This chapter discusses the nuts and bolts of player modeling from the three perspectives of input, output, and algorithm. Input is subdivided into gameplay input, objective input (including physiological signals), game context, and player profile. The model’s output, or that which is predicted, can be of various types. Some player models simply learn embeddings that can be used to segment players, other models predict high-level player behavior or various aspects of player experience. The chapter also contains a discussion of principles and tools for labeling data for player modeling. A multitude of machine learning methods can be used to model players, and this chapter discusses supervised learning, reinforcement learning, and unsupervised learning, as well as manual model authoring.