Wimbledon: what the data says about some tennis myths

Wimbledon: what the data says about some tennis myths

Wimbledon: what the data says about some tennis myths

Novak Djokovic serves
Do new balls help the player serving? And are you more likely to serve an ace at gamepoint?
Location: All England Club dates: June 27-10 July
Coverage: Live on BBC TV, radio and online with extensive coverage on BBC iPlayer, Red Button, Connected TVs and mobile app.

During the first set of the first match at Wimbledon next week, the commentator will invariably mention that the server will use the ‘new balls’.

After the first seven games, and then every nine games after that, the old balls are removed and replaced with new ones. The popular opinion is that the new, firmer balls give an immediate advantage to the server, something that diminishes game by game until they are replaced.

But is this true?

Dutch professors Franc Klaassen and Jan R. Magnus decided to find out if this was the case and set to work investigating as many tennis ‘myths’ as possible.

“The list [of myths] was created by watching tennis on TV and listening to commentators,” explains Professor Klaassen. “And we combined that with our own extensive experience as a tennis player.”

What followed was 25 years of research, academic papers and a book.

The first work included data from 86,298 points played at Wimbledon and expanded to over 100,000 points played at other Grand Slam events, with data collected from the 1990s to the present day.

So what did they find and will we look at the match in a different way?

The Myth: All Points Are Equal

One of the fundamental questions arising from research from the 1970s was whether or not points are independent of each other.

Does the previous point have any influence on the point currently being played?

The findings were clear and they confirmed that not all points are equal, although the difference between them is small.

This information was accompanied by the idea that an ace could be worth more than one point. The prevailing wisdom was that serving an ace gave the server so much more confidence that they were more likely to win the next run.

The data showed that in men’s singles there was an effect and after an ace the server was more likely to win the next point to the beat of the ace worth 1.04 points, and in women’s singles the effect was worth 1.01 points.

These margins are small, but sometimes that’s enough to change the outcome of a point and thus a game and ultimately a game.

The myth: the seventh game of the set is the most important

If the points are not even, what about games?

The professors knew that legendary BBC commentator Dan Maskell would often say that the seventh game of a set was the most important, and that whoever won it had the best chance of winning the set.

That turned out not to be true, and that the most important matches are the numbers 11, 12 and 13 (the tiebreak) when the set goes that far.

The seventh game could come with the scores at 5-1, 4-2, 3-3, 2-4 or 1-5. You might think that at 3-3 the seventh game is important, but data shows this is not the case. In fact, the seventh game has less impact on the set than the sixth.

The professors devised an “importance” measure to rank each game from 1-100. Game six scored 43.0 in men’s singles and 41.3 in women, while game seven scored 37.6 in men and 36.6 in women.

The “interest” was the probability of the server winning the set in that game given that they win the game, minus the probability of the server winning the set given that they lose the game.

The myth: players are likely to lose their serve once they break their opponents

Klaassen and Magnus analyzed what is happening at the ‘big points’ and chose breakpoints and gamepoints to consider.

They found that in both the men’s and women’s singles, the number of aces increased at gamepoint, but so did the number of double fouls.

The result was that the total number of points won on service at gamepoint was slightly lower than on all other points. Male servers won 62.2% of the points at the breakpoint compared to 65% of all points, while women gained 53.5% at the breakpoint compared to 56.4% of all points. †

When the match point is a match point, or even a championship point, the stakes are markedly higher.

The elevated aces and double fouls at this point were never more tense than at the close of the 2001 final between Goran Ivanisevic and Pat Rafter. The big serving Croat led 8-7 in the fifth set, but served two aces and three double errors at crucial moments before finally taking the crown in his fourth final.

At the breaking point, players serve fewer aces, pay more attention to getting the ball into play, and play more carefully.

So does a player who wins a service break relax and the player who just broke tries extra hard to make sure there is a breakback right away?

The answer was no. In general, a player is more likely to keep their own serve after making a break.

Men have an 81.8% chance of holding the serve, which rises to 83.3% after a break. For women, this is normally 64.3% and rises to 68.3% after a break.

Men Women
Aces on point of play 10.9% 4.2%
Aces on all points 9.1% 3.3%
Double fouls on match point 5.7% 5.3%
Double faults on all counts 5.5% 5%

The myth: whoever wins the fourth set has the best chance of winning the fifth

Having previously looked at the effect of one point on the next, they looked at the effect of winning one game on the next.

Are players getting momentum, or as they called it, a ‘winning mood’?

This was the most surprising finding.

“The winning vote is much smaller than we thought,” explains Professor Klaassen. “Yes, there is momentum. Doing well at recent points indicates that the player will also do well at current points. But this momentum is small.”

How would this translate into the idea that a player who wins a set to even win a match and needs a deciding fifth (men’s) or third (women’s) set would have the advantage?

The data was split into seeds and non-seeds and gave clear results.

When playing with two seeds, the winner of the fourth set will most likely lose the fifth (an 11.1% chance that the same player will win sets four and five).

But when two non-seeds are playing, the opposite is the case, as the winner of the fourth is likely to win the fifth as well (a 60% chance that the same player will win sets four and five).

When a seed plays a non-series number, the winner of the fourth set is irrelevant, as the seeded player has a greater chance of winning the last set regardless of what happened before (an 80% chance that the seeded sets will be four and five win).

These numbers are the same for women entering a third set.

Looking at the 20 men’s Wimbledon finals since the Millennium, six have gone the distance to five sets. In each set, the player who won the fourth set to tie then lost the fifth.

Roger Federer has won this way twice (2007 v Nadal and 2009 v Roddick), but three times he lost (2008 v Nadal and 2014 and 2019 v Djokovic).

The myth: it is an advantage to serve with new balls

The data helps explain a lot about what’s happening in tennis, but what about the new balls?

Klaassen and Magnus found that men serve slightly more aces when using the new balls (10% with new balls, 9.1% if not), but in addition there is an increase in double fouls (5.9% with new balls, 5 .5% without ).

So the new balls are faster to produce the aces but harder to control hence the double faults.

It’s not an easy answer, there are shades of gray. However, when the commenter reminds us that the server has the new balls, we know that “something” more than usual is going to happen.

Around the BBC iPlayer bannerAround the BBC iPlayer footer