The strongest program that play chess960 is NOT Rybka

[Tony Thomas]

Hello

Unfortunately no PGN:

http://www.chesstigers.de/partien/ccm7/ ... 5/live.htm

Round 1
Shredder - Jonny 1-0
Spike - Rybka 1-0

http://www.chesstigers.de/ccm7_index_ne ... ng=0&kat=0

In german, but if you scroll down you can read 3. Livingston C960 Computer World Champion ....

There you see the "online" button - which is offline at the moment.

BYe
Ingo

Looks like I was wrong.

[Uri Blass]

Yes I always thought that Spike has tremendous potential. It is IMHO one of the best performers on slow hardware and seems to have great chess instincts that are not based on search ... but based on knowledge. Still I would expect Rybka to win a match against any of them ... although some losses are inevitable.

Can you give examples what do you think that spike knows that other programs do not know?

I think that if spike has knowledge that other programs do not have than the other programs have tremendous potential because the programmers can implement that knowledge in their programs and make an improvement when poor spike is going to get no improvement from the same thing that it already has.

Uri

[Gandalf]

I think that if spike has knowledge that other programs do not have than the other programs have tremendous potential because the programmers can implement that knowledge in their programs and make an improvement when poor spike is going to get no improvement from the same thing that it already has.

The same can be said of Rybka, or pretty much any chess playing program.

[IWB]

Hello

Here are the games from "www.chesstigers.de":

[Event "Livingston Chess960 Computer World Championship "]
[Site "Chess Classic Mainz"]
[Date "2007/8/15"]
[Round "1"]
[White "Deep Shredder"]
[Black "Jonny Mainz"]
[Result "1-0"]
[Variant "chess 960"]
[SetUp "1"]
[FEN "brkrnbnq/pppppppp/8/8/8/8/PPPPPPPP/BRKRNBNQ w DBdb - 0 1"]
[Board "1"]
[Input "DGT3067"]
[Owner "GRENKELEASING AG"]

1. b4 b5 2. f3 Bd5 3. a4 bxa4 4. e4 Bc6 5. Nh3 e6 6. Ba6+ Bb7 7. Qf1 Nd6 8. e5
Nf5 9. g4 Nh4 10. Bd4 c6 11. Ng5 Nh6 12. Ne4 Re8 13. Qd3 Kc7 14. Bxa7 Ng6 15.
Nc5 Nxe5 16. Qd4 Ba8 17. Nxa4 d5 18. Qxe5+ Bd6 19. Qe3 Kd7 20. Nd3 Rbd8 1-0

[Event "Livingston Chess960 Computer World Championship "]
[Site "Chess Classic Mainz"]
[Date "2007/8/15"]
[Round "1"]
[White "Spike 1.3 MP"]
[Black "Rybka"]
[Result "1-0"]
[Variant "chess 960"]
[SetUp "1"]
[FEN "brkrnbnq/pppppppp/8/8/8/8/PPPPPPPP/BRKRNBNQ w DBdb - 0 1"]
[Board "2"]
[Input "DGT3159"]
[Owner "GRENKELEASING AG"]

1. b4 Ngf6 2. f3 g6 3. e4 Bg7 4. Bd4 b6 5. Ba6+ Bb7 6. Bxb7+ Rxb7 7. Be3 e5 8.
Nd3 Nd6 9. Nb2 a5 10. bxa5 bxa5 11. Ne2 O-O 12. Nc3 Rfb8 13. Qf1 c6 14. Qa6 h6
15. Nca4 Bf8 16. Nc5 Rb4 17. a3 R4b5 18. a4 Rb4 19. Nb3 Nfe8 20. Bc5 Nc7 21.
Qd3 Ne6 22. Bxb4 axb4 23. O-O Qf6 24. Qe3 Qd8 25. Rbd1 Qc7 26. Rfe1 Ra8 27. Kh1
c5 28. d3 h5 29. Ra1 Be7 30. Qf2 h4 31. a5 Nb5 32. Nc4 Ra6 33. Nb6 Nc3 34. g3
Qc6 35. Rf1 h3 36. f4 exf4 37. gxf4 f5 38. Qg3 Kg7 39. Qxh3 fxe4 40. Qg2 e3 41.
Qxc6 dxc6 42. f5 Nf8 43. fxg6 Nxg6 44. Rg1 Ne2 45. Rg4 Nef4 46. Nc4 e2 47. Rag1
Kh6 48. Nc1 Kh5 49. Ne5 e1=Q 50. Rxe1 Bf6 51. Nxg6 Kxg4 52. Re4 Bg5 53. Ne2 Kf5
54. Ngxf4 Rxa5 55. Ng3+ Kf6 56. Nfe2 Kg7 57. h4 Bd8 58. Kg2 Kh7 59. Kh3 Kg7 60.
Re8 Bb6 61. Nf5+ Kf7 62. Nd6+ Kg7 63. Nf4 Ra7 64. Kg4 c4 65. Ne6+ Kf6 66. Ng5
Re7 67. Nde4+ Rxe4+ 68. dxe4 1-0

[IWB]

No guarantee but if I got everything right the first day was as follows:

Round 1

Shredder - Jonny 1:0
Spike - Rybka 1:0

Round 2

Rybka - Jonny 1:0
Shredder - Spike 1:0

Round 3

Rybka - Shredder 0.5 : 0.5
Spike - Jonny 0.5 : 0.5

Not sure about the colors, but 99% about the results!

This means:

Shredder 2.5
Spike 1.5
Rybka 1.5
Jonny 0.5


Bye
Ingo

[M ANSARI]

Yes I always thought that Spike has tremendous potential. It is IMHO one of the best performers on slow hardware and seems to have great chess instincts that are not based on search ... but based on knowledge. Still I would expect Rybka to win a match against any of them ... although some losses are inevitable.

Can you give examples what do you think that spike knows that other programs do not know?

I think that if spike has knowledge that other programs do not have than the other programs have tremendous potential because the programmers can implement that knowledge in their programs and make an improvement when poor spike is going to get no improvement from the same thing that it already has.

Uri

Hi Uri ... yes I can. I will have to look it up but some time back I posted a game between a 4 core Shredder against a single core Spike. In that game Spike was black and in an open position it completely outplayed Shredder by playing simple strong moves. Shredder was trying to find tactical solutions for a win yet Spike simply played solid strong moves without having to wade through tons of search. This totally nullified the hardware advantage of Shredder and Spike won very convincingly. I have also games against Rybka where Spike somehow seems to have much better instincts in some positions ... especially positions where tactics are useless and long term planning is more important. I would like nothing more than to look and find those games but it will take some time.