Credit: Justin Smith
I’m counting down the 128 best players of the last century. If this is your favorite player, Congrats! (or: I’m sorry.)
* * *
Roger Federer [SUI]Born: 8 August 1981
Career: 1999-2022
Plays: Right-handed (one-handed backhand)
Peak rank: 1 (2004)
Peak Elo rating: 2,383 (1st place, 2007)
Major singles titles: 20
Total singles titles: 103
* * *
Was Roger Federer clutch?
The conventional wisdom, I think, says he was not. The consensus view of Federer is that he was brilliant, inspiring, one of the greatest of all time, but… maybe a little tight in the big moments.
The headline stat is that he lost 24 matches after reaching match point. The list includes some very high-profile clashes, including a 2005 Australian Open semi-final against Marat Safin, the 2006 Rome final against Rafael Nadal, and the 2018 Indian Wells final against Juan Martín del Potro. It is impossible to forget Fed’s three failures to put away Novak Djokovic, first at the US Open in 2010 and 2011, then for the 2019 Wimbledon title.
We can cut Roger a little slack for the 2011 loss–that was The Return–but in that match and others, he squandered plenty of opportunities by missing forehands of his own.
Flip the result of those three Djokovic matches, and Federer retires with at least 21 majors–probably 22 or 23–and he departs the scene with Novak stuck at 18. The Austrian coach Günter Bresnik goes so far as to call Fed an “underachiever,” suggesting he should’ve gotten to 30 grand slams.
It’s not just match points, either.
You may recall the years-long gnashing of teeth and rending of garments over Federer’s inability to convert break points. He lost to Tommy Robredo at the 2013 US Open, seizing just 2 of 16 break chances. At the time, it felt like a trend. Or maybe a disaster movie. A few months later, he converted only 1 of 10 in a loss to Lleyton Hewitt.
At moments like that, the best you could say about Roger was that he was so good, it didn’t matter that he choked away all those opportunities. You could emphasize (as I did at the time) that service returns weren’t the centerpiece of his game, so we shouldn’t have expected him to capitalize on many of his break chances, especially compared to guys like Djokovic and Nadal.
Whether the focus was on break points, match points, or something else, we were always left with a paradox. Clutch performance seems to be a key component of athletic greatness. Federer is, without question, one of the all-time greats. Could a man with a penchant for choking really fill up so many pages of the record books?
* * *
In a roundabout way, it’s an immense compliment to Federer that this narrative has taken hold. To say that a 20-major winner was an underachiever is to imply extraordinary things about his talent. He made the game look so easy that his physical capabilities seemed limitless. By that reasoning, any failure had to be chalked up to his mind.
Roger drove otherwise sensible people to near-poetry. A few lines from David Foster Wallace’s famous essay, Roger Federer as Religious Experience:
He is never hurried or off-balance. The approaching ball hangs, for him, a split-second longer than it ought to. His movements are lithe rather than athletic. Like Ali, Jordan, Maradona, and Gretzky, he seems both less and more substantial than the men he faces.
Early in his career, Federer didn’t connect with fans the way he did later. He looked so casual, spectators weren’t sure he was giving full effort. One man who understood what was going on beneath the surface was Pete Sampras, who Roger beat in an era-defining match at Wimbledon in 2001. Sampras told Christopher Clarey, author of the recent Federer biography The Master:
It might look like Roger and I are not trying or we’re not that into it. We’re just very efficient: our movement and our games and our strokes. It’s like one swing of the racket, one forehand, one serve, and boom, it’s done, while most other players are grinding, grinding, grinding.
Fans in the 1990s sometimes struggled to recognize Pete’s effort, but no one ever doubted he was clutch. Sampras readily admitted that he coasted through some return games, saving energy for higher-leverage moments just as Jack Kramer and his followers had done for half a century. Then, at break point, or in tiebreaks, it was clear he could access another level.
Federer was, in a perverse sort of way, too good for that. He grew up playing on clay, and though he was no Djokovic on return, he had the skills to back himself on every point. In the 2009 Wimbledon final against Andy Roddick, he struggled for 30 games to earn a break in the marathon fifth set. Roddick was one of the best servers in the sport’s history, shooting bullets on the tour’s fastest surface, yet he managed to win just 3 of 15 service games to love. The deciding-set score of 16-14 is an example of tennis’s minuscule margins. But in that set, Federer won 26 return points to Roddick’s 16.
The American never did reach match point that day. He did, however, get that close against Roger at the 2006 Masters Cup in Shanghai. Roddick won the first set and took a 6-4 lead in the second-set tiebreak. After Federer erased those two chances, A-Rod earned another at 7-8 on Fed’s serve. The Swiss pulled out the breaker, 10-8, then hit three aces in the final game to secure a 6-4 third set.
In other words, Federer wasn’t the only man to become vulnerable with match point on his racket. Though Roger lost 24 of those heartbreakers, he won 22 matches after coming within one point of defeat himself.
* * *
We’re not going to settle the question of whether Roger was clutch by counting just the handful of matches when a match point was defied. Yes, Fed’s rivals managed better–Djokovic has lost only three matches from match point up while coming back from the brink 15 times–but in a career spanning more than 1,500 matches, a 48% win rate in 46 of them barely moves the needle.
So we need to take a wider view. There are more routine forms of pressure that have a bigger cumulative impact.
In at least one important way, Federer was clutch. His career tiebreak record was 466-247, good for a winning percentage of 65%. That’s the best of all time.
We’d expect someone at Roger’s level to win a comfortable majority of his breakers. While big serving does not predict tiebreak success, excellent tennis playing does, because–well, duh, of course it does. Federer was better than his average tiebreak opponent, to the extent that if he played as well in tiebreaks as he did in the rest of the relevant matches, he’d win about 60% of them.
That difference–between the expected 60% and the actual 65%–means that he won 33 tiebreaks that he “shouldn’t have.” In the 30-plus years for which we have the necessary stats, only John Isner has snuck off with more of these “unexpected” tiebreaks. A few more players–including Djokovic–have outperformed at a slightly higher rate than Federer did, but even in that category, Roger is near the top of the leaderboard.
We can’t directly translate those bonus breakers into wins and losses. The Swiss would’ve pulled out some of the matches even after losing a tiebreak set, and stealing a single tiebreak still leaves open the possibility of defeat. But if we’re talking about performance under pressure, we can’t ignore tiebreaks just because they didn’t involve a match point. Federer may well have added a couple dozen victories to his career total purely on the basis of his better-than-expected tiebreak prowess.
The same sort of logic is required to quantify a player’s performance on break point, both serving to save them and returning to convert them.
Facing break point, Federer won about 3% more often than he did at other moments against the same opponents. Most tour regulars manage to raise their game at that juncture, but not quite by that much. While Nadal and Djokovic outscore him on this measure, many of his strongest peers–including Hewitt, Andy Murray, and David Ferrer–do not. At the very least, his performance facing break point is not a strike against him.
Even attempting to convert break points, Roger was better than his reputation. He was about 1% less effective returning on break point than at other scores. Over more than 10,000 career break chances, that’s just a few points each year that he should have won but didn’t. It’s easy to point to dreadful performances like the loss to Robredo, and 2013 was one of a couple of abysmal seasons. In the big picture, though, Federer turned break points into breaks about as often as you’d expect from a returner of his caliber.
* * *
Break points and tiebreaks are better tools than anecdotes and match points. But they are still limited. Bill Tilden told us a century ago that the most important points are 30-15 and 15-30, and those, to take just one example, don’t show up in any of the numbers I’ve cited so far.
My best attempt at an all-encompassing clutch stat is something I call Balanced Leverage Ratio, or BLR. It’s possible to quantify the leverage–the importance–of every point, from love-all in the first game to 6-all in the deciding set tiebreak. It’s possible to win a match despite losing more than half of the points, as long as you win enough of the higher-leverage ones.
To calculate BLR, we find the average leverage value of the points a player won, then do the same with the points he lost. BLR is the ratio between the two, adjusted slightly to balance serve and return performance. If a player performs equally well regardless of the impact of the moment, BLR is 1.0. If he excels when the pressure ramps ups–essentially the definition of clutch–BLR is greater than 1. If he crumbles when the points matter more, BLR is less than 1.
It’s possible to calculate BLR only when we have the point-by-point sequence of an entire match. We’re lacking that for much of Federer’s early career, and even when the data is “out there”–as it is for most professional matches of the last several years–it isn’t yet wrangled into a form for easy data analysis. Still, I have a collection of this data for tens of thousands of matches, including over 400 of Roger’s, mostly from the 2010s.
Conveniently, the dataset also has about 400 matches each for Nadal, Djokovic, and Murray.
If you’re scrolling down just a paragraph or two at a time, take a moment and predict how you think BLR rates the four men.
Here is how they stack up, measured by the ratio between the importance of points they won and points they lost:
Player BLR Federer 1.08 Nadal 1.07 Murray 1.06 Djokovic 1.05
This stat, broad as its purview is, doesn’t prove that Roger was the clutchiest of them all. The margins are too small, and for most definitions of clutch, we’re not that concerned about how stars perform in routine matches–the majority of the material fed into this particular calculation. Still, like the tiebreak and break point numbers, BLR bolsters the case that, at the very least, Federer wasn’t not clutch.
The man played over 235,000 tour-level points. We can surely look past a few dozen lapses on match point.
* * *
If you’re unconvinced by all this, there’s an easy way out.
Who cares if Federer was clutch?
I find it fascinating that Fed fans, loyal as they are, have largely conceded the greatest-of-all-time debate to his rivals. A few points remain in Roger’s favor, like his record eight Wimbledon titles and 103 tour-level titles, a mark that trails only the somewhat padded tally of Jimmy Connors. But Djokovic and Nadal have pulled ahead in the grand slam race, and Federer finished his career with losing head-to-head records against both.
So the case for the Swiss, such as it is, rests on the sort of fuzzy arguments that use words like “transcendent.” I get it. I’ve spent much of this year reading about life-changing moments where a young fan saw Tilden, or Don Budge, or Lew Hoad, or Sampras and concluded that tennis could not possibly be played any better. Federer probably had that effect on more people–myself among them–than anyone else in history.
Poll today’s ATP and WTA locker rooms and ask everyone for his or her first idol. Rafa, Novak, and Serena have their fan clubs, to be sure. But Roger wins by a landslide.
For years, the otherworldly skills that so influenced a generation of players rendered clutch irrelevant. Federer won his first seven major finals without going to a fifth set. The 2005 Australian defeat to Safin was the only match point loss of his first decade on tour with any lingering impact on his legacy. In 2005, he won 81 of his 85 matches. 62 were straight-set victories. Who needs clutch when you’re playing like that?
One of the odder Federer statistics is his record in so-called “lottery” matches. 36 times in his career, Roger lost a match despite winning more than half of the points played. These things tend to even out, but not for him. When he was outpointed, he won only 11 matches, two of them by retirement.
With small samples like 47 lottery matches or the 46 reversals from match point up or down, it’s tough to separate skill from luck. The big picture says Federer was just fine under pressure. A few painful memories, especially against Djokovic, suggest he could be fragile, particularly when the moment was heightened in ways that statistics might not be able to capture.
Federer got his share of luck in his two decades on tour. He could have used a little more. Most often, though, his margin of safety was substantial enough that it didn’t matter.
When Roger won the 2009 French Open to complete his career grand slam, Andre Agassi was there to present the trophy. “A lot of people say it’s better to be lucky than good,” the American quipped. “I’d rather be Roger than lucky.”
* * *
Previous: No. 6, Serena Williams
Next: No. 4, Novak Djokovic
Subscribe to the blog to receive each new post by email: