Author: blass uri
Date: 02:43:29 10/25/98
Go up one level in this thread
On October 25, 1998 at 04:41:34, Amit Kagan wrote:
>The following game is taken from Mark Young's latest match.
>
>[Event "30"av"]
>[Site "?"]
>[Date "1998.10.24"]
>[Round "10"]
>[White "Junior 5.0"]
>[Black "Fritz 5.00/16Bit"]
>[Result "0-1"][ECO "C82"]
>[WhiteElo "2430"]
>[BlackElo "2576"]
>[PlyCount "92"]
>{W=24.5 ply, B=17.0ply} 1. e4 {0} 1... e5 {0} 2. Nf3 {0} 2... Nc6 {0} 3. Bb5 {0
>} 3... a6 {0} 4. Ba4 {0} 4... Nf6 {0} 5. O-O {0} 5... Nxe4 {0} 6. d4 {0} 6...
>b5 {0} 7. Bb3 {0} 7... d5 {0} 8. dxe5 {0} 8... Be6 {0} 9. c3 {0} 9... Bc5 {0}
>10. Nbd2 {0} 10... O-O {0} 11. Bc2 {0} 11... Bf5 {0} 12. Nb3 {0} 12... Bg6 {0}
>13. Nfd4 {0} 13... Bxd4 {0} 14. cxd4 {0} 14... a5 {0} 15. Be3 {0} 15... a4 {0}
>16. Nd2 {0} 16... a3 {0} 17. Nxe4 {0} 17... axb2 {0} 18. Rb1 {0} 18... Bxe4 {0}
>19. Rxb2 {0} 19... Qd7 {0} 20. Bd3 {White out of book 0} 20... Bxd3 {0} 21.
>Qxd3 {0.27/14 53} 21... Rfb8 {Black out of book 0} 22. Rxb5 {0.27/15 65} 22...
>Rxb5 {0.38/14 98} 23. Qxb5 {0.16/14 41} 23... Nxe5 {0.44/15 133} 24. Rb1 {
>(Qb7) 0.18/15 87} 24... Qxb5 {0.16/14 89} 25. Rxb5 {0.19/14 19} 25... c6 {
>0.28/13 105} 26. Rb6 {(Rb4) 0.13/17 87} 26... Nc4 {0.09/13 76} 27. Rxc6 {
>0.10/18 89} 27... Kf8 {(f5) 0.25/12 60} 28. Rc5 {0.16/16 92} 28... Rxa2 {
>0.34/12 55} 29. h4 {(g4) 0.18/15 74} 29... Ra5 {0.31/13 76} 30. Kh2 {
>(Rxa5) 0.09/15 45} 30... Rxc5 {(Ke7) -0.59/15 38} 31. dxc5 {-1.78/20 25} 31...
>Nxe3 {-0.66/13 44} 32. fxe3 {-1.89/20 28} 32... Ke7 {-1.00/17 36} 33. g4 {
>(Kg3) -1.89/26 62} 33... Kd7 {(h6) -1.06/17 38} 34. Kg3 {-1.91/24 42} 34... Kc6
>{-1.16/16 39} 35. g5 {(Kf4) -2.13/26 88} 35... Kxc5 {-1.78/16 20} 36. Kf4 {
>-2.12/25 39} 36... Kc4 {-3.66/15 24} 37. Ke5 {-2.34/26 35} 37... Kd3 {
>(h6) -5.25/14 12} 38. Kxd5 {-2.90/28 48} 38... Kxe3 {-5.25/16 9} 39. Kd6 {
>-3.61/31 46} 39... Kf4 {-6.00/20 11} 40. Ke7 {-3.49/27 18} 40... Kg4 {
>-8.97/22 13} 41. Kxf7 {-3.20/28 12} 41... g6 {-9.41/24 15} 42. Kg7 {
>-10.47/38 41} 42... Kxh4 {-9.44/22 12} 43. Kh6 {-10.47/36 34} 43... Kg4 {
>-9.97/25 81} 44. Kxh7 {-10.55/39 45} 44... Kxg5 {-#11/24 21} 45. Kg7 {
>-10.62/38 38} 45... Kf5 {-#11/21 10} 46. Kf7 {(Kh6) -10.64/33 39} 46... g5 {
>-#10/19 9} 0-1
>
>
>30. Kh2 is a horrible move.
>To me, it is surprising that a very decent chess program makes such a move,
>since after two (obvious and only) captures, the position becomes a won pawn
>ending for black.
>I thought that capture-recapture extensions are supposed to help rejecting such
>moves in a matter of seconds (at low depth).
>Any opinions?
1)Junior uses the half ply depth for capture-recapture moves(the depth is the
number of important moves like capture-recaptures+2*the number of not important
moves it sees).
It is not enough and I think it should use negative depth for moves that leads
to a pawn ending.
2) There is an evaluation problem.
Junior should evaluate the pawn ending as winning for black and avoid Kh2
because the black king is closer to the white pawns than the white king.
Junior evaluates it as better for white because white has a passed pawn.
I hope this problem will be solved in the next version.
Uri
> results with other programs?
>(I should note that Junior5 avoids 30. Kh2 at depth=16, but IMO it is too slow.)
>
>
>One more thing. From posted games of Junior against CSTal and Fritz, it seems to
>me that Junior's opening book leads to wild positions, that Junior, like most
>(if not all) other programs, can't defend well.
>Any opinions on this? Would using "optimized" book of other program would help?
>
>
>Amit
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.