Author: Uri Blass
Date: 01:07:26 01/21/02
Go up one level in this thread
On January 21, 2002 at 03:52:29, David Rasmussen wrote: >On January 21, 2002 at 00:23:08, Russell Reagan wrote: > >>On January 20, 2002 at 16:21:02, David Rasmussen wrote: >> >>>On January 20, 2002 at 15:38:45, martin fierz wrote: >>> >>>>On January 20, 2002 at 09:01:50, David Rasmussen wrote: >>>> >>>>>There must be a value system of material that takes care of all special cases. >>>>> >>>>>1,3,3,5,9: >>>>> >>>>>Has the following problems: >>>>>3 pawns for bishop or knight is almost always a bad idea. >>>>in the middlegame, yes, in the endgame no. >>>> >>>>>2 knights/bishops for rook and pawn is almost always a bad idea. >>>>in the middlegame yes, in the endgame, maybe not, if the rook has a passer. >>>> >>>>>2 rooks for queen is often not a good idea. >>>>not at all, except if the queen can mate you. >>>> >>>>>3 knights/bishops for a queens is often not a good idea. Then again, often it is >>>>>:) >>>>it is a good idea more often than not. >>>> >>>>all these things are heavily dependent on the rest of the material on the board. >>>>i don't think you have a chance... >>>> >>>>cheers >>>> >>>> martin >>> >>>I am not talking about things that depend on a lot of other stuff. Evaluation >>>should take care of this. I am talking about Crafty that explicitly checks for >>>some material special cases _and nothing else on the board_, and decides if it >>>should give a penalty/bonus. I say that I think it can be done with the values >>>alone. Why not? >>> >>>/David >> >>You cannot assign one set of values to the chess pieces and use that for >>material evaluation. Material evaluation changes depending not only upon which >>of the other pieces are on the board, but upon positional factors. Accurately >>assessing material value can only be done by assessing those positional values >>along with the other material on the board (which would be one complicated >>algorithm), or by breaking material evaluation down into it's atomic parts, >>which I believe is the correct way to do it. You have to evaluate chess upon the >>elements of the game, which are determined by the rules of the game. Deep >>insight into this area is the only way to allow for an efficient evaluation of >>any element of a chess position. >> >>Russell > >For the 1000th time... I am not suggesting to drop any complex evaluation terms, >and just have magical material values. I am suggesting to use other values than >1,3,3,5,9. If you don't believe that material values are important why do you >use 1,3,3,5,9 (or whatever)? If you _do_ think it matters, why do you think >1,3,3,5,9 is magical? > >/David Nobody said that 1 3 3 5 9 are magical. The point is that the first thing that we need is a clear definition what are the values. I use piece square table and the value of the piece is included in the piece square tables. I have not one number for a knight but 64 numbers for knight based on the square of the knight so it is not clear for me what do you mean by the value of the knight in my program. Things are going to be more complex when you add mobility and other factors. A possible way to try to get an estimate is simply to take a lot of random positions from games and calculate static evaluation with the knight and static evaluation without the knight and to decide that the average difference is the value of the knight but before discussing about the value of the pieces we should have a clear definition what is the meaning of the value of the pieces. 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.