Effort/performance tradeoff.

<p>The left panel presents normalized RMS tracking error against RMS control input velocity for young subjects (1–12), showing the existence of a tradeoff between effort and performance. The displayed line traces an effort/performance tradeoff implied by the model, and borne out in the data for the young subjects. Subjects well fit by the optimal control model (see the optimal control results section) are marked with circles. Subjects 1 and 10 (shown with triangles) are outliers in this panel and were not well fit by the optimal control model. They did not reproduce the outlying behavior shown here during repeated trials as shown in <a href="http://www.ploscompbiol.org/article/info:doi/10.1371/journal.pcbi.1000708#pcbi.1000708-Sherback1" target="_blank">[25]</a>. The effort/performance tradeoff line is obtained by varying the control cost parameter in the optimization while also adjusting the noise intensity to account for the multiplicative or “signal dependent” nature of the noise. The right panel shows that the elderly (13–24) are generally unable to obtain as favorable an effort-performance tradeoff. The displayed reference line is not a fit to the data. Rather, it is the same line as in the young data, provided for comparison. Again subjects well fit by an optimal control model are indicated with circles, while subjects 19, 22, and 24 are indicated with triangles.