At what time control are humans best vs. computers?

Sort:
BigSplat2018

We all know the best computers can beat the best humans. It's been that way for 20 years.
But where is the difference smallest? Do the electronic geniuses shine most at bullet, blitz, rapid, or traditional?

I seem to remember that computers were kicking human butt at Blitz before they were beating us at other time controls, but that memory might be faulty. Or things may have changed.

I'd like to see that tested. Get Magnus or Fabiano or Hikaru vs. AlphaZero, with time odds--is the carbon-based life form most competitive at 3 vs. 5? Or 30 mins vs. 50? Etc.

demneptune

Well, interesting question! I do hope you get some better answers than mine. Traditionally, the shorter the time control, the greater the computer's skills, so the first loss by a grandmaster to a computer was, I believe, Michael Stean (in 1977) losing a 5 minute game. He had thrashed the machine positionally, and thought it was time to open up the tactics - and then the computer came into its own. I'm fairly sure Kasparov lost to Fritz occasionally before his loss to Deep Blue, at 5 minute games, but I don't know the deets.

kindaspongey

http://www.chessgames.com/perl/chessplayer?pid=15423

http://www.chessgames.com/perl/chessgame?gid=1472109

http://www.chessgames.com/perl/chessgame?gid=1472110