TalkChess.com

Bug was interesting. It distroyed time management badly, but in test positions you can't see anything wrong! It was MC formatting error fixed 6.4.

Interesting test! Middlegame phase of the game, roughly moves 15-22. So after 23 it's endgame?

Here https://nextchessmove.com/dev-builds.

SF 5.4.2019 was buggy and -100 ELO to SF10!

In Komodo team Larry is 70+ and Mark 60+ I think . But probably same as here: It is a general truism in mathematics that mathematicians do their best work in their 20s and 30s.

Has Komodo really max_depth higher than 99/128? For solving 82 mate you need 164 plies. Normal SF and Houdini have 128 limit AFAIK.

OUCH: Testrun of Stockfish 190331 finished - bad regression: -9 Elo to Stockfish 190310! But now your list has +21 to SF10. That's more realistic, when NCM and framework tests (20000 and 40000 games) have both stalled to +16 ELO month or two ago.

Houdini 6 gets fast mate in 17 in normal mode, but is it correct?! Analysis by Houdini 6 x64-pext: 1.Kd2 Rh2 2.Be4 Kb8 3.Ke2 Kc8 4.Kf2 Kd8 5.Kg1 Rxg2+ 6.Kxg2 Ke8 7.Rg1 Kf8 8.Rh3 Kg7 9.Rxh4 gxh4 10.Kh3 Kg8 11.Kxh4 Kg7 12.g5 Kg8 13.gxf6+ Kf8 14.Kg5 Ke8 15.Rh1 exf6+ 16.Kxf6 Kd8 17.Rh8# +- (#17) Depth: ...

Correct 0.5.2 is again 80 ELO better .

And remember, that Stockfish dev is 1167872 bytes! How much is current Lc0? 100 megabytes + cuda things? SF should get 1/100 time handicap .