Author: Peter Fendrich
Date: 11:20:13 09/27/99
Go up one level in this thread
On September 27, 1999 at 10:03:05, Georg v. Zimmermann wrote: >I would like to know why people do even look at a list like the SSDF, with less >than 200 games played for the top programs. > >I did run a test crafty16.19-cometB06 and got the following results (from nunn >positions,nothing changed but time control, more details in "forget about >computer-computer matches"): > >15 min games 15,5 : 4,5 > >14 min games 14 : 6 > >5 min games 9,5 : 10,5 > >I thought that this shows that maybe less than 200 games played between 2 >programs does not tell you which is how much stronger. SSDF played less than 200 >overall for one prog. ! > >Now since some much more experienced people still take the SSDF list seriously >I'd like to know what is wrong about my conclusion. > >Best Regards, > >Tec-- ...why do people take human ELO lists seriously? Each ELO is based on less games and what is messured is a "moving target" over time, as opposed to the programs that are stable over time. //Peter
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.