Author: Peter Kappler
Date: 17:14:09 08/25/02
Go up one level in this thread
On August 25, 2002 at 19:04:34, Gian-Carlo Pascutto wrote: >books Gent 2002 > 1234567890123456789012345678901234567890 >1 Sjeng 12.14 Maastricht 2500 10½110½110½1½½½½½½1½½000½10½½½110½½10½01 >21.0/40 >2 Sjeng 12.14 Maastricht AH 2500 01½001½001½0½½½½½½0½½111½01½½½001½½01½10 >19.0/40 > >Both programs are identical, but the first one is using Fritz7.ctg (pro book) >and the second one is using Shredder6.ctg (random book). > >I played 3'+2" on an Athlon XP 1800, 128M hash per engine. > >The difference is on the order of 20 points, so certainly nothing significant >(may just be random noise, even). > >My impression was that the one using Fritz7.ctg came out of the opening better >(with white) or equal (with black), but the engines made enough mistakes that >this was 'washed out' most of the time. > Yes, I have noticed this effect, too. IMO, most amateur engines (even the stronger ones like Sjeng) make enough middlegame and endgame mistakes that games are rarely won because of small opening advantages. Would be interesting to test this theory by repeating your experiment with weaker amateur programs, and some top commercial programs, too. -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.