Figure 2.

Optimal task strategy. (A) For trials in which the penalty value was 0, the optimal strategy was to rotate the payoff display so that the center of the reward region was aligned with the mean orientation of the texture (indicated here by the black arrow). (B) In non-zero penalty conditions, as the task became more difficult due to higher penalty, increased spread s_l of stimulus line orientations, and/or decreased distance between the payoff and penalty regions, the optimal strategy required the observer to rotate the payoff display so as to move the penalty region further away from the mean texture orientation.