TalkChess.com

The name of your program makes me think of the British Broadcasting Company (BBC). Exactly, it's intended for youtube purposes - if peoaple search for BBC (British Broadcasting Company) they would probably find my engine as well. May be one day I will monetize my work via ads. A certain irony in th...

Yes! And I have raised the same concerns here also. 1 tier testing at bullet time controls, and only testing with 1 core with Stockfish NNUE. Will give you a over inflated impression of Stockfish NNUE's true strength. And the reason is simple, Stockfish NNUE does not scale like other engines. With ...

According to Ordo we now have 8CPU as +48 Elo to 1CPU. The ratings list will not show that because it uses bayeselo - that I think will show it as +30 Elo. However 1CPU has currently not played the GPU engines that 8CPU has. That is planned for the coming week, and I suspect that may depress the 1CP...

If SF 8-core has 'underperformed' in terms of Elo because it had too many NN opponents, this can be considered evidence for the above hypothesis. There will be a wider range of opponents in the update later today. It was probably unwise to publish just the games we did, but that is how we work, we ...

Here is a download from my site containing current cli and gui from current alweys test branch (EXPER1). https://rwbc-chess.de/Downloads/cutechess-win64.7z This is an overview about already incorporated changes in this branch: https://github.com/alwey/cutechess/commits/EXPER1/projects/gui Note that...

If someone wants the current one with lots of new features/user settings I can upload it, but I have only compiled WIN binaries. I would like this, thanks. Is it just a matter of overwriting the existing cutechess.exe or is there more involved than that ? I currently have x64, Version 1.1.0 (noted ...

We will put notes against those two highlighting the controversy surrounding them, as we did with Rybka. There wasn't time this week.

It is certainly possible, I used to run Gaviota many years ago in HG's tournaments from xboard on a Linux box. But Miguel gave me the scripts etc so I don't have any knowledge of the technicalities.

In a quick FRC bullet test, Ethereal 12.50 performed 34 elo better than Ethereal 12.25 against SlowChess 2.3, so FRC play looks clearly improved for this version (though my error bars are high, I only did the 1920 games for 12.50) We're showing more than +50 Elo so far at FRC compared to 12.00 whic...

Thanks, so it seems RTX2060 is a good match for 1 fast thread. Still, it's close, RTX2080 is not even 2 times faster. CCRL seems to show 50 Elo advantage of SF 12 over Lc0 1541 2080, which I guess are either $$$ or an artifact of the rating list with many much weaker opponents. SF12 has not actuall...