Category: Groundstrokes

WTA Decisions From the Backhand Corner

Earlier this week I presented a lot of data about what happens when men face a makeable ball hit to their backhand corner. That post was itself a follow-up on a previous look at what happened when players of both genders attempted down-the-line backhands. You don’t need to read those two articles to know what’s going on in this one, but if you’re interested in the topic, you’ll probably find them worthwhile.

Decision-making in the backhand corner is one of the biggest differences between pro men and women. Let me illustrate in the nerdiest way possible, with bug reports from the code I wrote to assemble these numbers. My first stab at the code to aggregate player-by-player numbers for men failed because some men never hit a topspin backhand from the backhand corner. At least, not in any match recorded by the Match Charting Project. The offending player who generated those divide-by-zero errors was Sam Groth. In his handful of charted matches, he relied entirely on the slice, at least in those rare cases where rallies extended beyond the return of serve.

Compare with the bug that slowed me down in preparing this post. The problematic player this time was Evgeniya Rodina. In nine charted matches, she has yet to hit a forehand from the backhand corner. If your backhand is the better shot, why would you run around it? Of the nearly 200 players with five charted matches from the 2010s, Rodina is the only one with zero forehands. But she isn’t really an outlier. 23 other women hit fewer than 10 forehands in all of their charted matches, including Timea Bacsinszky, who opted for the forehand only four times in 32 matches.

Faced with a makeable ball in the backhand corner, men and women both hit a non-slice groundstroke about four-fifths of the time. But of those topspin and flat strokes, women stick with the backhand 94% of the time, compared to 82% for men.

A few WTA players seek out opportunities to run around their backhands, including Sam Stosur and Polona Hercog, both of whom hit the forehand 20% of the time they are pushed into the backhand corner. Ashleigh Barty also displays more Federer-like tactics than most of her peers, using the forehand 13% of the time. Yet most of the women with powerful forehands, like Serena Williams, have equal or better backhands, making it counter-productive to run around the shot. Serena hits a forehand only 1% of the time her opponent sends a makeable ball into her backhand corner.

Directional decisions

Backhand or forehand, let’s start by looking at which specific shot that players chose. The Match Charting Project contains shot-by-shot logs of about 2,900 women’s matches from the 2010s, including 365,000 makeable balls hit to one player’s backhand corner. (“Makeable” is defined as a ball that either came back or resulted in an unforced error.)

Here is the frequency with which players hit backhand and forehands in different directions from their backhand corner. I’ve included the ATP numbers for comparison:

BH Direction               WTA Freq  ATP Freq  
Down the line                 17.4%     17.4%  
Down the middle               35.2%     29.5%  
Cross-court                   47.3%     52.9%  
                                               
FH Direction               WTA Freq  ATP Freq  
Down the line (inside-in)     35.2%     35.1%  
Down the middle               16.2%     12.8%  
Cross-court (inside-out)      48.4%     51.8%

Once a forehand or backhand is chosen, there isn’t much difference between men and women. Women go up the middle a bit more often, which may partly be a function of using the topspin or flat backhand in defensive positions slightly more than men do. I’ve also observed that today’s top women are more likely to hit an aggressive shot down the middle than men are. The level of aggression and risk may be similar to that of a bullet aimed at a corner, but when we classify by direction, it looks a bit more conservative. That’s just a theory, however, so we’ll have to test that another day.

Point probability

Things get more interesting when we look at how these choices affect the likelihood of winning the point. On average, a woman faced with a makeable ball in her backhand corner has a 47.2% chance of winning the point. (For men, it’s 47.7%.) The serve has some effect on the potency those shots toward the backhand corner. If the makeable ball was a service return–presumably weaker than the average groundstroke–the probability of winning the point is 48.2%. If the makeable ball is one shot later, an often-aggressive “serve-plus-one” shot, the chances of fighting back and winning the point are only 46.3%. It’s not a huge difference, but it is a reminder that the context of any given shot can affect these probabilities.

The various decisions available to players each have their own effect on the probability of winning the point, at least on average. If a woman chooses to hit a down-the-line backhand, her likelihood of winning the point increases to 53.0%. If she makes that shot, her odds rise to 68.4%.

The following table shows those probabilities for every decision. The first column of percentages, “Post-Shot,” indicates the likelihood of winning after making the decision–the 53.0% I just mentioned. The second column, “In-Play,” is the chance of winning if she makes that shot, like 68.4% for the down-the-line backhand.

Shot      Direction  Post-Shot  In-Play  
Backhand  (all)          48.5%    55.2%  
Backhand  DTL            53.0%    68.4%  
Backhand  Middle         44.6%    48.8%  
Backhand  XC             49.9%    55.8%  
                                         
Forehand  (all)          56.3%    56.1%  
Forehand  DTL (I-I)      61.4%    73.7%  
Forehand  Middle         45.7%    50.3%  
Forehand  XC (I-O)       56.2%    64.4%

The down-the-line shots are risky, so the gap between the two probabilities is a big one. There is little difference between Post-Shot and In-Play for down-the-middle shots, because they almost always go in. For the forehand probabilities, keep in mind that they are skewed by the selection of players who choose to use their forehands more often. Your mileage may vary, especially if you play like Rodina does.

Cautious recommendations

Looking at this table, you might wonder why a player would ever make certain shot selections. The likelihood of winning the point before choosing a wing or direction is 47.2%, so why go with a backhand down the middle (44.6%) when you could hit an inside-in forehand (61.4%)? It’s not the risk of missing, because that’s baked into the numbers.

One obvious reason is that it isn’t always possible to hit the most rewarding shot. Even the most aggressive men run around only about one-quarter of their backhands, suggesting that it would be impractical to hit a forehand on the remaining three-quarters of opportunities. That wipes out half of the choices I’ve listed. And even a backhand wizard such as Simona Halep can’t hit lasers down the line at will. The probabilities reflect what happened when players thought the shot was the best option available to them. Even though were occasionally wrong, this is very, very far from a randomized controlled trial in which a scientist told players to hit a down-the-line backhand no matter what the nature of the incoming shot.

Another complication is one that I’ve already mentioned: The success rates for rarer shots, like inside-in forehands, reflect how things turned out for players who chose to hit them. That is, for players who consider them to be weapons. It might be amusing to watch Monica Niculescu hit inside-out topspin forehands at every opportunity, but it almost certainly wouldn’t improve her chances of winning. You only get those rosy forehand numbers if you can hit a forehand like Stosur does.

That said, the table does drive home the point that conservative shot selection has an effect on the probability of winning points. Some women are happy sending backhand after backhand up the middle of the court, and sometimes that’s all you can do. But when more options are available, the riskier choices can be more rewarding.

Player probabilities

Let’s wrap up for today by taking a player-by-player look at these numbers. We established that the average player has a 47.2% chance of winning the point when a makeable shot is arcing toward her backhand corner. Even though Tsvetana Pironkova’s number is also 47.2%, no player is average. Here are the top 14 players–minimum ten charted matches, ranked by the probability of winning a point from that position. I’ve also included the frequency with which they hit non-slice backhands:

Player                     Post-Shot  BH Freq  
Kim Clijsters                  53.4%    77.6%  
Na Li                          53.2%    87.5%  
Camila Giorgi                  52.9%    93.8%  
Patricia Maria Tig             52.1%    66.1%  
Simona Halep                   52.1%    83.6%  
Belinda Bencic                 51.5%    91.7%  
Dominika Cibulkova             51.3%    70.1%  
Veronika Kudermetova           50.9%    73.9%  
Jessica Pegula                 50.7%    73.7%  
Su-Wei Hsieh                   50.6%    81.8%  
Dayana Yastremska              50.6%    87.6%  
Anna Karolina Schmiedlova      50.3%    87.4%  
Serena Williams                49.9%    89.2%  
Sara Errani                    49.8%    70.0%

These numbers are from the 2010s only, so they don’t encompass the entire careers of the top two players on the list, Kim Clijsters and Li Na. It is particularly impressive that they make the cut, because their charted matches are not a random sample–they heavily tilt toward high-profile clashes against top opponents. The remainder of the list is a mixed bag of elites and journeywomen, backhand bashers and crafty strategists.

Next are the players with the best chances of winning the point after hitting a forehand from the backhand corner. I’ve drawn the line at 100 charted forehands, a minimum that limits our pool to about 50 players:

Player                Post-Shot  FH Freq  
Maria Sharapova           69.0%     4.1%  
Dominika Cibulkova        65.1%    10.5%  
Ana Ivanovic              64.7%    11.1%  
Yafan Wang                64.4%     8.8%  
Rebecca Peterson          63.4%    15.2%  
Simona Halep              63.1%     6.8%  
Carla Suarez Navarro      63.0%     7.7%  
Andrea Petkovic           62.3%     5.3%  
Christina McHale          61.9%    15.2%  
Anastasija Sevastova      61.3%     4.2%  
Petra Kvitova             60.8%     4.6%  
Caroline Garcia           60.7%     7.5%  
Misaki Doi                60.5%    17.0%  
Madison Keys              59.3%     9.3%  
Elina Svitolina           59.1%     3.9%

Maria Sharapova is the Gilles Simon of the WTA. (Now there’s a sentence I never thought I’d write!) Both players usually opt for the backhand, but are extremely effective when they go for the forehand. Kudos to Sharapova for her well-judged attacks, though it could be that she’s leaving some points on the table by not running around her backhand more often.

Next

As I wrote on Thursday, we’re still just scratching the surface of what can be done with Match Charting Project data to analyze tactics such as this one. A particular area of interest is to break down backhand-corner opportunities (or chances anywhere on the court) even further. The average point probability of 47.2% surely does not hold if we look at makeable balls that started life as, say, inside-out forehands. If some players are facing more tough chances, we should view those numbers differently.

If you’ve gotten this far, you must be interested. The Match Charting Project has accumulated shot-by-shot logs of nearly 7,000 matches. It’s a huge number, but we could always use more. Many up and coming players have only a few matches charted, and many interesting matches of the past (like most of those played by Li and Clijsters!) remain unlogged. You can help, and if you like watching and analyzing tennis, you should.

Weighing Options From the Backhand Corner

A few weeks ago, I offered a “first look” at the down-the-line backhand. I offered a stack of Match Charting Project-based stats showing how often players opted to play that shot, what happened when they did, how lefties differ from righties, and which players stood out thanks to the frequency or success of their down-the-line strikes.

Like Richard Gasquet returning a serve, we need to take a step back before we can move forward. Rather than continuing to focus solely on the down-the-line backhand, let’s expand our view to all shots played from the backhand corner. The DTL backhand is only one choice among many. A player in position to go down the line has the option of a cross-court shot or a more conservative reply up the middle. She also might run around the backhand entirely, taking aim with a forehand up the line (“inside-in”), down the middle, or cross-court (“inside-out”).

Every shot is a choice, and one of the roles of analytics is to analyze the pros and cons of decisions players make. Ideally, we would even be able to identify cases in which pros make poor choices and recommend better ones. We’re still many steps away from that, at least in any kind of systematic way. But thanks to the thousands of matches with shot-by-shot data logged by the Match Charting Project, we have plenty of raw material to help us get closer.

The first choice

In 2,700 charted men’s matches from the last decade (happy new year!), I isolated about 450,000 situations in which one player had a makeable ball in his backhand corner, excluding service returns. The definition of “makeable” is inherently a bit messy. For today’s purposes, a makeable ball is one that the player managed to return or one that turned into an unforced error. With ball-tracking data, we could be more precise, but for now we need to accept this level of imprecision.

Of the 450,000 makeable backhand-corner balls, players hit (non-slice) backhands 63.7% of the time and (non-slice) forehands 14.3% of the time. The remaining 22% were divvied up among slices, dropshots, and lobs, and we’ll set those aside for another day.

Here’s how 2010s men chose to aim their backhands from the backhand corner:

  • Down the line: 17.4%
  • Down the middle: 29.5%
  • Cross-court: 52.9%

And their forehands from the same position:

  • Down the line (inside-in): 35.1%
  • Down the middle: 12.8%
  • Cross-court (inside-out): 51.8%

The inside-in percentage is a bit surprising at first, though we need to keep in mind that it’s 35% of a relatively small number, accounting for only 5% of total shots from the backhand corner. Less surprising is the much higher frequency of shots going cross-court. Not only is that a safer, higher-percentage play, it directs the ball to the opponent’s backhand (unless he’s a lefty), which is typically his weaker side.

Point probability

Shot selection is only a means to an end. More important than deploying textbook-perfect strategy is winning the point, and that’s where we’ll turn next.

The average ATPer has a 47.7% chance of winning the point when faced with a makeable ball in his backhand corner. Of course, any particular opportunity could be much better or worse than that. But again, without camera-based ball-tracking data, we can’t make more accurate estimates for specific chances. We can get some clues as to the range of probabilities by looking at how they vary at different stages of the rally. When a player has an opportunity for a “serve-plus-one” shot in the backhand corner–the third shot of the rally–his chances of winning the point are higher, at 51.1%. On the fourth shot of the rally, when pros are often still recovering from the disadvantage of returning, the chances of winning the point from that position are 45.4%. Context matters, in large part because context offers hints as to whether certain shots are better or worse than average.

So far, we have an idea of the probability of winning the point before making a choice. There are two ways of looking at the probability after choosing and hitting a shot: the odds of winning the point after hitting the shot, and the odds of winning the point after making the shot. The second number is obviously going to be better, because we simply filter out the errors. By excluding what could go wrong, it doesn’t give us the whole picture, but it does provide some useful information, showing which shots have the capacity to put opponents in the worst positions.

Here are the point probabilities for each of the shots we’re considering. For each choice, I’ve shown the probability of winning the point after hitting the shot (“Post-Shot”) and after making the shot (“In-Play”).

Shot      Direction  Post-Shot  In-Play  
Backhand  (all)          48.2%    54.2%  
Backhand  DTL            51.4%    64.6%  
Backhand  Middle         44.2%    48.2%  
Backhand  XC             49.5%    54.6%  
                                         
Forehand  (all)          55.1%    63.0%  
Forehand  DTL (I-I)      58.5%    69.0%  
Forehand  Middle         47.3%    52.0%  
Forehand  XC (I-O)       54.9%    61.9%

Forehands tend to do more to improve point-winning probability than backhands, though the down-the-middle forehand is less effective than a backhand to either corner. Again, this is context talking: A player who runs around a backhand just to hit a conservative forehand may have misjudged the angle or spin of the ball and felt forced to make a more defensive play. Still, it’s a relatively common tactic on slower clay courts (on clay, it is almost twice as common than tour average), and it may be used too often.

The most dramatic differences between the two probabilities are on the down-the-line shots. Both forehand and backhand are aggressive, high-risk shots, something reflected in the winner and unforced error rates for each. 9% of all shots from the backhand corner are winners, and another 11% are unforced errors. Of down-the-line shots, 23% are winners and 19% are unforced errors. While the choice to go down the line isn’t superior to other options, both the forehand and backhand are devastating shots when they work.

Player by player

Let’s tentatively measure “effectiveness” in terms of increasing point probability. Setting aside the complexity of context, which won’t be the same for every player, the most effective pro is the one who makes the most of a certain class of opportunities.

Here are the 10 best active players (of those with at least 20 charted matches) who do the most when faced with a makeable ball in their own backhand corner. Keep in mind that the average player has a 47.7% chance of winning the point from that position:

Player                Post-Shot  
Rafael Nadal              52.9%  
Diego Schwartzman         52.4%  
Novak Djokovic            52.3%  
Nikoloz Basilashvili      51.9%  
Andrey Rublev             51.8%  
Kei Nishikori             51.5%  
Gilles Simon              51.2%  
Pablo Cuevas              50.9%  
Alex De Minaur            50.0%  
Pablo Carreno Busta       49.6%

The Match Charting Project data might understate just how effective Rafael Nadal, Novak Djokovic, and Kei Nishikori are from their backhand corner, since a disproportionate number of their charted matches are against other top players. In any case, it is no surprise to see them here, along with such backhand warriors as Diego Schwartzman and Gilles Simon.

This list is limited to the tour regulars with at least 20 matches charted. One more name to watch out for is Thomas Fabbiano, with only 12 matches logged so far. In that limited sample, his point probability from the backhand corner is a whopping 59.2%. He isn’t quite that much of an outlier in reality, since his charted matches include contests against Ivo Karlovic, Reilly Opelka, and Sam Querrey, opponents whose ground games leave a bit to be desired. But his overall figure is so far off the charts that, even adjusting downward by a hefty margin, he appears to be one of the more dangerous players on tour from that position.

Forehands and backhands

Let’s wrap up by looking at something a bit more specific. For backhands and forehands (without separating by direction), which players are most effective after hitting that shot from the backhand corner? We’re continuing to define effectiveness as winning as many points as possible after hitting the shot. I’ll also show how often each of the players opts for their effective shot, giving us a glimpse at tactical decisions, not just tactical success.

Here are the best backhands from the backhand corner. It was supposed to be a top ten list, but I think you’ll understand why I struggled to cut it off before listing the top 16 players, roughly one-fifth of the 75 players with at least 20 charted matches:

Player                 Post-shot  BH Freq  
Diego Schwartzman          52.8%    74.0%  
Rafael Nadal               52.7%    64.7%  
Novak Djokovic             52.7%    76.1%  
Kei Nishikori              51.7%    74.0%  
Gilles Simon               51.4%    88.0%  
Andrey Rublev              51.1%    67.1%  
Pablo Carreno Busta        51.1%    75.3%  
Nikoloz Basilashvili       51.0%    75.0%  
Alexander Zverev           50.8%    75.1%  
Alex de Minaur             50.6%    74.8%  
Daniil Medvedev            50.6%    87.2%  
Juan Martin del Potro      50.3%    49.1%  
Pablo Cuevas               50.2%    60.6%  
Andy Murray                50.1%    65.0%  
Richard Gasquet            49.9%    75.8%  
Stan Wawrinka              49.8%    63.4%

The “BH Freq” column–for backhand frequency–really demonstrates the range of tactics used by different players. Gilles Simon and Daniil Medvedev opt for the topspin backhand almost every time, rarely slicing or running around the shot. At the opposite extreme, Juan Martin del Potro hits a topspin backhand less the half the time from that position. Perhaps because of his selectiveness–dealing with awkward positions by slicing–he is effective when he makes that choice.

Now the best forehands from the backhand corner:

Player                 Post-shot  FH Freq  
Gilles Simon               63.1%     6.7%  
Rafael Nadal               61.9%    16.6%  
Benoit Paire               61.9%     1.5%  
Kei Nishikori              61.2%    10.4%  
Andrey Rublev              61.0%    20.1%  
Casper Ruud                60.8%    27.1%  
Marton Fucsovics           60.5%    16.3%  
Novak Djokovic             60.0%     9.7%  
Daniil Medvedev            59.8%     3.3%  
Pablo Cuevas               58.9%    20.9%  
Sam Querrey                58.2%    15.6%  
Felix Auger Aliassime      57.7%    16.0%

This list is more of a mixed bag, in part because there are so many fewer forehands from the backhand corner. Benoit Paire’s numbers are based on a mere 21 shots. I wouldn’t take his effectiveness seriously at all, but it’s always entertaining to see evidence of his uniqueness. At the opposite end of the spectrum is Casper Ruud, who runs around his backhand more than anyone else in the charting dataset except for Jack Sock and Joao Sousa. (Neither one of which is particularly effective, though presumably they do better by avoiding their backhands than they would by hitting it.)

One name you might have expected to see on the last list is Roger Federer. He’s around the 80th percentile in the forehand category, winning 56.9% of points when hitting a forehand from the backhand corner. He’s good, but not off the charts in this category. Like Nadal and Djokovic, he might look better if these numbers were adjusted for opponent, because so many of his charted matches are against fellow elites.

Next

There’s clearly a lot more to do here, including looking at probabilities for direction-specific shots, isolating the effect of certain opponents, and trying to control for more of the factors that aren’t explicitly present in the data. Not to mention extending the same framework to other shots from other positions on court. Stay tuned.

Tramlines and Wide Groundstrokes

The NextGen Finals are played on an unusual court, in that the surface is marked only for singles matches, leaving out the “tramlines” that define the doubles alleys. Virtually all tennis events includes doubles, as well, so this is rarely an option. The ATP has skipped tramlines at season-ending events before, but at the end of the 2010s, the singles-only court is exclusive to the NextGen Finals.

One might reasonably wonder whether the unique paint job has any effect on play:

I discussed this on a recent podcast with Erik Jonsson, and we tentatively concluded that tennis pros (even young ones) with thousands of hours of playing experience shouldn’t be affected by a tweak to the appearance of the court. But why speculate when we can look at some data?

The Match Charting Project, my volunteer-driven effort to log shot-by-shot records of professional tennis matches, notes various details about errors–forced or unforced, and “type”–net, deep, wide, or wide-and-deep. MCP contributors didn’t immediately take to the NextGen Finals–before this week, the 2018 final was the only charted match out of the 6,600 matches in the dataset–but 2019 was different. We now have shot-by-shot stats for 8 of the 15 matches played in Milan last week. (Big thanks to Carrie, who took charge of Alex de Minaur’s entire run to the final.)

Quantifying wide errors

We’re interested in the frequency of wide errors, which isn’t quite as simple as it sounds. I chose to focus only groundstrokes, and I also excluded forced errors–shots on which the player might not have much control of the direction of the ball.

Here are three metrics we could use for the frequency of wide errors:

  • Wide errors per point
  • Wide errors per unforced error
  • Wide errors per “makeable” groundstroke–that is, groundstrokes that were either unforced errors or put in play

Wide errors per point is probably too crude, but it does have the advantage of simplicity. Wide errors per unforced error might have some value, telling us in what direction a player was most aggressive. The last, wide errors per makeable groundstroke, is probably the best representation of what we’re looking for, as it tells us how frequently a player tried to hit a shot and it went wide.

Here are de Minaur’s numbers for his five 2019 NextGen matches, along with his hard-court aggregates from 28 other charted matches in the last two years:

          Wide / Pt  Wide / UFE  Wide / GS  
NextGen        2.7%        1.5%      21.7%  
ATP Hard       3.0%        1.4%      21.4%

At least for Alex, the tramlines don’t seem to make much of a difference.

Let’s look at the slightly larger group of players. We have eight matches, which means 16 records of one match for a single player, including at least one for each of the eight guys who qualified for Milan. Here are the three wide-error rates for the NextGen Finals matches, along with the same players’ wide-error rates for other charted hard court matches in the last two years:

          Wide / Pt  Wide / UFE  Wide / GS  
NextGen        3.2%        1.8%      19.5%  
ATP Hard       3.2%        1.8%      23.1%

For our first two metrics, there is absolutely no effect. Tramlines or no tramlines, wide errors mark the end of 3.2% of points, and 1.8% of total unforced errors. (The 3.2% figure is per player, meaning that 6.4% of points were ended with a wide error.)

The third metric, though, is more interesting. On tour, these players make a wide error on 23.1% of their “makeable” groundstrokes. That number dropped by more than one-seventh, to 19.5%, on the tramline-free court in Milan. At the same time, the overall rate of unforced errors (not just wide errors) increased compared to the same players’ efforts on hard courts at other events.

Deep mind

I see two possible explanations for such a substantial drop. First, we don’t have much data, and maybe it’s just a fluke of a small sample. Some of the difference can be traced to Ugo Humbert, who didn’t make a single wide error in his one charted NextGen Finals match. (Humbert’s usual wide-error rates are close to average.) Without a lot more matches played on tramline-free surfaces–not to mention charts of those matches–we won’t be able to draw a firm conclusion.

Second, it could be a real effect stemming from some aspect of the conditions in Milan. The lack of tramlines really might, as Lisa puts it, “focus the mind.”

Compared to other innovations trialed at the NextGen Finals, the singles-only court gets very little press. But unlike, say, the towel rack or the shot clock, it might just have a small effect on play.