Originally posted by no1marauder Berliner's was 82% in Kepler's analysis. Here's the entire (interesting) thread: http://www.timeforchess.com/board/showthread.php?threadid=97088&page=1#post_1817591
I went on and improved on that later. Using a variety of engines and longer times I got Berliner up to 85% over a much larger sample of games. His performance in the world championship is also "better" than my original analysis. I suspect that this is the result of the engines getting closer to human play if given greater resources such as processor power, time etc rather than the moves mysteriously becoming more engine-like.
Basically i borrowed my son's multi core PC and ran HIARCs, Rybka and Shredder. Give an engine five minutes a move on a fast machine and it is surprising how human they start to appear.
What was the percentage match up of Kramniks games against topalov in their reunification match? Topalov seemed to think it was high enough to indicate engine use, but few people believe him, so I was thinking maybe that might be a good benchmark as to how close a strong human player can match up to an engine without using one.
Originally posted by Tyrannosauruschex What was the percentage match up of Kramniks games against topalov in their reunification match? Topalov seemed to think it was high enough to indicate engine use, but few people believe him, so I was thinking maybe that might be a good benchmark as to how close a strong human player can match up to an engine without using one.
That I don't know. I think I used those games but didn't keep details of each game separate. I don't recall Kramnik's match rate being massively high in any game to be honest.
Originally posted by Squelchbelch I can see the 2008 Championship win being awarded 2 or 3 times over the next year or so unless swift action isn't taken.
If game moderation will not work then it will not be possible...... 😕