Annotated Game #186: All rook endgames *should* be drawn

This last-round tournament game is another illustration of why persistence and active play can pay off (even if it does not actually, in this case).  Some relative weak opening play in an unfamiliar position leads to a failure to falsify a key move, which lands me significant material down.  Deciding that there was at least some hope for a kingside attack and pressure in compensation, I continue playing and the initiative shifts to me, despite being objectively lost.  Eventually my opponent can't take any more pressure and simplifies to what should be a drawn rook endgame...which isn't, however, in the end.  A good lesson on weak pawns, rook activity and other elements of rook endgames...which apparently all really should be drawn.

[Event "?"] [Site "?"] [Date "????.??.??"] [Round "?"] [White "ChessAdmin"] [Black "Class A"] [Result "0-1"] [ECO "A25"] [Annotator "ChessAdmin/Komodo 11.2"] [PlyCount "184"] {[%mdl 8192] A25: English Opening vs King's Indian with ...Nc6 but without early d3} 1. c4 e5 2. Nc3 Nc6 3. g3 Nf6 4. Bg2 Bb4 {the second most common move in the position, according to the database. The bishop needs to be developed, but this doesn't have as much bite as when the d-pawn has already advanced, since there's no pin on the Nc3.} 5. Nd5 {the standard reaction.} Nxd5 6. cxd5 Ne7 7. a3 (7. Nf3 {is played most here and scores quite well, over 72 percent.}) 7... Bc5 8. b4 Bb6 9. e3 {I'm playing a lot of pawn moves here and neglecting development. The idea is to bring the knight out via e2 rather than f3, to avoid it being harrassed by the e-pawn. It's rather slow, however.} (9. Bb2) 9... d6 10. Ne2 O-O 11. Bb2 Bg4 12. Qb3 $146 {inviting the minor piece trade on e2, which I think would be better for me. The king would be safe and my light-square bishop would then be unopposed.} (12. d4 {is the move that is begging to be played here.} f6 13. h3 Bf5 14. e4 Bg6 15. O-O Rc8 16. Kh2 Re8 17. f4 c6 18. dxe5 fxe5 19. fxe5 cxd5 20. exd6 Qxd6 21. Nf4 Bf7 22. e5 Qd7 23. Qb3 g5 24. e6 Bxe6 25. Nxe6 Qxe6 26. Rae1 Qd6 {Butala,M (2246) -Banic,S Ljubljana 2001 1-0}) (12. h3 $5 {also looks fine.}) 12... Qd7 { obviously aimed at exchanging off the Bg2 and taking advantage of the resulting light-square weakness on the kingside.} 13. h3 Bh5 (13... Bxe2 14. Kxe2 f5 15. f4 $11) 14. f4 $6 {this is not terrible, but not ideal either. It weakens the king position and now the exchange on e2 is better for Black.} (14. g4 Bg6 15. f4 $11 {is an improved version of the idea. My knight would be happy to go to f4 after an exchange.}) 14... Bxe2 15. Kxe2 f6 {blunting future potential threats on the long diagonal.} 16. Raf1 $6 {here I completely miss Black's imminent threats. The g3 square is now unprotected, thanks to the f-pawn advance.} Nf5 {with a simple fork threat that I handle terribly.} 17. Rf3 $4 (17. Kf2 {I'm uncomfortable here and will have to focus on defense, but the engine considers the position equal.} Rae8 18. Re1 $11) 17... e4 $19 { what explains not seeing this response to my last move? Probably over-focusing on the tension between the e5 and f4 pawns and looking at the exchange possibilities there. Also not seriously focusing on my opponent's possibilities in order to falsify my move.} 18. Rf2 Nxg3+ 19. Ke1 Nxh1 20. Bxh1 {although an exchange and a pawn down, I decide to fight on. Black's kingside is looking a little open now and I thought my only chance would be to try to attack.} Rae8 21. Rh2 (21. f5 $19 {is the engine's improvement, preventing Black from occupying the f5 square and seizing more space. The square f4 is also now available for a rook transfer.}) 21... Qb5 22. Qd1 $2 {another bit of territory lost, notes Komodo via the Fritz interface.} (22. a4 {would be an opportunity to regain some material.} Qd3 23. Qxd3 exd3 24. a5 Bxe3 25. dxe3 Rxe3+ 26. Kd2 $17) 22... Qxd5 23. Qg4 Qe6 {by this point I really am lost.} 24. f5 {another desperate move, to prevent a queen exchange.} Qf7 {here my opponent starts losing the thread, playing very conservatively instead of putting the final nail in my coffin.} (24... Qa2 $5 25. Rg2 Qb1+ 26. Kf2 Re7 $19) 25. Bxe4 {A bit of material back and a psychological victory, even if it doesn't change the objective evaluation.} d5 (25... Qb3 {my opponent still does not see the strength of this penetration idea with the queen.} 26. Rg2 Re7 $19) 26. Bd3 c6 27. Rg2 Bc7 28. Bd4 b6 29. h4 {although I'm still losing badly, at least it's nice to have some initiative. Watching the h-pawn grind towards him made my opponent react sub-optimally.} h5 {this stops the forward motion of my h-pawn, but also leaves the h5 pawn without much support.} 30. Qf3 Be5 { the best move, exchanging off one of my good pieces and getting closer to an easily won endgame.} 31. Bxe5 Rxe5 32. Rg6 {blocking the h5 pawn's protection.} Qe7 {a direct approach, doubling on the e-file, but not the best.} (32... c5 $5 $19 {would get Black's pawns rolling and support d5-d4.}) 33. Qxh5 Rf7 { my opponent continues to react conservatively to my threats, giving me a bit of hope.} 34. Qg4 Re4 $2 {my opponent decides to simplify rather than endure further kingside pressure. Komodo now evaluates the position as equal.} (34... c5 {is still a good option.} 35. h5 c4 $19 {and now ...c3 is threatened, undermining support for e3.}) 35. Bxe4 $11 Qxe4 36. Qxe4 dxe4 {we now have a rook endgame that is perfectly fine for me.} 37. Rg4 {conservative play.} (37. h5 $5 {the engine identifies the correct plan, which is to exchange off the weak h-pawn. It doesn't have to be done immediately, but there's no reason to wait. Unfortunately, I did not identify this as a strategic need.} Kh7 38. Kf2 Rd7 39. h6 Rxd2+ 40. Kg3 gxh6 41. Rxf6 h5 42. Re6 $14) 37... Re7 38. Ke2 Re5 39. Rf4 Kh7 40. d4 (40. Rg4 {was probably the best route to a draw, just shuffling the rook.} Kh6 (40... Rxf5 41. Rxe4 $11) 41. Rg6+ Kh7 (41... Kh5 $6 42. Rxg7 $14) 42. Rg4) 40... exd3+ 41. Kxd3 {while the engine still rates it as equal, my job is now more difficult, with more open lines in the center and a backward e-pawn.} Rd5+ 42. Ke2 {I pick the wrong side, concerned about my weak pawns. Now Black can more easily activate his queenside majority.} (42. Kc3) 42... c5 43. e4 (43. b5 Kh6 $11) 43... Rd4 44. Ke3 Rd1 {things are still equal, if uncomfortable for me. Black has the iniative.} 45. bxc5 {not an objectively bad move, but it gives me considerably fewer options to combat Black's queenside threats.} (45. Ke2 $5 {is much more flexible, harrassing Black's rook and allowing my rook space on the third rank.}) 45... bxc5 46. Ke2 $6 (46. Rf3 {is the only move that holds equality and is not obvious to find.}) 46... Rd4 $15 47. h5 {now I start crumbling, not having any better ideas.} (47. Ke3 {this is supposedly better according to the engine, but it still looks difficult.} Ra4 48. Rf2 Rxa3+ 49. Kf4 $15) 47... Rc4 (47... a5 $17) 48. Rf3 $2 (48. Kd3 {and White could well hope to play on, comments Komod.} Rd4+ (48... Ra4 49. Rg4 Rxa3+ 50. Kc4 $11 {Black will not be able to make progress on the queenside with separated pawns and the rook in front, versus my king close at hand.}) 49. Kc2 $11) 48... Rxe4+ $19 {now all of my pawns are weak, vulnerable and isolated. Black has a winning game now.} 49. Kf2 Rh4 50. Kg3 Rxh5 51. Kg4 Rg5+ 52. Kf4 g6 53. fxg6+ Kxg6 54. Rc3 Re5 55. Rc4 Rd5 56. Ke3 f5 57. Ra4 { the best chance, with a temporary defense along the fourth rank, but my opponent correctly takes the time to reset his rook position.} Rd7 58. Ra6+ Kg5 59. Rc6 {I figured I needed activity and threats to have any sort of chance to resist.} (59. Ra4 {is safer but still very much losing.}) 59... f4+ 60. Ke2 Re7+ 61. Kf2 Re5 62. Rc8 {with the idea of pursuing an "annoying rook checks" strategy to harrass my opponent.} a5 63. Rg8+ Kf5 64. Ra8 c4 65. Rc8 Re4 66. Rc5+ Re5 {this was an unnecessary concession by my opponent.} (66... Kg4 67. Rc6 (67. Rxa5 f3 $19) 67... Re3 68. Rxc4 Rxa3 $19) 67. Rxc4 {I began to have a bit of hope again now.} Rb5 $6 {this allows the next move with tempo.} (67... Kg4 $5 68. Rc8 a4 $19) (67... Rd5 $5) 68. a4 $17 Rb2+ $2 {now according to the engine, I can draw.} 69. Kf3 {Exerts pressure on the isolated pawn} (69. Ke1 $5 ) 69... Rb3+ 70. Ke2 Re3+ 71. Kf2 Re4 {Black threatens to win material: Re4xc4} 72. Rc5+ {here I thought I had to avoid the rook trade, not having fully calculated out the resulting K+P endgame.} (72. Rxe4 Kxe4 73. Ke2 {will also draw. For example} Kd4 74. Kf3 Kc3 75. Kxf4 Kb4 76. Ke3 Kxa4 77. Kd2 Kb3 78. Kc1 Kc3 79. Kb1 $11) 72... Re5 {Black threatens to win material: Re5xc5} 73. Rc3 $2 {this unnecessarily limits my rook movement and loses to Black's next.} (73. Rxe5+ {works, similar to the above variation.} Kxe5 74. Kf3 $11) (73. Rc1 $11 {and now my rook can check on g1 if Black's king goes to g4.}) 73... Kg4 74. Rc8 Re4 75. Rg8+ Kf5 76. Rf8+ Kg6 77. Kf3 Rxa4 $19 78. Rc8 Kf5 79. Rf8+ $2 Ke5 80. Re8+ Kd5 81. Rd8+ Kc6 82. Rh8 Rd4 83. Rh5 Kb6 84. Rh6+ Kb5 85. Rh5+ Kb4 86. Rh8 a4 87. Rb8+ Ka3 88. Rb7 Rb4 89. Rd7 Kb2 90. Ke2 a3 91. Rd2+ Kb3 92. Rd3+ Ka4 (92... Ka4 93. Rd1 a2 94. Rd8 Ka3 95. Rd3+ Rb3 96. Rd1 f3+ 97. Kf2 Rb1 98. Rd3+ Kb4 99. Rd4+ Kc3 100. Ra4 a1=Q 101. Rxa1 Rxa1 102. Kxf3 Ra4 103. Ke3 Kc2 104. Kf2 Kd2 105. Kf3 Rd4 106. Kg2 Ke3 107. Kg3 Ke2 108. Kg2 Rd3 109. Kh1 Kf2 110. Kh2 Rb3 111. Kh1 Rh3#) 0-1