Over the

Horizons
Stefan Bcker

Play through and download
the games from
ChessCafe.com in the
DGT Game Viewer.

Nimzowitsch and the Dutch Defense
Aron Nimzowitsch called the Dutch Defense "very modern" in Wiener
Schachzeitung 1913 [1]. In his harsh and long review of Tarrasch's new Die
moderne Schachpartie (1912), he gladly cherished many lines that his
"chosen arch-enemy" (their conflict had already begun in 1904 [3]) dared to
criticize. Siegbert Tarrasch had claimed that the Dutch Defense gave Black a
weak game right from the start ("Die hollndische Partie gibt dem
Nachziehenden von vornherein ein schwaches Spiel.").
But when Simon Alapin asked in WSZ 9-12 (1913) what Nimzowitsch had in
mind against the Staunton Gambit 1 d4 f5 2 e4 fxe4 3 Nc3 Nf6 4 f3,
Nimzowitsch avoided a difficult discussion (a decade before O. H. Krause
found 4Nc6!) with the snippy hint (WSZ 15-18, 1913) that the Dutch can
also arise via 1 d4 e6 2 Nf3 f5. In Copenhagen 1922, he would indeed choose
this classical Dutch treatment against E. Jacobsen. It didn't take long,
however, until Nimzowitsch came into contact with another, just as
controversial version of the Dutch Defense. In Copenhagen 1923, he faced the
surprising opening novelty:
Aaron Nimzowitsch Jrgen Mller
Copenhagen 1923
Dutch Defense [A81]
1 d4 f5 2 c4 Nf6 3 Nc3 d6 4 Nf3 Nc6!

[FEN "r1bqkb1r/ppp1p1pp/2np1n2/5p2/2PP4/
2N2N2/PP2PPPP/R1BQKB1R w kq - 0 5"]
The new idea, developed by the Danish player and theoretician Orla Hermann
Krause. "Obviously Black intends to play e7-e5," writes Nimzowitsch, adding
that here a preparation of the advance by means of Qe7 isn't necessary (his
article [2] had started looking at 1 e4 e6 2 Qe2 c5 3 d3, followed by f4, since
in A. N.'s view the Krause Dutch is a kind of "Chigorin Reversed"),
confirmed by the short variation 5 g3 e5! 6 dxe5 dxe5 7 Qxd8+ Kxd8. In the
resulting position, Nimzowitsch doesn't see an advantage for White.
5 Bf4 h6 6 h4 Ng4
Nimzowitsch [2]: "All this according to Dr. Krause. Willy-nilly now I had to
play:"
7 d5
Purchases from our shop help
keep ChessCafe.com freely
accessible:

Dutch A80-A85
by Boris Schipkov

Killer Dutch
by Simon Williams

Starting Out:
The Dutch Defense
by Neil McDonald

[FEN "r1bqkb1r/ppp1p1p1/2np3p/3P1p2/2P2BnP/
2N2N2/PP2PPP1/R2QKB1R b KQkq - 0 7"]
The "critical position," says Nimzowitsch [2]. However, after Mller's reply
7Nce5? he could have achieved a significant advantage: 8 e4 fxe4 9 Nxe5
dxe5 10 Bg3, analysis by Nimzowitsch [2]. In the game he lost his advantage
by exchanging the bishop: 8 Bxe5 dxe5 9 e4 e6 10 Nh2, but won nevertheless
(1-0, 49).

Orla Hermann Krause (1867-1935)
Here Nimzowitsch's article [2] abruptly jumps into a correspondence game
Nimzowitsch Krause, also from the position above. He doesn't say so, but
it seems probable that the theoretician Dr. Krause was eager to demonstrate
the correctness of his idea and informed the winner of the debut game about
7e5!. So let's follow the improved course, in another duel:
Aaron Nimzowitsch Orla Hermann Krause
Correspondence 1923
Dutch Defense [A81]
7e5!

[FEN "r1bqkb1r/ppp3p1/2np3p/3Ppp2/2P2BnP/
2N2N2/PP2PPP1/R2QKB1R w KQkq e6 0 8"]
A significant improvement upon Mller's move. "Much stronger," admits
Nimzowitsch [2]. He explains why the pawn sacrifice is "playable, even
absolutely correct": 8 dxc6 exf4 9 cxb7 Bxb7 10 Qd4 Qd7!, for example:
(a) 11 Qxf4 Be7 followed by Bf6 and a "formidable set-up," Nimzowitsch [2].
(b) 11 Rd1 (11 0-0-0 "leaves f2 unprotected" [2]) 11a6 12 Rh3 0-0-0,
"already threatening Be7, since Qxg7 would fail to Nxf2 Kxf2 Bh4+ followed
by Qxg7," Nimzowitsch [2].
8 dxe6 Bxe6 9 g3!
According to Nimzowitsch [2] this "extraordinarily interesting pawn
sacrifice" has its point in the strange bishop manoeuvre Bf1-h3-g2 in moves
eleven and twelve. Altogether the chances are roughly balanced, and if we
check the further game with a computer, the next moves are not entirely
forced. But to retain some of the spirit of the old magazine, let's just repeat
Nimzo's exclamation marks:
9Bxc4 10 Nd4 Qd7 11 Bh3!! g6 12 Bg2!!

[FEN "r3kb1r/pppq4/2np2pp/5p2/2bN1BnP/
2N3P1/PP2PPB1/R2QK2R b KQkq - 0 12"]
Forcing g7-g6 is essential, says Nimzowitsch, since after
12d5 13 Nxc6 bxc6 14 b3
the continuation 14Bb4 15 Qd4! attacking Rh8 (thanks to the weakening
g7-g6) would lose a piece, says Nimzowitsch. Which for the software
translates to "150-0-0 16 bxc4 c5 17 Qd3 d4 (winning back the piece,
but) 18 0-0 dxc3 19 Qf3! c6 20 a3 Ba5 21 Rab1 with a mating attack.
14Ba6 15 Rac1 Bc5! 16 0-0 g5! 17 Nxd5!

[FEN "r3k2r/p1pq4/b1p4p/2bN1pp1/5BnP/
1P4P1/P3PPB1/2RQ1RK1 b kq - 0 17"]
17gxf4 18 Rxc5 0-0-0!! 19 Rxc6 Bb7 20 Qc2
20 Rc5! secures an advantage.
20Bxc6 21 Qxc6 Qxc6 22 Ne7+ Kb7 23 Nxc6 Rd2 24 gxf4 Kb6 25 a4
Rxe2 26 b4!
So far the source [2]. The rest of the correspondence game is unknown.
Nimzowitsch's evaluation "approximately equal" seems to be correct. A nice
game, but the theoretical value should not be overestimated. In the last
diagram, for example, Black has the stronger 12Nd8! (an observation of
mine in [5]) which consolidates an extra pawn. One move earlier 11g5!
(instead of 11g6) 12 Bxg4 fxg4 13 hxg5 was a strong alternative, though
overlooked in my earlier article [5], due to the in-between move 13Bg7!.
And there are further good options for Black in moves 10-12.
In the Nordic master tournament Copenhagen 1924 the two Danish players
had another opportunity to test the original set-up. Knowing his opponent's
plan, Nimzowitsch moved his bishop even earlier to the square f4, to prevent
the advance e7-e5:
Aron Nimzowitsch Orla Hermann Krause
Copenhagen 1924
Dutch Defense [A81]
1 d4 f5 2 Nf3 Nf6 3 Bf4 d6
"I believe that here Black can well transpose to the usual treatment e6, b6,
Bb7," remarks Nimzowitsch, hitting the nail on the head. In almost every line
of the Dutch Defense a bishop is miserably placed on f4. Only in the Krause
Variation the bishop f4 makes some sense, since Black's set-up greatly
depends from the success or failure of the advance e7-e5.
4 e3 h6 5 h4 Nc6 6 d5

[FEN "r1bqkb1r/ppp1p1p1/2np1n1p/3P1p2/5B1P/
4PN2/PPP2PP1/RN1QKB1R b KQkq - 0 6"]
6e5!?
According to [2], under these circumstances the sacrifice "may not be quite
correct." White's move-order is too unusual anyway to have much theoretical
relevance. However, it is interesting to compare the move with other cases of
Krause's e7-e5 motif.
7 dxc6 exf4 8 Bb5 b6
! by [2]. But Nimzowitsch doesn't give the precise refutation of Krause's later
idea 8Kf7. Neither 9 exf4 (Krause) nor 9 0-0 (Nimzowitsch [2]) nor 9 Qd4
(Nimzowitsch in Die Praxis meines Systems) really works. The computer
suggests the convincing 9 Nc3! fxe3 10 Ne5+ Kg8 11 Bc4+ Kh7 12 Nf7 with
the strong threat of 13 Ng5+! (+/-).
9 exf4 d5

[FEN "r1bqkb1r/p1p3p1/1pP2n1p/1B1p1p2/5P1P/
5N2/PPP2PP1/RN1QK2R w KQkq - 0 10"]
Black has no real compensation. Considering White's extra pawn on the
kingside, the best continuation now would have been 10 Qe2+ followed by
Ne5 and 0-0-0. Instead he chose to castle short, when his doubled f-pawn
doesn't seem so useful: 10 0-0? Bc5 11 g3 Ne4 12 Kg2. After the natural 12
0-0 Black would still have a reasonable position. But 12g5? 13 Ne5 0-0 14
hxg5 hxg5 15 Qh5 +- was unnecessarily cooperative (1-0, 28).
This game didn't diminish Nimzowitsch's respect for Krause's system. From
Kagans Neueste Schachnachrichten [2]: "We dare to predict that the treatment
of the Dutch Defense launched by Krause (and developed by me) will replace
the common line (f5 and e6) shortly." Which may be a surprise for the
readers. Even the Hort-Antoshin system, a "subtler" way to realize the
immediate advance e7-e5 (by means of c6 and Qc7, instead of Nc6) is more
often seen today. It has become difficult to find information on the Krause
Dutch in recent opening works, while books on the Classical Dutch,
Leningrad or Stonewall abound.
Obviously I am not so optimistic to share Nimzowitsch's conclusion, in the
very last sentence of his article [2], that one day Krause's System might be the
most popular reply against 1 d4. But the other extreme, to ignore it
completely, seems equally ridiculous to me. The Krause Dutch deserves
(some) attention. There is still room for doubt, and for experiments. My
following loss was hardly the last word.
Paul van der Sterren (2535) Stefan Bcker (2341)
Den Bosch 1999 (""Toernooi voor de Toekomst"), round five
Dutch Defense [A04]
1 Nf3 f5 2 g3
In round three of the same tournament, I had already used Krause's system, so
my opponent was warned. To postpone c4 and Nc3 and include the flexible
g3 is probably the most critical line.
2 d4 Nf6 (move orders are important; 2d6 3 Nc3! is well covered by
Richard Palliser in "Don't be Tricked into the Leningrad" [8]) 3 c4 d6.
Reaching an earlier game of mine, by transposition: 4 Nc3 Nc6 5 Bf4 h6 6 d5
e5

[FEN "r1bqkb1r/ppp3p1/2np1n1p/3Ppp2/2P2B2/
2N2N2/PP2PPPP/R2QKB1R w KQkq e6 0 7"]
7 dxe6 Bxe6 8 e3 g5 9 Bg3 Bg7 10 Qd2 0-0 11 0-0-0 a5 12 h4 g4 13 Nd4
Nxd4 14 exd4 d5 15 c5 Ne4 16 Nxe4 dxe4 17 d5 Bf7 18 Bf4 Qf6 19 Be3
Rfd8 20 Bd4 Qxd4 21 Qxd4 Bxd4 22 Rxd4 Rxd5 23 Rxd5 Bxd5 24 Kd2
Bxa2 25 Ke3 Rd8 26 Be2 Bb3 27 Ra1 a4 28 Bb5 Kg7 29 Bxa4 Ra8 30 Ra3
Bxa4 31 b3 f4+ 0-1, Marian Bcker, Detmold 1983.
2Nf6 3 Bg2 d6 4 d4 Nc6 5 Bf4!

[FEN "r1bqkb1r/ppp1p1pp/2np1n2/5p2/3P1B2/
5NP1/PPP1PPBP/RN1QK2R b KQkq - 0 5"]
There are several alternatives, but the text move must be critical.
In round three of the same tournament I was more successful: 5 d5 Nce5 6 0-0
Nxf3+ 7 exf3 (Black isn't worse; Euwe [3] assessed a similar situation as
advantageous for White) 7g6 8 c4 Bg7 9 Nc3 0-0 10 Re1 c6? (10Re8 11
f4 Bd7 12 Qb3 Rb8, about =) 11 f4 (11 Bg5 +=) 11Bd7 12 Qb3 c5 13 Qxb7
Rb8 14 Qa6 Rb4

[FEN "3q1rk1/p2bp1bp/Q2p1np1/2pP1p2/1rP2P2/
2N3P1/PP3PBP/R1B1R1K1 w - - 0 15"]
15 Qxa7? (underestimates Black's plan; 15 a3! Rb3 16 a4! Ng4 17 Nb5 was a
refutation) 15Ng4 16 h3 Bxc3 17 bxc3 Ra4 (an unusual situation: for the
two sacrificed pawns Black seems to have sufficient compensation) 18 Qb7
Nf6 19 Bd2 Kf7 20 Reb1 Qa5 21 Qb6 Qa8 (21Qxb6 was an option) 22 Rb2
Ra3 23 Bf1 Qa4 24 f3 h5 25 Bd3 (offering a draw) 25Ra8 26 Be1 Ra6 27
Qb8 Qa5?! (risky: Black avoids the draw) 28 Qh8 Ra8 29 Qh6 Rg8 30 Re2
Qd8 31 Bd2 e6 32 dxe6+ Bxe6 33 Rae1 Bc8 34 Bc1 Ra7 35 Qg5 Qd7 36 a3
Rc7 37 Be3? (a serious error; 37 Bc2 Qc6 38 Re3 +/- protects f3 more
efficiently, with good winning chances) 37Qc6 38 Rf1 Re7 39 Bd2 Rxe2
40 Bxe2 Re8 (40Qa4!?) 41 Bd3 d5 42 Rf2 Qe6 43 cxd5 Qxd5 44 Bf1 c4 45
Qh6 Bd7 (45Qc5! =+) 46 a4 Bxa4 47 g4 hxg4 48 hxg4 Bd1 49 gxf5? = (49
g5! Nh5 50 Bc1! would have refuted Black's attack, e.g. 50Re1 51 Ba3 Ng7
52 Qh8 Re8 53 Rd2! +-) 49Rg8 50 Rg2 Bxf3 51 fxg6+ Ke7 52 Rg5 Qxd2
53 Re5+ Kd8 54 Qg5 Qd6 55 Rc5? Rxg6 01, Donaldson (2419) Bcker
(2341), Den Bosch 1999 ("Toernooi voor de Toekomst").
5h6
5Nd5?! appears a bit slow, after 6 c4 Nxf4 7 gxf4 White is better, for
example 7g6 8 h4 or 7e6 8 d5. But at least Black has conquered the
bishop pair, this set-up could be worth a second look. The continuation 5g6
6 d5 Na5 is equally risky. The Krause motif would be sounder than usual after
5Rb8 6 d5?! e5, but almost everybody will prefer 6 0-0! and wait for an
explanation of Black's mysterious rook move.
6 h4
6 d5 is another dangerous possibility: 6Nb8 7 c4 g5 8 Bd2!? Bg7 (after 8
Ne4 White can even play 9 Ba5!? Bg7 10 Nd4 0-0 11 Nc3) 9 Bc3 +=.
6Ng4
Renewing the threat e7-e5.
7 d5 e5

[FEN "r1bqkb1r/ppp3p1/2np3p/3Ppp2/5BnP/
5NP1/PPP1PPB1/RN1QK2R w KQkq e6 0 8"]
There is no obvious refutation of 7Nce5, but White has many alternatives,
for example 8 Nc3 c6 9. 0-0 (9 Qd4 Qc7! 10 e4 g6 11 exf5 c5 12 Qd2 gxf5
+=) 9cxd5 10 Bxe5 (10 Qxd5) 10dxe5 11 Qxd5 e4 12 Qxd8 (12 Nd4 e5
+=) 12Kxd8 13 Rad1+ Kc7 14 Nd4 e6 15 Ndb5+ Kb6 16 Rd8 a6 17 Nd6
Kc7 +=.
The text move follows Krause's basic concept, seen in the older games above.
But against the modern set-up with a bishop on g2 and no pawn on c4, the
Krause approach is perhaps too risky. A repertoire which chooses the Krause
Dutch only in the "safer cases" could be a more pragmatic solution.
8 dxc6!
Clearly stronger than 8 dxe6 Bxe6 9 Nd4 Nxd4 10 Qxd4 c6 11 e4 Qa5+ 12
Bd2 Qb6 13 Bc3 0-0-0 14 Qxb6 axb6 15 Bd2 fxe4 16 Bxe4 d5 17 Bd3 Bc5 18
Be3 Nxe3 19 fxe3 Bxe3 20 Nd2 Rhf8 21 Rf1 Bh3 22 Rxf8 Rxf8 23 Nf1 Bf2+
24 Kd2 Re8 25 a4 Kc7 26 Be2 Bf5 27 Bf3 Kd6 28 Rc1 g6 29 b3 h5 30 Be2
Re6 01, Calapso Kostic, Merano 1926.
8exf4 9 gxf4
White can also force matters by playing 9 Nd4! fxg3 (9b6 would be similar
to the passive game continuation) 10 cxb7 (not 10 fxg3? Ne3) 10gxf2+ 11
Kd2 Rb8 12 Nc6! (12 bxc8Q Qxc8 offers Black more chances) 12Bb7
(what else?) 13 Nxd8 Bxg2 14 Ne6 Rb4!? (14Rxb2 15 Nc3) 15 c3! Rxb2+
16 Kc1 and White is better.
9Be7?
A mistake; overlooking two reasonable alternatives:
(a) 9Rb8 10 Nd4 Be7 11 cxb7 Bxh4 12 bxc8Q Bxf2+ 13 Kd2 Qxc8 14 c3

[FEN "1rq1k2r/p1p3p1/3p3p/5p2/3N1Pn1/
2P5/PP1KPbB1/RN1Q3R b k - 0 14"]
Black has some active play, and White's king isn't exactly in safety. For a
tournament game this situation may offer sufficient compensation, though it
seems dubious whether it is entirely correct: 14Be3+ 15 Kc2 Bxf416 Rh3,
or 14Bxd4 15 Qa4+ followed by Qxd4, or 14Qa6 15 b3.
(b) In the post-mortem my opponent recommended 9bxc6!, which indeed
was my best chance: 10 Nd4 d5 11 Nxc6 Qd6 12 Ne5 (12 e3 Qxc6 13 Bxd5
Qd6 14 Bxa8 Qxd1+ 15 Kxd1 Nxf2+ leads to an acceptable ending: +=), and
Black can choose between the solid 12c6 13 Qd4 Rb8 += and the sharp
12Nxe5!? 13 fxe5 Qxe5 14 Qxd5 Qxb2 with hair-rising complications.
10 Qd3 (10 Nh2) 10b6 11 Nc3 Bf6 12 Kf1!?
There was nothing wrong with 12 0-0-0 0-0 (12Nxf2? fails to 13 Qe3+) 13
Rdf1 etc.
1200 13 Rd1 Kh8 14 Rh3 a5 15 Nd5!

[FEN "r1bq1r1k/2p3p1/1pPp1b1p/p2N1p2/
5PnP/3Q1N1R/PPP1PPB1/3R1K2 b - - 0 15"]
15Bxb2 16 c3 Ba3 17 Kg1 a4 18 h5 Bc5
Black's defensive task remains difficult; the bishop c5 is out of play.
19 e3 Ra5 20 c4 Rf7 21 Nh4 Qe8?! 22 Rg3 a3 23 Ng6+ Kg8 24 Bf3 Ra4 25
Qc2
A database erroneously gives 25 Qc3 as the game continuation.
25Ra5 26 Bxg4 fxg4

[FEN "2b1q1k1/2p2rp1/1pPp2Np/r1bN3P/
2P2Pp1/p3P1R1/P1Q2P2/3R2K1 w - - 0 27"]
27 e4?!
27 f5! Bxf5 28 Nge7+! Rxe7 29 Nxe7+ Qxe7 30 Qxf5, followed by Rxg4,
would have won immediately.
27Kh7 28 f5 Qxc6 29 Nde7 Qd7? (29Qa4) 30 e5 Rxf5 31 Nf8+ Kh8 32
Nxd7 1-0
This article has concentrated on Krause's counter-motif, directed against the
bishop on f4. In my opinion it is one of the hardest tests for the correctness of
Krause's Variation. However, there are many other set-ups available to the
first player. For example, in Kaissiber 11 (1999), Hans Berliner proposed 1 d4
f5 2 g3 d6 3 Nf3 Nc6 4 d5 Ne5 5 e4!? Nf6 6 exf5 Bxf5 7 Nd4 ("Die System-
Methode gegen die Hollndische Verteidigung"), to which Bent Larsen
replied 7Be4! in Kaissiber 12 (1999), p. 18. But White still has 7 Nxe5,
with a small edge. The battle goes on.
Sources
[1] A. Nimzowitsch: "Entspricht Dr. Tarraschs Die moderne Schachpartie
wirklich moderner Auffassung?," in Wiener Schachzeitung 5-8 (1913), pp. 73-
84
[2] A. Nimzowitsch: "Moderne Phantasie ber ein Tschigorin'sches Thema,"
in Kagans Neueste Schachnachrichten 1925, pp. 1-12
[3] M. Euwe: Theorie der Schacherffnungen, Teil VI-VII, Berlin 1965 (2nd
ed.)
[4] S. Bcker: "Dutch Defense: Krause's Variation," in Myers Openings
Bulletin 38 (1986), pp. 2-5
[5] S. Bcker: "Aaron Nimzowitsch: Triumphe dank Dr. Krause," in
Kaissiber 10 (1999) pp. 26-59
[6] S. Bcker: (commented game) in Toernooiboek voor de Toekomst, Zeist
2000
[7] P. Anderberg: "Neues zum Nimzowitsch-Tarrasch-Konflikt," in: Kaissiber
26 (2007), pp. 50-55. How the conflict between Nimzowitsch and Tarrasch
began.
[8] R. Palliser, S. Williams, J. Vigus: Dangerous Weapons: The Dutch,
London 2009
Send your games or comments to [email protected].
Comment on this month's column via our Contact Page! Pertinent responses
will be posted below daily.



[ChessCafe Home Page] [Book Review] [Columnists]
[Endgame Study] [The Skittles Room] [Archives]
[Links] [Online Bookstore] [About ChessCafe.com] [Contact Us]
2010 BrainGamz, Inc. All Rights Reserved.
"ChessCafe.com" is a registered trademark of BrainGamz, Inc.