Author: Uri Blass
Date: 07:39:57 09/10/02
Go up one level in this thread
On September 10, 2002 at 10:22:52, Thorsten Czub wrote: >On September 10, 2002 at 09:44:23, Uri Blass wrote: > >>Here I do not agree. >> >>If it is losing against the old version in a long match it means that there is a >>problem. > > >nonsense. >it's not important how a version plays against the version before. >there will be no championship where somebody participates with YOUR older >versions. I believe that significant improvement should be against everything and if the program is losing against the old version then it means that there are some type of positions when the new program is weaker. > >>no >>Games between x and x+1 give you more than final result and you can learn from >>them. > >no - you cannot. the program versions are too similar. > >you cannot come to senseful data. for the same reason it is important >that people are not relatives when saying or swearing something at court. >the first question always is: is this guy your relative. and it is asked because >the similarity of familie members, their relationship changes the judgement, >stops the lawyers from finding out the truth. I disagree. If the new version is losing against the old version then there is a reason for it. Looking at the games and analyzing positions that the new version blundered can help to find the reason and fix the problems. 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.