• 1 Post
  • 28 Comments
Joined 1 year ago
cake
Cake day: October 25th, 2023

help-circle





  • They’re a better indicator of how good someone is versus their ATP ranking. Rank is solely a function of performance in tournaments irrespective of the quality of opponents you play and how well you perform against them. Thus with rank you get people like Ruud who coast to slam finals without playing anyone and end up #2 in the world whereas their Elo is still in the teens and more predictive of how they’d fare head-to-head with another like player








  • What’s your context for saying a 3 percentage-point increase is “a lot?” Djokovic from his first full season (2005) to his second (2006) jumped from 50% win rate to 69%. Then the next year (2007) up to 78%. That’s a lot of change.

    A 3 percentage-point change is probably less than the standard deviation from year to year for a player who has effectively plateaued. Since 2011 when Djokovic broke through, the standard deviation of his win rate through 2023 is 4.35. That means his win rate just “randomly” changes by over 4 percentage points every year. Rafa’s standard deviation since 2005 to 2022 is over 5.

    Your win rate changing 3 percentage points doesn’t mean jack all.