The present paper provides a broad overview of the three main dynamic models-the rigid body model, the deformation (deformable body) model, and the fluid structure model-that are used when investigating physical exercise and sports techniques from a biomechanical or ergonomic standpoint. The benefits of such models and the precautions that should be taken in their use and interpretation are investigated, as are some potential future models. Rigid body system models, among which the mass-spring model and rink-segment model are the most common, have conventionally been the core models in the fields of physical education and sport. This type of model will likely continue to evolve and be applied to various problems. At the same time, it is true that rigid body models are ill-suited to the study of certain techniques. For this reason, deformation and fluid structure models have been developed for use in such research and for problem solving. Analyses using conventional physical education and sport models usually describe sporting techniques in detail by calculating joint torque based on kinematic data from actual movement coordinates in an inverse dynamic manner (inputting displacement and outputting force). This has, to date, produced outstanding results, providing a “description” of the action taking place. In future, analyses that incorporate direct dynamics (inputting force and outputting displacement) based on elicited force will also be required in conjunction with this descriptive approach. A combination of these two approaches will enable simple “predictions” to be carried out on an appropriate level. The ultimate goal of such predictions will be to optimize the performance of the technique. In future, therefore, it will be important to perform modeling that incorporates description and prediction, yielding a model for optimization of the action.