Cooperation is widespread in the animal kingdom. Studies have sought to explain the evolution of cooperation through kinship, reciprocal interactions, and spatial structures arising from limited dispersal of offspring. However, animals across taxa live in complex dynamic social groups that constantly merge and split. In such cases, these mechanisms cannot be invoked to explain cooperation. Using individual-based simulations, we show how emergent differences in the movement characteristics of cooperators and defectors can lead to the evolution of cooperation without communication, kinship, reciprocity and limited dispersal. I propose an empirical test of our model in a semi-virtual system with real fish predators and virtual simulated prey.