From The Guardian – Britain’s gambling industry is booming, and new figures show almost one in 20 iPhone owners use sports betting apps, but a UK technology firm is using behavioural data to combat the onset of gambling addiction.
Featurespace, the corporate spin-out of a University of Cambridge engineering department project, is using machine learning techniques to identify people showing patterns indicative of problem gambling, before consulting psychologists on the best and safest preventative action to take.
Gaming sites collect data on the betting patterns of every one of their customers, including the time of day, frequency and size of bets placed and the types of games an individual typically plays.
By analysing this information, Featurespace is able 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.
Initially a consultancy, the company began working on fraud-detection solutions after winning a contract in 2008, and it was from its work in this area with gaming companies that the idea of looking at addictive behaviour emerged.
“As we worked with the gaming industry, we know a lot of the companies really well, it was looking at one aspect of fraud – first party fraud – which is where a customer will charge back a transaction and say “well I didn’t make that”, though sometimes they actually did make that transaction but they may have been spending a lot more than they wanted to”, said David Excell, co-founder and CTO.
“So it’s that false claim, maybe out of desperation, that constitutes the fraud, and we decided to start looking at the protections in place for customers at the moment to stop this kind of thing happening. 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.
