It’s exactly this challenging question that Bournemouth University and VAIX.ai are seeking to answer with their unique player protection partnership announced today.
Preparations for the second annual Responsible Gambling Week are in full swing for next month. Yet with the final quarter left to run, 2018 is already reported as another record year for the industry experiencing the impacts of regulation. With the repeal of PASPA and the impending regulation of Europe’s second largest gambling market in Sweden, this is a hard trend that you couldn’t have escaped even if you’d been living under a rock for the last few years. Huge opportunities have emerged, but of course, travelling with them is the shadow of considerable challenges.
While it is both necessary and expected that revenues from problem gamblers will be reduced by increased protection measures, next-generation deep learning and AI algorithms mean that Operators can at the same time improve the experience of existing players, and increase revenue through enhanced entertainment where it is responsible to do so. AI algorithms mean that we can both provide better protection to the vulnerable, and lift the entertainment experience of other players, say Andreas Hartmann, Co-Founder and COO of VAIX.ai; and Raian Ali, The EROGamb Project Leader at Bournemouth University.
The challenge is manifest - Operators can’t afford to be a burden on society, but what about the burden to their businesses?
AH: Imagine you’re at an airport and the policy is that everyone with a beard must have their bag routinely searched. Even if you have a squeaky clean profile, merely sporting some designer stubble would have a negative impact on your experience - both in terms of long wait times, and the disturbance to you as a perfectly sensible traveler with no risk profile. Current rules-based and predictive analytics approaches to identifying problem gamblers (and fraud) aren’t really that much smarter and produce too many false positives. Continuing to increase such administrative measures under the banner of responsible gambling measures will not only fail to protect the vulnerable, it will also get in the way of enjoyment for the majority of players.
Maintaining such forests of rules has become an expensive and burdensome effort for Operators whilst showing limited effect on protection of the vulnerable. It is increasingly more difficult for Operators to show the necessary quality of duty of care when the accuracy of the rules cannot possibly capture the unique circumstance of each player. Surely the time and energy wasted could be more effectively applied, if only the signals in use were more accurate. Unfortunately for executives, rule-based and predictive analytics shallow AI systems also often convey a false sense of security through the mistaken belief that they should be sufficient. What’s happened now is the focus on responsible gambling has never been greater, which has in turn rendered the practice of extending rule forests an unsustainable approach.
RA: Prevention is always better than cure, so the earlier that gamblers can understand they are developing risky or problematic behaviours the better. While initiatives such as GamStop play an absolutely vital role in the player protection ecosystem, the unfortunate reality is that significant harm may have already occurred before the gambler takes steps to self-exclude.
At the moment, player protection features required by regulators rely on the gambler self-identifying their need to make use of them. They are also designed independently with very limited research input, and limited or no sharing of best practice between Operators.
What then, is the solution?
AH: In an industry which turns on good fortune, it’s an amazing stroke of luck that AI and deep learning have matured at just the right time for the industry to take a fresh perspective. By leveraging the expertise that has enabled VAIX to "deliver higher player entertainment via recommendations, predict player lifetime value, and churn, Operators can now call a halt to ineffective and expensive legacy approaches whilst achieving a vastly better result for all involved.
Gamblers exhibiting or developing problem behaviours can be helped far quicker and more effectively when a deep AI-driven vulnerable player identification system powers the Operator’s response method, presentation, and tone. And whilst many Operators have been concerned that more effective responsible gambling measures will hurt all players’ experiences, AI player protection measures can - perhaps counterintuitively - move completely out of the way for players who are still enjoying gambling as responsible entertainment.
With AI, the conflict with the business model is transformed into a state of balance with better player protection. Surprising, delighting, and entertaining customers can remain responsibly as a front and centre objective.
RA: We agree with the KPMG research consensus that an end-to-end methodology for safeguarding the protection of vulnerable customers is required, and also with the Department for Digital, Culture, Media, and Sport (DCMS) that correct next steps are to trial a range of harm minimisation measures within such a methodology.
If gamblers were well-informed in the right way, then they would consider themselves supported and well-equipped to control their gambling independently. With the right data together with effective analysis and user experience research, technology can be designed to enable gamblers to be more self-aware, and then to define and customise their problem gambling triggers.
Compared to other industries where harm can occur through misuse, the gambling industry has an unprecedented capability to track personal consumption and communicate data and information about it to gamblers, their surrogate counsellors, and intelligent software, in real-time. Our research group aims to make a significant contribution to overcoming challenges in player protection through such real-time access to data and the timely and interactive persuasive techniques towards a more conscious and controlled gambling.
Is the cost of smarter approaches generally misunderstood?
RA: For the majority of individual Operators, collaborating with universities to access their knowledge and research capabilities is often perceived as a nice idea, but unrealistic to execute in reality. Openly collaborating with other Operators to develop and share best practices in the area of responsible gambling comes with the same perceived challenges. Today with VAIX.ai, our Engineering and Social Informatics research group (ESOTICS) at Bournemouth University have embarked on a collaboration journey to prove these notions to be false in the player protection area of regulation.
Accessing data is just the first step where we need to research every single aspect of their usage to avoid causing further harm, e.g. lowering self-esteem or triggering overconfidence. This includes the timing, content, framing and presentation of such data-driven interactive and persuasive techniques and the profiling of players for their preferences on them. Research partnerships at scale are critical to the cost-effective delivery of deep insights that drive improvements in this area.
AH: Yes, such approaches often belie the deep investment requirements necessary for huge internal programmes of work that can only be afforded by the largest Operators. Either that, or costly specialist outsourced alternatives with complex integrations. In addition, Operators almost always experience challenges with acquiring bandwidth in their product roadmap to deliver projects.
At VAIX, we want to make things as simple as possible, so we set out to clear these barriers from the start. Sharing data with us is all that is required with very little tech involvement to make use of deep learning to detect and predict problem gambling behaviours. Deep AI is far more accurate and produces significantly less false positives than rules-based and shallow AI predictive analytics. As a result, Operators tend to pleasantly surprised at what is possible using VAIX.
Unlike other services providing proactive harm prevention services, VAIX offers a sustainable ecosystem of AI microservices that enable Operators to derive economies of scale from their data partnership with us. Most of the data required for our services is the same, so through our collaboration with Bournemouth University, Operators are able to proactively contribute in a meaningful way to responsible gambling research, whilst at the same time drive significant commercial value from their customer data in a sustainable business model.
Any final thoughts?
RA: We are delighted that our collaboration with VAIX will enable us to overcome a challenge we have experienced in connecting to, and analysing Operators’ wide range of gambling behaviour data in real-time. We also plan to suggest additional data to capture, with user consent, and enable further mechanics for responsible gambling, e.g. browsing history of the gambling site, time of access, device used and location.
AH: We’re very excited about our collaboration with Bournemouth University having a deep positive impact on player protection, both for society and for Operators’ reputations as trustworthy and responsible businesses.
When is the next opportunity to meet with you?
Raian and Andreas invite you to discuss the possibilities with them at the World Regulatory Briefing on October 17th
“Let’s Talk About Responsible Gambling” is the theme of this year’s Responsible Gambling Week