Hi everyone.
Does anyone want to be REAL NICE and help me analyze this game?
Executive summary: I lost an endgame I should have won. After long postmortem, I think I've discovered a 22 move multiple-queen-sac forced win if I hadn't blundered on move 33, but no chance I could have found this over the board, so I need a rule or principle to keep me from makign the same blunder in the future and I hope someone knows one, pretty please. :-)
Any ideas what went wrong here? I was black. I beat my opponent tactically around the board in the middlegame, and simplified to a king and pawn endgame with an extra two pawns and no pawn structure problems whatsoever. I should have won, right? Admittedly, my king was rather less centralized, as a casualty of the rush to simplify. But still. Two extra pawns. Pawn majorities on both sides. I should have won.
I've just started reading Nimzovich. Maybe that'll help.
1. e4 c5 2. Nf3 Nc6 3. c3 e6 4. Bb5 b6 5. Bxc6 dxc6 6. d3 Ba6 7. d4 cxd4
8. cxd4 Bb4+ 9. Bd2 Bxd2+ 10. Qxd2 Ne7 11. Qc2 O-O 12. Nc3 Ng6 13. g3 Qf6
14. Ne5 Nxe5 15. dxe5 Qxe5 16. f4 Qd4 17. Rc1 Bd3 18. Qd2 Rfd8 19. Rd1 Bxe4
20. Qxd4 Rxd4 21. Rxd4 Bxh1 22. Rd7 c5 23. Kf2 Kf8 24. Rd1 Bc6 25. Ke3 Ke7
26. Ne4 Bxe4 27. Kxe4 Rd8 28. Rxd8 Kxd8 29. Ke5 Kc7 30. h4 a5 31. g4 b5
32. f5 exf5 33. gxf5 Kd7 34. Kd5 Ke7 35. Kxc5 Kf6 36. Kxb5 Kxf5 37. Kxa5
Ke6 38. Kb5 Kd5 39. a4 f5 40. a5 f4 41. a6 f3 42. a7 f2 43. a8=Q+ Ke6 44. Qf3
{Black resigns} 1-0
(I'm really not sure why he so readily conceeded the 7th rank on move 24. I was gearing up for a long hard fight to get rid of that rook without losing any pawns. Although with the bishop on that diagonal, I could have rid myself of it fairly easily with the loss of only one pawn by ... Rb1, Rxa2 Rb2 at any point...)
So... what was the game losing blunder? My prime candidate for the real lulu is 38. ... Kd5, wasting a tempo AND putting the king in a nice spot where he could queen with check. I'm not sure what posessed me to do that. But, even without that, it still looks drawish. I don't know that I could do better than perpetual check after 39. Kc4 Ke5 40. Kd3 Kf4 41. a4 Kg3 42. a5 f4 43. a6 f3 44. a7 f2 45. a8=Q f1=Q+ 46. Kc3 Qe1+ 47. Kd3 Qd1+ 48. Kc4 Qc2+ -- and I can't force a queen trade, because then his outside passed pawn finishes me off.
So maybe my problem is earlier? 33. ... Kd7? 33. ... Kc6 would have prevented him from entering, but wouldn't I have been in a pinch of zugzwang then? Although maybe 34. h5 c4 35. Kd4 Kd6 36. a3 h6 and ... am I ok? Hell if I know. I don't even begin to know how to analyze these positions beyond who has the opposition? So I guess, since his next move is obviously a king move, I'd have to make whatever move keeps the opposition. 37. Kc3 Ke5 37. Ke3 Ke5 37. Ke4 Kc6 ? that last one's the toughie. 38. Ke5 Kc5? then 39. he goes back to Kc4 and it looks like a draw. I can't even pass a pawn: 39. ... c3 40. bxc3 b4 41. axb4+ axb4 42. cxb4+ Kxb4 and the stupid check on 41 costs me my passed pawn. Although... maybe 39. ... b4? Fails to 40. a4 c3 41. Kd3 cxb2 42. Kc2. Hmm... Then do I queen that pawn and force him to recapture on b1, then go gobble the kingside pawns? Ok, THAT at least looks like a winning line. 42. ... b1=Q+ 43. Kxb1 Kd5 44. Kc2 Ke5 45. Kd3 Kxf5 46. Kc4 Ke5 47. Kb5 . Kxa5 f4 49. Kb5 f3 50. a5 f2 51. a6 f1=Q+ And then, just to be pretty, sac the queen on a6 and promote the b pawn. 52. Kb6 Qxa6+ 53. Kxa6 b3 54. Kb6 b2 55. Kc6 b1=Q 0-1
okie... there's my forced win. I just figured this out as I'm writing it, over about half an hour with winboard. 22 moves from my blunder, with best play as far as I can determine (and I'm probably missing something), including the creation and sacrifice of two queens before the third one survives to deliver the win. 33. ... Kc6 34. h5 c4 35. Kd4 Kd6 36. a3 h6 37. Ke4 Kc6 38. Ke5 Kc5 39. Kc4 b4 40. a4 c3 41. Kd3 cxb2 42. Kc2. 41. Kd3 cxb2 42. Kc2. b1=Q+ 43. Kxb1 Kd5 44. Kc2 Ke5 45. Kd3 Kxf5 46. Kc4 Ke5 47. Kb5 . Kxa5 f4 49. Kb5 f3 50. a5 f2 51. a6 f1=Q+ 52. Kb6 Qxa6+ 53. Kxa6 b3 54. Kb6 b2 55. Kc6 b1=Q 0-1
Needless to say, even if I'm right, there's no way I could possibly have done all this calculation over the board, even though it was a 45 minute FICS game with a 12 second increment, and I basically had 40 minutes to do it in. Not without cheating and seeing the positions on a real chessboard and/or winboard, at least. I'm not Alekhine. Soooo... obviously, I made a mistake somewhere that wasn't sheer calculation. I had more pawns, I should have won. I blew a principle somewhere. Was it the rook trade? Should I have not simplified with my king so far from the center? Problem is, I saw no other way to get his rook off the only open file.
So... How should I have known to play 33. ... Kc6 without the advantage of the last half hour's winboard-aided calculation? What principle did I forget, or outright not know, or knowing, unconsciously violate? Someone? Help?
While I'm a bit too lazy right now to type out actual variations, I think your mistake was leaving your queenside pawns and heading to the kingside. You tried to do too much with your king, and you didn't put your position's strength (your two pawn majorities) to good use. The fact that your king isn't well centralized isn't terribly important, because you can create a passed pawn that forces your opponent to retreat his king. You should have first used your queenside majority to create a passed pawn. Next, make a passed pawn with your kingside majority. The win from here is easy; his king can't stop both threats. I hope this isn' t too confusing; maybe later I'll get around to giving an actual variation.
By the way, the rook trade, on your part, was a good idea. It was bad for him to accept it. Rook and pawn endgames are tougher to win than pawn endgames, so he should have kept the rooks.
Alright, I'll get a little more into the specifcs here. One of your first mistakes in this endgame (although it went unpunished) was 30...a5. Your opponent should have quickly played 31.a4 here. Now you can't push your b pawn, and so this majority loses some of its effectiveness. You should still win, but you don't want your opponenet to be able to make it more difficult. You should have played b5 first, then followed that up with a5.
Likewise, your opponent's pawn pushes on the kingside were all wrong. All they did was make his pawns weaker and even easier for you to get a passed pawn.
Also, all of your analysis seems to center around the getting the opposition. The opposition is mainly important in endgames where the material is a bit more balanced. You have two pawn majorities so the opposition carries a lot less relevance. Remember, opposition is only a means to an end, not the end itself.
Now, to the critical line, beginning with 33.gxf5. What you opted for in the game was a race. There's no need for a race here. You have both majorities, your opponenet won't get a passed pawn unless you let him. The win might look something like this: 33...Kc6 34.Ke4 (pawn moves don't do him much good right now) g6 35. hxg6 hxg6 36.fxg6 fxg6 (If 35.h6, then 35...g5 wins easily) 37.Ke5 g5 38.Kf4 Kd5, and from here the win is pretty obvious.
aaah. thank you very much. that, unlike my play in that endgame, makes a lot of sense. Obviously, I need to do a little more study in massed pawn endgames. Most of my K+P endgame studies have been with one pawn. Wasn't really sure about the proper approach to take with a bunch of pawns.
muchos gracias.