Lenten examines how umpires voted for the "best and fairest" player of each season in the Australian Football League (AFL). True to the Becker approach, Lenten looks at the market for personal characteristics that are not related to productivity. He finds that "Indigenous footballers" receive more votes - something that he views as arguably unexpected.
The productivity data includes scores, free kicks (for & against), and tackles. Lenten says that an "attractive feature of this AFL data is that, as the sport's fans would attest, each of these match statistics serves as a reliable and informative input into a given player's production function". I'm not so sure that he needed to add the following sentence where he says "This is not necessarily the case in other sports; for example, possession time in soccer, which can sometimes be misleading". Most sporting statistics can be misleading sometimes. This includes the issue of potential measurement error that Lenten returns to later.
The description of the statistics, and what they are attempting to capture, is thought provoking. The paper includes three measures dealing with the "disposal" of the ball. There is no variable for "passes". This probably speaks to the nature of the game. It is not too far removed from the sport of Gaelic Football that is played in Ireland. In fact, the countries play each other in a hybrid game. The word "disposal" could also be used in gaelic football. In fact, one can frequently hear fans roar "Get rid of it". It makes one think about the classification of what a player does with the ball. When does a "disposal" become a "pass"? Soccer statistics examine passes and pass completion rates. Soccer also counts "crosses". Is a "cross" somewhere between a "disposal" and a "pass"?
There are other thought provoking parts to this paper.
At the end of his Results section, Lenten says "The possible underlying reasons for the main result (other than discrimination) - for example, differences in style of play - are ultimately beyond the scope of this study, and can be left open to fan debate." I'm not sure exactly what Lenten has in mind by style of play but it made me think of team dynamics that might be more than the sum of individual characteristics. Like a macroeconomics that is not based on microeconomics underpinnings. My instinct, and sporting experience, suggests that such things matter. But my training in economics, including my admiration for the work of Mancur Olson, raise doubts.
Lenten follows the quoted sentence in the last paragraph with the following two sentences. "One possibility using economic reasoning - again presuming that the mean Indigenous player is more talented - is that greater talent extends to higher production levels 'off-the-ball', which are unmeasurable in an objective manner, such as the match statistics used here, yet may be unnoticed by umpires. Another possibility could be that Indigenous players tend to execute certain plays successfully (e.g. kicking goals") with higher degrees of difficult, a prospect that this particular sample would be unable to reveal."
These two sentences, and the one immediately before them, are worth pondering. The sports statistician might ponder how to capture the off-the-ball characteristics and the relative importance of these characteristics (see previous blog post here). The economic theorist might ponder what do we mean by "economic reasoning" and if it is any different from statistical reasoning. How has it changed over time and how has Gary Becker contributed to the way economists reason today? Lenten's paper gives us the opportunity to ponder these issues. It is well worth reading.