Author: chandler yergin
Date: 10:49:10 01/03/05
Go up one level in this thread
On January 03, 2005 at 09:56:27, Uri Blass wrote: >On January 03, 2005 at 07:26:07, chandler yergin wrote: > >>On January 01, 2005 at 19:30:12, Ulysses Omycron wrote: >> >>>>The Significant part of the Eval is the first digit! >>>>It shows Zero! (0.50) That means neither side has an advantage. >>>>The numbers after the Decimal are in hundredths of a Pawn,and based on >>>>Static Positional features in a Quiescent Position. These are Programed >>>>weights given for King Safty, Mobility, Pawn structure, strong/weak squares >>>>etc. etc. >>>> >>> >>>You are slightly right, but not at all. >> >>Closer than you are my friend (Read the Manual!) >> >>> >>>Draw score is 0.00 (Unless you tweak it), that means the Engine knows it's going >>>to draw (Though most of the time the 0.00 goes away after a while). A score >>>higher or lower than 0.00 means someone has an advantage... Even an advantage >of 0.10 (1/10 of a pawn) doesn't mean draw. >> >>At that point in time it does! >>As I pointed out; the 0.xx is based on Static positional programed weights! >>The slight = or - simpy means either White or Black has better King safty, >>mobilty better pawn structure etc. etc. >> >>The Significant point is the 1st digit! >>Until that shows a + 1.xx or - 1.xx or better; >>neither side has an advantage. >> Consider it a Draw! >> >>Even a Win is not certain until a + 3.00 or a -3.00 or better is reached. >> >> >> >>> >>>I think the main poster wanted to know if any engine give the CORRECT score of >>>0.00 and not something lower (Black has advantage) or higher (White has >>>advantage). >> >>He probably did.. he simply doesn't understand. >> >>You CANNOT say CORRECT SCORE! The engine evaluates based on how it is >>programmed. There is also a 'contempt' factor which can be programmed in. >> >>With pieces on the Board; one of them usually has slightly better, though >> not decisive, static positional features, as determined by the Programmer. >>Different engines will evaluate a position differently based on these factors. >> >> >>> >>>If 0.50 meant neither side has advantage, then an engine could pick any move at >>>random as long as it doesn't goes below/avobe the 0.50 margin to give the enemy >>>advantage. >> >>There maybe several moves that the engine evaluates the same; and will select >>the 1st one examined then. >> >>They wouln't since they'd lose, just try it in CB's friendly mode in >>>where choosing 0.50 has this effect, 1.00 is just slightly better than 0.99, >>>0.99 doesn't mean "neither side has an advantage" >> >>Yes it DOES! When you get to + 1.0 or - 1.0, >>then and only than does one have a pawn advantage! >> >>Read the Manual. > >The manual is not relevant here. The HELL it's NOT! Guess ya can't teach an old dog new tricks. > >Every programmer will tell you that when a program say +0.50 for white it means >that white has the advantage(it does not say if it is enough to win but it says >advantage). Well, they are WRONG! If you can't understand what the .XX IS, you shouldn't be Programming! > >Maybe it is better if people do not answer you because you always argue about >things that the posters do not mean. If they can't articulate clearly and precisely, than they don't understand what they are trying to say! >When posters say seeing a draw they mean seeing 0.00 score. Well they are wrong! They see.. but they don't understand! That's because they don't understand what the Eval really means. >When posters say finding a move they mean finding in one best mode. I know what they mean... > >You should stop give original explanation to things and talk in the terminology >that almost everybody is using. > >Uri I can't help if they are all wrong! ;) CY A Rose by any other name still smells.!
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.