Hey big spender! Yodo1 develops AI that’s on the money
A mobile game developer has created an artificial intelligence tool that can analyse player behaviour – with almost 90% accuracy – and predict who will spend the most money.
Henry Fong, CEO of Yodo1, recently spoke about how free-to-play games need to begin thinking about monetisation at the design stage, as opposed to factoring it in afterwards.
During his speech he illuminated on how ‘whales’ could spend up to several hundred dollars through microtransactions.
Yet in the case of one particular mobile game, a user had managed to rack up a bill of $150,000.
Perhaps more alarming is the fact he has been working on an artificial intelligence (AI) tool that analyses player behaviour and session times to determine who might become big spenders.
The bot can spot people who are likely to spend a lot of money with 87% accuracy – though the CEO believes this figure could be raised to as high as 95%.
It has been trained for roughly two and a half years with Fong even revealing that it is technically possible to build in logic to target different players with varying monetisation packages based on their behaviour.
He also admitted that implementing such a tool would be catastrophic from a business point of view because the backlash would result in a net loss for the studio as people boycott it.
“We don’t want to create a situation whereby different people pay different prices for the same thing,” he tells Kotaku Australia.
He then made reference to how tools such as this one would need to be heavily regulated as the video games industry continues to grow and impacts more and more people.
“We need to work with regulators to make sure that they have the full context of the industry and that any regulations work as intended and don’t break a bunch of other things by accident,” he adds.
Interested in reading more mobile gaming-related stories? Discover more about Google launching its own subscription service to rival Apple Arcade.