Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Is the NPS tend to grow at the end of the game?

Author: leonid

Date: 08:34:36 07/22/00

Go up one level in this thread


On July 22, 2000 at 09:20:23, Josť Carlos wrote:

>On July 22, 2000 at 08:46:25, leonid wrote:
>
>>Hello!
>>
>>Is the nodes per second number tend to grow when program search position more
>>advanced in the game? When there are less pieces on the board.
>>
>>I just speeded my program in around 30% and tried to find where my nps is now.
>>Went to the Fritz 6 counter (Fritz engin inside of Hiarcs package) and was
>>disappointed. Fritz still keep me at the safe distance at the end of the game.
>>This is due to the strange (for me) fact that Fritz numbers goes up at the end
>>of the game. This time in mine nps goes down. In mine program less are the
>>pieces on the board, less efficency in the search of the legal moves. Wanted to
>>see if there are some general tendency in all programs. Before asking here my
>>question, went to see Rebel 10. Its numbers looks like to be very stable. No big
>>difference between heavy initial and final positions. Once again difference with
>>my program.
>
>  My program (Averno) usually searches 300% faster (in nps) in the endgame. I
>never have thought about it cause I saw it natural, but if I think a little, I
>can give some reasons:
>
>  - My evaluation in the endgame is easier (no king-safety code).


But when you still have few pieces on the board? For instance some 11 pieces and
few of them can attack the king. Probably your procedure for the safety of king
check what kind of pieces are on the board and check only for the pieces that
can attaque. In my I use general procedure that will look for all possible
attaque even when few variation of the pieces are no more on the board.


>  - Less moves generated = less time spent generating moves. :)


If you generate few moves for given ply, your generation of the moves should be
more efficent when the data for legality of the moves will be used for greater
number of moves. Less number of moves use needed data, less efficent your move
generation should be. But I see that you go somewhere against this logic and
very successfully.

>  - Better use of trans/ref table (many more transpositions).

This could be the response. I have nothing like this.

Thanks!
Leonid.


>>Thanks for response,
>>Leonid.
>
>  Hope this helps.
>
>  Josť C.



This page took 0.08 seconds to execute

Last modified: Thu, 07 Jul 11 08:48:38 -0700

Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.