Five key ways data analytics is changing the way tennis is played

Michael Searles
Nitto ATP Finals - Day Four
World No1 Novak Djokovic is in action this week at the London ATP World Tour Finals (Source: Getty)

Data has an ever-increasing presence in tennis as coaches use new analytical tools to gain the slightest, but all-important, advantages for their players.

Leading figures from the sport discussed new trends this week at an event held by Infosys in London to coincide with the ATP Finals. Here are five key ways in which data is influencing the game.

Rally length

Rally length has been identified as a key metric for coaches as the industry moves away from an emphasis on “primitive data” about forehands, backhands and unforced errors.

In the majority of matches the winner will take more points than the loser from the much more common short rallies of 0-4 shots, but not the longer rallies of 5-8 or 9+ shots, which tend to be more evenly split between players – a trend recorded across multiple tournaments and grand slams.

“When you go to watch the game, you are going to remember that 22-shot rally, but those are the ones that don't really matter to winning and losing,” says Craig O'Shannessy, the ATP Tour's strategy analyst and one of Novak Djokovic's assistant coaches.

“By far the most common rally length in tennis is one. One shot goes in the court, sometimes it's an ace, most of the time it's a return error. I look at the short rallies because I know the data says: win the short rallies and you win the match.”

Focus on the opposition

Newly available data is driving a change in approach to matches and practice sessions. It has become as much about analysing opponents as it has improving a player's own game.

It makes more sense to focus on forcing the opponent into a mistake because rarely does a player have a perfect game.

“I think that's been a huge paradigm shift globally,” says O'Shannessy. “It's about making the opponent uncomfortable.”

Leading umpire Ali Nili agrees: “They can compare previous head-to-heads and every aspect of the game. We've come a long way from 'that was a bad day, we'll get him next time'. Now we can look at the data points and customise your practice to improve."

Era of the returner

Data gathered by the ATP and Infosys on all men's world No1s since 1991 reveals that the best returners of the ball, including the likes of Andy Murray, Novak Djokovic and Rafael Nadal, have all been part of a golden era during the last five years.

The drive behind improving the return is, of course, to win more points. By improving their average percentage of points won, those players have given themselves a statistical advantage to win more games and matches.

“[In 1991] you could become No1 in the world winning just 53 per cent of points. Now it goes higher to 55-56 per cent,” says O'Shannessy. “If you can explain to a player that on one of their best days at the office they could lose 45 per cent of the points they play, it helps stop an emotional roller-coaster.”

Cutting up the court

One method coaches are using is to cut the court up into segments, according to O'Shannessy, who says that the inside of the outer court is the most common place for balls to be hit from and to.

By analysing segments in which balls are most likely to land, coaches can establish positions that players enjoy receiving the ball and thus target areas that make it more challenging for the opponent to return from.

“I have a saying: how you hit the ball matters, where you hit it matters more. All the players here can hit the ball well,” O'Shannessy says. “We can use the data to understand our opponent and how to put a strategy together.”


Coaches are now able to craft patterns of play in order to deliver the highest number of points. Rather than just saying a player is more likely to win a point by serving out wide, the data can now go further and look at the next shot, the serve plus one.

“Nadal is about as good as it gets, hitting 78 per cent forehands on the first shot after the serve and winning 64 per cent of points when that happens,” O'Shannessy says.

The advancement in data analytics means coaches can now tailor patterns of play to specific opponents.

This doesn't mean they would use this every point though; players are able to conceal their “primary pattern” by using other patterns for the majority of the time, but when they need a point they will know the most likely method to win one.