No Human is now capable of or will there ever be a human capable of,
playing a chess position that will challenge todays top programs using quad or above hardware.
If you find a human chess position that a program will think for 3 minutes or more and then choose that position and stay with the same position, then the human does not understand the position and/or the human chose the position for the wrong reason.
You will never find a good test position analysing SuperGM (cough cough) games today or in the future. Since we have just about analysed all past human chess positions, our only hope for challenging chess positions in the future will be from stronger hardware, 4 socket, 8 socket, 8 core dies, program improvements, clusters etc.... and these will only come from computer vs computer games.
Be4,
kgburcham
SuperGM/GM moves past and present vs todays programs
Moderators: hgm, Rebel, chrisw
-
- Posts: 2016
- Joined: Sun Feb 17, 2008 4:19 pm
-
- Posts: 6340
- Joined: Mon Mar 13, 2006 2:34 pm
- Location: Acworth, GA
Re: SuperGM/GM moves past and present vs todays programs
[d]r3r1k1/pp1q1p2/2p2npb/PPPp4/3Pp3/2NNPpP1/1R1B1P1K/3Q3R b - -
"Good decisions come from experience, and experience comes from bad decisions."
__________________________________________________________________
Ted Summers
__________________________________________________________________
Ted Summers
-
- Posts: 20943
- Joined: Mon Feb 27, 2006 7:30 pm
- Location: Birmingham, AL
Re: SuperGM/GM moves past and present vs todays programs
Not so fast. What about Shirov's Bh3 sacrifice that is extremely difficult to find? Or many of the LCT positions, etc... There are lots of positions that humans have correctly solved that programs can't touch yet...kgburcham wrote:No Human is now capable of or will there ever be a human capable of,
playing a chess position that will challenge todays top programs using quad or above hardware.
If you find a human chess position that a program will think for 3 minutes or more and then choose that position and stay with the same position, then the human does not understand the position and/or the human chose the position for the wrong reason.
You will never find a good test position analysing SuperGM (cough cough) games today or in the future. Since we have just about analysed all past human chess positions, our only hope for challenging chess positions in the future will be from stronger hardware, 4 socket, 8 socket, 8 core dies, program improvements, clusters etc.... and these will only come from computer vs computer games.
Be4,
kgburcham
-
- Posts: 3245
- Joined: Thu Mar 09, 2006 9:10 am
Re: SuperGM/GM moves past and present vs todays programs
I agree with that and like the fact that you are careful enough to use the word "yet".bob wrote:
There are lots of positions that humans have correctly solved that programs can't touch yet...
I am however of the opinion that games between Rybka 3 and Naum 4 on a Quadcore machine are already of much better quality than games among top GMs.
Matthias.
My engine was quite strong till I added knowledge to it.
http://www.chess.hylogic.de
http://www.chess.hylogic.de
-
- Posts: 4367
- Joined: Fri Mar 10, 2006 5:23 am
- Location: http://www.arasanchess.org
Re: SuperGM/GM moves past and present vs todays programs
This game also is beyond most engines' comprehension:
http://www.chessgames.com/perl/chessgame?gid=1287604
http://www.chessgames.com/perl/chessgame?gid=1287604
-
- Posts: 12541
- Joined: Wed Mar 08, 2006 8:57 pm
- Location: Redmond, WA USA
Re: SuperGM/GM moves past and present vs todays programs
Analyze this game with a computer at tournament time control and see if it understands all of the moves:
[Event "Sigeman & Co"]
[Site "Malmoe SWE"]
[Date "2001.06.18"]
[Round "7"]
[White "Berg, E."]
[Black "Hector, J."]
[Result "1/2-1/2"]
[WhiteElo "2474"]
[BlackElo "2546"]
[ECO "C34"]
[Opening "KGA"]
[Variation "Schallop defence"]
[EventDate "2001.06.12"]
[PlyCount "85"]
1. e4 e5 2. f4 exf4 3. Nf3 Nf6 4. e5 Ne4 5. d4 d5 6. Bxf4 c5 7. Bd3 Qb6 8.
O-O cxd4 9. Nbd2 Bf5 10. Nb3 Bc5 11. a4 a6 12. a5 Qa7 13. Nxc5 Qxc5 14. b4
Qxb4 15. Bc1 Nc6 16. Ba3 Qxa5 17. Bf8 Qc7 18. Bxe4 Bxe4 19. Bd6 Qd7 20. Ng5
f6 21. Nxe4 dxe4 22. Qh5+ g6 23. Qh4 f5 24. g4 Qe6 25. Rfb1 Ra7 26. Rb6
Nxe5 27. Bxe5 Qxe5 28. Ra5 Qg7 29. Re6+ Kd7 30. Rae5 Kc8 31. Re8+ Rxe8 32.
Rxe8+ Kd7 33. Qd8+ Kc6 34. Re6+ Kc5 35. Qb6+ Kc4 36. Qb3+ Kc5 37. Qb6+ Kc4
38. Qxa7 Qh6 39. Qxb7 Qe3+ 40. Kf1 Qf3+ 41. Kg1 Qxg4+ 42. Kf1 Qd1+ 43. Kg2
1/2-1/2
[Event "Sigeman & Co"]
[Site "Malmoe SWE"]
[Date "2001.06.18"]
[Round "7"]
[White "Berg, E."]
[Black "Hector, J."]
[Result "1/2-1/2"]
[WhiteElo "2474"]
[BlackElo "2546"]
[ECO "C34"]
[Opening "KGA"]
[Variation "Schallop defence"]
[EventDate "2001.06.12"]
[PlyCount "85"]
1. e4 e5 2. f4 exf4 3. Nf3 Nf6 4. e5 Ne4 5. d4 d5 6. Bxf4 c5 7. Bd3 Qb6 8.
O-O cxd4 9. Nbd2 Bf5 10. Nb3 Bc5 11. a4 a6 12. a5 Qa7 13. Nxc5 Qxc5 14. b4
Qxb4 15. Bc1 Nc6 16. Ba3 Qxa5 17. Bf8 Qc7 18. Bxe4 Bxe4 19. Bd6 Qd7 20. Ng5
f6 21. Nxe4 dxe4 22. Qh5+ g6 23. Qh4 f5 24. g4 Qe6 25. Rfb1 Ra7 26. Rb6
Nxe5 27. Bxe5 Qxe5 28. Ra5 Qg7 29. Re6+ Kd7 30. Rae5 Kc8 31. Re8+ Rxe8 32.
Rxe8+ Kd7 33. Qd8+ Kc6 34. Re6+ Kc5 35. Qb6+ Kc4 36. Qb3+ Kc5 37. Qb6+ Kc4
38. Qxa7 Qh6 39. Qxb7 Qe3+ 40. Kf1 Qf3+ 41. Kg1 Qxg4+ 42. Kf1 Qd1+ 43. Kg2
1/2-1/2
-
- Posts: 16465
- Joined: Wed Aug 01, 2007 4:16 am
- Location: Canada
Re: SuperGM/GM moves past and present vs todays programs
A computer won't find 1..Kg7!! an excellent move if I'm not mistaken.AdminX wrote:[d]r3r1k1/pp1q1p2/2p2npb/PPPp4/3Pp3/2NNPpP1/1R1B1P1K/3Q3R b - -
Ted would you mind giving the game score or a link to this game?
Thanks!
Terry
-
- Posts: 12541
- Joined: Wed Mar 08, 2006 8:57 pm
- Location: Redmond, WA USA
Re: SuperGM/GM moves past and present vs todays programs
See:Terry McCracken wrote:A computer won't find 1..Kg7!! an excellent move if I'm not mistaken.AdminX wrote:[d]r3r1k1/pp1q1p2/2p2npb/PPPp4/3Pp3/2NNPpP1/1R1B1P1K/3Q3R b - -
Ted would you mind giving the game score or a link to this game?
Thanks!
Terry
http://rybkaforum.net/cgi-bin/rybkaforu ... l?tid=7234
-
- Posts: 16465
- Joined: Wed Aug 01, 2007 4:16 am
- Location: Canada
Re: SuperGM/GM moves past and present vs todays programs
Thanks Dan! Wow! Rybka got it!!!Dann Corbit wrote:See:Terry McCracken wrote:A computer won't find 1..Kg7!! an excellent move if I'm not mistaken.AdminX wrote:[d]r3r1k1/pp1q1p2/2p2npb/PPPp4/3Pp3/2NNPpP1/1R1B1P1K/3Q3R b - -
Ted would you mind giving the game score or a link to this game?
Thanks!
Terry
http://rybkaforum.net/cgi-bin/rybkaforu ... l?tid=7234
Still, I don't know how healthy it is to think like Fischer!?
-
- Posts: 814
- Joined: Sat May 09, 2009 4:51 pm
- Location: Toronto
Re: SuperGM/GM moves past and present vs todays programs
This reminds me a bit of Averbakh - Kotov, Zurich '53AdminX wrote:[d]r3r1k1/pp1q1p2/2p2npb/PPPp4/3Pp3/2NNPpP1/1R1B1P1K/3Q3R b - -
.