Players Gonna Pay: Characterizing gamblers and gambling-related harm with payments transaction data. (June 2023)

• Record Type: Journal Article
• Authors: Ghaharian, Kasra; Abarbanel, Brett; Kraus, Shane W.; Singh, Ashok; Bernhard, Bo
• Is Part Of: Computers in human behavior. Volume 143(2023)
• Publication Date: 2023-06
• Subjects: Online gambling -- Digital payments -- Gambling -- Gambling harm -- Cluster analysis -- Fintech; Interactive computer systems -- Periodicals; Man-machine systems -- Periodicals; 004.019
• DOI: 10.1016/j.chb.2023.107717
• Languages: English
• ISSNs: 0747-5632

Abstract

Abstract: Payment providers in the gambling industry facilitate the transfer of money to and from gamblers' wagering accounts. Payments transaction data is information captured from these transactions, such as the type, amount, time, and location. There is significant stakeholder interest in how technology-assisted payments may affect potential gambling harms, but research using actual payments transaction data is rare. This study considers the utility of payments transaction data in distinguishing subgroups of gamblers, and exploring potential markers of harm in gambling payments transactions. We benchmarked six cluster analysis methods using a dataset of 2, 286 online casino gamblers obtained from a U.S. gambling digital payments provider. The k -means algorithm with five centers was the optimal method. Two large clusters contained the majority of the dataset (87.9%) and characterized customers at a low risk of harm. Three smaller clusters comprised profiles of customers at a potential risk of harm – a high deposit-to-withdrawal ratio (8.4%), high activity, high intensity (2.5%), and high volume, high variability (1.2%). Our results establish the use of payments transaction data for identifying subgroups of gamblers, including potential risk groups that provide preliminary insights into payment behavioral markers of gambling harm. Highlights: Payments transaction data can be used to identify subgroups of gamblers. Cluster analysis revealed five distinct payment behavioral profiles. The majority of customers were classified into two lower risk clusters. Three smaller clusters comprised individuals at a potential risk of gambling harms. Possible markers of harm relate to deposit and withdrawal activity and declines.