Using predictive analytics to combat problem gambling

As part of our series of blogs on ‘future government operating systems’ we are asking a range of experts to examine the possibilities created for governments by emerging technologies such as machine learning, the internet of things and API platforms.

Using predictive analytics to combat problem gambling

Imagine if businesses could easily, automatically protect gamblers from addictive play, spot money-laundering and identify new fraud attacks as they happen - the impact for demonstrating control of customer populations to government legislators would be huge.

The pressure is on for businesses to use their vast volumes of data to adhere to public policy – minimising customer inconvenience while making it simple for government regulators to see that requirements are being met. The ability to identify normal behaviour for each individual (and spot when they deviate) has huge potential for helping both organisations and the government clearly visualise how regulatory requirements are being met – from monitoring responsible gambling, to anti-money laundering, identifying fraud, and protecting legitimate customers who are the backbone of sustainable business growth.  

But how can organisations draw definite conclusions from complex, varied sources of customer data and interpret these into useful policy-related actions?

The answer lies in revolutionary Adaptive Behavioural Analytics, applying cutting-edge machine learning and Bayesian maths principles to spot anomalies in real-time – an approach being pioneered at Featurespace via our ARIC engine (Adaptive, Real-time, Individual, Change-identification).

By automatically identifying the exact moment an individual’s behaviour changes, Adaptive Behavioural Analytics can help both organisations and government regulators - recognising gamblers at risk of addictive play, preventing fraud before it occurs, and even monitoring skin lesions for early warnings of melanoma. Compared to traditional, rules-based systems (which rely on knowing expected outcomes) this new approach is much more sophisticated, helping organisations intervene and communicate with their customers in an intelligent, proactive and targeted manner.

Combatting problem gambling with machine learning

In 2014, the Responsible Gambling Trust commissioned Featurespace and NatCen Social Research to investigate harmful patterns of machine play and draw implications for responsible gambling interventions. It was a ground-breaking first – achieving data collaboration from the five largest UK bookmakers, and allowing the actual gaming play of problem gamblers to be analysed and modelled.

Featurespace’s ARIC engine was able to learn the predictability of each individual player’s interactions, as well as understanding markers of harm (such as ‘chasing’ behaviours of faster gaming and higher stakes). Applying this Adaptive Behavioural Analytics approach distinguished between problem and non-problem players with a 66% improvement in accuracy compared to the baseline model derived from the current Code of Conduct defined by the Association of British Bookmakers. Gaming operators can use this approach for early identification of players most at risk, choosing the appropriate intervention method to prevent further harm.

Of course, an Adaptive Behavioural Analytics approach has wider implications beyond the Gaming industry – from banks fighting the increasing velocity of fraud while meeting FCA mandates to treat customers fairly, to new mobile payment platform Zapp analysing transactional behaviour for 18 million customers in real-time. Investigations have even been made into applications in the Health sector, such as monitoring skin lesion changes for early warning signs of cancer.

Embracing big data for public policy

With machine learning, the complexity of big data is no longer a problem for organisations – if anything, the more data the better for proactively understanding every individual, whilst keeping in-line with policy changes. Now that these behaviours can be understood, it’s up to organisations to make use of them, proactively indicating their ability to meet policy requirements to the government, and offering a better service to their customers.

Author

David Excell

David is the Co-founder and CTO of Featurespace