Tuesday 17 July 2007

GAME 008 - Training Game from 25th Dec 2006

The following training game was played on the Playchess.com server in late December in preparation for the Hastings weekend congress. I am black and try out a slightly odd early h-pawn push in a Caro-Kann advance. I found this analysis on my harddrive and decided to post it.





Dura (1764) - Kraken (1962)
Rated game, 45m + 20s Main Playing Hall, 25.12.2006


1.e4 c6 2.f4 d5 3.e5 h5 Not a great move; as is shown, Bf5 is still completely possible. [3...Bf5 4.Nf3 (If 4.g4 Then 4...Be4 5.Nf3 e6 6.d3 Bg6 (6...Bxf3 7.Qxf3 Qb6 This is also good for black.) 7.h4 (7.Be3 (Fritz prefers this)) 7...h5 (7...h6 This is inferior. 8.h5 Bh7) 8.g5 Qb6 This position is very good for black.) 4...e6 5.Be2 (5.h3 Be7 6.g4 Bh4+ 7.Ke2 (7.Nxh4 Qxh4+ 8.Ke2 Bxg4+ 9.hxg4 Qxh1–+) ) ] 4.d4 Bf5 5.Nf3 e6 6.Bd3 Bxd3 [6...Nh6 is well worth considering. E.g: 7.Ng5 Bxd3 8.Qxd3 Nf5 9.Nd2 Be7 10.Ndf3 c5] 7.Qxd3 Ne7 [Again 7...Nh6 very much suggests itself. This is probably superior to Ne7 because it aims at both g4 and f5, and also doesn't impede development of the other pieces such as the f8 bishop. In conclusion, Nh6 is more flexible and positionally more logical.] 8.Nc3 Nd7 9.Bd2 Nf5 10.Ng5 Be7 11.g3 I get the feeling this is a waste of a move. [11.0–0–0 is suggested by fritz.] 11...c5 A combination is missed. [11...Bxg5 12.fxg5 Qb6 Double attack d4 b2 wins black a clean pawn. e.g: 13.Ne2 Qxb2 14.0–0 Qb5 15.Qxb5 (or Qxd3 cd leaves white's pawn structure weaker) 15...cxb5 16.Rfb1 Rc8 etc(This is far inferior. The 'complications' of Rc8 shouldn't be avoided. 16...a6 17.a4 Nb6 18.axb5 Nc4 19.Bc1 a5 20.h4 Kd7 21.Ra2 Black has a much better position, but no longer has a material advantage.) ] 12.Ne2 Qb6 Black has a large advantage. 13.Bc3 This stands out as a positionally ugly looking move. [Fritz recommends: 13.Nf3 e.g. 13...Qxb2 14.0–0 Qb6 Black still has a large advantage.] 13...cxd4 This isn't bad, but black would have kept more advantage with 13...Bxg5 [13...Bxg5 14.fxg5 cxd4 15.Bxd4 (if 15.Nxd4 then 15...Nxe5 16.Qb5+ Qxb5 17.Nxb5 d4! 18.Bd2 (otherwise 18.Nxd4 Nxd4 19.Bxd4 Nf3+ 20.Kf2 Nxd4; 18.Bxd4 Nf3+ 19.Kf2 N3xd4) ) 15...Nxd4 16.Qxd4 Qxd4 17.Nxd4 Nxe5 18.0–0 Ke7 Black has a very strong position.] 14.Nxd4 Bxg5? A big mistake. This move, as opposed to some of the earlier instances, doesn't work now. The position shifts from black's favour to white's. Presumably I felt under some kind of positional pressure and felt obliged to come to some clarity by a sequence of trades. 14...Nxd4 for example would have been much better though. [14...Nc5 This is fritz's first choice. 15.Qd2 (I may however have been scared of this. 15.Qb5+ Qxb5 16.Nxb5 This move is tricky to deal with, but black doesn't lose any of his advantage. 16...Rd8 (not 16...Kd7 17.Nxf7) 17.Nxa7 Bxg5 18.fxg5 d4 19.Bb4 Rd5 20.0–0–0 Ne4 21.c4 Rxe5 I don't consider such a sequence in any way easy to calculate or foresee, but it must be possible.) 15...Bxg5 16.Nxf5 exf5 17.fxg5 Ne4! 18.Bd4 (18.Qxd5?? Qe3+ 19.Kd1 (19.Kf1 Qf2#) 19...Nf2#) 18...Nxd2 19.Bxb6 Ne4 20.Bd4; 14...Nxd4 15.Bxd4 Qb4+ (15...Bc5 16.0–0–0 (16.Bxc5 Nxc5 17.Qd4 Rc8) 16...Bxd4 17.Qxd4 Qxd4 18.Rxd4 The position is almost totally equal.) 16.Qd2] 15.Nxf5 exf5 16.fxg5 The resultant position isn't very pretty for black. I had the feeling I'd now be fighting for a draw. 16...Qe6 This was the only way I could see to defend the d5 and f5 pawns. It seems I chose the best move other than Nc5; a move that I still don't really understand. [16...Nc5 17.Qxf5 (Fritz gives the following sequence which I hardly understand, but despite being two pawns down some of the time, black is given as much better. 17.Qxd5 Rd8 18.Qf3 Ne4 19.Qxf5 Nxc3 20.bxc3 Qc5 21.Qf3 0–0 22.Rd1 Rxd1+ 23.Kxd1 Qxe5) 17...Ne4 18.0–0–0 Nxc3 19.bxc3 Qe3+ 20.Kb2 Qb6+ 21.Ka1 Qe6 22.Qd3 (22.Qxe6+ fxe6 23.c4 dxc4 24.Rhf1 The second example sequence from Fritz. I don't understand either of these fully, so 16..Nc5 would hardly be playable for me currently.) ] 17.0–0–0 The natural and best move, doubling up on d5. 17...Nb6 Slightly inferior to Nc6 but not by much. [17...Nc5 18.Qxd5 Qxd5 19.Rxd5 Ne4 20.Bb4 Rd8 21.Rxd8+ Kxd8 22.Rd1+ Kc8] 18.Bd4 [18.Rhf1 g6 19.Bd4 Nd7 20.Bg1 Nb6 21.Bc5 0–0–0 (Not 21...Nd7 22.Bd6) ] 18...Nc4 19.b3 Na5 20.Bc3 [20.Bc5 with a view to Bd6] 20...Nc6 21.Qxd5 Qxd5 22.Rxd5 0–0 Fritz puts this as a whole pawn under the move 22...h4, and gives a losing evaluation (almost 2 pawns). [22...h4 23.gxh4 Rxh4 24.e6 fxe6 25.Re1 Re4 26.Rxe4 fxe4 27.Rd2 e5 The comparatively good position of the black king is probably why this variation is preferred so much to 22...0–0] 23.h3 [23.Rf1 g6 24.h3 Rfd8 25.Rfd1 Rxd5 26.Rxd5 Rd8 27.Rd3 Kf8 28.Rxd8+ Nxd8 29.Bb4+ Ke8 30.Kd2 Nc6 31.Bc3] 23...Rac8 24.g4 This is a big mistake. Black has better drawing chances after this. [24.Rf1] 24...Ne7 Not much worse than the recommended fg [24...fxg4 25.hxg4 hxg4 26.Rg1 Rfd8 27.Rxd8+ Nxd8 28.Kd2 Kh7 29.Rxg4 Kg6 30.a3 Ne6 31.Rg2 it's obvious just looking at the position that black has decent drawing chances now] 25.Rd7 Fritz dislikes this NB trade. 25.Rd3 is no.1 25...Rxc3 26.Rxe7 Now we have entered the double rook ending, and plans must be made immediately. Much analysis can be done one this, testing different plans. It appears through analysis that this double rook ending should be drawn by black. 26...Rfc8 [Fritz's first choice: 26...hxg4 27.hxg4 fxg4 (27...f4 28.Rxb7 Rg3 29.g6 fxg6 30.Rd1 Rxg4 31.Rdd7 Kh7 32.Rxg7+ Kh6 33.e6 Re8 34.Rge7 Rxe7 35.Rxe7 Rg5 36.Rxa7 Re5 37.Kd2 Rxe6 38.b4 Rd6+ 39.Ke2 Rc6 40.Kf3 Rxc2 41.Kg4 (41.Kxf4 Rc4+ 42.Kf3 Rxb4) 41...g5 42.Ra6+ Kg7 43.Kxg5 f3 44.Rg6+ Kf7 45.Rf6+ Ke7 46.Rxf3 Rxa2= 47.Rd3 Ke6 48.Kf4 Rf2+ 49.Ke4 Re2+ 50.Kd4 Kd6 51.Kc4+ Kc6 52.b5+ Kb6 53.Rd6+ Kb7 54.Rd8 Rc2+ 55.Kb3 Rc1 56.Rf8 Kb6 57.Kb2 Rc5 58.Rb8+ Kc7 59.Ra8 Rxb5+) 28.Rxb7 g3 29.Rg1 Re8 30.g6 fxg6 31.Kd2 Rec8 32.Rxa7 Rxc2+ 33.Ke3 g2 34.e6 Kf8 35.e7+ Ke8 36.Kf4 Rb2 37.a3 Rcc2 38.b4 Rc4+ 39.Ke3 Rc3+ 40.Ke4 Rh3 41.Rc1 Re2+ 42.Kf4=] 27.Rh2 hxg4 28.hxg4 fxg4 29.Rg2 R8c7 [29...g3 was better. 30.Rd7 (30.Rxb7 is impossible due to the combination: 30...Rxc2+! 31.Rxc2 Rxc2+ 32.Kxc2 g2 33.g6 fxg6 34.Rxa7 g1Q) 30...b5 31.g6 Fritz evaluates this as dead drawn.] 30.Rxc7 Rxc7 31.Rxg4 Kh7 Inaccurate on account of 32.g6+, doubling black pawns and leaving a strong e5 passed pawn. [31...Re7 32.Rb4 Kh7 33.Rb5 Kg6 34.e6 fxe6=] 32.c4 This throws away the advantage resulting in a very drawish position. g6+ was the criticalcontinuation [32.g6+ fxg6] 32...Kg6 Black's moves are now not difficult to find. 33.Kc2 [Fritz recommends 33.Re4 Kxg5 34.e6 fxe6 35.Rxe6 with a position that should be drawn.] 33...Kf5 Black's king position is obviously very good now, and black is no longer under pressure. [fritz prefers 33...Re7 34.Rd4 Kxg5 35.Kd3 Kf5 36.Rd5 White is the one with the more difficult position. 36...Kf4 37.Kd4 g5 38.Rd6 Rxe5 39.Rd7 f5 40.Rxb7 a5 41.c5] 34.Rg2 Kxe5 35.Kc3 a5 A bit of an unnecessary committal move, but I probably felt I shouldn't let white gain too much ground on the queenside uncontested. [Fritz prefers 35...Kf5 allowing white queenside advance, but maybe the computer has it wrong as the following quick bit of analysis shows, (though it would be more illustrative to give the computer lots of time to analyse each move of this line, I have given it about 1 to 2 seconds per move)... 36.b4 Re7 37.g6 f6 38.b5 Re3+ 39.Kd4 Re4+ 40.Kd3 b6 41.a4 Ke5 42.a5 f5 43.axb6 axb6 44.Ra2 Rg4 45.Re2+ Kf6 46.c5 bxc5 47.Rb2 Rd4+ 48.Kc3 Rd8 49.b6 Kxg6 50.b7 Rb8 51.Kc4 Kh7 52.Kxc5 f4 53.Kb6 g5 54.Kc7 Rxb7+ 55.Kxb7 Kg7 56.Kc6 Kg6 57.Kd5 Kf5 58.Kd4 Kg4 59.Ke4 Kg3 60.Rb3++- White is won.] 36.a4 [36.Rd2 Maybe it is more consistent to play on the possible weakeness of the black q-side pawns right away; fritz: 36...Kf4 37.Rd5 b6 38.a3 Re7 but black is in no trouble here.] 36...Ke4 37.b4 This can't be a very good move. Black can now give white much weaker separated q-side pawns. 37...axb4+ 38.Kxb4 f5 39.gxf6 gxf6 Black now has the better position. 40.Kb5 Fritz considers this a bad move, incresing black's advantage quite considerably. 40...f5 41.c5 f4 42.Kb6 Apprarantly allowing winning tactics. [42.Kc4 is preferred, but the evaluation is already suggesting very good winning chances for black. 42...f3 43.Rg8 Re7 44.Rf8 Ke3 45.a5 f2 46.Kb5 Re6 47.c6 Rxc6 48.Re8+ Kf3 49.Rf8+ Kg2 50.Rg8+ Kf1 51.Rf8 Re6 52.Rf4 Ke2 53.Rf5 f1Q 54.Rxf1 Kxf1] 42...Rc6+ A missed opportunity, restoring the evaluation as a draw. [42...f3! 43.Rg4+ Ke3 44.Kxc7 f2 45.Rg3+ Ke4 46.Rg4+ Ke5 47.Kxb7 f1Q White has good practical drawing chances. 48.c6 Qb1+ 49.Kc7 Qd3 50.a5 Ke6 51.a6 Qxa6 52.Rg6+ Ke7 53.Rg7+ Kf6 54.Rd7 Qa7+ 55.Kc8 Qa8+ 56.Kc7 Ke5 57.Rd3 Qa6 58.Rh3 Qe2 59.Kc8 Kf4 60.c7 A well known technique is needed to even win this with just Q vs p, so the inclusion of a defending rook I'm sure makes this technically extremely difficult to win, if it is indeed possible to win at all. 60...Qe8+ 61.Kb7 Qb5+ 62.Ka7 Qc6 63.Kb8 Qb6+ 64.Kc8 Qb5 65.Rh6 Qb3 66.Kd7 Qd5+ 67.Rd6 Qf5+ 68.Kd8] 43.Kxb7 Rxc5 44.Kb6 [More accurate is 44.Ra2 , with the rook getting behind the passed pawn. 44...Rf5 45.a5 The position is considered to be dead drawn still, but as is seen, this results in a pure R+Q vs R+Q game, which in practice might become chaotic. 45...Rf7+ 46.Kb6 Rf6+ 47.Kc5 f3 48.a6 Rf5+ 49.Kd6 f2 50.a7 Rd5+ 51.Kc7 f1Q 52.a8Q for example: 52...Qc4+ 53.Kb7 Qb4+ 54.Ka7 Qc5+ 55.Kb7 Qb5+ The computer draws against itself. However, between human opponents, the player with the move may well have good chances of tricking his/her opponent.] 44...Rc3 [Naturally it makes more sense put the rook behind one's passed pawn. This doesn't alter the evaluation though. 44...Rf5 ] 45.a5 f3 46.Rg4+? According to Fritz, a losing blunder! Let us analyse (?-Ed.) 46...Ke3 47.Rg3 Kf4 48.Rg8 f2 49.Rf8+ Kg3 50.Rxf2 Kxf2 51.a6 Rb3+ 52.Kc7 Ra3 53.Kb7 Ke3 54.a7 Rb3+ 55.Kc8 Ra3 56.Kb8 Rb3+ 57.Kc7 Ra3 58.Kb7 Rxa7+ 59.Kxa7 ½–½

No comments: