Stochastic optimal control and the human oculomotor system

Neuroscientists have long been interested in how efficiently we solve probabilistic sensory problems. In order to explore analogous questions in the motor domain, we observed the eye movements of human subjects attempting to track a visual target which moved stochastically across a computer screen. The subjects' behavior was then compared to a mathematically-derived bound on the best performance possible in such a task. The subjects were able to perform surprisingly near the optimum under the conditions examined. These results constitute an important step in determining the efficiency of the nervous system in the context of ongoing behavior.