Working in the Data Mines: Uncovering Gameplay Gold

Lecturer: Darius Kazemi, Orbus Gameworks

When designing and maintaining an MMO, developers are faced with a litany of questions. Do mages level faster than fighters? What vehicles are being neglected? Which players are hubs of our social networks? What areas are densely populated? How can we detect gold farmers?

These questions are often answered with intuition—and these answers often introduce assumptions that, while intuitive, are also dead wrong. Data metrics can provide a wealth of quantitative information about player behavior, provided that there is a solid technical backbone and that an appropriate methodology is followed.

This session will cover real-world examples of how data metrics can crush commonly-held assumptions about player behavior and give designers a new perspective on how their virtual worlds really work. The speaker will show attendees how to ask the right questions of their game so they can get the most out of a metrics system, covering basic census information, combat metrics, macro- and microeconomics, and social network visualization. Also covered will be methods for correlating player behavior to subscription lifecycle information to stop player churn before it starts.

Finally, the lecture will touch on the ways that data metrics can assist behind the scenes in bug hunting, customer service, billing, and even network operations.

Intended Audience: Game developers who want to make decisions with quantitative backing.
Prerequisites: Attendees should have solid understanding of design for virtual worlds.
Format: Lecture
Date/Time: Thursday, 10:00 AM
Room: Parliament