Author: Mark Young
Date: 09:13:01 06/12/98
Go up one level in this thread
On June 12, 1998 at 10:55:32, Graham Laight wrote: >Looks like it's the end of the line for the SSDF list. > >It was due to update in early April. It's now mid June, and it still >hasn't updated. > >Anyway, the autotester methgodology entails the computers playing >thousands of games that no human will ever see. So what. You do not have to see the games, just the results. >Human chess players don't play thousands of games between 2 players, >just to make sure the result is exactly right. > You miss the point of why autotesting is done. You need a large number of games to show that one program is truly better then another. And its much better to have many games against more then just 2 players. >In sport, the winner is the person who wins on the day. Who produces a >little extra magic when it is needed. > Chess is not football. No one plays just one game of chess for a championship. >The autotester methodology is taking the game too far away from human >experience for it to be worth doing in the context of providing sport or >entertainment for mankind. Autotesting is not done for entertainment. >I challenge anyone to tell me why autotesting is a good thing to do! > It helps the consumer gauge what program is most likely to be stronger. >Having said that, I eagerly await the next SSDF list. :-)
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.