Friday, September 8, 2017

Bartling, Brandes, and Schunk (2015) on Soccer and Reference Points

Björn Bartling, Leif Brandes, Daniel Schunk, “Expectations as Reference Points: Field Evidence from Professional Soccer.” Management Science 61(11): 2646-2661, 2015; working paper version (pdf) here.

• Do soccer teams play differently, and less “rationally,” when they are in the loss domain relative to expectations?

• Betting odds give a measure of expectations for match outcomes in professional soccer; so, we can test if teams play differently, and less rationally, when they are in the loss domain (performing worse than expected).

• Indicators are the numbers of yellow cards and red cards, as well as substitution patterns.

• The rationality of any changed behaviors in the loss domain can be checked by how final outcomes are influenced.

If a team that is favored is behind, they are in the loss domain. These teams receive more cards (by 14%), and their coaches make more offensively-minded substitutions (by a large margin), than if the game situation were the same but the team was not in the loss domain. 

• Both the additional cards and the increase in offensive substitutions seem to lead to worse outcomes for teams. 

• The additional cards when a team is in the loss domain tend to be related to frustration-style events, such as dissent or violent conduct. The loss domain is psychologically more challenging and this leads to worsened decision making.

• For more on reference points and soccer, see the BEO post on the Dickson, Jennings, and Koop (2016) analysis of Domestic Violence and Glaswegian Football. 

No comments:

Post a Comment