Real world trial: Reduced spend after a targeted three-tier intervention

Session Title

Responsible Gambling: Player Outreach

Presentation Type

Paper Presentation

Start Date

28-5-2026 12:00 AM

Abstract

We examine players’ behavioural response to a targeted three-tier intervention across three brands. Operator teams sought to speak by phone with players identified as high risk over a 12-day period, to assess possible exclusion needs. Where contact could not be established, up to three messages were sent along with ongoing attempts to call the player, culminating in an imposed monthly loss limit if players could not be contacted. The study tracked 38,952 intervention event clusters, including follow-up messages where applicable, across 7,809 unique users over a 2-year period (2023-25). Behavioural change was measured across 5,10, and 15 day periods before/after an intervention cluster. The average pattern across players identifies a sharp reduction in average behaviour for 5-day analyses, with a gradual return to trend for 15-day analyses (remaining below baseline). Key findings include a 21% reduction in gambling days and 45% reduction in gambling day loss in the 5-day analyses, moderating to 17% and 36% reductions in the 15-day analyses. A significant proportion of players, 14%, did not play at all in the 5 days following the intervention, although all but 5% had played at least once by the time 15 days had passed. The presentation will dive into differentiated behavioural responses by interaction/message type, brand, and different observation windows, exploring implications for responsible gambling intervention effectiveness and recommendations for future research.

Author Bios

As principal data scientist at Playtech Software Limited, Sanjoy Sarkar is responsible to identify and implement automation using AI and machine learn technologies in the business areas within Playtech’s service operation and related product offerings. For last eleven years, he is associated with responsible gambling research, data analytics, AI model development and product design. Originally, from the world of information technology, where most of his career was spent dealing with data, its model(s), analytics and application, he brought wide data analytics experience into responsible gambling world, finding insights from data, building state-of-the-art AI models to identify players at risk of gambling harm.

Share

COinS
 
May 28th, 12:00 AM

Real world trial: Reduced spend after a targeted three-tier intervention

We examine players’ behavioural response to a targeted three-tier intervention across three brands. Operator teams sought to speak by phone with players identified as high risk over a 12-day period, to assess possible exclusion needs. Where contact could not be established, up to three messages were sent along with ongoing attempts to call the player, culminating in an imposed monthly loss limit if players could not be contacted. The study tracked 38,952 intervention event clusters, including follow-up messages where applicable, across 7,809 unique users over a 2-year period (2023-25). Behavioural change was measured across 5,10, and 15 day periods before/after an intervention cluster. The average pattern across players identifies a sharp reduction in average behaviour for 5-day analyses, with a gradual return to trend for 15-day analyses (remaining below baseline). Key findings include a 21% reduction in gambling days and 45% reduction in gambling day loss in the 5-day analyses, moderating to 17% and 36% reductions in the 15-day analyses. A significant proportion of players, 14%, did not play at all in the 5 days following the intervention, although all but 5% had played at least once by the time 15 days had passed. The presentation will dive into differentiated behavioural responses by interaction/message type, brand, and different observation windows, exploring implications for responsible gambling intervention effectiveness and recommendations for future research.