Author: Andreas Stabel
Date: 02:25:45 08/02/05
Go up one level in this thread
On August 02, 2005 at 03:59:23, Thorsten Czub wrote: >Genius (all Richard Lang programs) computed asymetrically. > >Instead of a normal capture search, it had static exchange evaluator and >this special search. plies 1-3-5-7-9 were pruned different then 2-4-6-8... > >the effect was that it did not always THE BEST move in ply 1-3-5-7-... > >result: boring playing style. no active but passive behaviour. > >the effect: computing against a NORMAL program, it happened very often that >genius did a move the opponent would NOT have imagined. > >breaks PB. > >CSTal was opposite. it pruned more in opposite plies. >result: active play. sacs. > >when CSTal sacced it mostly never thought the opponent would TAKE the sac. > > >There was a Saitek dedicated chess computer that was not only computing on ONE >PB move but on many. > >The effect was that almost all the time the machine had a move and the move came >in an instant. it was a good blitz opponent :-) > >no - the name of the machine was NOT Saitek Blitz. This was a strange machine in >design and technic, but it had not the DIFFERENT pb. > >don't forget that many people use chess engines on ONE PC and there most often >(depending on hardware used) >the PB is knocked off. Thanks - very interesting. Do you know if this was done explicitely to stop the opponent from predicting the move ? Wouldn't this technique also cause Genius and CSTal to not predict the opponents moves in a lot of cases ? Regards Andreas
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.