Decrease lmr depth if static eval decreases a lot

[Ergodice]

Decrease lmr depth if static eval decreases a lot

Passed STC
LLR: 2.93 (-2.94,2.94) <0.00,2.00>
Total: 60064 W: 15797 L: 15439 D: 28828
Ptnml(0-2): 236, 7080, 15106, 7310, 300
https://tests.stockfishchess.org/tests/view/67a2af9cfedef70e42ac3325

Passed LTC
LLR: 2.94 (-2.94,2.94) <0.50,2.50>
Total: 76794 W: 19740 L: 19337 D: 37717
Ptnml(0-2): 61, 8327, 21236, 8694, 79
https://tests.stockfishchess.org/tests/view/67a2c904fedef70e42ac374d

Passed Non-Regression VVLTC scaling check
LLR: 2.94 (-2.94,2.94) <-1.75,0.25>
Total: 29046 W: 7581 L: 7389 D: 14076
Ptnml(0-2): 2, 2557, 9213, 2749, 2
https://tests.stockfishchess.org/tests/view/67a54b591c4a3ea87241cb83

Bench: 2712545

[github-actions]

clang-format 18 needs to be run on this PR.
If you do not have clang-format installed, the maintainer will run it when merging.
For the exact version please see https://packages.ubuntu.com/noble/clang-format-18.

(execution 13164683912 / attempt 1)

[xu-shawn]

As with all depth-- patches, I think it is ideal to test this at VVLTC as well.

[Ergodice]

As with all depth-- patches, I think it is ideal to test this at VVLTC as well.

I disagree on this point since the tweak effectively just increases the reduction which scales positively in fact, but if it makes maintainers comfortable I can test vvltc.

[xu-shawn]

I understand that this patch may have some of its intricacies, but AFAIK we have not yet observed any depth-- tweaks that scales to VVLTC. So I think we should be careful here and make sure it doesn't introduce any scaling weaknesses.

[Ergodice]

added result of scaling check