TalkChess.com

Some people are writing about Multi-PV and I think it is worth a try. I proposed a method back in 2016 which is based in Multi-PV move-picking involving PRNG. I am not aware that anyone has implemented it so far... That's what I do in my engine. I described the technique on my website. See The MadC...

I have not been able to verify any strength gain from aspiration windows. I have decided to keep aspiration windows in my engine, though. They are useful for two non-competitive modes: MultiPV UCI_LimitStrength Why? Because unlike in competitive mode, I don't reduce the search horizon of root moves ...

I tried something similar to what Joost suggested. If it fails high reset the bound to inf and continue with the next iteration. But if it fails low expand at low side by 50, 100, inf, then research. Also start applying aspwin above iteration 3. That's encouraging. My test of skipping to next ply o...

The algo is simple, if there is an early sign of score instability reset the bounds to its original value as early as possible... if the score is already losing or winning... Thanks Ferdy for your explanation and testing. If I understand you correctly, that's the algorithm I've tried (except with a...

I just run some experimentations that strongly disagree with your findings... Are you sure your implementation is correct and optimal? Interesting. Thank you Richard for running these tests. I always am willing to admit the possibility I screwed up something in my code. I've never had any luck with...

Otherwise I feel I'm subverting the correctness of the alpha / beta algorithm. With pruning and reductions (especially LMR), that has gone out of the window long since anyway. Is that a bad pun? Window, ha ha. Well, there never was correctness anyway because that assumes perfect static eval which i...

One alternative could be to store the next best score from the previous iteration (and possibly the next best move so we can search it first) along with alpha and beta. In Principal Variation Search, we don't know the next best move. We only know it if MultiPV > 1. At least in MadChess. I do PVS ev...

Aspiration windows are tricky, and a bit of a kludgy device. The assumption is that disaster never happens. I agree with you, H.G. The technique seems kludgy to me. I never liked it. I added it years ago when I was new to chess programming, less disciplined about testing, and under the assumption i...

Dann Corbit wrote: ↑

Sun Dec 20, 2020 9:05 pm

Do you mean that your search window is +/- infinity for every search?

Yes, +/- infinity for every search (with PVS) is stronger than using aspiration windows. Presumably because the researches caused by failing high or low are too costly.

I should add MadChess uses PVS.