B0601
Title: A varying coefficient state-space model for investigating betting behaviour within in-play markets
Authors: Marius Oetting - Bielefeld University (Germany) [presenting]
Rouven Michels - Bielefeld University (Germany)
Roland Langrock - Bielefeld University (Germany)
Abstract: The potential effects of in-game dynamics on betting behaviour are investigated. Considering two comprehensive data sets from the 2017/18 Bundesliga season comprising in-play betting volumes and match events, we use state-space models to analyse the dynamics and drivers of betting volumes. Within this state-space framework, we use (penalised) B-splines to model the potentially time-varying effect of in-game dynamics as implied by measurable events such as shots and passes. Preliminary results suggest that volumes in the in-play market are driven by such in-game dynamics and that this effect varies throughout a match.