Author: Chris Carson
Date: 06:18:49 02/05/03
Go up one level in this thread
On February 04, 2003 at 21:05:27, Amir Ban wrote:
>[Event "X3D man-machine match"]
>[Site "New York City"]
>[Date "2003.02.02"]
>[Round "4"]
>[White "*Deep Junior 8"]
>[Black "*Kasparov, Garry"]
>[Result "1/2-1/2"]
>
>1. e4 c5 2. Nf3 Nc6 3. d4 cxd4 4. Nxd4 e6 5. Nb5 d6 6. c4 Nf6 7. N1c3 a6 8.
>Na3 Nd7
>
>{Kasparov took here 20 minutes to take Deep Junior out of book. Is that any way
>to spend your time ?}
>
>9. Nc2 Be7 10. Be2 b6 11. O-O Bb7 12. h3
>
>{It's already clear that this game is evolving in a different way than previous
>ones. Kasparov, whether out of choice or out of necessity, is not grabbing the
>initiative and is letting Deep Junior develop its game. He may be using Levy's
>"Do nothing, but do it well" advice, essentially meaning that if the program is
>given enough rope it will hang itself. This move is somewhat pointless, but in
>the end it didn't matter}
>
>12... O-O 13. Be3 Rc8 14. Qd2 Nce5 15. b3 Nf6 16. f3
>
>{An alternative was f4 followed by Bf3, but the great merit of this move is that
>it focuses the program on the queenside for the rest of the game}
>
>16... Qc7 17. Rac1
>
>{Suddenly things get exciting. This allows 17... b5, after which Deep Junior
>would have sacrificed a piece by 18. cxb5. Although the masters in the computer
>room shook their head in disapproval, I personally would welcome this discussed
>over the board. Kasparov did not take up the challenge}
>
>17... Rfe8 18. a3 Ned7
>
>{Same challenge again, declined again. This is a victory of sorts}
>
>19. Rfd1 Qb8 20. Bf2 Rcd8 21. b4
>
>{The commentary by Ashley and Seirawan devotes a lot of time to black's possible
>strikes against white's supposed weaknesses, but it's becoming hard to deny that
>white is taking the advantage}
>
>21... Ba8 22. a4 Rc8 23. Rb1 Qc7 24. a5
>
>{That's the right idea, but was it played too soon ? Deep Junior played this
>immediately to Qc7, after its evaluation dropped somewhat, but not enough to
>trigger more time}
>
>24... bxa5 25. b5 Bb7 26. b6 Qb8 27. Ne3 Nc5 28. Qa2 Nfd7 29. Na4 Ne5 30. Nc2
>Ncd7 31. Nd4 Red8 32. Kh1 Nc6 33. Nxc6 Rxc6 34. Kg1
>
>{This move as well as the 32nd betray an indecisiveness about how to proceed in
>this position. Computer programs do not play like masters. Masters often follow
>a plan that is set in operative terms. The program's plan is its evaluation
>function, which is an implicit plan and it's up to the search to find the best
>way to execute. Every move is a fresh look into that problem. Sometimes this
>search comes up blank}
>
>34... h6 35. Qa3
>
>{The new plan is to pressure the d6 pawn, which is going to be pinned in three
>different directions}
>
>35... Rdc8 36. Bg3 Bf8 37. Qc3 Ne5 38. c5
>
>{38. Bxe5 must have been the way to win, but Deep Junior's evaluation is already
>blinded by winning chances that will soon prove to be non-existent}
>
>38... Nd7 39. Qxa5 Nxc5 40. Nxc5 Rxc5 41. Qa4
>
>{Now if black tries to hold on to the pawn by 41... a5 42. Bb5 will be painful}
>
>41... R5c6
>
>{Kasparov now has time to think, and he spends more than 20 minutes over this
>move, and finds a way to hold the draw}
>
>42. Bf2 d5 43. Bxa6 Bc5 44. Bxc5 Rxc5 45. Bxb7 Qxb7 46. exd5 exd5 47. Qa7 R5c7
>
>{Deep Junior gets excited with +2 scores, but in the computer room we were less
>excited as we have been discussing the ensuing endgame for some time and
>recognizing it as a draw}
>
>48. Qxb7 Rxb7 49. Rxd5 Rc6 50. Rdb5 h5
>
>{The only chance for a win is if the rooks leave the b6 pawn to create mayhem on
>the kingside pawns. Kasparov shuts off all such possibilities}
>
>51. Kf2 Re6 52. f4 g6 53. Kg3 Kg7 54. Kh4 Kh6 55. R1b4 Rd6 56.g3 f6 57. g4 hxg4
>58. hxg4 Kg7 59. Rb3 Rc6 60. g5 f5 61. Rb1 {Draw Agreed} 1/2-1/2
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.