It is becoming standard practice to detect problem gambling with automated tools to fulfil regulatory demands. One way to accomplish this is to employ rule-based procedures based on common markers of harm, be it amount of time or money spent or a myriad of other factors. Realizing this involves too many ad-hoc rules and has one main weakness. The rule-based approach to detect problem gambling neglects to identify the nuances of and the interplay between risk markers. To overcome this short-coming/blind spot, some operators and RG vendors are looking to AI to solve the task of identifying customers with disordered gambling.
Prediction is the new black
We ultimately want to know who of two seemingly identical customers progress to develop disordered gambling and intervene before it even comes to that while leaving the other gambler uninterrupted. Said simply, we want to predict how gambling will develop for every single gambler. Needless to say, this is a difficult task – even the best experts would be challenged to answer this question. At a minimum, it would require detailed gambling trajectories but even experts would struggle to identify subtle differences in gambling style at an early stage. It would even be difficult to articulate what to look for, this would be more of a hunch. As a result, Mindway AI has decided to respond to this need and now offers a complete solution to predict rather than detect problem gambling. It integrates into operator work-flows as easily as the GameScanner solution.
Turning gambling data into prediction
Artificial intelligence has the potential to identify and relate complex relations between consecutive bets to the development of problem gambling behaviour. In practice, this can be done by working back in time: given a reliable assessment of current customer behaviour, such as assessments made by trained experts, artificial intelligence can be trained on earlier gambling data for these particular customers. Effectively, AI then has the potential to identify patterns from early on in the customer journey which later on lead to disordered gambling. This would be a strong tool for managing customers and keeping their risk at bay. More specifically, applications can range from among others monitoring and profiling gamblers to nudging and early intervention to keep potential future problem gamblers in the safe zone. Another real benefit is that low-risk customers can be left totally undisturbed as there is no need to interfere with their gambling or use resources on them. Prediction and precision become synonyms in customer communication from very early on in the customer journey.
The bigger picture
There is an often overlooked but potentially even more important advantage of such a system: The AI could give away exactly which patterns are associated with later problem gambling. It is humanly impossible to grasp the tremendous amounts of data vertically across risk factors and horizontally across time, so AI is needed to analyze and generate insight into complex interrelations. By applying AI, we have a strong source of new knowledge, going beyond what we scientifically and clinically realize today. We could identify problem gambling before it even turns into a problem. The 1-2% of gamblers who develop problem gambling may never get to that stage, as they will be under a closer watch with intervention ready to launch. This in turn can lead to new and more effective rules, procedures and algorithms for early detection and intervention and enrich our fundamental understanding and approach to a sustainable gambling industry.
Kim Mouridsen, Professor and Founder of Mindway AI