Author: Aaron Tay
Date: 10:39:13 01/16/02
Go up one level in this thread
On January 16, 2002 at 13:17:08, Severi Salminen wrote: >>Not sure what you mean by "understand" but I recall seeing many Winboard chess >>engines play b4 and follow up with b5 in QGD positons when out of book. >> >>I suppose some sort knowledge based on the pawn chain/structure would be >>sufficent..And once b4,b5 is played the rest would be handled automatically by >>search, bonus/penalties for backward,isolated pawns? > >Well, Requiem doesn't know anything about minority/majority attack, opposition, >weak/strong bishops, but still it sometimes seems to understand those aspects. >Requiem pushes many times pawns even without any knowledge. So unless you see >source codes, you can't possibly know what the program actually evaluates. You >are right: fast search makes many evaluation terms obsolete, IMHO. > > >Severi Which leads me back to what i said. "Understand" is a rather vague term. There might not be a specific "knowledge" about minority attack (whatever that means) but if the program plays as if it understands due to a combination of other factors, , who are we to judge whether it has or does not have that knowledge?
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.