Author: Keith Kitson
Date: 09:13:51 07/29/99
Go up one level in this thread
Robert, Your comments are interesting. I posted a message some time ago regarding what I thought were potential inconsistencies in the way people could test the programs. There are so many configuration parameters that are not quoted with game results that it is difficult to believe any results, or at least put them into context. Can we really take any results as being a true reflection of both programs playing at their strongest. That is supposing that the strongest settings are already known. So many variables are now being brought into play with the latest programs that i reckon it is difficult to determine which is the strongest setting without a long series of games on several settings. I am not paranoid about the integrity of testing, although it may seem that way. I am more concerned that lots of testers with very good intentions do not realise how they can affect the results of a program. They do not have enough experience to understand all of the variables that can affect a program. This being the case i find it very dangerous for any results posts to be taken seriously unless the full configuration under which the program was executed is also supplied. Keith
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.