Author: Uri Blass
Date: 15:42:25 04/24/02
Go up one level in this thread
On April 24, 2002 at 18:06:34, Jorge Pichard wrote:
>On April 24, 2002 at 17:53:13, Uri Blass wrote:
>
>>On April 24, 2002 at 17:36:50, martin fierz wrote:
>>
>>>On April 24, 2002 at 17:17:48, Otello Gnaramori wrote:
>>>
>>>>On April 24, 2002 at 16:56:10, martin fierz wrote:
>>>>
>>>>>impressive play by hiarcs! it outplayed smirin in a positional game as black -
>>>>>by far the best game of of a computer in the "comp vs smirin" match yet!
>>>>>
>>>>>aloha
>>>>> martin
>>>>
>>>>I agree with you,
>>>>The only strange thing was the final eval score of Hiarcs : over 2 pawns for
>>>>it...
>>>>
>>>>Otello
>>>
>>>not so strange :-)
>>>computers don't seem to understand the "fortress" concept.
>>
>>I expect higher intelligence auto response chess system to understand the
>>fortress concept.
>>
>>Intelligent program is a program that can understand everything that is trivial
>>to humans with rating of 1500 and I am sure that the draw is trivial to people
>>with rating of 1500 and they do not need previous knowledge in their evaluation
>>function.
>>
>>I think that there are no intelligent chess programs today.
>>
>>Uri
>
>
>I agree with you on this Uri, it is simply that chess programmers are feeding
>more chess knowledges to their programs and writing programs that are less
>buggy. Hopefully in the future you will achieve a program of Hiarcs 8 strength,
>since your program Movei is very decent for a starter. By the way how is your
>Opening Book for Movei progressing ?
>
>Pichard.
I hope that you are right but I do not expect hiarcs's level in the near future.
I still did not do a progress in the direction of an opening book but I got a
significant improvement in the level of movei by implementing the null move
algorithm.
I believe that my latest version is at least 100 elo better than the version
that I sent you so I may call it movei007.
A version that is probably also better than movei006 is playing today in the 5th
division of the winboard programs and has 8.5 out of 10.
You can see the results and download the games in the following link
http://home.hccnet.nl/leo.dijksman/index.html
I can add that I was not sure if recursive null move is really productive at
long time control inspite of the fact that it is productive at blitz based on my
tests of many blitz games so I tested it against chezzz at 400 minutes per
game(chezzz is a program of the 3th division) and movei produced the following
game(note that chezzz beated a previous version of movei at the same time
control):
I did not tell instructions for chezzz about hash tables but I know that chezzz
was using 8 Mbytes hash tables as default value when movei still does not use
hash tables.
Event "Computer chess game"]
[Site "URI-PC"]
[Date "2002.04.23"]
[Round "-"]
[White "moveinull224"]
[Black "Chezzz_088"]
[Result "1-0"]
[TimeControl "24000"]
1. e4 e5 2. Nf3 Nf6 3. Nxe5 Qe7 4. d4 d6 5. Nf3 Qxe4+ 6. Be2 Be7 7. Nc3 Qf5
8. O-O O-O 9. Re1 Re8 10. Bc4 Nc6 11. d5 Nb8 12. Bg5 a6 13. Re3 Bf8 14. Bd3
Qg4 15. h3 Qd7 16. Rxe8 Qxe8 17. Bxf6 gxf6 18. Nd4 Qe5 19. Qd2 Qg5 20. f4
Qh5 21. Qe3 Bd7 22. Ne4 Bg7 23. f5 Kh8 24. c3 Bh6 25. Qf2 Bg7 26. Re1 Qh6
27. Nc2 Bf8 28. Ne3 Qg7 29. Ng4 Be7 30. Re3 Qf8 31. Nexf6 Bxf6 32. Nxf6 Qh6
33. Ne8 b6 34. Nxc7 Ra7 35. Ne8 Ba4 36. Re4 Nd7 37. Rxa4 Qf8 38. Qe3 Ra8
39. Nc7 Rc8 40. Nxa6 Re8 41. Qd4+ f6 42. Nc7 Rb8 43. Ra7 Ne5 44. Na6 Ra8
45. Rb7 Qh6 46. Rxb6 Qc1+ 47. Kf2 Nxd3+ 48. Qxd3 Re8 49. Kf3 Re1 50. Rb8+
Kg7 51. Rb7+ Kh6 52. Kf2 Re5 53. Rb4 Re7 54. Kf3 Kg7 55. Nb8 Re1 56. Nc6
Rf1+ 57. Kg3 Qe1+ 58. Kg4 Qf2 59. Rb7+ Kf8 60. g3 Rg1 61. Rxh7 Rg2 62. Rd7
Qe1 63. Rxd6 Kg7 64. Rd7+ Kh8 65. Rb7 Qc1 66. h4 Rd2 67. Qf3 Rxb2 68. Re7
Qd2 69. Kh5 Rb7 70. Rxb7 Qe1 71. Kh6 Qd2+ 72. Kg6 Qxd5 73. Rh7+ Kg8 74.
Qxd5+ Kf8 75. Rh8#
{White mates} 1-0
I know that one game is not enough but I get the impression that the null move
algorithm is productive for my program at all time controls.
Uri
This page took 0.01 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.