The application of principles of motor behavior to teaching and training methodologies is problematic because of task specificity. Although it is possible to classify tasks per se, there are grounds, one of which is explored here, for holding that classifying human performance requires an account of the person performing the task. An important consideration of the person is the person’s skill level. An experiment is presented that demonstrates that classification of tasks changes depending on the level of the performer’s skill. Three common taxonomies were evaluated with respect to the way in which they would classify people at different levels of skill in bouncing a basketball. Nine participants ranging from absolute beginners to one basketball-player expert were examined bouncing a basketball at their preferred rhythm, and the kinematics of their arms, torso, and legs were recorded and analyzed by digitized video tape. The open-closed scheme, in which predictability of stimulus is the measure, was operationally defined by the variability of lateral and vertical ball position at ball strike and period of the ball. The continuous-discrete scheme was operationally defined by number of consecutive cycles a participant could produce in a session. The gross-fine scheme was operationally defined by the amplitude and coefficient of variation of amplitude of the body segments. It was found that task classification for the “task” of “bouncing a ball” changes for each of the three classification schemes according to the skill level of the performer. For example, for a beginning ball bouncer, the task is open, but for the expert, it is closed. The results suggest that as taxonomists develop classification schemes for tasks, one crucial dimension to consider is the level of the performer’s skill. It is assumed that training methodologies must be sensitive to task type, as different types of tasks may require different pedagogical approaches. As researchers develop criteria for training methodologies, then, they must ask whether the taxonomy used is sensitive to skill level. As practitioners implement training and teaching techniques, they might ask this same question. There are many ways to classify human performance, but for researchers and practitioners interested in rationalizing theories of motor control and training methodologies, taxonomies should be developed that countenance skill-performer-task interactions.Keyword(s): curriculum development, interdisciplinary, research