Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Crafty questions

Author: Robert Hyatt

Date: 10:34:45 08/04/05

Go up one level in this thread


On August 04, 2005 at 12:46:54, Madhavan wrote:

>On August 04, 2005 at 12:15:51, Robert Hyatt wrote:
>
>>On August 04, 2005 at 11:16:55, Kevin K wrote:
>>
>>>Hello, Dr. Hyatt.
>>>
>>>My question is about easy move in root.c:
>>>
>>>  if (sort_value[0] > sort_value[1] + 200 &&
>>>      ((To(rmoves[0]) == To(last_opponent_move) &&
>>>              Captured(rmoves[0]) == Piece(last_opponent_move)) ||
>>>          sort_value[0] < PAWN_VALUE)) <------ ?
>>>    easy_move = 1;
>>>
>>>I understand recapture is easy move.
>>>But, I don't understand why if sort_value[0] < PAWN_VALUE, it is easy move.
>>
>>Suppose your opponent appears to hang a piece out of the blue.  That is, your
>>last search produced a score of +.2, and then you made that move.  When your
>>opponent makes a move, and you start off with a score of +2, do you believe that
>>he hung a piece, or perhaps the piece is "poisoned"?  Here is a position from
>>Cray Blitz vs Belle, ACM 1981, that illustrates the problem.  Qxb6 loses badly.
>>Yet it appears to win material to a very shallow search.
>>
>
>where do I get your source code with commented style.I don't find any comments
>or explanation of "how,why,when and what" in the code

First place is to look in main.c for a description of changes, version by
version.  However, the code is about 50% comments so comments are everywhere.
I'm not sure I always explain the "why" for everything when it was something
that was done prior to crafty (the "easy move" stuff came from Cray Blitz...



>Tom and explained precisely how they did the code.
>
>Have you seen Fruit's code?what's missing part of Crafty relative to fruit?does
>fruit have more knowledge in it than crafty has?or what makes Fruit stronger
>than crafty?[IMHO Fruit is stronger,ofcourse but don't challenge me]


Have not looked at any fruit code, and have not looked at any fruit-crafty games
either so I am not very up-to-date on how fruit plays...



>
>
>>the above simply says "if the score is significantly plus, I'm going to be
>>suspicious that my opponent is not that stupid, and use all the time in case it
>>is a trap...



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.