Annotated Game #32: A somewhat embarrassing draw

This next tournament game features an unusual variation of the English Four Knights, where Karpov's 4...Be7 is followed up by Black exchanging knights on d4.  White gets a pleasant plus out of the opening, which after a series of subsequent exchanges on d5 rapidly turns into an ending.  White's outside passed pawn gives him all the winning chances, but the double rooks and bishops mean that it won't be easy for him.  White fails to maintain the tension and exchanges off the outside passer for Black's d-pawn, essentially ensuring the draw for Black, as the resulting bishop ending with 4 vs. 3 pawns on the kingside is very easy to defend.

Useful points from the game analysis:

• The early knight exchange on d4 does not appear to challenge White in this variation.
• Provoking the series of exchanges on d5 and going into a double rook and bishop ending appeared to be the correct decision, due to the weakness of Black's isolated queen pawn.
• White should have developed his rooks earlier and seized the c-file, although Black ultimately lets him do this anyway.
• Silly 18th move by White was due to a poor thinking process and not examining his opponent's potential responses (i.e. failure to falsify the candidate move).
• Unwillingness to preserve tension in the position on move 30 (a common amateur error) led to the disappearance of White's winning chances.

After my initial look at the game, I'd felt quite embarrassed, since I thought I had thrown away an easy win.  In fact, the win was likely there, but it wasn't so easy to realize (at least for a non-endgame expert) and the final position was in fact drawn, despite White's one-pawn advantage.  So I'm now just somewhat embarrassed by the draw.

[Event "?"] [Site "?"] [Date "????.??.??"] [Round "?"] [White "ChessAdmin"] [Black "Class C"] [Result "1/2-1/2"] [ECO "A28"] [Annotator "ChessAdmin/Fritz/Houdini"] [PlyCount "90"] [EventDate "2002.??.??"] {A28: English Opening: Four Knights Variation} 1. c4 e5 2. Nc3 Nf6 3. Nf3 Nc6 4. e3 Be7 5. d4 exd4 6. Nxd4 Nxd4 (6... O-O {is what appears in the opening manuals. This is the first time I had seen the text move; it is the second most common in the database (albeit by a long distance) following O-O.}) 7. Qxd4 (7. exd4 {interestingly doesn't appear at all in the database.}) 7... O-O 8. Be2 d6 9. O-O c6 {Covers b5+d5, notes Fritz.} 10. b3 d5 (10... Be6 {is most often played here, although we are now down to a handful of games in the database. The only one in there with the text move was a loss.}) 11. cxd5 $14 { the d5 square is a key one in the English, so White should always challenge for its control.} Nxd5 12. Nxd5 Qxd5 {now out of the database.} 13. Qxd5 cxd5 { and we find ourselves in an ending. White has all of the chances, due to Black's isolated queen pawn, although Black is far from losing.} 14. Bb2 Be6 15. Bf3 {ignoring the need to develop the rooks (a common amateur problem), as was typical of my play at the time.} (15. Rac1 Rac8 16. Rfd1) 15... Rfd8 16. h3 {gives the king some space, but in the absence of any threats on the horizon, perhaps not the best choice.} Rd7 {a major positional error, allowing White to seize the c-file.} 17. Rac1 Rad8 18. Rc2 {a rather silly move, due to} Bf5 19. Rcc1 {the rook embarrassingly retreats, not wanting to give up the c-file.} b6 20. Rfd1 $14 {The isolani on d5 becomes a target, says Fritz. About time the rook got into the action, says me.} Bc5 21. a3 a5 {Black did not wish to allow b4, evidently.} 22. Bd4 {this idea culminates on move 25.} Bxa3 23. Ra1 { White gets the initiative, says Fritz.} ({Houdini considers the sequence} 23. Bxb6 Bxc1 24. Bxd8 Rxd8 25. Rxc1 Be4 26. Bxe4 dxe4 27. Ra1 {as better. Black cannot protect both of his advanced pawns from White's rook.}) 23... Bc5 24. Bxc5 bxc5 25. Rxa5 Kf8 (25... c4 {is found as the best defense by the engines} 26. Rdxd5 Rxd5 {Houdini finds the key sequence here} 27. Bxd5 c3 {and the c-pawn ensures Black an even game.}) 26. Rxc5 Be6 27. b4 $18 Rb7 28. b5 { passed pawns must be pushed!} Rdb8 29. Rb1 Bd7 30. Bxd5 (30. Be2 {was necessary to preserve the outside passed pawn and real winning chances.}) 30... Rxb5 31. Rcxb5 Rxb5 32. Rxb5 Bxb5 {now looking like a draw in practical terms.} 33. g4 g5 34. Kg2 Ke7 35. Kg3 f6 36. f4 h6 37. fxg5 {this simplifies matters for Black.} hxg5 38. h4 gxh4+ 39. Kxh4 Kd6 40. Ba8 Ke5 41. Bf3 Bd3 42. Kg3 Bb1 43. Kh4 Bd3 44. Kh5 Bb1 45. Kh6 Bd3 1/2-1/2