When preparing for a local IM tournament here in Malmö a few months ago, I stumbled over an improvement for white in the Four Pawn Attack. The 4P is no doubt the most macho way of meeting the Alekhine, providing black with a large central pawn mass as a target for the levers c7-c5 and f7-f6. However, if black hesitates or mess up his move order he may very well end up on the wrong side of a barrage of pawn and piece sacs - making the white player appear a genius to all theoretically ignorant onlookers...
Since I intended to play the Alekhine as my main defence to 1.e4 but wasn't completely happy with my older preferences against the 4P, I decided to look into black's toughest defence / counter-attack 9.-Be7 which even seemed to result in a =+ according to the latest theory.
After spending a couple of weeks on the very complicated positions resulting from this subvariation, playing through games, analysing and comparing different sources. I found to my astonishment that Krizsány & Videki seemed to indicate in their book "Alekhine Defence" that white could get an advantage in an old game from 1987, which other sources dismissed as only equal. (If you can tolerate horrendous english and messy layout, that book contains some useful info, if you're ready to do some work finding it...).
When I took a closer look at that game, I unfortunately found a virtually winning improvement for white. And since I don't play the 4P myself (even if I was tempted to start doing so...), this article may inspire someone who does...
We'll be following the game Filipenko-Gleizerov, a game which is mentioned in most books on the AL.
1.e4 Nf6 2.e5 Nd5 3.d4 d6 4.c4 Nb6 5.f4 dxe5 6.fxe5 Nc6
[The main alternative is Mikenas' interesting idea: 6...c5 7.d5 e6 8.Nc3 exd5 9.cxd5 c4!? which of course is a critical try, but my gut feeling is that black is risking more than white after both10.Nf3 and 10.d6!? even if it's complicated as always.]
7.Be3 Bf5 8.Nc3 e6 9.Nf3 Be7!?
This is considered to be black's best reply, but both 9.-Bg4 and 9.-Qd7 are safe alternatives.
10.d5!?
I think this is white's only bid for an advantage. Most experts prefer the calmer 10.Be2 which leads to a quiet waiting for opportunities. The reason they prefer it is mainly because the complications after 10.d5 previously haven't worked out for white.
10...exd5
[The former main line was: 10...Nb4 which still looks playable if one is ready to give a piece for some pawns and complications.]
11.cxd5
[But not: 11.Bxb6? axb6 12.cxd5 Nb4 13.Nd4 Bg6 14.d6 which is excellent for black, who can give up his bishop on three different squares: d6, e7 and h4; since this will give black time to castle. A privilege for which the white king will pay dearly...]
11...Nb4 12.Nd4 Bd7 (the only move) 13.e6 fxe6 14.dxe6 Bc6 15.Qh5+!
This is probably the only move that will trouble a booked up second player in this variation.
|
|
[15.Qg4 is the traditional main line, but nowadays the main question is just how much better black stands after: 15...Bh4+ 16.g3 Bxh1
|
17.0-0-0
(17.Bb5+!? c6 18.0-0-0 0-0 19.gxh4 Df6 20.Bg5 De5 21.e7 cxb5 22.Nf5 Nxa2+ 23.Nxa2 Tfc8+ 24.Nc3 Rxc3+ this certainly looks promising, but it's possible that white actually has roughly even chances here...25.bxc3 Qxc3+ 26.Kb1 Db3+ 27.Ka1 Bd5 28.Nh6+ Kh8 29.Nf7+ Bxf7 30.Rd8+ Be8 31.Qf5 h6 32.Qf8+ Kh7 33.Qxe8 Dc3+ ˝-˝ Roth,E-Etmans,M/corr 1986 (33))
17...Qf6 18.gxh4 0-0 and black's position is good enough to deal with white's various tries, maintaining at least the slightly better chances.]
15...g6 16.Qh6!?
Instead of various X-ray threats aimed at the knight on b4 after the old 15.Qg4, white will now try to keep black's king in the centre. According to my database this position has only occured in two games so far - which may explain why somebody hasn't discovered my TN before...
|
|
16...Bf6
[To interpose: 16...Bh4+ doesn't seem to improve black's situation, after: 17.g3 Bf6
(17...Bxh1 18.gxh4 Df6 19.0-0-0 is +/- according to Krizsany/Videki. The following game seems to confirm that conclusion: 19...N6d5 20.Bg5 De5 21.Nxd5 Bxd5 22.a3 Sc6 23.Bd3 Se7 24.Bc2 c6 25.h5 1-0 Barten,R-Klein,A/corr TT 1993 (25))
18.0-0-0 De7 19.a3 S4d5 20.Nxc6 bxc6 21.Nxd5 Nxd5 22.Rxd5! white has a slight advantage according to Krizsany/Videki. Strangely enough, they then follow previous books on the Alekhine - giving the stem game as main line and an '=', despite that the positions from this variation and the main line is identical, except for the position of white's g-pawn which is hardly important.
Perhaps black should go for queenside castling with: 16...Qd6!? now Krizsany/Videki give 17.Qg7 as an easy win for white (I think 17.0-0-0!? is a better choice.) but after 17...0-0-0 18.a3 they only give: 18...N4d5?? 19.Ne4 +-
I think black can do a whole lot better than that, with either
A) 18...Nc2+ 19.Nxc2 Bh4+ 20.g3 Bxh1 21.gxh4 Qxh2;
or probably even better:
B) 18...Bh4+ 19.g3 (19.Kd2 Tdf8 20.axb4 Bf6) 19...Bxh1; ]
But let us return to Filipenko-Gleizerov, for a few more moves at least...
17.0-0-0 De7 18.a3 S4d5 19.Nxc6 bxc6 20.Nxd5 Nxd5 21.Rxd5 cxd5 22.Bb5+ Kd8 23.Rf1! Rf8
[Black already suffers from a severe lack of good alternatives:
23...d4
24.Rxf6 dxe3 25.Qg5 c6 26.Bxc6 ±; 23...Bg7 24.Qg5! +-]
24.Rd1!
The big novelty! Even if it does look rather obvious once one has come this far into the theoretical maze. Thanks to the short stop on f1, black obviously has no Qxe6 anymore. It seems to be yet another example that a threat (Rxf6) is stronger than it's execution.
|
|
[Instead, the stem game continued: 24.Rxf6 Rxf6 25.Bg5 Rb8 26.Bd7 Qc5+ 27.Kb1 Rxb2+ 28.Kxb2 Qd4+ 29.Kc2 Qc4+ 30.Kd2 Qd4+ 31.Ke2 Qe4+ with a draw, Filipenko-Gleizerov, Soviet Union 1987.]
And now I haven't been able to find a defence for black, for example:
24...c6
[24...d4 (24...Qd6 is of course met by 25.Rxd5 +-) 25.Bxd4 Kc8 (25...Bg5+ 26.Be3+) 26.Ba6+ Kb8 27.Bxf6 Qxf6 28.Qe3 +-;
24...Bxb2+ 25.Kxb2 Rb8 26.Rxd5+ (26.a4 Rxb5+ 27.axb5 Db4+) 26...Kc8 27.Ka2! +- ]
25.Bxc6 Rb8
[25...Rc8 26.Rxd5+ Kc7 27.Qf4+! Kxc6 28.Qe4 Kc7 (28...Kb7+ 29.Rc5+) 29.Rc5+ Qxc5+ 30.Bxc5 Kb8 31.Kb1 +-]
26.Rxd5+ +- Kc8 27.Bd7+ Kb7 28.Rb5+ Ka6 29.Rxb8 Rxb8 30.Qf4 and white is winning.
Thomas Johansson
This article of mine was previously published in Swedish at http://www.schack-kompaniet.com/veckans_tn!.htm on August 9th 2000. As my Swedish visitors probably already know.
Added June 2nd 2005 (link updated 8/sept-06 - also see Mark Morss' analysis in my Guestbook.); A new idea for black, overturning the whole variation!
