Moneyball Revisited
posted by Michael Kang
Michael Lewis’s bestselling book Moneyball occupies a unique convergence of academic, sports, and popular fascination. Moneyball profiles Billy Beane and his management of the Oakland Athletics baseball team, with particular attention to Beane’s use of cutting-edge quantitative analysis in an industry portrayed as bound by tradition and decisionmaking by anecdote. Moneyball garnered recent attention again after the movie version of Moneyball, starring Brad Pitt, suddenly halted production just five days before shooting was to begin in July. The event, or nonevent, brought forth several commentaries on Moneyball’s legacy, six years after its publication. Today’s post begins to explain my ambivalence about Moneyball’s place in the academic imagination; my next post continues by arguing that, perhaps to the surprise of its academic enthusiasts, Moneyball actually gets a good chunk of its baseball wrong and in the end, may tell a slightly different story than usually thought.
Baseball fans from outside academia would be shocked how influential and popular the book Moneyball has been within academic circles. Cass Sunstein and Richard Thaler wrote a book review of Moneyball for the Michigan Law Review, and professors have cited Moneyball as inspiration for new approaches to everything from faculty hiring to election administration to health care reform. There’s even a Moneyball-inspired blawg called Moneylaw. The great contribution of Moneyball was to puncture a certain overconfidence in untested conventional wisdom based on unsystematic anecdotal information. Moneyball offered a colorful example from baseball, now widely cited in academia, of how inefficiencies in markets can be exploited by canny operators who identify objective metrics of value underappreciated by traditional practices. As Sunstein and Thaler note, “If Lewis is right about the blunders and the confusions of those who run baseball teams, then his tale has a lot to tell us about blunders and confusions in many other domains.”
The problem with Moneyball is the hyperbole deployed to construct Lewis’s lesson of absolute quantitative triumph. A key element of Moneyball’s influence is the vividness and persuasiveness of Lewis’s account of the Oakland Athletics’ success, but it is so vivid and persuasive at least in part because it exaggerates the brilliance of Billy Beane and his quantitative approach to baseball.
A naive or otherwise uncritical reader might walk away from Moneyball in awe of the overwhelming genius of Beane and his methods. However, a shocking number of the predictive judgments described by the book as genius turned out not to be so smart at all, and the distorted descriptions of the then-present would stun a baseball-savvy reader today (e.g., consider Lewis’s repeated derision of Miguel Tejada, during a season when he won the league MVP and finished fourth in win shares). This sounds like petty criticism of the book, but it’s not. Lewis’s hyperbole is central to his persuasive case that baseball traditionalists didn’t just view baseball differently than Beane and his team, but that they view baseball backwardly, made misjudgments to which Beane would never fall prey, and clung to hoary myths despite their demonstrable wrongness.
Of course, my point is not that statistical approaches have no place in baseball and elsewhere. In fact, as someone sympathetic to statistical analysis and enthusiastic about the new metrics now available to baseball fans after the book’s publication, I always found annoying Moneyball’s overstated portrayal of Beane as nearly infallible. It would be wrong to replace an overconfidence in qualitative data with an overconfidence in quantitative data to the disdainful exclusion of other useful information, particularly after watching major financial institutions melt down because of similar overconfidence. I won’t rehash what is a familiar methodological debate for social scientists, but the modest and important point is that statistical approaches, while incredibly powerful, have at times their own weaknesses and certainly no exclusive claim to truth, as the book Moneyball often overclaims. As Ian Ayres explains in Super Crunchers, a less colorful but more carefully written book than Moneyball, “In the end, Super Crunching [i.e., quantitative data-based analysis] is not a substitute for intuition but a complement . . . . The future belongs to those who can comfortably inhabit both worlds.”
Where does Moneyball exaggerate Beane’s brilliance? In a future post, I’ll describe in detail Oakland’s 2002 draft, which is a central event in the book. However, most general assessments of Moneyball’s value as baseball philosophy have understood it as either validated by past successes, or undercut by Oakland’s current failures. And as Howard Bryant reports, “Around certain quadrants of baseball, there is no shortage of enjoyment in the belief that the team’s sub-.500 record over the past 2½ years — Oakland [at the time, owned] the third-worst record in the majors — is proof that the game is witnessing the denouement of the Moneyball legacy.” Nonetheless, the bottom line of wins and losses doesn’t necessarily attend to the strength of the causal relationship between Moneyball as baseball philosophy and those wins and losses. In other words, the better questions might be how much of Oakland’s past successes were attributable to Moneyball’s wisdom, and how much of Oakland’s current misfortune is attributable to Moneyball’s failings? As to the former, I am more skeptical than Lewis.
I agree with the consensus within professional baseball that the A’s success under Billy Beane is attributable more to their fortune in having excellent starting pitching than the Moneyball offensive strategies discussed so heavily in the book. Set aside the fact that Oakland has won only one playoff series, and no championships, during Beane’s entire tenure. From 2000 to 2004, the Athletics won at least 91 games during each regular season and made the playoffs four times in five seasons. However, each year during that stretch, the Athletics were led by the trio of Tim Hudson, Mark Mulder, and Barry Zito—three all-star starting pitchers who were improbably healthy every year and affordable. Oakland’s pitching shined during the period, never finishing lower than third in earned runs allowed and finishing first overall in 2002 and 2003. But since Hudson and Mulder were traded, Oakland has qualified for the playoffs only once in six seasons. Of course, the Athletics’ offense played a role in the team’s success, but its ranking among the 14 American League teams in runs scored was third in 2000, fourth in 2001, eighth in 2002, ninth in 2003, and ninth in 2004. Solid to mediocre. Oakland’s offensive performance seems quite unremarkable when you look closely, despite all the attention Moneyball brought it.
I suspect that specific details about Moneyball matter little to most of its academic enthusiasts. Moneyball simply provides a coathook from which to hang the argument that markets contain exploitable inefficiencies, or that quantitative analysis provides clear-eyed empirical truth ignored by the superstitious. These can be fair and useful points regardless whether Moneyball effectively cements its case about the Oakland A’s. Pointing out Moneyball’s hyperbole does not undermine the ultimate value of quantitative analysis in baseball or elsewhere. What is more, Beane is surely a good general manager who has had his run of successes. But a better, more accurate understanding of the underlying narrative in Moneyball presents a better, fairer assessment of what lessons Moneyball actually offers us.
There’s much more to say about Moneyball. Next time, more detail and what I think is the real lesson of Moneyball.
August 18, 2009 at 1:00 pm
Posted in: Current Events, Psychology and Behavior
Print This Post
Responses (7)
pe - August 19, 2009 at 4:17 pm
of those years when they were 3rd, 4th, 8th, 9th, and 9th in offense, what did they rank in payroll? That is sort of like the central premise of the book….a point you should at least consider addressing.
Michael Kang - August 19, 2009 at 5:24 pm
pe, thanks for the comment, but I think you may be sort of like missing my point. There’s no question that Oakland won a lot of ballgames with low payrolls during 2000-04, but most academic commentators cite that fact, credit Moneyball strategies, and call it a day. But that’s insufficient to show Moneyball strategies were largely responsible for those victories. My discussion of the offensive rankings suggests that perhaps the prominent Moneyball emphasis on on-base percentage, discussed very prominently as a focus of the book, wasn’t as responsible for those victories as Lewis represents.
You’re right, though, that cost is central to Moneyball, and I hope to talk more about it in my next post (which I’m struggling not to make so long!).
Andy - August 19, 2009 at 8:40 pm
Regarding your point about the strength of the A’s pitching during this period, I’m curious as to how much of a home field advantage the A’s had playing in one of the most pitcher-friendly parks in baseball. It makes alot of sense that the A’s focused on pitchers that did not walk many hitters or historically give up many HRs because their stadium has a very large foul territory and I believe is hard to hit the ball out of.
I also wonder how much money the other teams in their division were spending during this time (cannot remember if Moneyball mentioned this comparison as opposed to just the top spenders in baseball).
Assistant Village Idiot - August 20, 2009 at 6:05 am
That the Red Sox and Yankees also adopted OBP strategies, but could also outbid other teams watered down the market advantage of the A’s.
Part of the difficulty in seeing the Moneyball lesson – even for the A’s – is that OBP has come to be seen a player evaluation strategy in the present, when its better value is as a predictor of young players’ progress. Control of the strike zone is a good predictor of improvement in a 22 year old. Think Kevin Youkilis. It is far less effective in predicting the rate of decline in a 32 year old. Think David Ortiz.
While I agree that the lack of injury to A’s pitching was fortuitous, it is not altogether accidental in their strategy. Acquiring a lot of pitchers whith good strikeout-to-walk ratios and few homeruns creates a bench strength that means there is less pressure to overuse a star pitcher.
All in all, I agree with your assessment, however, and look forward to your evidence for other factors.
Linus - August 20, 2009 at 12:18 pm
My understanding is that the new “OBP” or undervalued, cheap skill available, is now defense. But of course, some of the large market teams have already been able to move ahead of the curve on that, and so it becomes harder for the smaller market teams to hunt for inefficiencies.
Linus - August 20, 2009 at 12:36 pm
Perhaps you’ll explain this in a future post. But you seem to indicate that the success or value of Moneyball doesn’t lie with the actual results of the philosophy and approach of Billy Beane, (namely the success of the A’s), but can be found elsewhere.
So a couple of questions then.
1) How do you judge the success or importance of Moneyball if not by the A’s performance on the field?
2) Would it be more accurate to say that much of the “paradigm ” shifting ideas attributed to Billy Beane and Moneyball, belong far more to bill james and SABR? It seems to me that the biggest real change in the As was to start taking sabermetrics seriously, and other teams began to legitimize that approach from Paul DePosta to the BoSox hiring the guru himself.
Finally, i would most definitely recommend reading “The Drunkard’s Walk” for anyone interested in quantative analysis and statistics in general, or baseball specifically. One of the best chapters is on how to evaluate the underlying performer based on performance. I suspect that most evaluations (either quantitative or qualitative) suffer egregiously by not considering confidence intervals properly.
Michael Kang - August 21, 2009 at 12:25 am
Linus,
To assess Moneyball’s value, I agree that we need to look at performance on the field, but looking just at the number of wins and losses plus payroll is insufficient. We need to look more closely at how well specific Moneyball strategies worked. For instance, in the next post, I try to do this by studying the results of Oakland’s 2002 “Moneyball draft” described glowingly in the book. By looking more closely at specific Moneyball strategies, we learn more than by looking just at wins and losses. It is a mistake, in my mind, to assume that Moneyball was a success simply because Oakland won with low payrolls in the early part of the decade, or to assume that Moneyball is a failure now simply because Oakland has been losing in recent years.
You’re right that Bill James was a sabermetric pioneer, and Beane and Lewis fully credit James’s critical importance. To be sure as well, the A’s before Beane, the Yankees, and perhaps other teams understood long ago certain Moneyball insights like the importance of walks and OBP. Beane’s innovation was embracing statistical analysis at the organizational level and relying on it so heavily, as described in the book. It was an important innovation, but I think Moneyball and its value gets oversimplified and somewhat overblown by commentators who may not follow baseball so closely.
Leave a Reply