Author: Jarkko Pesonen
Date: 02:53:41 02/19/02
Go up one level in this thread
On February 18, 2002 at 14:48:50, Chessfun wrote: >On February 18, 2002 at 14:42:17, Albert Silver wrote: > >>On February 18, 2002 at 14:25:05, Roy Eassa wrote: >> >>>Chess Tiger 14 is 100 points WEAKER than TheKing ?? >> >>Keep in mind a few things: >> >>- The engines are playing with ponder OFF, which although ideal for a single >>machine, can seriously affect the overall result. >>- The performance is for this event only and with varying time controls, so it >>can easily be the reverse in a next event. >>- The programs are only using 4 piece TBs. Christophe has already said that he >>thought Tiger's performance (he was commenting the SSDF when saying this) could >>be affected by the use of incomplete TBs. >>- The tournament's biggest purpose (unplanned) IMHO has been to show the >>strength of new settings for CM8000 showing it to be an engine of the highest >>quality suffering from less than optimal default settings. >> >>Still, I must admit to finding it fascinating. >> >> Albert > >Your missing a few other; >The relative few number of games. >The large + - of those ratings caused by the relative few number of games. > >Sarah. > > >> >>> >>> >>> >>>On February 18, 2002 at 11:38:43, Kurt Utzinger wrote: >>> >>>>Program Score % Elo >>>> >>>>1 TheKing312X 43.0/70 61.4 2467 >>>>2 Fritz7002 39.5/70 56.4 2437 >>>>3 Junior7 35.5/70 50.7 2404 >>>>4 ChessTiger14 31.0/70 44.3 2367 >>>>5 Hiarcs732 31.0/70 44.3 2367 >>>>6 Shredder6 30.0/70 42.9 2358 At the moment error bars are about +- 60-80 ELO points. Can anyone say how many would be needed to reduce errors further smaller to 40, 20 ,10 ,1 ELO points?
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.