Exact, Exacting: Who is the Most Accurate World Champion?

I meant, who are the most accurate.

I don't see to which depth or with which time the moves are analyzed. Is it exactly the same analysis we get for our games here, on lichess? Imo it should be longer in order to do justice to the more aggressive players.
Another aspect that would have to be dealt with: Ideally every analysis would start in the position that was the first "new" move. That would be a lot of effort, of course, but I don't see how the results can be fair without it. Or has this already been taken care of and I am missing something?

Very interesting. It confirms our assumptions about how the quality of the game has changed with the advent of computers.
However I think they should reduce clock times to recover human error and with it entertainment.
Sorry for my grammar; I use an imprecise translator. :)

Just FYI, the 1957 Botvinnik-Smyslov game should not have made the bottom 5 list, there is an error in the PGN:
lichess.org/study/MPkpQJkr/lS2KHaPC#63
32...Qxa7+ was the actual move, the d6 pawn remained to make Qd4+ impossible, not giving white a free rook multiple times.

I don't see to which depth or with which time the moves are analyzed. Is it exactly the same analysis we get for our games here, on lichess? Imo it should be longer in order to do justice to the more aggressive players. but Very interesting. It confirms our assumptions about how the quality of the game has changed with the advent of computers. I meant, who are the most accurate. Absolutely love this article. Very interesting and nice data visualisation!Thanks for your amazing analysis. Do you think that shorter time controls should at least try to spice things up? Maybe if both players had 90 minutes with no increment they would be forced to play more decisively? Perhaps this is better than this entire ultra-conservative match going to the tiebreak phase?As many people have touched upon, I feel that the ACPL is an imperfect measure of how "well" players have played, as opposed to accurate. It favors games where nothing much happened over very sharp ones because it would be much more of a feat to replicate such a super low ACPL in a Najdorf, where players are at some point on their own, as opposed to a position where both sides' plans are well trodden out. Also, a computer might regard it suboptimal to transform into a winning endgame, rather than continue sacrificing and mate in 13.A stronger engine is needed to show all their mistakes. Is it possible that the top players have memorized all the best moves by the latest engine? A stronger engine is needed to show all their mistakes. Is it possible that the top players have memorized all the best moves by the latest engine?

Oh yes

At what depht were the games analyzed?

@MFXX said in #55:
> I think you forgot to add the link

I think lichess silently filtered it for everything but me. Make your way through my bio on the profile :).

This was really interesting. I ignore a lot of articles that are similar to this in theme, but this was good. I don't know if I understand all of the conclusions. I don't know if they clearly defined terms like "science" and "art", but I like the data. Also I'm not sure if the number of moves per game being longer and short draws were factored in, in a way that I understand. But great job @LiChess bloggers!