MultiCut pruning

Programming Topics (Computer Chess) and technical aspects as test techniques, book building, program tuning etc

Moderator: Andres Valverde

Re: MultiCut pruning

Postby diepeveen » 28 Jun 2005, 10:55

Tried that multicut in Diep.
gives 0.5 ply deeper when it gets around 10-12 ply.

Problem is it reduces at cut nodes. IMHO nullmove is superior there and already reduces plydepth a lot by reduction factor.

You can get 1 ply deeper by using R=4 instead of R=3 with nullmove too, less risks than multicut.

Figured out you REALLY need to follow the definitions of Yngvi real well.

Multicut is by far the best algorithmic find of the past few years IMHO as it is a new method to cutoff.

however, reducing at all nodes somehow is more interesting than cutnodes. nothing beats nullmove simply and when used in combination with nullmove you do a kind of supernullmove.

Of course when you sacrafice a piece in order to get it deeper back in the tree, multicut in combination with nullmove is going to fail, to give a trivial example.

Nullmove with bigger R already isn't going to fail there.

p.s. hoped i replied in correct way, does not show up as a 'reply' to that multicut thread.
diepeveen
 
Posts: 116
Joined: 28 Jun 2005, 01:09
Location: Netherlands

Re: MultiCut pruning

Postby Volker Pittlik » 28 Jun 2005, 11:05

diepeveen wrote:...
p.s. hoped i replied in correct way, does not show up as a 'reply' to that multicut thread.


It does, only on a new page. If you would like to quote something press the Image button in the upper right.

Volker
User avatar
Volker Pittlik
 
Posts: 1031
Joined: 24 Sep 2004, 10:14
Location: Murten / Morat, Switzerland

Previous

Return to Programming and Technical Discussions

Who is online

Users browsing this forum: No registered users and 34 guests