In Sunday’s Miami Open final, Jannik Sinner posted some very odd stats in his straight-set loss to Hubert Hurkacz. He won a respectable 48.4% of his second serve points–three points behind the the ATP top 50’s average since the restart–but only 55.3% of his first serve points. First serve points are the bread and butter of the offensive game, and Sinner got only as much out of that as Casper Ruud derives from his second serve.
It’s not that Hurkacz has magical anti-first-serve powers, either–it was only the second time since the restart that he won more than 40% of first-serve return points. He barely won 30% against Denis Kudla in the Miami first round.
Sinner’s first serve is not typically so ineffectual, but that isn’t to say it’s particularly good. While the tour wins well over 70% of its first-serve points, the Italian won only 63.5% in the quarter-finals against Alexander Bublik (another man who would never be confused with Andre Agassi), and 64.6% in his semi-final match with Roberto Bautista Agut. In both of those contests, he held on to 57% of his second serves–an outstanding mark for a player’s weaker offering.
I keep returning to second-serve points won and the difference between firsts and seconds to emphasize that Sinner is doing a lot of things right. Winning so many second-serve points suggests that he understands the tactics that go into playing service points when the ball comes back. Yet he hardly reaps any extra benefits from landing his first serve.
If it looks like a clay-court specialist…
I calculated the difference between first-serve and second-serve points won–let’s call it WinDiff–for every men’s tour-level player-season between 2000 and 2021 with at least 15 completed matches. That gives us almost 2,500 data points. At one extreme is 2019 Sam Querrey, who won over 80% of his first serve points but only 47% of his seconds, for a WinDiff of 33.2 percentage points. At the other end is Juan Carlos Ferrero 2011 campaign, when he won 65.5% of his firsts and 57.0% of his seconds, for an 8.5-point WinDiff.
Querrey’s season was mediocre and Ferrero’s was pretty good, suggesting that there’s a sweet spot somewhere in the middle. Yet it’s possible to have outstanding seasons near either end of the spectrum. Last year, Alexander Zverev’s WinDiff was 32 (77%/45%), and he won 28 of 39 matches. And in 2008, a young Rafael Nadal had a WinDiff of 11.5 (71.9%/60.4%), which was good enough to give him his first year-end #1 ranking.
Finding Ferrero and Nadal in Sinner’s neighborhood starts to give us an idea of what kind of players have low WinDiffs. Out of 2,460 player-seasons, only 33 had smaller WinDiff’s than Sinner’s 11.4 percentage points for 2021 so far, and you might be able to identify a few common traits among the players who posted narrower gaps:
Yoshihito Nishioka, Juan Monaco, Filippo Volandri, Potito Starace, Flavio Cipolla, Albert Montanes, Pablo Cuevas, Diego Schwartzman, Damir Dzumhur…
Clay courters, short guys, Italians… you get the idea. Ferrero and Nadal offer profitable career paths for Sinner, but I’m not sure that “be like Rafa” is practical advice for anyone, no matter how talented.
Room for improvement?
When I mentioned the extreme stat from Sunday’s final in my Expected Points podcast yesterday, I concluded on an optimistic note. Sinner is 19, he just reached a Masters final, he’ll continue to work on his first serve, and as we’ve seen, the rest of his game is already top-notch.
But does the data bear out such a rosy outlook? Are there players who have emerged from the purgatory of a low WinDiff to get more out of their first serves?
The average WinDiff of the top 50 since the restart is about 21.5 points. The bad news is that only a few low-WinDiff players eventually reach that level. The good news is that a disproportionately weak first serve is apparently correctable, or–at least–the stat is noisy enough that some players regress toward the mean.
Going back to my set of 2,460 player-seasons, the 5th percentile was a WinDiff of about 14 points. 71 different players had at least one season below that threshold, and 20 of those guys have played at least 10 full seasons since 2000. Drawing the line at a decade’s worth of play is some serious selection bias, but if Sinner doesn’t stick around for another 7.5 seasons of 15 or more matches, that’s probably a sign of something else gone very wrong.
The following table shows those 20 players. For each, I’ve shown their highest single-season WinDiff since 2000 and the average across their 21st-century career. Remember that tour average is a bit above 20 points, and Sinner’s 2021 so far sits at 11.4.
Player Seasons MaxWD AvgWD
Juan Martin Del Potro 10 24.8 21.2
Pablo Cuevas 11 22.5 20.0
Fernando Verdasco 17 23.0 20.0
Albert Montanes 15 24.1 19.5
Philipp Kohlschreiber 16 23.9 18.2
Juan Ignacio Chela 13 22.1 17.7
Tommy Robredo 15 21.1 17.6
Jarkko Nieminen 14 21.4 17.4
David Nalbandian 12 22.2 17.1
Fabrice Santoro 10 21.0 17.1
Nikolay Davydenko 14 20.3 17.0
Dudi Sela 10 21.9 16.4
Albert Ramos 10 19.6 16.3
David Ferrer 17 19.2 16.2
Juan Carlos Ferrero 13 18.3 15.1
Mikhail Kukushkin 10 18.0 14.9
Rafael Nadal 18 17.5 14.8
Olivier Rochus 12 17.8 14.7
Filippo Volandri 10 16.2 13.2
Juan Monaco 13 15.9 12.7
Juan Martin Del Potro offers the brightest path, if one can emulate the results without the injuries. His WinDiff as as 17-year-old tour newbie was roughly 13 points, but he quickly landed near tour average. Sinner isn’t nearly as tall, so a better comparison might be David Nalbandian, who didn’t win nearly as many first-serve points as his fellow Argentine, but held on to enough. It’s certainly easier to look at Sinner and imagine a Nalbandian-like future than it is a Monaco- or Volandri-like one.
Another reason for optimism is that Sinner himself has already posted a 21-point WinDiff season last year, and this year’s weirdness is in large part due to his improvement against second serves, not a drastic drop in first-serve effectiveness. Maintaining his 56.5% rate of winning second-serve points seems unlikely only because it is so good. If he can manage that, he can survive with a modest first delivery so long as it’s in his typical high-60s range instead of the mid-50s that proved his undoing against Hurkacz.
Finally, the clay-centricity of the list above might be reason to pause before pegging Sinner as an eventual #1. But it also suggests that the teenager is developing exactly the right kind of game to excel on dirt. For the next couple of months, Italian fans will have plenty to get excited about.
