Annotated Game #102: In which I fail to crush my enemy

In this first-round game of an online 45 45 tournament game in the Slow Chess League, I singularly fail to crush my enemy as I should have.  Out of the opening, I take advantage of an overloaded bishop on g7 to win a pawn and then pick up an exchange with a skewer.  After this, however, I must give full credit to my opponent for his strong resistance and constantly seeking active ways to make threats.  This eventually pays off as I fail to find winning ideas at key points and barely manage to avoid a mate threat.  At the end, low on time and with an uncertain endgame, I go into a threefold repetition.  While not the result I wanted, it was great for training purposes and points out how I need to be more steely in the face of danger (real or imagined).

[Event "Slow Swiss #8"] [Site "Chess.com"] [Date "2013.08.22"] [Round "1"] [White "ChessAdmin_01"] [Black "MathBandit"] [Result "1/2-1/2"] [ECO "A24"] [WhiteElo "1250"] [BlackElo "1450"] [Annotator "ChessAdmin/Houdini"] [PlyCount "105"] [EventDate "2013.??.??"] [TimeControl "45"] {A24: English Opening vs King's Indian: Lines without ...Nc6} 1. c4 Nf6 2. Nc3 g6 3. g3 Bg7 4. Bg2 d6 5. Nf3 O-O 6. O-O e5 7. d3 Na6 {the knight is normally better off on c6 or d7. The difference her is that it can go to c5 and then on to e6 after it's kicked by b2-b4.} 8. Rb1 Be6 $146 (8... c6 {is suggested by Houdini, focusing on the fight for the d5 square.} 9. Bg5 $11) (8... Nc5 $5) 9. b4 {White moves ahead with his queenside expansion plan.} Qd7 {the threat here is to exchange off the Bg2, which is doing much more for White than its counterpart is for Black. I decide to pre-empt this possibility with the next knight move.} 10. Ng5 $16 {this opens up an attack on the b7 pawn and threatens to exchange on e6. Houdini already assesses the position as a pawn up equivalent for White.} Rab8 11. Qc2 (11. Nxe6 fxe6 12. Qb3 {is Houdini's preference.}) 11... Bg4 12. b5 (12. h3 $5 {I thought a good deal about this possibility, but in the end decided it didn't give White enough. The engine favors it, however.} Be6 (12... h6 {was what I thought my opponent would likely play.} 13. hxg4 (13. Nxf7 {I did not consider but is evaluted favorably by Houdini.} Bxh3 14. Nxh6+ Kh7 15. Qd2 $16) 13... hxg5 14. Bxg5 Nxg4 {at the time didn't seem to promise much, but it compares favorably with the game continuation, as White obtains the two bishops and keeps the initiative.}) 13. Qb3 $16 {keeping a lock on d5.}) 12... Nc5 $14 13. Be3 h6 14. Nge4 {although exchanging on c5 as recommended by the engine makes positional sense, I was hoping to target h6 later on.} (14. Bxc5 dxc5 15. Nge4 Qe7 $14) 14... Ncxe4 15. Nxe4 b6 $2 {the obvious way of protecting the a7 pawn, but now White takes advantage of the overloaded Bg7 and picks up the h-pawn.} (15... Nxe4 {played first would prevent the tactic and lead to equality.} 16. dxe4 b6 $11) 16. Nxf6+ $16 Bxf6 17. Bxh6 Rfe8 {missing the skewer.} (17... Rfd8 $16) 18. Bc6 $18 Qe6 {here I thought for a while in order to make sure Black did not have any tricks involving the hanging Bh6.} 19. Bxe8 Rxe8 20. Qd2 (20. Be3 Bh3 21. Rfc1 {would have been a simpler way to continue.}) 20... Qf5 {Black intends e4, notes Houdini. The e2 pawn is weak and White ends up bringing his bishop back to e3 in any case.} 21. Be3 e4 22. dxe4 Qe6 $2 {this provides a free pawn for White and with that a stronger position in the center.} (22... Qxe4 23. Rb3) 23. f3 Bh3 24. Rfc1 Kg7 {the idea being to eventually be able to protect the Bh3 by moving the rook to h8. However, this allows White to force a bishop trade and move closer to realizing what should have been a victory.} 25. Bd4 Bxd4+ 26. Qxd4+ Kh7 27. Re1 {this is not a bad move in itself, but it represents the failure to find the correct winning plan and the start of White's problems. I was unsure how to finish Black off, so decided to shore up e2 in case the e-file was later opened (following ...f5, for example).} (27. Rb3 {is the easiest way for White to make progress. The rook can then go to a3 and Black cannot prevent the queenside breakthrough, while g4 is also threatened. One possible continuation is} Qe7 28. g4 f5 29. exf5 gxf5 30. Qf4) 27... f6 {Black renews his plan to bring the rook to h8.} 28. Rbc1 (28. Kf2 { was something I actively considered and probably should have played here. It defends e2 again, is a safer square for the king, and renews the threat to the bishop after g4.}) 28... Kg7 29. a4 (29. Rc3 {again would work, bringing the rook to the third rank in order to swing it to the a-file. I unfortunately failed to find this idea, instead focusing on how the file could be opened with the a-pawn.} Qc8 $18) 29... Rh8 30. Qf2 {I continue to play defensively and again while not bad in itself, it represents how White is losing the thread of the game. I need to activate my rooks and make threats against Black, not allow him to reorganize his pieces.} g5 31. g4 {I'm forced to make the best move here, finally.} Rh4 {this was a clever way of ensuring the bishop's protection, but of course has the drawback of walling the pieces off on the h-file. I start taking Black's kingside threats (sacrifice on g4) too seriously and make a weak queen move, when sidestepping with the king would have been best.} 32. Qg3 $2 {I had thought this was a solid defensive move, but then realized that it offers a free pawn to Black and an out for this bishop. This marked a major turning point, as White is still winning afterwards, but Black has many more practical chances.} (32. Kh1 $18 {and White wins, says Houdini.}) 32... Bxg4 33. Kf2 (33. fxg4 $2 Rxg4) 33... Bh5 34. Rg1 Rh3 35. Qg2 Be8 36. f4 (36. Ke1 $142 Bh5 37. Qf2 $18) 36... Kf8 $2 { I was too focused on defense, thinking this would force a queen exchange, to pick up on the obvious tactic.} (36... Rh4 37. fxg5 f5 38. Qf3 Rxe4 39. Rg2 $18 ) 37. Qg4 $2 {throws away the win.} (37. f5 Qe5 38. Qxh3 Qf4+ 39. Qf3 Qh4+ 40. Ke3 $18) 37... Rxh2+ $16 {an excellent in-between move spotted by my opponent, and very disheartening for me. We now head into an endgame where move choices are not at all obvious and my time on the clock is limited.} 38. Ke1 $6 { there's no need to give up the pawn.} (38. Ke3 Qxg4 39. Rxg4 $16) 38... Qxe4 $11 39. fxg5 Qe3 {also a surprise move for me. My calculating and thinking process had broken down under pressure. However, I find the only defensive move.} 40. Rc2 Qf2+ 41. Kd2 $2 {the king has two squares, I was unable to calculate well at this point and just picked this one on general principles. Black now has a win.} (41. Kd1 Bh5 42. Qg3 Qd4+ 43. Rd2 Bxe2+ 44. Kc1 Qa1+ 45. Kc2 Bd1+ 46. Rgxd1 Qa2+ 47. Kc1 Qa1+ 48. Kc2 Qa2+ 49. Kc1 Qa1+ 50. Kc2 $11) 41... Rh4 $2 {I breathed a sigh of relief after this, since it allowed me to get the queens off the board with a small tactic.} (41... f5 {is winning for Black, as the queen can no longer protect e2.}) 42. Rg2 $18 {Houdini evaluates this as winning for White again.} Rxg4 43. Rxf2 Rxg5 44. Rxf6+ Ke7 45. Rf1 { again I'm unsure of exactly where to go, but this should be OK.} Bg6 46. Rc3 { the correct idea, activating the rook on the third rank, but now I fail to follow up on it.} Rg4 47. Re1 $6 {passive.} (47. Re3+ Be4 48. Kc3 $18) 47... Rd4+ $16 48. Ke3 Re4+ 49. Kf3 Rh4 50. Kg3 {at this point I have 5 minutes on my clock to my opponent's 20 and am not interested in taking any risks in playing out a difficult endgame, so I go for the draw by repetition. Ways White could have played on:} (50. Rcc1) (50. Rg1) 50... Re4 51. Kf3 Rh4 52. Kg3 {Twofold repetition} Re4 53. Kf3 1/2-1/2