Ioannis Simeonidis

Carlsen’s Neo-Møller
A Complete and Surprising Repertoire
Against the Ruy Lopez

New In Chess 2020

Contents
Explanation of symbols . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

Part I Main line: 3...a6 4.♗a4 ♘f6 5.0-0 ♗c5 6.c3 0-0 7.d4 ♗a7 . 15
Chapter 1 White’s critical reply: 8.dxe5 . . . . . . . . . . . . . . . . . . . . . 16
Chapter 2 Refutation refuted: 8.♗g5 exd4 9.cxd4 . . . . . . . . . . . . 27
Chapter 3 The 9.♕c1 variation: 8.♗g5 exd4 9.♕c1 . . . . . . . . . . . . 35
Chapter 4 White maintains the pin: 8.♗g5 exd4 9.e5 h6 10.♗h4 . 39
Chapter 5 White’s extra option: 8.♗g5 exd4 9.e5 h6 10.♗xf6 . . 45
Chapter 6 The ♖e1 variation: 8.♗g5 exd4 9.♖e1 . . . . . . . . . . . . . . 51
Chapter 7 Eighth move alternatives for White . . . . . . . . . . . . . . . 58

Part II Sidelines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
Chapter 8 Minor sixth move alternatives for White . . . . . . . . . . 74
Chapter 9 The pseudo-sac 6.♘xe5: 6...♘xe5 7.d4 b5 8.♗b3 . . . . . 82
Chapter 10 The pseudo-sac 6.♘xe5: 6...♘xe5 7.d4 b5 8.dxe5 . . . . 85
Chapter 11 The ‘Anti-Berlin’: 6.♗xc6 dxc6 7.d3 . . . . . . . . . . . . . . . 89
Chapter 12 Transpositions: 6.c3 0-0 . . . . . . . . . . . . . . . . . . . . . . . . . 93
Chapter 13 A serious system: 5.d3 ♗c5 6.c3 . . . . . . . . . . . . . . . . . . 98
Chapter 14 Fifth move alternatives . . . . . . . . . . . . . . . . . . . . . . . . . 106

Part III Completing the repertoire . . . . . . . . . . . . . . . . . . . . . . . . . 113

Chapter 15 The Exchange Variation: 4.♗xc6 . . . . . . . . . . . . . . . . . 114

Part IV Training material . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127

Chapter 16 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
Chapter 17 Solutions to exercises . . . . . . . . . . . . . . . . . . . . . . . . . . 136

Afterword . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149
Index of variations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150
Index of players . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152
Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157

5
Preface
In every opening there is a position that demonstrates the ideal set-up for
both sides. Not all variations lead to this position. There are only a few
for each side. Adopting the ideal approach means for me that the system I
choose to play will be one of the few systems in that opening that can lead
to an ideal position from my point of view, and therefore I will reject the
other moves as a choice for my opening repertoire.
Of course, that does not mean that I intend to go for dubious moves
that only contain traps and have surprise value but a poor positional basis.
The idea is to go for the objectively best moves. I was never a fan of either
modest or dubious systems.
But what are the criteria for such an ideal approach? The moves have
to be logical. The system has to obey the golden rules of opening play:
occupation of the centre and quick development. No concessions! I
refuse to lose tempi and will go for the most logical and ambitious set-up,
having the ideal position of the system in mind. Following the plan with
a maximum of accuracy is the objective. Part of the approach is adopting
systems in which the most natural human response is not the best.

A perfect example of following this approach as White is the Gligoric

Variation of the King’s Indian Defence: 1.d4 ♘f6 2.c4 g6 3.♘c3 ♗g7 4.e4 d6
5.♗e2 0-0 6.♘f3 e5.

TsLd.tM_
jJj._JlJ
._.j.sJ_
_._.j._.
._IiI_._
_.n._N_.
Ii._BiIi
r.bQk._R
After 7.♗e3 (the Gligoric Variation) 7...♘c6?! 8.d5 ♘e7 9.♘d2 White
achieves the ideal set-up. First 7.d5 also gives White the opportunity
to achieve the ideal set-up later on. But 7.0-0 does not! Now White has
already castled kingside and after 7...♘c6 8.d5 ♘e7, 9.♗e3 runs into 9...♘g4.
Now 10.♗g5 is not with tempo and Black can play 10...f5. That means that

7
Carlsen’s Neo-Møller

I usually play the Gligoric Variation and the Petrosian Variation (7.d5) and
not the main line with 7.0-0 against the King’s Indian Defence.

T_Ld.tM_
jJj.sJlJ
._.j.sJ_
_._Ij._.
._I_I_._
_.n.b._.
Ii.nBiIi
r._Qk._R
This is the ideal position for White. All of his pieces are actively placed
and looking in different directions, and he has various possible pawn
pushes to try to conquer more space, for example 9...♘d7 10.b4 or 9...c5
10.g4.

My aim in this book is to achieve such an ideal position with black in the
Ruy Lopez, where my pawns occupy a sufficient part of the centre and my
pieces are active or at least potentially so. Of course, this is much more
difficult to achieve for the second player than for the first player! But in
Carlsen’s Neo-Møller Variation, the subject of this book, I believe I have
found a way for Black to achieve this against the Ruy Lopez, and as you
will see this approach stands up to analysis.

Ioannis Simeonidis
Athens, November 2020

8
Introduction
The Spanish Torture
1.e4 e5 2.♘f3 ♘c6 3.♗b5!

T_LdMlSt
jJjJ_JjJ
._S_._._
_B_.j._.
._._I_._
_._._N_.
IiIi.iIi
rNbQk._R
The notorious ‘Spanish Torture’ means that Black has to defend a slightly
worse but solid position for a very long time without having the chance to
break free by simplifying the position with exchanges.
In the 1930s, according to Savielly Tartakower, having to defend the
black side of the Ruy Lopez was a torture. In the 1960s for Bobby Fischer it
was like milking a cow. His favourite game was to torture Black in the Ruy
Lopez. Bent Larsen suggested that the Open Variation (3...a6 4.♗a4 ♘f6
5.0-0 ♘xe4) was the correct way to handle the Ruy Lopez with black. Later
on, Anatoly Karpov tortured his opponents with both colours!
At the time, no top player would play any kind of system with ...♗c5
against the Ruy Lopez. In the 1990s, Garry Kasparov was the first World
Champion who had to face these ...♗c5 systems. Thanks to the games of
Vladimir Malaniuk, the Arkhangelsk Variation (3...a6 4.♗a4 ♘f6 5.0-0
b5 6.♗b3 ♗b7) became popular, and later on, in 1994, Vladislav Tkachiev
started using the Neo-Arkhangelsk, where Black plays 6...♗c5 instead
of 6...♗b7 and doesn’t fianchetto the queen’s bishop. Suddenly all these
systems started to develop, with top players like Viswanathan Anand,
Alexei Shirov, Michael Adams and Alexander Onischuk leading the way.
In 1995, Anand used the Neo-Arkhangelsk in his World Championship
Match against Garry Kasparov. In 1996 the Møller move-order – that is,
playing 5...♗c5 immediately without the inclusion of the moves 5...b5
and 6.♗b3 – started to be used as a way to reach the Neo-Arkhangelsk
by players like Vasily Ivanchuk, Shirov, Malaniuk and Onischuk. The
evolution had begun!

9
Carlsen’s Neo-Møller

3...a6 4.♗a4 ♘f6 5.0-0 ♗c5

T_LdM_.t
_JjJ_JjJ
J_S_.s._
_.l.j._.
B_._I_._
_._._N_.
IiIi.iIi
rNbQ_Rk.
The ...♗c5 complex in the Ruy Lopez
It is in the nature of the Ruy Lopez for White to seize the centre by
playing c2-c3 followed by d2-d4. For Black, placing the dark-squared
bishop on c5 against the Ruy Lopez was always controversial, as White
can now play c2-c3 and d2-d4 with tempo. Can Black afford playing the
ambitious ...♗c5 at any stage in the Ruy Lopez without ending up in an
inferior position?

Applying the ideal approach with black

In the Møller, compared to the Neo-Arkhangelsk, Black has not
committed to ...b7-b5 yet. Black has made no concessions and is getting
ready for the ideal set-up with ...0-0 and ...d7-d6. All the black pieces stand
on the best possible squares while he hasn’t committed to the weakening
move ...b7-b5, which only improves the placement of White’s light-squared
bishop.
Still, White has played his bishop to b5 to threaten ♗xc6 at certain
moments. Can White punish Black for ignoring this? White can also try to
punish Black with ♗g5 ideas, exploiting the absence of the dark-squared
bishop on e7.
In the closed lines of the Ruy Lopez, White usually manages to take
control in the centre and keep the tension. This is a price Black has to pay
for his temporary king safety and easy plans. Black will have to submit to
the Spanish Torture later on.
In the Møller Variation, White will manage to take the centre too, but
will not be able to keep it! Black applies huge pressure on White’s centre
and has concrete ways to break free and release the tension. The system’s
uniqueness is based on the unconventional type of play that arises. It
is one of the sharpest and most principled systems, based on concrete
tactics but good and solid, and so far undetected by the majority of the
chess world. It’s a new attempt to stop the Spanish Torture once and for

10
 Introduction

all, based on the ideal opening approach. If Black makes no concessions,

White will not manage to get an advantage and torture him.
The only question is whether the black set-up can be refuted. What
is the price Black has to pay for all this? He has to enter a series of
complications without fear or prejudice and play ...h7-h6 and ...g7-g5 at the
right moment despite the fact that he has already castled kingside. The
most amazing part is that everything works, which confirms that this
natural set-up has a solid positional base and deserves its place among the
most serious and most reliable set-ups against the Ruy Lopez.
Nowadays, 20 years later, the young generation, Magnus Carlsen,
Fabiano Caruana et al, have adopted the ...♗c5 complex against the Ruy
Lopez as an active way of playing, cutting the Gordian knot with a sword.
Pure Møller set-ups have long been a rare guest in top-level events. This
changed in 2018 when Carlsen adopted it against Sergey Karjakin. Up till
then, only Tamir Nabaty had used the system regularly and Onischuk had
played it in just a few games. Soon after, Laurent Fressinet, Igor Lysyj and
Antoaneta Stefanova started following Carlsen’s move-order.
But the development of this system has just started. Caissa still holds
many secrets from us. As Bobby Fischer said, ‘Chess is the search for truth’
and this is exactly what I have tried to do in this book: to find out the
truth about the Møller Variation, or better: Magnus Carlsen’s Neo-Møller!
It is Black’s most uncompromising system against the Ruy Lopez.
My intention is to provide a repertoire for Black after 1.e4 e5 2.♘f3 ♘c6
3.♗b5 a6 4.♗a4 ♘f6, with the general idea to place the black bishop on c5
on the next move, refraining from ...b7-b5 unless it is necessary. We also
cover the Exchange Variation, to provide a complete repertoire for Black
against the Ruy Lopez.
Up to now the Møller was used as a clever move-order to transpose to
a more favourable version of the Neo-Arkhangelsk. After 5.0-0 ♗c5 6.c3,
6...b5 was considered practically the only move for Black – other moves
were considered dubious. The originator of this line, Anders Møller,
played 6...♗a7!? 7.d4! ♘xe4?! in 1902. This does not equalize, but I guess
the correct move, 7...0-0, looked scary to him in view of 8.♗g5!. However,
against the main line 5.0-0, the basic position in this repertoire occurs
after 5...♗c5 6.c3 0-0 7.d4 ♗a7 (which after 8.♗g5 can transpose to the
‘scary’ line above). Black’s seventh move is our alternative to 7...♗b6,
which has also been played. We want to keep the bishop safer and retain
the opportunity to play ...b7-b5, although, as we will see, it is often wise to
refrain from this push! The Møller Variation is characterized by the moves
5...♗c5 6.c3 b5, so I think it is appropriate to call the line 5...♗c5 6.c3 0-0
7.d4 ♗a7 the Neo-Møller!

11
Carlsen’s Neo-Møller

In the year 2018, Carlsen deployed the Neo-Møller as Black against

Karjakin – with success. This game motivated me to take a fresh look into
this position and create this repertoire.
My interest in this line began 30 years ago, in the early 1990s, when I
was looking for ...♗c5 systems against the Ruy Lopez from the black side.
In the old days, all we had was a chessboard, pen and paper. No engines,
just games on paper. It was very hard to analyse these sharp positions that
looked losing from a human perspective. Also, you would have to be a very
strong human player.
All these ...♗c5 systems against the Ruy Lopez were a mystery to me.
Even later, when chess programs came to assist us in analysing such
positions, the result was poor. The positions were too complicated even for
computers. Again, not a lot of progress was made in many systems.
Only lately, in the past three years, the computing power has reached a
level where sophisticated programs like Houdini, Komodo and Stockfish
are able to come much closer to the truth than before. The recent release
of Stockfish 12 with neural networks embedded was another big step
forward for chess. Now we can shed more light on all these difficult
positions, and come closer to the truth.
Of course there are still many positions that can’t be properly evaluated
by engines. This problem will not be solved any time soon, since it is
impossible to create the correct algorithm that enables the engine to
find the best moves in any given position. But computing power will
continue to increase, so more and more difficult positions will be evaluated
correctly. In the opening it will always be a problem for engines to evaluate
positions correctly. There are simply too many pieces on the board!
At present, tournament players face a much more demanding situation
compared to the past. In the old days you could learn a dubious opening or
gambit and be successful with it for a long time. Now, on the next day (or
even immediately after the game with a mobile phone) you can find the
refutation of such a gambit in a few minutes. Therefore, investing time to
learn a dubious opening is no longer a good idea. Soundness has become
an issue. This is the reason why especially in master games we see all those
Berlins and Marshalls. It seems that correct openings are more fashionable
nowadays. No one wants to get into a worse position from the opening.
Another big problem is the fact that there is too much information. Sure,
you can find all these games, but which game is the correct one to follow?

How to read this book

The main body of this book consists of two parts with seven chapters
each. In the first part, we discuss the main line of the Neo-Møller with

12
 Introduction

6.c3 0-0 7.d4 ♗a7. The second part contains all the deviations for White
on the 5th and 6th moves. In the third part, we deal with the Exchange
Variation from Black’s perspective. The fourth part contains Exercises and
Solutions to these exercises, which will hopefully serve to help the reader
digest the material.

For club players

Against the moves 8.♗g5 and 8.dxe5, Black doesn’t have much choice. The
lines are sharp here, and you have to work on the positions in order to
perform well in tournaments. Against White’s other efforts, for example
8.♖e1, I suggest, whenever it is possible, to play one of two lines for Black,
both of which are equally good. Those who play the Arkhangelsk and Neo-
Arkhangelsk or the Møller with 6...b5 will feel at home in this repertoire.
Berlin and Marshall players will have an easy time to adjust as well. There
are many transpositions from the Berlin, especially in the 5.d3 systems,
and in many positions we have Marshall-type play where Black is a pawn
down but has very active pieces.
The critical moments for the club player are those where he has to make
the following decisions:
1) When to play ...b7-b5 ?
2) When to play ...h7-h6 ?
3) When to play ...g7-g5 ?
4) When to play ...exd4 ?
5) When to play ...0-0 or ...d7-d6 or ...d7-d5 ?
6) When to play ...dxc6 or ...bxc6 ?
You don’t have to learn both repertoires in the chapters in which two
reper­toires are suggested. I suggest you go with the one that looks easier to
you.

For correspondence players

You can trust the analysis. As for computer games, it is possible to improve
after a certain point. You can also follow human games, but here in general
you should look for improvements early on in the game.

How the book was made

Until two years ago, there was no path to equality available to Black in the
...♗c5 complex. Everything changed in 2018 when Carlsen took the 8.dxe5
line out of business in his game against Karjakin. This game inspired me
to enter a deep search into the 8.♗g5 line, looking for a way for Black to
equalize, in order to create a repertoire based on the general idea of not
playing ...b7-b5 if possible.

13
Carlsen’s Neo-Møller

I mostly used Houdini 6.03 as an analysis engine, but later I checked

my analyses with Komodo, Stockfish and Stockfish derivatives as well.
Difficult positions where checked in the ChessBase engine cloud with
the best available machines. Also I would like to thank my Engine Room
friends Stables and Basse1999 for making available to me their Xeon
machines for many hours, so I was able to double-check my analysis.
Not much can be found in chess literature about the Neo-Møller, so I
tried to investigate the position from my own perspective. My number one
priority was to create a sound repertoire that could stand the test of time
and can be used on a regular basis in practical play. It looked strange to me
that a natural move like 5...♗c5 could be dubious. I was sure that there had
to be a path leading to sound and interesting positions.
Yes, it is true that it looks like Black is overextending. This is the reason
why this variation is so sharp. Black is trying to place his pieces on the
best possible squares so that if White plays passively, Black can press for an
advantage. That means that White has to accept the challenge and go for
6.c3 followed by 7.d4 ! But with that, he creates a chaotic position, because
he can’t really hold the centre without making concessions. In the end we
will have tactical fireworks that may lead to a draw.
For me it is amazing that Carlsen’s Neo-Møller is so rich, and I was also
amazed that it has been hidden so long from the public eye, while this is a
sound and reliable system with plenty of unknown theory behind it!

Ioannis Simeonidis
Athens, November 2020

14
CHAPTER 2

Refutation refuted: 8.♗g5 exd4 9.cxd4

1.e4 e5 2.♘f3 ♘c6 3.♗b5 a6 4.♗a4 ♘f6 5.0-0 ♗c5 6.c3 0-0
7.d4 ♗a7 8.♗g5 exd4 9.cxd4

T_Ld.tM_
lJjJ_JjJ
J_S_.s._
_._._.b.
B_.iI_._
_._._N_.
Ii._.iIi
rN_Q_Rk.
In this chapter we will examine White’s most natural response 8.♗g5.
9.cxd4 (or 10.cxd4 if 8...h6 9.♗h4 is inserted) was considered the refutation
of the Neo-Møller for some time. It looked like the best choice for White
until Tamir Nabaty neutralized it with a queen sacrifice!

27
 Part I – Main line: 3...a6 4.♗a4 ♘f6 5.0-0 ♗c5 6.c3 0-0 7.d4 ♗a7

Anton Smirnov 2478 this system was not popular and

Tamir Nabaty 2622 did not have a name. It was just a
Pardubice 2016 (6) risky and suspicious sub-variation
of the Møller.
1.e4 e5 2.♘f3 ♘c6 3.♗b5 a6 The position is too complicated
3...♘f6 4.d3 ♗c5 5.♗g5 d6 6.♘bd2 even for engines and no human
0-0 7.c3 a6 8.♗xc6 bxc6 9.d4 exd4 player wants to take such a risk
10.cxd4 ♗a7 11.0-0 h6 12.♗h4 (Yu unless he knows exactly what he
Yangyi-Wang Hao, Huai’an rapid is doing. But if the human player
2017) transposes to the main game! knows the position is sound and
In Chapter 11 and 13 we will see knows how to counter White’s
other examples in which Black gets threats, the idea of catching the
a similar or favourable version of opponent unprepared in such a
the Berlin Defence after playing the mess is very appealing. A bad move
Møller. can lead to instant disaster!
4.♗a4 ♘f6 5.0-0 ♗c5 6.c3 0-0 7.d4 8...h6
♗a7 8.♗g5! 8...exd4!. We plan to reach the
position from the game via this
T_Ld.tM_ move-order. There is no OTB-game
lJjJ_JjJ with 8...exd4 in the MegaBase!
J_S_.s._ The reason we plan to adopt the
_._.j.b. 8...exd4 move-order is mainly to
B_.iI_._ confront the move ♖e1, which will
_.i._N_. be discussed in Chapter 6. The
Ii._.iIi position that Black will have to
rN_Q_Rk. play after 8...h6 9.♗h4 exd4 10.♖e1
is very difficult to handle. Also
The most principled and most the number of attempts White has
dangerous move, to which I have makes the position impractical and
devoted the next five chapters. difficult to play for Black.
What makes the Neo-Møller so Furthermore, Black has extra
unique is the type of play that arises op­tions against White’s set-ups,
in these positions. Just the fact which you will find in Chapters 3
that Black after kingside castling and 4. White has an extra option as
plays ...h7-h6 and ...g7-g5 is at least we will see in Chapter 5, but that
worrying! Can such a strategy be causes no problems for Black.
effective? We teach beginners to In conclusion, 8...exd4 is the more
do just the opposite: ‘You can go correct and precise move: 9.cxd4
...h7-h6, but don’t weaken your king (for 9.♕c1 see Chapter 3, for 9.e5
with ...g7-g5 !’ It is definitely risky, see Chapter 4 & 5 and for 9.♖e1 see
and this is the main reason why Chapter 6)

28
 Chapter 2 – Refutation refuted: 8.♗g5 exd4 9.cxd4

This is one of the most important

T_Ld.tM_ games in the 8.♗g5 variation. It
lJjJ_JjJ involves a queen sacrifice and
J_S_.s._ precise play is required from
_._._.b. both sides. The position is very
B_.iI_._ dangerous for White too!
_._._N_. 10...d6
Ii._.iIi Black is threatening ...b7-b5 and
rN_Q_Rk. ...♗g4. Also ...g7-g5 is in the air and
the e4-pawn is hanging. White has
analysis diagram
no time to defend against all the
9...h6! 10.♗h4 d6 11.♗xc6 bxc6 threats.
12.♘bd2 (12.♕c1 g5 13.♘xg5 11.♗xc6!
hxg5 14.♕xg5+ ♔h7 15.♖d1 ♖b8 A) 11.♕d3 is a tricky move. White
transposes to the ♕c1 variation, is planning to put his queen on
Chapter 3) 12...♖e8 13.♕a4 ♗d7 g6 later on if Black falls for the
14.e5 g5 transposes to the game trap! 11...g5! 12.♗g3 (12.♘xg5 hxg5
Smirnov-Nabaty. 13.♗xg5 ♘xd4 14.♘c3 c6 15.♖ae1
9.♗h4 exd4 ♔g7 16.♕g3 ♘h5 17.♕h4 f6 18.♗e3
9...d6?! does not work here like in ♖h8 19.♗d1 ♕e8 20.f4 b5 21.♕f2
the 7...♗b6 system: 10.♗xc6 bxc6 c5 22.♘d5 ♕f7 23.g4 ♗b7 24.gxh5
11.dxe5 dxe5 12.♘xe5䩲 Zieher- ♖he8 25.♗xd4 cxd4 26.♔h1 f5
Tabatt, Saarbrücken 2013. 27.♕g2+ ♔h7 28.♗b3 ♖g8 29.♕f3
10.cxd4 ♖ae8 Van der Velde-Visser, cr
2017) and now:
T_Ld.tM_ A1) 12...♘h5!? is an interesting try:
lJjJ_Jj. 13.♘c3
J_S_.s.j
_._._._. T_Ld.tM_
B_.iI_.b lJj._J_.
_._._N_. J_Sj._.j
Ii._.iIi _._._.jS
rN_Q_Rk. B_.iI_._
_.nQ_Nb.
Up to now 10.cxd4 looked like the Ii._.iIi
killer line of the 7...♗a7 system (or r._._Rk.
9.cxd4 if Black uses the 8...exd4
analysis diagram
move-order). Sutovsky’s games
seemed to take the line out of A11) 13...f5! 14.♗xc6 (14.exf5 ♔h8!
business! Nabaty did not give up 15.♘d5 ♗xf5 16.♕c3 ♗e4 17.♗xc6
and came up with a great novelty. bxc6 18.♘b4 c5! (activating the

29
Part I – Main line: 3...a6 4.♗a4 ♘f6 5.0-0 ♗c5 6.c3 0-0 7.d4 ♗a7

a7-bishop!) 19.dxc5+ ♕f6 20.♕xf6+ ready to break White’s centre with

♖xf6 21.♘d2 ♘xg3 22.hxg3 ♗h7 ...c7-c6!) 16.♕e2 c6! 17.♖fd1 cxd5
23.cxd6 cxd6 24.♖ad1 ♖af8 25.♘f3 18.exd5 ♘xd5 19.♘e4 ♕e6 20.♗xd5
♗e4) 14...bxc6 15.♕c4+ ♔h8 ♗xd5 21.♖xd5 ♕xd5 22.♘f6+
16.exf5 ♖b8 17.♕xc6 ♘g7 (Black has gxf6 23.♕g4+ ♔h7 24.♕f5+ ♔g8
more than enough activity for the 25.♕g4+;
pawn) 18.f6 ♘f5 19.♘d5 ♗b7 20.♕c4 C) 11.♖e1 ♗g4! transposes to
♗xd5 21.♕xd5 ♕xf6 22.♖ac1 ♕f7 Chapter 6;
23.♕c6 ♗b6 24.b3 ♖g8 25.♖ce1 ♖bf8 D) The natural 11.♘c3 has a trap
26.♕e4 a5; but after the correct 11...♗g4!
A12) The more direct 13...♘xg3 (11...b5?! does not work here! After
looks OK too: 14.fxg3! (14.hxg3 b5 12.♗c2 ♗g4 (12...g5 13.♘xg5)
15.♗b3 g4 16.e5 ♔g7䩱) 14...♘e5! 13.♘d5 g5 14.♕d3! gxh4 15.e5)
(a nice trick, exploiting the pin! only White can get into trouble:
Exchanging pieces is in Black’s
favour in this position) 15.♕d2 c6!? T_.d.tM_
(with the idea ...♕b6) 16.♔h1 ♘xf3 lJj._Jj.
17.♖xf3 ♗g4 18.♖ff1 ♕b6 19.♗d1 J_Sj.s.j
♗d7. _._._._.
A2) After 12...b5? comes the trap. B_.iI_Lb
The usual ...b7-b5 followed by ...♗g4 _.n._N_.
does not work here: 13.♗b3 ♗g4? Ii._.iIi
r._Q_Rk.
T_.d.tM_
analysis diagram
l.j._J_.
J_Sj.s.j D1) 12.♘d5? g5!;
_J_._.j. D2) 12.♕d3 ♗xf3 13.♗xc6 bxc6
._.iI_L_ 14.gxf3
_B_Q_Nb.
Ii._.iIi T_.d.tM_
rN_._Rk. l.j._Jj.
J_Jj.s.j
analysis diagram
_._._._.
14.e5! dxe5? 15.♕g6+; ._.iI_.b
A3) 12...g4! is a solid choice: _.nQ_I_.
13.♗xc6 bxc6 14.♘h4 ♖b8 15.♘d2 Ii._.i.i
(15.b3?! ♖b4) 15...♖xb2 16.♘c4 r._._Rk.
♖b5 17.♖ae1 ♖e8 18.e5 dxe5 19.♗xe5
analysis diagram
♗e6.
B) 11.d5 ♘e5 12.♘xe5 dxe5 13.♘d2 14...g5 15.♗g3 ♘h5⇆ 16.d5 ♕d7
b5 14.♗b3 ♗b7 15.a4 ♕d6 (Black is (16...f5 17.exf5 ♕f6 18.dxc6 ♕xf5

30
 Chapter 2 – Refutation refuted: 8.♗g5 exd4 9.cxd4

19.♖ad1 ♕xf3 20.♕xf3 ♖xf3 21.♔g2 15...♕b5 (15...♘h5!䩱) 16.♕xb5 axb5䩱

♖af8) 17.dxc6 (17.♔g2 c5 18.e5 dxe5 Alekseev-Nozdrachev, Yaroslavl
19.♗xe5 ♖ae8 20.♗g3 f5) 17...♕h3 2019;
18.♔h1 f5 19.exf5 ♗xf2; B) 12.♕c1 g5 transposes to 9.♕c1,
D3) 12.♗xc6 bxc6 13.♕a4 ♗xf3 Chapter 3: 13.♘xg5 hxg5 14.♕xg5+
14.gxf3 g5 15.♗g3 ♕d7. ♔h7 15.♖d1 ♖b8.
E) 11.h3?! g5!⇆.
11...bxc6 12.♘bd2 T_Ld.tM_
A) 12.♖e1 ♗g4⇆ 13.♕a4 ♗xf3 l.j._Jj.
14.gxf3. J_Jj.s.j
_._._._.
T_.d.tM_ ._.iI_.b
l.j._Jj. _._._N_.
J_Jj.s.j Ii.n.iIi
_._._._. r._Q_Rk.
Q_.iI_.b
_._._I_. White manages to hold everything
Ii._.i.i and 16.e5 is a serious threat:
rN_.r.k. 12...♖e8!
12...g5?! 13.♘xg5! hxg5 14.♗xg5
analysis diagram
♗xd4 15.♘f3 ♗xb2 16.♖b1 ♕d7
Already Black can hope for more 17.♖xb2 (17.e5 ♗xe5 18.♘xe5 dxe5
than equality: 14...♕b8! (14...g5!?N 19.♗xf6 ♕xd1 20.♖bxd1䩲) 17...♘xe4
15.♗g3 h5 (15...♘h5!? 16.♘d2 (16.♘c3 18.♗h6→ Sutovsky-Bartel, Biel
♕f6 17.♖ad1 ♕xf3 18.♖e3 ♕g4 2015.
19.♕xc6 ♘f4) 16...♘g7 17.♖ac1 ♘e6 13.♕a4!
18.d5 cxd5 19.exd5 ♘g7 20.♔g2 ♖e8 13.♖e1 (13.e5 dxe5 14.dxe5 g5⇆) 13...
21.♖xe8+ ♕xe8) 16.h3 ♕d7 17.♔g2 g5 14.♗g3
c5) 15.♘d2 (15.♗xf6 ♕xb2! 16.♘a3
gxf6) T_LdT_M_
l.j._J_.
Td._.tM_ J_Jj.s.j
l.j._Jj. _._._.j.
J_Jj.s.j ._.iI_._
_._._._. _._._Nb.
Q_.iI_.b Ii.n.iIi
_._._I_. r._Qr.k.
Ii.n.i.i analysis diagram
r._.r.k.
14...g4 (14...♘h5!? is the engines’
analysis diagram
choice: 15.♘c4 d5 16.exd5 ♖xe1+

31
Part I – Main line: 3...a6 4.♗a4 ♘f6 5.0-0 ♗c5 6.c3 0-0 7.d4 ♗a7

17.♕xe1 ♘xg3 18.hxg3 cxd5 19.♘ce5 14.e5

♗b6 20.♘c6 ♕d7 21.♘fe5 ♕d6 If 14.♖fe1 c5! and now:
Ivers-Opasnost, playchess.com A) 15.♕c4 cxd4 16.e5 dxe5 17.♘e4
2019) 15.♗h4! gxf3 16.e5 dxe5 ♖e6 18.♘xf6+ gxf6 19.♖xe5 ♖c6
17.♘e4 20.♕d3 ♕f8 21.♖d5 ♖d6䩱;
B) 15.♗xf6 ♗xa4 16.♗xd8 ♖axd8
T_LdT_M_ 17.d5 ♗b5 18.♖ac1 ♗d3 19.b3 (19.♖c3
l.j._J_. c4 20.b3 f5䩱) 19...f5;
J_J_.s.j C) 15.♕xa6 cxd4 16.♕d3 g5
_._.j._. 17.♘xg5 hxg5 18.♗xg5 ♖e5 19.♗h4
._.iN_.b ♕e7 20.♕f3 ♔g7 Huuskonen-
_._._J_. Kuosmanen, cr 2018.
Ii._.iIi
r._Qr.k. T_.dT_M_
l.jL_Jj.
analysis diagram
J_Jj.s.j
17...♘xe4! (like in the main game _._.i._.
Black sacrifices the queen) 18.♗xd8 Q_.i._.b
♘xf2 19.♕xf3 ♖xd8 20.♕xf2 ♗xd4 _._._N_.
Sasikiran-Puranik, Douglas 2019. Ii.n.iIi
r._._Rk.
T_LdT_M_
l.j._Jj. 14...g5 15.♘xg5
J_Jj.s.j 15.♗g3 is not consistent. Now Black
_._._._. is slightly better: 15...♘h5 16.♘e4?
Q_.iI_.b
_._._N_. T_.dT_M_
Ii.n.iIi l.jL_J_.
r._._Rk. J_Jj._.j
_._.i.jS
13...♗d7! Q_.iN_._
Only move! After this game came _._._Nb.
to my attention, I had no doubt Ii._.iIi
that Black’s set-up was sound. If r._._Rk.
an opening system is good then
analysis diagram
somehow everything works!
13...g5?! 14.♘xg5 hxg5 15.♗xg5 16...f5! 17.exd6 fxe4 18.dxc7 ♕c8
♖e6 (15...♖b8 16.♖ae1 ♖e6 17.e5 19.♘e5 ♘xg3 20.hxg3 ♕xc7 Yu
Sutovsky-Onischuk, Poikovsky Yangyi-Wang Hao, Huai’an rapid
2010) 16.f4 ♕f8 17.f5 Sutovsky- 2017.
Nabaty, Gibraltar 2012. 15...hxg5 16.♗xg5 dxe5 17.♘e4

32
 Chapter 2 – Refutation refuted: 8.♗g5 exd4 9.cxd4

B) 20.h3 ♖xe5 21.♕xa6 ♗b6.

T_.dT_M_ These kind of positions are very
l.jL_J_. complicated for both sides and it is
J_J_.s._ very easy to blunder if unprepared:
_._.j.b. 22.♔h2 ♖a5 23.♕e2 ♖d2 24.♕f3
Q_.iN_._ (White has managed to hold onto
_._._._. the f2-pawn, but the b2-pawn
Ii._.iIi will fall and so will the a-pawn)
r._._Rk. 24...♖xb2 25.♖ae1 ♘d6 26.a3 ♖a2
27.♖e7 ♖5xa3 28.♕f4 ♔g7;
17...♘xe4! 18.♗xd8 ♖axd8 19.dxe5 C) 20.♖ac1 is another interesting
♗f5! try: 20...♖xe5 21.♕xa6 ♗b6 22.♖xc6
After 19...♖xe5?! 20.♕xa6 White ♘d2! 23.♖xb6 (only move; 23.♖fc1?
is better since he can take control ♖de8) 23...♘xf1 24.♖h6 ♖e1
with ♖ad1 next. 25.f3 ♗h7 26.♕f6 ♖de8 27.h4 ♖8e2.
Black is threatening 28...♘e3! so
._.tT_M_ White has nothing more than a
l.j._J_. perpetual: 28.♕d8+ ♔g7 29.♕d4+
J_J_._._ ♔f8 30.♕d8+;
_._.iL_. D) 20.♖ae1 ♖xe5 21.♕xa6 ♗b6
Q_._S_._ 22.h3 ♖d2 23.b4 ♗e6 24.♔h2 ♗xf2
_._._._. 25.♖e2 ♗g3+ 26.♔g1 ♔g7! (a cold-
Ii._.iIi blooded move. Now the threat is
r._._Rk. 27...♖d4 or 27...♗d5!) 27.♖xd2 (only
move) 27...♘xd2 28.♖d1 c5! 29.bxc5
Black’s position is not worse. It ♘e4 30.♕a8 ♗f2+ 31.♔h2 ♗xc5
is only fair to say that whoever is 32.a4 ♗d6 33.♔g1 ♗c5+;
better prepared will win. Among E) 20.♕xc6 ♖e6 21.♕xc7 ♗b6
correspondence players, it will end 22.♕c4 ♖g6 (Black has multiple
in a draw. threats) 23.e6 ♖xe6 24.♖ae1 ♖d2
20.♕xa6 (even if you know the position
Grabbing a pawn and at the same is equal if both sides play the
time clearing the way for the best moves, this position looks
a-pawn. uncomfortable for White) 25.h3
A) With 20.g4 White weakens the ♖xb2 26.a4 ♖xf2 27.♔h2 ♔g7 28.a5
f3-square: 20...♗g6 21.♖ad1 ♘d2! ♗c5 29.♖xf2 ♗xf2.
(White has to give up the exchange) 20...♗b6 21.a4! ♖d2
22.♖fe1 ♘f3+ 23.♔f1 ♘xe1 24.♖xd8 21...♘d2 is also OK for Black: 22.a5
♖xd8 25.♔xe1 ♖d4! (the pawn on ♗d3 23.♕b7 ♗d4 24.♖ad1 ♗xf1
f2 is the target) 26.♕xa6 ♗b6 27.h4 25.♖xd2 ♗b5 26.g3 ♖b8 27.♕xc7
♖e4+ 28.♔f1 ♖d4; ♗xe5.

33
CHAPTER 10

The pseudo-sac 6.♘xe5:

6...♘xe5 7.d4 b5 8.dxe5
1.e4 e5 2.♘f3 ♘c6 3.♗b5 a6 4.♗a4 ♘f6 5.0-0 ♗c5 6.♘xe5
♘xe5 7.d4 b5 8.dxe5

T_LdM_.t
_.jJ_JjJ
J_._.s._
_Jl.i._.
B_._I_._
_._._._.
IiI_.iIi
rNbQ_Rk.
In this chapter we are going to examine another sharp attempt for White
to refute the Møller starting with 6.♘xe5. White wins the piece back with
tactics, but as it turns out they rather work in Black’s favour.
I remember working on the game Short-A.Onischuk, Wijk aan Zee 1997.
At the time, chess engines could not see much and no theory existed, so
what was really going on was a mystery to me. Nowadays modern chess
engines provide very accurate evaluations to this kind of positions.

85
 Part II – Sidelines

Steve Berger 2312 it seriously: 10...♘xf2 11.♖xf2

Jan Gustafsson 2554 ♗xf2+ 12.♔xf2 ♕h4+ 13.♔f1 ♗xd5
Böblingen 2002 (9) 14.♕xd5

1.e4 e5 2.♘f3 ♘c6 3.♗b5 a6 4.♗a4 T_._M_.t

♘f6 5.0-0 ♗c5 6.♘xe5 ♘xe5 7.d4 b5 _.jJ_JjJ
8.dxe5 J_._._._
This position can occur only in _J_Qi._.
the Møller and not in the Neo- ._._._.d
Arkhangelsk since there the white _._._._.
bishop is already on b3. IiI_._Ii
8...♘xe4 rNb._K_.
T_LdM_.t analysis diagram

_.jJ_JjJ 14...0-0 15.♘d2 ♕xh2 16.♘f3 ♕h5⇆

J_._._._ 17.♗e3 (17.♗d2 ♕h1+ 18.♘g1 ♕h5
_Jl.i._. 19.♘f3 ♕h1+ 20.♘g1 ♕h5 21.♗c3
B_._S_._ d6 22.a4 ♖ad8 23.axb5 axb5
_._._._. Magnum-Kerveros, playchess.com
IiI_.iIi 2019) 17...d6 18.♔f2 ♖fe8 19.a4 dxe5
rNbQ_Rk. 20.axb5 ♖ad8 21.♕c5 axb5䩱 Sareen-
Stefanova, Moscow 2006;
9.♕d5?! B2) 10.♕h5 ♕e7 (10...0-0!?N)
A) 9.♕g4 ♗b7 10.♕xg7? (10.♗b3 11.♘c3 ♘xc3 12.bxc3 0-0-0 13.♗f4
♕e7) 10...♕h4!; g5 14.♗g3 f5 15.exf6 ♕e4 16.♕f3
B) 9.♗b3 is the main move. After ♕xf3 17.gxf3 ♖hf8 Dzhumaev-
9...♗b7 White has several options Gareyev, Tashkent 2007;
but nothing seems to work! B3) 10.♕g4 ♕e7 11.♘c3 (11.♕xg7?
0-0-0 12.♕xf7 ♕xe5 13.♕f4
T_.dM_.t ♕e8 14.♗e3 ♖f8 15.♕h4 ♗xe3
_LjJ_JjJ 16.fxe3 ♖xf1+ 17.♔xf1 ♕g6
J_._._._ Von Rosenberg-Schmitz, cr 2008)
_Jl.i._. 11...♘xc3 (11...f5 12.exf6 ♘xf6 13.♕g5
._._S_._ 0-0-0 Escobar Forero-Henriquez
_B_._._. Villagra, Linares 2019) 12.bxc3
IiI_.iIi 0-0-0 (12...h5 13.♕f5 g6 14.♕f6
rNbQ_Rk. ♕xf6 15.exf6 d5 16.a4 0-0-0
Pavlicek-Lombart, cr 2012) 13.a4
analysis diagram
(13.♗g5 Kobalia-Belozerov, Tomsk
B1) 10.♗d5 leads to a very unba­ 2004) 13...♕xe5 14.axb5 axb5
lanced position. We have to take Halvax-Rabiega, Graz 2010;

86
 Chapter 10 – The pseudo-sac 6.♘xe5: 6...♘xe5 7.d4 b5 8.dxe5

B4) 10.♘d2 ♕h4 (10...♘xd2 ♘xc5 14.♕c4 ♘e6 Brooker-

11.♕xd2 0-0) 11.♕f3 f5 12.♘xe4 Lorenti, cr 2002.
♗xe4; 10.♕xb7 c6
B5) 10.♘c3 ♕h4!? (10...♘xc3N
would be my choice now. It is a T_.dM_.t
simpler alternative, e.g. 11.bxc3 0-0 _Q_J_JjJ
or 11...♕e7!?) 11.♗xf7+ (11.♕f3 0-0; J_J_._._
11...0-0-0!? could be playable too) _Jl.i._.
B_._S_._
T_._M_.t _._._._.
_LjJ_BjJ IiI_.iIi
J_._._._ rNb._Rk.
_Jl.i._.
._._S_.d White’s queen is trapped! Black can
_.n._._. catch it with ...♖a7.
IiI_.iIi 11.♘c3 ♘xc3 12.bxc3
r.bQ_Rk.
T_.dM_.t
analysis diagram
_Q_J_JjJ
11...♔xf7! (11...♔e7 12.♗e3 ♗xe3 J_J_._._
13.fxe3 ♘xc3 14.bxc3 ♕e4 15.♖f2 _Jl.i._.
♖af8? (15...♕xe5) 16.♕h5 B_._._._
Short-A.Onischuk, Wijk aan Zee _.i._._.
1997) 12.♕xd7+ ♔g8 (12...♕e7?? I_I_.iIi
13.♕f5+ ♔g8 14.♘xe4) 13.♗e3 r.b._Rk.
(13.♕e6+ ♔f8) 13...♕e7.
12...♖a7?!
T_LdM_.t 12...0-0!.
_.jJ_JjJ 13.♗g5 ♕xg5 14.♕b8+ ♕d8
J_._._._ 15.♕xd8+ ♔xd8 16.♗b3 f6 17.exf6
_JlQi._. gxf6 18.a4
B_._S_._
_._._._. ._.m._.t
IiI_.iIi t._J_._J
rNb._Rk. J_J_.j._
_Jl._._.
9...♗b7! I_._._._
The best! _Bi._._.
9...♕h4!? 10.♗e3 c6 11.♕d3 bxa4 ._I_.iIi
12.f4?! (12.♘c3 ♗xe3 13.♘xe4 ♗b6 r._._Rk.
14.♘d6+ ♔f8) 12...♕g4 13.♗xc5

87
Part II – Sidelines

18...d5 24.♖xa7+ ♗xa7 25.♖a1 ♗c5 26.♖a6

Now the ♗b3 is trapped like in the ♖c8 27.g3 ♔e6 28.♔g2 ♔d6 29.♖a1
game Winter-Capablanca, Hastings ♖e8 30.♔f3 ♔c7 31.♖d1 ♖e5 32.g4
1918/19! ♖e4 33.h3 ♗d6 34.♖d3 ♗c5 35.♔g3
19.♖fd1 ♔c7 20.♖e1 ♔d6 21.♖ed1 ♔d6 36.♖f3 ♔e6 37.♖f5 ♖e5 38.♖f3
♔e6 22.♖e1+ ♔f7 23.axb5 axb5 ♗d6 39.♔g2 c5 0-1

Conclusion
9.♕d5 falls into a trap. The shocking 9...♗b7!! kills the line instantly.
White’s queen gets trapped after taking the bishop.
9.♗b3 is the standard move here and after 9...♗b7 White has several
attempts. The position is pretty sharp, so Black must be ready to enter
complications. 10.♗d5 leads to a very unclear position where Black has a
rook and pawns versus two minor pieces, with chances for both sides.

88