Author: Robert Hyatt
Date: 13:56:24 04/30/04
Go up one level in this thread
On April 30, 2004 at 16:26:34, Bryan Hofmann wrote: >On April 30, 2004 at 14:51:37, Telmo C. Escobar wrote: > >>On April 30, 2004 at 06:43:19, Bryan Hofmann wrote: >> >>>On April 29, 2004 at 19:57:44, Telmo C. Escobar wrote: >>> >>>> >>>> Does somebody know whether there is some way to make Crafty be a bit more >>>>careful about its king? The following game features a common accident of >>>>Crafty's. A human player could be not able to calculate, several moves before, >>>>that 41...f5+! would be a winner, but we are used to think that such an >>>>adventurous king invites disaster. >>>> >>>> Telmo >>>> >>>> >>>>[Event "Computer chess tournament"] >>>>[Site "Athlon 1200"] >>>>[Date "2004.4.29"] >>>>[Round "04"] >>>>[White "Crafty 17.14 SMP"] >>>>[Black "Chessmaster 8000"] >>>>[Result "0-1"] >>>>[TimeControl "3600"] >>>> >>>>1. d4 Nf6 2. c4 e5 3. dxe5 Ng4 4. Bf4 Nc6 5. Nf3 Bb4+ 6. Nbd2 Qe7 7. e3 >>>>Ngxe5 8. Nxe5 Nxe5 9. Be2 O-O 10. O-O Bxd2 11. Qxd2 d6 12. Rfe1 Bf5 13. f3 >>>>Be6 14. c5 Rfd8 15. cxd6 cxd6 16. b3 f6 17. Rac1 d5 18. Rc3 Rac8 19. Rec1 >>>>Rxc3 20. Rxc3 Nc6 21. Rc1 d4 22. Bb5 Qb4 23. Qe2 dxe3 24. Qxe3 Qxb5 25. >>>>Qxe6+ Kh8 26. Re1 Qb6+ 27. Be3 Qa5 28. Re2 h6 29. Qf7 Rd1+ 30. Kf2 Ne5 31. >>>>Qe6 Qd8 32. Bf4 Nd3+ 33. Kg3 Nxf4 34. Kxf4 Qc7+ 35. Kg4 Rd8 36. g3 Qc1 37. >>>>Qf5 Qc7 38. h4 Qd6 39. f4 Qd1 40. Qc2 Qf1 41. Re7 f5+ 42. Kxf5 Qh3+ 43. g4 >>>>Qxh4 44. Qc5 Qf6+ 45. Ke4 b6 46. g5 Qg6+ 47. Qf5 Rd4+ 48. Kxd4 Qxf5 49. Re4 >>>>hxg5 50. fxg5 Qxg5 51. Re2 Kh7 52. Ke4 Qg4+ 53. Kd3 Qg3+ 54. Kd4 a6 55. Re3 >>>>Qf4+ 56. Kd3 Qf5+ 57. Kd4 Qc5+ 58. Ke4 Qc2+ 59. Ke5 Qxa2 60. Rh3+ Kg6 61. >>>>Rg3+ Kh5 62. Re3 Kg4 63. Re4+ Kf3 64. Rf4+ Kg3 65. Rf7 g5 66. Rd7 Qxb3 67. >>>>Kf6 Qc3+ 68. Kf7 b5 69. Re7 b4 70. Kg8 b3 71. Re8 >>>>0-1 >>> >>>You might want to consider using a more current version of Crafty. The current >>>release is 19.12. >> >> Well, I gave the above game because I was interested in it, but I remember >>watching games by more recent versions lost in similar way. It could be -I don´t >>know for sure- that king safety has not modified in recent versions. Maybe Bob >>Hyatt could say something about. >> >> I was informed that one can add a parameter in the crafty.rc configuration file >>that presumably gauges king safety up. >> >> Needless to say, I don´t know if experimenting with a single parameter will >>really improve results. Bet it doesn´t- yet I´m interested in seeing. >> >> Telmo > >You can read the documentation on "evaluation kscale" in the following; >ftp://ftp.cis.uab.edu/pub/hyatt/doc/crafty.doc.ascii I think there is a better option at www.cis.uab.edu, click on faculty and then hyatt, and in the "online paper" section there is an HTML version of the docs I think...
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.