New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Very bad Chess960 performance #203
Comments
|
Making the psq symmetric in chess960 by using (psq_left + psq_right) / 2 makes it even worse. |
|
There is propably something wrong with the (castle-) move handling or maybe incremental psq evaluation. Look at this log and the score in the first depths: |
|
Probably this one bad evaluation value for queen with no mobility VALUE(-4097, 83) which is the result of some fake tuning and has no relevance for normal chess but it has for FRC. |
|
New gauntlet with latest FRC evaluation fix: |
After implementing Chess960 tests show that Rubi plays extremely bad in this variant. Probably evaluation too much hardcoded for the normal chess setup. Improving Chess960 performance with more flexible evaluation may also help in normal chess.
Gauntlet normal chess:
Gauntlet Chess960:
The text was updated successfully, but these errors were encountered: