Yesterday I took my family to the Chicago White Sox game against the Milwaukee Brewers. On the way home, we learned the White Sox had just made a trade to acquire Kevin Youkilis from the Boston Red Sox. This piqued my interest because I remembered Youkilis is prominently featured in Moneyball: The Art of Winning an Unfair Game, Michael Lewis’ page-turner from 2003 about how Billy Bean, general manager of Major League Baseball’s Oakland Athletics, used empirical data to overcome the overwhelming financial advantage of large-market teams such as the New York Yankees. (Brad Pitt starred as Bean in the 2011 movie by the same name.)
In Moneyball, Lewis describes how Bean built a winning team on a baseball’s second-smallest budget by relying on data rather than baseball’s conventional wisdom. On the drive home I started thinking about whether there are any lessons business leaders could take from the book. In looking through the book again last night I found three:
Develop Proprietary Insight
The antagonist of Lewis’ story is conventional baseball wisdom circa 2000. The widely accepted belief was that teams were successful because their players had five basic “tools”—running, throwing, fielding, hitting, and hitting with power—that could be captured in common statistics such as speed in a sixty-yard dash, batting average, and number of home runs.
In fact, most teams were built on the premise that batting average was the primary precursor to runs and therefore to wins. However, Bean was the first GM to build a team on the insight that on-base percentage was more correlated with wins than batting average. Therefore, he started to select players who walked a lot (a walk improves on-base percentage but not batting average) and he was able to pay them reasonable salaries because of their unremarkable batting averages.
Baseball, Inc. had an incorrect mental model of what caused runs and Bean’s more refined model gave him leverage. A lot has been written recently about the importance of analytics and “big data,” but the larger point is this: organizations that want leverage need to seek out some proprietary insight—they need to have a better mental model than their competitors. That can come through scenario planning that helps management understand which industry uncertainties really matter, insightful analysis of corporate data, “voice of the market” research, or some (accurate) personal belief about what your customers really need.
Use Objective Criteria in Personnel Selection and Expect to Make Changes
The second Moneyball lesson that conventional managers can use is to reject received wisdom and intuition in personnel selection, which I wrote about previously and Wharton’s J. Scott Armstrong also does a nice job of summarizing. In Moneyball, Lewis writes that Bean was incredibly frustrated with his scouts’ selections for the amateur draft. These were often high school players, for whom limited statistics are available, or players who just “looked right.” Bean criticizes them for selecting players who would look good selling jeans instead of those who have proven something with their actual performance. Lewis says “what begins as a failure of imagination ends as a market inefficiency: when you rule out an entire class of people from doing a job simply by their appearance, you are less likely to find the best person for the job.”
Lewis quotes Bean saying “the draft has never been anything but a…crapshoot. We take fifty guys and we celebrate if two of them make it. In what other business is two for fifty a success?” Bean preferred to draft college players, who have more stats, collected for a longer period, against a wider variety of tougher players. They are also more than twice as likely to make it in the majors as those drafted directly from high school.
To be fair, managers of most organizations have few standard statistics with which to rate candidates. What chance, then, do business managers have when the characteristics of successful candidate are even harder to measure? For me, there are two takeaways: First, determine as best you can what characteristics lead to success and structure the interview process around measuring those characteristics, however imperfectly. Second, expect to make a lot of errors. Design recruiting and pay processes around the assumption that many choices will not work out.
Separate Individual Contribution from Luck and Circumstance
I once read that intelligence is the ability to keep two contradictory ideas in mind at the same time. As I re-read Moneyball, I initially thought I’d found a major contradiction but then realized I might just be dealing with great intelligence.
Bean and his lieutenants were major proponents of using statistics to make decisions about players rather than intuitive judgements about how the players “looked.” But, they also point out that statistics were imperfect, as they can be influenced by luck and circumstance. A pitcher can have a high earned run average because he gets called in when the bases are loaded. A batter may hit an abnormally higher (or lower) number of home runs one year completely due to random variation (call it luck, if you will).
You can think of similar situations in business. A sales person does not make her quota because the product is not competitive. A CEO makes millions of dollars on restricted stock because the overall market goes on a tear, even though his particular business is doing no better than average. The proponent of a new, innovative product is shown the door when it fails, despite the fact the eventual outcome was unknowable at the start.
A manager once told me that in the face of a bad situation, the best I could do was demonstrate to my company’s leadership that even though the outcome wasn’t what we hoped, they couldn’t think of a better approach. In his book, Lewis quotes Paul DePodesta, Bean’s assistant GM, as saying “it’s looking at a process rather than outcomes. Too many people make decision based on outcomes rather than process.” It almost seems as if he is advocating overlooking the statistics!
We live in era where the dominant theme is accountability. We do not want excuses from our executives, we want successful outcomes. A measure of our own aptitude, however, may be the extent to which we know when to judge others’ work based on outcomes and when to base it on the process they used.