Author: John Merlino
Date: 17:59:05 05/28/01
Go up one level in this thread
On May 28, 2001 at 16:15:55, Steven J. Brann wrote: >On May 28, 2001 at 13:27:47, John Merlino wrote: > >>On May 28, 2001 at 00:04:26, Graham Banks wrote: >> >>>I'm no expert on these. It would be interesting to get the programmers' >>>perspectives, eg, is it better to set them at odd or even? And I wonder how much >>>impact the difference say between 7 and 12 actually has on performance at the >>>various time controls. >>>Regards, Graham. >> >>These are VERY general rules of thumb: >> >>1) The faster the time control, the lower you want the SS value to be. >>2) The faster the processor, the higher you can make the SS value. >> >>As an example, at the most recent CSVN tournament, Johan used SS=10 with his >>Athlon 1.3GHz, and the time control was G/90. If he had been using his trusty >>550MHz, he would have used SS=6. >> >>jm > >What about the SS setting in the recent SSDF match between CM8K and Junior? Did >the SS setting remain at the default (7) for this match? What settings changes >(if any) were made for this encounter? If it remained at SS=7, why wasn't it >extended for the longer time control since there appears to be a fairly dramatic >increase in performance at long time controls? Were there any (other) changes >in settings for this match? > >Steve Firstly, the default for the Chessmaster personality is SS=6, and it was not changed for the SSDF matches. The only change I know of to the default personality was changing the hash table size to 32MB (I think -- maybe it was 16MB). jm
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.