All I see is that Roger tend to lose 5-setters. Unlike Novak and Rafa that are so pugnacious that they rarely lose a 5-setter.
No, because you´re not seeing the full universe of decisive-setters in this stat. Maybe Roger has a 90% of win rate on decisive-setters, but you cannot infer that number from the metric I presented. It’s just how hard the player fights on losses.
Yes but they are not completely unrelated. For example if a player chokes horribly in the decider of a match he should have won, that will make his score in your table go up, not down.
Sure, but what is a “match he should have won”? Of course it would be nice to see this metric disaggregated by opposite rank for example, or by win/lose first set. I think it just gives you a general overview of the picture.
Simply a match where that day he was the much better player and should have won, and he lost. That certainly happens in tennis, and fairly often at that. And that’s a solid counter-example to show your metric doesn’t work as intended at all, since that’s at least 1 common case that makes the player looks good in your metric while it’s not good at all for the player. if you ever did some high-school-level maths, you know a simple counter-example is sufficient to prove something doesn’t work as intended, and this is a pretty big one.
Are trying to say chokes don’t exist ?
Genuine question: how is this stat different from the overall win percentage? Doesn’t it give you an overview what are the opponent’s chances of winning against the high-ranked players?
This metric says the percentage of the total lost matches that the player has reached the decisive set. It gives you a very general overview of how difficult to beat the player is. Of course you then have to weigh for the overall win percentage, because if a player has 60% of decisive sets on lost matches but a 10% of overall win rate, it’s not a very difficult to beat player after all. It’s just one metric, you have to consider others also for a deeper analysis.