Why successive versions of Stockfish converge towards Lc0?

Discussion of anything and everything relating to chess playing software and machines.

Moderators: hgm, Rebel, chrisw

Vinvin
Posts: 5228
Joined: Thu Mar 09, 2006 9:40 am
Full name: Vincent Lejeune

Re: Why successive versions of Stockfish converge towards Lc0?

Post by Vinvin »

Laskos wrote: Thu Jun 18, 2020 9:41 pm
Raphexon wrote: Thu Jun 18, 2020 4:29 pm Because a lot of eval patches are written with (losses against) Leela in mind.
Hmmm, interesting, didn't know. How do they do it, a systematic analysis of errors made when playing Lc0? Doesn't seem easy. I think it's easier to do the opposite, to fit Lc0 to SF, either by training on SF games or implementing some tactical abilities of SF.
You can find a lot of lost games analyzed here : https://groups.google.com/forum/?fromgr ... ishcooking
SF as black loses to LC0 in CCC13 Heptagonal: Game 386: French: Wing Gambit
SF losses to Leelenstein in Games 51 and 52 of CCC14: Round 1 (15|5)
...