Getting the Right Numbers

The trailer came out recently for a new Brad Pitt movie (here), Moneyball, an adaptation of Michael Lewis’ book about the rise of the Oakland A’s baseball team, where General Manager Billy Beane was temporarily able to use innovative statistical analysis of players performance to gain an advantage over the opposition. Unfortunately for the A’s the benefits lasted only until the other teams caught on and started using Beane’s methods.

The use of data and statistics to understand performance in professional sport is not new and as technology improves the options to develop a sophisticated insight into performance increases. This has led to more and more teams hiring professional statisticians to develop a deeper understanding of the fundamentals of success on the field. At the forefront of this trend English football teams are making significant investment in technologies that enable the breakdown of the minutiae of the action on the pitch, this trend was recently documented in the Financial Times by Simon Kuper (here).

Whereas applying statistical analysis to baseball is a relatively straightforward process, the nature of the game makes it well suited to breaking down into to easily digestible chunks. Clearly, football is more complex, there are far more variables involved and the outcomes of different actions are less clear cut, however this increased complexity doesn’t mean that statistical analysis can’t work. Interestingly, what the analysts at football clubs are discovering is that the most obvious indicators of individual performance have little bearing on outcomes. Instead, a more sophisticated understanding of the game is required.

Kuper discusses this in terms of the evolution of statistics and analysis over the past 15 years, where clubs have moved on from looking at the more obvious indicators such as the distance covered covered by a player over the course of a match, where there is little correlation between the amount of running or effort a player puts in and results to more subtle but perhaps more meaningful indicators such as the amount of distance covered at maximum speed, which may have a greater correlation with positive outcomes.

Furthermore, there are many more variables in football that are harder to measure such as the ability to run into and find space and other aspects of positional awareness that are much more intangible than a completed pass or tackle which can be easily catalogued. For me the most important development cited by Kuper is a short sentence that indicates a potentially very interesting new avenue of analysis;

“And football’s data revolution has only just got going. Fleig thinks there is an exciting future in sociograms: who passes to whom, who tends to start a team’s dangerous attacks?”

By analysing the interaction of players and the role of team composition on performance, we can begin to move away from focusing on individual performance or excellence and instead begin to understand the much more important role of collaborative performance. What is really important in a team sport such as football is not who your best individual players are but what combination of players is best suited for a particular game. It is easy to measure who scores the goals, who has the hardest shot, who has the greatest athletic ability, makes goal saving tackles etc. What is key is to understand the effect of different players playing together has on the overall performance. For me, this understanding goes way beyond the football pitch and ultimately to the heart of organisational behaviour.

If Moneyball has taught us anything it is that it is possible to punch above your weight and compete against bigger, better funded and more powerful competitors. What this boils down to is the ability to create teams that perform at a far greater level than the sum of their individual parts.

Unfortunately the current reality is that most organisations are still tied to measuring performance in terms of individual behaviour and actions, what they need to do is take a leaf out of these football clubs and begin to analyse the social element of performance. By continuing to focus on the individual rather than what factors govern collective performance, organisations will remain stuck with the numbers that have little or no correlation to success. Unfortunately as has been proven on the football pitch, having a team of highly paid superstars is no guarantee of success.

This entry was posted in Uncategorized. Bookmark the permalink.

Leave a reply