Physical interaction between humans can be studied by haptically connecting individuals via robotic interfaces. Our previous work showed that pairs of individuals track targets more accurately while haptically connected during ankle movements; changes in performance were dependent on the ability of each partner and the virtual connection stiffness. We modelled the interaction as a series of springs, considering the stiffness of the connection and each partners’ ankle, to show that improvements were likely due to an attenuation of random tracking errors between partners. In this study, our goal was to expand this simulation by exploring other dyadic behaviors as inputs to our model. In particular, we focus on the ankle stiffness of partners to show how tracking strategies and physiology can influence performance during dyadic tracking tasks. These results provide additional context for our published model and potential explanations for upper- and lower-limb dyadic tracking behaviors in the literature.