Elo gain from futility pruning in others engines?

[algerbrex]

As some of you might know, I'm currently working on the release of Blunder 8.0.0, and part of that process involves more rigorous testing to optimize my implementation of various features.

The problem, of course, is that it's hard to know when I've maximized a feature's potential and should move on to the next feature. (I also recognize not everything will be perfect and improvements can and should be made later down the line.)

So for the current feature, I'm on, futility pruning, I'd like to ask others the biggest amount of Elo they've gained from it, so I can get a better idea of what I should be aiming for, optimally.

And hopefully, there won't be too many posts like this in the future.

[emadsen]

So for the current feature, I'm on, futility pruning, I'd like to ask others the biggest amount of Elo they've gained from it, so I can get a better idea of what I should be aiming for, optimally.

Way back when, I got 54 Elo from futility pruning.

[Brunetti]

Disabling it in my engine the loss is around 110 points (tested only wit fast tc 4"+0.1).

Alex

[algerbrex]

Disabling it in my engine the loss is around 110 points (tested only wit fast tc 4"+0.1).

Alex

Thanks Alex. I think I ended up getting around 55 Elo from it for me, so I may look into trying to optimize it more.

[algerbrex]

So for the current feature, I'm on, futility pruning, I'd like to ask others the biggest amount of Elo they've gained from it, so I can get a better idea of what I should be aiming for, optimally.

Way back when, I got 54 Elo from futility pruning.

Cool, that actually matches about what I got. 37 Elo from conservative futility pruning, and then another 19 Elo from tuning the reductions and depth parameters.