fail high handling with tranposition tables

Discussion of chess software programming and technical issues.

Moderators: hgm, Harvey Williamson, bob

Vempele

Re: fail high handling with tranposition tables

Post by Vempele » Fri Aug 03, 2007 5:24 pm

After searching 1.Kb2 Kb6 score 0 for all lines
second move to search is Kb8
What is alpha and beta after Kb8?
-1 and 0. Kb8 fails high and has to be researched with alpha = 0, beta = 30.

Uri Blass
Posts: 8015
Joined: Wed Mar 08, 2006 11:37 pm
Location: Tel-Aviv Israel

Re: fail high handling with tranposition tables

Post by Uri Blass » Fri Aug 03, 2007 5:53 pm

Vempele wrote:
After searching 1.Kb2 Kb6 score 0 for all lines
second move to search is Kb8
What is alpha and beta after Kb8?
-1 and 0. Kb8 fails high and has to be researched with alpha = 0, beta = 30.
This is the same idea that I use but I use it only for root moves.
I did not know that principle variation search means doing research also for non root moves.

I understand now the idea and I will try to implement it later and see if it helps.

Basically everytime when the window for the first move is bigger than 1
you use window of 1 for the second move based on the guess that it is going to fail low in order to do a research later and I can understand that in this condition most of the searches are with null window.

Uri

Post Reply