Better not touch the Dutch: How to demolish the Stonewall

Breaking the stonewall is an achievement in itself. But doing so while improving on what opening guru Boris Avrukh suggests as best play for White in his otherwise great “1. d4 Grandmaster Repertoire” series feels even better.

Also the middlegame is instructive. Instead of going for his obvious (and only) pawn break immediately White creates a second front first, ties down the opponents’ pieces on the other side of the board and only then goes for the break.

Hero (2.222) – Villain (2.121)
Correspondence Game Lechenicher Schachserver 2016
Dutch (Stonewall)


1. d4 f5


The Dutch Defense doesn’t deserve its shady reputation, depending on which system Black is heading for. The modern interpretation of the Stonewall for instance may even be one of the top notch defenses against 1.d4, regularly employed by Magnus Carlsen among others. The Leningrad, while being structurally somewhat dubious, is far from being refuted.

The reason why I wouldn’t touch the Dutch (nice rhyme, isn’t it?) in OTB chess is the enourmous number of sidelines White may choose beginning at move 2, each of them venomous. In this position for instance, the lines connected to 2.Bg5 and 2.Nc3 deserve attention. 2.e4!? and 2.g4!? may not lead to a white advantage, but you don’t want to navigate the resulting complications unprepared. And there’s more. Just look at Pähtz-Nakamura, played on the Isle of Man in late 2016, in which a 2.470 WGM outplayed Hikaru in one of these sidelines until the rating difference started to kick in.

2. g3

White wants to play main lines. He will delay the development of the Ng1 as long as possible since against both the Stonewall and the Leningrad the option Ng1-h3 may often be advantageous.

2… Nf6 3. Bg2 e6 4. c4


4… c6!

The “correct” move order to get a proper stonewall, delaying …d5 as long as White delays the development of the Ng1.

(4… d5 5. Nh3 is pleasant for White. Against …Bd6 he will go Bf4, often transfer the h3 knight via f4 to d3, shuffle the b1 knight to f3 and enjoy a beautiful grip on the crucial e5 square.)


5. Nf3

(5. Nh3 d6! is fine for Black.)

(5. Nc3 d5 6. Nh3 is a possibility, but c3 is not a square where the Nb1 usually shines in this structure. Often White keeps it on b1 in order to preserve the option of b2-b3 and Bc1-a3, exchanging the “bad” bishop. Another white setup is Nb1-d2, combined with b2-b3 and Bc1-b2 (compare Caruana-Carlsen below). At this point I wasn’t sure yet which setup to go for, and Nc3 just looks shady, so I finally touched and moved the Ng1.)

5… d5 6. O-O Bd6


The tabiya of the Dutch Stonewall. White now has several options. After some research I decided to try a system I had never looked into, recommended by opening guru Boris Avrukh in his grandmaster repertoire books. The lines and evaluations Avrukh gives seemed somewhat flawed to me (at least for correspondence requirements), but especially against what is supposed to be Black’s most beefy setup I found something promising. Fortunately my opponent played right into it.

7. Nc3

(7. b3 The most flexible and by far the main move. 7… Qe7 prevents White from going Ba3. 8. Bb2 (8. a4 insists on playing Ba3, not a bad try.) 8… b6 Back in the day Black preferred to castle quickly, then free the Bc8 via …Bd7-e8-h5. Magnus shows how the Stonewall is handled in this day and age. 9. Ne5 Bb7 10. Nd2 O-O 11. Rc1 a5 Still a highly theoretical position that has been reached hundreds of times. Black is fine, I suspect, as he was in Caruana, Fabiano – Carlsen, Magnus 0-1 (52 moves), Gashimov Memorial 2015.)

7… O-O 8. Qc2 Ne4 9. Rb1


b2-b4-b5 coming with (hopefully) some queenside initiative, the idea of White’s setup and the only way to justify the somewhat crude Nb1-c3.

9… Nd7 10. b4 b5


A radical way to stop the white queenside advance. In OTB chess this hasn’t been tried on the 2.600+ level, but in correspondence it is the main move, so I wasn’t caught by surprise. If Black can prove that he hasn’t overextended and can prevent White from
intruding he may be fine.

11. cxb5 Nxc3 12. Qxc3 cxb5


13. Bg5

Leads to a slight, but lasting White advantage in a static position. I wasn’t overly optimistic about my winning chances at this point, but I was sure that I’d be able to press for a while at least.

(13. Qc6 Avrukh suggests to take the queen fishing in the black camp and then sacrifice it, but objectively this isn’t anything. After for instance 13… Qb6 14. Qxa8 Bb8 15. Ng5 Bb7 16. Qxb7 Qxb7 17. Nxe6 Re8 18. Nc5 Qc6 things are very much unclear. Against a guy who can consult an engine this isn’t winnable while in OTB chess it may go either way.)

13… Qb6 14. Rfc1 Bb7 15. Qe3 Rfe8 16. Bf4 Bxf4 17. Qxf4!


Avrukh only gives 17.gxf4 and states that there’s not much going on, probably correctly so. But after 17.Qxf4! instead it seems to me that the white advantage is quite obvious. Is this advantage substantial or just symbolic due to the static nature of the position? That’s the question to be answered.

17… Rac8 18. e3


18… Qd8 


(In an earlier correspondence game White nurtured a nice advantage for quite some time, but couldn’t break through in the end: 18… Rxc1+ 19. Rxc1 Rc8 20. Rxc8+ Bxc8 21. Nd2 Qb8 22. Qh4 Nf6 23. Nb3 Qd6 24. Nc5 h6 25. Bf1 a6 26. Bd3 e5 27. dxe5 Qxe5, Alfredo – Pannullo, Umberto 1/2-1/2 (61 moves), Champion’s League email, 2010.)

19. Bf1 a6 20. Ne1


White needs to regroup his interfering light pieces, easier said than done. At this point I was dreaming to go Nd3, Ra1, threaten a4, have him go Nb6, install a knight on c5 and take it from there.

20… Nb6 21. Rxc8 Bxc8 22. Rc1 Nc4 23. Nd3


23… Rf8

(23… g5 immediately is a serious option, and probably Black should have opted for it.The move is less about attacking on the kingside and more about preventing White from developing a grip on both wings. At this point it had started to dawn on me that with a2-a4 as the only break available White will have trouble to seriously shake the black position. White needs to create a second frontline.)

24. h4!


The move itself is not that sophisticated, but I like to award me an exclamation mark here since the engine doesn’t see a necessity to prevent …g5/create a second frontline at all. After 24…h6 25.h5 White has to evaluate the fact that his h pawn looks quite isolated and may turn out to be a weakness in an endgame. But before any endgames, and that’s the second and much less obvious point of 24.h4, White has created himself a second possible break for the middlegame: g3-g4 which at the moment looks like an absurd plan but will become quite real soon.

24… h6 25. h5 Bd7 26. Nc5 Qc8 27. Be2 Be8


At first glance the structure suggests that it should be Black who can develop something on the kingside, but it is in fact the other way around.

28. g4!


Only this way White can crack the black fortress: First tie his pieces down on the kingside and the e6 pawn, then go for the queenside break.

28… fxg4 29. Qxg4 Rf5


Plans to free himself with 30…e5 and unclear play if White doesn’t prevent it. The trick is that after dxe5 there’s …Nxe3, and fxe3 fails due to the discovery …Rf1+.

30. Qh4 Bf7

Bg4 was a threat.

(30… e5 31. Bg4)

31. Bg4 Rf6 32. Ra1!


Step two: Only now that the black forces are tied down and discoordinated White goes for the break on the other wing.



Tricky response, setting up a nice trap should White go for a2-a4 prematurely (or play 33.Nxa6). It’s the same motif he already used with 29…Rf5.

33. Kg2!

Stepping out of the nasty …Rf1+ motif.

(33. a4? bxa4 34. Rxa4 Nxe3 and Black is better due to 35. fxe3 Rf1+)

33… Qe7


Let’s the pawn a6 be en pris. Black would be happy to sacrifice it.

34. a4

(34. Nxa6 e5 35. Nc5 exd4 with counterplay)

34… Be8 35. axb5


35… Bxb5

(35… axb5? 36. Ra8 +- Opening the a file is deadly, Black is running out of useful moves.)

36. Nxa6 Qa7

The pin doesn’t help Black, White is in time to recoordinate.

37. Qh1


37… Qf7

(37… Bxa6 38. b5 +-)


38. Nc5!

White gladly gives the f2 pawn in order to reinforce the grip on e6 and open the a file.

(38. Qe1? Rf5! and suddenly the position is unclear again. The weakened white squares around the white king guarantee Black a perpetual at least, should White take on f5.)

38… Rxf2+


39. Kg1

(39. Kg3 would be much worse since now 39…Rf6 holds, while after 39.Kg1 Rf6 White has 40.Qh3 winning (e3 is protected, after 39.Kg3 it is not).)


Black resigned, somewhat prematurely in my opinion. He is in deep trouble, though, for instance after  39… Kh7 40. Bxe6 Qf6 41.Ne4! it’s unlikely Black can save himself.


3.5 2 votes
Article Rating
Benachrichtige mich bei

Most Voted
Newest Oldest
Inline Feedbacks
View all comments

[…] opening books don’t meet correspondence requirements (including the Avrukh bibles as you can see here), this one does. It has introductional value of course, but is also so deep it touches the edges of […]


[…] Partie hatten wir vor einigen Wochen in einem Essay über die Holländische Eröffnung, speziell den Stonewall, erwähnt, aber nicht näher beleuchtet. Sie diente uns als Beispiel für […]


[…] außerdem Meisterpartien deutscher Spieler. Dazu mehrere Eröffnungsübersichten, kommentierte Fernpartien und Partien zwischen Engines. Einige dieser Werke haben auch unsere Freunde von ChessBase […]


[…] finden hier ein kleines Holländisch-Essay und eine Idee, wie sich der Stonewall knacken lässt: „How to […]