Author: Graham Banks
Date: 16:28:27 05/09/04
Go up one level in this thread
On May 09, 2004 at 09:58:27, George Tsavdaris wrote: >On May 09, 2004 at 06:34:05, Kurt Utzinger wrote: > >>On May 09, 2004 at 06:15:15, Graham Banks wrote: >> >>>I've discovered the reason for Hiarcs 9 poor performance and I'm too embarrassed >>>to divulge why, but it's my fault. I've pulled Hiarcs 9 out of the tournament >>>and will rerun all its games at the end. > > >What was the wrong thing with it's setup? > > >>> >>>Also when doublechecking everything else I found SmarThink 0.17a has only been >>>using 20mb hash as opposed to all other programs using 128mb! This was due to >>>the incorrect setting in its ini file. So I will also pull SmarThink 0.17a out >>>and rerun all its games also. >>> >>>So the tournament crosstable with 10 engines will conclude this week and then >>>the Hiarcs 9 and SmarThink 0.17a games will be rerun. >>>Very annoying and I apologise. However it will all be worthwhile in the end! >> >> >> Hi Graham >> Such things can happen. But it's fine that you are going to >> rerun all games. As far as SmarThink 017.a is concerned I >> do not think that there will be much difference in performance >> between 20 MB and 128 MB hash. The importance of big hash >> size is much overestimated. Interesting to see if your games >> will support my "theory". > > > Yeah, i also don't think the difference would be important. I believe that a >20 MB hash instead of 128 MB, means something about 1 point less out of 35 >games. Of course i may be wrong. > > >> Kurt & Rolf Chess: http://www.utzingerkurt.com Trouble is I'm a perfectionist, so I'll be rerunning the Hiarcs 9 and SmarThink 0.17a games so that the tournament was consistent for all programs. Don't want to say anything more about my error with Hiarcs 9 as I feel stupid enough already! Graham.
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.