Player types emerge from unsupervised learning
itu.dk
Lots of data about your playing style are stored on central servers when you play e.g. Tomb Raider. This data can be mined to discover player types, and automatically classify players, which can in turn be used to adapt games to suit you better.