 20.Bb2 N3a4 21.Bxg7 Rg8 22.Bd4+- Rg4 23.Rfb1 Re4 24.Qg5 Rxd4 25.Nxd4 Bc3
26.Qg8+ Ke7 27.Rxb7+ 1-0

Part 2
The Classical System

Nowadays the natural development 1.d4 d5 2.c4 dxc4 3.Nf3 Nf6 4.e3 e6 5.Bxc4 is living through a crisis. Whites
nightmare begins not later than on move 3 when 3...a6!? 4.e3 b5 (or 4...Bg4) throws him out of his repertoire. I
personally prefer the more challenging move order 3...Nf6 4.e3 Bg4. Even if he reached the main line after 3...Nf6 4.e3
e6 5.Bxc4 c5 6.0-0 a6, it would not promise him an advantage. His traditional weapons of choice, 7.Bb3 and 7.a4, are
convincingly neutralised. Thus the Whites only hope seems to be the Exchange Variation 7.dxc5. I admit that this
endgame could be really boring and Black must know by heart 2-3 important lines in order to escape long-term bind. I
tried to facilitate your task by cutting off half of the theory (I consider only 7...Bxc5, omitting the similar 7...Qxd1+
altogether). I have further filtered the abundant material to two main set-ups for Black, depending on Whites play
after 8.Qxd8+ Kxd8, youll only have to learn 9.Ne5 Ke7 10.Be2 Nbd7 and 9.Nbd2 Ke7 10.Be2 Bd7. In several lines,
I suggest ways to take over the initiative by pawn sacrifices.
The current state of this variation is reassuring for Black I have not seen any game which should be a cause of concern
to us.
Chapter 4. The Exchange Variation

Main Ideas

1.d4 d5 2.c4 dxc4 3.Nf3 Nf6 4.e3 e6 5.Bxc4 c5 6.0-0 a6 7.dxc5

The queenless middlegame is Whites most unpleasant weapon against the Classical System. Although first players
score more convincingly in nearly all the other major systems, our chances of winning a full point against a well
prepared opponent are minimal. Thats why some die-hard QGA fans often opt for alternative development with an
early ...b5 before ...c5. I do not approve of such an approach. The QGA is a solid variation which allows Black to play
chess without having to study too much theory. If you want to complicate things as early as possible, you should be
prepared to suffer in more irregular openings. Besides, maintaining the balance in an equal position is not as simple
as it may seem.
As a rule, play is not forced and both sides have to choose on every turn between multiple reasonable continuations.
That makes calculation difficult because there are too many sub-branches. Therefore, we have to rely more on strategic
factors and correct evaluations than on concrete move-by-move analysis.

I will attempt to offer you a straightforward set-up which nips Whites most dangerous ideas in the bud. For that, we
have to be well acquainted with them. Here is one mostly instructive game:

Gupta 2625 Gevorgyan 2182

Dubai 2015

1.d4 d5 2.c4 dxc4 3.Nf3 Nf6 4.e3 e6 5.Bxc4 a6 6.0-0 c5 7.dxc5 Qxd1 8.Rxd1 Bxc5 9.Be2 b5?!

First sin an early ...b5. It weakens simultaneously: the squares a5, c5, c6, a6, the long light-squared diagonal, the
whole queenside in short.

10.a4! b4 11.Nbd2 a5?

Second sin instead of developing, Black loses a tempo on another pawn move.

12.Nb3 Bb6 13.Ne5 0-0 14.Bf3 Ra7 15.Nc4

Look at the stand of Whites knights. They dominate Blacks dark-squared bishop from b3 and c4 (you can also see
them on d3-c4 in other games). As a bonus, the f3-bishop is ruling over the queenside.

15...Bd8 16.e4!

The last nail in Blacks coffin. White gains even more space and undermines the only stable black piece the f6-knight.
But the most important is that the queens bishop enters play from e3 with a killing effect. The game is over in just 16
16...Bb7 17.Be3 Ra8 18.Bc5 etc. 1-0 on move 25.

How to escape this positional vice?

1. Do not castle! Keep the king in the centre by:

7...Bxc5 8.Qxd8+ Kxd8

This way well be a tempo ahead in the event of mass exchanges along the open files.

2. Do not play ...b5!

This extended fianchetto is only good against a3-b4 from Whites side. Wise people say we should not move pawns
where we are weak. Instead we should aim to develop our bishop to d7 refraining from making a hole on c6. This is
closely connected with the next item on our to-do list.

3. Develop quickly, aiming to bring the kings rook to d8! Most Blacks troubles come from delayed development.
9.Nbd2 Ke7 10.Be2 Bd7
The point of our set-up is that we can meet 11.Ne5, 11.b3 and 11.a3 by 11...Bb5. That would solve one of Blacks most
difficult problem how to activate his light-squared bishop. You should not be afraid of the weak pawn on b5. It
controls the critical c4-square and opens the a-file. Besides, it can always move forward where it will be well protected.
For instance, in the following position which arises after 11.Ne5 Bb5 12.Bxb5? axb5,

White cannot even equalise: 13.Nb3 Bd6 14.Nf3 (c4 is forbidden!) 14...Nc6 15.Nfd4 b4 when it is already White to
worry about the future of his bishop.

11.Nb3 Bb6 12.Bd2 Nc6

Typical Position 1
You can see a detailed analysis of this position in the Step by Step section. Black will play ...Ne4, ...e5, ...Nd6. If
13.Bc3, then 13...Rhg8 unpins the knight and enables the above-mentioned set-up.

Theoretical Status

I recall two major White repertoire books where the authors rely on the Exchange System. Both prefer to discuss
irrelevant variations while timidly evading the most topical lines.
Jonathan Hilton and Dean Ippolito fleetingly mention in Wojos Weapons that White gets only a slight advantage at
best after 11...Ba7 while 11...Bb6 is granted a !? sign, but then they cite Gelfand-Shirov which saw 13.Bc3 Rhd8
etc. Neither a word about Topalovs 13...Rg8, nor an explanation why 11...Bb6 deserved attention.
Scherbakov in claims that White still kept some edge after 13...Rhg8 14.Rfd1 Rac8 as in Kozul-
Topalov without further elaboration. This statement is strange since Topalov did not face even a trace of difficulties in
that game and Scherbakov does not suggest any improvement of Whites play.
Thus we can say that Typical Position 1 is practically unexplored. I discuss it in the Step by Step section.

Kornev in A Practical White Repertoire with 1.d4 and 2.c4, vol.1 stands up for:

9.Ne5 Ke7 10.Be2 Nbd7 11.Nc4

Hilton and Ippolito also claim that this line is Whites best shot at a true advantage.
However, both books completely ignore the relatively new stand with ...b6, ...a5, ...Ba6. Both 11...b6 12.Nbd2 a5
13.Nb3 Ba6 14.Nxc5 bxc5 and 11...a5 12.Nc3 b6 13.Na4 Ba6 14.Nxc5 bxc5 lead to the same position:

Typical Position 2

Four players with an average rating of nearly 2700 scored only 1.5 point with White from here. If we add the drawn
game 10 Melkumyan-Laznicka, Legnica 2013, in which White obtained the best possible version of this pawn
structure, Blacks performance becomes even more convincing. The c5-pawn is extremely strong. It effectively cuts the
scope of Whites long-range pieces and ensures some space advantage in the centre. Blacks game is very easy since all
his pieces can manoeuvre freely.

Let me also mention the other retreat: 11.Nd3 Bd6 12.Nd2 Nc5 followed up by ...Rb8 and ...b6.

Important Decisions in the Opening

Many White players believe that their task in the Exchange system should be to chase the enemy bishops with their
knights until trading one of them. Then the bishop pair should assure them of a comfortable and stable edge. Black, for
his turn, should spare no effort to escape this scenario.
In fact, that is a wrong concept. We should think in concrete terms and assess all the positional factors in their integrity:
piece activity, space control, available targets. Here are some examples which should clarify my standpoint.


White has achieved his goal. His knight had landed on c5 and Black had to give up his bishop for it. Does that
promise White any advantage? Nope! Blacks pieces are compact and well coordinated, the e5-pawn is clogging both
the f3-knight and the e1-bishop which does not have other than defensive functions. Play may continue 18.Nd2 Nb4=.

Bu Xiangzhi Sengupta
Sharjah 2014
13...Bxc4! 14.Bxc4 Nc6. Here Black does not have concrete threats (yet!), but several factors are in his favour. First of
all, he is at least two tempi ahead in development. Second, the c1-bishop will remain passive for long, because Black
has a firm grip on e4. And third, the symmetrical pawn structure without weaknesses neutralises a great deal of the
bishops power. The game went 15.Rd1 Rhd8 16.Bd2 Ne4 17.Be1 g6 18.Kf1 Rxd1 19.Rxd1 Rd8 20.Rc1 when best is
20...Ne5 21.Be2 f5=.

Gelfand Shirov
blindfold, Monte Carlo 2001

In this position, on the contrary, Whites pieces are much more active. After 17.Nba5 Rab8 18.Nxc6+ Bxc6 19.Na5,
Blacks dark squares are weak. Compare the c3-bishop with that on e1 from the previous examples!

Now 17.Rac1 Nd5 18.Bd2 b6! would equalise, but White can part with both his bishops to enforce his domination all
over the board:
17.Bxf6+ Kxf6 18.Rfd1 Ke7 19.Bf3 a5 20.Bxc6 Rxd1+ 21.Rxd1 bxc6 22.f3.

I repeat once again: the main factor is piece activity. We should not aim for exchanges at all cost. It is often better to
shed a pawn for getting control on the open files:

Kozul Kveinys
Nova Gorica 2004

Kveynis opted for the passive 17...Nc5 to struggle after 18.Nxc5 Rxc5 19.Rxc5 Bxc5 20.Rc1.
Instead he had 17...Ne4! 18.Bxg7 f6 19.Bh6 Ba3 with compensation.

Here are two more examples of the same method:


14...Ne4!? when no matter which pawn White takes, he cannot easily disentangle his clumsy pieces (15.axb5 Bc3 or
15.Bxg7 Rhg8).


12...Bb7! 13.axb5 axb5 14.Rxa8 Rxa8 15.Bxb5 Ra1 and Black takes over the initiative, e.g. 16.Bd3 Bd6 17.Bb2 Ra2
18.Bd4 e5 19.Bc3 e4 20.Bc4 Nb6.

Middlegame Plans

Lets assume that the opening is behind, the position is symmetrical and more or less balanced, and both sides do not
have immediate threats.

Carlsen Nisipeanu
Medias 2011

White has a good version of the bishop pair, but no targets. His further task should be to provoke a weakness or extend
the scope of his bishops. Black, for his part, should seek to trade rooks. Without them, most endings are drawn. So
Carlsen refrained from putting his second rook on an open file, but started a kingside advance with 18.h3. Look at
game 11.
Note that if Black had an initiative himself, everything said for White would have been true for him, too, due to the
symmetrical pawn structure.

Another typical plan of White is to build up a mole against a fianchettoed light-squared-bishop under the shape of e4-
f3 pawns.

Aronian Karjakin
blitz, Zurich 2015
It is unpleasant to defend such positions. Should it happen, however, Black must seek to undermine the e4-pawn with
...f5, for instance: 22...Nd7! 23.Ke3 f5!.

Of course, Black could pursue the same goal:

Lokander Shalamberidze
Stockholm 2015

Black kept some pull after 21...Rc7 22.Nfd2 Rac8 23.Ne4 Rc2 24.g4 R8c4 25.Rxc2 Rxc2 26.Rd2 Rc4.
Chapter 4. The Exchange Variation
Step by Step

1.d4 d5 2.c4 dxc4 3.Nf3 Nf6 4.e3 e6 5.Bxc4 c5 6.0-0 a6 7.dxc5 Bxc5

7...Qxd1 assumes a different set-up for Black.

8.Qxd8+ Kxd8

Whites lead in development is transient. He should hurry to convert it in piece domination. The only way to set
concrete problems to Black is to put knights on c4 (via d2) and d3 (via e5 or e1). Then we might have to trade our dark-
squared bishop.

Main continuations are:

A. 9.Nbd2; B. 9.Ne5

9.Be2 Ke7 should transpose to the other lines.

Sometimes White fianchettoes his queens bishop, but this approach is completely harmless:

a) 9.b3 Ke7 10.Be2!?

Waiting to see what set-up Black will choose.
10.Bb2 Nbd7 11.Be2 b6 12.Nbd2 Rd8 transposes.
10...Bd7 would be pointless since White can answer 11.Nc3, depriving our bishop from the b5-square.
10...b6 11.Ne5 Bb7 12.Nd2, threatening Nd3, is a bit awkward.
If White tries to build the set-up from the previous note with 11.Ne1, we can exploit the temporary lack of
coordination between his pieces with 11...b5! 12.a4 Bb7 13.axb5 axb5 14.Rxa8 Rxa8 15.Bxb5 Ra1 and Black
takes over the initiative, e.g. 16.Bd3 Bd6 17.Bb2 Ra2 18.Bd4 e5 19.Bc3 e4 20.Bc4 Nb6.
11...b6 12.Nbd2 Bb7
12...Rd8!? is another possible move order. Its point is to enable ...Kf8 in the event of 13.Ne1 or 13.a3. It assumes
that the bishop would be more stable on e7 than on d6.
After 13.Ne1, 13...b5 is still good even though Black is a tempo down compared to the previous note. Then 14.a4
Bb4! 15.Ndf3 bxa4= has been tested, but instead Black can play sharper: 15...Ne4!? when no matter which pawn
White takes, he cannot easily disentangle his clumsy pieces (16.axb5 Bc3 or 16.Bxg7 Rhg8).
13...Rhd8. Black has successfully finished his development without allowing any concessions. See game 9 Karpov-
Ponomariov, Benidorm 2002.

b) 9.a3 Ke7
Whites set-up with b4 is the only one where I recommend ...b5. However, 9...b5 outright is probably not too
precise as White might reconsider and attack us with 10.a4.
10.b4 Bd6 11.Bb2 b5 12.Be2 Bb7 13.Nbd2 Nbd7
Black will follow up with ...e6-e5 aiming to gain space on the kingside. See game 12 Maki Uuro-Baburin, Fuegen 2006.

A. 9.Nbd2 Ke7 10.Be2

10.b3 Nbd7, followed by ...b6 transposes to 9.b3.

10.Ne5 does not make any sense 10...Nbd7 11.Nd3 Bd6 and White should switch to defence with 12.a4 in view of the
awkward position of the c4-bishop.


White had not played Ne5 earlier so why not exploit that by developing the bishop. Our idea is to meet 11.Ne5 by


11.b3 has more venom than on move 9 as Black has already committed his bishop to d7. The critical position then arises
after 11...Bb5 12.Nc4 Nc6 13.Bb2 Rhd8 14.Rfc1 Rac8 15.a4 Bxc4 16.Rxc4 Na5 17.Rh4 (17.Rxc5 Rxc5 18.Ba3 Ke8
19.Bxc5 Nxb3) 17...Kf8 18.Bxf6 gxf6 19.Bd1 Be7 20.Rxh7 Kg8 with a nice compensation for the pawn.

11.a3 can be met in the same fashion by 11...Bb5 12.Bxb5 axb5 13.Nb3 b6, followed up by ...Na6 and possibly ...b4.
(13...Bd6 14.Nfd4 b4 15.Nb5 Nc6 keeping the symmetry is also possible.)


I do not see any advantages of 11...Ba7 over the text so I choose to retreat to b6 from where the bishop controls a5.
Although 12.Bd2 Nc6 13.Na5 is harmless due to 13...Rhc8! 14.Nxc6+ (14.Nxb7 Rcb8=) 14...Bxc6 15.Ne5 Bb5
16.Bxb5 axb5, see game 13 Shirov-Motylev, Moscow 2001, the more passive stand on a7 might tell later in the
game. White should continue as in the main line:
12.Bd2 Nc6 13.Bc3! when Black will be choosing between 13...Rhd8 14.Rfc1 and 13...Rhg8 as played by Topalov in a
similar position. In that event the game may continue 14.Rfc1 (14.Rfd1 Nd5 15.Ne5 Nxe5 16.Bxe5 Ba4 17.Bf3 f6
18.Bxd5 exd5=) 14...Nd5 15.Bd2 Rac8 16.Nc5 Bxc5 17.Rxc5 f5.

The only argument against 11...Bb6 could be the line 12.Ne5, but well see that it does not pose any problems.


12.Ne5 Bb5 13.Nc4 Bxc4 14.Bxc4 Nc6 is an example where Whites bishop pair is innocuous because Black has a lead
in development and stable stands for his knights.

Bu Xiangzhi-Sengupta, Sharjah 2014, went on 15.Rd1 Rhd8 (exchange every rook you see!) 16.Bd2 Ne4 17.Be1 g6
18.Kf1 Rxd1 19.Rxd1 Rd8 20.Rc1 when simplest would be 20...Ne5 21.Be2 f5=.


Perhaps the most important position for our repertoire against 7.dxc5. If we count the tempi, well see that in terms of
development both sides are even. White does puts a rook on the open files first, but that is compensated by the fact that
our king is two(!) squares closer to the centre. More troublesome is the question how to neutralise a possible jump of
the b3-knight to c5. We must understand when it would be good to eat it with our bishop, and when it would be better
to live with it.


Obviously, whites bishop would be best placed on e1 out of reach of the f6-knight and not crowding the c-file.
However, the straight 13.Rfc1 unleashes the knight: 13...Ne4 14.Be1 f6 15.Bd3 Nd6 16.Nc5 Bxc5 17.Rxc5 e5=
18.Nd2 Nb4 19.Be2 Rac8.

Similarly, 13.Rfd1 Ne4 14.Be1 e5 (14...f6=) leaves Black even more active: 15.Nfd2 Nxd2 16.Nxd2 Rhd8 17.Nc4 Bc7
18.a4 Be6 19.Rxd8 Rxd8 20.b4 e4 21.b5 axb5 22.axb5 Bxc4! 23.Bxc4 Ne5.

13.Bc3 aims to drag the h8-rook to g8 by pinning the f6-knight to the g7-pawn. Then, after Black plays ...Ne4, White
will have Bd3, eyeing the h7-pawn.


Played by Topalov against Kozul. It is aimed against the manoeuvre 14.Nd2 which would now fail to 14...Nd5.
Shirov chose to save a tempo with:
13...Rhd8, but after 14.Nfd2 (14.Rfd1 Rac8 15.Nfd2 allows 15...Nd5!? 16.Bxg7 Rg8 with enough compensation),
he became clearly worse following 14...Be8?! 15.Nc4 Bc7 16.Rac1 g6 17.Nba5 Rab8 18.Nxc6+ Bxc6 when
19.Na5 would have fixed a clear edge. Besides, a retreat to e8 would have made sense if White had a rook on d1
to trade.
He should have opted for:
14...g5 unpinning the f6-knight and ensuring the f4-square in the event of e3-e4 after ...Nd5. Play may continue:
15.Nc4 Bc7 16.Nc5
Alternatives with Nca5 lead to full equality:
16.Rac1 Nd5 17.Bd2 b5 18.Nca5 Nxa5 19.Bxa5 Bxa5 20.Nxa5 Rac8 21.Nb7 Rxc1 22.Rxc1 Rc8 23.Rxc8 Bxc8
24.Nc5 e5 25.a3 f5 26.b4 Kd6;
16.Nca5 Nxa5 17.Nxa5 Nd5 18.Nxb7 Rdc8 19.Bd4 Bc6 20.Nc5 e5 21.Bc3 Nxc3 22.bxc3 Bd6.

Black was ready to regroup his pieces on the queenside: 17.Rac1 Nd5 18.Bd2 b6! (18...b5 19.Na3 Bd6
20.Rfd1!) 19.Nd3 (19.Ne4 Bb7 20.Nxg5 b5 21.Na3 Be5) 19...Bb7 20.Rfd1 Rac8= followed by ...b5, ...Ba8,
17...Kxf6 18.Rfd1 Ke7 19.Bf3 a5 20.Bxc6 Rxd1+ 21.Rxd1 bxc6 22.f3. Whites knights dominate the bishops and the
split queenside pawns will remain weak for good. Topalovs move offers Black an easier life.

14.Rfd1 Rac8

The point of 13.Bc3 is that 14...Ne4 15.Be1 e5 loses a pawn to 16.Bd3.

A critical position for this line.

15.Nfd2 is still bad due to 15...Nd5.
Kozul opted for the senseless 15.Ng5 which could only be justified after 15...h6?? 16.Nxf7. Topalov equalised with
15...Rgd8 16.Bf3 Be8. More ambitious was 15...Nd5!? 16.Bd2 h6 17.Ne4 Ne5 18.Rac1 when 18...f5! underlines the
fact that 19.Nec5 was not a threat due to 19...Bb5!.
I will focus on the more consistent:

15.Rac1 Ne4 16.Be1 e5

Advancing the central pawns looks positionally dubious since it gives a wider scope to Whites bishops. On the other
hand, Black gains space. He could also choose the safe and solid stand with e5-f6:
16...Nd6 17.Nc5 Bxc5 18.Rxc5 Rgd8 19.Nd2 e5 20.Nb3 f6
Now 21.Rcc1 Be6 22.Bf3 (22.Nc5 Bxa2 23.f4 b6 24.Nxa6 Bb3 25.Rd3 Nd4=) 22...Nc4, and 21.Rdc1 b6 22.R5c3
a5 are fine for Black. The only way to retain some pull is to open immediately the centre with:
21.f4! exf4
21...Be6 22.fxe5 Nxe5 23.Rxc8 Nxc8 24.Bb4+ Ke8 25.Rxd8+ Kxd8 26.Nc5 Bd5 27.e4 Bc6 28.Kf2 a5 29.Bc3
Ke7 30.Nd3 Nxd3+ 31.Bxd3.
22.exf4 Be6 23.Rc3 Nb5 24.Re3 Rxd1 25.Bxd1 Kf7 26.a4 Nd6 27.Nc5 Bd5 28.Rd3 Ne7. If we compare this position
with the main line, Blacks pieces are more passive although everything is protected.

17.Bd3 f5 18.h3

18.Bxe4?! fxe4 19.Ng5 Bg4 is awkward 20.Rd2 Bf5 21.Rc4 Na5.

18...Nd6 19.Nc5

19...e4 20.Nxd7 Kxd7 21.Bb1 Rgf8 22.Nd2 Ke7 23.a3 leaves the initiative to White.

20.Rxc5 e4 21.Bb1 Rgd8 22.Nd2 Be6

Blacks pieces are well coordinated which compensates the bishop pair advantage. The game is level.
White must hurry to undermine the centre with 23.f3 h6 24.Rdc1 exf3 25.Nxf3 Kf7 26.b3 Ne4 27.Bxe4 fxe4 28.Nd2
Bd5 29.Nc4

Now both 29...Bxc4 30.R5xc4 Re8 31.Kf2 Ne7 32.Rc7 Ke6 33.Bb4 Nc6 34.Rxc8 Rxc8 35.Rc4 Kd5 36.Kg3 Rd8
37.Ba3 Rd7 and 29...Be6 30.Nb6 Rc7 31.Re5 Rd1 32.Rxd1 Nxe5 33.Bg3 Rc6 are drawish.

B. 9.Ne5 Ke7 10.Be2

White shifts the bishop to the main diagonal and frees c4 for the knight at e5. The bishops retreat also discourages
10...b5 which could be attacked with 11.a4.

B1. 11.Nc4; B2. 11.Nd3

B1. 11.Nc4 b6

11...a5 12.Nc3 b6 13.Na4 Ba6 14.Nxc5 bxc5 leads to the same position.

12.Nbd2 a5 13.Nb3 Ba6 14.Nxc5 bxc5

Blacks play on the queenside is easy and straightforward:

a) 15.b3 a4= (15...Rhb8) 16.bxa4 Rhb8 17.Bd2 Ne4 18.Be1 Nd6 19.Rc1 Bxc4 20.Bxc4 Rxa4 21.Bb3 Ra6 22.f3 c4
23.Bxc4 Nxc4 draw, K.Georgiev-Laznicka, Pamplona 2009

b) 15.Bd2 Ne4 16.Rfd1 Ne5 17.Nxe5 Bxe2=, Giri-Dominguez Perez, Beijing 2013.

c) 15.f3 Komodo prefers this set-up and evaluates the position slightly in Whites favour, but the engine does not see
that White lacks an active plan. Black gets comfortable equality after:
15...Rhb8 (discouraging Bd2).
I see no reason to eliminate to a drawish ending with 15...Ne5.
In Rodshtein-Eljanov, Tsaghkadzor 2015, Black chose a wrong set-up: 15...Nd5 16.Bd2 N7b6 17.Rfc1 Nxc4
18.Bxc4 Rhc8 and lost a pawn after 19.Bxd5 exd5 20.Bxa5. This is a good illustration why Black needs a knight
on d7, but not on d5.
16.e4 a4 (threatening ...Rb4) 17.Rf2 Rb4 18.Ne3 Bxe2 19.Rxe2=

Now simplest is to activate the f6-knight by 19...Ne8 20.Rc2 Nd6 21.Bd2 Rb7 22.Bc3 f6 23.Rd1 Nb5.

See also game 10 Melkumyan-Laznicka EU-ch Legnica 2013 for details about this pawn structure.

B2. 11.Nd3 Bd6 12.Nd2 Nc5

12...b5 is more popular, but I do not consider set-ups with ...b5 if we have decent alternatives.

13.Nxc5 Bxc5 14.Nb3

14.a3 allows the manoeuvre ...Bc8-d7-b5, for instance: 14...Bd7 15.b4 (15.Bf3 Bb5 16.Rd1 Rab8 17.b4 Bd6 18.Bb2
Rhc8 19.Rac1 b6) 15...Bd6 16.Bf3 Bb5 17.Re1 Be5.
After 14.Bf3 e5 15.a3 Rd8 16.b4 Ba7, White does not have an obvious way of improving his pieces while Black could
advance on the kingside 17.h3 g5.

14...Bd6 15.Bf3

15.Bd2 is harmless, but you should avoid the following position:

15...Bd7? 16.Bf3 Rab8 17.Na5 b6 18.Nb7! (18.Nc6+=).
Safest is 15...Ne4 (or 15...b6 16.f3 Bd7 17.e4 Bb5) 16.Ba5 Bd7 17.Rfd1 (17.Bf3 Bc6 18.Rfc1 Bd5=) 17...Rhc8 18.Bf3
Bc6 19.Nd4 f5!
This is a typical way of keeping the long diagonal blocked. The pawn on e4 is not weak since Black is well coordinated
with all his pieces active and the king in the centre: 20.Nxc6+ Rxc6 21.Bxe4 fxe4=.

15...Rb8 16.Bd2

White cannot reach the position mentioned in the previous note because 16.Na5 will face 16...Bb4 17.Nc4 Bd7 18.a3


Ill repeat once again: 16...Bd7? 17.Na5! must be avoided.

Both 17...b6 18.Nb7! and 17...Nd5 18.e4! favour White.

However, 16...e5!? 17.Na5 e4 18.Bd1 Be6= is a fair alternative.

17.Nd4 Bb7=

White can get the bishop pair with 18.Nc6+ Bxc6 19.Bxc6, but Black is fine after 19...Rhc8 20.Rfc1 Nd5 or 20.Bf3 Be5
21.Rac1 Nd5.

Chapter 4. The Exchange Variation

Annotated Games

9. Karpov 2688 Ponomariov 2743

Benidorm 29.11.2002

1.d4 d5 2.c4 dxc4 3.Nf3 Nf6 4.e3 e6 5.Bxc4 c5 6.0-0 a6 7.dxc5 Bxc5 8.Qxd8+ Kxd8 9.Nbd2 Ke7 10.b3

The move order with Nbd2 + b3 has no venom. White simply develops his pieces and does not hinder us to do the same.
Note, however, that our main set-up with ...Bd7-b5 would be a bit slow here. It is effective only when White spent a
tempo on tricky manoeuvres like Bc4-e2 or Nf3-e5-d3.

10...Nbd7 11.Bb2 b6

This set-up is generally easier to defend than 11...b5, although the extended fianchetto is possible here. We commonly
refrain from it in order to avoid a future attack with a2-a4. However, this attack is more troublesome if Whites b-pawn
remains on b2 and the bishop on c1.

12.Be2 Bb7 13.Rfd1

Or 13.Ne1 Bb4!

13...Rhd8 14.Ne1

This excellent move concludes the opening battle in Blacks favour. Now the bishop has a safe retreat to e7, the g7-
pawn is protected so the f6-knight is unpinned. The game is completely balanced and even Black is a tempo ahead his
king being closer to the centre. The same idea was possible in the event of 14.a3.

15.Rac1 Rac8 16.Nd3 Be7 17.Nc4

Black could get this position with a clear tempo down from the line 7...Qxd1. This would give White time for 17.f3
when 17...b5?! is a typical strategic mistake. Blacks bishop was a poor sight after 18.Na5 Ba8 19.e4 Nc5 20.Ne5 and
he went to lose in Ernst-Van den Doel, Leeuwarden 2002, after 20...Ke8 21.Kf1 Rxd1+ 22.Bxd1 Bd8 23.b4 Bxa5
24.bxa5 Bb7 25.Bd4 Ncd7 26.Rxc8+ Bxc8 27.Nd3 etc. I would hate to struggle with a bishop which bites on e4-f3
pawns. Perhaps it is better to anticipate such situations and swiftly swap the bishop by ...Bb7-d5xc4 or ...Bb7-c6-b5xc4.
In the game, it is Black to move and the same bad move proves to be a decent equaliser. One more example that there
are no rules of thumb in chess! In the end, it all comes to concrete calculation.

17...b5! 18.Na5 Ba8 19.a3

Blacks pieces are too active and 19.f3 Nd5 20.e4 Ne3 would be promising for him. As to the a8-bishop, it could be
activated by ...f5. After the text, Ponomariov executes a mass elimination.

19...Nc5 20.Nxc5 Rxd1+ 21.Bxd1 Rxc5 22.b4 Rxc1 23.Bxc1 Ne4 24.Bb2 Bd8 25.f3 Nd6

White would have been better here had he a rook. Without it, Karpov takes the bold decision to give a pawn, but
preserve active pieces. The typical 26.e4 would have been a mistake due to 26...Bb6+ 27.Kf1 f5!.

26.Bd4 Bxa5 27.bxa5 Nc4 28.e4 Nxa5 29.a4 Bc6 30.Bc3 Nc4 31.axb5 Bxb5 32.Bb3 f6 33.Kf2 Ke7 34.Bxc4 Bxc4
35.Ke3 Kd7 36.f4 h5 37.Kd4 Bf1 38.g3 Draw.

The next game shows how White can fall in his own trap. In symmetrical positions, if White loses the initiative, he
easily risks to become the prey.

10. Melkumyan 2622 Laznicka 2679

EU-ch Legnica 14.05.2013

1.d4 d5 2.c4 dxc4 3.Nf3 Nf6 4.e3 e6 5.Bxc4 c5 6.0-0 a6 7.dxc5 Bxc5 8.Qxd8+ Kxd8 9.Be2 Nbd7 10.Ne1 b6 11.Nd3
a5 12.Nxc5 bxc5 13.Nc3
A very interesting pawn structure has arisen. It is important for our repertoire since it may arise after the moves 9.Ne5
Ke7 10.Be2 Nbd7 11.Nc4 b6 12.Nbd2 a5 13.Nb3 Ba6 14.Nxc5 bxc5

If we compare the two diagrams, we realise that the main difference is the placement of Whites knight on c4 instead of
c3. In my opinion, c3 is a much better place as the knight controls two critical squares from there e4 and a4. It can
also attack the isolated c5-pawn from a4. Finally, on c4 the knight is pinned. That allows the exchange operation
15.Bd2 Ne4 16.Rfd1 Ne5=.
Whites best try in the position of the last diagram may be 15.f3 when 15...Rhb8 16.e4 a4 17.Rf2 Rb4 18.Ne3 Bxe2
19.Rxe2. After the trade of the light-squared bishops, the game should enter a manoeuvring phase similar to the
annotated game, but with pawns on b2:a4 instead of b3:a5.

13...Ba6 14.Bxa6 Rxa6 15.b3 Ke7 16.e4

It is useful to take d5 away from the f6-knight. 16.Bb2 Rc8 17.Rac1 Nb6 18.a4 (18.Rc2 a4) 18...Nfd5 lets the knight to
b4, for instance: 19.Nb5 (19.Nxd5+ exd5 20.Bxg7 f6 21.Bh6 c4) 19...f6 20.Rfd1 (20.Ba3 Nb4) 20...Nb4 21.Rd6 (21.f3
c4) 21...c4.

16...Rc8 17.Ba3 Nb6 18.f3 Nfd7 19.Rfd1


The rest of the game is an illustration that White does not have any active plan. Although the engines stubbornly grant
him a slight edge, Melkumyan was obviously unable to generate any threat. Black did not even try to alter the pawn
structure with ...a4, although it would be a nice option after 20.Bb2 or 20.Nb5. Whites next move is aimed against that
push, but it was still possible: 20.Rab1 a4 21.b4 Nc4 22.bxc5 Nxc5 23.Bb4 Kf6=. It looks that Laznicka wanted to
prove that his position was fireproof and did not need any counterplay.

20.Rab1 Rc6

Black is planning to swap a pair of rooks with 21...Rd6. Only 21.Nb5 would prevent it, but then 21...a4 22.b4 Nc4 is

21.Ne2 Rd6 22.Rdc1 Rd2 23.Rb2 Rxb2 24.Bxb2 g6 25.a4 Rc8 26.Ba3 e5 27.Nc3 Ke6 28.Rd1 Rc6 29.Kf2 Rd6
Without rooks, the draw is inevitable.

30.Ke2 Rxd1 31.Kxd1 Kd6 32.Kc2 Kc6 33.Nb5 c4 34.Na7+ Kb7 35.Nb5 Kc6 36.Na7+ Kb7 37.Nb5 Kc6 38.Na7+
Kb7 39.Nb5 Kc6 40.Na7+ Kb7 Draw.

11. Carlsen 2815 Nisipeanu 2659

Medias 15.06.2011

1.d4 d5 2.c4 dxc4 3.e3 Nf6 4.Bxc4 e6 5.Nf3 c5 6.0-0 a6 7.dxc5 Bxc5 8.Qxd8+ Kxd8 9.Be2 Ke7 10.Nbd2 Bd7
11.Nb3 Bd6?!

This retreat (and 11...Ba7) invites the enemy knight on a5. Thats why I recommend 11...Bb6. Although that makes the
game irrelevant for our repertoire, it offers an instructive example of squeezing a small edge in a symmetrical pawns
structure. Remember that Black can also find himself in Whites shoes should his opponent lose a couple of tempi on
clumsy manoeuvres.

12.Na5! Ra7 13.Nc4 Bb5 14.b3 Rd8 15.Bb2 Bxc4 16.Bxc4 Nc6

Black had to give up his bishop, but this is not a big concession in symmetrical pawn structures. His real problem is the
lack of coordination between his pieces. Stayed his rook on c8 instead of a7, his position would have been satisfactory.
Now White preserves some initiative.

17.Rfd1 Raa8 18.h3

Every initiative is only temporary and it can easily evaporate. The bishop pair is a more stable factor, but Black does not
have any weaknesses so there is nothing to attack.

Carlsen embarks on a typical plan in such situation. He pushes his kingside pawns in the hope of provoking some defect
in the enemy pawn formation. Black should aim, for his part, to exchange material while keeping the structure
symmetrical. Thus, he should prefer to trade h-pawns, but not g-pawns because a pawn islet on the h-file would be a
more accessible target for the long-range bishops. Accordingly, 18...h5 would be the most natural reply although
18...h6 also seems adequate. Look at the famous game for the world title Kramnik-Kasparov, London 2000:
20.g4! h6 21.h4 Bc8! 22.g5 hxg5 23.hxg5 Nfd7! 24.f4 Ng6. Whites pawn advance looks daunting, but Kasparov does
not flinch and refrains from making any weaknesses. Now 25.Nd4 Rc7 26.Bh5 could be countered with 26...e5!.
Kramnik chose 25.Nf3 when simplest would have been 25...Rc7 26.f5 exf5 27.exf5 Nge5 28.Nxe5 Nxe5 29.Bf4 f6=.
Nisipeanus set-up with g6 weakens the dark squares and prolongs the main diagonal.

19.g4 h6 20.Bf1 Rac8 21.Rac1 Nd5 22.h4?!

This is too hasty. Carlsen underestimates Blacks counterplay on the other wing. Now 22...Ncb4 23.a3 Na2! would
have distracted White from the kingside. Thats why 22.Bd3 was more accurate, in order to meet 22...Ncb4 by 23.Bb1.

22...Ke8? 23.g5 hxg5

The open h-file finally costs Black the point, but 23...h5 looks an even more difficult decision. Putting all the pawns on
light squares makes them permanent targets. At some point White will organise the break f4-f5. To take stock, Blacks
last seven moves were a total failure. He has not traded even a pair of rooks, nor a piece, and has allowed his pawns to
be fixed on light squares. It is nearly always better to have these pawns on g7-f6 to restrict Whites minor pieces.

24.hxg5 Be7 25.Kg2 Nb6 26.Bd3! Nb4 27.Be4 Nxa2 28.Rxd8+ Kxd8 29.Rh1 Nd5 30.Ne5

White is already winning thanks to his hyper-active pieces. Ironically, the killing blow is delivered not by his bishop
pair, but by his knight.

30...f5 31.Bxd5 1-0

12. Maki Uuro 2392 Baburin 2541

EU-Cup 22nd Fuegen 2006

1.d4 d5 2.c4 dxc4 3.Nf3 a6 4.e3 e6 5.Bxc4 c5 6.0-0 Nf6 7.dxc5 Bxc5 8.Qxd8+ Kxd8 9.a3 Ke7 10.b4 Bd6 11.Bb2 b5
12.Be2 Bb7 13.Nbd2 Nbd7 14.Rac1 Rac8

The knight is heading a5. However, this manoeuvre could be effective only of White dominates in the centre or on the c-
file. As we know from game 9 Karpov-Ponomariov, when Blacks forces are well coordinated, the knight on a5 might
even prove to be cut off. A more strategic approach presents 15.Nd4, aiming to gain space with f3, e4. I was curious to
investigate possible counterplay with ...g5-g4, but my analysis convinced me that such an activity is not well grounded.
In a slightly different position, however, it might be a good option. So: 15...g5?! 16.f3?! Rhg8 17.N2b3 (17.e4 g4)
17...g4 18.Rxc8 Rxc8 19.fxg4 Ne5 20.h3 Ne4.

A curious position. Even further exchanges will not help White to escape the bind.
16.Bf3 Bxf3 17.N2xf3 g4 18.Nd2 Ne5 is also fine for Black.
Unfortunately, 16.N2b3 Bd5 17.Na5 creates the unpleasant threat of Nc6+ so Black does not have time for ...g4 and
White is the first to fulfil his strategic goal after 17...Ne5 18.f3 Rhg8 19.e4. There is nothing left, but 19...Bc4 20.Bxc4
bxc4 21.Rc2.

The good news is that after 15.Nd4, 15...Nb6 16.N4b3 Na4 17.Ba1 Ne4 is absolutely balanced.

15...e5! 16.Na5 Ba8 17.Rfd1 h6 18.Kf1 Ke6 19.Rxc8 Rxc8 20.Rc1 Rxc1+ 21.Bxc1

Black has a little more space which make his pieces more mobile. Perhaps he should have tried 21...Ne4 22.Nd2 f5.

22.h3 Ne4 23.Nd2 Nc3 24.Bd3 e4 25.Bc2 f5 26.Bb2 Be5 27.Nab3 Nb1 28.Bc1 Nxd2+ 29.Nxd2 Bd5 30.f3 exf3
31.Nxf3 Bf6 32.Kf2?!

32.Nd4+ Bxd4 33.exd4 gave good drawing chances as White could then tradde his light-squared bishop for the knight
and achieve opposite coloured bishops. The same idea was still good on the next move.


33.g4 (an active, but risky defence) 33...fxg4 34.hxg4 Kd6?! 35.Bf5?!

35.Nd2-e4-g3 aiming for Nf5 was a draw, but White chose to trade the wrong piece, remaining with a passive bishop.
No wonder he lost.

35...Be6 36.Bxe6 Kxe6 37.e4 Nd7 38.Ke2 Ba1 39.Ke3 Nf6

40.Nd4+ Bxd4+ 41.Kxd4 Nxg4 42.Kc5 Nf2 43.Kd4 Ng4 44.Kc5 Nf6 45.Kd4 Nd7 46.Bd2 Ne5 47.Be3 h5 48.Kc5
h4 0-1

13. Shirov 2706 Motylev 2641

FIDE-Wch k.o. Moscow 2001

1.d4 d5 2.c4 dxc4 3.Nf3 Nf6 4.e3 e6 5.Bxc4 c5 6.0-0 a6 7.dxc5 Bxc5 8.Qxd8+ Kxd8 9.Nbd2 Ke7 10.Be2 Bd7
11.Nb3 Ba7 (11...Bb6!) 12.Bd2 Nc6 13.Na5 Rhc8! 14.Nxc6+

It turns out that 14.Nxb7? Rcb8 15.Bxa6 blunders a piece due to 15...Bb6 16.b4 Rxa6 17.b5 Ra4+. This line suggests
that White should have played 13.Bc3.

14...Bxc6 15.Ne5
15...Bb5! 16.Bxb5 axb5 17.Rfc1 Ne4 18.Be1

The doubled b-pawns could have been be weak if there were more pieces on the board. Every exchange only underlines
the fact that Blacks king is closer to the battlefield, for instance, 18.Bb4+ Bc5 19.Nd3 Bxb4 20.Nxb4 Rc4. With his
next moves, Shirov tries to avoid exchanges.

18...Bc5 19.Nd3 Bd6 20.Rd1

Or 20.f3 Nc5.

20...Rc2 21.a3 g5

In the Exchange Variation, when one side is more active, but does not have concrete threats, commonly advances on the
kingside. However, 21...Rac8 deserved attention since Black would gladly trade rooks after 22.Rac1.

22.Rac1 Rac8 23.f3 Rxc1 24.Nxc1 Bc5 25.Rd3 Nf6 26.Nb3 Bb6 27.Bb4+ Ke8 28.Bc3 Nd5 29.Kf2

White would like to swap the bishops, remaining with a rook + pawn, but after 29.Bd4, Blacks rook would be too
active on the second rank 29...Rc2 30.Bxb6 Nxb6 31.Rd2 Rxd2 32.Nxd2 Nc4. White has to defend an unpleasant
29...Ke7 30.Ke2 Nxc3+ 31.Rxc3 Rxc3 32.bxc3 The ending is a draw. 32...Bc7 33.h3 Bd6 34.Nd4 Bxa3 35.Nxb5 Bc5
36.Kd3 Kd7 37.g4 Kc6 38.c4 f5 39.e4 fxg4 40.fxg4 Bf2 41.e5 Bg3 42.Nd4+ Kd7 43.Ke4 Bf2 44.Nb5 Bc5 45.Kd3
Kc6 46.Nc3 h6 47.Ne4 b5 48.cxb5+ Kxb5 49.Nf6 Kc6 50.Ke4 Draw.

Chapter 5. The 7.Bb3 Variation

Main Ideas

1.d4 d5 2.c4 dxc4 3.Nf3 Nf6 4.e3 e6 5.Bxc4 c5 6.0-0 a6 7.Bb3

This good prophylactic move anticipates ...b5 and waits for Black to define his set-up. Our task proves to be not trivial
as well see from the next survey.

a) 7...b5 8.a4 b4 has been out of fashion lately. I do not see how Black could get a satisfactory game after 9.e4! Bb7
10.e5 Ne4 11.Nbd2 Nxd2 12.Bxd2 cxd4 13.Ng5 Bd5
Khenkin-Gharamian, Germany 2009, tested 13...Nd7 14.f4 Nc5 15.f5 Nxb3 16.Qxb3 Qd5 17.Qg3 h6 18.fxe6
hxg5 19.exf7+ Kd7 20.Rac1 Qe6 21.Bxg5 Rc8 22.Qf4 Rxc1 23.Rxc1 Bc6 24.Qxd4+ Qd5 25.Qg4+ Kc7 26.e6
Bd6 when instead of 27.h3? Kb7 28.Rd1 Bc5+ 29.Kh2 draw (perhaps White had missed in his calculations that
the h3-pawn was actually unprotected so 29.Kh1?? would end with mate after 29...Rxh3+ or even better,
29...Qxg2+!), 27.h4 would preserve a dominating position.
14.Rc1! Nc6 15.Bxd5 Qxd5 16.Qf3 Qxf3 17.Nxf3 Kd7 18.Rfd1 Rc8
Black lacks a tempo to finish development. The best way to exploit that is 19.Be1! h6 (19...Be7 20.Nxd4!) 20.Nxd4
Ne7 21.Nf3+ Nd5 22.Rxd5+! exd5 23.e6+.

We should discard the typical QGA plan with extended fianchetto in favour of positions with an isolated d4-pawn.
However, the straightforward:
b) 7...cxd4 8.exd4 Nc6 9.Nc3 Be7 10.Bg5 0-0 offers White a free hand on the kingside. His queen can quickly reach a
dreadful position after 11.Qd2 Na5 12.Bc2 b5 13.Rad1 Ra7 14.Qf4 Nc4

The direct attack with 15.Qh4 g6 16.d5 Nxd5 17.Rfe1 f6 18.Bh6 Rf7 is seemingly insufficient for an advantage. The
game Cheparinov-Drenchev, Albena 2014, continued 19.Nxd5 exd5 20.Nd4 Rc7 21.Qg3 Bd6 22.f4 Nxb2 23.Bxg6
hxg6 24.Nc6 Bc5+ 25.Kf1 Qd7 26.Qxg6+ Kh8 27.Rxd5 Qxd5 28.Bg7+ Rxg7 29.Re8+ Qg8 30.Qh6+ Rh7 31.Qxf6+
Rhg7 with a quick draw. Moreover, only Black can play for a win here.

Im more worried though by a more prosaic Whites option. In the diagram position, he can eliminate to a grim
endgame with:
15.d5 Nxd5 16.Bxe7 Qxe7 17.Nxd5 exd5 18.Rxd5 g6

19.Re1 Be6 20.Rd4 Rd7 21.Rxd7 Qxd7 22.Rd1 Qe7 23.b3 Nb6 24.Qd6 Qxd6 25.Rxd6 Nd5 26.Nd4 Nb4.
The breakthrough d4-d5 is a constant threat in these IQP positions.

It is natural for Black to seek more cunning move orders which would force the enemy into concessions in order to get
the active isolated pawn. The most popular is:

c) 7...Nc6 8.Nc3 Be7, hoping for 9.Qe2.

You might wonder why I consider the set-up with Qe2 less dangerous than Qd2. I have mentioned already in line b one
reason from d2 the queen can go to f4 thus enhancing the impact of the d1-rook on the d-file. Another one is that Qe2
is linked with Rfd1 (because of the hanging d-pawn). Naturally, a set-up with Rd1+Re1 would have been more
threatening than Rc1+Rd1 if we take into account that Whites most dangerous threat is the break d4-d5 which opens
the e-file.
You can often read about typical positions and pawn structures, together with examples how to play in them. I think
such writings are only good to teach you how to convert a stable advantage in typical positions. In the early stage of the
game, where the dynamic factor is of paramount importance, our thinking must be utmost concrete. A single nuance of
the placement of one piece could dramatically change the evaluation. One tempo could make the difference between a
passive and unpleasant position with an isolated pawn, and a strong attacking set-up. The same considerations apply to
most openings of course. Thus I do not like the move order of line c, because White could switch to an unpleasant
version of the symmetrical queenless endgame with:
9.dxc5! Qxd1 10.Rxd1 Bxc5
Stayed the knight on d7, Black would have been fine. But on c6 it is one-too-many piece on the open c-file and he does
not have time for the natural development ...b5, ...Bb7.

So I recommend to delay both capturing on d4 and committing the knight to c6. That can be achieved with:
7...Be7 followed by 8...0-0 if White refrains from Qe2.

Now White can choose between two flavours of a symmetrical pawn structure with or without queens, and an IQP
with Qe2.

Studying deeply the queenless endgame after 8.Nc3 0-0 9.dxc5 has no practical value as it has occurred in only one
game. We should remember to develop our queens knight to d7: 9...Qxd1 10.Rxd1 Bxc5
11.Bd2 Nbd7! 12.Rac1 b5.

A. 8.Nc3 0-0 9.e4 Nc6! 10.dxc5 Nd7 11.Ba4 Nxc5 12.Bxc6 bxc6 13.Be3 Qc7

Blacks bishop will enter play from a6, see game 14 Moiseenko-Ganguly, Fujairah City 2012.

B. 8.Qe2 cxd4! 9.exd4

After 9.Rd1 Nc6 10.Nc3! 0-0 11.Nxd4 we should not capture on d4. Instead 11...Qc7! 12.Nxc6 Qxc6 13.e4 b5 14.Bg5
h6! 15.Bh4 Bb7 16.Rac1 Qe8 is balanced.

9...Nc6 10.Rd1 Na5 11.Bc2 b5 12.Nc3 Bb7 13.Bg5 0-0

Black has a good version of the isolator. Statistics show only 44% for White. His problem is that the attack with:

14.d5 is ineffective due to 14...exd5 15.Nd4 g6 followed by ...Re8, ...Ne4.

The other straight approach is 14.Ne5 when Black can either seek exchanges with 14...Nd5 15.Bxe7 Qxe7 16.Ne4
Nc6= or keep the tension with 14...Rc8 15.Rd3 Nc4 followed by ...g6.

Finally, 14.Bxf6 Bxf6 15.Be4 Qb6 is a solid balanced position.

Thus after 7.Bb3 Be7 8.Nc3 0-0! the ball is in Whites court.

Theoretical status

The line 7.Bb3 is Avrukhs recommendation in his White repertoire book. However, he completely misses the retort
7...Be7!. The bishop move, especially in conjunction with 8.Nc3 0-0!, is indeed seldom seen. However, it is not a
revolutionary new idea, but rather a sophisticated move order which dodges Whites most dangerous plans.
Chapter 5. The 7.Bb3 Variation
Step by Step

1.d4 d5 2.c4 dxc4 3.Nf3 Nf6 4.e3 e6 5.Bxc4 c5 6.0-0 a6 7.Bb3


I will analyse further: A. 8.Nc3; B. 8.Qe2.

8.dxc5 Qxd1 is similar to 8.Nc3 0-0 9.dxc5, but Blacks king could remain in the centre.

A. 8.Nc3 0-0 9.e4

9.Qe2 cxd4 should transpose to line B.

9.dxc5 is seldom seen. 9...Qxd1 (or 9...Bxc5 10.Qe2) 10.Rxd1 Bxc5

The only game in my database, Kasimdzhanov-Ivanchuk, Moscow 2007, went on 11.Bd2 Nbd7! (the point of delaying
...Nc6) 12.Rac1 b5 13.Ne2 Bb7 14.Ne5 Rfc8 15.Nxd7 Nxd7 16.Ba5 Nf6 17.Nf4 Rc6 18.Nd3 Be7 19.Ne5 Rxc1
20.Rxc1 Bd6 21.Nd3 Rc8=.
The engines prefer 11.e4 (11.h3 b5 12.e4 Bb7 13.e5 Nfd7 14.Bf4 Be7=), but without queens, White space advantage
does not bring him dividends: 11...Ng4 12.Rf1 b5 13.h3 Nf6 14.e5 Nfd7 15.Bc2 Bb7 16.Be4 Bxe4 17.Nxe4 Be7
18.Bf4 Nc6 with comfortable development.


9...cxd4 makes sense only if Black wanted to meet 10.Nxd4 by 10...e5. However, the d5-square ensures White some
pull, for instance: 11.Nf3 Nc6 12.Be3 Bg4 13.Rc1 Rc8 14.Qxd8 Rfxd8 15.Nd5 Nxd5 16.Bxd5.

10.dxc5 Nd7 11.Ba4

Black has less space so exchanges suit him, for instance: 11.Na4 Nxc5 12.Nxc5 Bxc5 13.Bg5 Qxd1.

11.Be3 Nxc5 12.Bc2 b5 13.Qe2 Bb7 14.Rac1 Qb8! solves all Blacks problems 15.Bb1 Nd7 16.Nd2 Rd8. If White
attempts to display activity, Black can counter-attack: 17.f4 Bc5 18.Bxc5 Nxc5 19.Nxb5 axb5 20.Rxc5 Qa7 21.Nb3
Nd4 22.Qf2 Nxb3 23.axb3 Rd7.
11...Nxc5 12.Bxc6 bxc6 13.Be3 Qc7
Black has enough counterplay. See game 14 Moiseenko-Ganguly, Fujairah City 2012.

B. 8.Qe2 cxd4! 9.exd4

Recapturing by knight is less challenging:

a) 9.Nxd4 0-0 10.Nc3 e5 11.Nf3 e4 12.Ng5 Bf5 13.Bc2

Black can now preserve material balance by 13...Bd6 14.Ncxe4 Nxe4 15.Nxe4 Bxh2+ 16.Kxh2 Bxe4 17.Bxe4 Qh4+
18.Kg1 Qxe4=, but I like the more enterprising pawn sac:
13...Bg4 14.Qd2 Qc7 15.Ngxe4 Rd8 16.Nxf6+ Bxf6 17.Qe1 Nc6 with a long-term initiative, e.g. 18.f3 Nb4
19.Bb1 Be6 20.Ne4 Be7 21.Bd2 Nd5.

b) 9.Rd1 Nc6 10.Nc3!
White awaits ...0-0. In the event of 10.Nxd4 Nxd4 11.exd4 Black has the more useful move 11...Bd7! 12.Nc3
Bc6 blocking the isolator.
10...0-0 11.Nxd4

The older continuation 11...Nxd4 12.exd4 Nd5 13.Qf3 has disappeared from tournaments.
12.Nxc6 Qxc6 13.e4 b5 14.Bg5 h6!
14...Bb7 is satisfactory, but why not make a luft for free.
15.Bh4 Bb7 16.Rac1 Qe8! 17.e5
After the timid 17.Bg3, it would be good to unload a little the position with 17...Rd8.
17...Nd7 18.Bxe7 Qxe7

The position after 19.Ne4 has also been tested without ...h6 when Black had the only move ...Bxe4. Even then, he
remained very solid. With ...h6 included, he can take the pawn: 19...Nxe5! 20.Nd6 Ng6 21.Qg4 and defend
successfully with 21...Kh7.
19...Nc5 20.Bc2 b4 21.Ne4 Bxe4 22.Bxe4 Nxe4 23.Qxe4 Rac8= Bake-Heilala, ICCF email 2007.

9...Nc6 10.Rd1


The old school preached that an isolated pawn should be blockaded, but that would leave White with extremely active
pieces after 10...0-0 11.Nc3 Nb4 12.Ne5. Modern chess thinking puts the accent on dynamic factors. ...Na5 is
contending for the d5-square by driving back the b3-bishop. At the same time the knight is heading for c4 with a direct
threat to b2. Note that we should do that before castling!

11.Bc2 b5 12.Nc3 Bb7 13.Bg5

White has also tried to leave the bishop on c1: 13.Ne5 Rc8 14.a3 0-0 15.Rd3 Nc4 and it is unclear what White could do
from here. For instance, 16.Rg3 Qxd4 17.Bh6 loses by force to 17...Nxe5 18.Rxg7+ Kh8 19.Rd1 Qc5 20.Rd5 Bxd5
21.Qxe5 Be4+.

13...0-0 14.Ne5

14.Bxf6 Bxf6 15.Be4 presents a more restrained approach. White wants firstly to take control of the centre and then
prepare a gradual attack on the kingside or play along the c-file.
15...Qb6 16.Rd3 Nc4 17.Bxb7 Qxb7 18.Ne4 Be7 19.b3 Nb6 20.Nc5 Qd5 21.Rc1, Ikonnikov-Rabiega, Germany 2010.
Best continuation here is to pin the f3-knight with 21...Qh5! 22.Qe1 Nd5 23.Ne5 Nf4 24.Re3.
The typical pawn sacrifice 14.d5 exd5 15.Nd4 (15.Qd3 g6 16.Re1 Re8) has no venom: 15...g6 16.Rd3 (16.Qf3 Re8
17.Re1 Ne4 18.Bxe7 Qxe7 19.Qf4 Nc4) 16...Re8

Cori Tello-Estrada Nieto, Lima 2013, saw further the desperate 17.Nf5 gxf5 (17...Bf8) 18.Rg3 Bd6 (18...Kh8)
19.Bxf6+ Bxg3 20.Bxd8 Rxe2 21.Nxe2 Bxh2+ 22.Kxh2 Rxd8.


14...Nd5 15.Bxe7 Qxe7 16.Ne4 Nc6 has been known to be equal. The text aspires to the advantage.


15.Rac1 looks senseless as it does not pose any threat. Black can either follow up with his plan 15...Nc4, or unload the
tension with 15...Nd5 16.Bxe7 Qxe7 17.Ne4 Nf4 18.Qe3 Nxg2 19.Kxg2 f5 20.f3 fxe4 21.Bxe4 Bxe4 22.fxe4 Nc4
23.Nxc4 Rxc4=, Borroni-Blauhut, ICCF email 2004.


Now 16.Re1 Nd5 would unload the position so White has to complete his rooks lift with Rg3 or Rh3. Black will then
play ...g6 and his castling position will be unassailable while the white pieces might prove to be poorly coordinated for
defence. Possible continuations are:
a) 16.Rg3 g6 (16...Qxd4? 17.Bxf6 Bxf6 18.Bxh7+) 17.Bh6 Re8
17...Qxd4 18.Nxg6 (18.Bxf8 Bxf8 19.Nxc4 Rxc4) 18...hxg6 19.Bxg6 fxg6 20.Qxe6+ Kh8 21.Bxf8 Bxf8
22.Rxg6 Re8.
18.Nxg6 hxg6 19.Bxg6 fxg6 20.Qxe6+ Kh8 21.Qf7 Rg8 22.Re1 Nd6 23.Qxe7 Nh5 24.Rh3 Re8 25.Qxd8 Rcxd8.

b) 16.Rh3 g6 17.Rd1 (17.Bb3 Nh5;

17.Bh6 Qxd4) 17...Nd5 (17...Nxb2) 18.Nxg6 hxg6 19.Bxg6 Bxg5+, Wojtaszek-Dominguez Perez, blitz, Beijing 2014.
Chapter 5. The 7.Bb3 Variation

Annotated Games

14. Moiseenko 2702 Ganguly 2619

Fujairah City 18.11.2012

1.d4 d5 2.c4 dxc4 3.Nf3 Nf6 4.e3 e6 5.Bxc4 c5 6.0-0 a6 7.Bb3 Be7 8.Nc3 0-0 9.e4 Nc6 10.dxc5 Nd7 11.Ba4 Nxc5
12.Bxc6 bxc6

At first glance, Blacks split queenside pawn structure should favour White. Then we see the hole on d3 and the
beaming a6-f1 diagonal. Blacks queens rook also has bright prospects on the b-file.

13.Be3 Qc7 14.Qc2 a5 15.Rfc1

Perhaps the rook would stay better on the open file, but 15.Rfd1 Ba6 16.Bxc5 Bxc5 17.Na4 Bd6 18.Nc5 Bb5 is
balanced, e.g. 19.a4 Bxc5 20.Qxc5 Be2=.

15...Ba6 16.Bxc5 Bxc5 17.Na4 Be7 18.Nc5

White cannot hope to win the c-pawn since 18.b3 c5, followed up by ...c5-c4, would open the position in favour of the
black bishops.

18...Rfb8 19.Nxa6

Moiseenko acknowledges the fact that he does not have any active plan. He could have waited with this exchange for a
move or two, but after 19.b3 h6 20.h3 Rb4, the pressure on e4 would force him to trade his knight. Perhaps he wanted
to avoid 19.b3 Bb5 20.a4 Bxc5 21.Qxc5 Bd3 although White would have enough compensation for the pawn after
22.Re1 Rxb3 23.Rad1.

19...Rxa6 20.Qe2 Rab6 21.Rc2 a4 22.g3 h6 23.Rd1 Bf6 24.e5 Be7 25.Rdd2 Bb4 26.Rd4 Qb7 27.Qd3 Bf8 28.Rd7 Qc8

29.Nd2! Rb5 30.Nc4 Rd5= was more accurate.

29...R6b7 30.Rxb7 Qxb7 Draw.

Black could play on with 31.Rc2 c5 32.Qe2 Qd5.

Chapter 6. The 7.Bd3 Variation

Main Ideas

1.d4 d5 2.c4 dxc4 3.e3 Nf6 4.Bxc4 e6 5.Nf3 c5 6.0-0 a6 7.Bd3

This system has seen better times when Kramnik used it several times. Lately, only Dreev has been employing it
successfully at a top level. To me, it looks dried out as White has not shown new ideas for more than 10 years. That
does not mean the retreat to d3 is bad or refuted. White opts for a calm natural development with an emphasis on the
middlegame. At the same time he sets up two positional traps. If Black plays ...b5, then a4, followed up by Rxa4,
activates the queens rook which might go to d4 or to the kingside. Another point of Bd3 is that the bishop prevents a
trade of queens and enables dxc5 aiming for e4-e5. We should keep that in mind and avoid too flexible moves, e.g.
7...Nc6 8.Nc3 Be7?! 9.dxc5!.

Black has at least two decent answers. If you are familiar with the Nimzo, you could enter a fairly good version of an
IQP position with:

A. 7...cxd4 8.exd4 Nc6

You have probably read a lot about playing against an isolator. To me, Blacks primary goal should be not to blockade
the d4-pawn, but to restrain the opponents activity. Thus our agenda should look as follows (in this order!):

1. Castle quickly.

2. Trade dark-squared bishops with ...Nh5 or ...Nd5.

3. Develop the light-squared bishop on d7 or b7 depending on the situation.

I also recommend the stand with Nc6+Nd5 (it takes less time and controls e5) instead of Nd7-b6-d5.

It is more difficult to give a clear advice when and whether at all to fianchetto the queens bishop. Lets consider three
main scenarios.
1. White leaves his bishop on c1 9.Nc3 Be7 10.a3 0-0 11.Bc2
White has neglected the fight for the centre so we can safely adopt the most active way of developing our bishop:
11...b5 12.Qd3 Bb7 13.Re1 Na5! 14.Ne5 Rc8 15.Qh3 Nc4.

2. White plays Be3 9.Nc3 Be7 10.Be3. Again, the extended fianchetto is a good set-up: 10...0-0 11.Rc1 Nb4 12.Bb1
b5 13.a3 Nbd5. A more sophisticated version of this set-up is analysed in game 16 Botvinnik-Euwe.

3. The most active and popular option is Bg5. It is our main line.

9.Nc3 Be7 10.Bg5 0-0 11.Rc1

We are at a big juncture now. While Black has many options, most of them allow White to enjoy a slight, but lasting
pull in a safe position. I recommend to reduce the tension by:

11...h6 12.Bh4 Nh5 13.Bxe7 Nxe7

This way we keep the blockade on d5. See game 15 Radjabov-Bacrot.

Besides IQP positions, Black can also choose a line which is easier to play:

B. 7...Nbd7!? 8.Re1 Be7!? 9.e4 cxd4 10.e5 Nd5

Note, however, that it is easier for both sides! Play may quickly steer into a sharp endgame after:


Or 11.Nxd4 Nc5 12.Bc4

11...Nc5 12.Nxd4 0-0 13.Bxd5 Qxd5 14.Nf5 Bd8 15.Nd6 Bd7 16.Qxd5 exd5 17.Rd1 Ne6 18.Be3 f6=.

Chapter 6. The 7.Bd3 Variation

Step by Step

1.d4 d5 2.c4 dxc4 3.e3 Nf6 4.Bxc4 e6 5.Nf3 c5 6.0-0 a6 7.Bd3

A. 7...cxd4; B. 7...Nbd7

7...Nc6 8.Nc3 cxd4 transposes while 8...Be7 is at least arguable due to 9.dxc5.

A. 7...cxd4 8.exd4 Nc6

It is tempting to continue development with ...b5, ...Bb7 and ...Nbd7 which is a typical set-up against an isolator, but
White can hinder it: 8...Be7 9.Ne5! Nc6 (9...0-0 is a bit passive, e.g. 10.Nc3 Bd7 11.Be3 Nc6 12.a3 Rc8 13.Re1)
10.Nxc6 bxc6 11.Be3 Nd5 12.Nc3 0-0 13.Rc1. Blacks pawn structure assures him of counterplay and good control of
the centre, but still it is very static.

9.Nc3 Be7 10.Bg5

It is ineffective to delay this move. Black can safely fianchetto his bishop after 10.a3 0-0 11.Bc2 b5 12.Qd3 Bb7 13.Re1
(hoping for 13...Rc8? 14.d5! exd5 15.Bg5) 13...Na5 14.Ne5 Rc8 15.Qh3 Nc4. The point is that White cannot break
through on the kingside without a rook lift or d4-d5.

10.Be3 is innocuous. It makes more sense after Bg5 h6 when Black will lack the defence ...g6 due to the weakness of
the h6-pawn. 10...0-0 11.Rc1 (11.a3 b5 12.Ne4 Nxe4 13.Bxe4 Bb7 14.Qc2 Rc8 15.Bxh7+ Kh8 16.Qd3 g6 17.Bxg6
fxg6 18.Qxg6 Qd5=) 11...Nb4 12.Bb1 b5 13.a3 Nbd5 with a fine game, for instance: 14.Nxd5 Qxd5 15.Ne5 Bb7
16.f3 Rac8 or 14.Ne5 Bb7 15.Qd3 Nxc3.

10...0-0 11.Rc1

Black has tried 9(!) moves here.

a) 11...Bd7 12.Re1 Rc8 is too passive and gives White a free hand on the kingside. He can follow up with 13.a3 or even
13.h4!? Nb4 14.Bb1 Bc6 15.Ne5 as in Ivanchuk-Harikrishna, Tripoli 2004.

b) 11...b5 and 11...Nb4 offer White a slight, but lasting pull after 12.Bxf6!? Bxf6 13.Ne4 (respectively 13.Be4).

c) 11...b6 looks like an improved version of the previous line since Black keeps control of c5. However, White retains
an initiative in the typical IQP position which arises after 12.Bb1 Bb7 13.a3 Rc8 14.Qd3 g6 15.Rfe1

Black lacks room for manoeuvering:

15...Rc7 16.Ba2 Rd7 17.Rcd1; 15...Na5 16.Bh6 Re8 17.Ne5 Nc6 (17...Nd7 18.Qg3 Bf8) 18.Qh3; Perhaps he
should trade a few pieces with:
15...Nd5 16.Bh6 Re8 17.Nxd5 Qxd5 18.Qd2 Na5 19.Be4 Qd8.

In my opinion, Black should aim to unload the position by trading minor pieces.


A typical method of trading bishops, borrowed from the Nimzo-Indian.

11...Nd5 pursues similar ideas, but I would like to avoid 12.Nxd5
I do not believe that Black could be worse with a bishop pair and no weaknesses after 12.Be3 Nxe3! [12...Ncb4
13.Bb1 b6 (13...b5 14.Ne4) 14.a3 Nxc3 15.bxc3 Nd5 16.Qd3 g6 17.Bh6 Re8 18.c4] 13.fxe3 Nb4 14.Bb1 Nd5
15.Qd2 Nxc3, but White has another decent alternative:
12.Bxe7 Nce7 13.Ne5 Bd7 14.Be4 Bc6 15.Nxc6 gives White a structural advantage and active pieces although
Black easily drew in Naroditsky-Rapport, Riga 2014.

White gets a temporary initiative after both:

12...Bxg5 13.Nxg5 Qxg5 14.Nb6 Rb8 15.Qc2 g6 16.Qc5 and:
12...exd5 13.Bxe7 Qxe7 14.Re1.

11...h6 followed by 12...Nh5 retains control of d5.


12.Be3 deprives White of the option of Bxf6 which is the main retort against the plan with ...b5, see game 16
Botvinnik-Euwe, Hastings 1934.


12...Nd5 13.Bg3

13.Bxe7 Nxe7

The position is balanced and both sides have many options of equal worth. White could also try 14.Re1 b6 or 14.Bb1
Bd7 15.Ne5 Nf6 16.Qd3 Bc6=. Blacks aim is to keep the blockade on d5 while seeking a way to develop his light-
squared bishop.

14...Nf6 15.Ne4

Another possible line is 15.Re1 Bd7 (15...b6!? 16.Qf3 Ra7) 16.Ne4 Nxe4 17.Bxe4 Bc6 18.Nxc6 bxc6 19.Bxc6 Rb8
when Black should regain the pawn soon 20.b3 Rb4 21.Qe2 a5 22.d5 Nxd5 23.Red1 Qd6.


A tense middlegame fight is ahead.

Black has plenty of defensive resources as ...f6 or ...Nf5. See game 15 Radjabov-Bacrot, Beijing 2014.

B. 7...Nbd7
This move is dubious against 7.Bb3 because Blacks knight lacks a stable retreat square after e3-e4-e5. 7.Bd3 removes
the hit on d5 so:
8.e4 cxd4 9.e5 is not devastating due to 9...Nd5. In practice White has even tried 10.Bc4 (10.Be4 Nc5), but it offers
Black too many free tempi after 10...Nb4! 11.a3 Nc6 12.Re1 b5.

8.Qe2 followed by Rd1 is somewhat slow. Black can answer 8...b6 9.Nc3 Bb7 10.Rd1
Dreev beat Ivanchuk in Astana 2012 with 10.a4 Bd6 11.Rd1 0-0 12.dxc5 Nxc5 13.e4

13...Qc7! 14.Bc2 Ng4 15.h3 Ne5 16.Nd4. Here 16...Ng6 would have forced White to justify his previous play
with 17.g3, but a further advance would be impossible as 17...Rfd8 18.f4? would stumble into 18...e5.
10...Qb8 (This is safer than 10...Qc7 which, however, brings Black even better results.) 11.h3
11...Be7 (It is interesting to insert ...b5, for instance: 11...b5 12.a3 Be7 13.dxc5 Nxc5 14.Bc2 Nce4 15.Nxe4 Bxe4
16.Bxe4 Nxe4 17.Nd2 Nc5=) 12.dxc5 Nxc5 13.Bc2 Nce4 14.Nxe4 Nxe4 (14...Bxe4) 15.Nd2 Nf6 16.e4 Qc7=. The
point is that after 17.Nf3 e4-e5 is not a threat in view of ...Bxf3.
8.a4 b6 9.Qe2 can be met by the following set-up: 9...Qc7 10.Nc3 Bd6 11.h3 Bb7 12.Bd2 0-0 13.Rac1

13...e5 14.dxe5 Bxf3 15.Qxf3 Nxe5.

This survey of Whites options shows that his most dangerous plan is based on e3-e4-e5, but its direct execution is
ineffective. Therefore, White players should prepare it:

8.Re1 Be7!?

8...b5 is still dubious due to 9.a4 even though Re1 might prove to be a superfluous move. I think that it is safer to
develop the kingside, although:
8...b6 is a fair alternative: 9.e4 cxd4 10.e5 Nd5 11.Nxd4 Nc5 12.Bf1 Bb7

13.Nd2 Be7 14.Qg4 0-0 15.N2f3 Kh8 commonly leads to draws by repetition of moves, e.g. 16.Ng5 h6 17.Ngf3 a5
18.Qh5 Kh7=. It is more cunning to put in 13.a3 when 13...Be7? 14.b4 Nd7 15.Qg4 0-0 drops the exchange to 16.Bh6.
Black would have to find some waiting yet developing move as 13...Qd7 when 14.Qg4 h5 would be fine for Black, but
14.Nd2 Be7 15.Qg4 is yet to be tested.

9.e4 cxd4 10.e5 Nd5


11.Nxd4 gives Black a wider choice:

a) 11...Nc5 12.Bc4 (12.Bf1 0-0 13.Qg4 f5 14.exf6 Nxf6) 12...0-0 transposes to the main line.

b) 11...Qc7 12.Be4 Qxe5 13.Bf3 Qd6 14.Nf5 exf5 15.Nc3 N7b6 16.Nxd5 Nxd5 17.Bxd5 0-0 18.Bf4 Qxf4 19.Rxe7
Be6 20.Bxe6 fxe6 21.Qb3 b5 22.Qxe6+ Kh8 23.Re1 h6 leads to a boring drawish endgame.

c) 11...Qb6 12.Nf3 (12.Bc4 Bc5 13.Re4 0-0 14.Nc3 Bxd4 15.Rxd4 Nxc3 16.bxc3 Nxe5 17.Bb3 Qa5) 12...Nc5
13.Bc4 0-0 14.Bxd5 exd5 15.Nc3 Be6 16.Re2.

11...Nc5 12.Nxd4 0-0 13.Bxd5 Qxd5 14.Nf5 Bd8 15.Nd6 Bd7 16.Qxd5 exd5 17.Rd1 Ne6 18.Be3 f6

Black has sufficient counterplay.

Chapter 6. The 7.Bd3 Variation
Annotated Games

15. Radjabov 2734 Bacrot 2718

Rapid Beijing 12.12.2014

1.d4 d5 2.c4 dxc4 3.e3 Nf6 4.Bxc4 e6 5.Nf3 a6 6.Bd3 c5 7.0-0 Nc6 8.Nc3 cxd4 9.exd4 Be7 10.Bg5 0-0 11.Rc1 h6

12.Bh4 Nh5 13.Bxe7 Nxe7 14.Ne5 Nf6 15.Ne4 Nfd5 16.Nc3

It is a matter of taste, but in my opinion Blacks game is easier since he has a clear target on d4:
16.Re1 b6 17.Ng3
Or 17...Nf6 18.Nh5 Nxh5 19.Qxh5 Bb7 when the rook lift 20.Re3 is ineffective due to 20...Nf5 21.Bxf5 exf5
22.Rg3 f4 23.Rg4 Rc8 24.Rd1 Qf6.
18.Qf3 Rc8 19.Qe4 g6.

Alternatively, 16.Ng3 f6 17.Nf3 Nf4 18.Bc4 Bd7 19.Qd2 Ned5 20.Rfe1 is balanced.
Radjabovs move is a little provocation perhaps Blacks best retort is 16...Nf6, but Bacrot wants to take over the
psychological initiative.

16...Bd7 17.Be4 Bc6

I prefer to keep the pawns flexible so I like 17...Rc8.

18.Nxc6 bxc6

Many games have proved that this structure is passive, but very firm. It is important that Black got rid of his light-
squared bishop. His remaining knights do not hinder rooks manoeuvres.

19.Qd2 Qd6 20.Rfd1 Rfb8

Whites only weakness is d4 so the rook should go directly to d8 and the other one to b4.

21.b3 Nxc3 22.Rxc3 Rd8 23.h3 Rab8 24.Rc4 Rb5 25.Bf3 a5 26.Qc3 Nd5 27.Qc1 Rb6
White has reached the maximum in this structure, but that should be insufficient for winning the c6-pawn.
Bacrots next manoeuvres are rather clumsy and he drops a pawn.

28.g3 Nb4 29.Rd2 Rdb8?! 30.Kg2 Qd8?! 31.Rc5 Rb5 32.Bxc6 Nxc6 33.Rxb5?

This leads to an easy draw. White had to keep both rooks by 33.Rxc6. Later White even offered the opponent the
initiative, but the draw was the natural outcome.

33...Rxb5 34.Qxc6 Rd5 35.a3 e5 36.b4 axb4 37.axb4 Rd6 38.Qb7 exd4 39.Ra2 d3 40.Rd2 Qg5 41.Qb8+ Rd8 42.Qf4
Qb5 43.Qe4 g6 44.g4 Rd5 45.Kg1 Kg7 46.Kg2 Qd7 47.Kg1 h5 48.gxh5 Rg5+ 49.Kh2 Qd6+ 50.f4 Rd5 51.hxg6
fxg6 52.b5 Rd4 53.Qe5+ Qxe5 54.fxe5 Kf7 55.b6 Ke6 56.b7 Rb4 57.Rxd3 Rxb7 58.Rd6+ Kxe5 59.Rxg6 Draw.

16. Botvinnik Euwe

Hastings 28.12.1934

1.c4 c6 2.e4 d5 3.exd5 cxd5 4.d4 Nf6 5.Nc3 Nc6 6.Bg5 e6 7.Nf3 dxc4 8.Bxc4 Be7 9.0-0 0-0 10.Rc1 a6 11.Bd3 h6

This plan was dubious without the insertion of 11...h6 12.Be3 due to strategic reasons. White could take on f6 and play
Be4. Now it is still dubious, this time due to tactical reasons. The pawn on h6 is a good target for sacrifices and White
could have pose problems on move 14.
The best move was 12...b5! when 13.Ne4 Bb7 14.Nc5 Bxc5 or 13.a3 Bb7 14.Qd2?! Na5! are fine for Black.
Euwes method is good against set-ups with Be3 and Blacks pawn on h7.

13.Bb1 b5 14.Ne5?!

14.Qd2! would have pinpointed the vulnerability of the h6 pawn. Then 14...Re8 looks like an only move and 15.Ne4
Nbd5 16.Nc5 would give White an edge after 16...Nxe3?! 17.fxe3 Bd6 18.Qc2!. 16...Bxc5 17.dxc5 e5 would be
playable, but Blacks knight on d5 instead of the blockading square c6 would make the difference in Whites favour
compared to the position from the previous note.

14...Bb7 15.Qd2 Re8

Black has successfully consolidated and his chances are not worse, at least. Perhaps White should embark on exchanges
with 16.a3 Nbd5 17.Ne4 Nxe3 18.Qxe3 Nxe4 19.Bxe4 Bxe4 20.Qxe4=. Botvinnik overestimates his attacking

16.f4? Nbd5 17.Nxd5 Qxd5! 18.f5 Bd6! 19.fxe6 Rxe6 20.Bf5 Re7

White has exhausted his active options and his centre is desperately hanging. He decides to seek salvation in an
endgame, but Euwes technique proves to be up to the task.

21.Bh3 Bxe5 22.dxe5 Qxe5 23.Bf4 Qd5 24.Qxd5 Nxd5 25.Bd2 Rae8 26.b3 Re2 27.Rf2 Nf6 28.Ba5 Rxf2 29.Kxf2
Ne4+ 30.Kf1 Ng5 31.Bd7 Re7 32.Bf5 Re5 33.Bb1 Be4 34.Bxe4 Nxe4 35.Rc6 Rf5+ 36.Ke1 Rf2 37.a4 Rxg2 38.Rxa6
bxa4 39.bxa4 Rxh2 40.Ra8+ Kh7 41.Bb6 Ra2 42.a5 h5 43.a6 h4 44.a7 h3 45.Bg1 Nf6 46.Kd1 Ng4 47.Re8 h2 0-1

Chapter 7. The Old Main Line

Main Ideas

1.d4 d5 2.c4 dxc4 3.Nf3 Nf6 4.e3 e6 5.Bxc4 c5 6.0-0 a6 7.Qe2 b5 8.Bb3 Bb7

This had been the big Main line during decades in the 20th century. The topical question was where should Black put
his queen after:

Note also the move 9.a4 when simplest is 9...b4 10.Nbd2 Be7 11.Nc4 0-0 12.Rd1 Qc7 13.Bd2 Nbd7 14.Rac1 a5!
followed by ...Nb6, ...Ba6, see game 17 Jussupow-Anand, Dortmund 1997.

9...Nbd7 10.Nc3

All three options, 10...Qb8, 10...Qb6, and 10...Qc7 have their fans even today. At first it seemed that Whites only plan
was to push d5 and play the following position:
10...Qc7 11.d5?! exd5?! 12.Nxd5?! Nxd5 13.Bxd5 Bxd5 14.Rxd5 Be7
It could br slightly different flavours with Blacks queen also on b8 or b6, but the overall evaluation is the same
Black is fine. He can follow ...Nb6, ...Qb7, ...Rfe8 and his pawn majority on the queenside would be a strong trump
once Whites attempts to attack on the other flank were neutralised.
Then it turned out that Black had not a perfect move order for reaching the diagram position. All his possible captures
on d5 could face a pawn sacrifice which drastically changed the pace of the game:
10...Qc7 11.d5 exd5 12.e4! d4 13.e5;
Taking on d5 by piece: 11...Nxd5 12.Bxd5 Bxd5 13.Nxd5 exd5, allowed the intermezzo 14.a4! which damaged Blacks
queenside pawn formation.

10...Qb6 is not perfect either, as White can try 11.a4 c4 12.Bc2 b4 13.a5!? Qc7 14.Na4 with complications.

The boom in popularity of the Meran Variation inspired Black to meet d4-d5 by ...c5-c4. With this idea in mind,
10...Qc7 looks best. Indeed, 11.d5 c4 12.dxe6 fxe6 would have been very nice for him. However, 10...Qc7 allows
another tactical blow: 11.e4! cxd4 12.Nxd4 Bc5
13.Bxe6!?. Latest analyses suggest that it leads to a forced draw (if you remembered the long lines!).
Thus the safest way to play for a win remains:


11.d5 c4!? 12.dxe6 fxe6 13.Bc2 Bd6

Blacks game is quite easy. If White develops his bishop on e3 (after e4), the manoeuvre Nc6-e5-g6 ensures control of
the dark squares.
Alternatively, 14.Ng5 0-0 15.f4 Nc5 16.b4, allows Black to weaken the enemy light squares with:
16...Nd3! See game 19 Baranowski-Cipka, Lechenicher SchachServer 2011.

Chapter 7. The Old Main Line

Step by Step

1.d4 d5 2.c4 dxc4 3.Nf3 Nf6 4.e3 e6 5.Bxc4 c5 6.0-0 a6 7.Qe2

7.Nc3 b5 8.Bb3 Bb7 transposes after 9.Qe2 while 9.e4 cxd4 10.Nxd4 Nc6 11.Nxc6 Qxd1 12.Rxd1 Bxc6 13.f3 Bc5+
14.Kf1 Rc8 is equal.



8.Bd3 does not fit to the move Qe2. Black should enter the IQP position after 8...cxd4 9.exd4 Be7 when 10.a4 bxa4 is
11.Rxa4 Bb7 (or 11...0-0 12.Nc3 Bb7) 12.Nc3 0-0 13.Be3 a5 14.Ne5 Nc6 15.Bb5 Nxe5 16.dxe5 Nd5 17.Nxd5
11.Nc3 Nc6 12.Nxa4 0-0 13.Be3 Bb7 14.Rfc1 Nb4 15.Bc4=.


Main branches now are:

A. 9.a4; B. 9.Rd1

A. 9.a4
Black now has a fair choice between
A1. 9...Nbd7 and A2. 9...b4.
The former triggers very sharp complications where the cost of every move is high. The latter leads to typical QGA
positions and a balanced quiet game.

A1. 9...Nbd7 10.axb5 axb5 11.Rxa8 Qxa8 12.Nc3 b4 13.Nb5 Qb8

White is slightly ahead in development, but his c1-bishop is passive. Therefore, his next move looks a must. Moreover,
practice has proved that taking on e4 is very risky, to say the least.

14.e4 cxd4 15.Nbxd4

15.Nfxd4 Bc5 16.Bxe6 stumbles into 16...0-0!? (16...fxe6 17.Nxe6 Qe5) 17.Bc4 Bxe4

15...Bd6 16.Bxe6

16.Rd1 0-0 17.Nxe6 fxe6 18.Bxe6+ Kh8 19.Qd3 loses to 19...Nxe4 20.Bxd7 Bxh2+ 21.Nxh2 Nxf2.

The only decent alternative to the piece sacrifice is:

16.Re1!? when the threat of 17.e4-e5 forces Black to grab the e4-pawn. It seems that both captures are possible and lead
to complications no less hazy than the main line:

a) 16...Nxe4 17.Bxe6! (17.Ba4 Bd5 18.Bc6 Qb6 19.Bxd5 exd5 20.Nf5 0-0 21.Be3 Qb8=) 17...fxe6 18.Nxe6 Ndf6
19.Nfg5 h6;

b) 16...Bxe4 17.Ng5 Nc5 18.Nxe4 Nfxe4 19.Bc2 Nf6 20.Bg5 Nd5 21.Qh5 g6 when instead of 22.Qh6 Bf8 23.Qh3
Bg7 24.Ba4+ Kf8 25.Nc6 Qa8 26.Bb5 h6, Marcos Nozar-Hyllengren, ICCF email 2011, White should have tried
22.Bxg6+ fxg6 23.Qf3 Qa8 24.Nxe6 with tangled play.

16...fxe6 17.Nxe6

This move was introduced in the game Kramnik-Anand, rapid Mainz 2001, which Black won. Its impact was so strong
that nobody has repeated Whites line ever since. In fact, things are not so bad for White. After 18.Nxg7+ Kf7 19.Nf5
Bxe4 20.Nxh6+! Kg7, Jussupow and Kindermann suggest 21.Qd2! (instead of 21.h3?). Forced variations do not show
a clear advantage for either side. Black has a wide choice: 21...Kf8, 21...Rh7, 21...Bxf3 with mutual chances.

A2. 9...b4 10.Nbd2 Be7 11.Nc4 0-0 12.Rd1 Qc7 13.Bd2 Nbd7

This is actually a sideline of the 7.Bb3 b5 system where Whites best approach is to push e3-e4 early and attack on the
kingside. The set-up with Qe2+Rd1 is totally harmless. Whites biggest problem is the awkward dark-squared bishop. It
does not have prospects and Black could underline this fact by bolstering the b4-pawn, for instance: 14.Rac1 (14.a5?!
Bd5 15.Nfe5 cxd4 16.exd4 Nxe5 17.dxe5 Ne4) 14...a5 15.Nfe5 Rfd8. By all means, play is not forced and any sensible
development should suit Black. For middlegame plans, see game 17 Jussupow-Anand, Dortmund 1997.

B. 9.Rd1 Nbd7 10.Nc3

10.a4 allows several good retorts as 10...b4 or 10...Bd6. But I prefer the more flexible 10...Qb8 11.axb5 axb5 12.Rxa8
Bxa8 13.Nc3 b4 14.Nb5 Be7.

10.e4 cxd4 11.e5 Bxf3 12.gxf3 Nh5 is double-edged.


Black also achieves a positive score after:

a) 10...Qc7
The only drawback of this move is the forced draw looming in the line 11.e4
11.d5?! is a mistake. Black gets an improved version of the main line with 10...Qb8 since the queen is better
placed on c7. He should continue 11...c4! [11...exd5 12.e4! dxe4 (12...d4 13.e5) 13.Ng5 c4 14.Ncxe4 Nxe4
15.Nxe4 0-0-0 16.Bc2 when even the best 16...g6 (16...Qc6? 17.Qg4) 17.b3 Re8 18.bxc4 Nc5 19.cxb5 Bxe4
20.Bxe4 Rxe4 21.Be3 is dangerous for Black.] 12.dxe6 fxe6 13.Bc2 Bd6.
11...cxd4 12.Nxd4
Black cannot prevent Bxe6 (12...Nc5 13.e5), for instance:
12...Bd6?! 13.Bxe6 fxe6 14.Nxe6 Bxh2+ 15.Kh1 Qb6 16.Nxg7+ Kf7 17.Be3 with an attack;

12...Bc5 13.Bxe6 (13.Be3=) 13...fxe6 14.Nxe6 Qe5 15.Nxg7+ Kf7 16.Nf5. Modern engines calculate this position to a
draw: 16...Rad8 17.Nh6+ Kg7 18.Be3 Bxe3 19.Qxe3 Bxe4 20.Nxe4 Nxe4 21.Ng4 Qf5 22.Qh6+ Kf7 23.Qh4 Qe6
draw, Dutra Neto-Kharitonov, LSS email 2008.

b) 10...Qb6 also gives White a tempo (for a4-a5) 11.a4 c4 12.Bc2 b4 13.a5 Qc7 14.Na4 Qxa5

15.e4 (or 15.Qxc4) 15...Rc8.

c) 10...Bd6 11.e4 (11.dxc5 Bxc5 12.h3 0-0 13.e4 Qc7 14.Bg5 h6=) 11...cxd4 12.Rxd4 (12.Nxd4 Qb8 13.g3 0-0=)
12...Bc5 13.Rd3 Ng4 14.Bg5
Most popular now is 14...Qb6 when critical is 15.Nd5 Bxd5 16.exd5 Nxf2 (16...Bxf2+ 17.Kf1 Nc5 18.h3 Nxd3
19.hxg4) 17.dxe6 Nxd3+ 18.Kf1 fxe6 19.Qxd3 with hazy complications.

Lets now return to 10...Qb8 which may be not the most principal, but it prevents Whites most dangerous tactical


11.e4 is ineffective as 11...cxd4 12.Nxd4 Bd6 13.Bxe6 fxe6 14.Nxe6 is not a double attack (compared to 10...Qc7)
14...Kf7 15.Ng5+ Kg8.

11.Ne5 is well met by 11...Bd6 12.Nxd7 Nxd7 13.d5 c4.


Transferring play to the Meran structures is more interesting than taking on d5:
11...Nxd5 12.Nxd5 Bxd5 13.Bxd5 exd5 14.Rxd5?! is known to be safe for Black after 14...Be7 15.e4 Qb7
16.Bg5 f6 17.Bf4 0-0. White should try to attack, but 18.Nh4 Nb6 19.Rd3 Rfe8 20.Nf5 Bf8 21.Re1 Rad8 does
not look convincing for him.
However, the intermezzo 14.a4! cripples Blacks queenside pawn formation and assures White of a small, but
lasting pull 14...bxa4 15.Rxd5 Be7 16.Qc2 Qb3 17.Rd2 Nb6 18.Qe4 f6 19.Rd3.

12...dxe4 13.Bc2! is slightly better for White: (13.Ng5 is less precise: 13...c4 14.Ncxe4 Nxe4 15.Nxe4 Qe5
16.Bc2 0-0-0 17.g3 Qe6 18.Bf4 Bb4 19.Nd6+ Bxd6 20.Qxe6 fxe6 with a balanced endgame in Civitillo-
T.Carlsen, ICCF email 2007.) 13...Be7 14.Nxe4 Nxe4 15.Bxe4 Nf6 16.Bf4 Qc8 17.Bxb7 Qxb7 18.Bd6 Kf8
19.Qd2 Kg8 20.Bxe7 Qxe7 21.a4. While Black is preoccupied with completing an artificial castling, White
regains the pawn.
13.Nd5 Bd6 14.e5 Bxe5 15.Nxe5 Qxe5 16.Nc7+ Ke7 17.Re1
17.Qxe5+ Nxe5 18.Nxa8 Rxa8 19.Re1 Kd6 20.Bf4 Nfd7 is risky for White, Likas-Kavaliauskas, ICCF corr
17...Qxe2 18.Rxe2+ Kd8 19.Nxa8 Bxa8 20.Bxf7 favours White.
18.Qxe5+ Nxe5 19.Rxe5+ Kd6 20.Bf4 Kxc7 21.Rxc5+ Kb6 22.Rxc8 Rxc8 23.Be5 Rd8. The strong d-pawn balances
Whites bishop pair.

12.dxe6 fxe6 13.Bc2 Bd6


This set-up is connected with f2-f4. The other sensible plan is:
14.e4 0-0 when White will attack on the queenside with a4. He could execute it either outright, or after some

15.a4 Ne5 16.axb5 Nxf3+ 17.gxf3 Bxh2+ 18.Kg2 axb5 19.Rxa8 Bxa8 draw in view of 20.Nxb5 Qxb5 21.Kxh2 Qh5+
22.Kg2 Qg6+ 23.Kf1 Qh5=, Johnston-Eldridge, ICCF email 2011.

15.h3 Qc7 16.a3 Rae8 17.Be3 Ne5

Black has good counterplay against the enemy king, see game 18 Sasikiran-C. Hansen, Skanderborg 2003.

Note that the immediate:

14.a4 is not threatening to win a pawn so Black can castle 14...0-0 15.axb5 axb5 16.Rxa8 Bxa8 17.Nxb5 Bxf3 18.gxf3
Bxh2+ 19.Kg2 Qxb5 20.Kxh2. Whites king is stripped of pawn protection which is a good ground for future tactics.

14...0-0 15.f4 Nc5 16.b4


Black has the more active pieces. See game 19 Baranowski-Cipka, Lechenicher SchachServer 2011.

Chapter 7. The Old Main Line

Annotated Games

17. Jussupow 2640 Anand 2765

Dortmund 06.07.1997

1.d4 d5 2.c4 dxc4 3.e3 Nf6 4.Bxc4 e6 5.Nf3 c5 6.0-0 a6 7.Qe2 b5 8.Bb3 Bb7 9.a4 b4 10.Nbd2 Be7 11.Nc4 0-0
12.Rd1 Qc7

12...Nbd7 is also possible. It gives White extra options, but they are not really dangerous. For instance, White can try to
push e4-e5:
13.Nfe5 Qc7 14.e4 cxd4 15.Bf4 Qc5! 16.Rac1 Nxe5! 17.Nxe5 Qb6 18.Nc4 Qd8.
13.a5 This set-up is dubious against 12...Qc7 since Black would lead his queens knight to c6. But even with ...Nbd7,
Black is fine after 13...Qc7 14.Bd2 Bd5 or 14...Rfd8.

13.Bd2 Nbd7 14.Nfe5

White does not have a clear plan. His pieces are passive, especially pitiful is the Bd2-bishop. Black needs only to put in
...a5 and the poor bishop will be biting a granite. The other one, on b3, is also idle. Only the X-ray along the c-file
could be a cause for concern, but White is unable to capitalise on it:
14.Rac1 a5 (or 14...Rac8) 15.Nfe5 Rfd8 (15...Rac8=) transposes.

14...Rfd8 15.Rac1 a5 16.Be1

White is the proud owner of the c4-square, but Tarrasch claimed: if one piece is badly placed, your whole game is bad.
Anand aims to trade knights and leave the opponent with his useless bishops.

16...Nxe5 17.dxe5

17.Nxe5 Nd7 continues the same strategy. Play is equal after 18.Nxd7 Rxd7 19.dxc5 Rxd1 20.Bxd1 Rc8 (or 20...Rd8
21.c6 Bxc6 22.Qc2 Qb6=) 21.Qb5 Bd5.


Black has a pawn majority on the queenside, but he has to unblock the c4-square. He can pile hits on it by ...Ba6, ...Nb6.
It would be useful to remove, even temporary, the rook from c1 17...Rxd1! 18.Rxd1 Nd5. Whites task is not easy.
19.Nd6? simply drops a pawn to 19...Bxd6 20.exd6 Qxd6 21.e4 Nf4. And a neutral move like 19.h3 stumbles into
19...Ba6 20.Qf3 Bxc4 21.Bxc4 Nb6 22.Bb5 Rd8 when suddenly the queenside pawns become fearsome. If White
attempts to block them with b3, Black can even sacrifice one with ...c4 in order to get a b-passer.

18.f4 Ba6
18...Nb6 is roughly equal 19.Rxd8+ Qxd8 20.Rd1 Qc7 21.Nxb6 Qxb6 22.f5!? (or 22.Rd7 c4 23.Bxc4 Bc5 24.f5 Bc8
25.fxe6 Bxd7 26.exd7 Rd8 27.e6 fxe6 28.Qg4 Bxe3+ 29.Kh1 Kf7 30.Qh5+ Ke7=) 22...Ba6 23.Qf3 Rd8 24.Rxd8+


White should have stayed active with 19.Qf3

Then 19...Nb6 is not too impressive due to 20.Nxa5 Nxa4 21.Rxd8 Rxd8 22.Nc6. Alternatively:
19...Bxc4 20.Bxc4 Nb6
White cannot keep the blockade as 21.Rxd8+ Rxd8 22.b3? Nxc4 23.Rxc4 Rd3 is lost for him. Therefore, he
should continue:
21.Bb5 c4 22.Kh1 g6 23.Rxd8+ Rxd8 24.e4 with mutual chances. Black can always sacrifice a pawn to break Whites
pieces coordination and take over the initiative (for instance, c4-c3 or ...Bc5).


19...Nb6! is better for Black because he would have practically an extra pawn after 20.Nxb6 (relatively better is
20.Nxa5 Rxd1 21.Rxd1 Nxa4) 20...Qxb6 21.Bc4 Rxd1 22.Rxd1 Bxc4 23.Qxc4 Rd8 24.Rxd8+ Qxd8
25.Qc2 c4 26.Qxc4 Qd1 27.Kf1 Qxa4.


20.Bxc4! Nb6 21.Bd3 wins an important tempo and White keeps the blockade after 21...g6 22.b3.

20...Nb6 21.Rxd8+ Qxd8 22.Qb5 g6?!

22...c4! would have crowned Anands strategy 23.Bxc4 Rc8 24.b3 Rc5 25.Qa6 Qc7 26.e4 Nxc4.

23.Bd1 c4 24.Bf3 Rb8 25.Qxa5

25...c3! 26.bxc3 (26.b3 Qd3) 26...b3 27.Rb1

27.Qb5 Nd5 28.Qd3 Qb6 29.Bxd5 exd5 leads to a curious illustration on the theme bad bishop. White does not have
a plan for improvement despite his 2 extra pawns.

27...Qd3 28.Rxb3?

28.Rd1 Qc2! 29.Qa7 Rd8 30.Rxd8+ Bxd8 31.Qb8 Qd3 32.Be2 would have kept the balance. After the text, Anand gets
a winning attack.
28...Qxe3+ 29.Kf1 Bc5 30.Qb5 Rd8 31.c4 Rd3 32.Qe8+ Kg7 0-1

18. Sasikiran 2679 C.Hansen 2616


1.Nf3 d5 2.d4 Nf6 3.c4 dxc4 4.e3 e6 5.Bxc4 a6 6.0-0 c5 7.Qe2 b5 8.Bb3 Bb7 9.Rd1 Nbd7 10.Nc3 Qb8 11.d5 c4
12.dxe6 fxe6 13.Bc2 Bd6 14.e4 0-0

15.h3 Qc7 16.a3

16.Nd4 Rae8 17.Kh1 is not threatening anything. Black could rearrange his Q+B battery with the manoeuvre 17...Bc5
18.Be3 b4 19.Na4 Ba7 20.Nf3 Bb8.

After 16.Bg5, it is possible to trade bishops by 16...Bf4=. However, it would be more aspiring to keep more pieces for
an attack, e.g. 16...Bc5 17.Rd2 Nh5 18.Nd4 Nf4 when 19.Bxf4 Rxf4 20.Nxe6? fails to 20...Rxf2.

16...Rae8 17.Be3 Ne5 18.Nd4 Ng6

All the black pieces have targeted the enemy king so White should urgently seek counterplay on the kingside.
The obvious move is 19.b3 using the fact that 19...cxb3 20.Bxb3 Qxc3 is bad owing to 21.Rac1 Qa5 22.Nxe6 Qxa3
23.Nc5+ Kh8 24.Rxd6 Nxe4 25.Nxe4 Qxb3 26.Ng5.
However, Black can afford to ignore the threat on c4: 19...Nf4 20.Bxf4 Bxf4 21.bxc4 Re7!
21...bxc4 22.Rab1 would give White an initiative, but it turns out that the pawn will not run away. Blacks bishop
pair offers him plenty of threats in the centre:
22.Rab1 Bh2+ 23.Kh1 Be5 24.Rb4 Rd8 25.Qe3 Qc5 26.Nf3 Rxd1+ 27.Nxd1 Qxe3 28.Nxe3 Bd6 29.Rb3 Nxe4.
Sasikiran opts for a4 instead, but first he fixes the future target on b5.

19.b4 Qe7 20.Rab1 Kh8 21.Kh1 Rc8!

Hansen protects in advance the c4-pawn which will become weak after the imminent a3-a4. His idea is to counter 22.g3
Bb8 23.a4? by 23...a5!.

22.Nf3 Nh5
Blacks play is natural while White does not have a clear plan. The a4-idea has failed so he should try to hinder Blacks
advance by harassing his pieces with 23.Qd2! Rcd8 (23...Bb8 24.Qd7 trades queens) 24.Bg5 Nf6 25.Qe3. Instead
White persists with his initial plan and his position becomes critical.

23.Bg5 Qc7 24.a4 Nhf4 25.Bxf4 Nxf4 26.Qe3 Bc6 27.axb5 axb5 28.Nd4 Rb8 29.Rd2 Bd7

29...Rf7! was more consistent. The point is that 30.g3 is better for Black after 30...Nxh3 31.Nxc6 Qxc6 32.Kh2 Nf4
33.Rxd6 Qxd6 34.gxf4 Rxf4 35.e5 Rh4+.



It is difficult to refrain from this move, but objectively, 30...Ng6 31.Bd1 Ne5 32.Be2 g5!? was better.
Both players had missed the excellent defence 30...Nxh3 31.e5!! which ensures the vital tempo 31...Bxe5 32.Kg2 Nf4+
33.gxf4 Bxf4 34.Qe4 g6 35.Rdd1 with mutual chances, e.g. 35...e5 36.Nd5 Qb7 37.Rh1 Rf7 38.Kf1.

31.Kg2? Nf4+ 32.gxf4 Bxf4 33.Qe2 Rf6 34.Rh1 Bxd2 35.Qxd2 Rbf8 36.Nd1 Qe5 37.Ne2 Bc6 38.Qd4 Qg5+
39.Ng3 e5 40.Qc3 Qg4 41.Qxe5? Rxf2+ 42.Nxf2 Qf3+ 43.Kh3 Bd7+ 44.Ng4 Bxg4+ 45.Kh4 Bd7 0-1

19. Baranowski Cipka

Lechenicher SchachServer, 2011

1.d4 d5 2.c4 dxc4 3.Nf3 Nf6 4.e3 e6 5.Bxc4 a6 6.0-0 c5 7.Bb3 b5 8.Qe2 Bb7 9.Nc3 Nbd7 10.Rd1 Qb8 11.d5 c4
12.dxe6 fxe6 13.Bc2 Bd6 14.Ng5 0-0 15.f4

15...Nc5 16.b4 Nd3!

Practical chess has only seen 16...cxb3?! 17.axb3 h6 18.b4?! hxg5 with an edge. The email game Class-Mueller, 2008,
showed the correct way of handling this position: 18.Nf3! Qc7 19.Bb2 b4 20.Na4 Nxa4 21.Rxa4 with the better pawn

17.Bxd3 cxd3 18.Qb2?!

18.Rxd3 Bxb4 19.Rb1 Bxc3 is equal, but the text turns to be even worse for White.

18...e5 19.Rxd3 exf4 20.exf4 Qc7

Time to take stock. Blacks strong bishop pair more than compensates for the missing pawn. White has only moves to
stay in the game since the threats ...Ng4, ...Rae8 could be decisive.

21.Ne6 Qe7 22.Qe2 Rfe8 23.f5 Bxb4 24.a3 Bxc3 25.Rxc3 Bd5 26.Re3 Bc4 27.Qf3 Qa7 28.Kh1

White has survived the first wave. His next task is to maintain the material balance. It is not trivial since his pieces are
28...Bxe6 29.fxe6 Qd4 30.Rb1 Rac8 31.h3 Rc2 32.Re1 Qh4

A critical moment. Perhaps it was time to convert the initiative into something more tangible. 32...Qd5! 33.Qxd5 Nxd5
was winning the e6-pawn. White cannot defend against ...Rc6 or ...Nc7.

33.Re5 Qd4 34.Re1 Rc3 35.Qb7 Qd5 36.Qxd5 Nxd5

Black has finally come to the same idea, but with a rook on c3. This allows White to simplify further. His next move
would have been impossible with Blacks rook on c2 due to ...Nd5-c3xa4.
37.a4! bxa4 38.Re5 Nc7 39.e7 Nb5 40.Re6 Rc2 41.Bf4 a3 42.Rxa6 Rb2 43.Re1 a2 Draw.
Chapter 8. The 7.a4 Variation
Main Ideas

1.d4 d5 2.c4 dxc4 3.Nf3 Nf6 4.e3 e6 5.Bxc4 c5 6.0-0 a6 7.a4

This system has been one of Whites main weapons for many decades. Khalifman chose it for his repertoire series
Opening for White According to Kramnik. However, it has nearly disappeared at top level in the last 5 years. The
reason could be that White does not score well in the IQP structure after:

7...Nc6 8.Qe2

After 8.Nc3?! Be7=, White cannot even reach it.

8...cxd4 9.Rd1 Be7 10.exd4 0-0 11.Nc3 Nd5!

To be sure, the overall of 49% is deceiving since it is due mostly to the older main line:

12.Bd3 Ncb4 13.Bb1 where Whites attack is easily tamed by a timely ...f7-f5. This defence works well after both
possible developments of the black bishop:
13...Bd7 14.Ne5 Rc8 15.Ne4 Nc6 16.Qh5 Be8 17.Qh3 f5! or
13...b6!? 14.Ne5 Bb7 15.Ra3 Rc8

Blacks only problem here is the possible draw by perpetual following 16.Bxh7+ while 16.Nxd5 Qxd5 17.Rg3 f5 is fine
for him, see game 21 Haufe-Michalek, corr Switzerland, 2003. He can shun the draw by playing ...f5 earlier, e.g. on the
15th move, but objectively it is a bit premature.

Another forced draw is looming after:

12.Qe4 Nf6 13.Qh4 (13.Qf4 Nd5) 13...Nd5 14.Qg4 Nf6 15.Qg3 Nh5
White can fight on with 16.Qh3 Nf6 17.Bg5 Nb4 18.Qg3, but the hit on g7 is clearly insufficient for an advantage:
18...Re8 19.Ne5 Nfd5 20.Bh6 Bf8 21.Rac1 f6=.

One of the last fresh ideas for White was Kramniks h2-h4 in his world title match against Kasparov in 2000.

12.Bb3 Re8!? (12...Ncb4 is another decent line) 13.h4!?

Whites strategic idea is to push the h-pawn up to h6, then repeat the exercise on the other wing by a4-a5 and play on
the dark squares. I recommend to cut across his plans by 13...h6 (The same treatment works against 12.h4 h6, see game
22 Giobbi-Feldborg, ICCF email 2007)

You should remember two things about this position:

1. Keep control of e5. Accordingly, do not play Ncb4, keep the knight on c6.

2. Develop the bishop on d7.

Possible lines are 14.Bc2 Qc7 15.Bd2 Ndb4 16.Be4 Bd7 or 14.h5 Bd6.

Finally, lets consider the quiet development:

12.Bd2 Ncb4 13.Rac1 Bd7

We follow the principle of the spring and try to remain flexible. Our bishop will lay in ambush on e8, for instance:
14.Ne5 Rc8 15.Bb3 (or 15.Nxd5 Nxd5 16.Bb3 Rxc1 17.Rxc1 Be8) 15...Nf6 16.Bg5 Be8.

The 7.a4 system is essentially restrictive. It is not based on concrete straightforward variations so it demands more
understanding than memorisation. Here are some general principles which may be useful while playing against the
isolator on d4. Note that they are applicable only to the concrete position with a white pawn on a4. Every IQP position
should be treated only in the context of its concrete opening variation since little nuances in the piece placement may
change significantly the plans of both sides.

1. I recommend to develop the c8-bishop to d7, and not to b7. The only exception is when White has already played
Bc4-d3 so taking on d5 is less dangerous. We should avoid positions like this:

Polugaevsky Hort

Manila 1976
Stayed the pawn on b7, Black would not have serious problems. In the diagram position, however, the c6 and b6-
squares are weak and that assures White of a long-term pressure down the c-file. The game continued 23...Be6 24.h3
Rc8 25.Rxc8 Qxc8 26.Bf4 Qb7 27.Ng4 Kh8 28.Bd6 Bxg4 29.hxg4 Bxd6 30.Qxd6 a5 31.Rc1 h6 32.Rc7.

2. Having developed the bishop to d7, it is often better to retreat it to e8 rather than offering it for an exchange from c6.

Poluljahov Yakovich
Krasnodar 2002

This pawn structure is very solid, but also very drawish.

Compare it with:

Ubilava Semkov
Elenite 1986
15...Be8! The light-squared bishop could be useful on both sides. More importantly, Black needs c6 for his pieces.
Having a pawn on this square would serve well for a blockade, but not for an attack.
16.Qh3 f5!.

3. Besides developing his light-squared bishop, Black should also decide how to defend his kingside. In many lines ...f5
nips the enemy attack in the bud. Although ...g6 is often sufficient, too, ...f5 is a better long-term solution.
The previous example was somewhat extreme as 16...f5 gave Black the better game outright.
In the following position White has achieved his goal he set up the battery B+Q against Blacks king. However, that
does not guarantee him even full equality:


18...f5 19.Qe2 Nb4 20.Bb3 Bc6.

To finish with that theme, lets see how a world champion treats this structure!

Najdorf Smyslov
Sao Paulo 1978

It is a know fact that the knights are very important for Whites attack. Still, Smyslov decision is surprising: 17...Bxe5
18.Qxe5 and the opponents signed a draw!
Now I would like to offer you an alternative repertoire against 7.a4 which is more suitable for playing for a win:


This way Black avoids forced draws in several lines of the IQP position. A finer point is that he also shuns symmetrical
structures as he is prepared to recapture on c5 by the b-pawn!
In my opinion, only 8.e4!? can seriously challenge Blacks idea. Taking the gift would be horrible, but 8...cxd4 9.e5
Nfd7 10.Qxd4 Bb7 11.Qg4 Qc7 12.Nc3 is unclear thanks to my novelty:
12...h5! which breaks the coordination between the enemy pieces.

The main line is 8.Qe2 Bb7 9.Rd1 Qc7 10.Nc3 Be7!

The game I. Sokolov-Edouard, Linares 2014, went on 11.Bd2 0-0 12.Rac1 Rd8 13.Bd3 Nc6 14.Ne4 Nxe4 15.Bxe4

Note Blacks last move. In those positions, the pawn sac a4-a5 is perhaps Whites biggest positional threat. It might
look totally incorrect, but the following examples show that White usually obtains tangible initiative after it.

18.a5!? Nxa5 19.b4 Nc6! 20.bxc5 bxc5 21.Qc3 Nb4 22.Bb3 Bb7 23.Qc4 a5 24.e5 Be7=.


13.a5!? Nxa5 14.Ne4 Nxe4 15.Bxe4 Bb7 16.Bxb7 Nxb7 17.b4 Be7 18.Rfc1 Qb8 19.Rxa6.

16.a5! bxa5 17.Bc2! 0-0 18.Rd4 Rfc8 19.Rc4 Qe8 20.Rxa5.

Chapter 8. The 7.a4 Variation

Step by Step

1.d4 d5 2.c4 dxc4 3.Nf3 Nf6 4.e3 e6 5.Bxc4 c5 6.0-0 a6 7.a4

A. IQP positions; B. Black plays ...Qc7; C. 7...b6

A. 7...Nc6

It is not advisable to open the opponents bishop by an early exchange on d4. I noted in the chapter about 7.Bb3 that we
should produce an isolator on d4 only after White has committed himself with the passive set-up Qe2+Rd1.
We are not afraid of 8.dxc5 because the endgame after 8...Qxd1 9.Rxd1 Bxc5 is balanced, e.g. 10.b3 b6 11.Ba3 Be7.
Besides, trading queens is far from obligatory as 8...Bxc5 9.Qe2 is a really bad version of the Furman system where a4
looks superfluous.


After 8.Nc3 Be7, White cannot transpose to the main line anymore.
The point is that 9.Qe2 (9.d5 exd5 10.Nxd5 0-0! 11.Nxe7+ Qxe7 is fine for Black, e.g. 12.h3 Rd8 13.Qe2 Be6) 9...cxd4
10.Rd1 (well eagerly eat the pawn after 10.exd4?) runs into 10...e5! 11.exd4 exd4 when best is 12.Nxd4
12.Be3? is often seen in blitz 12...Bg4 13.h3 Bxf3 14.Qxf3 0-0.
12...Nxd4 13.Qe5 Qd6! 14.Qxd6 Bxd6 15.Rxd4 Bc5=. The top level game Ivanchuk-Wang Hao, Beijing 2013, went on
16.Rd1 0-0 17.Bf4 Bg4 18.Re1 Rac8 19.Nd5 Rfd8 20.Nxf6+ gxf6 21.b3 Kg7 22.h3 Bf5 23.Rad1 Bb4 24.Rxd8 Rxd8
25.Rc1 Ba3 26.Re1 Bb4 27.Rc1 Ba3 28.Re1, draw.

9.Ne5 could be met by either 9...Bd7 or 9...cxd4 10.Nxc6 bxc6 11.exd4 a5.

Of course, White can also take on c5 aiming to outplay the opponent in the drawish endgame. See game 20 Anand-
Kramnik, rapid Monte Carlo 2011.


8...Qc7 is analysed in line B.

9.Rd1 Be7 10.exd4 0-0 11.Nc3 Nd5

We have reached the main juncture of this line. In my opinion, Whites most challenging plan is connected with h2-h4!?
A1. 12.h4. This idea is also seen after A2. 12.Bb3. White also scores well with the seldom seen A3. 12.Bd2. A4.
12.Qe4 most often leads to repetition of moves. Finally, the old A5. 12.Bd3 has faded out of fashion.

Alternatively, 12.Ne5 neglects the rule that White should not easily trade his knights when having an isolator. Black
should not fear a kingside attack after 12...Nxc3 13.bxc3 Nxe5 14.dxe5 Qc7 15.Bd3 g6.

Or 12.Ne4 b6 (12...Nf6=) 13.Ne5 Bb7.

A1. 12.h4!?

Curiously, after 12.Bb3 Re8, the advance of the h-pawn is Whites main weapon while only 4 games featured it on
move 12. In my opinion, this idea deserves attention.
Hundreds of games confirmed the firmness of Blacks castling position against a piece attack. Apparently without the
help of pawns Whites threats are insufficient. Therefore, I think that we should seriously consider the march of the h-
pawn. Whites idea is to either weaken the dark-squares with h5-h6, or provoke a lever for attack in the event Black
plays h6 himself. The latter looks very dangerous, indeed, but it is also the most principled retort.


I do not like too much the thematic 12...Ncb4 as we might need the knight to other places, for instance, ...Na5 or
...Nxd4!. After 13.h5 b6 14.h6 g6 15.Bd2, White has a pull on the queenside and along the c-file. The weakness on f6
and g7 restricts Blacks counterplay. Of course, we can meet 13.h5 by 13...h6, but then White has 14.Ne5
Black is unable to complete his development. 14...Bd7 turns to be shaky due to 15.Bxh6 gxh6 16.Qg4+ Bg5 17.Bxd5,
e.g. 17...f5 18.Qf3 exd5 19.Na2 Nxa2 20.Qxd5+ Kg7 21.Qxd7+ Qxd7 22.Nxd7 Nb4 23.Nxf8.
Or 14...b6 15.Nxd5 exd5 16.Bb3 Bg5 17.f4.
Perhaps best is 14...Bd6, but 15.Bd2 keeps the tension.


We are on uncharted territory here. See in the Annotated Games section the only example I could find, game 22
Giobbi-Feldborg, ICCF email 2007.
In line A2 I analyse the same position with Bb3 Re8 included.

A2. 12.Bb3 Re8

The logic behind this strangely looking manoeuvre is that White has just made a waiting move, hoping to see 12...Ncb4
which would allow 13.Ne5. Therefore, Black decides to wait, too. Besides, ...Re8 undoubtedly improves his position so
it is not a waste of time.

The classical response 12...Ncb4 is also holding, though:

13.Ne5 Bd7 14.Qg4
14.Nxd5 Nxd5 enables the rook lift 15.Rd3, but Black possesses sufficient defensive resources 15...Bc6 16.Rg3
(16.Nxc6 bxc6 17.a5 Qc7 is very solid.) 16...Bf6 17.Bc2 g6 18.Bh6 Re8 19.Re1 Qb6 20.Rg4 Ne7 21.h4 Nf5.
14...Nf6! 15.Qg3 Bc6

16.Nxc6 bxc6 17.a5 only gives Black another target to attack the a5-pawn, for instance: 17...Nbd5 18.Qf3 Qc7
19.Bd2 Rab8! 20.Rac1 (20.Na4 runs into 20...Ng4 21.g3 Nge3 underlying the hanging state of the b3-bishop)
20...Rfd8 21.Nxd5 Nxd5 22.Rc2 Bb4=.
16...Ne8 17.Rac1
17.Nxc6 bxc6 18.Bf4 Nd5 is balanced. (Another possible set-up is ...Nd6-f5 18...Nd6 19.Be5 Nf5 20.Qg4 Nd5
21.Rac1 Ra7) 19.Nxd5 cxd5 20.Rac1 Bd6 21.Be5=, Martello-Pettersson, ICCF email 2010.
The position after 17.Rac1 appears to be the critical for the 12...Ncb4 line.
The isolator threatened to move so it should be blocked! 17...Rc8 18.Qg4 Kh8 19.d5!? offers White an initiative:
19...Bxd5 20.Nxd5 Rxc1 21.Bxc1 exd5 22.Bg5 (22.Qf5 Nf6! 23.Bg5 Qd6=, Magerramov-Ibragimov, Moscow
1991) 22...a5 23.Nd7 Bxg5 24.Nxf8 Nd6 25.Nd7 f5 26.Qg3 Ne4 27.Qb8.
18.Nxd5 Nxd5 19.Bd2 Nef6 20.Bc2 Rc8 21.Qb3 Qb6 22.Qxb6 Nxb6=.


Kramnik introduced this idea in his match for the world title in London 2000. Kasparov chose the natural retort:
13...Ncb4 14.h5 b6 15.Ne5 (15.h6!?) 15...Bb7 when 16.a5! revealed Kramniks deep strategic plan to exploit the weak
dark squares all over the board.

White retains a pull after 16...b5 17.h6 g6 18.Ne4 Rc8 19.Nc5. Commentators suggested possible improvements for
Black, but in my opinion Whites game is easier as he has the more active pieces.

I do not like 13...Na5 14.Bc2 Nb4 15.Ng5 Nxc2 16.Qxc2 Bxg5 17.hxg5 either so I recommend to keep the enemys
pieces at bay with:


This set-up has been tested only in email games. ...Bxh4 is not a big threat since White would get a strong attack by
lifting his rook via d3-h3. However, White must free the f3-knight which can then jump to e5 with Qg4 on the agenda.
He can do it by:

a) 14.g3 Nf6 15.Be3 Qa5 16.Rac1 Rd8 17.Bc2 b6, Pommrich-Foulds, ICCF email, 2011. Perhaps it is safer to develop
the bishop on d7 14...Bb4 15.Bd2 Bd7 16.Qd3 Nf6, Pommrich-Rosen, ICCF email, 2010, e.g. 17.Ne5 Rc8 followed
up by ...Qe7.

b) 14.h5 Bd6
Now 15.Ne5 could be met by 15...Qh4. The battery 15.Bc2 Bd7 16.Qd3 is effectively tamed by 16...Nf6 (we see
the merits of ...h6!) 17.Ne4 Nxe4 18.Qxe4 f5 19.Qe2 Nb4 20.Bb3 Bc6.
15.Bd2 and Black is at a crossroads. Perhaps safest is 15...Bd7
Klausen-Blauhut, ICCF email 2007, saw 15...Bb8 16.Rac1 Nf6 17.Ne4 Nxd4 18.Nxf6+ Qxf6 19.Nxd4 Qxd4,
but White has a serious initiative for the pawn.
15...Bc7 is similar: 16.Re1 (16.g4 Na5 17.Bc2 Nb4 18.Bb1 Nb3 19.Ra3 Nxd2) 16...Nf6 (16...Nf4 17.Bxf4 Bxf4
18.d5) 17.Ne4 Nxh5 (17...Nxd4 18.Nxf6+ Qxf6 19.Nxd4 Qxd4) 18.Rad1.

16...Be7 is a bit inconsistent with Blacks 14th move, but it is unclear how White could develop his attack. For
instance, 17.g4 Na5 18.Bc2 Rc8 19.Nc5 Bxc5 20.dxc5 Rxc5 looks very shaky, but the engines find adequate
defence against any Whites attempt to exploit the hanging state of Blacks pieces 21.Qd3 Nf6 22.g5 hxg5
23.Bxg5 Rd5 24.Qe2 Qc7 25.h6 Qc4.
After 16...Qe7, the straightforward 17.g4 Bf4 18.g5 does not work in view of 18...hxg5 19.Nexg5 Nxd4.
I chose for a main line:

14.Bc2 Qc7

In Donnelly-Kuehnel, ICCF email 2008, Black developed his light-squared bishop on b7, but fell under attack after
14...Bd6 15.Bd2 Ndb4 16.Bb1 b6 17.Ne4 Bc7 18.Ra3 Qe7 19.Ne5 Rd8 20.Bxh6.
It is better to stay more compact with ...Bd7, ...Rad8, for instance:

15.Bd2 Ndb4 16.Be4 Bd7

Black has finally led out his bishop and he will soon target the isolated pawn. Whites only active move is 17.Ne5, but
then both 17...Nxe5 18.dxe5 Bc6 and the thematic 17...f5 18.Bxc6 Nxc6 19.Bf4 Nxe5 20.Bxe5 Bd6 21.Rac1=, Sabaev-
DAdamo, ICCF email 2010, are balanced.

A3. 12.Bd2 Ncb4

Black could also borrow the idea from the previous line: 12...Re8 13.Rac1 Bf8 14.Ne4 Bd7, Korobov-Kharlov, Linares

13.Rac1 Bd7

Whenever possible, I prefer this development over ...b6.


A solid, but less flexible approach is 14...Bc6 15.Nxc6 Nxc6 16.Nxd5 exd5 17.Ba2 Qd7. The text keeps the option of
...Be8. It is always more interesting to play against an isolator, if only we could afford it.


After 15.Nxd5 Nxd5 16.Bb3 (16.b3 Ba3 17.Rb1 Be8), Black plays 16...Rxc1 17.Rxc1 Be8. He does not have any
weaknesses and the bishop could return to play from both wings.

15...Nf6 16.Bg5 Be8

Black maintains the tension without taking excessive risks. He can meet 17.Qf3 by 17...Bc6 or even 17...Rc7.

A4. 12.Qe4 Nf6 13.Qh4 (13.Qf4 Nd5) 13...Nd5 14.Qg4

If White offered a draw with 14.Qe4, we could play on with 14...Ncb4 15.Ne5 Ra7
15...b6 is objectively better, but 16.Nc6 Nxc6 17.Nxd5 exd5 18.Bxd5 Bg4 19.Re1 Bd7 20.Bxc6 Bxc6 21.Qxc6
Qxd4 22.Be3 Qxb2= leaves only scarce winning chances to Black.
16.Qg4 Kh8 17.Qh3 b6 with complex play, e.g. 18.Ne4 Qe8 19.Be2 Bd7.

14...Nf6 15.Qg3 Nh5 16.Qh3 Nf6

White has transferred his queen to h3, but it is arguable that it is better placed here since it is blocking the h-pawn.
Besides, Black is threatening 17...e5. 17.Ne5 is innocuous due to 17...Qc7 so the next move seems obligatory.

17.Bg5 Nb4

17...e5 is ineffective owing to 18.Qh4.

17...h6? 18.Bxh6! is horrible.


The threat on g7 is easily deflected 18.Qg3 Re8 19.Ne5 (19.Rac1 Nfd5 20.Bxe7 Rxe7 21.Re1 Nf6 22.Ne5 Bd7
23.Bb3) 19...Nfd5 20.Bh6 Bf8 21.Rac1 f6 22.Nd3 Nxd3 23.Bxd3 Qd6=, Nava Baro-Koch, IECG email, 1998.

18...Nfd5 19.Bd2 b6

The other development, 19...Bd7?, is impossible since the bishop is hanging after 20.Bxd5.

20.Rac1 Bb7
Black has finally finished development and he is not running any immediate risks. Play may continue 21.Qg3 Nf6
22.Bh6 Ne8 or 21.Ne4 Rc8 22.a5 Nc6.

A5. 12.Bd3 Ncb4 13.Bb1


Now that Whites bishop is not eyeing the d5-knight, this fianchetto is probably best, but 13...Bd7 is absolutely safe,
too. For instance:
14.Qe4 g6 15.Ne5 Bc6 (Petrosian drew against Botvinnik with 15...Bf6 16.Qf3 Bg7 17.h4 Nc6) 16.Qf3 Nxc3
17.Nxc6 Ne2+ 18.Qxe2 Nxc6 19.Be3 Nb4 20.Be4 Nd5 21.Rac1 Qd7 22.Bh6 draw, Makeev-Steiger, ICCF
email 2008.
14...Rc8 15.Qh5
15.Ne4 brings another piece to the kingside, but the thematic defence ...f5 becomes even more effective
15...Nc6 16.Qh5 Be8 17.Qh3 f5! 18.Nxc6 Bxc6 19.Nc5 Bxc5 20.dxc5 Nf4, Ubilava-Semkov, Elenite 1986.

Here 15...g6 16.Qh3 Bc6 is sufficient, but it is instructive to see that a world champion trusts in 15...f5!? even though it
is without a tempo: 16.Bd2 Bf6 17.Qe2 Bxe5 18.Qxe5 draw, Najdorf-Smyslov, Sao Paulo 1978.


14.Qe4 should not be a problem of course 14...f5 (14...g6 15.Bh6 Re8 16.Ne5 Bb7) 15.Qe2=;
14.a5 bxa5 15.Ne5 Bb7 16.Ra3 might end in a draw after 16...Rc8 17.Bxh7+ (17.Ne4 f5) 17...Kxh7 18.Qh5+ Kg8
19.Nxd5 (19.Ne4? Rxc1 20.Rxc1 Nf4) 19...Qxd5 20.Rg3= followed by Rxg7. Black can play on by 16...f5!? or 16...g6.

14...Bb7 15.Ra3 Rc8

15...f5 has been tested successfully in two games. I would not claim White is better, but his game is easy, for instance:
16.Nc4 Rf6 17.Nxd5 Bxd5 18.Bf4 followed up by Rc3. Still, it is a serious option against a weaker player when a draw
by perpetual check after 15...Rc8 16.Bxh7+ would be unwelcome.


16.Bxh7+ Kxh7 17.Qh5+ Kg8 18.Nxd5 Qxd5 19.Rg3 Rxc1 20.Rxg7+ is a draw (20.Rxc1 Bf6 or 20...Qe4 is unclear).

16...Qxd5 17.Rg3 f5 18.Bd2 Nc6

Black does not have any problems without knights on the board, see game 21 Haufe-Michalek, corr Switzerland, 2003.

B. 7...Nc6 8.Qe2 Qc7 9.Nc3


B1. 10.dxc5; B2. 10.Rd1

Minor alternatives are:
a) 10.b3 This continuation dooms the c4-bishop. The arising hanging pawns in the centre do not promise White an
attack since he is left with a passive bishop on b2 10...0-0 11.Bb2 cxd4 12.exd4 Na5 13.Ne5 b6 14.Rad1 (14.Rfd1
Bb7 15.Rac1 Nxc4 16.bxc4 Rfd8 17.f4 Rac8 18.Rd3 Ne8, Mayer-Thompson, Canada 1989) 14...Nxc4 15.bxc4 Bb7.

b) 10.Bd2 0-0 11.Rac1

11.Rfc1 commits both rooks to the queenside. Typical response is 11...b6 12.dxc5 bxc5.
11.Rfd1 hopes for 11...b6 12.d5. Black has a choice between: 11...cxd4 12.exd4 e5; 11...Bd7!?; the waiting
11...b6 12.Bd3 (12.dxc5 bxc5 13.Bd3 Rb8) 12...Bb7 13.Ne4 Nxe4 14.Bxe4 Qe7 15.dxc5 Bxc5=.

c) 10.h3 0-0 11.dxc5 Bxc5 12.e4 could be met by either 12...Nd4 or 12...Nd7=.

B1. 10.dxc5 Bxc5 11.Bd2

a) 11.e4 has been proved to be harmless. After 11...Ng4 12.g3, Black has at least two decent choices:
12...0-0 13.Bf4 e5 14.Nd5 Qd6 15.Bd2 Nf6 16.b4 Bd4 17.Rab1 Bg4=, Nilson-Andersen, ICCF email 2011;
12...Nge5 13.Nxe5 Nxe5 14.Bf4 g5.

b) 11.b3 0-0 12.Bb2 is a bit awkward as it might lead straight to a draw. In Mengual Bolo-Karasova, ICCF email 2010,
both sides made all the obvious moves to reach an endgame with opposite colour bishops: 12...Na5 13.Nd2

13...Bd7 14.Nce4 Nxe4 15.Nxe4 Be7 16.Rac1 Nxc4 17.Rxc4 Bc6 18.Qg4 e5 19.Ng3 g6 20.Nf5 Kh8 21.Qg3 f6
22.Nxe7 Qxe7 23.f4 Kg8 24.fxe5 fxe5 25.Rxf8+ Rxf8 26.Qxe5 Qxe5 27.Bxe5 Rf5 28.Bd4 Rg5 29.Rc2 Kf7 30.h3 h5
31.Rd2 draw.
Black could try to refine the move order, starting with 12...Bd7. Now if White attempts to transpose to the above-
mentioned game with 13.Nd2, Black will be fine after 13...Ne5! as the knight is much stronger in the centre than on a5.
We should, however, consider 13.Bd3! Bd6 14.Ne4 Nxe4 15.Bxe4 f5 16.Bd3 Be8 17.Bc4 Qe7 18.Nd4 although Black
does not have serious problems here.

11...0-0 12.Rfc1

In some lines White might need a rook on the a-file.

12.Rac1 could be met as in the main line with 12...Ne5!? 13.Nxe5 Qxe5 14.Kh1 Bd7 15.f4 Qh5=.

12.Bd3 does not make much sense after Bd2, but I found an email game which continued 12...b6 13.Rac1
13.a5 Nxa5 14.Ne4 is an interesting attempt to seize the initiative. 14...Nxe4 15.Bxe4 Bb7 16.Bxb7 Nxb7 17.b4
Be7 18.Rfc1 Qb8 19.Rxa6. We see an example where White made a good use of his clever 12th move. Black
must display some ingenuty in order to get counterplay 19...Rc8! 20.Rxc8+ Qxc8 21.Rxb6 Nd6 22.Be1 Nc4
23.Rb5 Qc7. Whites rook on b5 does not have any retreat square and Black can simply ignore it.
13...Bb7 14.Ne4 Nxe4 15.Bxe4 Qe7 16.Rfd1 Rac8=, Van Tricht-S.Larsen, ICCF corr 2003.

This natural move is a novelty. Practice has only seen 12...Rd8 and 12...Bd7.
The beautiful game Leitao-Molina, Rio de Janeiro 2013, reveals Whites main threat the advance e4-e5. After 12...Rd8
13.Bd3 Ne5 14.Be4 Nxf3+? (This presents White with a clear tempo, but 14...Qb8! 15.Nxe5 Qxe5 16.Bf3 Rb8 17.g3
h6 18.e4 still favours him.) 15.Bxf3 Qe5 16.g3 Bb4 17.Rd1 Rb8 18.e4 g5 19.h4 h6 20.hxg5 hxg5 21.Be3 Rxd1+
22.Rxd1 Be7 23.Bd4 Qc7 24.e5 Nd7 25.Be4 with a rout.

In my opinion, it is better to play ...Ne5 before White had time for Bd3-e4.

13.Nxe5 Qxe5

I do not see how White could make progress from here. The thematic 14.g3 Bd7 15.e4 Bc6 16.Bf4 is level after
Or 14.a5 Bd7 15.Na4 Bd6 16.f4 Qf5 17.Bd3 Bb5.

B2. 10.Rd1 0-0 11.h3

Black was threatening to meet any Whites move with ...e5, for instance: 11.Bd2 cxd4 12.exd4 e5 13.Bg5 exd4 14.Bxf6
dxc3 15.Bxc3 Bg4 or 11.b3 cxd4 12.exd4 e5.


After 12.d5 exd5 13.Nxd5 Nxd5 14.Bxd5 Bb7 15.e4 Rae8 16.Be3, it is good to swap dark-squared bishops with
16...Bf4 17.Bxf4 Qxf4 18.Qd2 Qc7=.


I like this capture. It offers a good way to strategically unbalance the game. Besides, 12...Bxc5 is also solid: 13.e4 Bb7
13...Nh5?! 14.Be3 Nf4 15.Qd2 Bxe3 16.Qxe3 Bb7 17.Rac1;
13...Nd7 14.e5 Ncxe5 15.Bf4 Nxf3+ 16.Qxf3 e5 17.Qxa8 Bb7 18.Nd5 Qc6 19.Qa7 Ra8 20.Qxa8+ Bxa8
14.e5 Nh5= McCraw-Nagley, IECC email 2000. The knight looks shaky on h5, but it would cost dearly in the event of
15.Kf1 Rfd8 (15...Rad8!?) 16.g4 Rxd1+ 17.Nxd1 Na5 18.Ba2 Qc6 19.Kg2 Rf8 20.gxh5 f6.

13.e4 Ne5 14.Bg5 Nxf3+ 15.Qxf3 Be5

Black is somewhat better. Skembris-Gavrilakis, Kallithea 2009 continued 16.Rac1 Bb7 17.Bxf6 Bxf6.

C. 7...b6!?

This is a sophisticated version of line B. Black wants to avoid symmetrical pawn structures which arise after dxc5. His
idea is to recapture on c5 by ...bxc5 which generally leads to a complex balanced game. Otherwise Black follows the
set-up with ...Qc7, ...Nc6 with one notable difference Black leads out his bishop to e7 instead of d6. This is aimed
against d4-d5 when the bishop would be more stable on e7.
In my opinion, the only way for White to exploit the delay of castling is C1. 8.e4!? which is practically unexplored
while the most popular is C2. 8.Qe2.

8.Nc3 Bb7 commonly transposes to line C2.

C1. 8.e4!? cxd4

It is strange that a super GM proved to be totally unprepared against the pawn sacrifice. In Kamsky-Edouard, Bilbao
2014, Black took the Greek gift, to struggle after:
8...Nxe4 9.d5 Be7 10.Re1 Nf6

11.Nc3 0-0 12.Bf4.

I also considered 8...Bb7 (by analogy with the line 7.e4 b5 8.Bd3 Bb7!), but then 9.d5!? exd5 10.exd5 guarantees White
a lasting pull, for instance: 10...Bxd5 11.Qe2+ Qe7 12.Qd3 Bxc4 13.Qxc4 Qb7 14.Ng5 Qd5 15.Re1+ Be7 16.Ne4
Qxc4 17.Nd6+ Kf8 18.Nxc4 Nbd7 19.Nc3 h6 20.h4 Bd8 21.Bf4 g6 22.Rad1.

9.e5 Nfd7

10.Bf4 Nc6 11.Nxd4 is certainly less dangerous. Besides 11...Nxd4, Black can safely take the pawn: 11...Ndxe5
12.Nxc6 Qxd1 13.Rxd1 Nxc6 14.Nc3 Na5 15.Be2 Bb7.

10...Bb7 11.Qg4 Qc7 12.Nc3 h5!

A critical moment. Black must disrupt the coordination of Whites pieces by exploiting the hanging state of the c4-
bishop. The only game I found, Ghaem Maghami-Ganguly, Abu Dhabi 2007, saw 12...Nc6?! 13.Bf4 h5 14.Qh3 with
strong pressure. After the text, Whites queen must go to f4, thus blocking natural development of the c1-bishop.

13.Qf4 Nc6 14.Re1 Be7

This is a rich position which offers plenty of chances to the better player. Black has active pieces and his only concern is
his king. It cannot remain in the middle, but it is difficult to say which is the best place to hide it. An obvious solution is
long castling, for instance:

15.Bd3 Nc5 (15...Rd8 16.Qg3 Kf8 is possible!) 16.Bc2 0-0-0 17.h4 f6.
Castling short is also an option, e.g.
15.Bd2 0-0 16.Rac1 Rad8.
Even ...h4 followed by ...Rh5, ...g6 deserves attention.

C2. 8.Qe2 Bb7 9.Rd1

(9.Nc3 Qc7 10.Rd1 Be7)

9...Qc7 10.Nc3 Be7!

Blacks move order is aimed against
11.d5 which would now lead to favourable for him exchanges after 11...exd5 12.Nxd5 Nxd5 13.Bxd5 Bxd5 14.Rxd5 0-
0 15.e4 Re8=.
The other thematic pawn advance 11.e4 does not bring White good results either: 11...cxd4 12.Nxd4 Nc6 13.Nxc6
Qxc6 14.Bd3 Nd7 15.Be3 0-0 (pay attention to the pawn sac 15...Nc5 16.a5!? which gives White an edge after
16...bxa5 17.Bc2 0-0 18.Rd4 Rfc8 19.Rc4 Qe8 20.Rxa5) 16.Rac1 Nc5. When the queens rook has left the a-file, this
is possible. 17.b4 Nxd3 18.Qxd3 Rfd8 19.Nd5 Qe8=.

After 11.dxc5, 11...Bxc5 may be objectively the best option. However, 11...bxc5 is sharper and more challenging.
Whites only active plan then is 12.e4 Nc6 13.e5 Nd7

Here the pawn sac 14.Ne4 Ncxe5 15.Bf4 0-0 is unclear while 14.Bf4 0-0 15.Rac1 offers Black a wide choice: 15...Nb6
16.Bd3 Rad8 when a6 is no weaker than a4; 15...Na5; 15...Rfd8.

11.Bd2 0-0 12.Rac1 Rd8 13.Bd3 Nc6 14.Ne4 Nxe4 15.Bxe4 a5!=

The last touch. Black prevents the most unpleasant threat in such positions, the push a4-a5, and he is ready for a long,
balanced game, I. Sokolov-Edouard, Linares 2014.
Chapter 8. The 7.a4 Variation
Annotated Games

20. Anand 2817 Kramnik 2785

Monte Carlo 13.03.2011

1.d4 Nf6 2.Nf3 d5 3.c4 e6 4.e3 c5 5.Nc3 dxc4 6.Bxc4 a6 7.a4 Nc6 8.0-0 Be7

9.dxc5 Qxd1

In the chapter for the Exchange Variation, I recommended to leave the opponent take on d8 himself. The same approach
is viable here, too. Black has a comfortable game with queens on the board. Korchnoi-Ivanchuk, Odessa 2007, saw
9...0-0! 10.Qc2 Bxc5 11.Rd1 Bd7 (11...Nb4!?) 12.Ne4 Nb4 13.Nxf6+ Qxf6 14.Qe2 Bc6 15.e4 Rfd8 16.Rxd8+ Qxd8
17.Bd2 Nc2 with an initiative.

10.Rxd1 Bxc5 11.Bd2 Bd7

The fianchetto is a solid equaliser: 11...b6 12.Bd3 Bb7 13.Ne4 Be7 14.Nxf6+ Bxf6 15.Rac1 0-0 16.Be4 Rfc8.


12.a5 looks more consistent. After 12...Bb4 13.Na4 Ke7 14.Nb6, Rublevsky opted for 14...Rad8 and signed a draw
following 15.Be2. The game might have continued 15...Bxd2 16.Nxd2 Nb4=. Morozevich preferred 14...Bxd2
15.Nxd2 Rab8 16.Ne4 Rhd8 when 17.Nc5 Ne5 18.Be2 Bc6 is equal. Instead 17.Nxf6 would split the kingside pawns,
but it is unclear how White could exploit it.

12...0-0 13.Ne4 Be7 14.Rac1 Rfd8

Whites queenside pawns are somewhat weak. Perhaps White should think about maintaining the balance with
exchanges 15.Nxf6+ Bxf6 16.Be2, borrowing from the opponent the defence of the b2-pawn, but Anand is still trying
to fight for the advantage.

15.Nc5?! Be8 16.Be2 Rd5!

It turns out that the b7-pawn is immune and the initiative passes to Black.

17.Nb3 Rad8 18.Be1 Rxd1 19.Rxd1 Rxd1 20.Bxd1 Nb4 21.Na5?

Apparently Anand could not brace himself for a passive defence with 21.Nc1 and seeks to reduce the material, but this
cost him a pawn:

21...Nd3 22.Bc3 Ne4 23.Nxb7 Ndxf2 24.Be2 Bxa4 25.Bd4

25.Bxa6 Nxc3 26.bxc3 Nd1

25...Bb5 26.Bxb5 axb5 27.Kf1

Black is a clear pawn up, but in a rapid game, Kramniks famous technique betrayed him.

27...Nd3 28.Ke2 Nb4 29.Bb6 Bf6 30.Bd4 Be7 31.Bb6 Nd5 32.Ba5 Bf6 33.b3 Bb2 (33...b4!) 34.Nd2 Nxd2 35.Bxd2
Be5 36.h3 f5? (36...Kf8) 37.Nd8 Draw.

21. Haufe 2381 Michalek 2442

corr Switzerland, 2003

1.d4 d5 2.c4 dxc4 3.Nf3 Nf6 4.e3 e6 5.Bxc4 c5 6.0-0 a6 7.a4 Nc6 8.Qe2 cxd4 9.Rd1 Be7 10.exd4 0-0 11.Nc3 Nd5
12.Bd3 Ncb4 13.Bb1 b6 14.Ne5 Bb7 15.Ra3 Rc8 16.Nxd5 Qxd5 17.Rg3 f5 18.Bd2 Nc6

Black removes the last white knight and then he could hope to start play of his own. Wegman-Weber, ICCF email 2003,
went from here 19.Bc3 Bh4 20.Rf3 Bf6 21.b4 Ne7 22.Ba2 Qd6 23.Rh3 Bd5 when Blacks knight is obviously stronger
than the enemy c3-bishop. Haufe decides to kill it right away.

19.Nxc6 Rxc6 20.Qe5 Bf6!?

20...Qxe5 21.dxe5 Rc7 22.Ba2 Kf7 23.Bc3 b5 24.axb5 axb5 25.Rgd3 Rfc8 26.Rd4 is equal.

21.Qxd5 exd5 22.Bc3 Bc8= 23.Ba2 Rd8 24.Rf3 g5

Suddenly Blacks weak f5-pawn threatens to move forth and open the way to the even weaker c8-bishop. It
becomes evident that Blacks bishops hide significant potential.

25.Re3 f4 26.Ree1 Bf5 27.Rd2 h5 28.f3 Kf7 29.Bb3 Re6 30.Rdd1 Rxe1+ 31.Rxe1 Kg6 32.Rd1 g4

Whites defence is difficult because he lacks any counterplay. He cannot trade light-squared bishops as it would untie
the black rook.
33.Kf2 Kg5 34.Kf1 h4 35.fxg4 Bxg4 36.Rd2 h3 37.Kg1 Bf5 38.gxh3 Be4
Blacks advance is straightforward and implacable. Nothing can stop the passed f-pawn so Black simply ignores h3.

39.Bd1 b5 40.axb5 axb5 41.Be2 Rb8 42.Bb4 f3 43.Bf1 Rc8 44.Bd6 Kh5 45.Bf4 Rc1 0-1
22. Giobbi 2288 Feldborg 2274
ICCF email 2007

1.Nf3 d5 2.d4 Nf6 3.c4 dxc4 4.e3 e6 5.Bxc4 c5 6.0-0 a6 7.a4 Nc6 8.Qe2 cxd4 9.Rd1 Be7 10.exd4 0-0 11.Nc3 Nd5
12.h4 h6 13.h5


White has several active plans. One of them is to transfer the queen or the d1-rook on the kingside after Ne5. Another
one is to build up a Q+B battery on the b1-h7 diagonal.Black aims to disturb all these ideas.
Now, if White makes a passive move like 14.Bd2, Black can wait with 14...Re8 intending to meet 15.Qd3 by 15...Nf6
16.Ne5 Nxh5 and 15.Bb3 by 15...Bd7 16.Bc2 Nf6.

14.Ne4 Qc7

14...Bd7 15.Ne5 Be8 is a decent alternative. The queen move enables ...Rd8 and ...Nf4. White suddenly faces
difficulties with his next move. 16.Nc5? simpy loses a pawn to 16...Nxd4.

15.Be3 Rd8 16.Rac1 Bd7 17.Ne1


This thematic idea is commonly Blacks main defence against threats against his king. The current situation did not
demand such drastic measures, though. A good, flexible option was 17...Be8 18.Nd3 Qe7 and Whites weaknesses on
the queenside become tangible. For instance, 19.Nec5 b6 20.Ne4 Na5. The game text forces play.

18.Nc5 f4 19.Bxd5 exd5 20.Nxd7 Rxd7 21.Nd3

21.Bd2 Bxd2 22.Qxd2 was roughly equal. White would have enough counterplay with Rc5, b2-b4-b5. White wants to
keep the bishops, but his decision is based on a bad assessment. It is obvious that the d2-bishop does not have

21...Bd6 22.Bd2 Rf7 23.Qg4 Qc8

Black has a fine game so Giobbi should have traded queens seeking further exchanges. But White overestimates his
position again.

24.Qf3 Qf5 25.Bc3 Re8 26.b4 Re4

Blacks rook arrives on time to hit d4. Now the lesser evil was 27.b5 axb5 28.axb5 Nxd4 29.Bxd4 Rxd4 30.Ne1 Re4
31.Qb3 Rc7 32.Rxc7 Bxc7 33.Nf3 Qxh5 34.Rxd5. Instead Giobbi goes into a hopeless position, probably due to some

27.Nc5? Bxc5 28.dxc5 d4 29.Ba1 Qd5 30.Rb1 Rf5 31.b5 axb5 32.axb5 Nd8 33.Qb3 Qxb3 34.Rxb3 Rxc5 35.Bxd4
Rd5 36.Rbd3 Ne6 37.Bc3 Rxb5 38.Rd5 Rxd5 39.Rxd5 Rc4 40.Be5 Rc1+ 41.Kh2 Rc5 42.Rxc5 Nxc5 0-1

Chapter 9. The 7.b3 Variation

Main Ideas

1.d4 d5 2.c4 dxc4 3.Nf3 Nf6 4.e3 e6 5.Bxc4 c5 6.0-0 a6 7.b3

This insidious variation is surprisingly fashionable at top level. White hopes to trick us with the move order and lure us
into a favourable for him line of the Exchange Variation. I suggest to pay him back in the same coin and transform the
game into the Furman System with colours reversed! Take care not to rush with ...Qe7 though.

7...cxd4! 8.Nxd4 Bd6! 9.Bb2 0-0

Next comes ...e5 with an excellent game:

10.Be2 e5 11.Nf3 Nc6;

10.Nd2 e5 11.N4f3 Nc6 (more precise than 11...Qe7 12.Ng5)

Black has more space, see game 23 Davidov-Raijmaekers, ICCF email 2008.
Chapter 9. The 7.b3 Variation
Step by Step

1.d4 d5 2.c4 dxc4 3.Nf3 Nf6 4.e3 e6 5.Bxc4 c5 6.0-0 a6 7.b3

This line is much more tricky than 7.a3 which does not have any sensible idea. Black can follow the same set-up as
against the Old Main Line: 7...b5 8.Ba2 Bb7 9.Nc3 Nbd7 10.Qe2 (or 10.Re1 Be7 11.e4 cxd4 12.Nxd4 Ne5 13.Bf4
Ng6 14.Bg3 0-0 15.e5 Nh5!) 10...Bd6 11.e4 cxd4 12.Nxd4 0-0 13.Rd1 Qb8 14.g3 Rd8 15.f3 Bc5 16.Be3 Qa7=,
Poluljahov-Yuneev, St Petersburg 1994.


When choosing 7.b3, White is secretly hoping to get an improved version of the Exchange Variation.
7...b5 8.Be2 Bb7 9.dxc5 Bxc5 10.Qxd8+ Kxd8 11.Bb2 Nbd7 12.Nbd2 Ke7 13.a4 bxa4 14.Rxa4;

In the chapter about 7...dxc5, I recommended to meet that set-up by ...b6. So 7...b6 looks logical. Indeed, 8.Bb2 Bb7
9.dxc5 Bxc5 would transpose while 9.Nbd2 Nbd7 10.Qe2 Be7 11.Rfd1 0-0 12.dxc5 Bxc5 is symmetric. However,
7...b6 could face 8.Ba3!? Nbd7 9.d5 with an initiative. Black cannot unload the position with 9...exd5 (9...Nxd5
10.Bxd5 exd5 11.Qxd5 Ra7 12.Bb2) 10.Bxd5 Ra7 11.Bb2 Be7 12.Nc3 0-0 13.Qc2 Nxd5 14.Nxd5 Nf6 since
15.Nxe7+ Qxe7 16.Ng5 would win the h7-pawn.

8.Nxd4 Bd6!

The passive move 8...Bd7 is much more popular, but practical experience has seen Black struggling after it. He remains
solid, but cramped. Computers may be able to defend successfully such positions, but I would not like to stay the whole
game under pressure without real chances for a counteratack. A good illustrative example is the game Vitiugov-Balogh,
Tsaghkadzor 2015, which continued 9.Bb2 Nc6 10.Be2 Be7 11.Nd2 0-0
12.Rc1 (12.Bf3 Rc8 13.Rc1 Nb4 14.Nc4 b5 15.Ne5 Rxc1 16.Bxc1 Be8 17.Bb2 Qc7 18.Nd3) 12...Nxd4 13.Bxd4 Nd5
(13...Bc6 14.Nc4 Nd7 15.Qd2) 14.Nc4 Bc6 15.Qd2 Bb4 16.Qb2.

I strongly recommend to play more actively. A natural follow up of 8...Bd6 is ...e5, ...Bg4(f5).

9.Bb2 0-0


In P.Nikolic-Ivanchuk, Moscow 1995, White traded dark-squared bishops with 10.Be2 Qc7 (10...e5!? 11.Nf3 Nc6)
11.Nf3 e5 12.Ba3 Bxa3 13.Nxa3
Black already has the slightly better game. The imminent ...e4 will give him a spatial advantage and a nice outpost
on d3.
13...Nc6 14.Qc2 Bg4 15.Rfd1 Rac8 16.Qb2 e4 17.Ne1 Bxe2 18.Qxe2 Ne5 19.h3 Rfd8 20.Rxd8+ Rxd8 21.Rd1 Nd3
(21...Nd5) 22.Nac2? (P.Nikolic cracks under pressure. 22.Qc2 Qe7 23.Nb1 was equal.) 22...Nd5 23.Nd4 Nc3
24.Qc2 Nb4 25.Qc1 Rxd4 0-1.

10...e5 11.N4f3 Nc6!

Black has more space and well coordinated pieces. See game 23 Davidov-Raijmaekers, email 2008.
Chapter 9. The 7.b3 Variation
Annotated Games

23. Davidov Raijmaekers

ICCF email 2008

1.d4 d5 2.c4 dxc4 3.Nf3 Nf6 4.e3 e6 5.Bxc4 c5 6.0-0 a6 7.b3 cxd4 8.Nxd4 Bd6 9.Bb2 0-0 10.Nd2 e5 11.N4f3


Black plays more often 11...Qe7, but it gives White a slight initiative after 12.Ng5 Bg4
12...Bf5 13.Qf3 Bg4 (13...Bg6 14.Nde4 Nbd7 15.Nxd6 Qxd6 16.Qxb7) 14.Qg3 Bf5 15.f4 exf4 16.exf4 is
dangerous for Black.
13.Qb1 Nc6 14.Bd3 h6 15.Nh7 Rfd8 16.Nxf6+ Qxf6 17.h3, Terentiev-Khismatullin, Sochi 2004, although 17...Be6!
18.Ne4 Qh4 would have kept the balance. In these lines, Black makes two moves with his queen.
After 11...Nc6!, 12.Ng5 is less effective as Black can retreat the d6-bishop to c7 in the event of 12...Bg4 13.Qb1
13...Bc7 14.h3 Bh5 15.Bd3 h6=.
Even simpler is 12...Bf5 13.Qf3 Bg4! 14.Qg3 Bf5= since 15.f4?! exf4 16.exf4 is dubious due to the hanging state of the
d2-knight (16...Nh5! 17.Qf3 Bc5+).


12.Be2 Be6 13.Nc4 Bxc4 14.Bxc4 is fine for Black.

The game Schneider-Degerman, Stockholm 1998, lasted only a few more moves: 14...Qe7 15.Qc2 Ba3 16.Bc3 Bb4
17.Bb2 Ba3 18.Bc3 Bb4 draw.
14...e4 15.Nd4 Nxd4 is also possible, but why to prolong the diagonal to the enemy dark-squared bishop?!

12...Qe7 13.Rd1


This is a typical way of contending the e4-square in the Furman System. Black should not let the white queen there
since it would be attacking both the e5- and b7-pawn. For instance, 13...Bg4 14.Ne4 Nxe4? (14...Rad8 is close to
equal.) 15.Qxe4 Bh5 16.a3, threatening Bd5.


Whites bishop is not much better on e4 14.Bd3 Rd8 15.Ne4 Nxe4 16.Bxe4 Bd7 followed up by ...f6. White lacks a
clear plan as 17.Nd2 will be parried with 17...b5.

14...Bf5 15.Nxf6+ Qxf6 16.Bd3 Bxd3 17.Qxd3 Rfd8 18.Qe4

The opening stage is over. Black is at least equal and if he managed to lift the blockade of e4, he would even enjoy
some initiative. His natural answer is to double the rooks on d7. Raijmaekers makes instead a superfluous move which,
however, does not change significantly the character of play.

18...Qe7 19.h4 Rd7 20.h5 Rad8 21.hxg6 hxg6 22.g3 Qe6 23.Kg2 Be7 24.Rxd7 Rxd7 25.Rh1

White has enough counterplay on the h-file to maintain the balance. The only way for Black to aspire to the advantage
would be 25...f5 26.Qb1 Bf6 (26...Qd5 27.Rh6 Kf7 28.Rh7+=), but 27.e4!= stops Blacks further advance in the centre.

25...Bf6 26.Bc3 Rd5 27.Qc4 b5 28.Qe4 Rc5 29.Rc1 Ne7 30.Bb2 Rxc1 31.Bxc1 Qd5 32.Qb4 Nc6 33.Qg4 Kg7 34.e4
Qd1 35.Be3 Qd3 36.Bg5 Bxg5 37.Nxg5 Qd8 38.Qh4 Qf6 39.Qh7+ Kf8 40.Qh6+ Kg8 41.a3 Nd4 42.b4 Nc2
43.Qh7+ Kf8 44.Qh6+ Kg8 45.Qh7+ Kf8 46.Qh6+ Kg8 47.Qh7+ Kf8 Draw.
Chapter 10. The 7.e4 Variation
Main Ideas

1.d4 d5 2.c4 dxc4 3.Nf3 Nf6 4.e3 e6 5.Bxc4 c5 6.0-0 a6 7.e4

This variation is similar to the Furman Variation. Also similar is our reaction we ignore Whites advance and hurry to
develop our queenside.

7...b5 8.Bd3 Bb7 9.e5

9.Re1 Be7! 10.a4 0-0!? 11.axb5 axb5 12.Rxa8 Bxa8

13.e5 Nfd7 14.Bxb5 Qb6 offers Black strong counterplay.


The main equaliser is 9...Nd5 10.a4 c4 11.Be4 Be7, see game 24 Fateev-Peigney, ICCF email 2006. Retreating the
knight to d7 may have a surprise effect. It leaves more strain in the centre which limits the enemy options on the

10.a4 bxa4 11.Ng5 Be7 12.Qh5 g6 13.Qh6 Bf8 14.Qh4

Whites centre is about to crack so he should be happy to draw with 14...Be7 15.Qh6=. The only way to keep the battle
on is to sacrifice a pawn 14.Qh3 cxd4 15.f4 Bd5 with unclear consequences.

Chapter 10. The 7.4 Variation

Step by Step

1.d4 d5 2.c4 dxc4 3.Nf3 Nf6 4.e3 e6 5.Bxc4 c5 6.0-0 a6 7.e4


Taking on e4 may lead to boring drawish endgames after 7...Nxe4 8.d5 (8.Qe2 Nf6 9.d5 Nxd5 10.Rd1 is tangled)
8...Be7 9.dxe6
White might also complicate things with 9.Re1 b5 10.Rxe4 bxc4 11.Nc3.
9...Bxe6 10.Bxe6=.

8.Bd3 Bb7 9.e5

9.Re1 is an important alternative. White hopes to get pressure on the queenside with a4 without conceding the d5-
square. A good illustration of his idea is:
9...cxd4 10.a4 bxa4 11.Rxa4 (11.Nxd4 Bc5 12.Rxa4 0-0) 11...Nfd7 12.Nxd4 Nc5? (12...Be7 13.Bc2 Nc6=),
Moiseenko-Nisipeanu, Eilat 2012, when 13.Bc2! would have been rather awkward for Black. I suggest not to
define the centre yet:
The point is that 10.dxc5 runs into 10...Bxe4
Ponomariov opted for 10...b4 and signed a draw against Aleksandrov.
It is dubious to relieve the tension in the centre since it gives White a free hand on the kingside 10...c4?! 11.Bc2
0-0 12.Nc3 b4 13.e5 Nfd7 14.Ne4 Bd5 15.Nfg5 h6 16.Nh3.
11.axb5 axb5 12.Rxa8 Bxa8

White can win a pawn with 13.e5 Nfd7 14.Bxb5 Qb6 15.Bxd7 Nxd7 16.dxc5 Nxc5, but in that event he would
be struggling for a draw owing to his uncoordinated pieces.
13.dxc5 Bxc5 14.e5 Ng4 is worse so he should probably opt for:
13.Nc3 b4
13...c4 14.Bb1 b4 15.e5 Bxf3 16.Qxf3 Nd5 is roughly equal.
14.Nb5 (14.e5 bxc3) 14...c4
14...cxd4 15.e5 Nfd7 16.Nfxd4 Nc6 17.Nxc6 Bxc6 18.Qc2 Qa8 19.Bxh7+ Kh8 is preferable for Black according
to the engines, but the line 20.Re3 Bxg2 does take guts to enter it over-the-board.
15.Bxc4 Bxe4=.

White should preserve his bishop for an attack so 9.Bg5 is strategically dubious. Black answers 9...Be7! 10.e5
10.Bxf6 Bxf6 11.e5 Be7 12.dxc5 Nd7! is pleasant for Black, e.g. 13.b4 Bxf3 14.Qxf3 Nxe5 15.Qe4 Nxd3
16.Rd1 f5.
10...Nfd7 (10...Nd5 is also possible) 11.Bxe7 Qxe7 12.a4 0-0! 13.axb5 axb5 14.Rxa8 Bxa8 15.Bxb5 Bxf3 with


9...Nd5!? has been tested more extensively and it is considered to be the main line. White has tried:

a) 10.Bg5 Qb6 11.dxc5 Bxc5 12.Nc3 h6 followed up by ...Nxc3.

b) 10.Nbd2 Nd7 (10...cxd4 11.a4 Nb4 12.Be4 Bxe4 13.Nxe4 Qd5 14.Qxd4 N8c6 15.Qxd5 exd5 16.Nc3 Nc2 17.Rb1
d4 should probably be drawn.) 11.Re1 h6;

c) 10.a4 c4 11.Be4 Be7

Practical experience favours Black. See game 24 Fateev-Peigney, ICCF email 2006. My suggestion keeps the centre
under tension and aims to surprise the opponent.

10.Ng5 Be7 11.Qh5 g6 12.Qh6 Nxe5 was unclear in Latier-W.Arencibia, Salamanca 1998. Stronger is 12...Bf8!
13.Qh4 cxd4.


10...cxd4 11.axb5 Bc5 12.Nbd2 axb5 13.Rxa8 Bxa8 14.Ne4 0-0 15.Bf4 (15.Nfg5 h6) 15...Bd5 deserves attention.


Moiseenko chose this move against Zeng, Fujairah City 2012.

11.Qxa4 Nc6 12.Be4 Qc7= defends everything.

11...Be7 12.Qh5
White got a big advantage in the stem game after 12...Bxg5? 13.Bxg5 Qc7 14.Nd2 Bd5 15.Ne4. The obvious
improvement is:

12...g6 13.Qh6 Bf8 14.Qh4

14.Qh3 cxd4 15.f4 Bd5


Chapter 10. The 7.4 Variation

Annotated Games

24. Fateev Peigney

W-ch WS/M/073 email 2006

1.d4 d5 2.c4 dxc4 3.Nf3 Nf6 4.e3 e6 5.Bxc4 c5 6.0-0 a6 7.e4 b5 8.Bd3 Bb7 9.e5 Nd5 10.a4


The necessity of this move is the biggest difference with 9...Nfd7 when 10.a4 could be answered by 10...bxa4. In the
diagram position, 10...bxa4?! allows a queens lift to the kingside with 11.dxc5, followed by Qd1xa4-g4. It might seem
that the whole Blacks set-up with ...Nd5 and ...c4 is very risky, to say the least. In principle, White gets all the
preconditions for a terrible attack a stable advantage in the centre, all his pieces are targeted on the enemy king.
However, we always should weigh all the factors in their integrity. Here, White lacks any easy way to transfer his
queen to the kingside. Stayed that piece on g4 or h5, Black would have been nearly lost.

11.Be4 Be7 12.Qe2

The only way for the queen to cross over to the kingside is through e4 so it makes contact with this juncture point.
Summerscale-Reefat, Edinburgh 2003, saw instead 12.Nbd2 Nc6 13.Bb1 Na5 14.axb5 axb5 15.Qe2 h6 16.Bc2 Nc6
17.Rxa8 Bxa8
It becomes clear that 18.Qe4 is ineffective due to 18...Ncb4 19.Bb1 g5. The game went on:
18.Be4 g5! 19.Nb1 g4 20.Ne1 Nxd4 21.Qxg4 h5! with a strong attack 22.Qd1 Nb3 23.Nf3 Rg8 24.Bh6 Nc5.

12...Nc6 13.axb5 axb5 14.Rxa8 Bxa8


Perhaps it is more realistic to eliminate to a draw with 15.b3 cxb3 16.Qxb5 Qb6 17.Qxb6 Nxb6 18.Nbd2 Kd7 19.Nxb3
Nxd4 20.Bxa8 Ne2+ 21.Kh1 Rxa8=.

15...Nc7 16.Be3 0-0 17.Ra1 Qd7 18.Bb1

An alternative is 18.h4 Rb8 19.Ng5 Bxg5 20.hxg5 Ne7 where Blacks knight is obviously better than the e3-bishop.
18...Rb8 is also a natural answer to 18.Bb1.

18...Bb7 19.h4 Ra8 20.Rxa8+ Bxa8 21.Qd1

21.Qc2 g6 22.h5 Nb4 23.Qd2 Nbd5 24.Bg5 Nxc3 25.bxc3 Bxf3 26.gxf3 b4 gives Black the upper hand.

21...Nb4 22.Ng5 Bxg5 23.Bxg5 h6 24.Be3 Qd8 25.Qg4 Kf8

White has finally managed to lead out his queen, but he cannot generate dangerous threats. At best, he can hope to
maintain the balance. So he opts for a waiting game.

26.Kh2 Nbd5 27.Bd2 Nxc3 28.bxc3 Nd5

Black cannot break through without ...b4 so he should have tried 28...Bc6 29.h5 b4 30.cxb4 Qd5 31.Be3 c3 with an

29.h5 Qc7 30.Qg3 Qa7 31.Qh4 Ne7 32.f3 Bc6 33.Qe1 b4 34.cxb4 Qxd4 35.Qe3 Qb2 36.Bh7 Ke8 37.Bc3 Nd5
38.Qc5 Qxc3 39.Qxc6+ Ke7 40.Qc5+ Draw.
Chapter 11. The Furman Variation
Main Ideas

1.d4 d5 2.c4 dxc4 3.Nf3 Nf6 4.e3 e6 5.Bxc4 c5 6.Qe2


I do not consider in this book the Steinitz system which aims to produce a white isolator on d4 by 6...Nc6. Besides, my
recommendation against the Exchange Variation with 6.0-0 a6 7.dxc5, namely 7...Bxc5, could be met by 8.Qe2 so we
need to be prepared for the Furman Variation which is defined by 6...a6 7.dxc5.

7.dxc5 Bxc5 8.0-0

The idea of the famous Soviet player and coach Semen Furman is very logical. White is planning to use his extra tempo
(even two, if we count ...a6) in a symmetrical position to get an attack after e3-e4-e5, Bd3, Nb1-c3-e4. His design is
beautifully simple and straightforward. It had been bringing excellent results until Black discovered that he should not
try to hinder e4-e5 at all. The classical chess rule to counterattack in the centre has proved valid once again. It turned
out that Whites attack could be ignored. Blacks best defence is quick development. Thus, against 8.e4, we should
bring our queenside pieces into play: 8...b5! 9.Bd3 Nc6 (threatening 10...Nb4) 10.Nbd2 Ng4 (with a tempo!) 11.0-0
Nge5 12.Nxe5 Nxe5 13.Bc2 0-0=.


The main equaliser is 8...Nc6 9.e4 b5 10.Bd3 Nd4, see game 25 Radjabov-Kasparov, Linares 2003. It is arguable which
8th move is stronger. In my opinion, castling is simpler, but it allows a forced draw.
9.e4 b5 10.Bd3 Nc6 11.e5 Nb4 12.Rd1 Nxd3 13.Rxd3 Nd5

A critical position for 8...0-0. White can force a draw by 14.Ng5 Be7 15.Nxh7 Kxh7 16.Qh5+ Kg8 17.Rg3 f5
18.Rxg7+ Kxg7 19.Bh6+ Kh7.

Another way to finish peacefully the game is 14.Nc3 Nxc3! 15.Rxc3 Be7 16.Rd3 Qc7=.

To take stock, the Furman Variation aims to reach a symmetrical position with a pawn on e5. However, it does not
promise White an advantage. Black should put his knight on d5 (but not on d7!) and trade it immediately, should an
enemy piece appear on c3.
In the Annotated Games section I analyse another, more sophisticated way to meet Whites plan which is also safe
Chapter 11. The Furman Variation
Step by Step

1.d4 d5 2.c4 dxc4 3.Nf3 Nf6 4.e3 e6 5.Bxc4 c5 6.Qe2


I would be happy to play against an isolated pawn with passive white pieces, as in the event of 6...cxd4 7.exd4 Nc6
8.Be3?!, but unfortunately 8.0-0! Nxd4? 9.Nxd4 Qxd4 10.Rd1 Qb6 11.Nc3 gives White a terrible attack.

Judge by yourself:
11...a6 12.Be3 Qa5 13.Nd5 Nxd5 14.Rxd5 Qc7 15.Rc1;

11...Be7 12.Be3 Qa5 13.Bb5+ Bd7 14.Bxd7+ Nxd7 15.Qf3

11...Bd7 12.Bg5 Bc6 (12...a6 13.Bxf6 gxf6 14.Nd5 Qd8 15.Qf3+; 12...Be7 13.Rxd7 Kxd7 14.Bb5++) 13.Nb5! Be7
14.Be3 Qa5 15.a3 Nd5 16.b4 Qd8. In The QGA (2008), Semkov proposed here 17.Bxd5 Bxd5 18.Bc5. Even stronger
is 17.Bc5.

7.dxc5 Bxc5 8.0-0

8.e4 is less flexible. We should face it the same way as in the main line:
8...b5! (Do not try to hinder e4-e5 with 8...Qc7) 9.Bd3
The bishop is not threatening anything from b3: 9.Bb3 Bb7 10.e5 Nd5 11.0-0 0-0

10.e5? Nb4;
10.Nc3 is well met by 10...e5 (this move is less effective after 10.Nbd2 since White would have Nb3 with control
over c5 and d4) 11.a4 (11.Bg5 Bg4) 11...Bg4 12.axb5 Nd4 13.Qd1 axb5 14.Rxa8 Qxa8 15.Nxb5 Bb4+, Troia-
Hoynck van Papendrecht, ICCF email 2008, when 16.Kf1 is the only move, but Blacks initiative more than
compensate for the pawn.
10...Ng4 11.0-0 Nge5 12.Nxe5 Nxe5 13.Bc2 0-0 14.Nb3 Bb6 15.Be3
White has tried in practice 15.Rd1 and 15.Bf4, but in both cases 15...Qf6 gives Black the upper hand.
15...Bxe3 16.Qxe3
Black has traded two minor pieces and he should not fear an attack anymore. He can now complete development with
16...Nc4!? 17.Qc3 Bb7 18.Nc5 Qe7=.
In the email game Joppich-Povchanic, ICCF 2004, Black entered the forced line 16...Qf6 17.Na5 Ng4 18.Qe2 Qxb2
19.e5 g6 20.Rab1 Qxa2 21.Ra1 Qb2 22.Rfb1 Qc3 23.Rc1 h5 24.h3 Qxe5 25.Qf3 Qh2+ 26.Kf1 Bd7 27.hxg4 hxg4
28.Qxg4 Rac8 and singed a draw after 29.Nb3 Rc4.


This seems to me easier to play than 8...Nc6 or 8...b5. The point is that Black does not define yet the stand of his
queenside knight and could put it on d7 in the event of 9.a3 b5 10.Bd3 Bb7 11.b4 Bd6 12.Bb2 and now 12...Nbd7=.
The flipside is that White has a forced draw.

The main equaliser is 8...Nc6 9.e4 (9.Rd1 Qe7) 9...b5

10.e5 bxc4 11.exf6 Qd3 evens the game: 12.Qxd3 cxd3 13.fxg7 Rg8 14.Rd1 (14.Bh6 Nd4) 14...Bb7 15.Rxd3
Nb4 16.Rc3 Rd8 17.Nbd2 Nd3 18.Nb3 Bb4 draw, Vitiugov-Stevic, Budva 2009.
Kramnik-Ponomariov, Dortmund 2012, saw 10...Bb7 11.Nc3 (11.e5 Nd5 12.Rd1 0-0!) 11...Nd4 12.Nxd4 Bxd4
13.e5 Nd5 14.Nxd5 Bxd5 15.Rd1 Bxb3 16.axb3 Qb6 17.Bd2 0-0 18.Ba5 Qa7 19.Rd2. Here Ponomariov chose
19...f6 20.exf6 Rxf6 21.Rxd4 Qxd4 22.Bc3 Qf4 23.Bxf6 Qxf6=, 19...Bb6 20.Bxb6 Qxb6 21.Rd6 Qc5 was also
11.Nxd4 Qxd4

White cannot rip dividends from his lead in development. See game 25 Radjabov-Kasparov, Linares 2003.

9.e4 b5 10.Bd3 Nc6 11.e5 Nb4


12.Bc2 Nxc2 13.Qxc2 Nd7 14.Ng5 g6 15.Qe4 Qc7 16.Qh4 looks dangerous, but in fact Blacks game is even better,
for instance: 16...h5 17.Re1 Bb7 18.Nc3 Be7 19.Qg3 Qc4 20.Nce4 Qc2 21.Bf4 Qxb2 22.Rad1 Bd5 23.Nc3 Nb6
24.Nge4 Bxe4 25.Nxe4 Rfd8.

12...Nxd3 13.Rxd3

13.exf6 Nxc1 14.Rxc1 Bb6 offers Black a strong bishop pair.

13...Nd5 14.Nc3

14.Ng5 Be7 15.Nxh7 is a draw. [15.Qh5 allows Black to defend the h7-square by queen 15...Bxg5 16.Bxg5 Qc7
17.Rh3 (17.Bh6 Ne7) 17...Qc2]

15...Kxh7 (15...Qc7 16.Nc3 Kxh7 17.Rh3+) 16.Qh5+ Kg8 with perpetual check after either 17.Rh3 f5 or 17.Rg3 f5
18.Rxg7+ Kxg7 19.Bh6+ Kh7=.


The game Brewer-Mostowik, ICCF email 2010, saw 14...Bb7 15.Ng5 Qb6 16.Qh5 Bxf2+ 17.Kh1 h6 18.Nge4 f5
19.Nxd5 Bxd5 20.Nf6+ Rxf6 21.exf6 Qb7 22.Rxd5 exd5 23.fxg7 Bd4 24.Bxh6 Qf7 25.Qxf7+ and White went on to
win. I think that Whites knight should be destroyed: 15...Nxc3 16.Qc2 Ne4 17.Rxd8 Rfxd8 18.Kf1 Nxg5 19.Bxg5
Rd5 20.Rd1 Bd4 21.Bf4 Rc8 22.Qe2 which keeps Black in the game. However, the sooner we kill the knight, the better.

15.Rxc3 Be7 16.Rd3 Qc7=.

Chapter 11. The Furman Variation
Annotated Games

25. Radjabov 2624 Kasparov 2847

Linares 03.03.2003

1.d4 d5 2.c4 dxc4 3.Nf3 Nf6 4.e3 e6 5.Bxc4 c5 6.Qe2 a6 7.dxc5 Bxc5 8.0-0 Nc6 9.e4 b5


10.Bd3 is not too popular. The obvious answer is 10...Nb4 11.Rd1 Qb6 12.Bg5 when 12...0-0 13.Bxf6 gxf6 14.Ne1
Kh8 15.a3 Nc6 16.b4 Bd4 17.Ra2 Rg8 was fine for Black in the email game Kalinaj-Hanzalik, 2009.
Top players, however, prefer 10...e5, e.g. 11.Bg5 (11.Be3 Bxe3 12.Qxe3 0-0 13.a4 Nb4) 11...Nd4 12.Nxd4 Bxd4,
Korchnoi-Karpov, Zuerich 2006.

10...Nd4 11.Nxd4 Qxd4 12.Be3

12.Nc3 Qe5 13.Be3 Bxe3 14.Qxe3 Ng4 15.Qh3 h5

Black is not fully developed yet, but the pair Qe5+Ng4 binds the whole white army. f4 would lose the exchange, while
I.Sokolovs 16.Rae1 is not threatening anything Van Wely simply answered 16...Bb7. White kept the balance in
Szymanski-Young, ICCF email 2008 with 16.Bd1 g6 17.Bf3 Kf8 18.Rad1 Bb7 19.a3 Kg7 20.Be2 Rad8 21.Qg3 Qxg3
22.hxg3 e5 23.Rxd8=.

12...Qe5! 13.Bxc5 Qxc5


After the bishops trade Whites only advantage is the better development and 14.e5 would lose it on the spot after
14...Nd7 15.Re1 Bb7 16.Nc3 0-0=, Rogozenco-Lintchevski, St Petersburg 2012.

14...Bb7 15.Rac1
In later games White tested 15.Rad1 when Black should answer 15...Rc8
Huzmans old recommendation 15...Qe5 defies common sense: 16.f4 Qc5+ 17.Kh1 b4 18.e5! bxc3 when
19.Ba4+! Kf8 20.f5!! is awkward.

Now the direct 16.e5 Nd7 forces White to spend a tempo on 17.Rfe1 since the f-pawn is pinned. After 17...Nb6
18.Rd3 0-0 19.Qh5 Nc4 20.Rg3 g6 21.Qh6 Rc7 Black is safe, Schwenck-Ruefenacht, ICCF email 2012. 22.Rh3
f5 23.exf6 Rxf6 offers Black counterplay on the f-file. This example helps us understand Tkachievs

16.Kh1!? 0-0 17.e5

17...Nd7 is impossible, 17...Qc6?! is bad since after 18.f3 Nd7 19.Ne4 the knight arrives at d6. Fortunately, the
position after
17...Nd5! 18.Nxd5 Bxd5 19.Bxd5 exd5 is balanced:

20.f4 Qc4=.



16.Bc2?! e5;
16.Rc2 Qe5=;
16.e5 Nd7 17.Qd2 Qxe5 18.Rfe1 is not winning since 18...Qh5 19.Ne4 (19.Bd5 Nc5) 19...0-0 saves the day.

16...Qb4 17.Rfd1

Svidler-Dominguez Perez, Foros 2007, did not offer any real improvement 17.e5 Ng4! 18.Qg5 0-0 19.h3 Nh6 20.Ne2


Black has accomplished his two most urgent tasks he has castled and also he has neutralised e4-e5. Indeed, 18.e5?!
Ng4 19.Qe2 Rc5 takes over the initiative. His only problem now is that the b7-bishop is biting on granite.

18...Rc7 19.Ne2 Qxd2 20.Rxd2 Rxc1+ 21.Nxc1 Rc8 22.Rc2

White could still have tortured the opponent for a while by 22.Nd3 Kf8 23.Bd1 Ke7 24.b4, but Black should be able to
defend, of course 24...Rc3 25.Nc5 Bc8 26.Rd3 Rc1=.

22...Nd7 23.Nd3 Kf8 24.Kf2 Ke7 25.Rxc8 Bxc8 26.e5 f6 27.exf6+ gxf6 28.f4 a5 29.Ke3 h6 30.a3 Kd6 31.g4 e5
32.h4 exf4+ 33.Kxf4 Nc5 34.Bc2 Nxd3+ 35.Bxd3 b4 36.axb4 axb4 37.Bf5 Bb7 38.Bc2 Bc8 39.Bf5 Bb7 Draw.

Part 3
Deviations From The Classical System

In this part, I consider:

Chapter 12 the gambit line 3.Nf3 Nf6 4.Nc3;

Chapter 13 3.Nf3 Nf6 4.Qa4+;
Chapter 14 rare lines: A. 3.Nf3 Nf6 4.g3?;
B. 3.Nc3?!;
C. 3.Nf3 Nf6 4.Na3?

You better study these lines carefully as they are not so rare in rapid chess and you cannot sidestep them with ...Bg4.
Chapter 12. The Gambit Line
Main Ideas

1.d4 d5 2.c4 dxc4 3.Nf3 Nf6 4.Nc3

This is the only true gambit line in the QGA. It is a rare guest in modern chess due to many reasons. Black has found
lately a safe way to equalise by quickly returning the pawn with ...Nb6.
Another argument against the gambit is that it is impractical for White to have it as a main repertoire. He should be
ready to spend a lot of time on home analysis as it gives Black an enormous choice. While Black can prepare for the
game in minutes. For instance, he could simply transpose to the Slav with 4...c6. The Chigorin fans could omit any
studying altogether as 4...Nc6 would be the main line of that opening. Notice that 4....Nc6 5.e4 is the subject of
Chapter 2, while 5.Qa4 Nd5 transposes to Chapter 13. Youll only need to take a look at the sideline 4....Nc6 5.d5 Na5
6.Qa4+ c6
7.b4 b5 8.Qxa5 Qxa5 9.bxa5 b4 10.Nd1 cxd5 with a tangled game, e.g. 11.g3 e6 12.Bg2 Rb8 13.Be3 (13.Nd4 Bc5)
13...Rb5 14.Bxa7 Rxa5.
Even 4...e6 is possible the Vienna Variation.
The main equaliser, however, is:

4...a6! 5.e4 (5.a4?! Nc6 6.e3 Na5!) 5...b5 6.e5 Nd5 7.a4 e6 8.axb5

White has regained the pawn, but he paid a high positional price he has surrendered the central light squares. Although
the e5-pawn ensures him a space advantage on the kingside, his chances for an attack are slim since Blacks pieces are
active and exert strong pressure on the d4-pawn. In the diagram position, Blacks most precise move is


Then 9.Bd2 Nb6 will force White sooner or later to defend d4 by Be3. Thus Black is winning a tempo on development
compared to 8...Nb6 9.Be3 (in the worst case Black could transpose to 8...Nb6 by ...Bb4-e7). Still, 8...Nb6 is three
times more popular than 8...Bb4. Why?! I see two plausible explanations.

1. 8...Bb4 concedes a forced draw after 9.Qc2 Nb6 10.Qe4 Qd5 11.Qg4 axb5!! 12.Rxa8 Qxa8

13.Qxg7 Qa1 14.Qxh8+ Kd7 15.Kd1 Bxc3 16.bxc3 Qa4+ 17.Ke1! Qa1=.

2. Ten years ago the diagram position was considered promising for White as he had won crushingly most games with
13.Be2. However, the current evaluation of this position is a slight plus for Black after 13...g6! Also the more natural
13...Kf8 is holding firmly.
I analyse 8...Nb6 in game 26 Markus-Stevic, Bol na Bracu 2014. Lets return now to our main line:

9.Bd2 Nb6 10.Be2 Bb7 11.bxa6 Rxa6 12.Rxa6 Nxa6 13.0-0 0-0 14.Be3


Chapter 12. The Gambit Line
Step by Step

1.d4 d5 2.c4 dxc4 3.Nf3 Nf6 4.Nc3

4...a6! 5.e4

White has also tried:

a) 5.e3 b5 6.a4 b4 7.Nb1 e6 8.Bxc4 c5 9.0-0 Bb7 10.Nbd2 offers Black comfortable equality. He can either take on d4,
or keep the tension with 10...Nbd7.

b) 5.a4?! Nc6
6.Bg5 h6 7.Bh4 (7.Bxf6 exf6 8.e3 Na5) 7...Na5 8.e4 c6;
6.a5 e6 7.e3 Bb4 8.Bxc4 0-0 9.0-0 Nxa5 10.Bd3 c5;
6.e4 Bg4 7.Bxc4 Bxf3 8.gxf3 Qxd4 9.Qb3 0-0-0 10.Bxf7 Nd7; (10...e5 11.0-0 Qb4)
Why not! The alternative 6...Bg4 7.Bxc4 e6 is only equal 8.h3 Bh5 9.0-0 (9.g4 Bg6 10.Nh4 Nb4 11.Nxg6 hxg6
12.Kf1 c5) 9...Bb4=.
7.Ne5 Be6 8.e4 (8.Be2 g6)

The forced 8...Nb3 9.d5 Nxa1 10.dxe6 Qxd1+ 11.Kxd1 0-0-0+ 12.Ke2 Nb3 13.Be3 fxe6 14.Nf7 Rg8 is also in
Blacks favour, K.Arkell-Fridman, Douglas 2014.
9.Be3 g6.

5...b5 6.e5 Nd5 7.a4

Well soon see that the text gives Black a tempo for development with 7...e6 so the knight jump 7.Ng5 deserves more
attention. After 7...e6 8.Qh5, Black has two possible set-ups:

a) 8...Qd7 9.Be2 (9.Nxd5 exd5 10.e6 Qe7) 9...Bb7 10.0-0 g6 11.Qh3 Be7 12.Qf3!? Rf8.

b) 8...Qe7! keeps an eye on g5 and enables the following combination:

9.Be2 Nc6 10.Be3 g6 11.Qf3 Nxd4 12.Bxd4 Qxg5 13.Nxd5 exd5 14.Qxd5 Bb4+ 15.Kf1 Rb8 16.Qc6+ Bd7 17.Qxc7
Rc8. Black returns the material, but takes over the initiative.


7...Nxc3 and 7...Bb7 also have adherents, but I do not enjoy positions where the opponent has a tangible initiative for
just a pawn. It is enough to see the recent game Debashis-Edouard, Caleta 2015: 7...Bb7 8.e6 fxe6 9.Ng5 Nxc3 10.bxc3
Qd5 11.Qg4 Nd7 12.Nxe6 Rc8 13.axb5 axb5 14.Ra7 to smell a rat. If one of the most ardent QGAs protagonists
lately falls into a pit like this, something is dubious about that line.

8.axb5 Bb4

The only drawback of this move is that White can make a forced draw in line A. If this result is unacceptable for some
reason, 8...Nb6 is to be preferred. Play is similar to line B, only Blacks bishop goes to e7. see game 26 Markus-Stevic,
Bol na Bracu 2014.
A. 9.Qc2; B. 9.Bd2

A. 9.Qc2 Nb6 10.Qe4 Qd5 11.Qg4 axb5!! (Fominyhs idea) 12.Rxa8 Qxa8 13.Be2

White can win a rook with 13.Qxg7 Qa1 14.Qxh8+ Kd7, but that is only enough for a draw after 15.Kd1 Bxc3 16.bxc3
Qa4+ 17.Ke1! Qa1.

After 13.Be2, Black has an ample choice. Only 13...0-0? has not caught up as Black should not be able to survive the

13...Rg8 14.0-0 Bxc3 15.bxc3 Nc6 16.Bd1 Ne7 17.Bc2 was first seen in Timman-Kasparov, Prague 1998. Kasparov
suggested 17...Bb7! as an improvement and it passed the test in Bazant-Svihel, Czech Republic 2000 18.Bxh7 Bxf3
19.gxf3 Rf8 (19...Rh8!? 20.Qxg7 Rxh7 21.Qxh7 Qxf3) 20.Qxg7 draw.

Later games also saw:

13...Kf8 14.0-0 Bxc3 15.bxc3 h6 16.Qh4 Nc6

17.Nd2 Bb7 18.Bf3 Nd5 19.Bxd5 exd5 20.f4 b4 21.f5 bxc3 22.Nf3 Kg8 23.Qh5 Qf8 24.e6 f6 draw, Glembek-
Aliekhin, ICCF email 2008.
17...Ne7 18.Bf3 Bb7
18...Nbd5 19.Qh5 (19.Ne4) 19...g6! 20.Qh3 Kg7 is interesting.
19.Bxb7 Qxb7 20.Ba3 Nbd5 21.Ne4 b4 22.Bxb4 Nxb4 23.cxb4 Qxb4 24.Ra1 Qb8, Grischuk-Karjakin, rapid
Moscow 2012.

The engines have no prejudices and their brute force calculation proves that Black can play for a win with the
horrible move:


It was successfully tested in email games. See game 27 Nemchenko-Lins, ICCF email 2012.

B. 9.Bd2 Nb6
This is a calm variation where Blacks play is simple and easy. He targets the d4-pawn aiming to win it or at least trade
it for the c7-pawn. Then e5 will be Blacks next mark. The full control of d5 secures him from trouble even if he loses
the c4-pawn. Move order is not too important.


10.b3 drops the d4-pawn to 10...Bb7! 11.bxa6

Or 11.Bxc4 Bxf3 12.Qxf3 Qxd4.
11...Bxf3 12.Qxf3 Qxd4.

10.bxa6 at best could transpose to the main line after 10...Rxa6 11.Rxa6 Nxa6 12.Be2. However, Black can also
recapture on a6 by bishop 11...Bxa6 12.Be2 Nc6

Normal continuation here would be 13.Be3 Ne7 while in Chernin-Zilberman, Rishon Le Ziyyon 1994, Whites attempt
to trade dark-squared bishops left him with very passive pieces 13.Nb1?! Be7 14.Bc3 0-0 15.0-0 Qd5 16.Ne1 Rd8
17.Nc2 Bb5 18.Bf3 Qd7.

White could refrain from taking on a6. A very interesting top level email game shows that Black is OK even if he
weakened his castling position with ...g6 11.Be2 Rxa1 12.Qxa1 0-0 13.0-0 Nc6 14.Be3 Bb7 15.Qb1 Ne7 16.Ng5 g6

17.Nce4 Nf5 18.Nf6+ Kg7 19.Bf3 Bxf3 20.Nxf3 Be7 21.Ne4 draw, Soltau-Gil, ICCF email 2010.
In my opinion, it is more human to prepare ...c5, for instance: 11...0-0 12.0-0 Bb7 13.Be3 Rxa1 14.Qxa1 h6!? 15.Qb1

10...Bb7 11.bxa6 Rxa6 12.Rxa6 Nxa6 13.0-0 0-0 14.Be3


Another solid approach is 14...Be7 15.Nd2 Qa8 16.f3 Rd8 17.Nxc4 Nxc4 18.Bxc4 c5 19.Ne2 cxd4 20.Nxd4 Nb4=.
Whites extra pawn is irrelevant here.

15.Nb5 Bd5 16.Nd6 f6

Black has sufficient counterplay, Barnaure-Rogozenco, Predeal 2006.

Chapter 12. The Gambit Line

Annotated Games

26. Markus 2609 Stevic 2613

Bol na Bracu 17.05.2014

1.d4 d5 2.c4 dxc4 3.Nf3 Nf6 4.Nc3 a6 5.e4 b5 6.e5 Nd5 7.a4 e6 8.axb5 Nb6


It is better to see ...Bb7 before taking on a6 as 9.bxa6 Rxa6 10.Rxa6 allows, besides 10...Nxa6, also 10...Bxa6 and
White is left without any sensible plan. The next few moves of both sides are easy 11.Be2 Be7 12.0-0 0-0 13.Be3
Nc6, but here White should probably start thinking of how to maintain the balance. Either 14.b3 Nb4 or 14.Qa1 Bb7
15.Rd1 Nb4 16.Nd2 Nc2 17.Qa7 Nxe3 18.fxe3 Bc6 are nice for Black.
Even worse would be to trade dark-squared bishops with 10.Bg5 Be7 11.Bxe7 Qxe7 12.Be2 0-0 13.0-0 Bb7 14.Qc2
Rxa1 15.Rxa1 Nc6 16.Ne4 Nb4 17.Qb1 Bd5.
The text enables the threat Nf3-d2xc4. This is the main difference with the line 8...Bb4 9.Bd2, but still it is insufficient
for an advantage.


The set-up from the 8...Bb4 line does not work here 9...Bb7 10.bxa6 Rxa6 11.Rxa6 Nxa6 (11...Bxa6 12.Be2 Be7 13.0-
0 0-0 14.Qb1 Nc6 15.Qe4 invites Whites queen to the kingside.)
More precise than 12.Nd2?! Nb4 13.Nxc4 Nxc4 14.Bxc4 Bxg2 15.Rg1 Bc6 followed by ...g6 or
12.Bxc4? Nxc4 13.Qa4+ c6 14.Qxc4 Nb4 15.Qb3 Ba6.
12...Nb4 does not save the pawn either 13.0-0 Be7 14.Nd2 Qa8 15.g3 0-0 16.Nxc4 Nxc4 17.Bxc4 Bg2 18.Re1
Bh1 19.Bf1 Rb8 20.Qa4.
13.Nd2 0-0 14.Nxc4 Nxc4 15.Bxc4.

10.Rxa8 Nxa8 11.Nxb5 Nb6


12.Nd2 does not win the pawn due to 12...Ba6 13.Nc3 Nc6. Note that Black should not afford to lose his castling rights
so he must meet 14.Nde4 by some useful waiting move like 14...Bb7, but not 14...Be7?! 15.Qg4! Kf8 16.Qd1.

12...Be7 13.0-0 0-0 14.Qc2 Nc6 15.Qe4 Nb4 16.Qg4


The whole Blacks defence is based on this move.

17.exf6 Bxf6 18.Ne5

Najer-Sulskis, Linares 2001, saw 18.Rd1!? N4d5 19.Qe4 when 19...Qe8 20.Nc3 Bb7 looks playable. The activity of
Blacks pieces compensates for the weak c4- and e6-pawns.

18...Nc2 19.Qe4 Nxe3

20.Qxe3 Nd5 21.Qe4 Bxe5 22.Qxe5 Rf5 23.Qe4 Rf4 24.Qc2 c6 25.Na3 Rxd4 26.Nxc4 Ba6 simplifies to a level
position, but the text leads to an easy draw.

20...Qd5 21.Nc3 Qxe4 22.Nxe4 Bxe5 23.Rxf8+ Kxf8 24.dxe5 Nd7 25.Bxc4 Nxe5 26.Be2 Ke7 27.Kf2 Bb7 28.Nc3
Bc6 29.g3 h6 30.h4 g5 31.hxg5 hxg5 Draw.

27. Nemchenko 2454 Lins 2454

ICCF email 2012

1.d4 d5 2.c4 dxc4 3.Nf3 Nf6 4.Nc3 a6 5.e4 b5 6.e5 Nd5 7.a4 e6 8.axb5 Bb4 9.Qc2 Nb6 10.Qe4 Qd5 11.Qg4 axb5
12.Rxa8 Qxa8 13.Be2


Humans do not play like that as the complex of weak dark squares on the kingside looks terrifying. 13...Kf8 is certainly
more natural and safe. However, the text is probably the best way to solve the big technical problems which Black
would face if he tried to convert his extra pawn. If he just stayed in the centre, Black should be out of risk. But his only
sensible attempt of playing for a win is to invade the queenside with his queen. Then Whites attack on the opposite
part of the board could become crushing.

Practical experience has proved that Blacks king often should seek escape to the queenside anyway. The logic behind
13...g6 is to abandon the kingside outright and prepare ...Kd7 and even ...Kc6.


In Dillenburg-Chukanov, ICCF email 2011 White decided to transfer his queen to the queenside and he quickly faced
difficulties after 14.0-0 Bxc3 15.bxc3 h6 16.Qf4 N8d7 17.Qd2 Na4 18.Qa2 Ndb6 19.Qa3
19...Qe4 20.Re1 Qc2 21.Bd2 Bb7 22.Rc1 Qb3 23.Qa1 Qb2.

14...h6 15.Nge4 Bb7 16.0-0

16.Nf6+ Kd8 17.0-0 Bxc3 18.bxc3 N8d7 19.Nxd7 Kxd7 20.Qf4 Rh7!? shows an important defensive stand. The
comically looking rook protects the whole flank all by itself.

16...Bxc3 17.Nxc3

A critical position. In my opinion, best is 17...Bc6, intending to block the centre with ...Nd5, for instance: 18.Be3 N8d7
19.Rc1 Nd5 20.Nxd5 Bxd5.

17...b4 18.Nb5 Kd7 19.Qf4 Rh7

Black needs only one more move to cement his centre so the following sacrifice is a must.

20.d5 Bxd5 21.Be3 Qa6 22.Bxb6 Qxb5 23.Be3 Kc8 24.h4 (24.b3 Qb7) 24...g5 25.Qg4

25.hxg5 hxg5 26.Qxg5 Qe8 suddenly turns the tables on the kingside as both g- and h-file are vulnerable. The game
now walks to a draw.

25...Rh8 26.Rc1 gxh4 27.Qxh4 Rg8 28.Bxc4 Bxc4 29.Qxc4 Qxc4 30.Rxc4 Nd7 Draw.

Chapter 13. 4.Qa4+

Main Ideas

1.d4 d5 2.c4 dxc4 3.Nf3 Nf6 4.Qa4+

This check gives us the possibility of counter-attacking the centre by:

4...Nc6! 5.Qxc4 e5!?

This sacrifice is often a shock for White. In practic e first players always take by pawn which immediately hands Black
the initiative:

6.dxe5?! Be6
White must find several accurate moves in order to keep the balance.

See game 28 Ahlander-Brynell, Skara 2002 for 7.Qa4 Nd7.

7...Ng4!? 8.Qxd8+ Rxd8 9.Bf4 Nb4 10.Na3 f6!? 11.exf6 c6! with an initiative.

6.Nxe5! would be a novelty, and it leads to dynamic play after 6...Nxe5 7.dxe5 Be6! 8.Qc2 Ng4 9.Bf4 g5! 10.Bg3 Bg7
11.e3 Nxe5.

5.Nc3 is another major line. Black can fight for the initiative with 5...Nd5! 6.Qxc4 Ndb4! 7.Qb3

7...Be6! Do not lose the initiative for the d4 pawn yet! 8.Qa4 Bd7 9.Qd1 e5 10.dxe5 Bf5 11.Bg5 f6.
Chapter 13. 4.Qa4+
Step by Step

A. 1.d4 d5 2.c4 dxc4 3.Nf3 Nf6 4.Qa4+


Of course, 4...c6 is a solid alternative. Any fan of the Slav could even save time on preparation for the a4-check since
the arising side-line is considered harmless. Still, 4...Nc6 is the most testing (and sharp!) retort.


5.e3 Nd7 practically ssures Black of the bishop pair advantage since 6.Qxc4 would be met by 6...e5!. So: 6.Bxc4 Nb6
7.Qc2 Nxc4 8.Qxc4 Be6 (8...Bg4 9.Nbd2 e6 10.Ne5) 9.Qe2 Bg4 10.h3 Bh5 11.Nc3 e6 12.0-0 Bd6 13.Rd1
We could further unbalance the game with 13...f5 followed up by long castling.

5.Qxc4 is popular, but it offers Black a pleasant choice:

a) 5...Be6 6.Qa4 a6 7.Nc3 Bf5

8.e3 b5 9.Qb3 e6;
8.g3?! b5 9.Qb3 e6.
8...e6 (8...b5 9.Qd1) 9.Bg5 Be7=.

b) 5...e5!? with the following branches:

6.dxe5?! Be6
White is already on the defensive! After 7.Qb5 or 7.Qa4, 7...Nd7 sets concrete threats. See game 28 Ahlander-
Brynell, Skara 2002.
It is more reasonable to trade queens:
7.Qh4 Ng4!?
Adding more oil into the fire. 7...Bb4+ 8.Bd2 Ng4 9.Qxd8+ Rxd8 10.Bxb4 Nxb4 11.Na3 Nc6 regains the pawn
with good position, but the text is more fun.
8.Qxd8+ Rxd8 9.Bf4 Nb4 10.Na3

10...f6!? (10...Nxa2=) 11.exf6 c6!

White must defend with only moves now.
12.e3 Nd5 13.Be2 Nxf4 14.exf4 Bb4+ 15.Kf1 Nxf6 16.g3 Bh3+
Or 16...Bxa3 17.bxa3 Bd5.
17.Kg1 0-0. White is still in serious danger.
6.Nxe5! Nxe5 7.dxe5 is safer:

I have found only one email game which, however, saw 7...Ng4 8.Bf4?! (8.Nc3! c6 9.Qe4 Bc5 10.e3 is in
Whites favour.) 8...c6 9.e3 Qa5+ 10.Nc3 Nxe5=.
Or 8.Qb5+ Nd7 9.Qxb7 Rb8 10.Qe4 Bb4+.
8...Ng4 9.Bf4 g5!
It is not enough to regain the pawn with 9...c6 10.Nc3 Qa5 11.e3 Nxe5, because White will have a mobile pawn
majority on the kingside, e.g. 12.Be2 Ng6 13.Bg3 Be7 14.0-0 0-0 15.a3.
10.Bg3 Bg7 11.e3 Nxe5 12.h4 (If White refrained from this move, Black could attack later with ...f5-f4) 12...g4 13.Be2
0-0 14.0-0 Qf6

Blacks pieces are active which is a good sign. Play may continue:
15.Nd2 Qg6 16.Qxc7 Rac8 17.Qxb7 Qc2 18.Rfd1 Rfd8 19.Rac1 Qxc1 20.Rxc1 Rxc1+ 21.Nf1 Ng6 22.h5 Nf8.

5...Nd5 6.Qxc4

6.e4 Nb6 7.Qd1 Bg4 8.d5 Ne5 9.Bf4 resembles the 3.e4 Nc6 variation with one difference the knight is on b6 instead
of f6. We can exploit that by the typical Chigorin idea:

9...Bxf3 10.gxf3 Qd6 11.Be3 (11.Bg3 0-0-0 12.f4 Ned7) 11...0-0-0 with an initiative.


6...Nb6 is more popular, but I do not like Blacks position after either retreat of Whites queen:
7.Qd3 e5 8.dxe5 Qxd3 9.exd3 Nb4 10.Kd2 Bg4 11.a3 Nc6 12.Kc2;
7.Qb3 e5 8.dxe5 Be6 9.Qc2 Nd4 10.Qe4 Bc5 11.Nxd4 Qxd4 12.e3 Qxe4 13.Nxe4 Bb4+ 14.Bd2 Bxd2+ 15.Nxd2 0-0-
0 16.Be2 Rd5 17.f4 Rc5 18.Bd1 Rd8 19.Rf1 Nd5 20.Nb3 Rb5 21.Kf2, Landa-Godena, Reggio Emilia 2010.

7.Qb3 Be6!

7...Nxd4 8.Nxd4 Qxd4 9.a3 is slightly better for White as well as 7...e5 8.d5!.

8.Qa4 Bd7

Black is not forced to repeat moves after 9.Qb3. He can keep on playing with 9...e5 10.dxe5 Be6 11.Qa4 a6 12.Bd2 b5
13.Qd1 Nd4 14.Nxd4 Qxd4 15.e3 Qxe5.

9...e5 10.dxe5 Bf5 11.Bg5 f6

Black owns the initiative.

Chapter 13. 4.Qa4+
Annotated Games

28. Ahlander 2419 Brynell 2506

Skara 2002

1.d4 d5 2.Nf3 Nf6 3.c4 dxc4 4.Qa4+ Nc6 5.Qxc4 e5 6.dxe5 Be6 7.Qa4 Nd7


Perhaps the best defence is 8.Bd2 Nc5 9.Qd1

Blacks play is similar after 9.Qc2. He castles long and counts on more active pieces: 9...Qd7 10.e4 0-0-0 11.Be2
Nd3+ 12.Qxd3 Qxd3 13.Bxd3 Rxd3 The bishop pair and active rook more than compensate the doubled pawn,
for instance: 14.0-0 Bg4 or 14.Nc3 Bc5 15.Rc1 Rhd8.
Black has several other promising ideas:
9...Ne4 10.e3 Qd7 11.Nc3 Nxd2 12.Qxd2 Qxd2+ 13.Kxd2 0-0-0+ 14.Kc1 a6.
9...Qe7 10.Bg5 (10.Bc3 Rd8 11.Nbd2 Ne4 12.e3 Nxc3) 10...f6 11.exf6 gxf6 12.Be3 Ne4
13.Nc3 Nxc3 14.bxc3 Qa3 15.Qc1 Qxc1+ 16.Rxc1 Bxa2 17.g3 a5.
10.g3 0-0-0 11.Bg2 Ne4 12.0-0 Nxd2 13.Nbxd2 Nxe5 14.Nxe5 Qxd2 15.Qa4 Qd4

White escapes with a draw after 16.Bxb7+! Kxb7 17.Qc6+ Kb8 18.Rad1 Qxd1 19.Qb5+ Ka8 20.Qc6+.

8...Nc5 9.Qd1

9.Qc2 Nb4 10.Qb1 Qd7 11.a3 Nc6 is difficult for White.

9...Qxd1+ 10.Kxd1 0-0-0+ 11.Bd2


Whites queenside is chronically weak so Black is not forced to seek quick ways of regaining the pawn. He could
calmly spend a tempo on prophylaxis like 11...h6!, intending 12...Nd7. Brynells move is threatening ...Bh6, but if
Whites king escapes from the pin on the d-file with 12.Kc1, Blacks best follow-up would be again 12...h6.

12.g3 Bh6 13.Ke1 Bxd2+ 14.Nxd2 Nxe5 15.Bg2 Rhe8 16.Nde4?!

16.b4 Ncd7 17.Rc1 was more stubborn. The knight swap unloads the e-file and the e8-rook also finds employment.

16...Nxe4 17.Bxe4 f5 18.Bc2 Nc4 19.b3 Na3 20.Rc1 Rd6?!

20...Bd5! 21.Rg1 Bc6 22.Kf1 b5 retained the initiative. Black discovered this idea one move later.


21.f3 would have wiped most of Blacks advantage. For instance, 21...Rc6 is parried with 22.Bd3 so Black would have
to worry about his knight on a3. Of course he could trade it for Whites knight, but then the play would be roughly
equal since Blacks kingside pawns limit its own bishop.

21...Bd5 22.Nxd5 Rxd5 23.Bd1 Red8 24.h5?? 0-1

White resigned before 24...Rxd1+.

Chapter 14. Rare Variations

Step by Step

In this chapter, I consider three minor variations:

A. 1.d4 d5 2.c4 dxc4 3.Nf3 Nf6 4.g3?; B. 1.d4 d5 2.c4 dxc4 3.Nc3?!; C. 1.d4 d5 2.c4 dxc4 3.Nf3 Nf6 4.Na3?

A. 1.d4 d5 2.c4 dxc4 3.Nf3 Nf6 4.g3?

The kings bishop fianchetto is extremely popular. White has been testing it against virtually all Blacks defences
against 1.d4, even against the Chebanenko and the Slav. But it simply does not work against the QGA. You do not have
to remember anything special since all the natural retorts are pleasant for Black:

a) 4...Nc6 5.Bg2 (5.Nc3 e5=; 5.Na3 e5; 5.Qa4 Nd5 6.Qxc4 Nb6 7.Qd3 e5 8.Nxe5 Nb4 9.Qd1 Qxd4)
5...e5 6.Qa4 exd4 7.Ne5 Qd6 8.Nxc6 Bd7 9.Qxc4 Bxc6 10.Bxc6+ Qxc6 11.Qxc6+ bxc6 12.0-0 0-0-0, Devereaux-
Berkes, Gibraltar 2009.

b) 4...c5 5.Qa4+ Nc6 6.dxc5

6.Bg2 cxd4
a) 7.0-0 Bd7 8.Qxc4 e5 9.e3 (9.Ng5 Qe7 10.b3 Na5 11.Qd3 Bc6 12.Ba3 Qd7 13.Bxf8 Kxf8) 9...Rc8.
b) 7.Nxd4 Qxd4 8.Bxc6+ Bd7 9.Bxd7+ Qxd7 10.Qxc4 Rc8 11.Qf4 Nd5.
6...e6 7.Bg2 (7.Qxc4 Qa5+) 7...Bxc5 8.0-0 0-0 9.Qxc4

9...Qd5 10.Nbd2 Rd8=, Miton-Ljubojevic, Linares 2013. Black solves the problem of his queens bishop by pushing
...e5: 11.b3 e5 12.Bb2 e4 13.Ng5 Qxg5 14.Bxf6 Qxf6 15.Nxe4 Bxf2+ 16.Nxf2 Be6.

The most principled approach, however, is to keep the extra pawn. When White undermines the queenside with b3, we
take on b3 and push ...b4:

4...b5 5.Bg2 Bb7 6.a4 (6.0-0 e6) 6...a6 7.0-0

After 7.b3, simplest is 7...b4 8.bxc4 c5. The same idea is effective after 7.Nc3 7...b4 8.Nb1 c5.


The main position for this line.


8.Nc3 b4 9.Nb1 c5;

8.Bg5 Nbd7 (8...Be7 is also possible) 9.b3 (9.Nc3 Qb8 or 9...b4) 9...cxb3 10.Qxb3 b4 11.a5 Be7.
8...cxb3 9.Qxb3

White is struggling for the draw after 9.axb5 axb5 10.Rxa8 Bxa8 11.Qxb3 b4 12.Bf4 Bd6.


Black has a sound extra pawn which, on top of all, hinders Whites harmonious development. The game may continue
10.a5 Be7 11.Bf4 (11.Bd2 Bd5 12.Qb2 c5 13.dxc5 Nc6) 11...0-0 12.Rc1 Bd5 13.Qb2 Ra7 14.Nbd2 Nbd7.

B. 1.d4 d5 2.c4 dxc4 3.Nc3?!

I decorate this move with a ?! sign because it practically deprives White of any opening advantage. For instance, if
Black answers 3...Nf6, Whites options would be:

a) 4.Nf3 transposes to Chapter 12.

b) 4.e3 e6 5.Bxc4 c5 6.Nf3 transposes to an innocuous line of the Classical System.

c) 4.e4 e5=.

d) 4.Bg5 c6 5.a4 h6.

Besides, Black has an interesting way to confront the opponent on virtually uncharted territory by:


It is all the more effective since White answers in the overwhelming majority of games:
a) 4.a4?! when Black is faced with a pleasant choice:
4...Nc6 5.Nf3 (5.e3 Na5 6.Nf3 Bg4) 5...Bg4 6.e3 Na5 7.h3 Bxf3 8.Qxf3 e6 keeps the extra pawn;
4...e5 counts on quick development and safety: 5.Nf3
Or 5.d5 Bb4 6.e4 Qh4 7.Qc2 Nf6 8.Bd2 Bxc3;
5.dxe5 Qxd1+ 6.Nxd1 Nc6 7.Nf3 Nb4 8.Ne3 Ne7 9.Bd2 Nf5 10.Rc1 Nxe3 11.Bxe3 Bd7;
5.e3 Nc6 6.Nf3 exd4 7.exd4 Na5 8.Ne5 Nf6 9.Nxc4 Bb4 10.Nxa5 Bxa5 11.Bc4 Qe7+ 12.Be3 Be6.
5...exd4 6.Qxd4 Qxd4 7.Nxd4 c5 8.Nf3 Nc6 9.Nd5 Bd6 10.e4 Nf6.

b) 4.e3 Nf6 5.Bxc4 b5 (5...e6 6.Nf3 b5 7.Bd3 Bb7) 6.Be2 Nbd7=.

4.Nf3 b5


5.e3 Bb7 6.Be2 e6 7.a4 c6 (or 7...b4 8.Nb1 c5 9.Bxc4 Nc6) 8.0-0 Nf6 offers Black a sound extra pawn.
5.e4 Bb7 6.a4 b4 7.Bxc4 e6 8.Qb3 Nc6!? is unexplored and it leads to dynamic positions with mutual chances 9.a5
Nxd4 10.Nxd4 Qxd4 11.Be3 Qd7 12.f3 Bd6 13.0-0-0!? (13.Na4 Qe7 14.e5?! Bxe5 15.Nc5 Bc8, Pedersen-Sadler
Cannes 1995) 13...Bc6 14.e5 bxc3 15.Qxc3 Ne7 16.exd6 cxd6 17.Qxg7 Rg8 18.Qxh7.

5...b4 6.Ne4 Nd7!

Played in Karpov-Portisch, Tilburg 1983. 6...e6 looks more natural, but then 7.Bg5! is awkward.
The rest is satisfactory for Black:
7.Ne5 Be7! (7...Bb7? 8.Ng5 f6 9.Nef7! Qe7 10.e4 fxg5 11.Nxh8 Nc6 12.Bxc4 g6 13.h4) 8.e3 Nf6 9.Nxf6+
Bxf6 10.Qf3 Qd5 11.Bxc4 Qxf3 12.Nxf3 Bb7=;
7.e3 Bb7 8.Ng3 c5=.
7...f6 8.Bd2 Nc6 9.e3 Na5 10.Rc1 Bb7 11.Nc5 Bxc5 12.dxc5.
8.Ng3 Qa5 9.e4 c3 10.Qb3
In this chaotic position, Whites pieces are much better coordinated:
10...Nc6 11.d5 Be7 12.Be3 or 10...cxb2 11.Qxb2 Be7 12.Bxe7 Nxe7 13.Nd2 Qb6 14.a5 Qd6 15.Nc4.


I have also analysed 7.Ng3 c5 8.e4 cxd4 9.Qxd4 e6 10.Bxc4 Bb7 11.Bf4 Ngf6 12.0-0 Bc5 13.Qd3 0-0=.

7...c3 8.bxc3 bxc3 9.Ne4 Ngf6 10.Nxc3 e6=

Black is even a tempo ahead now. The stem game finished in a draw after:

11.e3 Bb4 12.Bd2 c5 13.Be2 0-0 14.0-0 Bb7 15.Rb1 Rb8 16.Na2 Ba5 17.Qe1 Bc7 18.dxc5 Nxc5 19.Bb4 Bd6 20.Bxc5
Bxc5 21.Qc3 Qe7 22.Qa5 Nd5 23.Rb3 Bb6 24.Qd2 Rfd8 25.Qb2 Bc6 26.Nb4 Nxb4 27.Rxb4 Bxf3 28.Bxf3 Bd4
29.Rb7 Bxb2 draw.

C. 1.d4 d5 2.c4 dxc4 3.Nf3 Nf6 4.Na3?

Grivas spent a lot of ink in New in Chess YB to convince first players to try this, but I think that Tarrasch is right once
again about the knight at the edge. Black immediately takes over the initiative with:

4...e5 5.Nxe5

5.Nxc4 e4 6.Nfe5 b5.

5...Bxa3 6.bxa3

6.Qa4+ b5 7.Qxa3 Qxd4.

6...b5 7.a4 a6 8.e3 0-0!

Grivas considers only 8...Bb7 9.axb5 axb5 10.a4 Nbd7 11.Bb2, but omits here 11...0-0 with a possible continuation
12.axb5 Rxa1 13.Bxa1 Nxe5 14.dxe5 Qxd1+ 15.Kxd1 Ra8 16.Bd4 Ne4 17.Rg1 c3 18.f3 Ra1+ 19.Kc2 Nd2.
However, 9.Rb1!? opposes the b7-bishop and is unclear.

9.Be2 (9.axb5 axb5 10.a4 Nd5) 9...Nd5 (9...Bb7 10.0-0 Nbd7) 10.Qd2 (10.Qc2 c5;10.Bd2 c5) 10...Nd7.
Part 4
Alternative Repertoires

I called the variations of this part alternative, but in fact they are my main weapon. It is really difficult to beat a decent
player with Black in the Exchange Variation. For those of you who think the same way, Chapter 16 offers a dynamic
alternative, based on ...Bg4. You can enter it either via 3...a6 see Chapter 17, or after 3...Nf6 4.e3 Bg4.

3...a6 could also be used against higher rated players to neutrlalise their greater knowledge and skill with 4.e3 b5!?.

First players often prevent both options by the cunning move order 3.e3. It is a strong medicine, indeed, but I like to
play against an isolator. See Chapter 15.
Chapter 15. 3.e3
Main Ideas

1.d4 d5 2.c4 dxc4 3.e3 e5

This system might be totally indifferent to players who like the Classical System. They simply meet 3.e3 by 3...Nf6
4.Bxc4 e6. However, if you have the Alekhine Variation with ...Bg4 in your repertoire, as in my case, or you would like
to play bluntly for a draw with 3.Nf3 a6 4.e3 b5, then the move order with e3 becomes of paramount importance.
Even if you are completely comfortable with the Classical System, you may want to exploit Whites move order by
3...e5!?. Indeed, playing against an isolated pawn on d4 gives Black more chances to win than in the Exchange System
with 7.dxc5.
Nowadays there is a strong tendency among White players to head for the Classical System through 3.e3. Thus you
could learn well only two major variations, 3.e4 Nc6 and 3.e3 e5, to be fully prepared for 90% of your practical games.
This fashion is encouraged if not established by the major repertoire books as Kornevs A Practical White Repertoire
vol.1 and the older Avrukhs 1.d4 Volume 1.

Theoretical status

Both above-mentioned books recommend:

4.Bxc4 exd4 5.exd4 Bd6! 6.Nf3 Nf6 7.0-0 0-0 8.h3 Nc6 9.Nc3 h6 10.Qc2 (The only way of preventing ...Bf5!)
10...Nb4 11.Qb1 Be6! 12.Bxe6 fxe6 13.Re1
Kornevs only comment here is: White has easy play against Blacks weakness on e6: 13...Qd7 14.Bd2 Nbd5 15.Qd3
Rad8 16.Re2 Qf7 17.Rae1 Tkachiev-Sulava, Gonfreville 2006.
It is easy to improve this line the b4-knight should remain in place to deny the d3-square to the white queen. Correct is
14...Rae8! 15.Re2 Nc6 16.Qd3 e5=.

I have always wondered why does Avrukh prefer Whites position. His line goes 13...Qe8 14.Bd2 Nbd5 15.Re2 Qf7
16.Ne5 Qh5 17.Qd3 Qf5 18.g4 Qxd3 19.Nxd3

19...Rae8? 20.Rae1 Kf7 21.Nb5 a6 22.Nxd6 cxd6 23.Bf4 winning a pawn. Instead of blundering material, Black should
simply put the other rook on the e-file: 19...Rfe8 to stand at least equal after 20.Rae1 Kf7 21.Nb5 a6 22.Nxd6+ cxd6. I
was ready to test this against Golod in 2013, but our game finished in a draw after 13.Bd2 Qe8 14.Re1 Nbd5.
Obviously my opponent liked Blacks position, too, since a couple of weeks later he entered the same line against
Avrukh, this time with Black! He only slightly altered my set-up choosing ...Qd7 instead of ...Qe8 which does not
change the strategic canvas of the game. In short, both 13...Qd7 and 13...Qe8 are perfectly comfortable for Black and
most top-level games finish with quick draws. For more details, see game 32 Avrukh-Golod, Jerusalem 2013.

Problems and critical lines

Recently 7.Qe2+ Qe7 8.Qxe7+ Bxe7 9.0-0 is fashionable.

White counts on his lead in development. I suggest a nearly unexplored approach:

9...Nbd7!? 10.Re1 Nb6 11.Bb3 Nfd5 12.Nc3 c6

My idea is simple to trade light-squared bishops by ...Be6, ...Nc7. The most White can do is to transfer his queens
knight to c5 where we kill it at once. In many lines we could leave our king in the middle.

Another line which needed fixing in my opinion is when White develops his bishop allowing ...Bg4:
8.Bg5 h6 9.Bh4 Nc6! (9...Bg4 10.h3) 10.h3
We need energetic actions here: 10...g5! 11.Bg3 Bxg3 12.fxg3 g4! 13.hxg4 Bxg4 14.Nc3 Kg7! with mutual chances.

The 3.e3 e5 system leads to lively open play which should suit the QGA fans taste. White is running short on new ideas
and his only hope for a small advantage seems to be the endgame after 7.Qe2+ where my idea ...Nb8-d7-b6 needs
testing. It is indicative that the older lines bring White modest 52% in practical chess and 51% in email games. The
mere fact that many theoreticians base their repertoires on such a harmless move order suggests that the QGA is in
excellent shape!

Chapter 15. 3.e3

Step by Step

1.d4 d5 2.c4 dxc4 3.e3 e5


4.Nf3 exd4 5.Bxc4 allows 5...Bb4+ 6.Bd2 Bxd2+ 7.Qxd2 when the endgame after 7...dxe3 8.Bxf7+ Ke7 9.Qxd8+
Kxd8 10.fxe3 is level:
10...Nh6 11.Bb3 Re8 12.0-0 c6 13.Nd4 Bd7 14.Nc3 Na6 15.Rac1 Nc5 16.Bc2 g6=;
10...Nf6 11.Nc3 c6 12.e4 Ke7 (12...Rf8 13.Bb3 Nbd7 14.0-0 Re8 15.Rae1 Ne5=) 13.Bb3 Nbd7 14.0-0 Nc5 (14...Ng4
15.e5 Nc5=) 15.Bc2 Be6 16.Nd4 Rad8 Campora-Moreno Ruiz, Madrid 2007.

5.Nxd4?! c5 6.Nb5 a6 7.Qxd8+ Kxd8 can only be in Blacks favour.

4....exd4 5.exd4 Bd6!

5...Nf6 is less accurate due to 6.Qb3 Qe7+ 7.Ne2 Qb4+ 8.Nbc3 Qxb3 9.axb3! and the endgame is pleasant for White.

6.Nf3 Nf6 7.0-0

7.Qe2+ has been topical lately. White aims for a slightly better riskfree endgame. Black cannot avoid it since:
7...Be7?! 8.0-0 0-0 9.Nc3 is unpleasant. For instance: 9...Bg4 10.h3 Bxf3 11.Qxf3 Nc6 12.Rd1 Nxd4 13.Qxb7 c5
14.Bf4 Bd6 15.Bxd6 Qxd6 16.Nb5 Qf4 17.Nxd4 cxd4 18.Rd3.
7...Qe7 8.Qxe7+ Bxe7 9.0-0
Blacks task is not trivial. After 9...0-0 10.Re1, he must make yet another move with the bishop to d6 and his lag
in development may become embarrassing.
I have analysed with my students 9...Nc6 and it turned well in game 29 Cvitan-D.Marjanovic, Zadar 2013.
However, later we found improvements in Whites play so things are not too clear. Im going to recommend
another way which is straightforward and easier to play:
9...Nbd7!? 10.Re1 Nb6 (I tested this line against Kharitonov in Paleochora 2015, but we signed here a draw.) 11.Bb3
Nfd5 12.Nc3 c6

Black has delayed castling in favour of development. If he succeeded in connecting rooks, he would be fine, for
13.h3 Be6 14.Ne4 (14.Bg5 Bd6) 14...0-0 15.Nc5 Bxc5 16.dxc5 Nd7 17.Nd4 Nxc5.
13.Bg5 provokes 13...f6 which is a useful move anyway 14.Nxd5 (14.Bd2 Kf7 15.a4 a5 16.Ne4 Be6 17.Nc5
Bxc5 18.dxc5 Nd7 19.Rac1 Nc7 20.Nd4 Bxb3 21.Nxb3 Ne6=) 14...Nxd5 15.Bd2 Kf7 16.Bc3 Re8 17.Nd2 Be6
18.Nc4 Rad8 and Black was on top in Djoric-Stojanovic, Serbia 2005. Whites most challenging idea is to fix a
target on a5:
13.a4! a5 14.Bd2 Be6 15.Ne4

It is still early for 15...0-0 due to 16.Nc5 Bxc5 17.dxc5 Nd7 18.Rac1 Nc7 19.Nd4 Bxb3 20.Nxb3 Ne6 when
White can provoke another weakness by 21.f4 g6 22.f5! gxf5 23.Rf1.
The text discourages 16.Nc5 in view of 16...Bxb3 17.Nxb3 Nc4! 18.Bc3 Nd5= so Whites only chance to force
play is:
16.Bf4 Bxb3 17.Bxc7 Nc8

A critical position. I have analysed from here:

a) 18.Re3 Bd5 19.Nc3 Bxf3 20.Rxf3 0-0;

b) 18.Ra3 Bd5 19.Rae3 Bxe4 20.Rxe4 Kd7 21.Bf4 f6=;

White also has two ways of winning a pawn when Black gets excellent piece play:
c) 18.Nc5 Bd5 19.Nxb7 f6 or 19...Bxf3 20.gxf3 Ra7 21.Nd6+ Kd7;

d) 18.Bd6 Nxd6 19.Nxd6+ Kf8 20.Nxb7 f6 21.Nc5 Bd5

White has two weak pawns on b2 and d4. Black can put his rooks on b8 and d8 or activate his bishop via e7-d8-b6.

9...Nbd7!? is practically unexplored, but it gives Black better chances than the common 9...0-0.


White now chooses either the active approach with A. 8. Bg5, or B. 8.h3

8.Nc3 Nc6 commonly transposes to the main lines.

A. 8.Bg5 h6 9.Bh4 Nc6! (9...Bg4 10.h3) 10.h3

10.Nc3 Bg4 solves the opening problems at once: 11.h3

11.Re1 a6 12.Bd5 cannot be serious. The game Voiculescu-Galanov, email 2011, went on 12...Bxf3 13.Qxf3
Nxd4 14.Qd3 c5 15.Rad1 Ra7 16.a4 Bb8 17.Bc4 Qd6 18.Bg3 Qd8 19.b4 Bxg3 20.hxg3 b6 21.bxc5 bxc5
22.Bxa6 Qb6 draw.
11...Bxf3 12.Qxf3 Nxd4 13.Qxb7 Rb8 14.Qxa7 Ra8 15.Qb7 Rb8=.


We could insert 10...Na5 11.Be2 first, but I think that rejecting the white bishop to e2 is not worth the tempo. White
retains an initiative following 11...g5 12.Bg3 Bxg3 13.fxg3 Qd6 14.Kh2! (Practice has only seen 14.g4?! Nd5)
14...g4 (14...Nc6 15.Nc3 Kg7 16.Qd2) 15.Ne5.

11.Bg3 Bxg3 12.fxg3


This move occurred in Babula-Schlosser, Pardubice 2007, but it turned into a rout after 13.hxg4 Bxg4 14.Nc3 Re8?
15.Qd3 Kg7 16.Rad1 Bh5 17.Nh4 Bxd1 18.Nf5+ Kh8 19.Qxd1 Ne4 20.Bxf7 Nxc3 21.bxc3 Rf8 22.Nxh6 Qg5 23.Rf5
Qe3+ 24.Kh2 Qxh6+ 25.Rh5 1-0.
Alternatives are:

a) 12...Qd6. Now 13.g4 Be6 14.Bxe6! is slightly better for White, but 13...a6, intending ...b5, is more complex.
However, Blacks king is rather weak and that is a solid ground for a pawn sacrifice: 13.Nc3! Qxg3 14.Nd2 when
14...Bxh3 15.Qf3 wins a piece although for 3 pawns.

b) 12...Nh5 13.g4 Nf4 looks solid, but that is only on the surface. White will trade the proud knight by Nc3-e2 and the
black king will remain a bit shaky.

13.hxg4 Bxg4 14.Nc3 Kg7!

This improvement assures Black of a good game. The d4-pawn is hanging so 15.d5 Bxf3 16.Qxf3 Ne5 17.Qf4 looks
Blacks knights dominate the dark squares. He can keep the tension with 17...Re8 or offer a draw endgame with
17...Qd6, e.g. 18.Nb5 Qb6+ 19.Kh1 Ng6= 20.Qxc7 Qxc7 21.Nxc7 Rac8 22.d6 Rcd8 23.Rad1 Ne4 24.Rxf7+ Rxf7
25.Ne6+ Kh7 26.Nxd8 Rd7 27.Rd4 Nxg3+ 28.Kh2 Nf5 29.Be6 Rxd6 30.Rd5 Rxd5 31.Bxd5 b6 32.Nc6 a5 33.Be4

B. 8.h3 Nc6 9.Nc3 h6

Both sides have fulfilled their compulsory program and now the c8-bishops fate will decide the outcome of the opening
battle. If it reaches f5, the game will be fully balanced. Black aims for the set-up ...Bf5, ...Qd7, ...Rad8. For instance:

a) 10.a3 Bf5 11.Re1 Re8

Black can delay this move in favour of ...Qd7!? see game 31 Timman-Morozevich, Moscow 1994.
Or 12.Be3 a6 13.Rc1 Qd7 14.Bf1 Rad8 15.b4 Qc8.
12...Qxe8 13.Nb5 a6 14.Nxd6 cxd6. Whites bishop pair is ineffective because Black has an excellent outpost for his
pieces on d5 15.Be3 Rc8 16.Rc1 Qd7 17.Bb3 Be4 18.Nd2 Bd5 19.Bc2 Re8 20.Bd3 Ne7 draw, Ozmen-Dos Santos,
ICCF email 2008.

b) 10.Re1 Bf5! 11.d5 Ne7 12.Nd4

The idea of this retreat is to hinder Ndb5 which gave White a space advantage in Kir.Georgiev-Kogan, Barcelona 2008,
after 12...Bh7 13.Ndb5 a6 14.Nxd6 Qxd6 15.Qf3. From d7 the bishop controls b5 so 13.Ndb5 would drop a pawn to
13...Bxb5 14.Nxb5 Nexd5. Black will play ...a6 to any other move, consolidating, e.g. 13.Be3 a6 14.Qf3 Ng6.

The only way to prevent ...Bf5 is:

10.Qc2 Nb4 11.Qb1 Be6!

Straightforward and good. It is better to neutralise the threat Bxh6 than provoke it with 11...c6?! 12.Bxh6! gxh6
13.Qg6+ Kh8 14.Qxh6+ Nh7
This position has been tested in several games. I have also analysed it in detail and my verdict is that it is defendable by
a computer, but from a practical standpoint it is difficult for Black. See game 30 Iotov-Ruefenacht, ICCF email 2009.

12.Bxe6 fxe6 13.Re1

Black has an easy game. His only weakness on e6 can be defended with ...Re8, ...Qd7 (or f7). The point is that a knight
on d5 will often dominate Whites dark-squared bishop. Now both 13...Qd7 and 13...Qe8 are good retorts. For more
details, see game 32 Avrukh-Golod, Jerusalem 2013.
Chapter 15. 3.e3
Annotated Games

29. Cvitan 2558 Dej.Marjanovic 2269

Zadar 16.12.2013

1.d4 d5 2.c4 dxc4 3.e3 e5 4.Bxc4 exd4 5.exd4 Nf6 6.Nf3 Bd6 7.Qe2+ Qe7 8.Qxe7+ Bxe7 9.0-0 Nc6

Attacking the d4-pawn. This idea is most effective against mundane continuations like:
10.Re1 Bg4! 11.Ne5
Or 11.Bb5 Bxf3 12.gxf3 a6 (12...0-0-0 13.Bxc6 Bb4 14.Bxb7+ Kxb7 15.Rd1 c5=) 13.Bxc6+ bxc6;
11.Nbd2 0-0-0.
11...Nxe5 12.dxe5 Nd7 13.Nc3 c6 14.Bf4 0-0 15.Ne4 Rad8 16.Bg3 Nc5 17.Nd6 Be6 18.Bxe6 Nxe6 19.Nxb7
Rb8 is also interesting.
12.Nxg4 loses material to 12...Nxg4 13.Re4 Nc2 14.Rxg4 0-0.
However, White has more testing continuations:


Of course White is much better developed after 10.d5 Nb4 11.Nc3 c6 12.dxc6 Nxc6 13.Re1 0-0, but I has not found
anything alarming in this position.


The point of Blacks idea is that 11.Nb5 c6 12.Nc7+ Kd8 13.Nxa8 Nxc4 is in his favour.

In my opinion, 11.Bb5+, taking away the retreat square c6, is more unpleasant. Then 11...c6 12.Bd3 Be6 13.Re1 0-0-0
14.Ng5 Rxd4 15.Bc2 Bb4 16.Nxe6 fxe6 17.Rxe6 Nc4 is playable although Whites bishop pair gives him a slight

11...Be6 12.Nb5

Or 12.Re1 0-0-0 13.Ng5 Nc6 14.Nxe6 fxe6 15.Rxe6 Bb4 16.Bg5 Nxd4.


The play is balanced. Black is ready to castle queenside and he does not fear 13.Be4 due to 13...a6 14.Bxd5 axb5
15.Bxe6 fxe6 16.Bf4 Kd7 17.Ne5+ Kc8=.

13.Bd2 Nc6 14.Rfe1 0-0-0 15.a3 a6 16.Nc3 Bf6 17.Be3 Nb6 18.Ne2 Kb8 19.h3 Bc4 20.Red1 Rhe8 21.Rac1 Bd5

30. Iotov Ruefenacht

ICCF email 2009

1.d4 d5 2.c4 dxc4 3.e3 e5 4.Bxc4 exd4 5.exd4 Nf6 6.Nf3 Bd6 7.0-0 0-0 8.h3 Nc6 9.Nc3 h6 10.Qc2 Nb4 11.Qb1 c6
12.Bxh6 gxh6 13.Qg6+ Kh8 14.Qxh6+ Nh7

White has full compensation. The forced line goes:

15.Ne4 Be7 16.Ne5 Qxd4 17.Nxf7+ Rxf7 18.Bxf7 Qxe4 19.Rae1 Qh4 20.Qxh4 Bxh4 21.Re4 Bf6 22.Rxb4

Objectively, this very sharp endgame should be drawn. However, Whites play is crystal clear as he is just pushing his
pawns forward while Black must constantly generate threats to hinder the opponents plan. The cost of his mistakes is
much higher. Ill cite for the record the email game Pierzak-Broniek, ICCF email 2011: 22...Kg7 23.Bh5 a5 24.Re4 Bf5
25.Re3 Kh6 26.Be2 Bd4 27.Re7 b6 28.g4 Bc5 29.Re5 Bg6 30.h4 Bf7 31.g5+ 1-0. The game text is a minor
improvement as 22...a5 23.Re4 Bf5 24.Re3 could be met by 24...Bxb2.

22...a5 23.Rf4 Kg7 24.Bb3 Ng5 25.h4 Ne6

Perhaps 25...Be5! 26.Rc4 Nf7 27.Re4 Bf6 is more accurate.

26.Rg4+ Kf7 27.a4 b5 28.Rc1

Black is clearly worse. Only the limited material can save him. Over the board, he will often lose this position.

28...bxa4 29.Rxa4 Rb8 30.Rxc6 Rxb3 31.Rxc8 Rxb2 32.Ra8 (32.Rxa5!) 32...Be5 33.g3 Bd4 34.Rxd4 Nxd4 35.Rxa5
Nf3+ 36.Kg2 Ne1+ 37.Kf1 Nd3 38.Rf5+
This is a draw!

38...Ke6 39.Rf8 Ke7 40.Rf3 Ne5 41.Re3 Kf6 42.Re4 Nd3 43.Re2 Kf5 44.Rxb2 Nxb2 45.f3 Nc4 46.Ke2 Ne5 47.h5
Nf7 48.Ke3 Kg5 49.g4 Nd8 50.Ke4 Draw.

31. Timman 2635 Morozevich 2575

Moscow 1994

1.d4 d5 2.c4 Nc6 3.Nc3 dxc4 4.Nf3 Nf6 5.e3 e5 6.Bxc4 exd4 7.exd4 Bd6 8.0-0 0-0 9.h3 h6 10.Re1 Bf5 11.Be3 a6
12.a3 Qd7

The move order is not very important. Black can play first 12...Re8 since 13.Nh4 Bh7 14.Qf3 Qd7 15.g4 only imitates
activity without threatening g4-g5 due to the hanging state of the d4-pawn. In Short-Bareev, Pula 1997, Black
underlined this weakness with 15...Rad8 16.Red1 Bf8=.


Otherwise White risks to remain with a weak isolated pawn, for example: 13.b4 Rfe8 14.Rc1 Rad8 15.Bf1 Ne4 16.Nxe4
(16.Na4 Bf8 17.Nc5 Nxc5=) 16...Bxe4.

13...Ne5 14.Bf1 Rfe8 15.Bd4 Nxf3+ 16.Qxf3 Rxe1 17.Rxe1 Re8 18.Re3 Rxe3 19.fxe3
White probably hoped here for the timid 19...Be7? 20.e4 Bg6 21.e5 with a huge spatial advantage, but Morozevich has a
surprise in store:

19...Bg6! 20.Bxf6 gxf6 21.Qxf6 Qe8 22.Qd4 Qe7 23.Nd1 Draw.

Blacks peaceful decision is a little strange since he does not risk anything in the final position and he can provoke
another weakness in Whites camp by 23...Bh5! 24.g4 Bg6, but in Olympiads many external factors can influence the
outcome of the games. It is generally thought that a draw with Black benefits the team as a whole.

32. Avrukh 2576 Golod 2578

Jerusalem 2013

1.d4 d5 2.c4 dxc4 3.e3 e5 4.Bxc4 exd4 5.exd4 Nf6 6.Nf3 Bd6 7.0-0 0-0 8.Nc3 Nc6 9.h3 h6 10.Qc2 Nb4 11.Qb1 Be6
12.Bxe6 fxe6

The speculative sacrifice 13.Bxh6 is not enough even for a draw13...gxh6 14.Qg6+ Kh8 15.Qxh6+ Nh7 16.Ne4 Bf4
17.Qxe6 Qe8.
13.Bd2 should transpose to 13.Re1. Actually, Gleizerov-Varga, Predeal 2006, saw 13...Qd7 14.Ne4, but the game
finished here in a draw.


In my opinion, 13...Qe8!? is equally good. It may lead to the same position as 13...Qd7 if Black follows up with ...Qf7.
An additional option could be ...Qe8-h5. White has tested:

a) 14.Bd2 Nbd5. Black should aim to remain with a d5-knight against the d2-bishop.

Avrukhs novelty. 15.Qd3 Qf7 16.Re2 (16.Ne5 Bxe5 17.dxe5 Nb4 18.Qc4 Nfd5) 16...Nh5 was even in
Tkachiev-Golubovich, Pula 2000.
15.Ne4 Nxe4 16.Qxe4 Rf6 allows a more active set-up with the rook before the queen.
15...Qf7 16.Ne5
Or 16.Qd3 Rad8 17.Rae1 Rfe8 18.Qb5 c6.
16...Qh5 17.Qd3 Qf5 18.g4 Qxd3 19.Nxd3

Of course not 19...Rae8? since we might need a rook on d8 for defending d6.
20.Rae1 Kf7 21.Nb5 a6 22.Nxd6+ cxd6 23.Bf4 Rad8=, intending ...Nc7.

b) 14.Nh4 Qh5 15.Ng6 Rfe8 16.Bf4 White wants to activate his prospectless bishop, but the simple 16...Rad8 17.Be5
Nc6 gives counterplay: 18.Bxd6 cxd6 19.Nf4 Qf7 20.Qd3 e5.

c) 14.Ne4 Nbd5 15.Nc5

We gladly eat the knight here:
15...Bxc5 16.dxc5 Nd7 17.Qc2 c6 18.Be3 Rxf3! 19.gxf3 Qh5 and Black took over the initiative in Ramirez Alvarez-
Morozevich, Bled 2002.

14.Bd2 Rae8

The b4-knight should remain in place to deny the d3-square to the white queen. 14...Nbd5 15.Qd3 Rad8 16.Re2 Qf7
17.Rae1 was slightly better for White in Tkachiev-Sulava, Gonfreville 2006.


This trade of knights does not leave any hope for an ad vantage. Avrukh recommends against 13...Qe8 to keep the
tension by Re2. Lets consider it here:
15.Re2 with two decent options:
a) 15...Nc6 16.Qd3 e5 17.dxe5 Nxe5 18.Nxe5 Rxe5=

b) 15...Qf7 16.Ne5 Bxe5 17.dxe5 Nfd5 18.a3 Nc6 19.Qe4 Rd8 20.Nxd5 Rxd5 with slightly better chances for Black,
Sebenik-Stevic, Rijeka 2010. As a rule, the knight is better than the dark-squared bishop.

15...Nxe4 16.Rxe4 Nd5 17.Qd3

Both sides lack a plan for improvement since any pawn advances would only weaken their positions. Jankovic-Stevic,
Otocac 2010, finished in a draw after 17...Rf6 18.Rae1 Qf7 19.Qe2 a6 20.Ne5 Bxe5 21.Rxe5 c6 22.f3. Golod does not
object to a draw, either:

17...Rf5 18.Rae1 Nf6 19.R4e2 Nd5 20.Re4 Nf6 21.R4e2 Nd5 22.Re4 Nf6 Draw.

Chapter 16. The 4...Bg4 Variation

Main Ideas

1.d4 d5 2.c4 dxc4 3.Nf3 Nf6 4.e3 Bg4 5.Bxc4 e6

This variation has remained for a long time in the backyard of the QGA theory despite that Spassky and Smyslov played
it regularly. In fact it scores exactly the same percentage as the Classical System with 4...e6. Furthermore, recent
developments have revived several important lines and currently Black does not have any serious theoretical problems.
I have analysed and enriched this variation together with my pupils and now Im completely confident in it. I especially
like it for playing for a win.


The bishop is more active on g4 and it has a strong impact on Whites centre, namely on the d4-pawn.
This variation is considerably less explored than the Classical System. There are only a few top-level games so
White players do not have reliable examples to follow. Thus they have to rely on their own judgement in home
preparation. On top of all, the play is not forced and both sides possess plenty of logical moves and various move
Learning overhead is modest you have to know plans and not long forced variations.
The play is more dynamic and we avoid endgames with symmetrical pawn structures


We have to be very careful with the move orders. Some of the main lines assume pawn sacrifices, although we get
excellent compensation.

To sum up, this variation should attract players who would like to avoid deep theory and early simplification and who
like to take strategic risks. For me, it is a perfect weapon for opens although I choose it even against 2600+ opponents.
Our goal is to achieve the following set-up:

6.Nc3 Nc6 7.0-0 Bd6 8.h3 Bh5 9.Be2 0-0 10.b3 a6

We can build it against any move order while White may vary with an extended fiachetto a3+b4 or with a knight on d2
instead of c3. Ill discuss later the best timing for h2-h3. In the diagram position, White has two more easy moves:

11.Bb2 Qe7 12.Rc1 Rad8 and it is unclear how he could stop the break ...e6-e5.
For instance: 13.Nd2 Bg6 14.Nc4 e5 15.Nxd6 exd4!

Let us ponder how White could deviate from the above scenario:

1. e3-e4? at any moment drops a pawn to ...Bxf3, followed up by ...Nxd4. This is the point of putting the knight on c6
instead of d7!

2. White can chase our bishop with 6.h3 Bh5 7.g4 Bg6 8.Ne5 Be4! 9.f3 Bd5 (9...Bxb1 10.Rxb1 Nfd7 is also unclear)
10.Qa4+ Nfd7 11.Bxd5 exd5 12.Qb5 Bd6 13.Qxb7 Nb6.

3. White pins our knight with 7.Bb5 Bd6

8.Qa4 Bxf3 9.gxf3 0-0 10.Bxc6 bxc6 11.Qxc6 Rb8 gave me a nice compensation in game 34 Edouard-Delchev,
Linares 2013.
8...Nd7! 9.Be3 0-0 10.h3 Bh5 11.0-0 Nb6 intending to strike at the centre with ...f7-f5 to almost any move:
12.Be2 f5

13.exf5 exf5 14.Ng5 Bxe2 15.Nxe2 Qf6!;

12.Re1 f5;
12.Rc1 f5. That gives Black counterplay.

4. White wins the b7-pawn: 6.h3 Bh5 (this is considered an improved version of 6.Qb3) 7.Qb3 Bxf3 8.gxf3 c5! 9.Qxb7

Of course, this is the acid test of the 4...Bg4 variation, but I could not dream for more with Black we have an easy
game due to Whites permanently crippled kingside pawns. It is well known that the pair Q+N is very effective against
a vulnerable king so we should not be afraid of trading out bishop for the c3-knight. For instance:
10.Nc3 cxd4 11.exd4 Bd6! 12.Qa6 Bb4 13.a3 Bxc3+ 14.bxc3 0-0 15.0-0 Nb6 16.Be2 Nfd5.
As a rule, in this line we play in the centre and along the c+d files, but keeping an eye on the enemy king.

5. White hinders ...e6-e5 by the manoeuvre Nb1-d2-c4.

We have a choice against this approach.
In general, Black can meet any passive Whites moves (like Bc4-e2, Nb1-d2 or 6.0-0 a6 7.a3) by ...c5, for instance:
6.0-0 a6 7.a3 c5 8.dxc5 Qxd1 9.Rxd1 Bxc5

Obviously, we have nothing to fear from this position except for the fact that it is a far cry from our initial idea of
getting complex play with an asymmetrical pawn structure.

I believe that Black can stick to his universal set-up with:

6.Nbd2 a6 7.0-0 Nc6 8.b3 Bd6 9.Bb2 0-0 10.h3 Bh5 11.Be2 Qe7 12.Nc4

Here we cannot achieve ...e5, but we can shift the focus on the kingside instead with 12...Nd5 13.Rc1 f5!?, intending

Move Order Tips

Black often enters the Bg4 Variation via the move order 1.d4 d5 2.c4 dxc4 3.Nf3 a6 4.e3 Bg4 5.Bxc4 e6

The idea is to sidestep the sharpest lines with Qb3 or Bb5. However, I would love to get an attack with Black, even at
the cost of a pawn. Besides, in the diagram position, 6.Qb3 is still possible. Although White does not win a pawn after
6...Bxf3 7.gxf3 b5 8.Be2 Nd7 9.a4 b4, his chances seem somewhat better.

The main question after ...Bg4 is when exactly White should play h2-h3?
The short answer is: the sooner the better!

If White lingers too long with it, Black could reconsider and exchange on f3 or retreat to f5. This is especially valid in
the symmetrical endgame when Black had played ...c5.
Another argument in favour of h3 is that it indirectly hinders ...e5:

12...e5? loses a pawn here to 13.Nxe5.

If we elaborated a little deeper, I should mention that playing h3 on move 6, after 1.d4 d5 2.c4 dxc4 3.Nf3 Nf6 4.e3
Bg4 5.Bxc4 e6, would enable another set-up for Black 6.h3 Bh5 7.Nc3 Nbd7 8.0-0 Bb4

This would have been dubious stayed the bishop on g4 due to 7.Qb3 c5 8.Ne5 (with a tempo!). Still, the set-up with
...Nbd7, is more passive than ...Nc6, and I do not see a reason to prefer it even if it were possible see game 35
Prohaszka-Delchev, 13th EICC 2012.
Perhaps Whites most accurate sequence is 6.Nc3 Nc6! 7.h3.
Chapter 16. The 4...Bg4 Variation
Step by Step

1.d4 d5 2.c4 dxc4 3.Nf3 Nf6 4.e3 Bg4 5.Bxc4 e6

White chooses here from 5 main continuations and he has also tried 7(!) other rare moves.
A. 6.Qb3; B. 6.h3 I consider here some variations which are specific to an early h3; C. 6.Nbd2; D. 6.0-0; E. 6.Nc3.

A. 6.Qb3 Bxf3 7.gxf3 c5! 8.Qxb7 Nbd7

Black has full compensation for the pawn, based on the structural flaws of Whites pawn formation.


Alternatives are:
9.Nc3 cxd4 10.exd4 Bd6 11.Ne4 (11.f4 0-0 12.Bb3 Qa5) 11...Nxe4 12.fxe4 Rb8 13.Qa6 Qh4 with initiative;

9.f4 cxd4 10.0-0 The most recent game, Cossin-Edouard, Montpellier 2015, went: 10...Bc5 11.Qf3 dxe3 12.fxe3 0-0
13.Nc3 e5 14.Kh1 Qe7 with complex play. It was stronger to insert 10...Rc8! 11.Bb5 Bc5.

9...Bxc5 10.f4 0-0

10...Nd5 is also possible 11.Bxd5 Rb8 12.Qc6 (12.Qa6!?) 12...Rc8 13.Qa4 exd5 14.Nc3 0-0 15.Nxd5 Nb6 16.Nxb6
Qxb6 17.0-0 Qg6+ 18.Kh1 Rfd8.


The famous game Foguelman-Bronstein, Amsterdam 1964, went 11.0-0 Nd5 (cutting off the queen) 12.Rd1?! (the only
way to return the queen was 12.Qb3 intending 12...Rb8 13.Qd3. Black can develop an attack with 12...Qh4 13.Bxd5
Rab8 14.Qd1 exd5, followed up by ...Rb8-b6-h6, ...Nd7-f6.) 12...Rb8 13.Qc6 Qh4 14.Nc3
Here, instead of 14...Rb6 15.Qxd7 Nxf4! which also led to a beautiful victory, Black was winning directly with
14...Ne5!! 15.fxe5 (15.Qxc5 Nxc3 16.bxc3 Qg4+) 15...Nxc3 16.Re1 (16.bxc3 Qg4+) 16...Ne4+.

We see that it is safer to delay castling.

11...Nb6 12.Bb3 Nbd5 13.Nxd5 Nxd5 14.Bd2 Rb8 15.Qa6

Now 15...Nxe3 leads to perpetual check while 15...Qf6!? keeps the tension.

B. 6.h3 Bh5 7.Qb3

Alternatives are:
a) 7.g4 Bg6 8.Ne5 is dubious due to 8...Be4! 9.f3 Bd5 (9...Bxb1 10.Rxb1 Nfd7 is also unclear) and White cannot
exploit the manoeuvre of Blacks bishop:

10.Qa4+ Nfd7 11.Bxd5 exd5 12.Qb5 Bd6 13.Qxb7 Nb6;

10.Bb5+ c6 11.Be2 c5 12.e4 Bc6 13.d5 exd5 14.g5 Nfd7 15.Nxf7 Kxf7 16.exd5 Ne5 17.dxc6 Qxd1+ 18.Bxd1 Nbxc6;
10.Be2 c5 11.Nc3 cxd4 11...Nc6!? 12.Nxd5 Qxd5 13.e4 Qxd4 14.Nxc6 bxc6 15.Qb3 12.Qxd4 Nc6 13.Bb5 Qc7
14.Nxd5 (14.e4? 0-0-0) 14...Nxd5 15.Bd2, Landa-Delchev, Spain 2015, 15...Be7!.

b) 7.Be2 is well met by 7...c5 8.0-0 Nc6, although:

7...Nc6 is also possible 8.Nbd2 Bd6 9.Nc4 0-0

I discuss a similar set-up in line C. The difference is that Black has saved ...a6.
10...b5! 11.Nxd6 (11.Ncd2 a6 12.0-0 Qe8! 13.b4 e5) 11...cxd6 12.0-0 Qb6=.
c) 7.Nc3 Nc6! transposes to our main plan in line E.
7...Nbd7 8.0-0 Bb4 is possible, but Black remains somewhat passive. I consider it in more detail in game 35
Prohaszka-Delchev, 13th EICC 2012.

d) 7.0-0 a6 8.Nc3 Nc6 should transpose to line D1.

The old Soviet books claimed that 6.h3 Bh5 7.Qb3 was the refutation of 4...Bg4. Practical experience also seems to
favour White. However, modern engines prove that Black has sufficient compensation for the pawn after:

7...Bxf3 8.gxf3 c5 9.Qxb7 Nbd7

I consider this sacrifice in line A. The insertion of h3 eliminates some variations based on perpetual check since the
pawn controls g4, but the long-term compensation due to Whites compromised pawn structure remains. I would even
say that Whites castling position is even weaker:

B1. 10.Nc3; B2. 10.dxc5

B1. 10.Nc3 cxd4 11.exd4 Bd6! 12.Qa6

White has also tried a number of other moves:

12.Bd2 0-0 13.Nb5 Nb6 14.Bb3 a6 15.Nxd6 Qxd6;
12.f4 0-0 13.Qf3 Qa5 14.Bb3 Rab8 15.Rb1 Qf5 16.0-0 Nh5;
12.Rg1 0-0 13.Bd3 (13.Bh6 Nh5) 13...Kh8;
12.Ne4 Nxe4 13.fxe4 (13.Qxe4 Rc8 14.Bd3 Qb6 15.0-0 Nf6 16.Qh4 Qb7) 13...0-0.
12.Nb5 Bb4+ 13.Kf1 0-0 14.a3 Be7.

12...Bb4 13.a3 Bxc3+ 14.bxc3 0-0 15.0-0 Nb6

15...Nd5!? 16.Bd2 N7b6 17.Ba2 Qh4 is also promising.

16.Be2 Nfd5 17.Qd3 Rc8 18.Bd2 Qh4 19.f4 Nc4 20.Bc1 Na5, Yuan-Alekseenko, Moscow 2012.

B2. 10.dxc5 Bxc5 11.f4 0-0 12.0-0 Nd5 13.Qb3


Black should play all over the board. He must aim to keep both central files open for his rooks so it is better to take on
d5 by piece rather than by the e6-pawn. 13...Qh4 14.Bxd5 is not too clear.

14.Be2 e5! 15.fxe5 Qg5+ 16.Kh1 Qf5 17.Bg4 Qxe5

Black has a strong initiative.

C. 6.Nbd2 a6

White was threatening 7.Qa4+ or 7.Qb3.

7.0-0 Nc6

The standard equaliser against moves which do not attack directly the centre as Nc3 is 7...c5, for instance: 8.dxc5 Bxc5
9.a3 Nc6. However, I play ...Bg4 in order to obtain more complex play than in the Classical System. Thus I do not
mind taking some risks.

White can also regroup his minor pieces with 8.Be2 Bd6 9.Nc4 0-0 10.h3 Bh5 11.b3 when 11...b5! 12.Nxd6 cxd6
13.a4 Qb6 would be equal.

8...Bd6 9.Bb2 0-0 10.h3 Bh5 11.Be2 Qe7


12.Ne5 Bxe2 13.Qxe2 Bxe5 14.dxe5 Nd7 15.f4 Rfd8 16.Nc4 a5 17.Ba3 occurred in Levin-Alekseenko, St Petersburg
2015, when 17...Nb4 would have been equal.

To 12.Rc1, I suggest to test 12...Nd5 intending ...f5. If 13.Ne5 Bxe2 14.Qxe2 Nxe5 15.dxe5 Ba3=.

12.e4 is possible, but it is unclear what to do after 12...Bf4!?.

12...Nd5 13.Rc1 f5!?

Later we can transfer the queens knight via the route ...Nc6-d8-f7-g5.

D. 6.0-0 a6

This is more flexible than 6...Nc6 7.h3 Bh5 8.Nbd2 Bd6 9.Be2 intending Nd2-c4 as in line C.


7.Nc3 Nc6 8.Be2 Bd6 9.b3 0-0 10.Bb2 Qe7 11.h3 transposes to line E.

If White attempts to trick us with peculiar move orders, we can attack the centre with ...c5, for instance:
7.Nbd2 c5 8.Qa4+ Nbd7 or:
7.Be2 c5 8.h3 Bh5 9.dxc5 Qxd1 10.Rxd1 Nc6. Then Portisch-Tal, Wijk aan Zee 1968, finished in a draw after 11.b3
Bxc5 12.Bb2 0-0 13.Nc3 Rfd8 14.Ne1.
7.h3 Bh5 8.b3 is game 36 Korchmar-Delchev, Paleochora 2015


The set-up with 7...Bd6 8.Nc3 Nc6 is also effective 9.b4 0-0 10.Bb2 e5 11.d5 Ne7.

7...c5 is a decent way to equalise. 8.dxc5 Qxd1 9.Rxd1 Bxc5

If now White rushes with 10.b4, Black will attack it with ...a5, e.g. 10...Be7 11.Nbd2 (11.Bb2 0-0 12.Nbd2 a5) 12...a5.
And in the event of 10.Nbd2 0-0 11.Be2 Be7 12.Nc4 Nc6, he does not have any problems.


8.Nbd2 should also be met by 8...c5 since 8...Bd6 9.h3 Bh5 10.e4 e5 (10...Bf4 11.Bxd5 exd5 12.Qb3 Nb6 13.Re1 0-0
14.Nf1 offers White the better pawn structure.) 11.dxe5 Nxe5 12.g4 Bg6 13.Nxe5 Bxe5 14.Nf3 Qe7 15.Qe2 h5
16.Nxe5 Qxe5 17.f4 Qxe4 18.f5 Qxe2 19.Bxe2 Bh7 20.Rd1 hxg4 21.hxg4 0-0 22.Bf3 c6 leaves the initiative to White.


This is already indispensable as 8...Bd6 9.h3 Bh5 10.e4 e5 11.g4! gives White the upper hand after 11...Bg6 12.Nxe5
Nxe5 13.dxe5 Bxe5 14.f4 Qd4+ 15.Qxd4 Bxd4+ 16.Kh2 h5 17.f5 Bh7 18.g5.

9.h3 (9.d5 Nb6!) 9...Bh5 10.Be2 Be7

10...cxd4 11.exd4 Bd6 12.d5 exd5 13.Nxd5 0-0 14.Nf4 gains the bishop pair advantage in an open position.


The symmetry is harmless for Black, for instance 11.dxc5 Bxc5 12.b4 Ba7 13.Bb2 0-0=.

11...exd5 12.Nxd5 0-0 13.b4 Bg6 14.Nxe7+ Qxe7 15.Bb2 Rfd8=.

E. 6.Nc3 Nc6!

Only this move addresses Whites two most dangerous threats:

1. to gain the bishop pair after:

6...a6 7.h3 Bh5 8.g4 Bg6 9.Ne5 Nbd7 10.Nxg6 hxg6 11.Bf1 see game 33 Delchev-Huch, Regionalliga 2014.

2. to push e3-e4, for instance:

6...Nbd7 7.0-0! (the point is to castle before playing h3 in order to have Ne5 with a tempo in some lines) 7...Bd6
7...Bb4 8.Qb3 is awkward 8...c5 9.Ne5 or 8...Bxf3 9.Qxb4.
8.h3! Bh5
8...Bxf3 9.Qxf3 c6 is the lesser evil, but my game against Swiercz taught me a lesson the hard way.
9.e4 e5 10.g4 Bg6 11.dxe5 Nxe5 12.Nxe5 Bxe5 13.f4 is a rout.

From c6, the knight is eyeing d4. Thus 6...Nc6 7.0-0 Bd6 8.h3 Bh5 9.e4? would drop a pawn to 9...Bxf3 10.gxf3 Nxd4.

Main branches are:

E1. 7.0-0; E2. 7.Bb5

For 7.Qa4 Bxf3 8.gxf3 Bd6 9.Bb5, see game 34 Edouard-Delchev, Honor 2013.

E1. 7.0-0 Bd6 8.h3 Bh5 9.Be2 0-0


10.e4? is still impossible due to 10...Bxf3! so White should seek a way of developing his dark-squared bishop.

10.a3 does not fit well with Nc3. If White liked this extended fianchetto, he should have refrained from 6.Nc3 in order
to have Nbd2-c4.


Mind the trap 10...e5 11.Nxe5!

11.Bb2 Qe7 12.Rc1

Or 12.Ne5 Bxe2 13.Nxe2 Bxe5 14.dxe5 Nd7 15.f4 Rfd8=.

12.d5?! exd5 13.Nxd5 Nxd5 14.Qxd5 Bg6 15.Nd4 Nb4 16.Qc4 (16.Qxb7 Be4) 16...Qe5 (16...b5) 17.g3 Nc2 18.Nxc2
Qxb2 19.Rac1 Rad8, Popchev-Lekic, Serbian Open 2014.


White does not have a clear plan.
13.Ne5 Bxe2 14.Nxc6 Bxd1 15.Nxe7+ Bxe7 16.Rfxd1 Rd7 was roughly even in Leitao-Kunin, Barcelona 2006.

13.Na4 Nb4 14.a3 Nbd5 15.b4 Ne4 16.Nc5 Bxc5 17.bxc5 Bxf3 18.Bxf3 Qh4 19.Qc2 offers Black an easy game. He
can try to mount an attack with 19...Ng5 or enter a balanced endgame with 19...f5 20.Bxe4 Qxe4.

13.Nd2 Bg6 14.Nc4 e5 15.Nxd6 is well parried by 15...exd4!.

E2. 7.Bb5 Bd6


8.h3 Bh5 9.g4 Bg6 10.Nh4 Be4!;

8.Qa4 Bxf3 9.gxf3 0-0 10.Bxc6 bxc6 11.Qxc6 Rb8 gave me a nice compensation in game 34 Edouard-Delchev, Linares

8...Nd7 9.Be3 0-0 10.h3 Bh5 11.0-0

After 11.Be2 Nb6 12.0-0, Black can try the thematic break 12...f5!;

11.g4 leads to complications which are favourable for Black 11...Bg6 12.h4 Bb4 13.Nd2 Nf6 14.Bxc6 bxc6 15.f3 h5
16.g5 Nd7.


Blacks play is similar to the Grnfeld defence. His next step will be ...f7-f5. Note that this break is less risky here since
the g-pawn is on g7 and Blacks castling position is not weakened.
Instead, 11...a6?! would be justified after 12.Bxc6 bxc6 13.Qe2 e5 14.Rad1 exd4 15.Bxd4 Re8 16.Rfe1 c5 17.Be3 h6
18.b3 a5 19.Nb5 Re6, but 12.Be2! is stronger.


In practice White has tested:

12.Rc1 f5 with branches:

13.exf5 exf5 14.Bg5 Bxf3 15.Qb3+ Bd5 16.Nxd5 Nxd4 17.Bxd8 Nxb3 18.Bxc7 Nxd5 19.Bxd6 Nxc1 20.Bxf8 Kxf8
21.Rxc1 Rd8=;

13.e5 Be7 14.Bxc6 bxc6 15.Ne2, Teske-Bentivegna, Sliema 2014, 15...Nd5! 16.Rxc6
or 16.Nf4 Nxf4 17.Bxf4 g5 18.Bd2 (18.Be3 Qd5 19.Qb3 Qxb3 20.axb3 f4 21.Bd2 c5) 18...g4 19.hxg4 fxg4
20.Nh2 Qxd4 21.Bh6=.
16...Qd7 17.Rc1 f4 18.Bd2 Bg6 with full compensation.
I have also analysed 12.Re1 f5 13.Bxc6 bxc6 14.exf5 exf5 15.Qb3+ Bf7 16.Qc2 Nd5 17.Ne5 f4 18.Bc1 Bxe5 19.dxe5
f3 20.e6 Bg6 with a tangled position.



13.e5 Be7 14.Qd2 Nb4;
13.Ng5 Bxe2 14.Nxe2 (14.Qxe2 Nxd4) 14...Qd7 15.Qb3 Rae8 16.e5 h6 17.Nxe6 Qxe6 18.Qxe6+ Rxe6 19.Nf4 Ree8
20.exd6 g5 leads to a drawn rooks endgame after 21.d5 gxf4 22.Bxb6 cxb6 23.dxc6 bxc6, e.g. 24.Rac1 c5 25.b4 cxb4
26.Rc7 Rf7 27.d7 Rd8 28.Rd1 a5=.

13...exf5 14.Ng5

Kinsman-Brynell, Wrexham 1998, saw 14.d5?! Bxf3 15.Bxf3 Ne5 16.Nb5 when the d5-pawn would have been hanging
after 16...Rf7.

The computer suggests 14.Re1 when

14...f4 15.Bd2 Kh8 16.Ne4 Bxf3 eliminates to a drawish position after 17.Bxf3 Nxd4 18.Bc3 Nxf3+ 19.Qxf3
Nd5 20.Rad1 Nxc3 21.bxc3, for instance: 21...b6 22.Nxd6 cxd6 23.Re6 Qc7 24.Rexd6 Qc4. However, I prefer
the more dynamic continuation:
14...Qf6 15.a4 (15.Bg5 Qg6 16.Bd2 Rfe8; 15.d5 Bxf3 16.Bxf3 Ne5=) 15...Na5 16.Bg5 (16.Nb5 Bb4) 16...Qf7 17.Ne5
Bxe2 18.Qxe2 Qb3 with complex play.

14...Bxe2 15.Nxe2
I have been following the game Kuljasevic-Kadric, Split 2013. Now the best set-up is 15...Qf6! 16.Qb3+ Kh8 17.Nf3
Rae8 18.Rae1 f4 19.Bd2 Qf5 with mutual chances.
Kadric played instead 15...Qd7 16.Qb3+ Kh8 17.Ne6 and White retained some pull.
Chapter 16. The 4...Bg4 Variation
Annotated Games

33. Delchev 2613 Huch 2251

Regionalliga 19.01.2014

1.Nf3 d5 2.d4 Nf6 3.c4 dxc4 4.e3 Bg4 5.Bxc4 e6 6.Nc3 a6?! 7.h3 Bh5 8.g4 Bg6 9.Ne5

The possibility of this move accounts for my preference for 6...Nc6.

9...Nbd7 10.Nxg6 hxg6 11.Bf1

This paradoxical manoeuvre came to the fore after the game Kasparov-Petrosian, Tilburg 1981. Although the future
world champion lost to the former one, the postmortem proved that he had a winning attack after 11...c6 12.Bg2 Qc7
13.0-0 Be7 14.f4 Nb6 15.g5 Nfd7 16.Qg4 0-0-0 17.Rb1 Kb8 18.b4 Nd5 19.Na4 f5 20.Qg3 Nxb4 21.Bd2 Nd5 22.Rfc1
Ka7 23.Qe1 Ba3 24.Rc2 Qd6 25.Rb3 Qe7 26.Qe2 Rb8 27.Qd3 Bd6 28.Nb2 Rhc8 29.Nc4 Bc7 30.a4 b5 31.axb5 cxb5
12...Bd6 was suggested as an improvement, but in my opinion it does not solve Blacks problems either. White should
answer 13.Bd2! awaiting Black to define the position of his king 13...Qe7 (13...Qc7 14.Rc1) 14.Qb3! Rb8 (14...0-0-
0?! 15.Rc1! Bc7 16.Ne2) 15.0-0-0.
The evaluation of the position on the first diagram is very important for the theory of the 4...Bg4 variation. I believe it is
very pleasant for White so 6...a6 should be archived in favour of 6...Nc6!.

11...Bd6 12.Bg2 Rb8

Preserving the option of ...c5.

13.g5 Nd5 14.h4 Nxc3 (14...N7b6 15.Ne4) 15.bxc3 c5 16.a4 e5 (16...0-0? 17.f4) 17.Rb1 Qc7

18.dxe5 would have fixed a clear advantage, but I was afraid of technical difficulties. For example,
18...Bxe5 (18...Nxe5 19.f4 Nc4 20.Kf2) 19.Qc2 (19.Bd2!? c4 20.f4 Bd6 21.h5 Nc5 22.hxg6 Rxh1+ 23.Bxh1 fxg6
24.Bd5 is more double-edged)

19...c4 20.f4 Bd6 21.Qe4+ Kf8 22.Rxb7 Rxb7 23.Qxb7 Qxb7 24.Bxb7 Nc5 25.Bd5 Nxa4 26.Bxc4 Nxc3 27.Bxa6 wins
a pawn, but still not the game.

18...b5 19.axb5 axb5 20.Bd2 Qb6 21.Qe4 Kd8?!

The natural 21...Kf8, having in mind 22.Qc6 Ke7, would offer more resistance. Now I keep on queens and an attack.
The rest of the game was not of highest quality, but my opponent was the last to err:

22.Qd5 Ke7 23.Bf1 Qa6 24.Rh3 Qa4 25.Rf3 Rhf8 26.Bh3? (26.dxe5! Nxe5 27.Rf4) 26...Qc4 (26...Qa8!) 27.Qc6
f5? 28.gxf6+ Nxf6 29.Ra1 Rfd8 30.Ra7+ Kf8 31.dxe5 1-0

34. Edouard 2665 Delchev 2637

Linares 2013

1.d4 d5 2.Nf3 Nf6 3.c4 dxc4 4.e3 Bg4 5.Bxc4 e6 6.Nc3 Nc6 7.Bb5

G.Grigorov tried this idea after 6.h3 Bh5 7.Nc3 Nc6 8.Bb5 Bd6 and now 9.g4 Bg6 10.Nh4 Be4 11.Nxe4 Nxe4 12.Nf3
0-0 13.Qc2 f5 14.gxf5 exf5 15.Bxc6 bxc6 16.Qxc6 Qf6 17.Bd2 Rae8 gave me a big edge.

7...Bd6 8.Qa4 Bxf3 9.gxf3 0-0 10.Bxc6 bxc6 11.Qxc6 Rb8

The position is similar to the lines with Qb3, but Black misses ...c7-c5. In return, he has castled and his rook is active on
b8. Four days after the current game, Nikcevic-Kadric, Croatia 2013, saw 12.b3 Rb6 13.Qc4 Qa8 14.Ke2
14...e5 15.d5 c6 16.Ne4 Nxe4 17.fxe4 cxd5 18.exd5 when Black has 18...f5 19.Bd2 Rc8 20.Qd3 e4 with a winning
attack. Instead Kadric opted for 18...Rc8 19.Qd3 e4 (19...Rb4! 20.Bb2 Qb7 21.Rhc1 e4 22.Qd2 Qa6+ 23.Ke1 Rbb8)
20.Qxe4 Re8 and went on to win after 21.Qc4 Rc8 22.Qd3 Be5 23.Rb1 Rd8 24.e4 f5 25.exf5 Rxd5 26.Qc4 0-1.
Edouards move seems more challenging:

12.f4 Nd5!

I remembered this method from the classical game Foguelman-Bronstein, Amsterdam 1964 (with 6.Qb3). The idea is to
cut the white queen off from its king.

13.Qc4 Qd7

My pieces are harmoniously placed while Whites bishop does not have any prospects and the king is rather awkward in
the middle. The extra pawn is hardly worth the practical difficulties White is experiencing.

14.a3 Rfc8

White is lacking useful moves.

15.Qd3 Rb3

The computer prefers to open files with 15...Nxc3! 16.bxc3 c5, but I was tempted by the idea of imposing a blockade
on the light squares.

16.Qc2 Rcb8 17.Ne4

A miserable attempt to lift the blockade of b3.


The beginning of a logical finish. White must deal with the threat 18...Rxe3+. As long as 18.Kf1 Qa6+ 19.Kg1 (19.Kg2
f5) 19...f5 20.Nd2 Rxe3 21.fxe3 Nxe3 is a rout, Edouard decides to castle.


I regain the pawn, using the fact that White would be tied up and down after 19.Nd2 Ne2+ 20.Kh1 Qa6 (or 20...Qa5
21.Nxb3 Qh5 22.f4 Ng3+ 23.Kg2 Nxf1 24.Kxf1 Qb5+ 25.Kf2 Qxb3 26.Qxb3 Rxb3 with the better endgame)
21.Nxb3 Nxd4 (21...Ng3+ 22.hxg3 Qxf1+ 23.Kh2 Rxb3 24.Qxb3 Qxf2+ 25.Kh1 Qf1+ is only a draw) 22.Qd1 Nxb3
23.Rb1 Qb5. White is unable to escape the bind, e.g. 24.Rg1 g6 25.f3 Nxc1 26.Qxc1 Qe2 27.Rg2 Qxf3 28.Qd2 Rb3
29.Re1 Be5.
Alternatively: 19.Rd1 Nd5 20.Rd3 Qb5 21.Qxb3 Qxb3 22.Rxb3 Rxb3 23.Nd2 Rb8; 19.exf4 Rg3+.

19.Nc3 Nh3+ 20.Kg2 Qc4

In the postmortem, Edouard pointed out that 20...Qc6+ was even stronger 21.f3 R3b5 22.e4 Rh5.

21.f4 g5 22.Qd1
The following exchange sac finishes the battle:

22...Rxb2+ 23.Bxb2 Rxb2+ 24.Kxh3 Qxc3 25.Qg4 Qxe3+ 26.Qg3 Bxf4 27.Qxe3 Rxh2+ 28.Kg4 h5+ 29.Kf3 Rh3+ 0-

This game significantly boosted the interest towards 4...Bg4. Instead of some meagre dozen of games per year, I
counted 75 in 2014. Curiously, I saw Edouard in 2015 defending Blacks side of the pawn sac, although in a different

35. Prohaszka 2553 Delchev 2622

13th EICC, 30.03.2012

1.d4 d5 2.c4 dxc4 3.Nf3 Nf6 4.e3 Bg4 5.Bxc4 e6 6.Nc3 Nbd7?!

This move should be played only after the insertion of 6.h3 Bh5, but I was not aware of that during the game. The
difference could be seen after 7.0-0! when Black does not have a good place for his bishop 7...Bd6 8.h3 Bh5? fails to
9.e4 e5 10.g4! while 7...Bb4 8.Qb3 c5 stumbles into 9.Ne5. Apparently my opponent (and many other players!) was
not familiar with this line either as he followed up with:

7.h3?! Bh5 8.0-0 (8.g4 Bg6 9.Nh4 Be4) 8...Bb4


9.Be2 0-0 10.Qb3 c5 11.dxc5 Qa5 is roughly equal, but:

9.a3 Bxc3 10.bxc3 0-0 11.Bd3 retains a slight edge 11...Bg6 12.Bxg6 hxg6 13.Re1 c5 14.e4 Qa5 15.Bd2 Qa6,
Lugovoi-Sulava, Saint Vincent 1999, 16.a4! although Black does not have any weaknesses.

9...c5 10.a3

Now 10.Ne5 does not hit the bishop so Black can castle 10...0-0 11.Ne2 a5! 12.Nf4 a4 13.Qc2 cxd4.
10.d5 Bxf3 11.gxf3 Ne5 12.dxe6 Nxf3+ 13.Kg2 Nh4+ gives Black counterplay.

10...Bxc3 11.bxc3 Qc7 12.a4 0-0 13.Nd2 Nb6

After 13...Rfc8 14.a5 b6 15.axb6 axb6 16.Rxa8 Rxa8 17.f3, Blacks pieces will be severely restrained.

Now Black escapes the bind. The critical line is 14.Bb5! a6 (14...Rfc8 15.a5 Nbd5 16.e4 c4 17.Qc2 Nf4 18.Re1 and the
c4-pawn is doomed to fail) 15.Bd3 Bg6 16.Bxg6 (16.e4 c4) 16...hxg6 17.Rb1 (17.Ba3 Nbd7 18.a5 Rfc8 19.Qa4 g5)
17...Nbd7 18.Qxb7 Qa5. White will return the pawn and keep a small advantage with 19.Qb3 Rfb8 20.Qa3 Rxb1
21.Nxb1 Rc8 22.Nd2 Nd5 23.c4 N5b6 24.Qb3.

14...Rfc8 15.Rfc1 Nxc4 16.Qxc4 Nd7 17.Nb3 Bg6 18.dxc5 b6 19.Rd1 Bc2 20.Rd2 Bxb3 21.Qxb3 Nxc5 22.Bxc5
Qxc5 23.a5 bxa5 24.Rd7 Rab8 25.Qa4 Qf5 26.Qf4 Qxf4 27.exf4 Rxc3 28.Rxa5 g6 29.Raxa7 Rf8 30.g3 h5 31.h4 Rc5
32.Kg2 Rf5 33.Re7 Rd5 34.Rab7 Rf5 35.Ra7 Rd5 36.Kf3 Rb5 37.Rac7 Ra5 38.Rb7 Rd5 39.Kg2 Ra5 40.Rbc7 Rf5
41.Ra7 Draw.

36. Korchmar 2406 Delchev 2606

Paleochora, 21.07.2015

1.d4 d5 2.c4 dxc4 3.Nf3 Nf6 4.e3 Bg4 5.Bxc4 e6 6.0-0 a6 7.h3 Bh5 8.b3

8...c5 is a solid equaliser, but with more than 200 Elo points in my favour, I had to seek complications.
If White had committed his knight to c3, my only active set-up would have been based on ...Nc6. His passive
development gives me another option which looked more flexible to me. Let me remind you, that ...Nbd7 is bad against
Nc3 due to e4 followed up by g4.

9.Bb2 Be7

9...Bd6 10.Nbd2 0-0 11.e4 Be7 was more tangled, but I decided to push ...c5. Since 9...c5 10.d5 would be awkward, I
wait for 10.Nbd2.
On the other hand, 10.Nc3 0-0 11.e4 b5 (11...e5 12.g4 exd4 13.Qxd4) 12.Bd3 c5 13.e5 Nd5 14.Nxd5 exd5 15.dxc5
Nxc5 would suit me perfectly since it leads to a dynamic position with mutual chances.

10.Nbd2 0-0 11.Be2

11.e4 is just a weakness because the white pieces are not set up for an attack. A good retort would be 11...c5 12.e5 Nd5
13.Ne4 cxd4 14.g4 Bg6 15.Qxd4 Nb4.
After the text, I could not find another useful waiting move and braced myself for the drawish ...c5. The computer
hints that 11...Nd5 or 11...c6 + ...a5 are also playable.
11...c5 12.dxc5 Bxc5 13.Nc4

13.a3 Qe7 14.b4 Bd6 15.Nc4 Bc7= occurred in Nikolac-Santos, Almada 1988.

13...Qe7 14.Nh4

This blunt play for a draw surprised me. I expected 14.a3 Rfd8 15.b4 Ne5.

14...Bxe2 15.Qxe2 b5 16.Ne5

16.Na5?! Rac8 17.a4 Bb6 18.axb5 does not win a pawn due to the double hit 18...Bxa5 19.Rxa5 Qb4 20.Ra4 Qxb3
21.Ra3 Qxb5 22.Qxb5 axb5.

16...Nxe5 17.Bxe5 Nd5

Perhaps 17...Nd7 18.Nf3 Nxe5 19.Nxe5 Qc7 gave more chances.

18.Nf3 Bd6 19.Bxd6 Qxd6 20.Qb2 Rac8 21.Rac1 b4 22.Nd4 a5

Im sightly better, but 22...Nc3 23.Ne2 Qc6 24.Nxc3 bxc3 25.Qc2 is drawish. My further attempts to set problems to
the opponent failed.

23.Ne2 Qa6 24.Rxc8 Rxc8 25.Rc1 h6

Or 25...Rxc1+ 26.Qxc1 g6 27.Qc2 Nc3 28.Nxc3 Qc6 29.Qe4.

26.e4 Rxc1+ 27.Qxc1 Qxe2 Draw.

Chapter 17. 1.d4 d5 2.c4 dxc4 3.Nf3 a6

Main Ideas

1.d4 d5 2.c4 dxc4 3.Nf3 a6

This is a very special line for the theory of the QGA. It has the potential to become the established killer of the Classical
System. The efforts of Stevic and Svidler in the last 2-3 years have promoted this humble move to Blacks main
equaliser against 3.Nf3 at top level! Blacks idea is not to keep the extra pawn, but to suck the life out of the position
as early as possible. While this is a completely alien ideology to me, it may be very effective at highest level. If you
check the stats, White scores only 52% compared to 59% against 3...Nf6. Any serious player should seek a workaround
when building his White repertoire. That explains the extreme popularity of 3.e4 and 3.e3 lately.
When we come to details, well see that the whole dispute is focused on one line:

4.e3 b5 5.a4 Bb7 6.b3 e6 7.bxc4 bxc4 8.Bxc4 Nf6

Obviously, the a6-pawn is the only target, but recent developments show that even taking it, White might end in a dead
draw position.
Blacks desired set-up is simple ...c5, ...Nbd7, ...Be7, ...0-0 followed up by ...cxd4 and total exchanges.

Whites only logical retort is to put Nc3, hoping to meet ...c5 with d4-d5. However, 9.Nc3 fails to 9...Bb4 10.Qb3 c5=.
So White should try:
9.0-0 when 9...Nbd7?! would enable 10.Nc3! c5 11.Rb1 Qc7 12.d5 Nb6 13.Rxb6!. The solution is to anticipate Nc3 by
9...c5! 10.Ba3 (10.Nbd2 cxd4) 10...Nbd7 11.Nbd2

11...Be7 12.Rb1 Rb8 13.Nb3 cxd4 14.Bxe7 Qxe7 15.Nfxd4 Qd6.

Since the onus was clearly on White, the first players put heir hope for an advantage in an early Rb1. It was meant to
drag Blacks pieces to vulnerable stands in the event of ...Ra7 or ...Qc7, but Black has a surprise in store:
9.Nbd2 Nbd7 10.Rb1 Rb8! 11.0-0 Be7 12.Qe2 0-0!
Practical experience has seen Black equalising easily after 13.Bxa6 (13.Bb2 c5 14.Bxa6 Bxa6 15.Qxa6 Ra8)
13...Bxa6 14.Qxa6 Rxb1 15.Nxb1 Qa8. With correct play Black should be able to regain the pawn.

Chapter 17. 1.d4 d5 2.c4 dxc4 3.Nf3 a6

Step by Step

1.d4 d5 2.c4 dxc4 3.Nf3 a6

A. 4.e4?!; B. 4.e3

A. 4.e4?! b5 5.a4

5.Nc3 Bb7 transposes to Chapter 14 line B, while 5...e6 6.a4 b4 7.Ne2 c5 is rather risky due to 8.d5 exd5 9.Nf4 dxe4
10.Qxd8+ Kxd8 11.Ng5 Ra7 12.Bxc4 Nh6 13.Be3.

5...Bb7 6.axb5

6.Nc3 b4 7.Bxc4 e6;
6.b3 e6 7.axb5 (7.bxc4 Nf6!? 8.axb5 Nxe4 9.Qb3 axb5 10.Rxa8 Bxa8) 7...axb5 8.Rxa8 Bxa8 9.bxc4 Bxe4 10.Nbd2
Bb7 11.cxb5 Nf6 12.Bd3 Nbd7 13.0-0 Nb6 14.Nb3 Qd5 is unclear, Beil-Krnavek, Olomouc 1976.

6...axb5 7.Rxa8 Bxa8 8.Nc3 c6

8...e6 9.Nxb5 Bxe4 10.Bxc4 Nf6 11.0-0 c6 12.Nc3 Bg6 13.Ne5 Bd6 is quite solid, but why not to keep the pawn?!
In this position White has tested 6 different moves, without any success.

The tricky 9.d5 is tamed by 9...Nf6 10.Bf4 cxd5! 11.exd5 Nbd7! 12.Be2 b4 13.Nb5 Nxd5 14.Bg3 Qc8 15.0-0 f6
16.Re1 Kf7.

9.e5 is also too committal 9...e6 10.Be2 Ne7 11.0-0 Nd5 12.Ng5 Be7 13.f4 Nxc3 14.bxc3 c5, Knaak-Gruenberg,
Dresden 1985.

The most popular answer is:

9.Be2 e6 10.0-0 Nf6 11.Bg5

11.Qc2 Be7 12.e5 Nd5 13.Qe4 is a principled queen lift, but in fact White does not have serious threats after 13...h6
14.Qg4 Nxc3 15.bxc3 g6 or 13...h5 14.Bd2 Nd7 15.Ra1, Rustemov-Grabarczyk, Germany 2010, 15...Nc7 16.Qf4

11...Nbd7 12.Qa1 h6

Black comfortably completes development.

B. 4.e3 b5 5.a4 Bb7 6.b3

6.Nc3?! b4 7.Nb1 Nf6 8.Bxc4 e6 9.0-0 Nbd7 10.Nbd2 c5 offers Black 2 extra tempi compared to the common lines of
the Classical System.

6.axb5?! axb5 7.Rxa8 Bxa8 8.b3 Nf6 is dead equal. Youll understand why when you finish studying the main line
where the only problems for Black ensue from the weak pawn on a6.

6...e6 7.bxc4 bxc4 8.Bxc4

After 8.Ba3, simplest is 8...c5 9.Bxc4 cxd4 10.Bxf8 Kxf8 11.0-0 Nc6 12.exd4 g6 13.Nc3 Kg7=.


Now White faces an important choice. In my opinion, his strongest option is:

9.Nc3 gives a tempo for 9...Bb4 followed by 10...c5=.

9.Qb3 Bxf3 10.gxf3 c5 is even more pleasant for Black.

Whites main alternative to 9.Nbd2 is 9.0-0 when 9...c5! has been bringing Black good results.
Svidler insists on his idea of delaying ...c5 (see the main line) and plays 9...Nbd7 when 10.Nbd2 transposes.
However, I do not like the more active continuation 10.Nc3! c5 (10...Be7 11.Rb1 Ra7 12.Ne5 Nxe5 13.dxe5
Nd7 14.f4 0-0 assures White of a lasting space advantage) 11.Rb1! cxd4 (The point of Whites idea is seen in the
line 11...Qc7 12.d5 Nb6

13.Rxb6! Qxb6 14.dxe6 fxe6 15.Qe2 Qc6 16.e4!! with a crushing attack.) 12.exd4 Qc7 13.Qe2 Bxf3 (13...Be7
14.Bf4!) 14.Qxf3 Rc8 15.Bxa6 Qxc3 16.Bxc8 Qxc8 17.Bg5 Be7 18.a5. Black is unable to coordinate his pieces.
10.Nbd2 cxd4 11.Rb1 Qc7 is discussed in game 37 Dronov-Lafarga Santorroman, ICCF email 2009.
10...Nbd7 11.Nbd2

11...cxd4 12.Bxf8 Nxf8 13.Nxd4 N8d7 14.Rb1 Rb8 15.Qe2 Nc5 16.N2b3 Nxa4 17.Ra1 Nc3 occured in two
drawn games, but instead of 18.Qb2 Ncd5 19.Na5 Ba8 20.Qa3 Qe7 21.Qa4+ Qd7, Lalic-Stevic, Bol 2014,
White could set difficult problems with 18.Qe1!, e.g. 18...Nce4 19.Bxa6 0-0 20.f3 Bxa6 21.Rxa6 Rb6 (21...Qc8
22.Qa5) 22.Rxb6.
Or 12.Qe2 cxd4 13.Bxe7 Qxe7 14.Nxd4 0-0 15.Rab1 Ne5=.
12...Rb8! 13.Nb3 cxd4 14.Bxe7 Qxe7 15.Nfxd4 Qd6. Black should gradually level the game, for instance 16.Na5
Ba8 17.Ndc6 Qxd1 18.Rfxd1 Bxc6 19.Nxc6 Rc8 20.Na5 0-0.

9...Nbd7 10.Rb1 Rb8!

Black used to defend with 10...Ra7, but his pieces remain passive and uncoordinated. White maintains a lasting pull
after 11.0-0 Be7 12.Qe2 c5 13.Bb2 0-0 14.Rfc1 a5 (14...Qa8 15.Bd3) 15.Bb5 cxd4 16.Bxd4 Bxf3 17.Qxf3,
B.Socko-Westerberg, Sweden 2015.

11.0-0 Be7 12.Qe2 0-0


It is indicative that Topalovs second, Cheparinov, was totally helpless against Svidlers pawn sac in Tromso 2014. His
improvement 13.Bb2 was too lame and the game was even after 13...c5 14.Bxa6 Bxa6 15.Qxa6 Ra8 16.Qc4
Or 16.Qb5 Ra5 17.Qb3 (17.Qc6 Qc8=) 17...Qa8 18.Ra1 Rb8 19.Qc2 cxd4 20.Nxd4 Rc8=.
16...Nb6 17.Qc2 Nxa4 18.Ba1 cxd4 19.Bxd4 Qc8=.

13...Bxa6 14.Qxa6 Rxb1 15.Nxb1 Qa8 16.Qc4!

Another encounter was Grischuk-Svidler, Khanty-Mansiysk 2015, where White practically offered a draw with 16.Qxa8
Rxa8 17.Ba3 Bxa3 18.Nxa3 Rxa4 19.Nb1 c5=. By all means it is more testing to keep the extra pawn.

16...c5 17.Qa2!?
17.Ba3 Nb6 regains the pawn 18.Qb3 Nxa4 19.Nbd2 Rb8 20.Qc2 Rc8 21.Rc1 Nd5 22.Nb3 c4 23.Bxe7 Nxe7
24.Nbd2 Nd5=.


Perhaps this is the only way to restore the material balance. 17...Nd5?! 18.Bd2 cxd4 19.Nxd4 Rc8 20.a5 keeps the
extra pawn.

18.Bd2 Ne4 19.Rc1 Rc7 20.Nc3

20.Be1 Ra7 21.a5 is only a temporary defence 21...Bd8 22.Qa4 Bxa5 23.Bxa5 Nef6 24.dxc5 Rxa5 25.Qd4 Qc6
26.Qd6 Qxd6 27.cxd6 h6 28.Rd1 Ne4= followed by ...Rd5 or ...Ra6.

20...Nxd2 21.Qxd2 cxd4 22.Nxd4 Nb6

It is time to sign a draw.

Chapter 17. 1.d4 d5 2.c4 dxc4 3.Nf3 a6
Annotated Games

37. Dronov 2675 Santorroman 2646

ICCF email 2009

1.d4 d5 2.c4 dxc4 3.Nf3 a6 4.e3 b5 5.a4 Bb7 6.b3 e6 7.bxc4 bxc4 8.Bxc4 Nf6 9.0-0 c5


White aims to put in Rb1 quickly. I believe that 10.Ba3 is a sterner test
of Blacks set-up.

10...cxd4 11.Rb1 Qc7 12.exd4 Bd6

Black is behind in development and any further delay of castling may be punished with a break in the centre. Balutescu-
Sladek, ICCF email 2009, saw 12...Nc6?! 13.Bb2 Be7 14.d5! exd5 15.Bxf6 Bxf6 16.Bxd5 0-0 17.Qc2 Rac8 18.Ne4
with an edge.
12...Be7 could be a decent alternative to the game. It looks that Blacks knight on c6 is in a predicament after 13.Qb3
Bd5 14.Ba3 0-0 15.Rfc1 Bxa3 16.Qxa3 Nc6,
but the unprotected white queen enables the trick 17.Ne5 Bxc4 18.Rxc4 Qe7, which saves the day. The high-level email
game Martin Clemente-Povchanic, 2010, was drawn a few moves later 19.Qc1 Nxe5 20.dxe5 Nd5 21.Ne4 Qd8
22.Nc5 draw.

13.Ba3 Bxa3 14.Qb3 0-0 15.Qxb7 Ra7 16.Qb3 Bd6 17.Rfc1 Qe7

Blacks pieces may look passive and stranded, but they are defending each other well. That gives him time to

18.Bf1 Nbd7 19.Nc4 Nd5 20.Na5

In this line, Blacks second big point of concern besides the a6-pawn, is the c6 square. Santorroman displays that in fact
Whites activity is totally harmless.

20...Nb8 21.g3 g6 22.Ne5 Bxe5 23.dxe5 Rc7 24.Rxc7 Qxc7 25.Nc4 Nd7

The weakness on a6 is balanced by the vulnerable e5-pawn.

26.Qb7 Qxb7 27.Rxb7 Nc5 28.Ra7 f6

Threatening ...Rf8-f7-f8.

29.Nd6 fxe5 30.Nb7 Nxb7 31.Rxb7 Rc8 32.Bxa6 Rc1+ 33.Kg2 e4

Blacks threats along the second rank compensate Whites passer, e.g. 34.Bb5 Ra1 35.Ra7 Ra2 36.Kg1 Ra1+.

34.Bf1 Draw.

Ods and Ends

QGA fans often have to face Queens Pawn Openings. First players suppose that if we like open centre, it would be a
wise idea to deprive us of the possibility to take on c4. So they play anything but 2.c4.
I also have something in store for them. They commonly expect that well rush to lead out our bishop to f5. Not that it is
a bad move, of course, but after all, Black has so many other decent set-ups! I suggest a very clear and easy to play
scheme against 1.d4 d5 2.Bf4 which does not require more than 10 minutes of preparation.
Against 1.d4 d5 2.Bg5 I recommend drastic measures 2...f6! intending to take over the initiative right from the

I hope that this part of the book will help you win the psychological and, maybe, the opening battle against such anti-
QGA opponents.
Chapter 18. 1.d4 d5 without 2.c4
Main Ideas

Every QGA player will often have to

deal with set-ups where White refrains from c4. Commonly you would have a wide choice against them, but I would
like to offer you some advice against the most fashionable of those lines.

Do not try to transpose to the QGA against the Rti move order:
1.Nf3 d5 2.c4 dxc4?! 3.e3 Nf6 4.Bxc4 e6 5.0-0 c5 6.Qe2 a6 7.Rd1

This position is unpleasant for Black. If you push 7...b5, White will attack it with a4, Nc3, aiming to provoke ...b4. Then
he will limit the scope of your b7-bishop with pawns on d3+e4, leaving you without an active plan. The best retort is to
gain space in the centre with 2...d4! 3.e3 Nc6.

1. If White plays Bf4, simplest is to aim for the following position:

1.d4 d5 2.Bf4 Nf6! 3.e3 e6!? 4.Nd2 Bd6 5.Bg3 c5 6.c3 Nc6 7.Ngf3 0-0 8.Bd3 b6
The point here is to meet 9.e4 by 9...Be7!. See game 39 Kamski-Nakamura, Saint Louis 2015.

2. In the Colle System 1.d4 d5 2.Nf3 Nf6 3.e3 c5 4.c3 e6 5.Nbd2,

White hopes to open the centre by taking on c5, followed up by e3-e4. We can address that with 5...Nbd7 6.Bd3 b6
when e3-e4 is not effective. White usually builds up a Stonewal formation with 7.Ne5 Bb7 8.f4 g6 9.0-0 Bg7. As
Petrosian said, the main thing a player should do against the Stonewall, is not to hinder the opponent to play it! Then
Blacks game plays it by itself.
See my game 40 Jussupow-Delchev, Top 16 France 2006 where I achieved the main strategic goal to push ...e5.

3. 1.d4 d5 2.Bg5 offers a chance to gain space on the kingside with 2...f6!?

Then 3.Bh4 Nh6 is nice for Black. We meet 3.Bc1 and 3.Bf4 with 3...Nc6, ...Bf5/g4, ...e6 with rich play. Our king will
be safe on the queenside.
Chapter 18. 1.d4 d5 without 2.c4
Step by Step

A. 1.d4 d5 2.Bf4 Nf6! 3.e3

3.c4 could be met by 3...e6 4.Nc3 c5 5.e3 Nc6 6.Nf3 (6.Nb5 Qa5+ 7.Qd2 Nb4) 6...cxd4 7.exd4 Bb4.


I propose you a simple and clear plan which should let you play chess without fearing long home analyses.
3...c5 may be the most challenging retort to Whites quiet development, but you should be well prepared. White can
then test your knowledge in two sharp lines:

4.Nc3! This attack became popular after Rapports victory over Sutovsky in Trompso 2014: 4...cxd4
Two false trails are 4...Nc6? 5.Nb5 and 4...a6?! 5.dxc5.
To 5.Nb5, Avrukh gives 5...Qa5+ 6.b4 Qxb4+ 7.c3 dxc3 8.Nc7+ Kd8 9.Nxa8 e5 10.Bxe5 c2+ 11.Qd2 Bf5.
5...Bg4!? is also interesting, see game 38 Delchev-Forchert, Bad Wiessee, 2014.
6.Nf3 Nc6 7.Ne5 e6 8.Nxc6 bxc6 9.Na4 Bd6 10.Bxd6 Qxd6 11.Bd3 0-0 12.0-0 e5, Rapport-Bu Xiangzhi,
Tsaghkadzor 2015.

4.dxc5 Nc6 5.Bb5 e6 6.b4 a5 7.c3 Bd7

Meister quickly became worse against me in 2015 after 8.Nf3 axb4 9.Bxc6 Bxc6 10.cxb4 b6 11.Ne5 Ba4 12.Qc1
bxc5 13.Nc3 Bd7 14.bxc5 Bxc5 15.0-0 Ba3 16.Qd2 Qa5.
8...Ne4 9.f3 e5 10.fxe4 exf4 11.Ne2 (11.exd5 Qh4+) 11...f3 with sharp play.

A calmer option is 4.c3 Nc6 5.Nd2 e6 (5...cxd4 6.exd4 Bf5=) 6.Ngf3 Bd6 7.Bg3 0-0.


The main protagonists of this system prefer the knight development over 4.Nf3 since the latter offers Black the
additional option of 4...Bd6 5.Bg3 Ne4 6.Nbd2 f5 7.c4 0-0. However, I believe that playing the Stonewall (any version
of it!) is always a bad idea. It is better to transpose to the main line with 4...c5! 5.c3 (5.Nc3 a6) 5...Nc6 6.Nbd2 Bd6.

In all those d-pawn systems, White can always return to c4 so we should consider 4.c4. In a blitz game against
Nakamura, Anand answered 4...Be7, transferring the game into the QGD. I recommend instead 4...c5!,

aiming to get a position from the Panov Attack with the senseless move Bf4:
5.Nc3 cxd4 6.exd4 Bb4 (or 6...dxc4 7.Bxc4 Bd6) 7.Nf3 0-0 8.Rc1 b6 9.Bd3 dxc4 10.Bxc4 Bb7 11.0-0 Nc6.
5...cxd4 6.exd4 Bb4+ 7.Nc3 0-0 8.a3 (8.c5 b6; 8.Rc1 b6) 8...Bxc3+ 9.bxc3 Nc6 10.Be2 (10.Rc1 Na5) 10...Na5=.

4...Bd6 5.Bg3

A standard retort. 5.Ngf3 Bxf4 6.exf4 is best met by 6...b6! 7.Bd3 c5 8.c3 0-0 (already threatening to trade the bishops)
9.Qe2 a5 10.0-0 Ba6.


Or 6.dxc5 Bxc5 7.Ngf3 0-0 8.a3 (8.c4 Qb6 9.Qc2 Nc6 10.cxd5 exd5) 8...b6 9.c4 when both 9...Bb7 and 9...Ba6 10.Be2
dxc4 11.Nxc4 Qxd1+ 12.Rxd1 Rc8 are playable.

6...Nc6 7.Ngf3 0-0 8.Bd3

White leads out his bishop on d3 semi-automatically, but I see a lot of reason in 8.Bb5 as played by Kamsky in 2015.
We can meat it by 8...Bxg3 9.hxg3 Qb6 10.Bxc6
10.a4 cxd4 11.exd4 (11.cxd4 Bd7) 11...a6 12.Bxc6 bxc6 13.b4 (13.a5 Qxb2) 13...a5 14.0-0 Ba6=.
10...bxc6 11.Qc2 (11.Rb1 cxd4 12.cxd4 c5)

Kamsky-Sevian, Chicago 2015, went 11...Rb8 12.Rb1 cxd4 13.cxd4 Ba6 14.Ne5 when 14...Qb5!? 15.Nb3 Ne4 is an
interesting exchange sacrifice. It is unclear how White could connect his rooks. 16.Nd7 h6 17.Nxf8 Kxf8 18.Nc5 leads
to a perpetual after 18...Qb4+ 19.Kd1 Nc3+ 20.Kc1 Nxa2+=.

More accurate is 11...cxd4! 12.cxd4 Ba6 with the following options:

a) 13.Nb3 h6 14.Nc5
14.Ne5 Ne4 15.0-0-0 (15.Nd7 Qb4+) 15...Bc4 16.Nd7 Qa6 17.Nxf8 Rxf8 18.Kb1 Be2.
b) 13.Ne5 Rfc8 14.Nb3 Ne4 15.Nd7 Qb4+ 16.Nd2 c5 17.a3 Qa5 18.Nxc5 Rab8;
c) 13.a3 Qb5 (13...Rfc8 14.b4) 14.Nb3 Ne4.
Blacks counterplay along the f1-a6 diagonal compensates for his worst pawn structures.


8...Qe7 is a decent alternative. Then

9.Ne5 is the only way to stop ...e5 since 9.0-0 Bxg3 10.hxg3 e5 and 9.dxc5 Bxc5 10.0-0 e5 11.e4 Rd8 are fine for
9...Nd7! 10.Nxd7 Qxd7!
10...Bxd7? 11.Bxd6 Qxd6 12.dxc5 Qxc5 13.Bxh7+ Kxh7 14.Qh5+ Kg8 15.Ne4 suddenly wins as after 15...g6
16.Nxc5 the bishop is hanging on d7. After the text, the same line would lead only to a draw.

11.0-0 b6! 12.Bxd6

12.dxc5 bxc5 13.e4 Rb8;
12.Qh5 f5 builds up the stonewall where Black needs his dark-squared bishop. Thus 13.Bb5 should be met by
13...Be7! 14.Nf3 Bb7 15.Ne5 (15.Ng5 h6 16.Nf3 Bf6 17.Be5 Qf7 18.Qxf7+ Kxf7 19.Bxf6 gxf6) 15...Qc8
16.Bxc6 Bxc6 17.Bf4 Be8 18.Qf3 Qd8 19.h3 g5 20.Bh2 Rc8.
12...Qxd6 13.f4 f5. Black did not have any problems in Grischuk-Hammer, Stavanger 2015.

9.e4 Be7!
Until recently, I believed that 8...b6 was a mistake due to 9.e4 dxe4?! 10.Nxe4 Nxe4 11.Bxe4 Bb7 12.Qa4 Rc8 13.dxc5
Bxc5 14.Rd1 Qf6 15.0-0, Sedlak-Prusikin, Regionalliga 2014. However, the amazing bishops retreat maintains the
tension. See game 39 Kamski-Nakamura 2015 for further details.

B. 1.d4 d5 2.Nf3 Nf6 3.e3 c5

If you have mastered the ...Bg4 system, you could answer 3...Bg4 4.h3 (4.c4 dxc4) 4...Bh5, but the text is undoubtedly


The fianchetto 4.b3 offers us several good plans. The most natural one is 4...Nc6 5.Bb2 Bg4.

The Colle formation looks very passive at first glance, but in fact White nurtures rather aggressive plans:

1. To open up the centre by taking on c5, followed up by e3-e4.

2. To build a stonewall attacking formation with Ne5, f4.

I will follow the course of my game 40 Jussupow-Delchev, Nice 2006.

4...e6 5.Nbd2

My plan is to fianchetto the f8-bishop which is the main set-up against the stonewall regardless of the colours. Since the
c5-pawn will be hanging and 5...cxd4?! would be a strategic mistake as it prolongs the diagonal to the c1-bishop, I
protect it with my queens knight.

5...Nc6 6.Bd3 Bd6 7.0-0 0-0 8.Qe2 leads to the Meran with colours reversed where Whites extra tempo ensures him a
pleasant game, e.g.: 8...Qc7 9.dxc5 Bxc5 10.e4 or 8...e5 9.dxc5 Bxc5 10.e4.

5...b6!? is an improved version of the above line. 6.Ne5 Bd6 7.f4 (7.Bb5+ Nfd7 8.f4 0-0 9.Qh5 Qe7 10.Bd3 g6 11.Qh6
f6; 7.Bd3 0-0 8.0-0 Ba6 9.c4 cxd4 10.exd4 Qc7) 7...0-0 8.Bd3 Ba6 is balanced.

6.Bd3 b6

This move order has two advantages:

1. It neutralises the plan with e3-e4 as Black will gladly trade his light-squared bishop for its more active protagonist on
2. It enables ...bxc5 in the event of White taking on c5.

6...g6 is of course also possible. Astasio Lopez-Granda Zuniga, Linares 2013, went 7.0-0 Bg7 8.e4 dxe4 9.Nxe4 Nxe4
10.Bxe4 0-0 11.Be3 Qc7 (11...cxd4 12.Bxd4 e5 13.Be3 Qc7) 12.Qd2 Nf6 (12...cxd4 13.Nxd4 a6 14.Bh6 Nf6 15.Bxg7
Kxg7 16.Bc2 Rd8 17.Rad1 Bd7) 13.Bd3 b6 (13...Rd8 14.Bf4 Qe7 15.dxc5 Qxc5) 14.Bf4 Qe7=.

6...Bd6 is the most popular set-up, but after 7.0-0 0-0 8.Re1 White will play e4 and take by rook. His game would be too
easy I think.


The point of my last move is that 7.e4 is less dangerous since 7...cxd4 8.cxd4 (8.Nxd4 Nc5 9.Bb5+ Bd7) 8...dxe4
9.Nxe4 Bb7 10.Nxf6+ Qxf6 11.Bg5 Bxf3 12.Qxf3 Qxf3 13.gxf3 Bd6 is at least equal.
7.0-0 Bb7 8.Re1 (8.Qe2 Be7 is similar 9.e4 dxe4 10.Nxe4 Nxe4 11.Bxe4 Bxe4 12.Qxe4 0-0= or 10...0-0 11.Nxf6+
Bxf6) would make me lead out the bishop on e7 8...Be7 (8...g6 9.e4 cxd4 10.Nxd4 Bg7 11.e5 would be awkward).
After 9.Ne5 Nxe5 10.dxe5 Nd7 11.f4 0-0 Black is fine:
White would be clearly worse after 12.e4?! c4 13.Bc2 Nc5;
12.Qh5 f5 13.exf6 Nxf6 14.Qh3 Bc8 15.Nf3 e5 16.Bf5 exf4 is roughly equal.

7...Bb7 8.f4 g6 9.0-0 Bg7

Both sides have defined their plans. See the rest of the game in the Annotated Games section.

C. 1.d4 d5 2.Bg5
I have always been happy to meet this move by 2...h6 3.Bh4 c6 followed by 4...Qb6. However, Semkov claims that the
bishops move is outright dubious and it should by punished with:


It is rare to play for an advantage since move with Black, but White must be already careful! For instance, his most
natural retreat 3.Bh4 puts the bisop in a precarious situation and we start the chase with 3...Nh6 4.e3?! Nf5 5.Bg3 h5!

As blunt as this idea looks, it is still very effective. White is losing material:
6.Be2 h4 7.Bf4
7.Bh5+ Kd7 8.Bg4 e6 9.Bf4 g5 10.e4 dxe4 11.Bc1 c6.
7...g5 8.Bh5+ Kd7 9.e4 dxe4 10.Bc1 c6.

More cautious alternatives are 4.c4 Nf5 5.Bg3 c5 and 4.f3 Nc6 5.Nc3 (5.c3 e5) 5...Nf5 6.Bf2 e5 7.e3 Be6 8.g4 (8.Bb5
Qd6) 8...Nfe7, but Black certainly possesses the initiative.

Probably 3.Bc1 is Whites best option, but the mere fact that we are seriously discussing such moves hints that the
bishops sortie on g5 was premature. Black has several worthy continuations. I like 3...Nc6 4.e3 (4.Nf3 e5) 4...e5
5.Nf3 Bg4. Perhaps we could assume that main line is:

3.Bf4 Nc6 4.Nf3 (4.e3 e5) 4...Bf5 5.e3 (5.c4 e5) 5...e6 6.Bd3 Bxd3 7.Qxd3 Qd7 8.Nc3 g5 9.Bg3 0-0-0

The left wing is symmetrical while on the kingside Black has some space advantage so he cannot be worse here.

Chapter 18. 1.d4 d5 without 2.c4

Annotated Games

38. Delchev 2633 Forchert 2376

Bad Wiessee, 31.10.2014

1.d4 d5 2.Bf4 Nf6 3.e3 c5 4.Nc3 cxd4 5.exd4 Bg4!?


6.Bb5+ Nc6 7.Qd2 Rc8 8.Nge2 e6 9.Ng3 Bh5;

6.Nge2 Nc6 7.f3 Bd7! (7...Bf5 8.g4 Bg6 9.h4 h6 is also possible, but there is no reason to give White free tempi for his
kingside advance) 8.Qd2 a6 9.0-0-0 e6
Black can attack with pawns ...b7-b5-b4, or with pieces, following Sicilian models ...Rc8, Nc6-a5.

6...Bd7! 7.g4 e6 8.a3 Be7 9.Bd3 Nc6 10.Nge2 a6

Both sides have finished development, but Whites set-up lacks harmony. His d4-pawn is weak (Black could bring
another hit on it with ...b5, ...Qb6), his king does not feel safe in the middle. On the opposite, Black could always castle
short. Perhaps I should have chosen here a plan of regrouping, for instance, with 11.Qd2 b5 12.Nd1 and castle

11.Bg3 b5 12.h4 Qb6 13.Bf2 Bd6

This is not a mistake, but the bishop stood well enough. Black had more useful moves as ...0-0 or ...Rc8 or even ...Qb7.

14.Qd2 Qc7 15.Kf1

The critical moment of the game. It may seem that White has a space advantage on the kingside, but in fact I have
nothing but weaknesses there. Forchert could have underlined that by 15...h5! 16.g5 Ng8 and if 17.g6, then simply
17...f6 with the better game. Instead, he just gave me the possibility of trading my clumsy c3-knight for a bishop.

15...b4? 16.axb4 Nxb4 17.Nb5 Bxb5 18.Bxb5+ Ke7 19.Ba4 Rhb8 20.h5

All my pieces miraculously became active and dangerous. Black did not cope with this turn of events and quickly fell
under attack:

20...Nd7? 21.Qg5+ Kf8 22.h6 g6 23.Bxd7 Qxd7 24.Qf6 Qb5 25.Qg7+ Ke8 26.Qxh7 Nxc2 27.Rd1 Bf4 28.Qg7 Ke7
29.h7 Rh8 30.Bh4+ Kd6 31.Qxf7 g5 32.Bf2 Raf8 33.Qg6 Be3 1-0

39. Kamsky 2683 Nakamura 2798

ch-USA Saint Louis 03.04.2015

1.d4 Nf6 2.Nf3 e6 3.Bf4 c5 4.e3 Nc6 5.c3 d5 6.Nbd2 Bd6 7.Bg3 0-0 8.Bd3 b6 9.e4


This surprising retreat levels the game.

9...dxe4 10.Nxe4 Nxe4 11.Bxe4 Bb7 12.dxc5 Bxc5 13.Qa4 Rc8 14.Rd1 Qe7 15.0-0 Rfd8 16.Rfe1 leaves White with
the more active pieces, but:
9...Nh5!? 10.Bxd6 Qxd6 11.e5 Qc7 was also possible since White cannot save his strong light-squared bishop from


10.dxc5 bxc5 11.0-0 Re8 12.Re1 Qb6 13.b3 Ba6=;
10.exd5 Qxd5! 11.Bc4 Qf5 12.dxc5 Bxc5
13.Qe2 Rd8 14.0-0 Bb7 15.Nb3 Na5 16.Nxa5 bxa5=.

10...Nh5 11.0-0 Bd7

11...c4!? 12.Bc2 b5 13.a3 a5 gives Black an initiative, but Nakamura prefers to keep the tension in the centre.


12.Ng5 Nxg3 13.Bxh7+ Kh8 14.Nxf7+ Rxf7 15.Bg6 fails to 15...Rf5 16.hxg3 Rg5.

12...Rc8 13.a3 Nxg3 14.hxg3 f5


Perhaps White should have blocked the enemy bishop pair with 15.exf6 Bxf6 16.Rc1 cxd4 17.c4!. Then 17...dxc4
18.Bxc4 Na5 19.Bd3! would be fine for White so Black should stay with something like 17...g6.
Kamskys decision puts him on the defensive.

15...bxc5 16.b4 g5 17.Nh2 d4

Nakamura is a bit impatient and overaggressive. He could have waited for a better moment to open the centre. Now
White gets counterplay.

18.b5 dxc3 19.bxc6 cxd2 20.Re2 Bxc6?! (20...Be8 21.Rd2 Qc7) 21.Rxd2 (21.Bc4) 21...Bd5


Apparently both players evaluated the position after 22.Bc4 Bxc4! 23.Rxd8 Rfxd8 in Blacks favour although a
computer would probably keep the balance. OTB, Blacks passer, supported by 4 long-range pieces, does look
daunting. After the text, however, Nakamuras advantage is beyond any doubt. Obviously, he wanted so much to show
off, that he found a way to sack his queen on move 55 and it cost him a half point.
22...c4 23.Bc2 f4 24.Qe2 Qa5 25.Nf3 Rcd8 26.Rdd1 Rd7 27.a4 Rfd8 28.Rab1 a6 29.Rdc1 Ba3 30.Rd1 h6 31.Kh2 Bf8
32.Rdc1 Qc5 33.Kg1 Qc6 34.Rd1 Bg7 35.Nd4 Qc7 36.Nf3 Rb8 37.Rxb8+ Qxb8 38.Be4 Qc8 39.Bc2 Qc6 40.Nd4
Qc7 41.Nf3 Rd8 42.Bb1 Rb8 43.Bc2 a5 44.Rc1 Qb7 45.Qd1 Qa8 46.Qe2 Rb4 47.Rd1 Qc6 48.Nd4 Qc7 49.Re1 Qb7
50.Rd1 Bf8 51.Nb5 Bxg2 52.Rd8 Bh1 53.Qf1 Rb2 54.Nd6 Qd5 55.Nf7

Apparently, Nakamura missed 57.Qb1. Instead, 55...Qc6 was winning.

56.Rxd5 Bxd5 57.Qb1 Kxf7 58.Qxc2 Kg7 59.Qd2 Bb4 60.Qd4 c3 Draw.

40. Jussupow 2605 Delchev 2661

Nice 06.05.2006

1.d4 Nf6 2.Nf3 d5 3.e3 c5 4.c3 e6 5.Nbd2 Nbd7 6.Bd3 b6 7.Ne5 Bb7 8.f4 g6 9.0-0 Bg7

White has scored only 27% in 22 games in this position. Of course I would not claim that Black is better, but lets say
that he has a clear plan to follow. He will contend the e5-square by ...Ne8, ...f6 or ...Nxe5, ...f6.


Jussupov has been for decades one of the main protagonists of the Stonewall with Black so he knows perfectly all the
plans and nuances. With his last move he opts for the old treatment of the Stonewall. White displays activity on the
kingside, but forgets his dark-squared bishop.
The modern approach is to fianchetto it on b2 with b3, Bb2, a4/c4. Then White will be ready for fighting all over the
board. For instance:
10.b3 0-0 11.Bb2 Rc8 (I see no logic in simplifications with 11...Ne4) 12.Qe2 Ne8

13.a4 is also possible, but it gives us a free hand in the centre. We should ignore the threat a4-a5 and complete our
regrouping with 13...Nd6 (or 13...f6 first) 14.a5 f6 15.Ng4 e5!.
13...cxd4 14.exd4 dxc4 15.bxc4 Nd6 16.Rac1 Qc7. The hanging pawns are potentially weak if White fails to move them
forward. I do not see how he could do that in the current position.

In Comas Fabrego-Illescas Cordoba, La Massana 2012, White tried to combine both plans: 10.b3 0-0 11.Qf3 Rc8
12.Bb2 Rc7 (protecting the bishop to avoid c3-c4, but the same result could be achieved by 12...Ne8 13.c4?! cxd4
14.exd4 dxc4) 13.Rac1 Ne8 14.Qh3 Nd6 15.c4. Now the best continuation is 15...cxd4 16.exd4 Nxe5 17.fxe5 Nf5
(hitting d4 and intending 18.g4 Nxd4 19.Bxd4 dxc4).

10...0-0 11.g4?

This is a strategic mistake. If White did not want to play b3, he could have resorted to 11.b4 Qc7 12.a4= when 12...cxd4
13.cxd4 Qc3 does not gain the b4 pawn due to 14.Ra3! Qxb4? 15.Rb3 winning material.

11...Nxe5 12.fxe5 Nd7 13.Qg3

Even if it were White to move and he put in g4-g5, Black would have been slightly better after ...Nd7-b8, ...Ba6.

13...f6! 14.exf6 Nxf6

14...Rxf6 15.Nf3 (15.Rxf6 Nxf6 16.Nf3 Ne4 17.Qg2 Qd7) 15...e5 was even more direct.

15.b3 Qe7

Interesting options were 15...Qb8 16.Qg2 e5 and 15...e5 16.Bb2 exd4 with some edge in both lines.

16.Bb2 e5

I have achieved the ultimate goal against the Stonewal to break through the wall with ...e5. Now I have more space
and a better dark-squared bishop. Unfortunately, my next move lets some of the advantage slip away.

17.Rae1 Qe6 (17...Rae8!) 18.h3 Rae8 19.Qxe5

19.c4 exd4 20.exd4 Qc8 is dangerous for Whites king since the h1-a8 diagonal will soon open.

19...Qxe5 20.dxe5 Rxe5 21.c4 Re7 22.g5 Nd7

Another inaccuracy. The knight would stand better on h5.

23.Bxg7 Rxf1+?

After 23...Kxg7! 24.Rxf8 Kxf8 25.cxd5 Bxd5 26.h4 Ne5, I would retain a lasting pull with no risk. I simply missed the
idea of Bd3-f1-g2.

24.Bxf1 Kxg7 25.Bg2 Re5 26.h4

White is already slightly better. Although the rest of the game was quite interesting, it is of no importance to the
opening. I had to continue here 26...Nb8 27.Kf2 Na6 28.a3 Nc7.

26...h6 27.Nf3 Re7 28.Nh2 hxg5 29.cxd5 Ne5

The only defence was 29...Nf8 30.hxg5 Re5 31.e4 (31.Rd1 Rxe3) 31...Bxd5 32.Ng4 Re7.

30.hxg5 c4 31.bxc4 Nxc4 32.Ng4?! (32.Rc1!) 32...Bc8 33.Nf6 Rxe3?

I missed 33...Re5= 34.Rc1 b5 35.d6 Rxg5 36.d7 Bxd7 37.Nxd7 Nxe3.

34.Rc1 b5 35.a4?! a6 36.axb5 axb5 37.Ra1 Nd6 38.Ra7+ Kf8 39.Kf2 Re7 40.Ra8 Kf7 41.Ne4 Nxe4+ Draw.

Index of Variations