Connect with us


Statistical poverty in tennis

Matt Zemek



Susan Mullane - USA TODAY Sports

If you know anything at all about professional basketball, and really, if you know anything at all about major team sports in the 21st century, you know this is the age of analytics.

You know that there are new ways to chart, measure, map, and define sports performance. There are new ways to put numbers to athletes’ or teams’ tendencies, inclinations, habits, strengths, and weaknesses.

Sports have always been a business, but the 21st century has become significant in the evolution of sports because applications of mathematical studies and scientific evaluation have changed how sports organizations think.

Analytics has drastically changed the way baseball is played here in the United States. Analytics is changing how various teams draft players in all of the North American team sports — baseball, basketball, football, and ice hockey (MLB, NBA, NFL, NHL).

Analytics tells sports teams which kinds of players are worth X amount of dollars, and which players are worth obtaining in a given round of a draft, and which players are worth seeking in free agency or trades as opposed to a draft out of college.

So much more information is available today to anyone who wants to study sports. The people who study sports the most are the executives who run sports teams. If you are in charge of selecting, trading and paying players for your sports franchise, which is worth at least nine figures (100 million dollars) if not 10 figures (the billions of dollars, such as the Dallas Cowboys or the New York Yankees), you HAVE to do your homework. You can’t rely on hunches.

The availability of information — certainly in the cases of the four North American team sports, and also in soccer, where (though I am largely ignorant of the finer details of the sport) there is something called a “heat map” of on-field activity — has revolutionized sports in the 21st century.

Moreover, in tennis, people such as Craig O’Shannessy are changing how tennis players get coached by bringing advanced study into scouting reports and game planning. There ARE people who study tennis itself with the depth worthy of the era of analytics.

Why, then, is tennis on television — tennis as presented to the general public — so lacking in statistical information?

Sunday’s long and complicated Wimbledon final between Novak Djokovic and Roger Federer is a great case in point.

One of many essential big-picture realities about this match — and how utterly close it was — is that 421 points were played in this match, and the two points everyone has talked about are the two championship points Federer lost in Game 16 of the fifth set.

Two points out of 421. Because of the history, the stakes, the importance, the drama, and everything else which was part of this match, two points out of 421 gained an enormous place in tennis history, changing a 21-15 major tally for Federer into a 20-16 tally which gives Djokovic a clear path to the No. 1 spot on the major titles list.

Wouldn’t it be great if any two points in a tennis match could be contextualized in ways which can make fans and analysts see how routine — or not routine — they really are?

I know for a fact that the forehand Federer missed on championship point No. 1 was a carbon copy of an inside-out forehand he had missed earlier in the match. I don’t know if he missed that shot three or four times, but I absolutely know he missed it at least one other time before 8-7, 40-15.

Wouldn’t it be great to have that kind of information on hand, such that TV could display it in the moment?

In NBA basketball, there is something called a shot chart. These have been posted for years, well before 2019:

The purpose of a shot chart in basketball is to show where teams (and individual players) make or miss shots.

For those of you who don’t follow NBA basketball very closely, NBA 3-point lines are different from FIBA (international basketball) 3-point lines in that they are not equidistant from the basket at all spots on the court.

The FIBA international 3-point line IS equidistant, but the NBA line flattens out in the corners, meaning that the distance on a corner 3-point shot is shorter than in the middle of the court, facing directly at the basket, or on the wings, when the flat line in the corner curves toward the middle of the court.

For this reason, NBA teams emphasize not only shooting more 3-point shots, in order to get three points from a possession instead of two; they also tell their best 3-point shooters to take more corner threes than midcourt threes.

Analytics tells them which kinds of shots are the best ones to take.

There is absolutely NO reason for tennis to not chart stats the same way AND provide them for public consumption on TV, such as in the Djokovic-Federer match.

There is no reason tennis CANNOT have readily available stats during matches. Unlike basketball, a continuous-action sport in which play can and does continue for three- or four-minute segments without a stoppage, tennis involves constant stoppages between each and every point played. This is a sport in which a paid team of statistical record keepers — say, six people — can follow the different statistical components of a match and complete a much fuller match profile.

I have been on record in the past as saying that tennis should offer live-match statistical charting of game points won and lost, not merely break points. Accordingly, tennis should also measure 30-30 points and deuce points won or lost by players. Those are longstanding deficiencies of in-match statistical tracking on television and modern media.

What Sunday’s Fedole match impressed upon me — getting back to the reality that two points out of 421 took on outsized historical importance — is as follows:

A) Chart forehand and backhand winners and errors by direction or shot choice. Provide winner-error splits for crosscourt and down the line as a starting point, ideally to deuce and ad courts as well. Also include drop shot winner-error totals and overheads made, missed, or incompleted (by incompleted, I mean that the shot was not an error but didn’t finish the point).

B) Every point has a specific length of shots: one, five, 10, 15, whatever. Given this reality, where are totals not just for points, but for SHOTS, broken down into games, sets, and matches?

There were 421 points in Fedole on Sunday. We should be able to get the number of shots played in the match, whatever it was. I would think that the average rally length was around 5 or 6 shots. If that was in fact the case (but I could be wrong), that means over 2,000 shots were taken in this match.

Why don’t we have a “shooting percentage” or “swing percentage,” then, in tennis? Why aren’t we able to note how many “live swings of the racquet” occur in a match like this, which puts one or two points in a fuller context?

Federer-Djokovic was a tennis feast, and yet tennis lingers in profound statistical poverty, certainly in public. Advanced stats might be kept behind the scenes by people who coach or consult players, but those advanced stats certainly don’t make their way into the public domain.

We are all poorer for it… and in 2019, it should not be the case.

Well, that’s tennis for ya.

Matt Zemek is the co-editor of Tennis With An Accent with Saqib Ali. Matt is the lead writer for the site and helps Saqib with the TWAA podcast, produced by Radio Influence at Matt has written professionally about men's and women's tennis since 2014 for multiple outlets: Comeback Media, FanRagSports, and independently at Patreon, where he maintains a tennis site. You can reach Matt by e-mail: You can find him on Twitter at @mzemek.

Advertisement Big Savings for Big Fans at