Scandinavian 4...Qh5: coffeehouse bluff or serious weapon?
Schliemann Mann
3.d4 Bg4, was popular and successful enough in the 1990s that many
players took care to avoid it as White by switching to 3.Nf3. (I
recall an article title in Kaissiber along the lines of "3...Bg4:
Victim of Its Own Success?") The point is that after 3.Nf3 Bg4? 4.Bb5+
Nbd7 5.c4 a6 6.Ba4, Black doesn't get the active counterplay for the
pawn that is typical of the Scandinavian gambit lines.
In 1998 NCO was able to relegate the above line to the first footnote
on p.126, note that 3...Nxd5 4.d4 transposed back to the traditional
main line of the 2...Nf6 variation, and leave it at that. They didn't
even bother to note that 3...Qxd5 4.Nc3 Qa5 5.d4 tranposes to the main
line of the 2...Qxd5 Scandinavian.
But one wouldn't expect devotees of the Icelandic Gambit (3.c4 e6) and
the Jadoul Variation to give up or be satisfied with main line
Scandinavians, and indeed I've noticed a new variation making more and
more appearances in recent years: 1.e4 d5 2.exd5 Nf6 3.Nf3 Qxd5 4.Nc3
Qh5!?
[1.e4 Nc6 players can try a similar line in response to 2.Nf3 which
may transpose, 2...d5 3.exd5 Qxd5 4.Nc3 Qh5, but should be aware of
Parma-Larsen, Zagreb 1965(!): 5.Nb5! Kd8 6.d4 Nf6 7.Bf4 Nd5 8.Qd2 a6
9.Nc3 Nxf4 10.Qxf4 e6 11.O-O-O Bd6 12.Qe3 which White won in 38 moves.
The only losses White has suffered in this line are when White
deviated from Parma's play, not Black improvements on Larsen.]
Black's plan seems to be to load up on the central dark squares with
Bg4, Nc6, e5 and O-O-O, whether White plays d4 or not. The hallmark of
the line is a refusal to retreat the Bg4 if attacked by h3: for the
time being the h-pawn is pinned to the rook, and if White castles
Black can sac the bishop for two pawns with Bxh3 or more often just
leave it there and if hxg4 Nxg4, attack the Nf3 with e4 or Nd4 and
break through on h2.
A brutal recent example of the line was Rodriguez Coarasa - Campos
Calvo in the Pablo Gorbea memorial in Madrid, September 1, 2003: 1.e4
d5 2.exd5 Nf6 3.Nf3 Qxd5 4.Nc3 Qh5 5.d4 Bg4 6.Be2 Nc6 7.Be3 0-0-0
8.Qd2 e5 9.0-0-0 exd4 10.Nxd4 Nxd4 11.Bxd4 Bc5 12.Qe3 Rhe8 0-1. Black
moves his queen twice, then everything else is a picture of model
efficiency in development: each minor piece to an active square, pawns
to d5 and e5, rooks to d8 and e8, a couple of exchanges on d4, and
boom, the game's over.
Obviously White's play in that game was horrible. My question is, is
there a good source for "theory" in this line, for those looking for
the best lines for both White and Black? Is Black's bishop sacrifice a
coffeehouse bluff good only for blitz and rapid chess, or a serious
weapon for slow time control chess as well?
The few times top-level players have faced the line, they've handled
it successfully with 5.Be2 Bg4 6.O-O Nc6 7.h3, putting the question to
the bishop as soon as possible. The most popular response has been
7...O-O-O, but in Godena-Damaso, Lisbon zonal 1993, White grabbed the
piece and survived a wild sacrificial attack, marching his king to b5
with queens still on the board: 8.hxg4 Nxg4 9.d3 e5 10.Ne4 Nd4 11.Re1
f5 12.Ng3 Nxf3+ 13.Bxf3 Qh2+ 14.Kf1 Nxf2 15.Qe2 Bc5 16.Qxe5 Nxd3
17.Qxf5+ Kb8 18.cxd3 Qg1+ 19.Ke2 Rhe8+ 20.Be4 Qxg2+ 21.Kd1 Qxg3 22.Bf4
Rxd3+ 23.Kc2 Qf2+ 24.Kxd3 Rd8+ 25.Bd5 Qg2 26.Re8 Qf3+ 27.Kc4 b5+
28.Kxb5 1-0.
Interestingly, the line was repeated through Black's 14th move in
Jactel - Le Duin, Paris 2002, when White deviated with 15.Kxf2 and
also won: 15...Bc5+ 16.Be3 Bxe3+ 17.Rxe3 f4 18.Re1 Qxg3+ 19.Kg1 Rd6
20.Bh5 Rg6 21.Bxg6 hxg6 22.Qf3 Qh2+ 23.Kf2 Qh4+ 24.Ke2 Qg5 25.Kd2 c6
26.Rad1 Rh2 27.Re2 Kc7 28.Kc1 c5 29.Kb1 b6 30.Rde1 b5 31.Qd5 1-0.
And Jonathan Rowson found yet another way to successfully consolidate
after 7...O-O-O 8.hxg4 Nxg4, vs. Vitor in the 1995 under-20 World
Championship: 9.Ne4 Nd4 10.Neg5 Nxf3+ 11.Nxf3 f5 12.g3 Qh3 13.Re1 e5
14.Bf1 Qh5 15.Bg2 Bc5 16.d4 Rxd4 17.Nxd4 Bxd4 18.Be3 Bxe3 19.Rxe3 Qh2+
20.Kf1 Nxe3+ 21.fxe3 Qxg3 22.Qd5 c6 23.Qe6+ Kb8 24.Qd6+ Kc8 25.Qe6+
Kb8 26.Qxf5 Qxe3 27.Re1 Qf4+ 28.Qxf4 exf4, and 1-0 in 49.
The fact that White has found two or three different ways to win here
doesn't bode well for 7...O-O-O. And retreat is out of the question:
in a game from last December's Pamplona Open, Baron Rodriguez - Argaya
Urdaniz, Black got gored after 7...Bd7? 8.d4 O-O-O 9.Ne5 Qh4 10.Nxf7
Nxd4 11.Bg5 Nxe2+ 12.Qxe2 1-0.
7...Bxf3 8.Bxf3 Qf5 looks like a possible alternative, though it was
no match for a GM in Nijboer-Vandevoort, Brussels zonal 1993: 9.d4 g5
10.Re1 Rd8 11.d5 Ne5 12.Be4 Nxe4 13.Nxe4 Bg7 14.c4 0-0 15.Qb3 h6
16.Bd2 b5 17.cxb5 Rxd5 18.Nf6+ exf6 19.Qxd5 h5 20.Bc3 g4 21.hxg4 hxg4
22.Bxe5 fxe5 23.Rac1 Bf6 24.Qe4 Qe6 25.Rc4 Bg7 26.Qc6 Qf5 27.Qxc7 Bf6
28.Qxa7 Rd8 29.Qe3 1-0. Black tried the curious 8...Qc5!? in Buckley -
Moreno Tejera in the 1999 under-14 WC and the game got rather crazy:
9.b4 Nxb4 10.Rb1 e6 11.Bxb7 Rb8 12.Qf3 Nd7 13.d4 Qf5 14.Qxf5 exf5
15.Bf3 Nxc2 16.Rxb8+ Nxb8 17.Nb5 Bd6 18.Bb2 Nb4 19.Nxd6+ cxd6 20.Ba3
a5 21.Bxb4 axb4. In the end Black drew a pawn-down N vs. B endgame.
7...Bxh3 8.gxh3 Qxh3 worked in Borda-Osorio, u-16 Pan-American ch
2001, but White's defensive technique was deficient to say the least:
9.Nh2 h5 10.f3 e6 11.Ne4 Rh6 12.d3 Nxe4 13.Bxh6 Bc5+ 14.d4 Bxd4+
15.Be3 Bxe3+ 16.Kh1 Ng3#.
All of this makes the following game very interesting, where Black
plays Bxh3 one move later after White played the apparently less
critical 6.d3, in Genin-Saatdjian, French under-20 championship,
Clichy, Dec. 2002: 5.Be2 Bg4 6.d3 Nc6 7.h3 0-0-0 8.0-0 Bxh3 9.gxh3
Qxh3 10.Ng5 Qf5 11.Nxf7 Qg6+ 12.Ng5 h6 13.Bg4+ Nxg4 14.Qxg4+ Kb8
15.Qe6 Rd6 16.Qxg6 Rxg6 17.f4 hxg5 18.fxg5. Black has gotten his piece
back and material is even, but White went on to win the endgame:
18...e6 19.Be3 Be7 20.Ne4 Nb4 21.Rfc1 Rh3 22.Bd2 Nc6 23.Kg2 Rh8 24.Rh1
Rf8 25.Raf1 Rxf1 26.Kxf1 Kc8 27.Kg2 e5 28.Kg3 Nd4 29.c3 Ne6 30.Kg4 Nf8
31.Kf5 Bd6 32.Rh8 Kd7 33.Rxf8 Bxf8 34.Kxg6 Ke6 35.c4 a6 36.a3 c5 37.b4
b6 38.Ng3 Kd6 39.Nf5+ Kc6 40.Nxg7 b5 41.Ne6 Bd6 42.Kf7 e4 43.cxb5+
Kxb5 44.dxe4 Ka4 45.bxc5 1-0. Was Black OK at some point in this game,
or just busted the whole way?
Finally, what about 5.Be2 Bg4 6.O-O Nc6 7.h3 e5? I have two recent
games with this move, but 8.Nh2?! isn't critical and 8.hxg4 Nxg4 9.Nh4
f5 10.g3 Nd4 11.Nd5, draw agreed, doesn't tell us much! No top-level
examples either. 8.d3 might permit an improved version of ...Bxf3 and
...Qf5.
Summary: the critical lines after 5.Be2 Bg4 6.O-O Nc6 7.h3 seem to be
(a) 7...Bxf3 8.Bxf3 Qf5 or Qc5; (b) 7...Bxh3 8.gxh3 Qxh3; (c) 7...e5
8.hxg4 Nxg4. With the seemingly inferior 7...O-O-O still played most
frequently, we have little in the way of practical examples.