Covariance decompositions and betting markets: early insights using data from French trotting
The literature on the economics of betting markets has developed largely independently of the part of the psychological literature on judgement and decision making that focuses on the evaluation of subjective probability assessments. The aim of this chapter is to indicate how a speciﬁc type of subjective probability evaluation can be applied to racetrack data and to note the additional insights that may thereby be gained. It is shown that one can both arrive at a summary measure of the overall quality of the betting on a set of races and establish the relative contributions to this overall achievement of different types of knowledge and skill, in particular the ‘discrimination’ and ‘calibration’ displayed by themarket/s. Furthermore, one can carry out this analysis for both different sub-markets and different types of event. Accordingly, it becomes possible for serious bettors to identify where their activities might be most proﬁtably deployed and for the operators of betting service (who have access to data not available here) to determine, on the basis of concepts not previously employed, which particular bets and events will maximise their turnover.