Search found 3 matches

by JollyJoker
Thu Mar 28, 2019 3:31 pm
Forum: Computer Chess Club: General Topics
Topic: New Cerebellum Library
Replies: 27
Views: 3508

Re: New Cerebellum Library

Are these evaluated using Stockfish? It would be interesting to see a comparison with a Leela version. Even more interesting if the AplhaZero team made a book with a trillion positions or so :D
by JollyJoker
Sat Mar 16, 2019 9:52 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Training the trainer: how is it done for Stockfish?
Replies: 22
Views: 5335

Re: Training the trainer: how is it done for Stockfish?

Comparing n-man to n+1 man, does n+1 man have a higher share of moves that don't worsen the outcome? Assuming we're considering that as the definition of a mistake, rather than having an opponent that can actually exploit the mistake.
by JollyJoker
Fri Mar 15, 2019 2:29 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Training the trainer: how is it done for Stockfish?
Replies: 22
Views: 5335

Re: Training the trainer: how is it done for Stockfish?

A simple approach would be to use LC0 for the first part of the game until sufficient material has gone from the board that SF is able to basically see to the end and then switch to SF. Certainly in the endgame it seems to make no sense to rely on LC0 if optimal match play is the goal. (But in the ...