Author: Mark Young
Date: 12:12:52 09/01/98
Go up one level in this thread
On September 01, 1998 at 02:09:28, Ed Schröder wrote: >>Is computer Vs computer testing now useless in gauging a chess program’s >>strength playing humans? > >Always been. I did not use to think that, and from other post here not all programmers would agree. > > >>When Crafty gets killed playing Junior 5 by a wide >>margin. And Fritz 5 draws a match with Rebel 10 even when Rebel 10 has a 2x >>hardware advantage. Is it time to abandon Computer Vs Computer testing all >>together? > >Nope, it remains fun. And a fast and easy way to get a lot of games. Games >are analyzed, weak points discovered and then improved. I think many >chess programmers work that way or at least it is a part of their testing >methods. > But have we come to rely to much on computer Vs computer testing to improve our programs in the mistaken belief that better computer Vs computer results equal better computer Vs human results. And if so does this not hurt computer Vs human program development. > >>Or are we going to have two standards to judge chess programs? > >This has always been the case. comp-comp is a different story than Man vs >Machine. You can not compare them. There can be easily a gap of 100-150 >elo points up, or even down. > >Best example Crafty. It seems to do a lot better against humans than against >computers. The same applies for Rebel although to a lesser extend. Rebel >since its existence always performed better against humans than against >computers. > > >> One chess program being the best playing other chess programs and one chess >>program being the best playing humans. > >True. > > >>And if so what is the best standard to judge a programs overall strength? > >Impossible, far too less Man vs Machine games. > > >>Is it better marketing to show you can destroy all other programs like Junior5 >>and Fritz 5 can do, or is it better to show you can beat a top grandmaster >>like Rebel 10 can do? > >If the topic is marketing the latter of course. Playing a top grandmaster >is not cheap, add up all the work (time) involved for a good preparation, >organization, promotion, etc. you get a pretty huge investment. There must >be a base to do such things other than to find out how good your program >is against such top players. Man vs Machine simply is the real work. Just >remember the two Deep Blue - Kasparov matches and their impact. > >- Ed -
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.