Grouping in animals is often regarded as a strategy to reduce risks of predation. Grouping animals exhibit a variety of movement patterns, from apparently incoherent swarms to spectacular coordinated movement of flocking birds and schooling fish. Investigation of costs and benefits of collective movement patterns in experimental or field systems is difficult due to various confounding factors. In this talk, I will present how we used a fusion of real predators with virtual prey to investigate predation risk driven selection for collective motion (based on Ioannou et al, 2012, Science). We found that virtual prey that attracted towards, and aligned its direction of motion with neighbours were at least risk of predation. Our work showed that coordinated collective movement may evolve in response to predation risk even when prey do not detect or respond to predation. I will discuss some of our ongoing work on employing virtual prey to study collective animal behaviour.