Annotated Game #356: Cramped to death in the endgame

This next tournament game features what should have been a comfortable draw as White in a King's Indian Defense (KID) by transposition - with some opportunities for advantage passed up along the way - but which devolved into a highly cramped loss. The key error, if not the final losing one, was exchanging queens on the c-file to reach a cramped minor piece ending. This was simply unnecessary and reflects a repeated tendency I have to worry excessively about an open file with an opponent's major piece (queen/rook) on it, even if I have all the penetration squares effectively under control. My opponent does an excellent job of demonstrating how squeezes can happen in endgames, although I had opportunities to relieve the pressure with an f-pawn break. Very instructive.

[Event "?"] [Site "?"] [Date "????.??.??"] [Round "?"] [White "ChessAdmin"] [Black "Class A"] [Result "0-1"] [Annotator "ChessAdmin/Dragon 3.2"] [ECO "E61"] [WhiteFideId "-1"] [BlackFideId "-1"] [PlyCount "78"] [GameId "2323483478960775"] 1. d4 Nf6 2. Nf3 g6 3. e3 Bg7 4. Be2 O-O 5. O-O d6 6. c4 Nbd7 7. Nc3 e5 8. Qc2 c6 9. b3 Re8 10. Re1 $6 $11 {this is simply an unnecessary precaution. White can do several other useful and more urgent things here.} (10. Ba3) (10. dxe5) (10. h3) 10... h6 11. Ba3 {now this is less effective.} (11. Bb2) (11. a4 $5) 11... Qc7 $6 {it looks perfectly reasonable to defend d6 in this way, but it gives White a freer hand in the center.} (11... e4 12. Nd2 Qe7 $11) 12. Rac1 (12. Rad1 {I also considered and the engine prefers. Instead of opposing the queen on the c-file, the rook's influence down the d-file can be felt more quickly.}) (12. h3 $5 {it's not a bad idea to get some luft for the king and take away the g4 square from Black.}) 12... a6 {now White has multiple strategic options available.} 13. d5 (13. dxe5 {there is often a tension between playing this and the text move in the KID setup. Here the engine prefers to exchange.} dxe5 (13... Nxe5 $6 14. Red1 $16) 14. c5 $14 {this is one pleasant possibility for White, constraining Black and controlling d6 and b6.}) 13... c5 $6 {this essentially justifies White's choice to advance, as now the d5 pawn is much stronger.} 14. Nd2 $14 {not a bad move, but} (14. e4 {is stronger. I've noticed a pattern in which I am reluctant to play e3-e4 when it is favorable.}) 14... Nh7 $6 {this is too passive.} 15. b4 {no pawn break, no plan. Although my edge is still small, I have the initiative on the queenside.} f5 $6 {consistent with how KID players normally want to advance on the kingside, but the action is currently on the queenside.} 16. bxc5 $16 Nxc5 17. Nb3 $6 {at the time, it seemed reasonable to bring the knight over to increase the pressure, but it is easily negated by Black's next move. I needed to strike immediately with} (17. Bxc5 Qxc5 (17... dxc5 18. e4 $16) 18. Na4 $1 $18 {with follow-up ideas like Qb3, Rb1 and c4-c5 to keep things rolling and dominate Black's backward queenside.}) 17... b6 $1 $11 18. Nxc5 {the point of bringing the knight over, but after} bxc5 {White's initiative on the queenside is gone and we are back to equality.} 19. Rb1 (19. e4 $5) 19... Bd7 20. Rb3 Rab8 21. Reb1 e4 22. Rxb8 {White has no prospect for obtaining an advantage, so I decided to head for a drawn endgame.} Rxb8 23. Rxb8+ Qxb8 24. Qb1 $6 {while not disastrous, this still gives Black more play than I should allow. I have made similar types of mistakes with heavy pieces on open files before, when my opponent stands better after the (unnecessary) exchange.} (24. Bc1 $11) 24... Qxb1+ 25. Nxb1 Ba4 {objectively Black is not really better, but certainly has the easier game, while I am cramped.} 26. Bc1 Ng5 27. Bd2 (27. f4 $5 $11 {this was probably a good time to make my position a little more flexible.} exf3 28. gxf3 $11) 27... Nf7 28. Ba5 (28. f4) 28... Kf8 29. Kf1 Ne5 30. Ke1 $6 {now my refusal to use the f-pawn lever is starting to make things worse.} (30. f3) 30... Bc2 $15 31. Nd2 $2 {unfortunately I am too cramped after Black's next, and he gains a winning advantage.} (31. Na3 $1 {this "only move" would leave the d2 square open for the king, after ...Nd3+}) 31... Nd3+ $1 $19 32. Bxd3 Bxd3 {the correct choice. Now I cannot do anything about the bishop on the d3 outpost and Black has a strategically won game.} 33. f3 Ke7 {good enough to maintain the advantage.} (33... exf3 $1 {my opponent probably thought that} 34. gxf3 Ke7 35. e4 {looked good for White, but now Black's kingside pawn majority will eventually win.}) 34. f4 g5 35. g3 gxf4 36. gxf4 Bf6 {unfortuantely, I have nothing left to do that can improve my position, while Black can further infiiltrate with his dark-square bishop. Eventually the c- or e-pawns will fall.} 37. Kd1 {this speeds up the loss a bit.} (37. Kf2 Bb2 $19 {and Black threatens to either exchange off the Ba4 and gain a 2v1 queenside majority, or exchange the Nd2 and then capture on c4.}) 37... Bh4 38. Bc3 Bf2 39. Nxe4 {pure desperation.} fxe4 0-1