Author: Ryan B.
Date: 02:48:28 02/18/06
Go up one level in this thread
On February 18, 2006 at 05:23:42, George Tsavdaris wrote: >On February 18, 2006 at 03:35:23, Uri Blass wrote: > >>On February 17, 2006 at 20:03:41, William Penn wrote: >> >>>I won't try to give examples here, but my impression is like Vincent's. I >>>regularly compare long analysis from Fruit and Rybka in infinite mode (usually >>>several hours run time) and their evaluations are remarkably similar, in >>>general. I haven't noticed any major exceptions. But that's just my impression, >>>I haven't researched or quantified it. >>>WP >> >> >>My impression is different based on analysis of correspondence games. > >Yes, i also have the same impression with you. > >But what is really important is what you said in another post: >Rybka is missing endgame knowledge that Fruit already has. So i don't think >there is anyone believe _and if we assume that Rybka is a Fruit-clone_ that >Vasik actually removed the endgame knowledge Fruit has. As you have shown in >some endgames, Fruit understands the positions while Rybka not due to missing >knowledge. >That can't be explained in another way than the obvious..... > Just a quick note, Gambit Fruit had less endgame knowledge than Fruit 2.1. > > Also William says about REMERKABLE similarities. All these are simple words >with no meaning, until he posts the positions that he has seen these remarkable >similarities as i haven't find any until now.... > Vincent did the same. Why he doesn't post the "0.1 position" in order to let us >see the truth instead of only hearing it.....? > > >> >>I clearly see cases of disagreement and it is not that one is always right. >> >>difference of more than 0.5 pawns is significant difference and there were cases >>when I saw it in practical analysis. >> >>The position that I gave is not from my correspondence games but from analysis >>of one of my tournament games when rybka understand immediatly that black is >>better and fruit see score that is close to draw (even at depth 3 that is >>probably depth 5 of other programs because i do not believe vasik's information >>about depth). >> >>I think that if you try unbalanced positions when one side has passed pawns and >>the other side has material compensation then you can see often significant >>difference between fruit and rybka and unbalanced positions happen. >> >>In one case there was no significant difference in evaluation but there was a >>significant difference in the suggested move and I believe that rybka's move was >>simply better. >> >>It is important to choose the better move when the evaluation of many moves is >>almost the same. >> >>Uri
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.