Sport competitions provide a rich test bed for investigating the complex adaptive behavior of interpersonal interactions. Most competitive sports can be characterized by game rhythms, e.g., the alternating volleys of tennis and squash or the ebb and flow movement from one end of the field to another in soccer or basketball. Further, this research has noted that the rhythmic play of individuals and teams tends to be synchronized and can be modeled by a coupled oscillatory regime similar to that used to model interpersonal interlimb coordination. Importantly, however, although much of sport competition consists of stable states of interpersonal coordination, these stable states are punctuated by intentionally introduced symmetry breaking perturbations that involve an individual trying to take advantage of another player/team to score points. A number of studies have investigated these behavioral bifurcations in attacker-defender dyads during squash, basketball and boxing. In our current work, we use a martial arts interaction as an exemplar system to develop a mathematical model of the symmetry breaking and switching dynamics in a joint action that is comprised of multiple behavioral states. Participants of different skill levels in the martial arts perform an Aikido technique in which a differential weighting of the wrists is used across trials to determine whether the relative timing of participants’ movements followed the patterns predicted by an oscillatory dynamical model. Understanding these steady-state dynamics of the initial part of this technique will provide a basis for investigating how an Aikido defender uses them the dynamics to “take the advantage” in subsequent parts of the technique and can provide a basis for a mathematical model of the symmetry breaking and switching dynamics involved.