Player Modeling

One of the sexiest aspects of digital environments and online worlds is the ability monitor player behavior in an unprecedented way. Every single action, decision and event can be recorded, aggregated and analyzed on a scale that would make Orwell look like an absentee baby-sitter.

However, before we get too excited and blow our Dystopia-whistle, there are a few efforts worth mentioning. For one, there’s a recent study [.pdf] that looks at self-organization in Tomb Raider: Underworld. By recording and analyzing the behavior of 1,365 people playing the game, they offer an automated approach to the “traditional user and play testing procedures.” And, notably, also evaluate whether players “play the game as intended by the game design.”

Elsewhere, in the magical world of business and economics, several companies are involved in the same effort. Recently I’ve been talking to Chethan Ramachandran, CEO at Turiya Ventures. His company has a team of hardcore data modelers on board who focus on behavior in virtual worlds. While I’m not allowed to say who, Turiya has been gaining traction with a bunch of big name companies. I’m curious to see their results, because they operate at a much larger scale.

This type of research is still in its infancy, but clearly offers a lot of potential toward understanding people’s behavior in digital environments and online communities.