Data Analysis And Soccer 
Before Germany-Holland at Euro 2012, German data analysts crunched the numbers on their opponents and spotted a weakness: The Dutch defenders often strayed too far apart. The German national team’s secret codebook says that the ideal distance between defenders in a back four is eight meters. The Dutch were regularly leaving larger gaps. In the group game in Kharkiv, Germany located those gaps and won 2-1.
History is written by the winners, and, with hindsight, it’s to forget the Germans. One bad night against the Italians in Warsaw, two childish positioning errors by German defenders, and they lost the semifinal. Yet, going into the semis, the Germans were the only team with four straight victories. Since they began serious data analysis in 2005, they have reached at least the semifinal of four successive tournaments. Almost unnoticed by fans and media outside Germany, data analysis became more influential than ever before at Euro 2012. Statistics are starting to change international soccer.
Like almost everything on earth, modern data analysis in sports originated in California. About 20 years ago, the Oakland A’s baseball team began finding new statistics to value players. For instance, hitters had always been judged on their batting average. The A’s found that on-base percentage -- in those days, an almost unknown number -- was a far better predictor of a player’s worth. They found that time-honored methods like stealing bases and sacrifice bunts made no sense. The man forever associated with this is the A’s general manager, Billy Beane. In the recent film Moneyball, based on Michael Lewis’ book, Beane was done the honor of being played by Brad Pitt.
Some years ago, on a romantic vacation in London with his wife, Beane discovered soccer and fell for it hard. Back home in California, he became friends with Jürgen Klinsmann, a German expat living in Huntington Beach. Klinsmann had noticed that American sports employed statisticians. He spent time at the A’s learning more.
Soon afterward, Klinsmann became manager of Germany. At 1:00 a.m. one night in 2005, Professor Jürgen Buschmann of the Cologne Sporthochschule (Higher School for Sports) was woken by a phone call from the great man. Buschmann sat straight up in bed. He agreed to help Germany with data analysis. He and his team of students became known inside German soccer as Team Cologne.
The world first got wind of Team Cologne at the World Cup 2006, when the Germany-Argentina quarterfinal went to a penalty shootout. In between penalties, the German goalie Jens Lehmann consulted a scrap of paper tucked into his sock. Composed with the help of Team Cologne, the note detailed the preferred corner of each Argentine penalty taker. Lehmann stopped two kicks and Germany won. The world was astounded. Until then, everyone had assumed that penalties were a lottery.
Still, even as data analysis helped the Germans, they learned the hard way that theory is not the same as practice. Before the Germany-Spain semifinal at the World Cup 2010, Team Cologne worked out exactly where Germany should position its lines. Analyzing Spain’s 1-0 victory afterward, the analysts realized that the Germans had been seven meters too far back. No wonder: The Spaniards don’t let you do what you want.
Just two years ago, data analysis was still not very widespread in international soccer. Penalties are the most obvious piece of data to analyze, yet big teams like England and Argentina entered their knockout games in 2010 with barely any statistics on their opponents’ penalties. England’s then-manager Fabio Capello rarely used the Football Association’s data analysts. Capello preferred working with his trusted Italian comrades, and they were not data people.
But at Euro 2012, data played a bigger role for more teams than ever before. Most visibly, before the penalty shootout in England-Italy, England’s goalies’ coach briefed Joe Hart on an iPad. It looked impressively high-tech, but one wonders about the content. The chief data analyst at one leading English club told me his doubts: “Did they have previous penalties [video] to show Joe Hart? Yes, I saw them on the pitch with their iPad. Do they do analytics on them and do more than show the last three or four penalties for each player? I don't know.” Certainly Hart frequently went the wrong way and didn’t stop a single penalty. A German goalie might have been better prepared
Still, the international trend for data analysis is upward. For Spain, Paco Jimenez -- who does the same job for Real Madrid -- oversees the effort. However, anyone analyzing opponents for Spain faces a peculiar hazard. Before the semifinal against Portugal, Jimenez lamented: “I hope that what happened against Ireland, Croatia and France doesn’t happen here -- that we study a team and then they totally change their game, trying to find solutions to stop Spain’s soccer.”
One key use of data analysis nowadays is in training. The soccer authorities do not allow players to wear GPS systems during matches, but teams do use a kind of GPS in training. That allows them to track how tired their players are. Each man’s distance is covered, as well as heart rate and speed. Because of the brutal Premier League season, England usually goes into tournaments with the most exhausted team. The English began using GPS in September 2010. By Euro 2012, they had built up enough data to know how vulnerable each player was to fatigue.
The French are also big users of GPS in training. During matches, analyst Thierry Marszalek says, “We note every ball touched by the players. We have statistics in real time. That allows us to give information.” For instance, if a key creative player is receiving fewer balls than usual, you might alter the team’s tactics mid-game. If someone is sprinting slower than usual (seven meters per second is soccer’s gold standard), then he may be tired and require substitution. But however much data analysis other countries were doing, the Germans continued to lead in this department at Euro 2012. Aided by a lot of coffee, a team of several dozen students spent several thousand hours analyzing Germany’s opponents. Team Cologne supplied the German coach, Joachim Löw, with a dossier several hundred pages thick on each opposing team. And Löw’s people, meeting every day in the Dwor Oliwski hotel in Gdansk, Poland, actually used the dossiers.
These contained all manner of useful information. What are the usual running and passing routes for each opposing player? How do you know when Cristiano Ronaldo is about to do a stepover? (Before Germany-Portugal, the German right-back Jérôme Boateng received an individual briefing on the great man’s patterns.) How many seconds does each opponent take to shift to defensive positions when he loses the ball? Which player most commonly starts the opposing team’s buildup?
Cynics will ask: If Germany was so clever, why didn’t they win the tournament? But data analysis alone does not decide soccer matches. Spain had better players than Germany, and Italy was better on the night. When I put it to Beane that numbers might be less decisive in soccer than in baseball, he replied that they didn’t need to be decisive. If they gave you an edge, that was enough -- then you had to use them. He explained, “If somebody’s right 30% of the time using gut feel, and you can find a way to be right 35%, you create a 5% arbitrage, and in sports that can make the difference between winning and losing.” To Beane, the logic is that all teams will end up having to use numbers. Certainly that is the way soccer has been heading. Expect data analysis -- which in soccer lags baseball by perhaps 15 years -- to make another jump by the 2014 World Cup.
There is so much still to be done. Just one example of easy progress that can be made: free kicks. Andrea Pirlo’s goal for Italy against Croatia might have seemed quite a common moment in soccer: A beautiful free kick that curves just inside the post. In fact, it was a rarity. It was the first goal direct from a free kick at a European Championship since Marek Heinz’s for the Czech Republic against Germany eight years earlier. This year, the data provider Opta Sports tweeted after the quarterfinals: “Only four of 37 attempts direct from free kicks have hit the target so far at Euro 2012.” In short, few kicks were even producing rebounds off goalies' fists.
The numbers suggest that shooting from free kicks -- an unquestioned method in soccer, much as sacrifice bunts once were in baseball -- is a terrible waste of a good ball. A free kick ought to be an excellent attacking opportunity. You have the ball near the other team’s goal, the opponents have to stand at least nine meters away, and they have to put a few players in the wall in case you shoot.
That gives you acres of space to pass to your men in the penalty area. Yet, players almost invariably shoot from the free kick. They are chasing glory, recalling those rare but memorable balls that plop wonderfully into the top corner. Data analysts should -- and increasingly will -- show them that most free kicks go nowhere. Coaches will start forbidding their teams to shoot from free kicks. Soccer will become more rational. One day, “Team Colognes” will be the norm.