Comment by NitpickLawyer

Comment by NitpickLawyer 15 days ago

> Oh boy, this absolutely does not work for chess at high levels.

Magnus himself said this. If he were to cheat, he'd only get 1-2 moves per game, and sometimes not even the moves explicitly, but merely the notion that "there is a very good / critical move in this position". That would be statistically impossible to accurately detect.

stevage 15 days ago

Well, statistics would be the only mechanism. If a player was on average playing at level X in one setting, but at a lower level Y in a setting where it was considered impossible to cheat, that's about as good as you can do.

But it's pretty impossible to point to a single move and say "that's definitely a cheat move".

Reply View 3 replies

datadrivenangel 14 days ago

You can look at moves as a series of probabilities. For each move, classify if it's more a blunder or inspired move and then look at people's games and see if they consistently have 1-2 moves that are much much better than their typical.

Reply View | 2 replies
- NitpickLawyer 14 days ago
  
  The problem is that at that level they're more likely to make the absolute perfect move than not. Super GMs often play 95-98%+ accuracy games.
  
  Reply View | 0 replies
- stevage 14 days ago
  
  But what if they only cheat occasionally? A top player would only need a handful of moves to go from say 3rd in a tournament to winning.
  
  Reply View | 0 replies