Author: Frank Phillips
Date: 10:30:37 01/09/03
Go up one level in this thread
On January 09, 2003 at 12:49:00, Uri Blass wrote:
>On January 09, 2003 at 12:44:57, Frank Phillips wrote:
>
>>On January 09, 2003 at 10:32:45, Uri Blass wrote:
>>
>>>On January 09, 2003 at 10:27:11, Richard Pijl wrote:
>>>
>>>>On January 09, 2003 at 10:09:26, Andrew Williams wrote:
>>>>
>>>>>On January 09, 2003 at 10:04:26, Richard Pijl wrote:
>>>>>
>>>>>>On January 09, 2003 at 09:35:16, Andrew Williams wrote:
>>>>>>
>>>>>>>Hi,
>>>>>>>
>>>>>>>Last night I was looking at a game between Hossa and PostModernist. PM lost
>>>>>>>after it took the pawn on a2 in this position:
>>>>>>>
>>>>>>>[D]r3r1k1/pbpp1pp1/1p3n1p/8/q1PPp3/2P1P1P1/P2BQP1P/1R2RBK1 b - - 0 19 am Qxa2
>>>>>>>
>>>>>>>
>>>>>>>How long does it take your program to ignore the Pawn here? Even after five
>>>>>>>minutes thinking, PM still wants to take it. The full game follows:
>>>>>>>
>>>>>>>[Event "ICC 20 5"]
>>>>>>>[Site "Internet Chess Club"]
>>>>>>>[Date "2003.01.03"]
>>>>>>>[Round "-"]
>>>>>>>[White "Hossa"]
>>>>>>>[Black "PostModernist"]
>>>>>>>[Result "1-0"]
>>>>>>>[ICCResult "Black resigns"]
>>>>>>>[WhiteElo "2303"]
>>>>>>>[BlackElo "2356"]
>>>>>>>[Opening "English: four knights, kingside fianchetto"]
>>>>>>>[ECO "A29"]
>>>>>>>[NIC "EO.04"]
>>>>>>>[Time "05:12:05"]
>>>>>>>[TimeControl "1200+5"]
>>>>>>>
>>>>>>>1. c4 Nf6 2. Nc3 e5 3. Nf3 Nc6 4. g3 Nd4 5. Bg2 Nxf3+ 6. Bxf3 Bb4 7. d4 e4
>>>>>>>8. Bg2 O-O 9. Bg5 Re8 10. O-O Bxc3 11. bxc3 h6 12. Bf4 b6 13. Re1 Bb7 14.
>>>>>>>Rb1 Nh5 15. Bd2 Qf6 16. Bh3 Qc6 17. e3 Nf6 18. Qe2 Qa4 19. Bf1 Qxa2 20. Ra1
>>>>>>>Qb3 21. Reb1 Qc2 22. Rc1 Qb3 23. Be1 Nd5 24. Rc2 d6 25. Rb2 Qxb2 26. Qxb2
>>>>>>>Nf6 27. c5 dxc5 28. dxc5 Red8 29. c4 Nd7 30. cxb6 cxb6 31. Bb4 Bc6 32. Bh3
>>>>>>>f6 33. Bd6 Re8 34. Ra6 Ne5 35. c5 Bd5 36. Bxe5 Rxe5 37. cxb6 axb6 38. Rxa8+
>>>>>>>Bxa8 39. Qxb6 Bd5 40. Qb8+ Kh7 41. Bd7 h5 42. Qc8 Kg6 43. Qf8 Bc4 44. Be8+
>>>>>>>Kh7 45. Bf7 Bxf7 46. Qxf7 Kh6 47. Kg2 Rb5 48. h3 Rg5 49. Qg8 Re5 50. g4 hxg4
>>>>>>>51. hxg4 Kg6 52. Kg3 Kh6 53. Kf4 Kg6 54. Qc4 Kh6 55. Qf7 Kh7 56. Qc7 Rg5 57.
>>>>>>>Kxe4 Kg6 58. Qd7 Re5+ 59. Kd4 Ra5 60. f4 Ra1 61. Qe8+ Kh7 62. g5 fxg5 63.
>>>>>>>fxg5 Rd1+ 64. Kc5 {Black resigns} 1-0
>>>>>>>
>>>>>>>
>>>>>>>Andrew
>>>>>>
>>>>>>On a PII-500 Baron 0.99.4 beta 4 discards Qxa2 after 2 seconds:
>>>>>>
>>>>>> ply time nodes score pv
>>>>>>
>>>>>> 4( 4)& 0 3815 75 a4a2 b1a1 a2c2 f1g2
>>>>>> 4( 6). 0 6542 75 a4a2 b1a1 a2c2 f1g2
>>>>>> 5( 6)& 0 11944 85 a4a2 f1h3 a7a5 h3f5 a8a6
>>>>>> 5( 6). 0 15654 85 a4a2 f1h3 a7a5 h3f5 a8a6
>>>>>> 6( 8)- 0 34821 50 a4a2 b1a1 a2c2 f1g2 d7d5 e1d1
>>>>>> 6( 8)& 0 39193 50 a4a2 e2d1
>>>>>> 6( 8). 0 39234 50 a4a2 e2d1
>>>>>> 7( 9)+ 0 46823 85 a4a2 b1a1 a2c2 a1c1 c2b3 c1a1 d7d5
>>>>>> 7( 9)& 0 63726 91 a4a2 b1a1 a2c2 e1c1 c2b2 f1g2 a7a5
>>>>>> 7(10). 1 103676 91 a4a2 b1a1 a2c2 e1c1 c2b2 f1g2 a7a5
>>>>>> 8(12)& 2 220687 57 d7d5 c4d5 b7d5 c3c4 a4a2 c4d5 a7a5
>>>>>> 8(12)& 4 399670 77 a7a5 b1a1 b7a6 e2d1 a4d1 e1d1 d7d5 c4d5 f6d5
>>>>>>c3c4
>>>>>> 8(12). 5 475311 77 a7a5 b1a1 b7a6 e2d1 a4d1 e1d1 d7d5 c4d5 f6d5
>>>>>>c3c4
>>>>>> 9(13)& 8 779985 78 a7a5 b1a1 b7a6 e2d1 a4d1 e1d1 d7d5 c4d5 a6f1
>>>>>>g1f1 f6d5
>>>>>> 9(14). 15 1404152 78 a7a5 b1a1 b7a6 e2d1 a4d1 e1d1 d7d5 c4d5 a6f1
>>>>>>g1f1 f6d5
>>>>>> 10(19)& 48 4152008 56 a4a5 e1c1 b7a6 e2d1 d7d5 d1b3 a8d8 d2e1 f6g4
>>>>>>c4d5
>>>>>> 10(19). 66 5713921 56 a4a5 e1c1 b7a6 e2d1 d7d5 d1b3 a8d8 d2e1 f6g4
>>>>>>c4d5
>>>>>> 11(21)& 156 13131496 40 a4a5 e2d1 f6h7 a2a4 b7c6 b1b4 a5f5 f1g2 d7d5
>>>>>>d1c2 d5c4
>>>>>> 11(21). 226 19485176 40 a4a5 e2d1 f6h7 a2a4 b7c6 b1b4 a5f5 f1g2 d7d5
>>>>>>d1c2 d5c4
>>>>>> 12(23)& 370 30991095 38 a4a5 e2d1 f6h7 a2a4 b7a6 d1b3 d7d5 c4d5 a6f1
>>>>>>e1f1 e8d8 b3c4 d8d5
>>>>>> 12(23)& 523 44028658 50 a7a5 e1c1 b7a6 e2d1 a4c6 d1b3 a5a4 b3b4 d7d5
>>>>>>c4d5 f6d5
>>>>>
>>>>>:-)
>>>>>
>>>>>Thanks Richard. Oddly enough, last night's work on trapped pieces was initiated
>>>>>by a game that PM lost to thebaron a few days ago when it heedlessly trapped its
>>>>>own bishop. It looks like I'll have to look at trapped queens too.
>>>>>
>>>>Baron does look at trapped bishops, but not at trapped queens. I think it is the
>>>>mobility score that might have done the trick here. Bad mobility for the queen
>>>>is penalized about -25 in the Baron.
>>>>
>>>>Richard
>>>
>>>I think that seeing the win of the queen is not simple but seeing a draw score
>>>for Qxa2 is a simple tactics for programs.
>>>
>>>I suspect that a program that likes Qxa2 with a positive score may have a bug in
>>>it's repetition detection.
>>>
>>>Uri
>>
>>
>>Are you sure this a draw ... if so then I need to look at my repetition code for
>>bugs.
>>
>>Frank
>
>It is not a draw but a program is supposed at least to see a draw and not
>positive score in case that it likes that move.
>
>The point is that in this position there is a simple tactic and a complicated
>tactics.
>
>simple tactics say that after Qxb2 black has no escape from draw by repetition.
>complicated tactics say that Qxb2 is losing the queen.
>
>programs need a long time to see the complicated tactics but seeing the simple
>tactics should not be hard task for them.
>
>Uri
Thanks Uri. In which case has anyone got any good position for testing
repetition detection. I use this rather simple one, which revealed problems
with my old method:
/*
* Draw by repetition check - Rxg7
* An important test.
*/
6k1/1p1q1ppp/p7/n2p4/3PrQ2/B3P1R1/6PP/6K1 w - -
Frank
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.