Flavio Cobolli’s Upside-Down Season

This has been a breakthrough campaign for Flavio Cobolli. He reached a major final in Paris and won enough matches throughout the year to reach #6 in the ATP rankings. He’s at the Tour Finals in Turin this week, a surprising result for someone who started the year outside the top 20.

Yet he has set a remarkable record in the process. Cobolli will almost certainly be the first top-tenner in the ATP matchstats era (since 1991) to win fewer than 50% of total points.

You probably know that it’s possible to win a match with fewer than 50% of points, or even games. Imagine a 7-6, 1-6, 7-6 contest, and you get the idea. Such a roller-coaster scoreline isn’t necessary, as long as the winner loses some games quickly but is dragged into deuces in the games they win. En route to the Acapulco title back in February, Cobolli won just 48% of points to beat Miomir Kecmanovic, 7-6, 3-6, 6-4.

But a season is a different story. Cobolli’s won-loss record entering the Tour Finals, 36-24, is weak for a top-eight guy. But it’s still a 60% win rate. It takes more than a couple of flukes like the Acapulco semi to get to a 60% match win rate with only 49.8% of points.

Here are the lowest total-points-won (TPW) rates for year-end top-tenners:

The Italian is in a class by himself. If he can win a match or two and hold off Arthur Fils to finish the year at #6, he’ll be the first top-sixer to go the full year with a TPW% under 51%, since Kafelnikov a quarter-century ago.

(Technically, Cobolli could also make this a moot point. If he wins his last two round-robin matches and ends up champion, he’d probably play around 600 more points. A 53.5% win rate for those four victories would push him up to 50.0% on the year. The odds are certainly against him, but he did win 53.8% against Tiafoe in Acapulco, and he drubbed Zverev to the tune of 57.4% in Munich. He also won 37.4% against Tiafoe at Indian Wells, so… yeah.)

Another way to put Cobolli’s season in context is to find the best years (by ranking) of players who finished with more points lost than won:

The 50.0% rates are rounded up from something just below that.

Cobolli’s #22 finish last year almost made the list. If you exclude Davis Cup, he ended 2025 just a hair below 50%. Fittingly, he polished off his season with a 1-6, 7-6, 7-5 victory over Jaume Munar, in which he won only 48% of points. But he won 58% and 52% in the previous two rounds–barely enough to end the season in the (point) win column.

How?

None of this answers the question of how the Italian got to a 60% win rate and a #6 (or #7, depending on the next few days) year-end ranking.

As for the ranking, that’s a combination of clustering wins–a full third of Cobolli’s ranking points come from the Roland Garros final–and a jam-packed field without many clear elites. In other words, the same factors that have allowed Ben Shelton to crack the top five with his own statistical quirk.

The 60% win rate comes from pulling out the close ones and losing badly more often than he wins easily. A simple way to see it is that in his wins, Cobolli earned an average of 52.8% of points. In his losses, he tallied just 44.1%. The margin in his losses was more than twice the margin in his wins. There were more victories, but not enough to pull the weighted average up over 50%.

A striking illustration is how Cobolli fared when he won between 48% and 50% of points:

Five wins out of seven! A typical player might pick up one or two wins out of a group like that, but five is rare. And two of them–the Wimbledon third round and Acapulco semi-final–were disproportionately valuable in terms of ranking points.

Of his 36 wins, only nine were in the “comfortable” range, with at least 55% of points won:

But nearly three-quarters of his losses were comfortable for the other guy:

This probably isn’t sustainable. Certainly the ranking isn’t, unless Cobolli sneaks into another French Open final–or just gets a lot better. Some of this might be a real tendency, where the Italian just doesn’t feel like it on some days, so instead of losing a close one, he flops out entirely. But at the very least, he isn’t going to keep pulling out so many matches in the 48-50% range.

The funniest outcome in 2027 is also the likeliest: Cobolli will finally raise his TPW a comfortable margin above 50%, but he’ll fall back out of the top ten. He’s due for a year of boring normalcy.

Ben Shelton’s Pliable Ceiling

Ben Shelton is ranked fourth in the world. In the (unlikely) event that Carlos Alcaraz founders in Turin and Shelton takes the Finals crown this week, he could even finish in the top three. He began the year ranked ninth. Has he gotten that much better?

A few years ago I asked, What Is Ben Shelton’s Ceiling? A big serve can get you a long way in this sport. But without a return game to match, the margins are perilously thin. In 2023, Shelton won 32.7% of his return points. Only one player has ever posted a number so low and finished in the top ten: John Isner in 2018. Last year, Shelton won 33.3% of his return points. That stands as the second-lowest RPW% for a top-tenner, at least since 1991, the span for which those stats are available.

The rule of thumb is that to be a top-five player, you need to win 35% of return points. To be top three, 37% is usually necessary, at least if your name isn’t Sampras or Roddick.

Well, Shelton is going to finish top five. Going into Turin, his RPW% was 32.9%.

How?

This isn’t supposed to be possible. Either the game has changed, or Shelton has done something unusually well, or clutch, or smart. Or lucky.

ATP tennis may have changed a bit. This year’s top 50 won 36.4% of return points, compared to 37.1% last year. Faster surfaces, if that’s what it is, might be suppressing Shelton’s RPW% by half a percentage point or so. That’s something, but it hardly explains everything.

At the individual level: Lucky is my default first guess–no reflection on Ben, it’s just that luck is the most common explanation for stuff like this. How about tiebreaks? He has won 61% of his breakers this year, better than his career average of 54%, if not overwhelmingly so. Maybe he “should” have lost one or two more. But ~60% is sustainable for a savvy big server: Milos Raonic’s career tiebreak win rate was 61%, and he thrice topped 65% for a season.

What about deciding-set tiebreaks? Shelton played a whopping eleven of them this year, winning seven. (Did someone say something about perilous margins?) A 6-5 record would be more likely than 7-4, but that’s just one win. If you flip the result of Shelton’s US Open quarter-final against Alcaraz, that’s 900 ranking points he doesn’t get. Take those away and he’s still #4 in the world.

Break points? The theory about guys like Sampras, Isner, and Nick Kyrgios is that they didn’t go full-out on all return points. So their RPW% doesn’t reflect their actual chances of breaking serve. Indeed, in 2023, Shelton won over 40% of his break point chances, breaking serve 18 more times than expected.

Kyrgios notwithstanding, most players don’t consistently over- or under-perform on break points. Sure enough, Shelton ended up winning fewer break points than expected in 2024–however much effort he expended, the results were worse when the stakes were higher. He returned to the positive column with another +18 in 2025. This year he won 36.5% of break points, good for 6.5 more breaks than expected. That’s pretty good, and with positive numbers in three or four seasons, he might really have this skill. Still, we’re talking about maybe two extra wins per year.

Clustering

The ATP rankings don’t care about RPW%, or points in general, or even won-loss records. The rankings disproportionately reward certain wins: late rounds and big events, preferably both.

It’s better, then, to cluster wins. All else equal, one title and one first-round loss is better than two quarterfinals. One good run, whether because of a hot streak or draw luck, is much more valuable at a major or a Masters. That’s why Roland Garros finalist Flavio Cobolli is playing in Turin this week.

Shelton went the distance at a slam, too. The US Open was one of his six finals this year. While he lost in New York, he went 4-2 in title matches. That’s hardly record-breaking, but for a player with a 45-19 record, it’s very good. Here is Shelton compared to four other players with similar records:

By most metrics, Shelton barely outplayed these guys, if at all. He lost more often at the start of tournaments: five losses in his first match at an event, seven more in his second. But when a big pile of ATP points were on the line, he usually won.

If you “de-cluster” Shelton’s results, you start to see just how much he was aided by the ATP algorithm (or his own clutch play in big matches, depending on your perspective). I’ve already mentioned that he would’ve ended up 840 points poorer had he lost to Alcaraz in New York. Had he lost the third-set breaker to Arthur Rinderknech in Paris, he’d have left with 100 points instead of 650. Replace those two matches with wins, say, against Faria in Cincinnati and Shevchenko in Miami, and he’d get only 40 of those points back. Same 45-19 record, but almost 1,500 fewer ranking points.

In 2026, that would still be good enough for 5th on the ranking list, so we might still be having this conversation. But Shelton’s record would be much less impressive. The counterfactual requires only swapping a handful of points.

Good not great

Fritz is a particularly good comparison. Shelton has has held 90.1% of his pre-Turin service games this year; Fritz held 89.8%. Shelton has broken 15.2% of return games; Fritz’s break rate was 15.8%. By those stats, the elder American had the (very slightly) better season, yet didn’t get the big results and had to scramble to end the year in the top ten.

Regardless of whether Shelton’s position at number four owes more to clutch play or luck, there’s just not much separating him from Fritz–or a number of other guys. There’s a clear ATP top tier of Sinner and Alcaraz, then a solid third-best in Alexander Zverev, who can’t hang with the top two but easily bests* the rest of the field.

* I almost wrote “outclasses” instead, but then I didn’t.

The third tier is much bigger. Check out the Elo table:

There’s essentially no separation between Fils, de Minaur, Medvedev, Jodar, and Shelton. The next five players aren’t far off either: These numbers give Fils a mere 56% chance of beating Lehecka, 53% to beat Davidovich Fokina. In the counterfactuals where Shelton loses to Alcaraz in New York and/or Rinderknech in Paris, he falls to the bottom half of this cluster.

None of this is meant to take away from Shelton’s accomplishments. He did win those clutch tiebreak points against Alcaraz, Rinderknech, and nearly everybody else on tour. When he faced Fritz head-to-head in finals, he won both times. The goal is to accumulate ATP points, and he did that better than almost anybody.

My point is to emphasize the narrowness of the margins, and the difficulty of pulling this off every year. Raonic was a brilliant tiebreak player, but when he won two-thirds of them in a season, you could rely on a regression the next year. Shelton has shown stretches of break point wizardry, but in 2024 he flopped on those clutch points. In 2026, he did everything right that a big-serving, 33% RPW player could do. Thanks to a crowded field, that earned him a precedent-busting spot in the top five. He might do it again, but it won’t be easy.

Does This Men’s Tennis Season Suck, Or What?

2026 is not going to go down as a great year in ATP history. We began the year poised for some scintillating Sincaraz. Instead, we got a nice step forward from Ben Shelton, a breakthrough from Rafael Jodar, and soon, an utterly unconvincing new number one, in Alexander Zverev.

Sinner and Alcaraz didn’t disappoint on court. They just weren’t able to stay on court. Alcaraz missed four and a half months, including two majors. Sinner’s knee has kept him out of action since he beat Zverev for the Wimbledon title. Both men played enough–and well enough–to make it clear they are the best two players on tour. But neither stayed healthy enough to earn the year-end number one ranking.

By process of elimination, that leaves Zverev. My Elo ratings put the German in third place, at 2,106. That’s 43 points behind Alcaraz and a whopping 190 points behind Sinner.

Kind of an awkward situation: The year-end number one is nowhere close to the year’s best player. Using yElo ratings–Elo ratings calculated based only current-year results; here’s 2026–Zverev trails Sinner by more than 200 points. I calculated year-end yElo ratings back to 1985. Over those 41 years, the yElo number one was the ATP number one 25 times. Of the others, there were only four occasions when the ATP number one trailed the yElo headliner by more than 100 points:

Year  ATP #1    yElo  yElo #1   yElo  Diff  
2026  Zverev    2118  Sinner    2323   205  
2022  Alcaraz   2116  Djokovic  2264   148  
1988  Wilander  2164  Becker    2282   119  
1992  Courier   2131  Becker    2238   107

Alcaraz’s 2022 season doesn’t quite feel like a fair comparison: He started the year outside the top 30, then his rating took a hit when he followed the US Open with four losses in ten matches. In any case, his underwhelming finish still left him closer to the yElo #1 than Zverev’s 2026 campaign puts him to Sinner.

The yElo of 2,118 is also weak in itself. Here are the lowest year-end yElos for full-season ATP year-end #1s since 1985:

Year  ATP #1            yElo  
2000  Gustavo Kuerten   2093  
2022  Carlos Alcaraz    2116  
2003  Andy Roddick      2117  
2026  Alexander Zverev  2118  
2002  Lleyton Hewitt    2119  
1998  Pete Sampras      2121  
1992  Jim Courier       2131  
1988  Mats Wilander     2164   
2001  Lleyton Hewitt    2176

Elo variants aren’t perfect for cross-era comparisons, but this list captures many of the less-convincing seasons from year-end number ones. Only Kuerten’s 2000 scored markedly lower than Zverev’s 2026. (In fairness, he could still improve it. In reality, he’ll probably lose to Jakub Mensik in Vienna next week.)

(Not) beating the best

The true knock on Zverev’s season is less abstract. For a number one (or even a number three), he hasn’t played many matches against high-quality opponents. Of those, he has won shockingly few.

Since 1985, the typical year-end number one has played 19.5 matches against top-ten opponents, winning 14.5 and losing 5. The average number one faces the rest of the top five 11 times, going 8-3.

To say that Zverev has not matched that standard is charitable:

That’s 3-8 against the top ten, 0-6 against the top five.

No year-end number one has posted as low as a .500 record against the top five since Federer in 2009. The last losing record was Sampras in 1998, when he managed only three meetings against such opponents.

For the last losing record against the top ten, we need to go all the way back to Stefan Edberg in 1991, when he went 5-8. And even he had three wins against the top five.

We can’t entirely blame Zverev for this. Most of the top ten wasn’t injured, they just lost too early to face the German. Zverev is an easy target, and he surely could’ve managed better, but the field underwhelmed as well.

Let’s be kind and say that the ATP is in a transition period. Jannik, my friend, we need you back.

Where Should You Hit An Approach Shot?

To the backhand, obviously. Or is it down-the-line, to cut down angles? Or maybe mix things up, so the opponent won’t start leaning the right way. Should the advice be the same for men and women?

Start with the basics. For the ATP, looking at all forehand, backhand, and backhand slice approach shots (excluding serve-and-volleys and net-rushes on returns), the Match Charting Project gives us over 84,000 data points since 2020. About 96% of approach shots land in the court. Of those, a bit more than one-third end the point immediately, either as winners or opponent errors. If the approach shot comes back, the net player wins 58.6% of the time. Altogether, when an ATPer hits an approach shot, he wins the point 70% of the time.

Of course, not all approaches are created equal.

Here are men’s approaches broken down by direction:

Direction  Share   Ends   Won%  
To FH      31.9%  37.4%  70.9%  
Middle      5.8%  23.9%  53.7%  
To BH      62.3%  30.8%  70.8%

Men overwhelmingly aim for the backhand. Approaches to the forehand are more likely to end the point immediately, but if the point goes longer, the netman fares better against the backhand. Apart from the weak results down the middle, we seem to have landed on a good balance. Maybe to-the-backhand is better against a naive opponent, but pros lean that way, knowing it’s the likely shot. To-the-forehands need to be better: ATPers miss those a bit more often, and judging from the winner/induced-error rates, those land closer to the lines.

If the forehand/backhand balance is an example of a well-executed mixed strategy, those down-the-middle approaches are … something else. The relatively poor results probably stem in part from circumstance. Players can be stranded in no-man’s land and hit an approach for a lack of decent options–those approaches are probably more likely to go down the middle. Apart from desperation shots, maybe there’s some benefit in an occasional approach at the feet, just to complicate the opponent’s guessing game. In any case, the down-the-middle share (under 6%) suggests that players know that approaches should usually go elsewhere.

The overall trends for women are similar. To-the-backhand is somewhat more successful (68.9% to 66.4% in overall win percentage), and down-the-middle approaches are even less so, at 49.6%.

What about the geometry?

All else equal, it’s best to approach from the middle, though it’s only slightly better than coming forward from the forehand corner. (For men: 70.8% win percentage to 70.3%, compared to 66.5% from the backhand corner.)

Down-the-line approaches make life harder for the opponent, but they are higher-risk. More errors, more winners, and mixed outcomes overall. For men, a down-the-line forehand approach beats a cross-court forehand approach by 1.7 percentage points. Among backhands, the risks and rewards cancel each other out. For women, there’s no net advantage from either wing.

If you’re in a corner, the important thing is to avoid approaching to the middle. You probably already figured that out, given the overall numbers. From a corner, the results are even worse. When men approach down the middle from a corner, they win only 50.8% of points. For women, it’s just 47.7%. Again, a lot of that disadvantage is situational. But that doesn’t mean it’s a good idea if you have a choice.

What about shot selection?

Forehands > backhands > backhand slices. Topspin backhand approaches are better than I would have guessed: ATPers coming in off the backhand win 67.9% of points, compared to 71.8% on forehands.

Slice approaches are a modern-game disaster. Once upon a time, slice approaches were a centerpiece of aggressive playing styles, hence the term “chip and charge” and all those graceful highlight-reel shots of Martina Navratilova and Pete Sampras. Now, baseline play is just too good. Men win barely half (50.4%) of backhand-slice approaches. As with down-the-middle approaches, some of that probably stems from net-rushing in suboptimal circumstances. But that’s doesn’t explain nearly all of it. Presumably slice approaches are better on fast surfaces, especially low-bouncing grass, than this dire average. Competitors themselves, though, are increasingly skeptical. Barely one in five backhand-side approaches are slices, as players tire of watching passing shots sail by.

(Edit! Presumably, you say? I got around to checking it. On grass, men’s slice approaches end up winning the point 53.6% of the time. Better than average, still not very good. For women, slice approaches have the same outcomes on grass as elsewhere.)

The shot-type breakdown is almost identical for women. WTAers win 68% of forehand approaches, 66.2% of topspin backhands, and 45.6% of backhand slice approaches.

Context matters

Major factors that I haven’t addressed–and can’t with any precision, given Match Charting Project data–are court position and quality of the previous shot. Clearly these things matter: Regardless of shot selection or direction, an approach has a better chance of success from inside the baseline against a weak reply than it does from behind the baseline against a Jannik Sinner cannonball.

One simple proxy for these factors is rally stage. More than half of men’s approach shots in the dataset occur on the third shot of the rally: the plus-one shot. On average, this is the best opportunity to come forward, as there’s only so much the returner can do with the return.

Here’s how those compare to subsequent stages of the rally:

Stage    Share   Ends   Won%  
shot 3   50.4%  33.9%  72.1%  
shot 4    9.4%  31.1%  65.8%  
shot 5+  40.2%  31.0%  68.0% 

A plus-one approach is rarely an awkward shot. Barring a net-cord or unexpected bounce, servers will only pounce on that next shot when the circumstances favor it. So while 72% isn’t quite the win rate for an approach shot when everything is set up just right, it’s pretty close. Net-rushers will always make errors, opponents will always come up with low-percentage passes made good, and so on. Still, by the standard of tennis’s small margins, a win rate in the 70s is huge.

Women don’t show quite the same splits. They win 67.7% of plus-one approaches, then 65.3% on fourth shots and the same 65.3% for everything beyond that. The serve doesn’t set up plus-one approaches to the same degree.

Try to distill all this to some simple advice, and you might end up with something generic like “pick your spots.” While not terribly helpful, it is true: Context probably matters more than forehand-versus-backhand or down-the-line-versus-cross-court.

As for concrete tips: If you’re forced to hit an approach up the middle, maybe don’t approach, unless the alternatives are even worse. And if you don’t need to go up the middle, hit your approach somewhere else. Anywhere else!

How Long Does Return Quality Matter?

A deep return is better than a shallow one. A return to a corner is better than one down the middle. There’s no doubt that return placement matters, but how long does it matter?

With every shot after the return, it is logical that the effect of the return would decay. A strong return might not go for a winner, but it probably induces a weak reply, making it more likely that the returner can end the point on her next shot. But if she doesn’t, if the rally continues to 6, 10, or more shots, at some point the quality of that return is no longer a factor. Either she sacrificed the advantage it gave her, or her opponent defied the odds to get back on neutral (or better) terms.

You guessed it: We can quantify this.

Drawing on the thousands of 2020s matches logged by the Match Charting Project, let’s start with the simplest comparison: returns that landed beyond the service line versus those that landed in the service box. Deeper is unquestionably better: For men, a back-half return means winning the point 50% of the time, while an in-the-box return is good for only 41.4%. Even controlling for match (opponent, surface, etc), the difference is 7.7 percentage points.

Most of that advantage shows up in just a few shots after the return. These plots show the percentage-point advantage of a deeper return at each stage of a rally:

Return depth matters more, and for longer, in the men’s game. A back-half return against a first serve is worth about 2.5 percentage points more to men than to women. Against first and second serves, return depth plays a role up to at least the eighth shot.

For women, this basic measure of return depth loses half of its predictive power once the rally reaches its fourth shot. By then, the server has either capitalized and ended the point, or she has squandered her advantage. (Or the return was good enough that she didn’t have an advantage to squander.) By the sixth shot the return is barely a factor at all: The confidence interval says the effect could be zero.

The server’s natural control of the point also explains the stair-step nature of the graph. A big part of the difference between return influence on 3-shot rallies and 4-shot rallies is that the points with the weakest returns become third-shot (plus-one) winners. There is no fourth shot. The effect is similar, if smaller, between the 5th and 6th shots as well, in some of the plots.

What about angles?

An ideal return forces the server to move away from the middle of the baseline. Intuitively, it seems like line-kissing returns affect the development of a point differently than deep ones do. To overgeneralize: If the server has to move, he could immediately be at a disadvantage. A deep return, by contrast, merely stops him from putting the point away.

The next set of plots show the difference between down-the-middle returns and crosscourt returns. (That is, crosscourt to the ~third of the court closest to the opposite sideline.) For men, the difference between crosscourt and middle isn’t as great as the gap between back-half and front-half. For women, it’s about the same. The return still remains a factor until the sixth or seventh shot:

(Ignore the women’s trend lines that head back upward for long rallies. That might be a selection effect. Maybe returners who neutralize the server’s advantage are more likely to come out ahead in marathon points.)

Down-the-line returns are a little different. The biggest effect of hitting in that direction is an increase in service winners: In the next set of plots, check out the diamonds in the 2* column for second serves. That column includes all non-ace points; the other data points on the graph depend on the server putting a second ball in play:

I expected that crosscourt returns would both be better than down-the-line returns and have a longer-lasting influence. I was wrong about the first. All else equal, down-the-line returns are a bit more effective, assuming they land in. Maybe servers expect a crosscourt return and lean that way. There’s a stronger case for the second. The advantage of a down-the-line return is gone for women by the sixth shot of the rally, the seventh shot for men.

Finding the backhand

Often, the returner’s goal isn’t to hit a particular corner, it is to target the server’s weaker side. For men, especially, that side is usually the backhand. Maybe there’s a detectable advantage–and ensuing influence–in finding the server’s backhand?

The evidence says: Not really. On men’s first serves, a little bit, and it’s gone after another stroke or two. The plots tell the story:

Of course, tennis is complicated and these are descriptive statistics, not recommendations. Returners have to weigh several factors (assuming they have time to think at all), not just which wing to target. If men tried to hit more returns to the backhand side, they’d have to do so with opportunities that aren’t as well-suited to a return in that direction, and that next marginal percentage point of returns-to-the-backhand might not be as good. In that sense, seeing near-equal outcomes for returns to either wing suggests that returners have found a reasonable, if not quite optimal, balance.

What about the extremes?

So far, we’ve been looking at one variable at a time. What’s the difference between a really well-placed return and a poor one? To put it another way, how much does return location matter at the extremes, and how long does the difference persist?

We can measure that by comparing “shallow-middle” (service box, middle third of the court) with “deep-corner” (back quarter of the court, outer thirds of the court) returns. The difference is substantial–nearly 20 points of win percentage for men on both first and second serves–but it doesn’t last much, if any, longer than the other effects we’ve seen:

For women, the advantage of a very well-placed return is almost entirely gone before the sixth shot of the rally. It lasts longer for men, with the confidence interval touching zero only on the ninth shot. Still, both server and returner have just a short window in which to capitalize on a weak or strong reply.

Putting it all together: A good or bad return affects the development of the rally even after more shots have been struck. But not forever: If somebody tells you that a return set up a winner on the 12th stroke of a rally, they’re overstating the case. A good service return can accomplish a lot, but its powers are not unlimited.

Do Groundstrokes Break Down?

The human brain is a trend-spotting machine. A couple of groundstroke errors in a row, and fans and commentators conclude that a player’s stroke has gone off the rails, perhaps because her opponent has exerted just the right pressure. Shots, we say, can “break down,” and even that opponents can “break them down.”

Except… an awful lot of the trends we spot are mirages, misguided attempts to make sense of noisy data. Yesterday, we learned that first serves aren’t streaky. For pros, missing one doesn’t make it more likely that you’ll miss the next; women even tend to make slightly more first serves after misses. So, what about groundstrokes?

It’s a trickier phenomenon to pin down, because unlike first serves, forehands (or backhands) are not all created alike. After missing one forehand, you might not see one on the next point or two. Or if you do, it could be at a different angle, a different height, a different court position, and so on. Still–and long-time readers are going to struggle to believe me on this–there is a detectable effect! Really!

Based on women’s matches since 2020 logged by the Match Charting Project, if a player makes a forehand unforced error on one point, she is 0.45 percentage points more likely to make a forehand unforced error on her next point with a forehand in it. It’s a tiny shift, representing about a 2% tick upward from the typical error rate around 20%, but it is real. The results are about the same (slightly weaker, but in the same neighborhood) for backhands. Men are in the same neighborhood, as well.

But wait: If we control for set, establishing our expected error rates for a single set rather than the entire match, the effect shrinks, and for men, it vanishes almost entirely.

This is good news and bad news. There’s little or no short-term influence of one error on the next opportunity to hit that shot. But error rates do drift over the course of the match.

Drift away

Splitting matches into chunks of 20 forehand points, error rates move around quite a bit more than that sub-half-point shift. Without controlling for set, the typical WTAer sees her forehand UFE rate jump around 1.4 percentage points more than would be expected from chance alone. That’s still not detectable within a single match, but it is a meaningful difference in the long run. Over the last 52 weeks, the range in UFE rate between the most error-prone player (Anna Kalinskaya) and the least (Belinda Bencic) is only 9 percentage points. 1.4 points represents a jump of several places at most positions on the leaderboard.

A bunch of subsequent tests reveal some aspects of that drift. For one thing, it’s usually not a single stroke that ebbs and surges. When a player starts making more errors, both wings suffer. There’s a hint in the men’s data that forehand and backhand error rates don’t quite move in lockstep, but the data is inconclusive.

It’s also clear that players seize on their opponents’ sloppiness. You can probably imagine situations where gaffes seem contagious and no one can keep the ball in the court, but typically, one player tightens up his (or her) game while the other one struggles. I don’t know which direction the causation points: Does one player “break down” the other? Does she take advantage of her opponent’s errors and play more conservatively? Anything along those lines is just speculation at this point.

The strongest effects, though, were somewhere I didn’t intend to look. Error rates shift chronologically, both across sets and within them, in predictable ways.

Tightening up

These graphs show the trend in relative unforced error rates by set, for both men and women:

On both wings, players make more errors in the first set. Women improve their forehand error rate by almost a full percentage point between the first and second sets. While there’s not much movement after the second set, the rates stay well below the first-set standard.

Before we speculate about the causes, let me show you another pair of plots, now comparing relative error rates within sets, by pairs of games:

Again, we start high then tack lower. Though it’s not an unequivocal downward trend, the first two games are consistently the most error-prone, and tiebreaks are the least. (Players become more conservative in tiebreaks by just about every possible metric, from serve speed to rally length to errors.)

Combine those two effects, and you have some sloppy opening games. Here’s one more plot, using WTA data, showing each set broken into chunks of 10 forehand points. The top left corner is the beginning of the first set; the bottom right is the end of the third. Red means more errors than expected (the numbers are percentage points), blue means less:

Every set tightens up as it goes along, but my oh my are those first ten points brutal. These shifts account for much of the “drift” within matches. Error rates move for reasons other than chronology, but chronology is a single factor that seems to affect a large number of players.

Sloppy or aggressive?

High error rates are not inherently bad: Jelena Ostapenko has made a career out of winning matches when she loses one of three points with a (wild) miss. The trick is to pair errors with winners, to take the good aggression with the bad.

That’s not what’s happening here! For women, on both wings, winner rates are at their lowest in the first set. Within sets, winner rates are lower in the first two games than they are until 5-5, when tiebreak caution starts to take over.

These numbers strongly point toward a “warm-up” effect at the beginning of the match and a smaller “re-set” effect at the start of each subsequent set. Maybe that’s understandable: No amount of actual warming up can prepare you for the reality of a new opponent’s individual mix of speed, spin, and angles. And despite the dark red of that upper-left-hand heatmap cell, we’re still only talking about a shift of one percentage point: Enough to pick up across 4,000 charted matches, but far too small to notice even in a fortnight of attentive tennis-watching.

What about the players?

Up to this point, we’ve been talking about averages. We don’t have enough data to identify tendencies for most players, but we can find some.

One of the strongest individual-player findings for women is that Coco Gauff’s error rates drift quite a bit within a match: about three percentage points compared to less than two for the average player. Gauff is slightly more prone than average to a sloppy start, as well, but the full-match metric excludes the first 20 baseline points, so her variance goes far beyond that.

By contrast, Aryna Sabalenka and Iga Swiatek are remarkably steady, showing effectively no in-match variation in error rates beyond what would be expected from chance.

Sabalenka also stands out in that her aggression (winners plus errors) goes up under pressure, defined as the ninth game of a set or later, with the score within one game. The effect is small–a bit more than one percentage point–but it is striking because the average trend under pressure is negative. Gauff, Elena Rybakina, and Jessica Pegula all score about one percentage point in the opposite direction, more conservative under pressure than the average player. Iga is almost precisely average.

I haven’t said much in this post about men, because the trends are generally the same across both tours. Still, there are some interesting individual numbers for ATPers.

Both Stefanos Tsitsipas and Grigor Dimitrov see even more error-rate drift within a match than Gauff. Lorenzo Musetti is the slowest starter, committing the most errors (relative to match averages) in his first 20 chances per match of all men for whom I have a lot of data. It’s suggestive, though hardly conclusive, that those three outliers all have one-handed backhands. I can sympathize: If I play a match on Monday morning, my backhand starts rounding into form on Wednesday afternoon. (Tsitsipas and Dimitrov, for what it’s worth, aren’t slow starters; their error-rate drift comes later.)

A slow start isn’t necessarily bad. Second on that list behind Musetti is Carlos Alcaraz, even though Carlitos scores about average for general within-match drift.

Jannik Sinner’s career of bloodless executions turns up in the data as well. His within-match drift is the lowest (tied with Felix Auger-Aliassime, surprisingly enough) of all players with enough data. In the big moments, he grinds you down, committing fewer errors under pressure and improving his winner-errors ratio. He isn’t the most conservative in the big moments, though: That honor belongs to Novak Djokovic, with the perpetually cautious Alexander Zverev right behind him.

Circling back to where we started: Pro groundstrokes don’t really break down. They take a few games to reach optimal form, and on occasion, they rattle a bit. Most of the apparent trends we spot as fans are fake. But in this case, there are indeed a few drops of water in the desert of statistical noise.

The Point After a Second Serve

Here’s a hypothesis for you: First serves are streaky.

It seems unquestionably true for amateurs. Our serves start clicking, or more commonly, fall apart entirely. Pros, though, are built different. After untold thousands of serves in practice, they should be steadier. The question, then, is whether they exhibit any streakiness at all, whether one missed first serve makes it more likely that they’ll miss the next one as well.

We can test this with the ever-growing dataset that is the Match Charting Project. Across over 4,000 charted men’s matches since 2020, the effect of one missed first serve on the next first serve is basically … zero. Not exactly: Men might be very (very) slightly anti-streaky, meaning that they are more likely to land a first serve on a point after a first-serve miss. But the effect is indistinguishable from zero, and the natural yo-yo of alternating deuce and ad courts erases anything that’s left. Some players have different results on the deuce and ad sides, so they are more likely to “recover” from a miss simply because they get to switch back to their preferred serve direction.

A few individual men show up as more streaky (or resilient, to use a more pleasant if less precise term for “anti-streaky”) than the norm. In the 2020s data, Novak Djokovic comes out as one of the most streaky, while Carlos Alcaraz scores among those at the other end of the spectrum. The magnitude, though, is tiny: We’re looking at shifts of about one percentage point in first-serve-in rate after a missed first serve. There’s not even enough data to be sure there’s an effect at all. Even if there is, that one percentage point translates to about one extra missed (or made) first serve every four matches, on average. It’s certainly not something you could detect by watching a match, or even a full tournament.

The story is somewhat different for women. Again using Match Charting Project data from the 2020s, spanning another 4,000-plus matches, we find that women make more first serves than expected on the point after they miss one. “Resilient” or “anti-streaky,” it doesn’t hold for every single player, but it’s a clear trend across the tour. As with the men, deuce/ad explains some of it, about one-fifth of the overall effect. What about the rest?

It would be one thing if there was simply no streakiness: We could give credit to the players for having trained themselves to the point that they are human ball machines. Maybe there’s something psychological, where players see a mistake, make adjustments, and temporarily become even better servers than their typical selves. More likely, though, it’s tactical. My instinct is that the average woman’s first-serve rate goes up because, having just missed a first serve, she becomes more conservative.

The evidence supports my guess. On first-serve points after a first-serve miss, women win 0.3 to 0.4 percentage points less often, compared to other first-serve points within the match. It’s possible that the additional lost points don’t stem directly from serve tactics, but the persistence of the effect points strongly in that direction.

We can even be a bit more specific. The difference is not speed: I checked the first-serve speeds after first-serve misses from Wimbledon and the US Open between 2021 and 2024. Controlling for serve direction, women served just as hard on those follow-up points, possibly even a tiny bit harder. If the point-outcome data can be trusted, the conservatism comes from serve placement, not direction (wide/body/T) or speed.

The net effect of all this is basically neutral. It’s good to decrease the odds of another second serve, but a slightly weaker first serve reverses some of that benefit. After missing a first serve, women servers win points at about the usual rate.

Double faults

Everything we’ve seen so far is a tiny effect, and it’s virtually no effect for men. The psychological or tactical shift is a bit clearer when we shift from missed first serves to double faults.

After a double fault, both genders dial things back with their next first serve. Men drop about 0.25mph from their average first serve. Roughly half of that comes from picking safer targets: They are less likely to fire one down the T, where margins are smaller. Women show a much bigger effect, dropping just over 1mph from their first serve. More than half of that comes from the direction. They, too, opt for body or wide serves as they seek to get things back on track.

Here, the results are far from neutral. There’s some degree of momentum regardless of which serves go in. Both men and women are a bit more likely to win a point after winning the previous point. But a double fault is recognizably different than a point lost by other means. The average woman’s first-serve winning percentage drops by almost a full point after a double fault, compared to other ways they could have lost the previous point. A couple of studies of mine suggest that 1mph in serve speed will move first-serve win percentage between 0.2% and 0.5%. So serve speed explains part of the win-percentage drop after double faults; other manifestations of serve caution, like placement, may account for the rest.

The men’s serve-speed drop is much smaller, so it’s no surprise that the point-result effect is smaller as well. The data doesn’t allow us to be as confident about this one as a real effect, as opposed to just suggestively-oriented noise.

My initial hypothesis was wrong, but the investigation turned up some momentum (and anti-momentum) effects that are worth more attention. Should players try to keep their first-serve risk level steady, or is a conservative offering a good way to get back on track? Is it possible to go wide or to the body without giving up quite so much? Should returners lean more in that direction, knowing that a double fault could influence the next serve? Beats me!

Do Pushers Grow Out Of It?

Is Victoria Mboko a pusher? I was tagged in a Twitter exchange the other day debating that exact question. From watching her play, I would hardly compare her to Ostapenko, but the p-word wouldn’t have come to mind.

Rally Aggression Score puts her at +6, on a scale designed to run from -100 to +100. Based on 27 charted matches from the last 52 weeks, that makes her about neutral, tactically similar (at least along this dimension) to the likes of Belinda Bencic and Karolina Muchova. Counterpunchers, maybe, not pushers.

On the other hand, “neutral” overstates it. Lowell West devised the metric about a decade ago, and the game has changed since then. The way I initially scaled it, Petra Kvitova flirted with triple digits, and that was about it. Now, Dayana Yastremska gets close to 200. Elena Rybakina, at exactly 100 in the last year, ranks only fifth among players with at least five charted matches. There’s nothing so extreme at the other end: Emma Navarro has averaged -68, while the most passive top-tenners are Coco Gauff and Mirra Andreeva at -20 and -30, respectively.

Mboko’s +6, while hardly vintage Wilander, is indeed more passive than the average WTA player in 2026.

Aggressive growth

Will the Canadian change tactics with age? These days, it seems like nearly everyone at least tries to take more chances and hit more winners. (If you don’t, your opponent will!)

I can’t tell Mboko’s future, but I can load up a database and run queries on it.

So, different question: Do players in general change tactics with age? It’s easy to think of examples: Serena Williams got more aggressive over the years. Ostapenko (remarkably) has done the same. I’m not sure there are equally prominent players who have moved in the other direction, but it does seem like some teenagers show up bashing balls, then settle for a more measured game in their 20s.

I took all women with at least 20 charted matches, grouped their rally scores by year, and figured out how much each year’s score differed from their career norms. Average those deviations across all players, and this is what you get:

From age 18 to 36–the entirety of most players’ careers–there’s nothing to see here. Plus or minus five points of rally aggression score is a rounding error, especially since these are non-random samples based on what Match Charting Project contributors wanted to chart and could find on video.

The spike at age 37 is also not very instructive: It’s basically Serena Williams, and we already knew that as she aged, she took increasingly few prisoners. From 2007 to 2009, her average aggression score was about +20. In 2019, when she was 37, it was +109.

Still, that climb at the left end of the graph is suggestive. It doesn’t literally apply to Mboko, who is already 19. But it supports a plausible narrative. When young women–especially the youngest prospects–arrive on tour, they aren’t as strong, or perhaps even as tall, as they will soon become. (Even Lilli Tagger recently gained a centimeter.) Their experience is disproportionately against juniors, who are even less physically imposing, relative to adult pros. The typical 16-year-old, no matter how talented, isn’t going to show up on tour and play like Sabalenka.

There are exceptions, of course. Maria Sharapova’s highest single-season aggression score was in 2004, when she was 17. Madison Keys arrived on the circuit playing essentially the same game she would play for the next decade. Iga Swiatek was more aggressive at age 18 than she has been since.

On average, though, the youngest players are more conservative–fewer winners, fewer errors–than they will become as they graduate from their teens. The mechanism could well apply to a 19-year-old, too.

The trend is null

We’re always talking about how players could develop and improve, or what their new coach brings to the table. Yet the undefeated champion of tennis forecasting is the null hypothesis. The null hypothesis, of course, means that you should’ve skipped this entire post and watched a Friends rerun instead.

It is boring, but the best way to predict how a player will look next year is to point at their current results and say, “yeah, just like that.” You can fiddle around the edges and maybe find players on the cusp of something or other, but … no, usually you can’t even do that.

Here are the year-to-year aggression trends of the women with the most matches in the MCP database:

See that trendline that goes from the lower-left corner up to the upper-right? No, you don’t, because it’s not there. The closest is Serena, who dipped to +12 in her late 20s before getting hyper-aggressive in her 30s.

To answer the question posed by my headline, then: No, pushers don’t grow out of it. The style of play you see today is a good predictor of the style of play you’ll see a decade from now.

Lucky for Mboko, then, that she isn’t a pusher. Throughout her career, I’ll bet she tries a lot of new things and incorporates plenty of fresh ideas from a handful of high-profile coaches. The end result may well take her to the top of the rankings: A 19-year-old ranked 6th on the Elo table has a bright future ahead. If she finds a game that takes her all the way to the top, odds are it will look a lot like what we’ve seen so far.

What Does Felix Auger-Aliassime Do So Right On Indoor Hard Courts?

Felix Auger-Aliassime has earned a reputation as a world-beater on indoor hard courts. He’s no Jannik Sinner–as Sinner reminded him all four times they met last year, twice indoors–but FAA is a fearsome customer against just about anybody else.

Last week the Canadian added to his indoor title haul with his second-straight championship in Montpellier. This time, he straight-setted Adrian Mannarino. While that win doesn’t particularly raise any eyebrows, the body of work keeps growing. It’s his eighth career title indoors, three of them at ATP 500s. Last fall in Paris, he also reached his second Masters final. (The first was in Madrid, the indoorsiest of the clay Masters.)

What’s the secret?

The conventional wisdom is that he has a big game, especially a deadly first serve. The controlled environment indoors, plus typically fast conditions, play to his strengths. The serves skid across the court even faster. His weaknesses are mitigated because the bounce is more predictable and because points are shorter.

All that sounds plausible. My only gripe is, couldn’t you say that about a lot of players? The whole paragraph applies, almost word for word, to Hubert Hurkacz, who has two Masters crowns on outdoor hard, plus a clay title, yet just a pair of indoor 250-level championships. What about Matteo Berrettini? The description might match him even better, yet the Italian has never won a title indoors. He has reached only one indoor 250-level final.

Before we go to the numbers, let me give you my seat-of-the-pants theory. FAA has huge weapons, but he doesn’t always play like it. He doesn’t consistently swat away easy plus-ones like Berrettini does. He gets sucked into long rallies, where he’s often at the disadvantage. Indoors, though, he knows what the tactics are, and he plays the way he should play. Indoor Felix, then, is the best Felix, both because his game is suited to the conditions and because he shows up with the right approach.

On the other hand, the last two points against Mannarino on Sunday were 8- and 18-shot rallies, respectively. So, you know, don’t trust my pants.

Numbers!

You can, however, trust the spreadsheets. The Match Charting Project has well over 100 Auger-Aliassime matches. Going back to 2020, the total includes 41 on indoor hard and 40 on outdoor hard, nicely suited for some comparisons. I was tempted to throw out the seven Tour Finals matches from the indoor tallies, because they skew the quality of the opponents, but the Canadian’s indoor averages are about the same with or without them.

Start with serve stats:

Surface       Unret%  <=3 W%  RiP W%  
Indoor Hard    35.7%   47.7%   55.5%  
Outdoor Hard   32.8%   43.4%   49.7%

About three percentage points more serves don't come back, and there's an even wider gap in points polished off on the serve or plus-one (the "<=3 W%" stat). The biggest gap here is in points won when the return comes back. Sub-50% is below average, especially for hard courts. 55% or better is very good, even in fast conditions.

Almost all of the indoor/outdoor serve differences are thanks to the first serve. FAA's second-serve numbers are about the same regardless of roof status.

Of course, Felix isn't the only guy on tour who wins more easy serve points indoors. I don't have comprehensive stats on the indoor/outdoor split, so I can't tell you the exact tour average. But we can compare how much Auger-Aliassime gains on serve to how much he gives up on return:

Surface        RiP%  RiP W%  
Indoor Hard   65.3%   48.8%  
Outdoor Hard  67.5%   43.2% 

He retrieves 2.2 percentage points fewer serves indoors--better than the 2.9-percentage-point difference he gains on serve. But when he gets the serve back, he's actually better indoors than outdoors! He gains five percentage points in that department on serve, and he gains the same margin on return.

This might dovetail with the conventional wisdom. His monster serve really pays off indoors. And predictable conditions give him a bit of cover on return.

Whatever the reason, Auger-Aliassime's groundstrokes are way more effective indoors. My Potency metrics, FHP and BHP, combine winners, unforced errors, and shots that set up winners and errors. They give you one-number estimates of how valuable each shot is, and... wow:

Surface       RallyLen  FHP/100  BHP/100  
Indoor Hard        3.7     +8.5     +0.0  
Outdoor Hard       3.9     +3.2     -5.8

His indoor points are a little shorter, but I assume that is typical. I would've guessed that the difference was greater.

The Potency numbers (expressed here as rates per 100 shots), tell a more emphatic story. A +3.2 FHP/100 is ok, not great. Tommy Paul and Ugo Humbert are in that zone. On the other hand, +8.5 is the 52-week average of Carlos Alcaraz. A -5.8 BHP/100 is near the bottom of the pack, below the likes of Ben Shelton and Grigor Dimitrov. By contrast, +0.0 is, as it sounds, a good solid average.

These numbers don't drill into the "why" questions that naturally follow. But they help us pick between theories. I suspect that much of the difference in groundstroke stats has to do with the shots he gets to hit. The winners are downstream of good serves. Auger-Aliassime picks up some aces, but he picks up more plus-one (or even plus-two) winners, and those make his forehand and backhand numbers look good.

The "indoor predictability" thesis also looks good here. Remember that everybody should benefit from that--and not everybody's numbers improve like Felix's do--but it may be that the Canadian is more-than-typically exposed by the vagaries of outdoor play.

All the angles

Quick thought experiment. Picture Roger Federer hitting an ace.

Now imagine Auger-Aliassime hitting an ace.

What specific serves came to mind? If you're like me, you pictured Federer shooting a bullet right down the tee. And then you visualized FAA hitting a flat bomb out wide.

Of course, both guys hit plenty of aces in every direction. The charting stats suggest that Felix has a slightly better chance of an ace when he goes up the middle. (Federer did too, by a bigger margin, as do most players.) Still, this indoor/outdoor split caught my eye:

Surface       Deuce Wide%  Ad Wide%  BP Wide%  
Indoor Hard         50.5%     47.8%     33.7%  
Outdoor Hard        46.9%     45.2%     41.0%

Each column shows how often Auger-Aliassime opted for a wide serve in various scenarios. The first-serve differences are probably more marked, because his second-serve tendencies are about the same.

Indoor, he goes wide more often--but less often under the pressure of break point. While the margins are rather slim, it seems like the wide serve becomes his bread-and-butter indoors, and he uses the tee serve to mix things up on break point--because he's hitting more wide serves the rest of the time.

Wide serves are more likely to come back, but they don't make the returner any more likely to win the point. Especially against Felix: His signature serve might not even be an ace, but a wide bomb that the returner just barely plops back over the net.

The fact that he hits more wide serves indoors explains a lot. He gets a few more unreturned serves (as everybody does, probably), but he gains more of an advantage on the serves that (weakly, oh so weakly) come back. His groundstroke stats sparkle, padded by those easy balls.

Here's one final comparison:

Surface       2ndAgg  
Indoor Hard       +7  
Outdoor Hard     +46

"2ndAgg" is the Aggression Score stat tailored specifically to second serves. A higher score means more double faults and more unreturned second serves. Lower means fewer risks on second balls. +7 is quite conservative: Only about a dozen players consistently score so low.

But--those careful second servers include Sinner, Hurkacz, and Berrettini. With a game like Auger-Aliassime's, the second serve isn't the time to take risks. And indeed, in all of his indoor finals, he has never topped a double-fault rate of 5%. In the Montpellier final, he missed his second serve just once, and he committed no double faults at all in the quarter- and semi-finals.

Here, finally, is some support of my seat-of-the-pants theory, that when Felix goes indoors, he plays the way he ought to be playing all the time. He stays within himself, which is still imposing enough to earn a lot of cheap points. It's not a particularly complicated story, and I'm still not convinced why it doesn't apply to a half-dozen other guys on tour. Maybe it is all about the wide serve, the signature shot that allows Auger-Aliassime to manage risk and put his opponents on the back foot, all at the same time.

Defanging the Ball Bashers With Sara Bejlek

On Saturday, 20-year Sara Bejlek won her biggest title–by far. She had just slipped out of the top 100, so after qualifying for the main draw in Abu Dhabi, she charged past the likes of Jelena Ostapenko and Clara Tauson, then secured the trophy with a 7-6, 6-1 win over almost-top-tenner Ekaterina Alexandrova.

It’s tough to overstate just how out of the blue this was for the Czech. By ranking, Ostapenko, Tauson, and Alexandrova represent three of her four highest-ranked victories. Her two other main-draw victims, Sonay Kartal and Ashlyn Krueger, also count among her top ten. And she beat Kartal 6-0, 6-2.

That’s a big-hitting set of opponents. Bejlek, by contrast, lacks the power weapons that are becoming standard on the WTA tour. As a left-hander, she practically begs us to call her “crafty.” One upset against that group is plausible enough: After all, Ostapenko’s low-percentage tennis invites chalk-defying outcomes. But so many?

The final

I don’t pretend to be an expert on the Czech’s game. She has only a couple dozen tour-level matches under her belt, so she’s a newcomer for most of us. That said, we now have a detailed match chart from Saturday’s final that offers some clues as to how Bejlek battled the barrage of ball-bashers in Bahrain Abu Dhabi.

The conditions helped. It was windy, and the conditions were slow. None of that favored Alexandrova, who likes predictable balls she can smack flat back across the net. The court speed not only made it difficult for Alexandrova to hit through the court, it gave her a little less pace to work with on Bejlek’s own balls. Surely the Russian must have wished she had played indoors in Ostrava(!!!) instead.

The lefty’s game plan seized on those advantages. She looped balls back down the middle. She sliced more than she had to, refusing to give her opponent a predictable bounce height. She mixed in some almost impossibly slow serves. She won the first point of the match with a dropshot-lob combo that, while she didn’t attempt many more, surely gave Alexandrova pause.

The central result was that the Russian just couldn’t hit winners. By my count, she ended with 13 winners against 37 unforced errors. Just as telling as the abysmal ratio was that the 13 winners represented less than 10% of total points. In her last 60 charted matches–going back to 2022–opponents have held her under 10% just six times. When they do, it’s usually because they take the racket out of her hands by playing hyper-aggressively themselves. Two of the opponents in question were Anisimova and Yastremska.

Bejlek, by contrast, gave Alexandrova ball after ball that looked like it should have been obliterated. In different conditions, or when the Russian was in better form, maybe the winner count would have been much higher. But from the first few games, it was clear that the Russian wasn’t confident in her ability to take control. Big, aggressive hitters usually have more influence on rally length than more passive opponents, but that wasn’t what happened on Saturday. The average point lasted 5.1 strokes, tied for third-longest among the nearly 100 charts we have from Alexandrova’s career.

Translated into tennis cliché: Bejlek let her opponent beat herself.

If we can extract one concrete skill from the Abu Dhabi final, it’s that Bejlek doesn’t let servers overpower her. 85% of Alexandrova’s serves came back, compared to a 52-week average of 75%. Again, we don’t have much data yet on Bejlek, but here’s another bit of evidence: In Madrid two years ago, she retrieved more than 80% of Rybakina’s serves. The Aussie Open champ is the toughest on tour to return, usually holding opponents to around 67%.

Despite all the frustrations and all the extra shots she had to hit, Alexandrova nearly pulled out the first-set tiebreak. She led 4-2 at the change of ends before getting dragged into a series of long rallies broken up only by a couple of well-executed short points from the Czech. Having dropped the 95-point slog that was the first frame, Alexandrova ran out of ideas. She simply watched the error count soar.

You don’t win slams by letting opponents beat themselves–marathon runners notwithstanding. But in the hands of someone persistent enough, it’s a game plan that can keep you in Bejlek’s new neighborhood of the top 40. With 500 points on the books from Abu Dhabi, the left-hander has the rest of the season to prove that she belongs.