Why Tennis Needs Its “Moneyball” Revolution
Posted by Brodie under: Strategy
On the surface, baseball is a simple game. One player throws the ball, the other player tries to hit it. You advance on the bases until reaching home plate, in which a run (a point, basically) is scored. Once three of your players are out, you switch from hitting to pitching with the opposing team. Nine times, 54 outs, and the game is over.
Due to the way the game is constantly starting and stopping (it is one of the few games that has no game clock) it is relatively simple to track what goes on in a game. Presumably, it would also be easy to track the things that help teams win. Getting on base, hitting batters home to score runs, hitting home runs. The better you are at the things that help your team win, the better of a player you are.
Or are you? The entire “Moneyball” story, book and movie revolves around the fact that scouts and managers relied (and still rely) on faulty statistics. Likewise, they rely on their eyes and the appearance to judge a player on whether or not he looks like he will be successful, despite the fact that physical appearance of perceived strength has nothing to do with how good of a baseball player you are.
Finally, it depicted baseball culture as an exclusive, old boys club made up of ex-players. If you hadn’t played the game, you probably just “didn’t get it”.
For baseball, a sport in which every pitch, plate appearance, hit and out is recorded, a statistical revolution made sense. It simply means looking at the information already presented to you in a different, more advanced way. In reality, this “Moneyball” revolution could have happened 40 years ago (of course, having computers to crunch more advanced numbers helps).
One would assume statistically analyzing sports such as soccer, tennis or hockey would be impossible in such fluid environments.Without going into detail, soccer is slowly proving this to be wrong. With advanced computer technology and companies like Opta able to sell their statistics to clubs and coaches alike, a statistical revolution has begun. Instead of relying on traditional ways of judging a player, or physical appearance, it is becoming possible to analyze players in ways that were never possible, such as distance run, times sprinted, pass percentage and so on.
Tennis has a history of statistics. Aces, first serve percentage in, winners and unforced errors. All give an indication of how a player played over the course of a match. However, tennis fans know that these always need a grain or two of salt. Tennis is not a sport of winning more points than the opponent, it is a sport of holding serve, and winning sets.
The invention of Hawkeye, and the excellent work by the people behind the scenes in tennis broadcasts are changing the way we consume the game on TV. The commentators, production truck, and Hawkeye guys are all connected by headset, and when, for example, the commentator says “wow, Djokovic has been having a lot of success on his backhand today” and a tidy graphic pops up showing that Djokovic has 20 backhand winners to Monaco’s 2, there is a very good chance that he was tipped off by someone in the production truck, or the Hawkeye truck. It means there are people tracking the stats for us, and keeping informed as the match progresses. It’s a start.
All of this is well and fascinating, but what does it all mean? I would argue, very little. My favourite example is the “% of points won at net”. Commentator A says to commentator B, “gee, Bob, Djokovic is having a lot of success at net, he’s 13 for 15. Do you think he should be trying to get to the net more?”
Outside of the nearly defunct serving and volleying, net play is not just a sudden choice by a player. Net play is usually created by playing a well constructed point in where the player gets to the net to cut off an angle to end a point they are dictating. “Hey, better rush the net now.” That is not how modern tennis works.
In other words, tennis is not just about how many aces, winners, or net points you can win. It is about what patterns of play prove effective for one player against another. Despite having very little statistical knowledge on the actual rallies themselves, tennis people still commonly refer to “high percentage” and “low percentage” shots, when really there is little percentage attached to these. It typically means a more difficult shot to take, or a shot that is at a lower or higher point of the net.
However, if it is a shot that you can only make 70% of the time, but win the rally 90% of the time while doing it, is it really a low percentage shot? This is a basic example. Every player has patterns of play that they prefer. This means that if you are able to make shots outside of your opponents preferred way of playing, you have a much higher chance of winning.
To take a real life example, in Toronto, I knew Mardy Fish was likely to get the better of Juan Monaco. Monaco struggles with hitting backhands down the line. He would much rather play the “high percentage” shot and hit it over the lower part of the net cross court to the opponent. However, this means he is hitting it back to Fish’s backhand, as both players are right handed. Fish is able to hit a down the line backhand with considerable effectiveness, which gets Monaco on the run. On a slower court, Monaco might be able to track the shot down, and grind the point out. However, on faster courts, he is immediately pushed into a defensive, running position. Tennis is all about creating and using space. Monaco tracking down a down the line Fish backhand leaves the rest of the court wide open. Even if Monaco tracks the ball down, Fish is on top of the rally and more likely to win it. If Fish’s shot is particularly good, he may even come into the net where he can cut off the angle and win the point immediately. Again, Fish’s percentage at the net might look quite good. This is due to the pattern of play that precedes a net rush.
While these tactics and observations could be made by the naked eye, the real effectiveness would be with the aid of Hawkeye. Baseball has had a similar revolution with the aid of technology in that each pitch can be measured in the zone in which it is thrown by the pitcher, as well as where it is hit by the hitter. With an average of four at bats over 162 games, and then over an entire career, it is very easy to figure out where certain batters prefer the ball to be thrown to them.
While it is interesting to study how far back Federer is when he returns a serve with a fancy graphic, this is entirely dependent on how well his opponent can serve, and the surface they are playing on. What would be more interesting would be, for example, the placement of the return of every second serve by Murray’s opponents. Ever. Not only can we judge the placement of Murray’s serves, but we can judge the speed and placement of the return. If we use statistics from every Murray match over the past three years, we would likely find some incredibly relevant statistics. For example, for every second serve from Murray to the deuce court, within a foot of the T, of all points won by the opponent, 60% of them were returned to Murray’s backhand, and only 40% to his forehand. Similarly, if you knew that of these points hit to Murray’s backhand, Murray then returned 80% of those shots cross court, you have an incredible tactical edge before the point has started.
In other words, we as an audience can gain some interesting knowledge over the course of the match, but the statistics are within the bubble of that match. A statistical look at patterns of play, even as simple as the example of the placement of the return of serve, over the course of several years that is player specific gives us much better insight into how a player plays and how they win and lose points, on average.
Finally, tennis has another problem very similar to baseball (and many other sports). The on court action, even the commentators, are run nearly entirely by ex-players. While baseball scouts may judge players on how they look instead of how statistically good they are, tennis players and coaches tend to be more interested in how the ball is hit instead of where the ball is hit. Of course, at lower and junior levels, this is very important. Even at the higher levels, errors are going to happen, and there will be good days and bad days where statistics don’t matter. But should players like Federer and Nadal not be as concerned about where they hit the ball as much as how they hit it, if not more? Players do get winners off certain wings because that is one of their best shots, but it tends to be more about where they are hitting the ball before the winner than how hard they are hitting the ball.
The most common response I get when telling people I am a big tennis fan and blogger is typically “oh, do you play?”. Perhaps it is because it is not a very popular sport in Canada. Regardless, the chance of being asked the same question if you were proclaim to be a hockey fan would be next to zero. Perhaps, that is why these types of changes take very long to come about. Sports are largely run players and ex-players, as well as businessmen (a point made by the whole “Moneyball” story). They are more concerned about the narrative and the physical part of actually doing the sport. They’re meant to play the game, not study it. Similarly, fans are interested in viewing, hopefully, a great piece of sports theatre, not a game of chess. It is meant to be human and random (or, for your favourite team, not random at all).
Tennis, baseball, and all other sports will always be human. Players have good days, players have bad days. Of this there is no doubt. However, that is not to say that with a keen eye, certain patterns can be noticed that weren’t there before, particularly over the average of large amounts of data. Much like baseball, for tennis, these observations are probably going to have to come from the nerds on the outside, and pushed through to players and ex-players who may be defensive to advanced statistics and analysis. As Italian ex-soccer manager Arrigo Sacchi once said, never a player himself, “I never realized that in order to become a jockey you have to have been born a horse first.”
Hawkeye has become an excellent tool to understand the patterns of play within any certain match. Likewise, those who work in the production trucks and at Hawkeye are doing an excellent job of enhancing the viewing of these matches and raising the level of knowledge of the average tennis fan. I would argue that if these statistics could be combined to reflect patterns of play over a large period of time, these statistics could be used by players and coaches alike to change the way we think about tactics in tennis.