|Zeit:||28. November 2023|
Abteilung SOWI VII
|Download als iCal:||
Ort: Seidenstraße 36, Raum: 2.067 (2. OG)
Im Rahmen des Kolloquiums der Abteilung SOWI VII freuen wir uns ganz herzlich, Dr. Thomas Krause (Universität Hohenheim) mit dem Vortrag "Predicting sports betting player suspensions by algorithm: potentials, limitations, and recommendations." bei uns begrüßen zu dürfen.
Predicting sports betting player suspensions by algorithm: potentials, limitations, and recommendations.
Dr. Thomas Krause
The location-, social-, and time-independent nature of online gambling poses new additional risks for susceptible problem gamblers. The constant access to online gambling, regardless of location and time, hinders effective social control and increases the potential risk for individual and societal consequences. However, the multitude of process-generated data and the use of machine-learning techniques offer new opportunities for player protection. Automatic detection of problematic and at-risk individuals based on non-reactive behavioural data, for instance, can facilitate early intervention or outreach and thus minimize potential harm.
This project focuses on the potential of automated early detection of individuals with problematic or pathological gambling behaviour in online sports betting based on non-reactive data that includes the extent, change, variability, and development of betting behaviour. To achieve this, data from the online sports betting market in the state of Schleswig-Holstein is analysed. The goal of this work is to estimate the precision of the prediction and provide recommendations for the selection of algorithms, performance metrics, data handling, and central prediction variables for regulatory authorities and providers. The presentation will provide initial results that demonstrate the potential of using non-reactive behavioural data for predicting player suspensions in online sports betting. Various aspects will be discussed, such as the handling of raw data, the performance of different machine learning algorithms, and central determinants for prediction. Political implications and regulatory aspects will also be debated to avoid and prevent addiction risks in online sports betting. Overall, this analysis provides a significant contribution to improving player protection in the liberalized German online gambling market and demonstrates the potential of machine learning techniques for the early identification of at-risk individuals in online sports betting.
Alle Interessierten sind herzlich dazu eingeladen. Eine Anmeldung ist nicht nötig.