Author: martin fierz
Date: 10:36:57 10/19/02
Go up one level in this thread
On October 19, 2002 at 11:23:34, Mark Young wrote: >A very disappointing end to the match right! >as a draw was accepted much too early by >the Fritz team with this type of endgame structure. There was a good chance to >gain the advantage for black in this endgame with little risk of losing. i don't think so... the IQP is not always weak... >The GM Kramnik vs. Deep Fritz match has put to rest some questions, or maybe >not! > >1. Do PC programs play at a GM Level? – Clearly Yes. > >2. Can PC programs win against a well prepared and motivated Grand Master? – >Yes. > >3. Is it easy to force the PC programs into unfavorable position were the human >has the advantage when the human player does not control the opening book? – No. > >4. Was Deep Blue 1997 was much stronger then Deep Fritz and other PC programs? - >No. > >5. Can a PC program compete with the worlds best chess players even when it is >for sale to the public and everyone has a chance to play against the program? – >Yes. > >6. Is it easy to trick a PC program with anti-computer tactics? – No. you can't conclude that, for two reasons: 1) kramnik never even tried to play classical anti-computer games (close the position and go for the king). 2) kramnik's anti-computer tactics consisted of exchanging the queen and beating fritz in the endgame. this worked *very* well until the fritz team changed to openings where the games stayed more complicated. so actually, he was beating the PC program with his own flavor of anti-computer tactics, but the fritz team saved fritz by changing the book! >7. Do computer programs have too many weaknesses in there play to let them be >competitive with the worlds best chess players. Clearly No. > >8. Will it take another 10 years before anyone can have a computer that plays as >well as Deep Blue? No. > >9. Do computers need to search many millions of nodes per second as Deep Blue >did to be competitive with the world’s best player(s)? – No. i have another conclusion to add to the list: 10. do todays super-GM-level computer programs still make mistakes that no 1600 player would make? yes. :-) aloha martin
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.