Author: Uri Blass
Date: 05:46:01 06/14/04
Go up one level in this thread
On June 14, 2004 at 07:47:50, Fabien Letouzey wrote: > >Hi Uri, > >> I think to look at fruit's source code to learn about it's evaluation. > >> Fruit nunn based rating increased from 2496 to 2642 based on the >> following link http://www.husvankempen.de/nunn/ratings/ratingsnunn2.htm > >> I understand that it is only thanks to better evaluation when it even >> has no king safety evaluation. > >You can get an idea of the improvement due to only evaluation changes >(plus activation of the single-reply extension) by comparing Fruit 1.0 >and the private version Fruit X 05/05 (or 05/09) which is still >playing in Heinz's Nunn Top 2 tournament. > >I later (Fruit X 05/21) added "time management", an often-overlooked >feature that seems to help a lot in fast games. There were also minor >speedups not to mention the use of a a special (beta?) version of the >Microsoft compiler thanks to Slater Wold. I remembered that you did not work on the search algorithm and forgot about better time management so you are right and it seems that I exagarated about it. > >Overall, too many differences to conclude about evaluation I am afraid. >From my point of view however, most of the time was spent on it. This >is what I mean by "mostly evaluation changes" (time management was a >one-day work plus a second one to fix it for pondering). > >> Did you look at Fruit1.5's source code to learn about it's evaluation. > >> I still did not look much at it but I guess that the fastest way for >> improvement should be to try to learn from it. > >My main point is that Fruit 1.0 had almost no evaluation to start >with. It should come to little surprise that there was a lot of room >for improvement there. In other words I had a problem (Fruit 1.0 >played extremely passively) and I "fixed" it (still not good, but more >active style). > >I suggest you don't lose (too much) time reading Fruit's evaluation >code. Feel free to ask questions though. thanks. questions and verification if I understand correctly. 1)Is it correct that Fruit does not evaluate positions when the king is in check? 2)Am I correct to assume that op[] is opening evaluation when eg[] is endgame evaluation? 3)What is the job of material_get_info? As far as I can see it is using material_comp_info I understand that material_comp_info does the following: 1)calculate drawn positions in simple endgames. 2)clauclating phase of the game between opening and endgame to decide about weight of endgame evaluation and opening evaluation 3)calculate pawn correction for rook and knights. It seems to me that you give bonus for knights if there are more than 5 pawns and you give panelty for rooks when there are more than 5 pawns. In other word the claim is that more pawns are good for knights and less pawns are good for rooks. 4)calculate bonus for pair of bishops 5)encouraging the better side to have more pawns and to trade other pieces(I will probably try to implement it in similiar way). I see that later in the evaluation you decide to divide the evaluation by 2 in case of opposite color bishops and the question is what is the definition of it(Do you consider only endgame with bishops to be opposite color bishop endgame or also endgame with rooks and bishops?) I still did not get to evaluation of pawn structure and evaluation of mobility that is also interesting for me because I do not think that I like my evaluation of these factors but I guess that I will do it in another post. > >Unfortunately I don't know what is the fastest way to progress for >you. You mentionned before that Movei was not using the transposition >table to full power, it sounds like something promising to fix?! > >Fabien. yes but it seems harder for me to do it. Movei is using evaluation that is dependent on the path and not only on the final position and the code of the search is too ugly to make it easy to make complicated changes without bugs. I believe that a small push of the evaluation to fruit direction can help. search changes can also help but it seems to me harder to write them without bugs. 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.