Acquisition of novel and complex motor skills: stable solutions where intrinsic noise matters less.

While professional quarterbacks like Peyton Manning make passing a football look effortless, it takes a lifetime of disciplined practice for these elite athletes to reach such a high level of skill. The principal responsibility of the quarterback in American football is to throw a forward pass to his teammate. This skill, however, is not as trivial as it may appear. To achieve a successful pass, he must accurately throw to a target that is moving, while he himself is in motion. But even before the pass is thrown, he must quickly decide which receiver will have the best chance of catching the ball, judging from the continuous movements of the defensive team. He must also choose what style of throw and what initial conditions (i.e., position, velocity) are appropriate for where he aims the ball. This decision-making process cannot last more than a fraction of a second or he risks being tackled by opponents twice his size. This ability to choose from a vast—in fact infinite—set of actions is an integral part of many sport skills. Such complexity is what makes events like the Olympic Games fascinating, not only to spectators around the globe, but also to movement scientists, who want to unravel how humans achieve coordinated actions. How does the brain acquire and control such complex spatiotemporal skills? It is somewhat sobering to then look into the laboratories and see how motor skill is typically studied. The vast majority of experiments have focused on highly simplified tasks, such as finger-to-thumb opposition (Karni et al. 1998), finger tapping (Wing and Kristofferson 1973), sequenced isometric contractions (Shim et al. 2005;