Hacker News new | ask | show | jobs
by lupire 1359 days ago
The sample size for that is absurdly small. (Carlsen only takes single classical losses seriously, since rapid/blitz is too random.)