Author: Richard Patterson
Date: 10:39:22 07/06/03
Go up one level in this thread
On July 06, 2003 at 12:47:46, Mike Byrne wrote: >it looks much different than the first tourney with 3 of the top names at top in >this one... > >For The King results I always have to check the status of the OPK --so that is >subject to change. > >I think in most cases one of the commercials will win and result like I had the >first time with so many of the Crafty versions doing well is rare. Also a top >program can do poorly once in a while - like Deep Sjeng/Ruffian - so there is >some randomness in results of just afew game ... > >seeing this randomness convinces me that you need at least 600+ games to start >getting results that are meaningfull when engines are this close and in any >given game, a weaker engine can beat a a stronger engine ..and sometimes can >lose to to 2 or 3 weaker engines in the same tourney. > >Fritz 8 is having a much different result in this tourney - all the settings >are the same btw >as the first tourney > > > >Blitz Secane, PA USA 2002 > > 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 >1 Fritz 8 * ½ 0 1 0 1 1 1 1 1 1 7.5/10 >2 Hiarcs 8 ½ * 1 1 ½ 0 1 0 1 1 0 6.0/10 >3 Chess Tiger 15.0 * 1 0 1 0 0 1 ½ 1 ½ 1 6.0/10 >4 Crafty 19.03 Nimzovich 1 0 0 * 0 1 ½ 0 1 1 ½ 1 6.0/11 >5 Crafty 19.03 ST 0 0 1 * 0 1 1 1 ½ ½ 1 6.0/10 >6 Crafty 19.03 Lasker 0 1 * 1 ½ 0 ½ 0 1 1 1 6.0/10 >7 Crafty 19.03 Capablanca 1 0 1 0 * ½ 1 0 ½ ½ 1 5.5/10 >8 Shredder 7.04 1 ½ 0 ½ ½ * 1 1 0 0 ½ 5.0/10 >9 Crafty 19.03 Morphy ½ 1 1 1 0 0 * 0 0 ½ 1 5.0/10 >10 Deep Junior 8 0 1 0 0 0 * 1 ½ 1 1 ½ 5.0/10 >11 Crafty 19.03 Petrosian 0 0 ½ ½ 1 1 0 * ½ ½ 1 5.0/10 >12 Crafty 19.03 Stein 0 1 0 0 1 ½ * ½ ½ 1 ½ 5.0/10 >13 The King ST 0 0 ½ 0 1 ½ 1 0 ½ * 0 1 4.5/11 >14 Ruffian 1.0.5 0 0 0 ½ 0 1 0 ½ ½ 1 * 3.5/10 >15 Crafty 19.03 Byrne 0 ½ ½ 0 ½ ½ ½ ½ 0 0 * 3.0/10 >16 Deep Sjeng 1.5 0 1 0 0 0 0 0 ½ 0 ½ * 2.0/10 Michael, I would agree that a higher reliability can be obtained by a greater number of games played. Once with black and once with white also may make a difference but I still think a even greater number of games should be played to get some sort of assurance. Richard
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.