Player segmentation: a practical mixed method approach
Data-driven player segmentations are one of the best ways to reduce the complexity of a community of millions of players into a manageable number of understandable player types, but their creation and application can be more art than science. This talk will cover the method I’ve used to build highly impactful player segments on Halo, Destiny, and Overwatch using machine learning, surveys, and user interviews. I’ll discuss how to build the right dataset, which tools to use, and how to judge whether your segments are really capturing the heart of your playerbase. Most importantly, I’ll cover my experience evangelizing the resulting player types to development teams. The best segmentation in the world is useless if it’s not internalized and acted upon by the design team, so my method (and my talk) has a heavy focus on producing segments that are both true reflections of the players and easy for non-researchers to understand and apply to make better games.
- Category: Talk
John has been a games researcher since 2003, and has worked on games ranging from small indie projects to massive blockbusters like Halo, Destiny, Hearthstone, and Overwatch. His specialty is combining analytics and traditional lab research techniques to produce insights that neither method could accomplish alone. He is a former chair of the Games User Research SIG of the IGDA and holds a Ph.D. in Behavioral and Brain Sciences from Duke University. He is currently the senior manager of user research at Blizzard Entertainment, responsible for research work on all of Blizzard’s game franchises.