Author: Bruce Moreland
Date: 09:59:02 07/22/98
Go up one level in this thread
On July 22, 1998 at 03:48:48, Howard Exner wrote:
>The net tournament of game/30 where the computers did so well.
>(Wasn't Ferret part of that computer group?)
People have brought that tournament up several times. I didn't regard it as very
significant at the time.
>Are the few examples I've given not worthy candidates of accomplishments
>made by chess programs?
>If Rebel-Anand isn't noteworthy which of the above are?
>
>I wasn't looking for some rigid test that the programs had to pass
>to be included. For example, I think an accomplishment would be
>citing some of the current ICC computer ratings. Put in an historical
>perspective that to me is quite an achievement.
I think the chess servers are also interesting because of the large volume of
games against strong players, and the overwhelming success that computers have
had there.
This was extremely surprising to me. When I started my program on ICC, in late
1994, I had no idea what to expect, as it had never played against a human other
than me before.
I never would have expected that all the programs there, including mine, would
give grandmasters a hard time, but this was true even then.
I just did a search on some of the old games that were played there, and the
first two I found are signifcant. Here is a two-game match at 2 8 between
GnuChess and GM Loek van Wely, from October 1994:
[Event "ICC 2 8 10/06/1994"]
[Site "Internet Chess Club"]
[Date "1994.10.06"]
[Round "-"]
[White "Zippy"]
[Black "KingLoek"]
[Result "1-0"]
[WhiteElo "2321"]
[BlackElo "2615"]
[ECO "B81"]
[NIC "SI.20"]
1. e4 c5 2. Nf3 d6 3. d4 cxd4 4. Nxd4 Nf6 5. Nc3 e6
6. g4 Nc6 7. g5 Nd7 8. Be3 Be7 9. h4 O-O 10. Qe2 a6
11. O-O-O Nxd4 12. Rxd4 b5 13. f4 Rb8 14. h5 b4 15. Nd1 e5
16. Rd2 exf4 17. Bxf4 Bxg5 18. Bxd6 Bb7 19. Bxb8 Bxd2+ 20. Qxd2 Nxb8
21. Qxb4 Qg5+ 22. Kb1 Ba8 23. Nc3 Nc6 24. Qa4 a5 25. Bb5 Nd4
26. Qxd4
{Black resigns} 1-0
[Event "ICC 2 8 10/06/1994"]
[Site "Internet Chess Club"]
[Date "1994.10.06"]
[Round "-"]
[White "KingLoek"]
[Black "Zippy"]
[Result "0-1"]
[WhiteElo "2643"]
[BlackElo "2293"]
[ECO "D20"]
[NIC "QG.04"]
1. d4 d5 2. c4 dxc4 3. e4 c5 4. d5 Nf6 5. Nc3 b5
6. e5 b4 7. exf6 bxc3 8. bxc3 Nd7 9. fxe7 Qxe7+ 10. Be3 Qe5
11. Qd2 Nb6 12. Rd1 Bb7 13. Nf3 Qxd5 14. Qb2 Qe4 15. Be2 Be7
16. O-O O-O 17. Ne1 Rfb8 18. Bf3 Qf5 19. Qe2 Bf6 20. Rc1 a5
21. Bxb7 Rxb7 22. Nf3 Re8 23. Rfd1 Nd5 24. Qxc4 Nxe3 25. Re1 Rb2
26. Rxe3 Rxe3 27. fxe3 Qg6 28. g3 Qf5 29. Rf1 Qc2
{White forfeits on time} 0-1
The first of these was won on a blunder, but GnuChess was up a pawn at the time.
Would you bet on GnuChess against a GM on 1994 hardware at 8-second increment?
Since then it has only gotten worse, there are very few humans who can come
anywhere near 50% against any of the computers, and contrary to popular
expectation, not just at 5 0, a large number of these games are 5 5 and 2 12.
bruce
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.