Author: David Dahlem
Date: 08:05:18 11/19/04
Go up one level in this thread
On November 19, 2004 at 10:46:49, George Tsavdaris wrote: >On November 19, 2004 at 10:16:11, David Dahlem wrote: > >>On November 19, 2004 at 07:18:28, George Tsavdaris wrote: >> >>> Actually, for an experienced Chess-engine user like you, it is easy to see that >>>when having a specific "small" score for many plies up to 16, which doesn't >>>change at all, then the position is most probably a draw if we are talking for >>>an edngame. (Smarthink finds the draw but it's materialistic DNA, prevents it >>>from showing a zero score.) >> >>A draw score is 0.00 as evaluated by engines. It's not up to the user to read >>the mind of an engine and interpret something different than the engines >>evaluation. An evaluation of +1.23 certainly can not be interpreted as a draw >>score!! >> > > What do you mean with "it's not up to user"? If you suggest that the user >should see ONLY the score of an engine in order to say if the position is >"white/black wins" or "white/black has the advantage" or "game is draw" then i >believe you are wrong. > Someone has to see BOTH the score and the tendency of the score to value a >position. > > For example if an engine for position A, gives the following: >Ply-12 = 6.00 >Ply-13 = 5.87 >Ply-14 = 5.48 >Ply-15 = 4.32 >Ply-16 = 2.52 > >while at position B, gives: >Ply-12 = 1.02 >Ply-13 = 1.37 >Ply-14 = 2.48 >Ply-15 = 2.32 >Ply-16 = 2.52 > >then you most probably(in most cases/positions) be wrong if you assume that both >positions are big advantages for white, since at P-A we have a huge fail low >that we can't be sure where it will end, while at P-B we have a more stable >situation, although the scores at both positions are the same. > > Another example is if an engine for position C which is an endgame with very >few pieces, gives the following: >Ply-10 = 1.12 >Ply-11 = 1.14 >Ply-12 = 1.14 >Ply-13 = 1.14 >Ply-14 = 1.14 >Ply-15 = 1.14 >Ply-16 = 1.14 >Ply-17 = 1.14 > >while at position D which is an endgame with very few pieces, gives the >following: >Ply-12 = -0.12 >Ply-13 = -0.16 >Ply-14 = 0.00 >Ply-15 = 0.37 >Ply-16 = 0.58 >Ply-17 = 1.14 > >then you most probably(in most cases/positions) be wrong if you assume that both >positions are advantages for white, since at P-C we have a stable "small" score >for the last 8 plies, while at P-B we have a fail high that may end to a score >+2.50 for example in favour of white, although the scores at both positions are >the same. > This stable "small" score shows that most probably white hasn't any really good >move to do, but only just play it's pieces around although it has a small >advantage(a Pawn more or some mobility more) and since most engines don't have >the ability to interpret that this small advantage will lead to a draw, they >just show this small advantage until we reach 50-move rule. I made my point, that +1.23 is not a draw score, and i stand by that. I'm not going to argue with you. Regards Dave
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.