Author: Terry McCracken
Date: 23:16:13 01/28/03
Go up one level in this thread
On January 29, 2003 at 01:33:38, Jouni Uski wrote: >According to SSDF list (and it's the best we have) Junior is 68 points after >Fritz! >And I don't believe, that it's better against humans than most before >statistically >proven evidence - needs about 200 games minimum... > >Jouni Ah...200 rating points, constitutes a full class difference, not 68 points! Also these ratings have an error rate of up to 50 points up or down, so even a programme rated 100 points higher than another, could be of equal strength. Rating stats, can be and often are in error, by this margin, whether it is man or machine. Although it may be shrinking, I feel. It's an inaccurate science, so don't take the SSDF as actual reality of playing strength with this variance or half this variance. It's only a very good estimate of true playing strength. Terry P.S. Deep Junior hasn't been beaten in any human match for 2 years now. Junior _is_ an excellent programme. Terry
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.