Author: Harvey Williamson
Date: 09:54:26 01/04/06
Go up one level in this thread
On January 04, 2006 at 12:21:40, Dann Corbit wrote: >On January 04, 2006 at 11:13:32, Eduard Nemeth wrote: > >>>>Next: Thoralf says that Fruit will played too agains weaker engines, but in my >>>>opinion should test next Programs like Hiarcs 10, Rybka 1.2, SmarThink 1.0 and >>>>Ktulu 7.5. >>>> >>>>The Freature is also: Not against weaker Engines but against stronger Engines. >>>> >>>>That is for me the reaso why I see this Rating List as a Joke - real! >>>> >>>>ED. >>> >>>I do not agree. >>> >>>part of the ability of chess program is in beating weak programs and if Fruit >>>beats Nimzo8(450 mhz) more convincingly than Fritz then it is something positive >>>about fruit. >>> >>>Uri >> >>Uri, see the all matches and You will see: Fritz 9 vs. Fritz532 (on P200!) 3-1. >>For me is this a Joke. Why play Fritz vs Fritz? Why not against the ChessMaster >>before? Many Things (in my eyes) are mysterious by SSDF. > >There is simply no doubt that the SSDF is the very best data available to tell >us the strength of programs. > >There is nothing I can see that is definitely wrong with their methodology. >Often, the complaints against them are very misguided. We should *expect* some >strange match results. But with hundreds of games those things even out. And >the use of opponents that are 200 Elo weaker or stronger is not a serious defect >(but huge Elo differences would be bad.) > >Of course, as with any experimental setup, it could be possible to improve it. >But there is nothing better to tell us a more accurate answer. For certain, the >SSDF tells us far more than contests like WMCCC and Paderborn. And far more than random games from the Playchess(Fritz) server.
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.