Author: Robert Hyatt
Date: 19:42:22 11/20/00
Go up one level in this thread
On November 20, 2000 at 22:05:12, Eduard Nemeth wrote: >Gambit Tiger 1.0 (i think chess Tiger 13.0 too) can not mates with Bishop and >Knight! > >I played from this FEN-Position a game with 1 minutes/move and play end draw ! > >I can this not right belived, but it is true ! See, and test youself: > >[D]8/8/8/3B4/8/2K1N3/8/k7 w - - 0 1 > >[Event "Test mates with N+B"] >[Site "My Town P600"] >[Date "2000.11.21"] >[Round "1"] >[White "Gambit Tiger 1.0"] >[Black "Eduard Nemeth"] >[Result "1/2-1/2"] >[Level " 1 Minute/move"] >[Opening "Not"] >[FEN: 8/8/8/3B4/8/2K1N3/8/k7 w - - 0 1] > >Here the notation on my P600, hash 48 MB : > >1. Be4 Ka2 2. Nc4 Ka1 3. Nd6 Ka2 4. Nb5 Ka1 5. Nd4 Ka2 6. Bf5 Ka1 7. Nf3 Ka2 >8. Ne5 Ka1 9. Be4 Ka2 10. Bd5+ Kb1 11. Nd3 Ka1 12. Be4 Kb1 13. Kb3 Ka1 14. Kc2 >Ka2 15. Nb2 Ka1 16. Nc4 Ka2 17. Bc6 Ka1 18. Ne5 Ka2 19. Be4 Ka1 20. Nf3 Ka2 >21. Nd4 Ka1 22. Kc1 Ka2 23. Bd5+ Ka3 24. Nc6 Ka4 25. Bc4 Ka3 26. Kc2 Ka4 27. >Kc3 Ka3 28. Bb5 Ka2 29. Kc2 Ka1 30. Bd3 Ka2 31. Bc4+ Ka1 32. Kc3 Kb1 33. Ne5 >Ka1 34. Nf3 Kb1 35. Kb3 Ka1 36. Bd5 Kb1 37. Be4+ Ka1 38. Bc6 Kb1 39. Bd5 Ka1 >40. Bb7 Kb1 41. Bc6 Ka1 42. Kc2 Ka2 43. Ne5 Ka1 44. Nc4 Ka2 45. Bf3 Ka1 46. >Kb3 Kb1 47. Nd2+ Ka1 48. Ne4 Kb1 49. Kc3 Ka1 50. Nf6 Ka2 1/2-1/2 > >Oh, draw. sorry but I'm "now" shocked ! > >Eduard First, it is a mate in 20: 1. Nc2+ Kb1 2. Bg8 Kc1 3. Ba2! Kd1 4. Nd4! Ke1! 5. Bd5 Kd1 6. Be4! Kc1! 7. Nb3+ Kd1 8. Bf3+ Ke1 9. Kd3! Kf2! 10. Nd4! Kf1 11. Ke3! Ke1! 12. Ne6 Kf1 13. Nf4! Ke1! 14. Bh5 Kf1 15. Be2+! Kg1! 16. Kf3! Kh2! 17. Kf2! Kh1 18. Kg3! Kg1 19. Nh3+! Kh1 20. Bf3# The main question is, did Christophe depend on tablebases to solve this? Others have done this in the past. IE I believe that Nimzo had problems with simple mates like KR vs K in the past, because the author assumed that basic table- base files would be available, and by taking out the redundant chess knowledge, the program became simpler and faster overall... It might be a case that is frequently over-worked here. Playing engines without a book? Perhaps the author doesn't test like that and didn't choose to write code to handle it. Playing without any EGTBs? Same issue. Playing Nunn tests? Perhaps the program doesn't like some of the positions, and since the author hand-crafted a book that wouldn't reach some of those positions, it really doesn't matter. Lots of things an author has to choose to include or exclude. I would personally assume that everyone would now use all the 3-4-5 piece files since they require 7.5 gigs of disk, and that is dirt-cheap at today's DASDI prices.
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.