Detection of harms
We have previously called for improvements in the way operators use data to identify harmful play and how they use this as the basis to intervene with players to reduce harms.4 Although many operators claim to use sophisticated data science techniques, we are concerned that the overall numbers of customers identified do not reflect that harm is suffered by a much wider range of players than just those who go on to self-exclude – which is the basis of many operators algorithms.
Operators must use their data to target interventions on a wider range of players who could be at risk of harm. Initial screening processes should only be treated as the starting point for further interaction. Consumers may be vulnerable due to a wide range of factors and circumstances. Thresholds and triggers should not be set too high, otherwise many players who might be at-risk of harm will not be flagged early enough. The second stage of screening, making use of increased customer interaction, should then focus on understanding the player better and intervening appropriately. The Gambling Commission’s work on affordability is vital to ensuring people suffering harms are flagged, including those who are spending amounts of money which may not be immediately eye-catching – but could nonetheless be harmful to many players.
We recommend that operators be required to make public their approaches and procedures for detecting harmful play. They should also publish their protocols for how these triggers are followed up with further customer interaction and interventions. The data science and algorithms developed by operators are generally treated as commercial intellectual property. This situation should change. Greater transparency would allow operators to learn from each other and would help build a wider evidence base on what constitutes best practice and which methods to detecting harmful play are effective and defensible.
The Gambling Commission should play a leading role in determining what is best practice in the use of data for harm detection. As the availability of data is a key rationale for the absence of limits for stake, prize and speed of play,5 we recommend that the Gambling Commission involves itself far more deeply in understanding how this data is used – with a view to stipulating minimum standards based on established best practice. This may require recruiting people with specialist skills and experience.
Previous pageTransparency and evaluation Next page
Effective interventions
Last updated: 25 March 2021
Show updates to this content
No changes to show.