Author: Pham Hong Nguyen
Date: 22:34:00 10/28/02
Go up one level in this thread
On October 29, 2002 at 00:34:41, Uri Blass wrote: >On October 28, 2002 at 20:28:43, Pham Hong Nguyen wrote: > >>Hi all, >> >>I read that someone suggests the margin between the evaluations and materials >>should not excess 1.5 of pawn. However, my margin is 1.5 of ... queen and I >>think that causes many problems for my program: difficult to apply razoring, >>lazy eval, cut off in qsearch... >> >>So my questions here are: >>- What is the margin of Crafty? >>- Could you share some experiences about your margin and its >>advances/disadvances? >> >>Many thanks for any answers. > >My margin is not defined because I did not define my values for material. >It is possible to do a linear regression in order to guess the evaluation based >on material but I did not do it. > >Pieces gwt different value on different squares that and I have not one value >for the pieces. >6 pawns in the 7th rank are usually evaluated as better than 8 pawns. > >If people tell you that the difference between evaluation and material should >not excess 1.5 pawns then they tell you that you cannot give the right score for >pawns in the 7th. >It also means that you cannot evaluate unstoppable passed pawn in pawn endgames. > >Uri Uri, I understand your way to evaluate because my margin is also much larger than 1.5 pawns. Just want to get some experiences about use of materials in some functions with which I have problems. For example, I cannot use the values of 1.5-2 pawns for razoring based on materials like many programs, I have to use value of rook, it turns out I gain very little. Do you do any thing based on materials? such as lazy eval, razoring,...?
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.