MONEYBALL MADNESS MARCH 1, 2013
In the summer of 2008, Kansas City-based sportswriter Joe Posnanski was invited to discuss Moneyball at nearby Fort Leavenworth. At first, he was incredulous. "I had literally no idea what I was supposed to talk about," he told me. But Posnanski duly attended the seminar-style session, which featured several Army officers, including a colonel, at the facility's University of Foreign Military and Cultural Studies. "They were really working with the various people on all sorts of out-of-the-box thinking—that was the general theme," he said. "They wanted to try to get into the mindset of a group of people that don't have the resources, and how they would try to overcome those things. They really were focusing on Moneyball and how the A's did it with less." As it turned out, he added, "We ended up just talking a whole lot of baseball."
We are living in the Moneyball age. Since the 2003 publication of Michael Lewis's book, which tells the story of the frugal Oakland Athletics and maverick general manager Billy Beane's use of sabermetrics (advanced baseball analytics) to make the playoffs three straight years, countless disciplines have seen disruptive, empirically minded, counterintuitive upstarts reinvent their respective landscapes. Hollywood producers are using Moneyball tactics to create hit movies. There's an Oakland Athletics of wine regions, the Loire Valley. Barack Obama's 2008 campaign had its own Beane, delegate chief Jeffrey Berman, and this past cycle you could hardly read about voter analytics without tripping over a Moneyball reference. Perhaps most famously, there is a Moneyball way of predicting political outcomes, and its pioneer, The New York Times' Nate Silver, used to work at Baseball Prospectus as a sabermetrician.1 This is to say nothing of the impact that sabermetrics, which is specific to baseball, has had on every other sport.
Counterinsurgency theory (COIN) 2 is, to oversimplify, the Moneyball approach to war. Posnanski being invited to speak at Leavenworth was one of three instances I found of Moneyball being taught at a U.S. military school over the past several years, and there almost certainly are more. At the School of Advanced Military Studies, also at Leavenworth, Professor Emeritus James J. Schneider assigned Moneyball to his class on military theory after hearing Lewis discuss the book on NPR. And Dr. Erin Simpson assigned the book to her students at the U.S. Marine Corps Base in Quantico, Virginia, and includes it in course packets prepared by her defense consulting firm, Caerus Analytics.
"There's a meta-story to Moneyball," Simpson said. Citing the 2007 troop surge in Iraq, which was in many ways COIN's high watermark in terms of both influence and success, she added, "I absolutely saw parallels to the COINdinistas in the pre-surge period and this effort of people like Billy Beane and others to say, 'Y'all are doing this wrong.'" As the 8th MIT Sloan Sports Analytics Conference begins today, it is worth appreciating that more than anything in baseball, or sports, or even numbers, the temperamental shift Simpson articulated, to a way of viewing the world that at once defies conventional wisdom yet seems totally commonsensical, is likely to be Moneyball's greatest legacy.
COIN, which has dominated high-level U.S. military thinking over the past decade, shares with sabermetrics a respect for statistical rigor and, more importantly, an openness to questioning accepted approaches and substituting counterintuitive alternatives. It is a retrenchment around basic principles and observable facts. "What we liked about Moneyball is that on the face of it, they looked at baseball and said, 'The things we believe we know are based on mistaken premises, and if we take a look at baseball differently, we will see solutions we would not see,'" said Col. Gregory Fontenot (Ret.), the director of the University of Foreign Military and Cultural Studies, who hosted Posnanski. "Counterinsurgency requires you to look at problems from perspectives you hadn't previously considered," he added. Fontenot's university is devoted to "red teaming," a practice of analyzing standard operating procedures from an outsider's perspective, and the red team who met with Posnanski was studying Moneyball for insight into the tactics of Iraqi insurgents, whose limited resources make them the A's to the U.S. Army's Yankees.
The most obvious similarities between COIN and sabermetrics are their statistical approaches. Beginning in the 1980s with the eccentric genius Bill James and continuing, in Moneyball's telling, to Beane, sabermetrics has been committed to collecting numbers from baseball's almost uniquely deep dataset, plugging them into ever more sophisticated formulas and letting those results guide decisionmaking. Similarly, adherents of COIN, who tend to be Baby Boomers (like Gen. David Petraeus) or younger, have been more willing to let the data guide them rather than use the data to fit preconceived conclusions, according to Jacob Shapiro, a former Navy officer and current Princeton professor.3 Sabermetricians have walks plus hits per inning pitched; COIN theorists have troops-per-square-mile.
The result has been an unprecedented use of data to combat insurgencies, chiefly in Afghanistan and Iraq. Eli Berman, an economist at the University of California-San Diego, was asked to examine data to find whether spending money on development programs helped reduce violence. The brass, he told me, already believed that "terrorist organizations that provide social services should be more deadly." But rather than merely rely on that instinct, as they would have in the past, they gave Berman the numbers and let him test the proposition quantitatively; he found the insight valid. (Sometimes the results are more surprising: I was told that one study concluded that insurgencies inflame as unemployment declines.)
But the more resonant similarity between sabermetrics and COIN is the broader mental shift they brought to their respective fields. The most famous sabermetric insight is that on-base percentage (the proportion of plate appearances that a batter gets on base) correlates more strongly with scoring runs than does batting average (which doesn't, for instance, account for walks). Arriving at this conclusion involved some number-crunching, but on-base percentage is an unsophisticated statistic that has been around for decades; and the proposition that a walk, which puts you on first base, is in many cases as good as a single, which also puts you on first base, isn't extremely subtle. Yet not 15 years ago, pretty much only the A's were deliberately accumulating workmanlike players with high on-base percentages—and doing so at bargain rates.
COIN is a similar alloy of numbers-backed, counterintuitive logic. When I asked Fred Kaplan, author of the COIN history The Insurgents, what COIN's central insight is, he replied, "You have to drain the swamp, you can't just swat the flies." COIN is less concerned with killing terrorists—in fact, it stipulates (and the stats show) that killing the bad guys can frequently net you more enemies—than protecting the local population and improving its infrastructure. The goal is to offer on-the-fence Afghans or Iraqis a better way of life if they do not support the insurgency. (Gen. David Petraeus, the most prominent COIN adherent, would ask, "What have you done for the people of Iraq today?") And it turned out that killing terrorists, particularly in Iraq, was not as efficient as co-opting local tribes, even just by buying them off.4 Though such thinking might seem rational, it constituted an about-face for the U.S. military. "There's a history written many years ago called The American Way of War—and it's, annihilate the enemy," Kaplan noted. "[Gen. William] Westmoreland was asked in Vietnam how you defeat an insurgency. His answer: Firepower."
Like sabermetrics, which was famously resisted by the grizzled, gut-driven scouts (albeit not as vociferously as is frequently remembered), COIN met opposition among the older, upper brass. "There [was] resistance to this at every level," Kaplan told me. "It's a generational thing. You had a group of officers coming up the ranks experiencing these kinds of conflicts"—in the Balkans, in Somalia, in Haiti—"and being told by their generals, 'We're not defining this as a war,' and they're thinking, 'It feels like war to me.'" Finally, that group of officers became the generals, and instituted COIN throughout the lower ranks. Once that happened, Beane's taming of the scouts in Moneyball felt familiar to Schneider. "It's almost analogous to a military situation," he said, "where the commander is confronting the colonels, [who are] fighting in the same old way."
Comparing sabermetrics with COIN does have its limits. In baseball, there is fair and foul; in war, all's fair. War is exponentially messier. It's also harder to measure: As Simpson said, "You're not going to have 3-D cameras on every hill in Helmand Province." Motivations and incentives are different. So are the stakes. While sabermetrics undoubtedly birthed a superior way of analyzing baseball, there are substantial questions about whether COIN, which is expensive and achieved questionable success in Afghanistan, is a sustainable doctrine.
And if we are living in a Moneyball age, we might be at its tail end in a certain sense: The period during which basic stats could be used to pluck the lowest-hanging insights is probably over—certainly in baseball, and likely in other fields as well, such as politics. Everyone knows about on-base percentage and voter engagement now.
The Moneyball revolution's lasting importance, then, is not these insights themselves, but a broader fealty to data and greater willingness to defy the established order of things. That, more than anything, is what the U.S. military—which, from ancient roads to the Internet, has developed so many technologies integral to human development—learned from the tidy, and usually inconsequential, world of sports. Five years later, Posnanski still seems bemused that the Army wanted to talk to him about baseball. Fontenot is not. "[Posnanski] was always puzzled why we were doing it," he said. "To me, it's pretty obvious."