Author: Marc van Hal
Date: 12:49:50 05/22/03
Go up one level in this thread
On May 22, 2003 at 12:34:52, Uri Blass wrote: >On May 22, 2003 at 12:14:54, Uri Blass wrote: ><snipped> >>Note that I do not say that testing against the previous version >>is enough and I agree that it is possible that there is no improvement >>inspite of positive results against previous personality. >> >>My opinion is that an improvement that I like >>should be also against previous version but not only >>against it. >> >>Uri > >I think about it and it seems that there is a contradiction in my opinion >because it is possible to say that there are cases when change from A to B is no >improvement but the opposite change is also no improvement and I do not say that >they are the same level. > >I think that there are cases when we only can say that we cannot compare. >If you score worse against the previous version but better against other >opponents then we cannot compare and the same for opposite change. > >I do not like a change for point A to B and I do not like a change from point B >to A and I see the target as trying to find point C that I like both relative to >A and relative to B. > >It is possible that it cannot be done only by changing parameters. > >Uri Corect I mentioned that before. Things which are bad in the raw engine can't be corected with different settings. Only slightly improved. But also Amir Ban came up with the same conclusion. Marc
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.