Three humans have a 60%+ match with Stocikfish

[Dann Corbit]

http://chess-db.com/public/enginefirstline.jsp

[Rebel]

Nice.

[JJJ]

And that doesn't say much. Carslen play 58% of his move like stockfish and is 2850 player , another player rated 2624 play the same %

Now let's see if they play like Houdini or Komodo instead. Would be fun to see Carslen playing more like Komodo.

[AdminX]

And that doesn't say much. Carslen play 58% of his move like stockfish and is 2850 player , another player rated 2624 play the same %

Now let's see if they play like Houdini or Komodo instead. Would be fun to see Carslen playing more like Komodo.

I think it says more about the engine that they analyze with and it's influence on their play.

[Dann Corbit]

I think this is the most logical explanation

[nimh]

I have data indicating that humans rated 1600-1700 have barely above 40%, whereas Houdini 1.5a and correspondence games from the ICCF W28 have around 75% coincidence with Komodo 8.

You've got to have a big strength difference and similar-type positions to get meaningful conclusions.

[Dan Cooper]

Stockfish must be the official trainer of China.

[bob]

Stockfish must be the official trainer of China.

I think it is simpler than that. I'd bet 10 years ago these numbers were HIGHER. As the computers were reaching real upper-GM levels. But as they get stronger, I'd expect the gap to be in a widening mode now as the humans simply can't stick with them tactically and the computers pull further and further away...