Static null move pruning

[Necromancer]

I'm trying to learn more about pruning methods to improve my engine.
Looking at CPW code (https://github.com/nescitus/cpw-engine/ ... search.cpp), I came across static null move pruning (line 277).

I have some questions:

1- Is this a standard pruning method, ie, does most engine use it?
2 - In line 289 they return static_eval - eval_margin. Can I just return beta? I'm not using fail-soft I guess.
3 - What can go wrong with this technique?

Thanks!

[ZirconiumX]

I'm trying to learn more about pruning methods to improve my engine.
Looking at CPW code (https://github.com/nescitus/cpw-engine/ ... search.cpp), I came across static null move pruning (line 277).

I have some questions:

1- Is this a standard pruning method, ie, does most engine use it?

Well, I call it "reverse futility pruning", because it's essentially futility pruning inverted. It's reasonably common, since it's cheaper than doing a full nullmove search at the leaves (or that's how it seems to be for me, anyway).

2 - In line 289 they return static_eval - eval_margin. Can I just return beta? I'm not using fail-soft I guess.

For fail-hard, returning beta is fine.

3 - What can go wrong with this technique?

I would say setting a suitable margin is a fine art that needs to be mastered. It should hopefully also go without saying that it shouldn't be used in check, too.

Thanks!