Author: Thorsten Czub
Date: 06:30:09 09/03/99
Go up one level in this thread
On September 03, 1999 at 02:35:06, Ed Schröder wrote: >On September 02, 1999 at 18:25:52, Thorsten Czub wrote: >>the series shows how the chessbase autoplayers is doing it. >>if you would have played the games each game with a reset >>pc, without autoplayer cable, your >10.game results would >>more like the first 10 games result. >>in fact your result exactly shows what we always said and talk >>about, but we have no evidence HOW they do it. >>i am sure ed knows HOW they do it. therefore next rebel century >>comes WITH autoplayer and new learning algorithm. >> >>ONE more thing: the cyrix cpu is not very fast. >>This hurts cstal, since cstal does many times less NPS than hiarcs. >>also 60/60 is not that good, 40/120 would be better. > >I think there is nothing wrong with the CB autoplayer. Just copy the >original opening book from cdrom (all learning is then deleted) and >start to (auto)play again. > >Ed if you do this after EACH game, than there is anything OK. But the ssdf-guys and the guy having played these games, has not done so. it is NOT the learning alone in my opinion. and for me it remains: it is cheating. i don't have these results playing manually. they play automatically and get these cheated-results.
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.