Begin collaborative research project to develop a methodology to detect harmful gambling behaviours and how problem gambling (i.e. vulnerability) can be detected during digital journeys based on customer behaviour. (City, University of London, LAB Group)
The online test records behavioural data (i.e. responses in the tasks/games), gambling severity data, and, most importantly, kinetic data. Kinetic data refers to metrics such as clicking, scrolling, swiping, mobile gyroscopic movements, and tapping.
In April 2022, we disseminated the test to many gamblers through charity networks (primarily GamBan and GamCare) and via research panellists. By early May 2022, we should have data from over 1000 participants.
Behavioural, gambling and mobile-specific kinetic data will be analysed via machine learning by partners at City. University of London in May 2022. Many of the machine learning algorithms and scripts have already been developed and will be implemented when participant data come in. The aim of using machine learning is to develop a model that can reliably predict whether someone is a vulnerable or non-vulnerable gambler based on their digital behaviour (i.e. how people click, scroll, swipe, tap, respond and move their phone). City and LAB will test the data using several models such as logistic regression, naive bayes, random isolation forest and timeseries autoencoder.
If we are able to build a model to accurately predict at-risk online mobile gambling behaviour, we are able to design mobile interventions that can nudge gamblers away from further harm and towards protection (e.g. one intervention might be a break-between-play or a headspace-style pop-up). If LAB and City reach this point, the next stage will be to implement said model in a live, digital environment with a gambling provider to examine how it works in the real world and can reduce UK gambling harms online.
The project is on-track to finish at the beginning of June 2022, though this depends on participant uptake complexity of machine learning modelling. A published report of the findings will be available upon request and project completion.
City, University of London and LAB Group have made strong project progress. The completed report on gambling using online language data analysis revealed key insights about vulnerable gambling behaviour. This was presented to the DCMS and well-received. In October and November, City and LAB Group will provide another online language data analysis report focussing on new behaviours especially around ‘mobile gambling’ using recent data.
City, University of London and LAB Group have made strong progress with the project. We have completed a data-driven stage of Comparative Linguistic Analysis Research to learn how gamblers think and discuss their behaviour, therefore revealing emerging trends and opportunities for intervention. We have also completed Qualitative Research to learn how gamblers interact kinetically when in online gambling journeys.
In the next stage of the project, City and LAB will conduct a large-scale quantitative test to collect kinetic and behavioural data which is then fed into machine learning analysis. The end goal from this is to develop new neuroscientific and behavioural digital tools to help prevent vulnerable gamblers from harm through dynamic digital solutions (e.g. changes to choice architecture) the industry can adopt quickly and easily.
City, University of London and the LAB Group have received £330,000 in regulatory settlements funding to conduct detailed research into new methods to help identify and protect problem gamblers in digital spaces.
The funding will allow City, University of London and LAB's research teams to jointly develop new neuroscientific and behavioural tools to help address the harmful effects of problem gambling and addicted behaviours, and identify solutions the industry can adopt quickly and easily.