The Kibitzer
by Tim Harding
Putting Dracula Back in His Coffin
Last month's Kibitzer column dealt with the Frankenstein-Dracula
Variation of the Vienna; in this promised follow-up we look to see
how White should play if Black avoids the most blood-curdling
variations.
After 1 e4 e5 2 Nc3 Nf6 3 Bc4 (or 2 Bc4 Nf6 3 Nc3) 3...Nxe4 4
Qh5 White has broken the rule taught to all beginners, not to bring
the queen out early. The main justification for doing so is that the
forced reply 4...Nd6 makes the black knight obstruct its own
d-pawn. Black's resulting difficulties in developing his queenside
are a recurring theme in this variation.
However, after 4...Nd6 5 Bb3 Black need not play the move
5...Nc6 which, as we saw last month, is in effect a sacrifice of the
exchange (sometimes even the offer of a whole rook). Instead he
can play the calmer (but not necessarily superior) 5...Be7. (See
Diagram)
From the diagram position, therefore, White's options are 6 Qxe5
and 6 Nf3, the latter being superficially the more dynamic.
The most immediate difference from the Frankenstein-Dracula
main line is that instead of trying to defend his extra pawn, Black
makes a kingside developing move that rules out the main
complications. Whereas 5...Nc6 led to a long forcing sequence
after 6 Nb5, there is little point in 6 Nb5 now. Admittedly, the
knight is still invulnerable on account of 6...Nxb5?? 7 Qxf7
checkmate, and 6...g6 is ineffective because 7 Qxe5 threatens the
h8-rook with check, but Black can simply castle -after which
White has nothing better than 8 Nxd6 with a deficit in development
and no problems for Black.
I think one of the main reasons that the Frankenstein-Dracula
Variation is not more popular for White is that it seems to require a
schizophrenic style shift: White must be equally prepared for
colossal complications after 5...Nc6 or peaceful symmetry after
5...Be7 6 Qxe5. Few players, perhaps, are equally happy in both
types of position.
Just as the possibility 5 Qxe5+ is not entirely innocuous, it is not
entirely certain that 6 Qxe5 will lead to a quick draw. After 6...0-0
7 d4 Nc6 8 Qf4 White retains some initiative since Black's
queenside development is not straightforward. The queen is well
placed here to exert some pressure without being attacked by
Black's minor pieces.
The book recipe for Black now is to develop his queen's bishop in
a fianchetto, at b7 or a6. Tseitlin and Glazkov's book "The
Complete Vienna" says that 8...b6 9 Nge2 Ba6 10 Be3 Nc4 11
0-0-0 d5 "led to approximate equality" in McCormick-Hartels,
Omaha 1959. Two recent Polish games continued similarly, but
with 11...N6a5 12 h4 d5. Now look at what happened in the second
of these.
R. Suder - W. Drygalski, Sponsorow 1995: 13 Ng3!
White improves on the drawn game Suder-Bielak, Jurajski 1994, in
which he had played 13 Qf3.
13...Nxb3+ 14 axb3 Nxe3 15 fxe3 Bc8 16 e4 Be6 17 Nf5 c6 18
exd5 cxd5 19 Nxe7+ Qxe7 20 g4 Rfc8 21 g5 a5 22 Qe5 Qc7 23
Rd2 Ra7
23...Qxe5 24 dxe5 had to be tried despite the weak d-pawn and the
fact that the N has good central squares.
24 Qe3 Qc6 25 h5 Re8 26 Qf3 Rf8 27 Rg2 Kh8 28 h6
Black didn't find a good defensive plan and is kingside is now
cracked open.
28...f5 29 g6! Rf6 30 hxg7+ Rxg7 31 Rxh7+ Rxh7 32 gxh7 Qd6
33 Ne2 Kxh7 34 Nf4 Rh6 35 Qg3 Qf8 36 Nh3 Kh8 37 Qe5+ 1-0.
This system for White was recommended almost half a century ago
by U.S. Master Weaver Adams in his "White to Play and Win"
book, or I should say in some editions of it. (Does anyone have a
copy to spare?) If Black plays 8...b5 instead of 8...b6, then a game
circulating in databases nowadays is J.Koch-Chabanon, Nantes
1993. White lost after 9 a3 which only encouraged Black to take a
queenside initiative. Instead White can follow the Adams formula
by 9 Nf3 Bb7 10 Be3 Na5 11 0-0-0 with some advantage
(Adams-Lyman, Boston 1945). In the 10th edition of "Modern
Chess Openings" Larry Evans preferred Black after 11...b4 (Lyman
played Nxb3+) 12 Ne2 Ndc4 but Weaver Adams and Berliner said
White had the edge.
Finally in this sub-variation, if Black wants to develop his bishop
"normally" he must move the knight again: 7...Ne8 8 Nge2 c6 9
d5! d6 10 Qf4 Nf6 11 h3 and White achieved an advantage in
Adams-Gates, Boston 1945.
So it seems that White has slight chances of an advantage with 6
Qxe5 but I don't think there is much doubt that the Frankenstein
Monster would have played 6 Nf3 in the first diagram position.
The other possibility, 6 d3, is unsound after 6...Nc6! (See Diagram)
(Position after 5...Be7 6 Nf3)
The options for Black after 6 Nf3 are to castle into an attack, or to
defend the e-pawn.
After 6...0-0 the obvious 7 Nxe5 is certainly playable, but most
attention has been given to the macho move 7 h4!?, which keeps
cropping up in this variation!
This is really a fun line to play and analyse. What a pity it cannot
be forced!
The classic example Gufeld-Tarve, Tallinn 1969, now went 7...Nc6
(7...g6 8 Qxe5 Bf6 9 Qf4 Re8+ 10 Kf1 Bg7 11 d4 Nc6 12 h5 b6 13
hxg6 was good for White in Roitov-Malevinsky, Leningrad 1960)
8 Ng5 h6 and now, as so often in open games, winning material is
a mistake. In Klaman-Nezhmetdinov, Baku 1951, Black took over
the initiative after 9 Nce4 Nxe4 10 Bxf7+ Rxf7 11 Qxf7+ Kh8 12
Nxe4 d5 and soon won.
So Gufeld improved by 9 Qg6 Bxg5 10 hxg5 Qxg5 11 Qxg5 hxg5
12 Nd5 (More accurate than 12 d3) 12...Nf5 13 d3 and now 13...d6
led to a draw in the very unconvincing game
Szewczyk-Zalachowski, cor 1974, but the usual move is 13...Ncd4
reaching the critical position of the next diagram. (See Diagram)
Gufeld now won by 14 Bxg5?! Nxb3 (If 14...a5 Tseitlin &
Glazkov recommend 15 g4! Nxb3 16 axb3 f6 17 gxf5 fxg5 18 Rg1
or 18 Nxc7 with some advantage) 15 Nf6+ gxf6 16 Bxf6 Ng7 17
axb3 Re8 18 g4 Re6 19 g5 b6? 20 Ke2 e4 21 d4 e3 22 f3 d5 23
Rh4 Ba6+ 24 c4! dxc4 25 Rah1 c3+ 26 Ke1 1-0.
This attacking idea continues to catch Black unawares, e.g. 13...g4
14 Bg5 Ncd4 15 Nf6+ gxf6 16 Bxf6 Ng7 17 Kd2 Nde6 (17...d6 18
Rh2+-) 18 Rh2 Nf4 19 Rah1 Ng6 20 Rh7 1-0 Leconte-Viot, FIDE
Paris open 1994.
The problem with this line is that it is unsound! At least two books
have published the refutation, 19...Ra6!, but the authors didn't state
categorically that White is busted so I am doing that now.
Surely the last diagram position cannot be bad for White?
However, as I stated in my book on the Vienna 20 years ago, 14
g4?! fails to 14...c6! 15 Nc7 Nxb3 (not 15...Rb8 16 gxf5 d5 17 f6
or 17 Bd2, with chances for both sides) 16 axb3 Nd4 17 Kd1 d6!
and now it's awkward for White as 18 Nxa8 is met by 18...Bxg4+.
It seems to me that White should just play 14 Nxc7 which is
prosaic but good, meeting 14...Rb8 by either 15 Bxg5 or 15 c3.
White restores material equality by picking up either the a7-pawn
or the g5-pawn and the resulting position with his rooks controlling
the open files is rather promising.
I am sure White has seen this in the previous games but didn't
select the move as he wanted to play for mate!
In view of this, Black is right to answer 6 Nf3 by 6...Nc6, when
Alekhine refuted 7 0-0 by 7...g6 8 Qh3 (8 Qg4 h5!) 8...Nf5 9 g4
Nfd4 10 Bh6 Bf8! against Jaffe back at Karlsbad 1911.
So White has nothing better than to meet 6...Nc6 by 7 Nxe5, a line
employed successfully by Alekhine (as White) in the 27th game of
his 1935 World Championship match. (See Diagram)
Euwe carelessly replied 7... Nxe5? and found it is not so simple to
equalise: 8 Qxe5 0-0 9 Nd5! Re8 10 0-0 Bf8 11 Qf4 c6 12 Ne3
Qa5 13 d4 Qh5 14 c3 Ne4 15 f3! (Preventing the freeing move
...d5) 15...Ng5 16 d5! cxd5 17 Nxd5 Ne6 18 Qg4 Qg6 19 Be3 b6
20 Rad1 Bb7 21 Qxg6 and after slips by both sides, White
eventually won. Nothing has happened since to change the
assessment of this line.
Back at the last diagram, Alekhine stated that 7...0-0 8 Nd5 Nd4!
(8...g6 9 Nxg6! and 8...Nxe5 are inferior.) 9 0-0 Nxb3 10 axb3 Ne8
would be about equal. Here Larsen suggested 11 Qe2! with the
point 11...Nf6 (11...Bf6 has also been seen.) 12 Nc6! dxc7 13
Nxe7+ Kh8 14 Nxc8 Qxc8 15 d3 but this is the sort of position that
Larsen can win in 80 moves while everybody else gives up trying
and offers a draw in 25. While it is not proven that White has much
here, it may not be an attractive line for Black to defend against
somebody who is good at endgames.
Nowadays the main line is reckoned to be 7...g6, when if White
plays 8 Nxc6 dxc6 he has solved Black's main problem in the
variation: the development of the c8-bishop. He has doubled the
c-pawns but this is hardly enough? As Black is about to gain tempi
against the white queen, the strategic retreat 9 Qd1 seems called for
but I see no advantage for White.
The sad truth seems to be that, so far as the Vienna is concerned,
the era of heroes and monsters has given way to rational defence
and in 1998 White cannot hope to get more than the faintest of
microscopic advantages after 5...Be7, 6...Nc6 if Black knows what
he is doing.
If White wants to play 3 Bc4, he must forget about romantic
mating attacks in the Gufeld-Tarve mould and concentrate on 6
Qxe5 which, as we saw earlier in this column, does seem to offer
some prospects.
Of course that won't suit Dracula and the Frankenstein Monster. A
bloodless draw or long endgame is not their aim when they sit
down to a game of chess. Maybe they'll have to try an Evans
Gambit next time!