The Cascade model of cognitive skill acquisition was developed to integrate a number of AI techniques and to account for psychological results on the self-explanation effect. In previous work, the authors compared Cascade's behavior in aggregate data collected from the protocols of subjects in a self-explanation study. Here, they report the results of a fine-grained analysis, in which Cascade's behavior is matched to the individual protocols of each of the subjects. The analyses demonstrate empirically that Cascade is a good model of subject behavior at the level of goals and inferences. It covers about 75% of the subjects' example-studying behavior and 60% to 90% of their problem-solving behavior. This research also led the authors to develop general feasible methods for matching a simulation to large protocols (approximately 3000 pages total). Finally, the analyses point out some weaknesses in the Cascade system and provide us with directions for future analyses of the model and data.