Soccer Sabremetrics 101

This year, the Columbia women’s soccer team wants to play moneyball.

Soccer may not have the statistical curiosities that changed baseball, such as DIPS, WARP, and WHIP (which could double as names for East Village clubs), but head coach Kevin McCarthy is spending his time turning the beautiful game into a numbers game.

“The old adage is that soccer is a sport where the game is the best teacher,” McCarthy said. “Most coaches are instinctual, and I count myself among them. But as I’ve learned, the more we can quantify, the more it supports my instincts. And when it doesn’t, we can correct the flaws. It just brings me to another level of instinct and understanding.”

In his 2003 book Moneyball, Michael Lewis tracked Oakland A’s General Manager Billy Beane and his team as they took a new approach to another game that, for all its number crunching, also boiled down to a coach’s gut feelings.

Things begin to change, however, when it comes to measuring soccer performance. It’s a game where discrete situations are nearly impossible to isolate. A home run is always just a home run—one pitch, one swing of the bat. One could expand the context to an entire at-bat. A goal in soccer, on the other hand, can be the culmination of several minutes of constant action involving 22 players.

Finding the right combination of measurable elements becomes a painstaking task. “We’ve stolen every model we can possibly find,” McCarthy said. “From UNC to Arsenal to Messiah College—a very good Division III program.”

Without a staff to spool through hours of game footage, creating a comprehensive statistical snapshot of a game is even harder. Nearly all the professional teams and many of the top college programs use state-of-the-art analysis software like ProZone and Gambreaker to generate customized reports. But at Columbia, funding is the barrier.

Still, McCarthy says Columbia’s method of trial-and-error has proven successful.
Besides shots, shots on goal, the win-loss percentages of 50-50 balls, and the time the ball spends in various quadrants of the pitch, the coaching staff has tailored a few important numbers to game situations. For Columbia’s offensive style of play, the time spent within six yards of the byline is key. Looking even closer at these situations, McCarthy examines how many of them led to a quality cross into the penalty area—a clear indicator of how many chances the Lions are creating.

“It’s a way of creating deliberate practice,” McCarthy said. “It’s also a way of creating feedback, which all performance-oriented people crave.”

The team has responded positively. “The players love having that knowledge,” he added. “When they can clearly see what’s expected, when I can show them numbers, it helps them and they can take ownership of it.”

When performance can be so objectively quantified, it increases accountability in the squad. But what might push one player to work even harder runs the risk of doing considerable damage to another’s morale. That’s why McCarthy knows he has to be careful in presenting such detailed analyses. “There are areas where evaluations are better done personally than publicly,” he said.

Whether these measurements will have a long-term impact on the team’s results remains to be seen. I am still of the mind that in a game like soccer, intangibles like momentum can have a far greater effect than in baseball. Some of my distinguished colleagues might add that there is no stat for heart—I wouldn’t go that far.

But McCarthy realizes that when it comes to the locker room before game time, there’s nothing like an old-fashioned, get-one-for-the-Gipper pep talk.

“Before the game we’re not going to sit and look at statistical charts,” McCarthy said. “That’s where we’re going to get the blood up and get ready to go.”