Author: James T. Walker
Date: 15:24:13 09/20/99
Go up one level in this thread
On September 20, 1999 at 17:56:37, eric guttenberg wrote: >I note that H7.32 scored 22.5-17.5 over F5.32 at 40/2. However I believe >several posts have shown F5.32 outperforming H7.32 at 40/2 in manually >played games(Shep's recent tournament comes to mind though I well understand >that it was only 10 or 11 games for each program). >Playing a game at 40/2 manually is so time consuming that there may never >be enough such games to permit any definite conclusions, but is there not >a suggestion that in some cases autoplayer results may differ from manual >results? >I do not want to be in Thorsten's corner on this (although there seems to >be plenty of room there) but this is not the first time that someone has >raised the question of whether the autoplayer can somehow affect the play >of a program. >Does anyone think that it is possible that, for example, H7.32 is slightly >stronger than F5.32 on autoplayer, but that F5.32 is slightly stronger >in manual games? > > >eric Hello eric, It's food for thought but I honestly don't believe it. I do believe that manually entering moves can affect the outcome of a game. I believe if it is done without intent the eventual outcome of a long match will be unaffected. Just by delaying a few seconds before relaying a move can cause a different move to be played. It usually only takes one different move to change a game entirely although it may not change the eventual outcome. One of the problems with manual games is they sometimes last 10-12 hours if played to the end. This leaves a lot of room for the operator to influence the game when taking care of eating or other duties or being distracted by a telephone. As I said if this is all random then the eventual outcome of a long match should be unaffected. Jim Walker
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.