Academic journal article
By Stewart, Mark F.; Mitchell, Heather; Stavros, Constantino
International Journal of Sport Finance , Vol. 2, No. 4
The best selling book Moneyball posited a theory on the success of a Major League Baseball franchise that used detailed match data to identify inefficiencies in the market for professional baseball players. These statistics were then exploited to the advantage of that team. An important part of this strategy involved using mathematical techniques to identify which player statistics were most associated with team success, and then using these results to decide which players to recruit. This paper uses a similar approach to analyze elite Australian Football, making use of various types of regression models to identify and quantify the important player statistics in terms of their affect on match outcomes.
Keywords: Moneyball, Australian Football, sports statistics
The use of statistics to assist sporting organizations in making personnel and coaching decisions is not a new phenomenon. They have, however, been given increased prominence with the release of books and the publication of websites that aim, in part, to describe advantages that may accrue to those sporting teams who best utilize these statistical methods. Michael Lewis' Moneyball (2003), which deals with baseball; The Wages of Wins by Berri, Schmidt, and Brook (2006), which focuses primarily on basketball; and the website Football Outsiders (http://www.footballoutsiders.com), which analyzes American football, are prominent examples of this.
The purpose of this paper is to explore the possibility of whether statistical methods can be used to assist in the recruitment of Australian Football League (AFL) players, particularly to establish if there are any market inefficiencies to be exploited. Using various regression models the individual player statistics that are most highly correlated with team success are selected and quantified. That is, the statistical modeling in this paper is able to show the statistical relationship between individual player statistics and team winning margins.1 This is something that has not previously been done for Australian Football.
Using the results from our model, club recruiting staff could use these statistics to identify potential players. These player statistics would be used alongside, or in place of, the traditional more subjective methods of selecting players that are currently utilized.
This paper will proceed as follows. In the next section some of the previous research using statistics to recruit elite sportsmen is summarized. Then the data used in this study is explained. After this the econometric estimation and results are outlined. This is followed by a discussion of the implications of our findings, and lastly some conclusions are drawn.
Lewis' (2003) popular publication Moneyball has been a significant catalyst in the increased attention given to statistical analysis and sporting organization decision-making. The book, which was among the top 10 on the New York Times best-seller list every week of 2004, chronicled the exploits of the Oakland Athletics Major League Baseball (MLB) team. In 2002, the Athletics, despite having close to the lowest player payroll, won the equal highest number of games throughout the regular season. This outcome, according to the theory suggested by the author, was directly related to a strategic statistical approach that sought to exploit perceived irregularities and inefficiencies in the baseball player labor market. By focussing on recruiting older players (college rather than high school) and emphasizing the importance of a certain narrow range of baseball statistical measures over traditional approaches, the Athletics were able to build a roster of players that performed very well for relatively little cost.
Although the details of how the Oakland Athletics specifically formulated their player statistical valuations was not presented in great detail, it appears that some form of regression analysis (as is the case in the current study) was used. …