Friday, April 27, 2012
What do the tennis stats really mean?
I've been messing around with tennis statistics lately. The sport has come a very long way in a relatively short time in that area, given that it's a global game with a baffling amount of data (with up to 128 players to track at even a single Grand Slam) -- along with a history of indifference to the art of statistics-based record-keeping.
Nobody would ever confuse tennis with baseball, the sport in which statistics have become an almost romantic dimension of the game. But the volume of statistics now generated by the ATP -- and, to a lesser degree, the WTA -- on a daily basis is impressive.
It also tends to support the theory that tennis statistics are in some ways inadequate when it comes to explaining the ebb and flow at the top of the game. That's because almost all good players are packed together so tightly that you invariably return to the idea that critical matches are determined by a handful of points (even that may be stretching it), which are won or lost for reasons that may have very little to do with anything you can quantify with stats.
Case in point: The ATP tracks 10 critical categories in its Ricoh Matchfacts module. It tracks them for retired as well as active players, in both "career" and "by year" categories. It will even break them down by surface.
This is a formidable bit of data mining, and it does turn up some mind-blowing results: Who knew that Guillermo Coria is the career leader in break-points won percentage (46 percent)? Or that the highest career first-serve conversion rate belongs to Gilbert Schaller (76 percent)?
But the problem arises when you see how close everyone else is to the leaders in these departments, and how many men end up tied with as many as a dozen others not very far behind the leaders. To wit: Eight different men, including Novak Djokovic and Rafael Nadal, are a mere percentage point behind Coria in the break points won category. Among them as well: Filippo Volandri. Say what?
In the 10 categories tracked by the ATP, the only one in which Djokovic places above Nadal is "Aces." You can't look at career stats for that category, because the two present-day stars are nowhere near done. (The top ace-maker, by the way, was Goran Ivanisevic, with 10,183.) In the year-to-date rankings, Djokovic is way ahead of Nadal, 140 aces to 82. But Djokovic is just No. 18 in the rankings, and Nadal is No. 41.
Djokovic and Nadal are among the very best who ever played (since the stats have been kept, anyway) in every percentage-based category. But Nadal is consistently ranked above Djokovic, albeit by more than one ranking spot in just three of the nine categories.
The rivals are ranked equally in two categories: Each man has converted break points at a 45 percent clip (along with six other players), and their first-serve points won stats are identical (72 percent). Interestingly, though, that makes them 11th-best on the list, with fully 87 players ahead of them.
If you take the long view, the stats bear out what most fair-minded pundits believe, and all those other career numbers (including Grand Slam events won) bear out. Nadal has been just that much better, on a consistent basis, in almost all departments. There's a reason he has so many Grand Slam titles, right? Even the amazing numbers Djokovic put up in his enchanted 2011 haven't quite turned the tide.
The most useful stat, in terms of showing a significant edge one way or the other? Nadal's first-serve conversion percentage of 69 percent puts him No. 4 on the all-time list; Djokovic is No. 9, with a 64 percent rate.
If you saw how many points Nadal won thanks to his first serve in last week's Monte Carlo final, you might even think it's worth wading through all those stats to come up with that telling nugget.