When Oakland A's owner Lew Wolff exercised his option to bring an MLS franchise back to San Jose, the once scorned fan base celebrated like the French when Lindbergh landed. But after the din subsided, the news that A's vice president and general manager Billy Beane would have some involvement with the revived franchise began to trickle out, leading to the inevitable question: Will Beane bring his "Moneyball" approach to MLS? The answer is an unequivocal "yes," and far from being the stuff of science fiction, in some ways the strategy is already in place.
"If you get the data, what you're trying to ultimately achieve is: What is the intrinsic value of this player's performance?" Beane said. "What is he paid, and what should he be paid?"
The A's analysis saw statistics like on-base percentage plus slugging percentage (OPS) take on increased importance. From a tactical standpoint, it meant relying on the team's big bats to produce runs and eschewing such well entrenched strategies like the hit-and-run and the sacrifice bunt. The end result was that the A's won a ton of games relative to their payroll, reaching the playoffs five times from 2000-06. The team slumped a bit in 2007, finishing under .500 for the first time since 1998, but that hasn't dimmed the enthusiasm of Beane, who in describing the methodology, sounds more like a financial analyst than a sports executive.
"I think the misconception about any statistical analysis is that you're not going to be 100 percent correct," Beane said. "What you're trying to do is create an arbitrage ... if you're right 25 percent versus 20 percent you've created a 5 percent arbitrage opportunity. That's really all you're trying to do."
But when it comes to soccer, such a strategy is up against what can be described as the Power of Too; as in too many players, in too many leagues, at too many levels in a game that has too many intangibles to easily render itself to statistics. But Beane contends that when it comes to soccer, such analysis is not uncharted territory.
"Let's face it, every business has metrics that can be used," Beane said. "It's just identifying the metrics that have the greatest weight and the greatest correlation to ultimately winning. It's a work in progress, and I say this with a tremendous amount of respect [for soccer]."
Indeed. On a recent trip to Europe, Quakes' GM John Doyle visited several British clubs, and was impressed by the precision and volume of data that was collected and analyzed. Closer to home, every MLS team has contracted with a company called Match Analysis to receive statistical and video breakdowns of every game. According to the company's president, Mark Brunkhart, the product has been used by some coaches and players to analyze their own week-to-week performance, as well as that of the opposition.
But the potential is there for additional uses as well. And in terms of player evaluation, the data, which records every touch a player makes in a game, reveals some interesting numbers. Although typical stats like a player's possession percentage are tracked, there is also one called shot creation, which records how many times a player was involved in an attack that led to a shot. (The runaway leader? David Beckham with over 11 shots created per 90 minutes. Among full-time players, the highest mark belonged to D.C. United's Christian Gomez at around 7.4 shots created per 90 minutes.)
That's not to say that there isn't room for improvement. As Houston head coach Dominic Kinnear put it, "You can win a header, and if it doesn't fall to your team, that counts as a loss of possession, so it's not 100 percent accurate. It's kind of a gray area."
The stats also don't take into account the relative risk or benefit of a particular event. Would Diego Maradona's epic slalom around five English players to score in the 1986 World Cup count as much as a tap-in? How do you measure heart, or what is happening away from the ball? Brunkhart contends, quite logically, that the analysis should be as objective as possible, and conclusions should only be drawn from a large body of data in combination with visual evidence. All of which points out that although soccer is still relatively immature in terms of its statistical analysis, it is further along than people think.
As for the Quakes, Beane insists that when it comes to his interactions with Doyle and head coach Frank Yallop, "I try not to get in their way." Mention the word "Moneyball" to Doyle and Yallop, and the urge to invoke their Fifth Amendment rights suddenly takes over. But one only has to look at San Jose's approach during the expansion draft to realize that their strategy bore some similarities to "Moneyball." Both Yallop and Doyle were holed up at Match Analysis' headquarters in Emeryville, Calif., late on the night before the draft, and used video and statistics to weigh the relative merits of the available players.
"You have a feeling about a player, and then you see how those feelings were backed up by data," Doyle said. "I think it was very useful to us to have that at our fingertips."
Yallop added that in the case of midfielder Ivan Guerrero, the information supplied played a part in his selection.
"We looked at [Guerrero's] stats and saw that he's was one of the best in possession in the league," said Yallop during the team's expansion draft announcement. "Getting the ball and keeping the ball is important, and Ivan does that."
The fact that San Jose's 10 picks made barely a dent in the team's salary cap (their combined salaries were around $712,000) would also indicate that a version of Beane's strategy has been implemented. Of course, it could be argued that given the paltry cap of around $2 million that's been in place since the league's inception, MLS teams have always been forced to employ a form of "Moneyball."
The Quakes' method is evolving through the use of more sophisticated statistics combined with video to confirm a team's findings. Of course the key for any analytical tool is for it to expand, develop, and be tested over time. A lack of data means the strategy won't work on a 17-year-old club player, nor will it be effective on an overseas performer not covered by the various data collection companies. Then there is the difference in playing styles from league to league that impact analysis.
To that end, Beane confirmed that he is working with Leeds University Business School professor Bill Gerrard in the hope of developing a proprietary system for evaluating soccer players, as well as looking to acquire additional sources of data.
"You don't create the template and then that's it," Beane added. "It's constantly evolving and changing as you try to ferret out things that can be applied. Statistical analysis is still very much fluid."
And when it comes to soccer, it now looks like it's here to stay.
Jeff Carlisle covers MLS and the U.S. national team for ESPNsoccernet. He can be reached at firstname.lastname@example.org.