Author: KarinsDad
Date: 18:16:41 06/08/99
Go up one level in this thread
On June 08, 1999 at 20:20:49, Peter Kappler wrote: >On June 08, 1999 at 17:31:47, Christoph Fieberg wrote: > ><snip> >> >>The gnawing question is: HOW HIGH IS THE MAXIMUM ELO NUMBER? >> > >I think this is an interesting question. > >Let's assume that Kasparov faced a computer opponent that played perfect chess - >i.e. it had the "ultimate tablebase". :-) > >Would Garry lose every game, or could he draw some games against a perfect >opponent? My guess is that the drawing margin in chess is large enough that >Kasparov would draw occasionally, but probably not more than 10% of the games. My guess is that he would loss them all since even he plays nowhere near perfect chess. Looking only 6 to 10 ply down (on average) means that he HAS to be making mistakes every game. They are just real subtle mistakes. And this can be illustrated by looking at Garry's endgames. Normally, he has an advantage walking into the endgame. But analyze his endgames where he does not have an advantage against a tablebase and you will see that he does not play perfect chess (or anywhere near). Garry playing against a perfect full game tablebase would be like me playing against Deep Blue. Not a contest. Ever. Just my guess... :-) KarinsDad :) > >A 10% win expectance correlates to approximately a 400 point ELO rating >difference, and since Kasparov is ~2850 FIDE (after his recent victories) this >would mean the "perfect" chess player would have a rating of at least 3250. > >Just my guess... :-) > >--Peter
This page took 0.03 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.