Comments on https://lichess.org/@/jk_182/blog/how-the-evaluation-and-clock-impact-results-of-blitz-games/I2kRp2sk
Wow, thanks for this analysis! So this (set of graphs which form a 3D topology) is what https://lichess.org/tutor compares my performance against, to find in what respects I am overrated or underrated!
Neat analysis, thanks for sharing! I've always been a bit dubious of the Lichess win percentage model (reporting a parameter estimate like "-0.00368208" to six significant figures doesn't help), but these figures show that some sort of logistic curve generally is an excellent approximation in a wide variety of contexts.
Such a neat analysis!!
very cool -- thanks for sharing!
This diagram makes no sense. Game score is binary, so how are there graduations? Do you mean average outcome? Also if black=no data the axis is labelled wrong in the first place