Author: Roy Eassa
Date: 08:04:43 09/22/03
Go up one level in this thread
On September 19, 2003 at 01:30:20, Russell Reagan wrote: >Are time controls too long for serious computer vs. computer chess games? > >I think that long time controls such as 40 moves in 2 hours aren't necessary on >today's hardware in serious computer vs. computer chess games. In 1999 the WCCC >was decided using a time control of 40 moves in 2 hours, 30 moves in 1 hour, and >the rest of the moves in 30 minutes. Given that hardware doubles every 18 >months, this is equivalent to a time control of 40 moves in 15 minutes, 30 moves >in 7.5 minutes, rest of the game in 3.75 minutes. These are considered blitz >time controls today, but these exact same time controls were used to decide the >world champion of computer chess, and this was only in 1999. Imagine the >equivalent time control to the time controls that decided some of the very first >WCCC/WMCCC events! > >I think time controls such as game in 60 minutes with a 30 second increment are >more than sufficient for serious computer vs. computer chess games, and this >gives the bonus of being able to play more games, which is prefered IMO when >trying to decide the superior engine. Game in 1 hour is equivalent to game in 8 >hours (!) in 1999, and I think everyone would agree that for a serious computer >vs. computer event, game in 8 hours is a little ridiculous. > >Correspondance chess is another matter entirely, since that involves human >opponents. Russell, Hurray for saying out loud what I've been wondering for years! As PCs get faster, why not use shorter and shorter time controls in matches between programs, in order to get more games in (statistically more significant) while _NOT_ reducing the average ply reached or the average nodes searched per move.
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.