Author: Pete R.
Date: 10:09:05 06/13/00
Go up one level in this thread
On June 13, 2000 at 02:48:22, Will Singleton wrote:
>On June 12, 2000 at 17:39:15, Pete R. wrote:
>
>>On June 12, 2000 at 05:10:54, Peter McKenzie wrote:
>>
>>>I watched following game on ICC this evening. The game was funny because I'd
>>>just been thinking that Chop (Beadle) sacs the exchange too often and was
>>>thinking of ways to fix this problem. When I saw Rxa4 come along in this game I
>>>though 'oh now, here we go again', but as events unfolded Chop whipped up a cool
>>>attack. These are the sort of games that keep me interested in this silly hobby
>>>of computer chess :-) I don't think you'll find the Rf8-f5-d5-d1 rook walk in
>>>too many text books.
>>>
>>>I don't know if giving up the exchange was sound (analysis welcome), but it
>>>certainly was entertaining. Hard luck to Amateur (& Will), I'm sure revenge
>>>will be swift :-)
>>>
>>>[Event "ICC 5 2"]
>>>[Site "Internet Chess Club"]
>>>[Date "2000.06.12"]
>>>[Round "-"]
>>>[White "Amateur"]
>>>[Black "Beadle"]
>>>[Result "0-1"]
>>>[ICCResult "White resigns"]
>>>[WhiteElo "2498"]
>>>[BlackElo "2508"]
>>>[Opening "Three knights: Steinitz variation"]
>>>[ECO "C46"]
>>>[NIC "KP.06"]
>>>[Time "04:21:45"]
>>>[TimeControl "300+2"]
>>>
>>>1. e4 e5 2. Nf3 Nc6 3. Nc3 g6 4. d4 exd4 5. Nxd4 Bg7 6. Nxc6 bxc6 7. Qf3 d6
>>>8. Bc4 Be6 9. Bxe6 fxe6 10. Bd2 Rb8 11. O-O-O Ne7 12. Na4 Rf8 13. Qe3 Be5
>>>14. g3 Qc8 15. Qe2 Qb7 16. b3 Qb5 17. c4 Qb7 18. h4 Bd4 19. Be3 Bxe3+ 20.
>>>Qxe3 Qa6 21. h5 g5 22. Qxg5 Rb4 23. Qe3 Rxa4 24. bxa4 Qxc4+ 25. Kb2 c5 26.
>>>e5 Nd5 27. Qd2 Qxa4 28. exd6 cxd6 29. Ka1 Rf5 30. Rde1 Kd7 31. Rb1 Nb4 32.
>>>Rb2 Rd5 33. Qe2 Kd8 34. f4 h6 35. Qe4 a5 36. Rhb1 Rd1 37. Qe2 Rxb1+ 38. Rxb1
>>>Nc2+ 39. Kb2 Na3 40. Rd1 Nb5 41. Qc2 Qa3+ 42. Ka1 Nd4 43. Qg6 Qc3+ 44. Kb1
>>>Nb5 45. a4 Qb3+ 46. Kc1 Qe3+ 47. Rd2 Nd4 48. Qf6+ Kd7 49. Qf7+ Kc6 50. Qe8+
>>>Kd5 51. Qa8+ Kc4 52. Qg2 Nb3+ 53. Kd1 Nxd2 54. Qxd2 Qxg3 55. Qxd6 Qd3+ 56.
>>>Qxd3+ Kxd3 {White resigns} 0-1
>>
>>I think most programs will choose this move. Fritz 5.32 and Hiarcs both still
>>evaluate this trade as being better for white. Hard to say. What seems clear
>>to me is that 43. Qg6? is a game-losing blunder. Instead 43. Qd3! should win
>>for white.
>
>I ran a couple programs on this, and I don't think there's any clear way to show
>that 43.Qg6 loses. Some programs I've tried don't see it as any worse than Qd3.
> That's not to say you are wrong. It's just that this type of position is tough
>for some programs to analyze correctly.
>
>Will
I'm not ragging on your program, I'm just analyzing the position. ;) It's
difficult to prove a win for black only because of 47. Kc2 instead of the 47.
Rd2? played in the game. After 47. Rd2? it's easy to show that black wins by
force no matter what. 48. Qf6+?! and 55. Qxd6? are strange, and though it makes
no difference in outcome I think it worth checking why your program would choose
Qxd6 at that point. But after 47. Kc2 white is able to scramble for a perpetual
check in a queen endgame. Black may be able to avoid this by playing 45...Qb4+
instead of Qb3+ when the slight difference in geometry looks like it may kill
the perpetual. This is the only difficulty in showing that all is lost after
43. Qg6, but in any case it is easy to show that white has handed the reins over
to black and can only scramble for a draw, when he could have won outright with
43. Qd3. I think that makes this the critical error in the game. FWIW Fritz
5.32 picks 43. Qd3 as best, giving Qg6 as over a pawn worse after a 12 ply
search or so.
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.