Author: Mark Loftus
Date: 10:17:27 08/20/98
Go up one level in this thread
On August 20, 1998 at 08:10:37, Robert Henry Durrett wrote: >On August 19, 1998 at 07:25:52, Mark Loftus wrote: ><snip> > >>I have a lot of respect for Mr. Silman and his down to earth style. Also I >>think people should realize that the ssdf list is not the gospel, this is not a >>knock on their hard work - but when computer programs go up against masters, IMs >>and GMs, it is a whole new ballgame. The problem with the use of lists like the >>ssdf is that the programs will be tuned to beat the other programs and not get >>the improvement in strength you think you are getting. However, if IM's and GMs >>say the program is strong or GM, then I believe it. I think the better measure >>of a programs strength is against humans. ><snip> > >Wouldn't it be a good thing if all of the games [or at least the results] played >by human masters and GMs against the main computer chess engines were compiled >in some database somewhere? Then a competent statistician, such as Arpad Elo >could create "best estimates" of the strengths of the chess programs, based >SOLELY on games with humans. Yes, that's a something I would like to see. The rating based on games against humans I think would be the best indicator of a programs strength. Now that there are no more Aegeon tournaments, something like this would be more valuable. Mark
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.