Author: Robert Hyatt
Date: 18:21:37 11/10/00
Go up one level in this thread
On November 10, 2000 at 20:27:00, walter irvin wrote: >why are computers held to a 1000 x higher standards than humans .like for >example when kramnik thumped kasparov .the match was very very short compared to >the matches people make computer programs play .and still people claim that oh >that is not enough games to tell if a program is stronger .if that is indeed the >case than every human world championship ever held was much too short to have >told who the better player was . Easy test: play 400 games. Then put the results in a long string of 0's and 1's (throw out the draws for simplicity here. Now say one program won 240-160. Look at the string of 0's and 1's. (0 means program a lost, 1 means it won, and program a is the match winner here.) Can you find a group of 10 consecutive zeros? Even though that program won? probably yes. Suppose that happened to be the 10 games you played in the 10 game match? You think B is absolutely destroying A. Computer games have far more luck involved than human games. Choosing an opening is risky. Stumbling into unclear positions is risky. Etc.
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.