Different handling of ALL / CUT nodes

Discussion of chess software programming and technical issues.

Moderators: hgm, Harvey Williamson, bob

Forum rules
This textbox is used to restore diagrams posted with the [d] tag before the upgrade.
Post Reply
User avatar
Greg Strong
Posts: 388
Joined: Sun Dec 21, 2008 5:57 pm
Location: Washington, DC

Different handling of ALL / CUT nodes

Post by Greg Strong » Sun Mar 22, 2009 6:39 pm

Has anyone tried hadling expected cut nodes differently from exected all nodes? For example, trying more aggressive move ordering or a more incremental approach to move generation?

Daniel Shawul
Posts: 3593
Joined: Tue Mar 14, 2006 10:34 am
Location: Ethiopia
Contact:

Re: Different handling of ALL / CUT nodes

Post by Daniel Shawul » Sun Mar 22, 2009 8:53 pm

I did not use node classification for the purposes you mentioned but
I used if for determining different bounds for deciding where to start the LMR reduction. Another one is for classifying nodes for SMP splitting purposes where you want to avoid unnecessary searches as much as possible.

Post Reply