Should Serena Be Seeded?

Italian translation at settesei.it

Serena Williams returned to professional tennis this month after more than a year of pregnancy, childbirth, and recovery. She took wild cards into both Indian Wells and Miami, competing as an unseeded player for the first time since August 2011. In her initial effort in California, she reached the third round before falling to sister Venus, and this week in Miami, she drew Indian Wells champ Naomi Osaka in her opening match and went home early, losing 6-3 6-2.

Seeing Serena without a number next to her name feels wrong. She left the tour for maternity leave just after winning last year’s Australian Open, a title that moved her back into the No. 1 ranking position. While she is clearly rusty–as she has been after previous absences–there’s little doubt she’ll quickly resume competing at a top-32 level (the threshold for an Indian Wells or Miami seed), if not considerably higher.

The brutal Miami draw and Serena’s ensuing early exit prompted all sorts of commentary, much of it calling for a rule change, some castigating the WTA for its lack of a maternity leave policy. The latter is not quite true: The WTA rulebook addresses absences for childbirth and treats returning players almost exactly as it handles women coming back from injury. Nevertheless, edge cases–like the greatest player in women’s tennis rejoining the tour without a single ranking point to her name–tend to put rules to the test.

Seedings are not just a convenient way to identify the top players on a printed bracket. They have an effect on the outcome of the tournament. In the March tournaments, seeded players get free passes to the second round. At every event, the seeding system keeps top players away from each other until the final rounds. Even minor differences, like the one between the fourth and fifth seeds, can have a major effect on two players’ potential routes to the title. This is all to say: Seedings matter, not just to returning players like Serena, but also to everyone else in the draw. While granting a seed to Williams right now may be the right thing to do, it would also push another seeded player into the unseeded pool, affecting that competitor’s chances at late-round ranking points and prize money. It’s important to acknowledge how the rules affect the entire field.

In a moment, I’ll outline various approaches the WTA could take to deal with future maternity leaves. I don’t have a strong opinion; there’s merit in each of them, as I’ll try to explain. What is most important to me, as a fan, is that any rules adopted are designed for the benefit of the whole tour, not just patches to handle once-in-a-generation superstars. Serena deserves a fair shake from the WTA, and her peers are entitled to the same.

1. Minor tweaks to the existing rule. The most likely outcome is almost always the status quo, and Osaka notwithstanding, the status quo is not that bad. The WTA rules allow for returning players (whether from injury or motherhood) to use a “Special Ranking” (SR) in eight events, including two slams.  The SR is the player’s ranking at the time she left the tour, and it determines whether she qualifies to enter tournaments upon her return. While Serena used wild cards for her two events thus far (more on that later), she could have used her SR for either or both.

In other words, new mothers are already allowed to pick up where they left off … with the important exception of seeding. Serena’s SR will allow her to enter, say, the French Open as if she were the No. 1 ranked player, but unless Roland Garros invokes their right to tweak seedings (like Wimbledon does), her seed will be determined by her actual ranking at that time. Since it’s only two months away, it’s very possible she’ll be unseeded there as well, making possible another nasty first-round matchup in the vein of the Simona HalepMaria Sharapova opener at last year’s US Open.

The debate over seeding boils down to “respect” versus “practicality.” Serena’s achievements and her probable quick return to greatness suggest that she “deserves” to be seeded as such. On the other hand, many players (including Sharapova, different as her situation is) have had a hard time returning to their previous level. The post-comeback results of Sharapova or, more recently, Novak Djokovic, indicates that a star’s ranking 12 months ago might not tell you much about how she’ll play now. Seedings exist partly to induce top players to compete, but also to increase the likelihood that the best women will face each other in the final rounds. By the latter criterion, it’s not clear that Serena (or any returning player) should immediately reclaim a top seed.

If the WTA does stick with this basic principle, I would suggest offering a few more SR entries–perhaps 12 instead of 8, and 3 slams instead of 2. Maternity leave necessitates more time on the sidelines than the six-month injury break required to qualify for the SR rule, and it may require still more time to return to form. The WTA might also convince the ITF to offer an additional few SR entries to lower-level events. Kei Nishikori came back from injury by playing a couple of Challengers; women might prefer to get their feet wet with a few ITF $100Ks before using their SR entries on top-tier events.

2. Link seeding to Special Rankings. The second option is essentially what fans wanted when they realized Serena might not make it to the Miami second round. Instead of using current ranking to determine seeding, tournaments could use SR for players who used them to enter the event.

There is a precedent for this: Monica Seles was given a top seeding when she returned from injuries sustained during her 1993 on-court stabbing. More than two years later, she came back as the top seed in Canada and the second seed at the 1995 US Open, where she lived up to that draw placement, winning 11 matches in a row before falling to Steffi Graf in the New York final.

The pros and cons of this route are the opposite of the first proposal. Giving players their pre-break seeding would show respect for their accomplishments, but since most players don’t come back from any length time off court the way Seles did, it’s possible the seedings would appear overly optimistic. (And yes, I realize the irony of saying so during the 2018 Miami tournament, when the top two seeded women won only one match between them.)

3. Devise a time-off-court algorithm. Players usually need some time to resume their former level, but their skill upon return has some relationship to how they played before. When I wrote about Sharapova’s return from her drugs ban last year, I showed that elite players who missed a year or more (for whatever reason), tended to play much worse than their pre-break level for their first five or so matches, and then a moderately lower level for the next 50. I measured it in Elo points: a 200 point drop at first, then a 100 point drop.

I don’t expect the WTA to adopt Elo anytime soon, but an algorithm of this sort could be based on any ranking system, and it represents a reasonable compromise between the first two positions. For someone as dominant as Serena, it would fulfill most of her fans’ wishes: A 200-point drop from her pre-break level would still leave her roughly even with Halep, meaning that a system of this sort would’ve made her the first or second seed in this month’s draws.

A better illustration of how the algorithm would work requires a player who didn’t so overwhelmingly outclass the rest of the field: If Wozniacki (current Elo: 2156) were to miss the next year, her seeding upon return would use an Elo 200 points lower, of 1956, dropping her to about 30th (assuming all the top players were competing). After the first five matches, when players usually start getting their groove back, her seedings would rise to around 15th. Several months in, her ranking would rise, and her seeding would no longer need to be adjusted.

The obvious flaw here is the level of complexity. My algorithm is approximate at best and would need to be improved for such an important role. The advantage, though, is that if an acceptable formula could be found, it would allow the WTA to offer a perfect compromise between the needs of returning mothers and the rights of the rest of the field.

And about those wild cards… 

I’ve mentioned that Serena used wild cards to enter both Indian Wells and Miami, even though she could have used her Special Ranking. Just about every WTA event would happily hand her a wild card, as they should. So in Serena’s case, the SR rule is largely irrelevant–if it didn’t exist, she could immediately resume a full schedule.

I also wrote that, as a fan, what matters to me is that all tour players are treated equally. Tournament entries are opportunities to gain ranking points, which in turn determine entries and seeding, which affects the likelihood of racking up wins and titles. Wild cards are often thought of as gifts, but we rarely acknowledge the effect that those gifts have on the players who rarely get them. Because tournaments understandably tend to hand out free passes to home-country players (like Donald Young) and marquee personalities (like Eugenie Bouchard), the wild card system introduces systemic bias into rankings and results. Wild cards can’t make a journeyman into a superstar, but they can boost a player from the top 200 to the top 100, or from No. 70 to No. 50. For some tour players, these differences really matter.

Thus, when a superstar or media darling–or just a player from a country that happens to host a lot of tournaments, like the United States–returns from maternity leave, injury, or a suspension, the regular rules don’t apply. Maria Sharapova was wild carded into most of the tournaments she wanted to play last year, while Sara Errani has spent the last six months playing ITFs, $125Ks, and qualifying. Sharapova gets to play matches with 100 ranking points at stake while Errani contests entire tournaments with less on the line.

Wildly different as their cases are, Serena’s situation with regard to wild cards is the same as Sharapova’s. Her allotment of SR entries doesn’t matter. But imagine if, say, Anastasija Sevastova or Magdalena Rybarikova took time off to have a child. They might get a few free entries into European international-level events, or maybe a wild card into a tournament they’ve previously won. But for the most part, a Sevastova or a Rybarikova–despite taking her hypothetical absence while a top-20 player–would be jealously protecting her eight SRs. She would need them.

Just to be clear, I’m not trying to say that Serena doesn’t “deserve” all the wild cards she’s going to get. Her achievements make it obvious that she does. On a tour where events can award draw places at their discretion, no one deserves them more. However, the very existence of those discretionary spots means that maternity leave means something very different for Serena than it would for the more anonymous players near the top of the WTA rankings.

How about this proposal, then: For players coming back from maternity leave, expand the number of SR entries from 8 to 12, and tack on another four free entries to ITFs, so that returning players can have a child knowing that they’ll be able to compete at the top level for nearly a season once they come back. But–they may accept no wild cards during that time. If they take a wild card, they lose their SRs. That proposal would put all players on an even keel: Close to a year of tournament entries at their pre-break ranking. It would give the next Serena-level superstar plenty of time to regain her lost status, and best of all, it would do the same for her lesser-known peers.

Podcast Episode 21: Talking About Talking About Tactics

Episode 21 of the Tennis Abstract Podcast, with Carl Bialik of the Thirty Love podcast, is our attempt to reconcile discussions of tennis tactics with those of tennis analytics. We start with our interest in the talented and crafty Alex De Minaur, then cover the oft-ignored tactical skills of Jo Wilfried Tsonga, the usefulness of tactical tips in the amateur game, and the constraints that coaches face when turning their observations into insights that players can use.

As a bonus, we also touch on this year’s new incentive to play doubles at Indian Wells, the crazy new proposal to overhaul Davis Cup, and the recent spate of tennis movies, especially Battle of the Sexes. In keeping with Jeff’s recent return to the U.S., it’s a super-sized episode, clocking in at just under 95 minutes.

Finally, we hope our audio quality is continuing to improve. We’re both now using proper mics, and the difference is, well, audible. Thanks for listening!

Click to listen, subscribe on iTunes, or use our feed to get updates on your favorite podcast software.

Update: Thank you to FBITennis for outlining this episode, with timestamps:

Alex De Minaur 1:38
Tsonga Decisionmaking 8:10
Pesky/Smart Players 12:30
What are “tactics”? How to analyze? 15:00
Testing effectiveness of shot patterns 20:43
DIY Tactics 27:00
Pros and Coaching Advice 34:40
Which analytics might be useful to pros? 39:58
Indian Wells Singles/Doubles Bonus 1:02:00
Reformatting Davis Cup 1:05:19
Tennis Movies 1:18:00

ATP Streaks of 2017

Italian translation at settesei.it

This is a guest post by Peter Wetz.

With the recent update of Jeff’s ATP and WTA GitHub repositories, we can take a look at notable streaks that happened in 2017. In this post I show matches won/lost and tiebreaks won/lost streaks of the 2017 ATP tour.

Let’s start with matches won:

Name               Start   End     Length
Rafael Nadal       04-17   05-15   17
Rafael Nadal       08-28   10-09   16
Roger Federer      06-19   08-07   16
Roger Federer      10-09   11-13   13
Roger Federer      03-06   03-20   12
Alexander Zverev   07-31   08-07   10
Rafael Nadal       05-29   07-03   10
Stan Wawrinka      05-22   05-29   10
Grigor Dimitrov    01-02   01-16   10

We see that, as far as streaks are concerned, the 2017 season was dominated by Roger Federer and Rafael Nadal. Rafa’s streak of 17 wins, which was halted by Dominic Thiem in the Rome quarterfinal, is the only streak containing three back-to-back tournament wins. Besides Roger and Rafa, only Alexander Zverev won two tournaments back-to-back.

When we talk about the less glamorous category of losing streaks, two names immediately should pop into our minds: Vincent Spadea and Donald Young. The former holds the record of 21 consecutive matches lost* on ATP level, and the latter holds one of the longer losing streaks (17 matches lost in a row) in recent years.

During the 2017 ATP season no player came close to any of these marks, but still there were a few moments where players seemed to have forgotten how to win a match. The following list shows all players with 8 or more consecutive matches lost.

Name                  Start   End     Length
Pablo Cuevas          05-29   10-23   10
Maximilian Marterer   02-06   08-28   10
Paolo Lorenzi         08-28   10-30    8 
Malek Jaziri          03-20   07-03    8 
Daniil Medvedev       07-31   10-09    8 
Stefanos Tsitsipas    02-13   10-02    8 

Regarding Maximilian Marterer‘s streak of 10 matches lost, we have to mention that he played a good season at the Challenger level. In between his losses at the ATP level there were deep runs at various Challenger tournaments. Still, it must be frustrating to lose your first round main draw match every time after having successfully gone through qualies. This fact accounts for 7 of his 10 losses at ATP main draws (the other 3 coming from entries as a wild card). Pablo Cuevas, the other player having lost 10 matches in a row last season, on the other hand, achieved a real losing streak with no Challenger level wins hidden among them.

Winning tiebreaks has been discussed on this blog a lot. One of the conclusions was that in the past three players–Roger Federer, Rafael Nadal, and John Isner–consistently outperform their tiebreak expectations. The list of consecutive tiebreaks won in 2017 supports this statement as can be seen in the following table.

Name           Start   End     Length
John Isner     05-15   05-29   11   
Roger Federer  06-19   08-07   8 
Roger Federer  03-06   03-20   8 
(Many tied)                    7

John Isner’s streak went over the course of 8 matches including 2 matches he lost, whereas Roger won all the matches in which he won the tiebreaks contributing to his streaks.

The list of consecutive tiebreaks lost looks as follows.

Name            Start   End     Length
Lucas Pouille   07-03   10-09   12 
Florian Mayer   01-02   07-03   11
Dusan Lajovic   03-06   07-24   8

Lucas Pouille holds the crown for most tiebreaks lost in a row in 2017. In fact he got really close to Robin Haase‘s infamous run of 13 tiebreaks lost.

Finally, I want to present an odd 2017 achievement by Nick Kyrgios: He is the only player ever to lose three matches in a row by retirement.

Date   Tourney   Matchup                           Result
07-31  W'ington  Nick Kyrgios vs Tennys Sandgren   3-6 0-3 RET
07-03  Wimby     Nick Kyrgios vs Pierre H Herbert  3-6 4-6 RET
06-19  London    Nick Kyrgios vs Donald Young      6-7(3) 0-0 RET

The list of 2018 streaks is shaping up nicely already: Doubles partners Oliver Marach and Mate Pavic opened their season with 17 straight wins, including three titles, finally ending with a loss in the Rotterdam final to Pierre-Hugues Herbert and Nicolas Mahut. Marach can even claim an 18-match streak, since he won a Davis Cup match for Austria last month while pairing with Philipp Oswald. Doubles data is tougher to come by, but it’s safe to say that the season-opening run for Marach/Pavic will have a prominent place in any summaries of this year’s ATP streaks.

 

* The list excludes one loss of Vincent Spadea at the 1999 World Team Cup in Düsseldorf, where he lost to Rainer Schüttler 5-7, 6-3, 1-6.

Peter Wetz is a computer scientist interested in racket sports and data analytics based in Vienna, Austria.

Trivia: Deja Vu All Over Again

Italian translation at settesei.it

In the last several days, Fernando Verdasco has seen a little too much of Diego Schwartzman. On Sunday in Rio de Janeiro, the two players met in the final of the 500-level clay court event, which Schwartzman won in straight sets. Both players immediately headed for the hard court tournament in Acapulco, where they drew each other in the first round. Verdasco lost again, this time winning six games instead of five.

The odds of this sort of final-to-first-round scenario, with back-to-back matches against the same opponent, is quite rare, and the surface switch makes this one even more unlikely. For one thing, the tour doesn’t move from one court type to another very frequently, and when they do, players don’t always travel through the same sequence of events. Another cause of improbability is that a pair of players who contest a final are usually pretty good, meaning that both of them are often seeded at their next event, making a first-round meeting impossible. In order to see a pair of consecutive matches like Schwartzman’s and Verdasco’s, we require synchronized schedules and a hefty helping of luck.

As Carl Bialik pointed out, this isn’t the first time Verdasco has played back-to-back matches in February against the same opponent, albeit on the same surface: He did so in 2011, dropping the San Jose final and then a Memphis first-rounder to Milos Raonic. Remarkably, when we broaden the search a bit, Verdasco’s name comes up twice more. In 2009, he lost to Radek Stepanek in the Brisbane final, then in his next event, the Australian Open, he beat Stepanek in the third round. (Radek played Sydney in the meantime, for what it’s worth.) And five years later, Verdasco overcame Nicolas Almagro to win the 2014 Houston title, then faced his countryman in his next event two weeks later, losing to Almagro in the round of 16. (Again, while they were back-to-back tourneys for Verdasco, Nico squeezed in a few matches in Monte Carlo in between.)

Back to the matter at hand: In the course of five decades of Open Era men’s tennis, just about everything has happened at least once before. But this exact scenario–two guys facing each other in a final, then a first round match the very next week on a different surface–is a new one. Relax any one of those constraints, and we see a few instances in the past.

Since 1970, there have been about 3,750 tour-level finals. Roughly one-third of the time, the two finalists ended up playing each other at least once more over the course of the season. 197 of those pairs drew each other in their very next event, and in another 62 of the finals, one of the players faced the other in his next tournament (though the other had played an event or two in the meantime, like Almagro and Stepanek). Several of the 197 duos played each other the next week, though it is a bit more common that there was a week off in between.

Of the 197 finalist pairs, 25 of them drew each other in the round of 32 or earlier in their following tournament, though not all of those were first-round matches. (Or, in the case of Andy Murray and Philipp Kohlschreiber in 2015 after contesting the Munich final, they played in Murray’s first Madrid match the following week but not Kohlschreiber’s, since Murray had a bye.) The most common round in which finalists met again was another final, which ensued about one-third of the time.

Dividing up the 197 pairs a different way, about one-fifth (39) played the follow-up match on a different surface. In only a few of these instances were the two surfaces hard and clay; a disproportionate number of these back-to-back matches happened in the 1970s and early 1980s, when carpet was regular feature on tour, so the hard-to-carpet or carpet-to-hard transition shows up in these results much more frequently than hard-to-clay or clay-to-hard. For any pair of surfaces in these 39 matches, only three occured in the round of 32, and none in the round of 64 or 128.

The three precedents for Schwartzman’s back-to-back wins all have several things in common. First, like Diego’s feat, the same player won both matches. The other two are unlike the Schwartzman double: In each case, there was a one-week break between the tournaments and one of the events was played on carpet.

The first similar achievement was recorded by Tom Gorman, who won consecutive matches against Bob Carmichael in 1976. The first was the Sacramento final (on carpet), followed by the first round in Las Vegas (on hard). Next up was Martin Jaite‘s pair of wins over Javier Sanchez in 1989. After triumphing in the Sao Paulo final (on carpet), Jaite won a hard-court first-rounder against the same opponent two weeks later. Finally, Fernando Gonzalez defeated Jose Acususo twice in a row in 2002, first in the clay-court final in Palermo, then a bit more than a week later on carpet in the first round in Lyon.

Like Schwartzman and his three closest predecessors, most of the finalists managed to defend their victory. Of the different-surface instances, the same player won both matches 26 of 39 times. When the two matches took place on the same surface, the title winner won the next match 101 of 158 times. Most recently, Yuichi Sugita failed to do so: After beating Adrian Mannarino for his first tour-level title in Antalya last summer, he met the Frenchman again in the Wimbledon second round and lost. In a more notable exception, Andre Agassi knocked out Petr Korda for the 1991 Washington title, then lost to Korda in his first match the next week in Montreal. (It wasn’t Korda’s first match, as he didn’t get a bye like Agassi did, but the extra effort paid off. The Czech reached the final.)

We could wait fifty years for an exact parallel of Schwartzman’s feat. Or we could set the bar a little lower and see a rematch almost immediately: Another of last week’s finalist pairs, Lucas Pouille and Karen Khachanov, followed up their Marseille title match with another meeting in the Dubai second round only three days later. Regardless of which standard you choose, there’s one person who would surely prefer to take a break from consecutive matches against the same opponent, and that’s Fernando Verdasco.