09 April 2012

Annotated Game #39: Rook Endings Are Always Drawn

The next tournament game, following on the heels of my most instructive loss, is typical of my play during that period of my chess career.  The opening is an English Four Knights with a slightly advantageous, or at least comfortable, position for White once the early middlegame is reached.  I don't have a particularly good idea of where to place my pieces, for example the bishop on move 12, the rook on move 14 and the queen on move 19.  These sorts of small positional errors accumulate and when White follows a skewed plan to push e4 using the f3 pawn as a support, tactics as a result appear for Black, who emerges up a pawn.

By the end of the exchanging sequence on move 26, we have a double rook endgame where Black has a central passed pawn and another one easily created on the queenside.  However, White is not lost, as his king is closer to the action and his rooks are active.  These sorts of endgames are notoriously difficult for the side with a material advantage to win; the move required for Black to win on move 35 would have taken a great deal of accurate calculation and nerves to play at the board.  White is able to exchange a pair of rooks and eliminate the central and queenside pawns in exchange for Black picking up two kingside pawns, which however means that Black's pawns will not be able to breach White's defenses.  By move 41 the draw has taken shape, with White's rook in an ideal position to prevent Black's king from getting into the fight.  By move 51 the draw is concrete, although it takes another 30 moves for my stubborn opponent to concede the fact (amusingly with both of us missing a threefold position repetition as well).

The main lessons I drew from the analysis of this game were:
  • Knowing an opening variation well is not enough, as some idea of the early middlegame requirements for the position is necessary to maintain momentum and accuracy of play.
  • Many times it is sufficient simply to follow a plan of improving the position of each of your pieces in order to increase their activity.  This is especially a good rule of thumb for when you know the opening well but lack experience with the resulting middlegame.
  • The defender in a rook endgame should never despair when he has active pieces that can get behind the enemy pawns and harass the king.  After all, all rook endgames are drawn, as the saying goes...

[Event "?"] [Site "?"] [Date "????.??.??"] [Round "?"] [White "ChessAdmin"] [Black "Class B"] [Result "1/2-1/2"] [ECO "A28"] [Annotator "ChessAdmin/Fritz/Houdini"] [PlyCount "161"] [EventDate "2003.??.??"] [TimeControl "240+2"] {A28: English Opening: Four Knights Variation} 1. c4 e5 2. Nc3 Nf6 3. Nf3 Nc6 4. e3 d5 5. cxd5 Nxd5 6. Bb5 {this is the key move in the variation, which changes the game from being a reversed Sicilian, and an important reason why I play the variation with 4. e3.} Nxc3 7. bxc3 Bd7 8. O-O Bd6 9. d4 exd4 10. cxd4 (10. exd4 {is rated roughly equal to this by Houdini and it's probably a matter of taste which is played.}) 10... O-O 11. Bd2 {now out of the database, with this simple developing move. Here a variety of moves have been played, with e4 being the favorite and h3 played by two GMs.} (11. h3 Qe7 12. Bd3 Nb4 13. Bb1 f5 14. Bb2 Nd5 15. Ne5 Bxe5 16. dxe5 Be6 17. a4 Rfd8 18. Qf3 g6 19. e4 fxe4 20. Bxe4 c6 21. Qg3 Qf7 22. f4 Bf5 23. Rae1 Re8 24. e6 Qc7 {1-0 Franco Ocampos,Z-Mompo Ballester,V/Mislata 1994/EXT 2000 (24)}) 11... Ne7 (11... Bf5 { the engines agree this is best for Black, seizing a key diagonal and preventing Rb1. It looks anti-positional because of the pawn structure weakness after} 12. Bxc6 bxc6 {but Black's bishop pair and piece activity more than compensate.}) 12. Bxd7 (12. Bd3 {is more patient and prepares Rb1 while seizing the b1-h7 diagonal.}) 12... Qxd7 13. Qb3 b6 14. Rfd1 {this puts the rook on what is currently a less-than-useful file.} (14. Rac1 {places the other rook on its most promising file and postpones deciding where the Rf1 should go.}) 14... Qf5 15. Bb4 Bxb4 16. Qxb4 Rfe8 17. Rac1 h6 {this doesn't appear to do much that's currently useful for Black.} (17... Qe6 $5) 18. Qa3 { the queen needs to be moved to a more useful square, this puts pressure on the a-file and helps cover d3.} Nd5 19. Qb3 {this negates one of the points of having played Qa3 in the first place, the queen exerting pressure on the a-file. Now Black is able to move his rook to a more useful square.} (19. Ne5 $5) 19... Rad8 20. Ne5 a5 21. f3 {White sees the value of pushing e4 but stubbornly refuses to move the rook, instead weakening his pawn structure and the e3 square.} (21. Re1 Qe6 $14 22. f4 c5 23. dxc5) 21... Qe6 (21... a4 { the engines have no problem spotting the weakness on e3 and this deflection tactic.} 22. Qa3 (22. Qxa4 Nxe3) 22... Nxe3 23. Qxe3 $11 f6 24. f4 fxe5 25. dxe5 c5 $11) 22. a3 {cleverly thinking this prepares the way for e4 by taking away the b4 square from the Nd5. This has tactical problems, however.} (22. e4 {is in fact best, as now there are no longer tactics involving Nxe3 and Black doesn't have time to undermine the d4 support point for the Ne5.}) 22... f6 ( 22... c5) (22... a4) 23. e4 {this just loses a pawn, reflecting a counting error on White's part (I used to be bad at counting tactics and it's still an area for improvement.)} (23. Nc4 {simple is best.} a4 24. Qb1 $11 {as} Nxe3 { fails to} 25. Re1) 23... fxe5 $17 24. exd5 Qxd5 25. Qxd5+ Rxd5 26. Rxc7 exd4 27. Kf2 b5 (27... Rc5 {would increase Black's advantage.} 28. Rb7 Rc2+ 29. Kg3 Ree2) 28. Rdc1 {doubling rooks here unfortunately does nothing for White.} (28. Rd3 {immediately blockades the d-pawn and makes Black's life harder.}) 28... Kh7 (28... Re3 {is the nastier response that Black missed.}) 29. Ra7 b4 (29... d3 {passed pawns must be pushed!} 30. Re1 Rc8 31. Rxa5 Rc2+ 32. Ke3 {forced, in order to prevent having to give up the rook for the d-pawn.} Rxg2 $19) 30. Rcc7 {so after a couple of inaccuracies from Black, White gets an (admittedly easily parried) threat going.} Rg8 31. axb4 axb4 32. Rab7 d3 33. Ke1 {this looks best, in order to stop the d-pawn, but eventually would lose with best play on Black's part.} (33. Rd7 {is what the engines show after some calculating, although with Black evaluated as still a pawn to the good.} Rxd7 34. Rxd7 $17) 33... d2+ 34. Kd1 b3 35. Re7 {otherwise Re8 and then Re1 wins for Black.} Rd3 {this move only draws.} (35... Rc8 {wins, although it would be relatively difficult for a human to submit themselves to the next sequence.} 36. Rxg7+ Kh8 37. Rh7+ Kg8 38. Rhg7+ Kf8 39. Rbf7+ Ke8 40. Re7+ Kd8 41. Rg8+ Kxe7 42. Rxc8 Rb5 43. Kxd2 b2 $19) 36. Re2 Rc8 37. Rxd2 Rxd2+ 38. Kxd2 Rc2+ 39. Ke3 Rxg2 40. Rxb3 Rxh2 41. Rb7 {despite Black's passed h-pawn, this should be a draw, since White's rook is ideally placed to harass Black's king and pressure the pawns from the side.} h5 42. f4 Kg6 43. Kf3 Ra2 44. Rb6+ Kh7 45. Kg3 (45. f5 {seems to resolve things quicker for White.}) 45... Ra1 46. Kg2 g6 47. Rb7+ Kh6 48. Rb6 Ra4 (48... Kg7 49. Rb7+ Kf6 50. Kg3 $15) 49. Kg3 (49. f5 Kg5 50. fxg6 Rg4+ 51. Kh3 Kh6 $11) 49... Rd4 (49... Ra3+ 50. Kg2 $15 {would keep alive Black's chances, at least temporarily.}) 50. Ra6 (50. f5 {White keeps missing this idea.} h4+ 51. Kf3 Kg5 52. Rxg6+ Kh5 $11) 50... h4+ (50... Rd3+ 51. Kg2 $15) 51. Kg4 {White now has a completely drawn position.} h3 52. Ra3 h2 53. Ra1 Rd5 54. Rh1 Rd2 55. Kh4 Re2 56. Kg4 Rd2 57. Kh4 Rc2 58. Kg4 Ra2 59. Kh4 Kg7 60. Kg3 Ra3+ 61. Kg4 Ra2 62. Kg3 Kf6 63. Rxh2 Ra4 64. Kg4 {there's now no point at all for Black to play on.} Ra1 65. Rb2 Rg1+ 66. Kf3 Rf1+ 67. Kg4 Rg1+ 68. Kf3 Kf5 69. Rb5+ Kf6 70. Rb2 {Threefold repetition, missed by both players.} Rd1 71. Rb6+ Kg7 72. Rb5 Kh6 73. Ra5 Rd7 74. Kg4 Rd1 75. Ra6 Rf1 76. Rb6 Rg1+ 77. Kh4 Re1 78. Kg4 Kg7 79. Kg5 Rg1+ 80. Kh4 Rh1+ 81. Kg4 1/2-1/2

No comments:

Post a Comment