The Moneyball Movement

A BRIEF summary:

Michael Lewis's book, Moneyball, chronicles how Okaland A's GM Billy Beane started a baseball revolution, eschewing traditional baseball scouts--old baseball players--and instead hiring economists to find the market inefficiencies and exploit them. Using models to discover trends that predict success, Beane focused on obtaining players who could hit, no matter how fat, ugly, or unathletic they were (sure turn-offs for traditional scouts) and stayed away from players if they came out of high school (sure attractions for said scouts, who searched for potential, which to them meant ahtlietc upside.) Beane's teams flourished--under a fraction of the payroll of other teams--before anyone realized what was going on.

This same type of thinking has transferred over to other sports. In Basketball, Houston Rockets GM Daryl Morey and Mavericks owner Mark Cuban have been among the many who have incorporated statistics into how they evaluate players and how they pay them. Some players, namely Shane Battier, have been symbols for this movement and represent a different way of thinking about the game. I covered it briefly in this post.

It's really fascinating. Everyone from academics to 14 year old kids love the game, and they have different languages for talking about it. Typical fans relish slam dunks and eye opening plays, but the more sophisticated fan will reference a player's rating on a scale he possibly created. The logic implies that we can quantify what could not previously be quantified. Who is to say that the thinking behind "so what if Iverson scores 22 ppg, Kirilenko is more valuable because he is responsible for 21 wins produced to Iverson's 10, largely becasue of Iverson's high turnover rate, low true shooting percentage, and Kirilenko's ability to shut down whoever he guards", can't transfer over to "so what if Mr. X makes 300,000 a year, he does it at an environmental cost of -41 on the footprint scale, Mrs. X is more valuable to the family because she has an 8.6 on the raising kids scale."

Some people, chronicled here in a NYT article, think we rely too much on statistics as accurate measures and consequently prescribe remedies prematurely.

We want tangible results for everything. Grades. GDP. Environmental Footprint. It's more comforting on the surface to have a statistical measure than to say something is important but we have no idea how important. The problem is that unmeasured attributes become unimportant attributes. In basketball, this means that blocks are traditionally valued higher than opponent FG%; points are valued higher than PER. In the economy, this means money brought in is more important than a mother's job at home. (What metrics are there for love and care)?

There seems to be a widespread agreement that we should have holistic measures of quality but also acknowledge the shortcomings of such quantification. Basketball's solution has been to create statistics which better portray what's important in a basketball game. Experts suggest that while statistics cannot replace watching what goes on in a game--that there are intangibles which cannot properly be assessed by numbers--statistics can watch every game, making them vital for multiple game trends. Luckily, basketball is just a game and a quantification mistake is not the end of the world. But a quantification flaw in evaluating student success, public policy, or the economy can have serious negative consequences.

I'm curious to hear what is the solution generally in terms of quantifying quality. I hope to learn more by studying economics.