Author: Chris Carson
Date: 06:45:37 06/26/01
Go up one level in this thread
On June 26, 2001 at 09:35:45, Martin Schubert wrote: >On June 26, 2001 at 09:20:46, Chris Carson wrote: > >>On June 26, 2001 at 09:06:42, Andrew Williams wrote: >> >>>Over the last month, there have been a number of such huge arguments. >>>I've always thought that it's a bit strange to get hung up on this >>>question. Five years ago, it was clear that the best micro programs >>>were not at GM strength. And presumably if you wait five years, PC >>>programs will have proven beyond any doubt that they are at GM strength. >>>Surely the time we happen to be living in is the best and most enjoyable, >>>because we're perhaps seeing a moment of transition between these positions. >>>Why then the need to convince anybody of anything when you can just sit >>>there and be proved right by waiting? Is there some particular benefit to >>>being able to say that PC programs are GMs *now*? >>> >>>Andrew >> >>I do not care if they are or they are not GM's. I do care what ELO strength >>they are. >> >The problem with the ELO is that there's the danger to compare things you can't >compare. ELO is one number for chess between humans. Computers play totally >different chess. So the ELO you get for a computer depends on what kind of games >you play. If the players are experienced in playing computers. >Of course if you let a top program play against players which are not very >experienced playing against computers you'll get a good rating. You wouldn't get >such a good rating when letting the program play against more experienced >players. >A lot of people are very interested in ELO. But it's just one number! You can't >put every knowledge about chess just in one number. So IMO it's far more >interesting to ask where can computers help the humans and where can't they >instead of just looking at one number. > >Martin > > >>Why not determine what the strength is now? I made a statement 2 years ago >>about the strength on 500Mhz machines and was challenged to provide more than >>one game to prove my point. I did that. >> >>Why is it so important to wait 5 years? >> >>If it is not important to you. Great, don't read the posts, I ignore a lot of >>stuff on this board that is not important to me. :) >> >>Best Regards, >>Chris Carson That is a good point. ELO is the measure for performance. That is what I am talking about, that is what I have always said. If you want to discuss something else such as "what are the weakness of programs or program x" that is a good topic, adds information, but is not part of the "are programs performaning at ELO 2600 or higher level. Best Regards, Chris Carson
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.