Annotated Game #56: Training game - King's Indian Attack

This training game (against Chessmaster personality "Turk") was the first time I had faced 1. g3 and perhaps the innocuous start by White made me more aggressive than usual, as I chose a somewhat dubiously threatening line with 5...Bc5 instead of transposing to the normal Caro-Kann line versus the KIA.  An inadvertent pawn sacrifice from Black makes the early part of the game interesting, where White needs to be careful in order not to give Black too much activity and developmental lead.

Positionally, Black is unreasonably fearful of having a two-pawn center and as a result does not achieve an optimal set-up in the middlegame.  I had worried that the pawn center could be too easily undermined by White, but analysis shows this was not the case.  This judgment error is a good example of  how an uncritical preference for a type of position (or against one) can lead to less effective play.

Around move 30 my calculation and judgment began declining rapidly when Black was faced with a menacing White pawn mass on the queenside.  At least I had a bailout into a drawn position, rather than fully collapsing into a loss.  "Turk" had played a typical computer handicap move on move 20 to give me the advantage, so I guess I just ended up returning the favor.

The training game was useful for highlighting the individual calculation and judgment errors mentioned above in the opening, middlegame and endgame phases, so was successful from that perspective.

[Event "Chessmaster: Grandmaster Edition Rated G"] [Site "?"] [Date "2012.06.03"] [Round "?"] [White "Turk"] [Black "ChessAdmin"] [Result "1/2-1/2"] [ECO "A00"] [Annotator "ChessAdmin/Houdini"] [PlyCount "87"] [EventDate "2012.??.??"] {A00: Irregular Openings} 1. g3 d5 2. Bg2 c6 3. d3 Nf6 4. Nd2 e5 5. e4 Bc5 ( 5... Bd6 {would transpose into the main Caro-Kann line against the King's Indian Attack, assuming White goes ahead with Ngf3. This is probably Black's best course of action here.}) 6. Ngf3 Qb6 $146 {this is aggressive-looking, but the threat to f2 is easily parried.} (6... O-O {appears to be the most interesting option here.} 7. O-O (7. Nxe5 {is tempting but doesn't fare too well, for example:} dxe4 8. Nxe4 Nxe4 9. dxe4 (9. Bxe4 Re8) 9... Qxd1+ 10. Kxd1 Bxf2) 7... Re8 8. Qe2 Bg4 9. h3 Bh5 10. Re1 Nbd7 11. Nf1 Qc7 12. g4 Bg6 13. Ng3 Nf8 14. Nh4 Ne6 15. g5 Nd7 16. exd5 cxd5 17. Bxd5 Nd4 18. Qd1 e4 19. Be3 exd3 20. cxd3 Ne5 21. Nxg6 hxg6 22. Rc1 Qd6 23. Be4 Rad8 24. Kg2 b6 25. f4 Nec6 26. Bxd4 Nxd4 27. Qg4 Ne6 28. Rf1 Be3 29. Rc6 Nxf4+ 30. Kh1 Qd4 31. Rc4 Qxb2 32. Qf3 Qd2 33. Rc7 b5 34. Rb7 Rc8 35. Bd5 Rc2 36. Bxf7+ Kf8 37. Ne2 Qxe2 38. Qxe2 Rxe2 39. Bxg6 Rc8 40. Bh5 Rd2 41. Rd7 Rcc2 42. Rd8+ Ke7 43. Re8+ Kd7 {0-1 Fernandez,D (2423)-Pruess,D (2400)/ICC INT 2005/CBM 108 ext}) 7. Qe2 (7. O-O Nbd7 $11) 7... Bg4 {this ignores the threat down the e-file and involves an inadvertent pawn sacrifice in the game line, which however is still good enough for equality. Houdni prefers a positional approach (8. h3) which would seize the two bishops and gain time for White in the opening.} (7... O-O) 8. exd5 (8. h3 Bxf3 9. Nxf3 Bd6 $14) 8... O-O {here the e5 pawn is defended indirectly, with the Re8 potential pin.} (8... Nxd5 $2 {doesn't work} 9. Nc4 ({ or simply} 9. Qxe5+) 9... Qd8 10. Ncxe5 Bxf3 11. Nxf3+ Be7 12. Qd2 $16) ({ Instead of} 8... cxd5 9. Qxe5+ Be7 10. O-O $14) 9. h3 (9. dxc6 Nxc6 {and Black is actually doing reasonably well, with his lead in development and piece activity compensating for the pawn.}) 9... Bxf3 $11 {the decision to exchange bishop for knight in these types of Caro-Kann positions is actually quite common. The downside to keeping the bishop would be its limited scope and loss of time as White could further threaten it.} 10. Nxf3 Nxd5 {one of those seemingly minor yet important recapture decisions. Black declines to have a two-pawn center, concerned it could be undermined or attacked easily.} (10... cxd5 {is preferred by Houdini. Having two central pawns is of course worth something.}) 11. Nd2 (11. O-O Nd7 $11) 11... Nf6 $14 {simply a waste of time. I was too concerned with avoiding having the two pawns in the center after an exchange on d5, which would a) not be so bad in itself, and b) require White to give up the two bishops.} (11... Nd7 {develops a piece instead.}) 12. Nc4 { the e5 pawn remains poisoned.} Qc7 13. c3 {Secures b4 and d4, but weakens d3. Castling immediately would be better.} (13. Nxe5 {still doesn't work, as I saw in the game:} Re8 14. Bf4 g5 $17) 13... Re8 {no more fooling around with tactical defenses for e5.} 14. O-O Nbd7 15. b4 Bf8 {a solid defensive move, bringing the piece back to help cover the king position.} 16. a4 Nb6 17. Bg5 Nxc4 {trading off White's superior knight.} 18. dxc4 Re6 {too artificial a move. I was looking to protect against doubling pawns on the f-file after Bxf6 and also to double rooks on the e-file, but there is nothing concrete gained by the move.} (18... Be7 {is the simpler choice.}) 19. Rfd1 h6 (19... Rae8) 20. Bxh6 $4 {computer handicap move.} (20. Bxf6 {is best according to Houdini, with the plan of following up with a gain of space on the queenside and the center.} Rxf6 21. a5 $14 Re8 22. c5 a6 23. Qe4) 20... gxh6 $19 21. g4 $2 Bg7 ( 21... a5 $5 {would apply some prophylaxis to the queenside, where White has some counterplay potential with his mass of pawns.}) 22. a5 Rd8 {Black starts to look to exchange down and simplify, with the material advantage.} (22... e4 {instead would shut out the Bg2 from the action.}) 23. Rdc1 $6 {perhaps "Turk" is looking for swindling chances, keeping both rooks on the board. Letting Black have the d-file is a problem, however.} Red6 24. Rd1 {deciding the d-file must be challenged after all.} (24. a6 {is no salvation, notes Houdini.} b6 25. Rd1 e4 26. Rxd6 Qxd6 $19) 24... Rxd1+ 25. Rxd1 Rxd1+ 26. Qxd1 Qd7 (26... e4 $5 {again would help neutralize the Bg2;}) (26... a6 {would be a good prophylactic move.}) 27. Qxd7 Nxd7 {Black still has a winning material advantage, but now White springs into activity with counterplay.} 28. a6 b6 { I had thought this would be sufficient to stem the tide on the queenside.} ( 28... bxa6 $142 $5 29. Bxc6 Nb8 $19 {is Houdini's line. A sample continuation:} 30. Be4 a5 {this is the idea, to undermine and attack the pawn group.} 31. c5 axb4 32. cxb4 Na6) 29. Bxc6 $17 Nb8 30. Bb7 Bf8 {not the best diagonal for the bishop, although it guards c5.} (30... Bf6 {keeps the bishop more active and with the threat of winning the c3 pawn after the e4 advance.}) 31. f3 f6 (31... e4 $5 $17 {I had seen this possibility earlier with the bishop on g7, but discarded it. The idea of moving Bg7 and then winning the c3 pawn is, remarkably, still the best.}) 32. Kf2 $15 Kf7 33. c5 $2 (33. Ke3 $11) 33... bxc5 $19 34. b5 Nxa6 $4 {here I didn't calculate correctly and thought that White would queen a pawn if the knight didn't prevent it.} (34... Nd7 {would have given Black the upper hand, notes Houdini. I had been concerned about the b6 advance clearing the way for the a-pawn; however, of course this would not work with the knight on d7.}) 35. bxa6 $11 {and now the opposite-color bishops ensure that it's a draw.} Kg6 36. Kf1 Kg5 37. Bc8 Bd6 38. h4+ Kxh4 39. Kg2 h5 40. gxh5 Kxh5 41. Kf2 Kg5 42. Ke3 Bc7 43. Bb7 Ba5 44. c4 1/2-1/2