Author: Steve Maughan
Date: 18:59:45 11/06/03
Go up one level in this thread
Terry, >>The results of test and matches are not opinions but facts. > >Yes, but the results are not scientific proof that Shredder 7.04 is in fact >stronger than any engine in the top ten! The difference is too small. At the 95% level Shredder is clearly better. According to the SSDF it leads Junior 8 by 28 ELO, having played 781 games. This difference is *NOT* too small - at least not at the 95% level. The error margin is +28 / -26. Of course it is invalid to add the error margins of Junior 8 and Shredder 7 because you'd be effectively increasing the confidence levels above 95%. Of course you could argue that 95% is not a high enough probability and you'd prefer 99.9999% - but then you're heading down the road of computer chess narcissism - an ideology that seems to be in favor here!?! Regards, Steve
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.