A new weapon in the war against addictive gambling

News on 1 Aug 2013

Featurespace, a British technology firm spun off from a University of Cambridge engineering department project, says that behavioural analytics allied to machine learning can help identify early stage problem gamblers and arrest the compulsive gambling process.

The company says that online gambling sites routinely collect data on the betting patterns of their punters, including details such as duration of sessions, times online and days spent playing, bet sizes and the types of games favoured, reports The Guardian newspaper.

This information can be used by Featurespace to build up a picture of what is normal for any given individual, and what would constitute erratic or uncharacteristically risky behaviour that might indicate the onset of a gambling problem.

David Excell, the chief technical officer of the company, told The Guardian that whilst working in the fraud detection sector of gaming, Featurespace realised that its technology could be used to combat problem gambling.

“We decided since we’re harvesting so much data for our fraud solution work, how can we use some of that to try to understand the player from a corporate social responsibility point of view, to understand “is that player in control?” and so on”, said Excell.

Excell says that tackling gambling addiction helps protect the operator by identifying potential problem gambling before it becomes an issue in terms of regulatory, financial or PR considerations.

Matt Mills, commercial director at Featurespace, told The Guardian that operators strive to identify and minimise the threat of compulsive gambling to reduce operational hassles and the risk that regulators will introduce more stringent regulations due to public pressure.

Excell says that in traditional fraud-detection systems, alarms are raised when certain thresholds are crossed, but with gaming addiction, this rule-based approach is all but ineffectual.

“Where our technology really is works is that we get to know the habits of each individual. We start to learn an individual’s playing patterns, and to what extent they are predictable. Our hypothesis around this is that where a player is in control of their gambling, their gaming patterns will be relatively regular,” he opines.

“That definition of ‘regular’ will be defined based on who they are, so if they play a lot, that’s not necessarily a problem, but if we start seeing more erratic behaviour, rash decisions, playing at random times of the day, alarm bells will start ringing”, said Excell.

Different financial circumstances also have to be factored in to the equation, Excell says: “You could have a city trader and someone on minimum wage, and their ability to absorb different losses would be completely different. You can’t just apply thresholds and rules across the spectrum”.

Once a problem gambler is identified, Featurespace and the operator have to be sensitive in the steps taken to help, working with psychologists to establish how best to communicate the punter concerned.

“Most operators regulated by the UK market will offer session limits, deposit limits and loss limits, so what they might do is contact the individual and ask if they are aware of this functionality. The key is to be more suggestive than authoritative”, said Excell.

Operators must also consider whether they have satisfied the “duty of care” requirement following the identification of a potentially addictive gambler.

The burgeoning sector of social gaming is one where Featurespace sees worrying patterns emerging. Although players are not obliged to part with any funds, upgrades or access to new levels can be bought for a nominal amount, but the costs quickly accumulate.

“We’re seeing people becoming addicted in-game, but it’s not actually gambling. There are considerable financial costs involved, but no regulation around what that actually means. Many games have no age restrictions, so you can have very young people paying their 69p to get to the next level”, said Excell.

“You only have to look at the App Store and see that the highest revenue earner is Candy Crush, which is making over $600,000 per day, entirely through people paying for additional in-game bonuses. This is a huge amount of money, and is completely unregulated.”

http://www.theguardian.com/news/datablog/2013/aug/01/uk-firm-uses-machine-learning-fight-gambling-addiction

Related and similar