It is a bit risky to naively "prune as much as possible" though? If my understanding is correct, futility pruning does not necessarily return the theoretically best move - i.e. the margin size is kind of a trade off between accuracy and speed.
This means the choice of margin size is to be optimised. Is my understanding correct?
Don't want to make a huge fuss about it for now, given that it is not turned on. I think we should have the functionality in first, and we can do testing later. Most important thing is that it is not breaking other things, and is doing what we think it should be doing.
This means the choice of margin size is to be optimised. Is my understanding correct?
Don't want to make a huge fuss about it for now, given that it is not turned on. I think we should have the functionality in first, and we can do testing later. Most important thing is that it is not breaking other things, and is doing what we think it should be doing.
_Originally posted by @ccjeremylo in https://github.com/KYLChiu/sporkfish/pull/96#discussion_r1477135886_