Author: blass uri
Date: 08:02:54 02/19/99
Go up one level in this thread
On February 19, 1999 at 08:35:56, Reynolds Takata wrote: >I don't mind if you fudge one little game and just declare it a draw, don't play >60 more moves if it gets to 100 draw it, everyone here would sympathize and >understand. I do not agree about it. I saw cases when computers lost a dead draw positions and I do not think that it was possible to be sure in more than 99% about the result. I can understand not playing 160 boring moves but in this case the result should be that the game was not completed and not a draw unless you can be sure in more than 99% about the result. There are only some cases when I understand adjudications in clearly winning positions or in known simple drawn positions when in both cases I can be almost sure(more than 99%) about the result. I can never be sure about the result before the game ends because there are many strange things that are rare but can happen to programs. 2 examples: Programs may go to an infinite loop and lose on time because of a bug. I found this bug in Genius3 (this is rare but can happen and it happened in one of the studies that I analyzed with it). Programs may have a bug in understanding the 50 move rule and allow mate in the last move(I think that Hiarcs6 lost against Fritz5 in the ssdf games because of this bug). Uri
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.