Author: Robert Hyatt
Date: 06:34:56 12/14/99
Go up one level in this thread
On December 14, 1999 at 06:03:57, Ed Schröder wrote: >>Posted by Robert Hyatt on December 13, 1999 at 22:32:01: >> > >>Here is one game I had an opportunity to watch this afternoon. What stands >>out is white playing Nxg6, opening the h-file, then following that with o-o >>and castling into a bristling attack. And then making it easy for black to >>fully open the h-file a couple of moves later. This was a slowish game (2 >>minutes on the clock, 12 second increment, played on FICS). I think you will >>agree that we don't normally see 2000 players take on a computer like this and >>have the computer 'self-attack'. >> >>There are other things that also need attention, but since I am not getting >>paid, this will be my sole voluntary contribution to the "tiger development >>effort." :) > >Thanks, hope you make that a habit :-) > > >>Bob >> >>[Event "FICS 2 12 game"] >>[Site "?"] >>[Date "1999.12.13"] >>[Round "?"] >>[White "chesspartner"] >>[WhiteElo "2295"] >>[Black "ceb"] >>[BlackElo "1922"] >>[Result "0-1"] >> >>1. e4 c6 2. d4 d5 3. Nc3 dxe4 4. Nxe4 Bf5 5. Ng3 Bg6 6. Bc4 >>e6 7. N1e2 Nd7 8. Nf4 Ngf6 9. Bb3 Bd6 10. c3 Qc7 11. Nxg6 hxg6 >>12. O-O O-O-O 13. Qf3 Rh7 14. Bg5 Rdh8 15. h3 Qa5 16. Bf4 Bxf4 >>17. Qxf4 g5 18. Qf3 g4 19. hxg4 Qg5 20. Rfb1 Nxg4 21. Bxe6 fxe6 >>22. Qf7 Qh6 23. Kf1 g6 24. Qxh7 Qxh7 25. b3 Qf7 26. Rb2 Ne3+ >>27. Ke2 Nxg2 28. Rf1 Nf4+ 29. Kd1 Nd5 30. c4 Nc3+ 31. Kd2 Qf3 >>32. Rc2 Ne4+ 33. Nxe4 Qxe4 34. Kc1 Qxd4 35. Rd1 Qa1+ 36. Kd2 >>Qd4+ 37. Kc1 Qf4+ 38. Kb1 Rh2 39. Rcd2 Ne5 40. Rc2 Qe4 41. Rf1 >>Rh1 42. Rxh1 Qxh1+ 43. Kb2 Nd3+ 44. Kc3 Qf3 45. b4 Ne1+ 46. Kb2 >>Nxc2 47. Kxc2 Qxf2+ 48. Kc3 c5 49. bxc5 Qxc5 50. Kd3 Kd7 51. Kc3 >>Kd6 52. Kd3 b5 53. cxb5 Qxb5+ 54. Ke3 Qc4 55. a3 Ke5 56. Kf3 >>Qd4 57. a4 Qxa4 58. Ke3 Qa3+ 59. Kf2 Qd3 60. Ke1 Qc2 61. Kf1 >>Kf4 62. Ke1 Kf3 63. Kf1 Qf2# 0-1 >> >>I didn't look at the eentire game, but Crafty strongly disagrees with Nxg6 >>and then after that, even more strongly disagrees with O-O. > >Rebel-Tiger changes after 7 seconds to the better 12.Qf3. I also think >that 15.h4 is a much better way to defend as it will be hard for black >to open the h-file then. > >Those blitz games :-) > >Ed I have said this _many_ times in the past. A chess server is a very _hostile_ place to play chess. If you can survive there, you have a pretty well-debugged program. In the case of move 12, I would be concerned that my program would castle kingside at _any_ shallow depth... because players like dlugy, Roman, Kaidanov, etc will roll you into a small ball at 1 0 or 2 0 if they catch you doing that. It seems to be a minor positional misunderstanding, IMHO. When Crafty castles, I ask this: I just castled kingside. What is my king safety score? What would be the king-safety score if I castled queen-side? What would be the king- safety score if I stay in the center. If the other two possible king positions look safer than where I am, I penalize the current score by an amount proportional to the difference of these scores. This prevents castling king-side, into a shaggy king position, when I might be able to castle queen- side later. Hsu implemented this into deep blue 2 as well. I did it like this because that is how I play as a human. I will only castle into a mess if I don't have time to castle queen-side, and if I get into tactical trouble in the center by staying there...
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.