Author: Vladimir Xern
Date: 19:52:08 07/24/05
...are blitz games against the aging (yet quite strong) amateur Delfi 4.5 and
are probably not good for much more than a few test positions and a laugh or
two.
First, a preface. Although they each borrow from each other, I believe that
three categories of computer-slaying have cropped up:
a) "Anti-chess" espoused by Pablo Restrepo, locking up a position and winning on
time (I think this is solely the domain of internet chess).
b) Specially prepared variations or traps (e.g., the "Nemeth Gambit") which are
repeatedly practiced against a computer until a satisfactory finished product
can be presented.
c) Outplaying the opposition in real time.
Obviously c) is the most desirable, yet is quickly becoming the least
obtainable. Between the other two, a) is more "boring" as many have said, yet,
I believe, more technically impressive. Each game is played in real time, such
that one cannot prepare for a specific opponent (Shredder, Fritz, etc.),
specific condition (computer speed), or specific variation, and follows only a
general strategy. Seeing the game in which Pablo defeated Fritz a queen down
was awe-inspiring. Of course it didn't have the thrill of, say, Tal demolishing
someone with a spectacular sacrifice, yet it represented a seemingly impossible
feat that was amazingly accomplished. That said, I think that b) is less
technically impressive because it is accomplished by preparing a trap for an
opponent to fall into under very controlled conditions (computer speed, time
control, opening variation, and first-move privilege). Nevertheless, the latter
are typically more visually appealing as games of chess, as I hope the following
games are.
I first got the idea for this post by noticing that Delfi's small book allowed
it to fall for a somewhat well-known trap. 6...Nxd5?? transposes into a losing
variation of the Lolli Attack.
[Event "Computer chess game"]
[Site ""]
[Date "2005.07.24"]
[White "Me"]
[Black "Delfi 4.5"]
[Result "1-0"]
[TimeControl "300+1"]
1. e4 e5 2. Bc4 Nf6 3. d4 exd4 4. Nf3 Nc6 5. Ng5 d5 6. exd5 Nxd5 7. O-O f6
8. Nc3 dxc3 9. Bxd5 fxg5 10. Re1+ Be7 11. Bxg5 Kf8 12. Qf3+ Bf6 13. Rad1
Bd7 14. Bc4 b5 15. Be6 Bxe6 16. Rxd8+ Rxd8 17. Bxf6 Bd5 18. Be7+ Kg8 19.
Qxc3 Nxe7 20. Rxe7 Bf7 21. h4 Re8 22. Rxc7 h5 23. Qf3 Rf8 24. Qb7 Rh6 25.
Qxb5 a6 26. Qd3 Rf6 27. Qd4 Re8 28. c4 Re1+ 29. Kh2 Rfe6 30. Qd3 R1e2 31.
c5 Re8 32. Qf3 Bg6 33. Qb7 Kh7 34. Rxg7+ Kh6 35. c6 Rxf2 36. c7 Rf6 37. Rd7
Bf5 38. Qf3 Kg6 39. Qg3+ Bg4 40. Rd8 Rff8 41. c8=Q Rxd8 42. Qxa6+ Rf6 43.
Qb5 Rdf8 44. Qg5+ Kh7 45. Q3e5 R8f7 46. Qe8 Rf8 47. Qe4+ Bf5 48. Qe7+ R8f7
49. Qgxf6 Rxe7 50. Qxe7+ Kg8 51. Qg5+ Kf7 52. Qxf5+ Ke7 53. Qxh5 Ke6 54.
Qg6+ Kd5 55. h5 Kd4 56. h6 Ke3 57. h7 Kd2 58. h8=Q Ke3 59. Qc3+ Kf2 60.
Qgg3+ Kf1 61. Qce1#
The nice-looking book move 8.Nc3 wins for white (8.Re1 transposes after ...Be7
9.Nc3, so check the pv for when it finds Nc3 with a winning score.)
[D] r1bqkb1r/ppp3pp/2n2p2/3n2N1/2Bp4/8/PPP2PPP/RNBQ1RK1 w kq -
The following game features a strange opening, strange sequence of sacrifices,
strange king-hunt, and a strange endgame (Post-game Fruit analysis says I
blundered on move 45, but at least Delfi didn't notice). I don't know *exactly*
what's going on or if any of my moves were good (in hindsight, I'd guess a 1-0
justifies the unsound, speculative means), so I'll leave it to the better
players to extract any possible test positions.
[Event "Computer chess game"]
[Site ""]
[Date "2005.07.24"]
[White "Me"]
[Black "Delfi 4.5"]
[Result "1-0"]
[TimeControl "300+1"]
1. e4 c6 2. d4 d5 3. Nf3 dxe4 4. Ng5 Bf5 5. Bc4 e6 6. Nc3 b5 7. Bb3 Be7 8.
Ncxe4 h6 9. Nxf7 Kxf7 10. Ng3 Bg6 11. O-O Nf6 12. f4 Nd5 13. f5 exf5 14.
Nxf5 Bf6 15. Nh4 Be4 16. Qg4 Re8 17. Bxh6 gxh6 18. Rae1 Qd7 19. Qh5+ Ke7
20. Rxe4+ Kd8 21. Qg6 Rxe4 22. Qxe4 a6 23. Nf5 h5 24. Ne3 Nxe3 25. Rxf6 Nd5
26. Rf5 Kc7 27. Bxd5 cxd5 28. Rxd5 Qg4 29. Qe7+ Kb6 30. Qc5+ Kb7 31. Rd6
Qd1+ 32. Kf2 Qd2+ 33. Kf3 Qd1+ 34. Ke3 Qg1+ 35. Kd3 Qf1+ 36. Kc3 Qc4+ 37.
Qxc4 bxc4 38. Rd8 a5 39. Rh8 Nc6 40. Rxa8 Kxa8 41. Kxc4 Nb4 42. Kb3 Kb7 43.
c4 Nd3 44. g3 Ne1 45. c5 Nd3 46. a3 Kc6 47. Kc3 Nf2 48. Kc4 a4 49. d5+ Kd7
50. b4 axb3 51. Kxb3 Ne4 52. Kc4 Nd2+ 53. Kb4 Nf1 54. a4 Ne3 55. d6 Kc6 56.
a5 Nc2+ 57. Ka4 Nd4 58. a6 Ne6 59. Kb4 Nd8 60. a7 Kb7 61. Kb5 Kxa7 62. c6
Kb8 63. c7+ Kc8 64. cxd8=Q+ Kxd8 65. Kc6 Kc8 66. d7+ Kd8 67. h4 Ke7 68. Kc7
Kf6 69. d8=Q+ Kf5 70. Qg5+ Ke4 71. Qxh5 Kd4 72. Kd6 Ke3 73. Qe5+ Kf3 74.
Qf4+ Ke2 75. h5 Ke1 76. h6 Kd1 77. h7 Kc2 78. h8=Q Kb3 79. Qb8+ Kc2 80. Qd4
Kc1 81. Qbb2#
{White mates} 1-0
Lastly..
Dear dissidents,
These games are provided as entertainment and proof of concept. Therefore,
I'm not obligated to play an arranged match against your Shredder on 8
processors, to waste 7 hours playing at classical time controls, upgrade my old
1.4 GHz computer to the bleeding edge just to prove myself, or otherwise fulfill
any of your condescending and antagonistic whims. I am not a grandmaster; the
purpose of these games is not to play like one, but to demonstrate a legitimate
victory. Do not make patronizing demands that do not correlate with the purpose
of my experiment and whose only effect is to bolster your own ego. Thank you.
Sincerely,
[Insert name of much-maligned computer slayer]
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.