Users of virtual worlds shape their behavior to what the environment supports. To ensure adequate transfer of training and to elicit realistic behavior in virtual environments, it is essential to understand the informational content of the perceptual array that users interact with. In the scientific community, two approaches are in use to resolve this matter. One, using regulatory models, assumes the availability of control-relevant information and focuses on dynamics. The other, using perceptual models, emphasizes the kinematics of locomotion and often ignores the dynamics of controlling perceptual cues through manipulation of locomotion. Information and dynamics in interaction, however, are intricately coupled and should be analyzed as a whole. It is shown that perceptual and regulatory models can be integrated by adopting a cybernetic approach, where humans are modeled as dynamic information-processing systems that interact with perceptual cues present in the environment.