Stockfish's blunders in the SSDF games

[syzygy]

Garry Kasparov still plays much better chess both strategically and positionally.

then why is he worse at the game than any superhuman engine

Because Uri (!= Uri Blass) posted this from the year 1997.

[Modern Times]

The GUI can have an impact for sure. There are 606 games on the CCRL 40/15 list for Stockfish 20230613 64-bit 4CPU. All of those were played under CuteChess GUI. Stockfish only has one loss in all those 606 games, and it was not a blunder due to time control issues. It was a loss as black to Komodo Dragon where Stockfish could not neutralise white's opening advantage. Are there any time control related blunders that caused SF to draw rather than win ? That I don't know. I think that is unlikely.

Probably setting the move overhead to a slightly higher number than the default mitigates any potential issue. I think I have mine set to 50 or 100 instead of 10.

[Eelco de Groot]

For CuteChess games you can have a lower Move Overhead probably than for instance ChessBase which I suspect SSDF may still use, I think in the past they had a lot of ChessBase stuff? For TCEC Stockfish move overhead used to be 1000 ms, when the Stockfish team forgot to ask or sent a version with 10 ms, I don't remember, Stockfish lost on time. It is a couple of years ago, they have very different hardware now and conditions will not be the same. When we had this discussion on ProDeo, someone only blamed the inferior GUI in TCEC, not Stockfish's fault. Can't be 10 ms is a bit ridiculous, you might as well set it to zero then for SSDF if it were 10 ms in 40 moves per 2 hours games. But what do I know. Stockfish losing some games is no big deal but the conditions should not favour other engines when they receive points from this, but not others.