Sign in / Join

Moneyball & Efficiency

I've quickly noticed that the title, Moneyball, incites strong emotions. In spite of this, I'll risk a venture into some of the issues intersecting with Skip's insightful post on Monday. These issues touch on the nature of analysis in economics and highlight areas where sports examples make a valuable contribution. I'm no behavioralist by a long stretch. The simplified analytics offered by economics, such as the basic axiom that individuals try to maximize subject to constraints, are powerful tools. On the other hand, my "Virginia Political Economy" roots have never led me to to think solely in terms of "what is, is therefore efficient." In a dynamic world with heterogeneous individuals holding diverse and limited knowledge sets, maxmizing does not, to me, imply the absence of inefficiencies or the stock market-like instant arbitraging of new information across all environments. .

Sports provide glaring examples of the opposite so, for me, the hypothesis and evidence supporting the Moneyball premise has not been a stretch. For example, whether one buys the best-teams-first story of racial integration that Bob Tollison, Bobby McCormick, and I presented (2002 AER) or not, there is no doubt that black players were a valuable but underutilized input before 1947. Managers may have been restrained in using them prior to 1947, but the utilization of this resource did not occur evenly or even quickly. Instead, it dispersed over a generation in an S-shape just like Zvi Griliches' hybrid corn example. The "what is is efficient" rebuttal might be that these delays only reflect other constraints and costs such as customer preferences -- maybe, but these rebuttals quickly lead down a circular and non-testable cul-de-sac. Going beyond race, the same kinds of dispersion over time in the adoption of new sports "technologies" where customer preferences do not matter: the "pro set", the 4-3 defense, the shotgun, passing offenses, and film analysis much less non-sports examples like new corn varieties.

The Oakland As would not be the first sports example of a team finding a competetive advantage through the use of new and improved methods of selecting players. Tom Landry and Tex Schramm accomplished a similar thing in better organizing and analyzing information on players including using (early) computers. Competitors, the media, and even some of their veteran players mocked this "quantification" of skill as absurd at the time, but now, most teams utilize latter-day versions of their original ideas.