Another interesting set of findings to lop on top of this is that dissatisfaction (diminishing probability x magnitude of reward) can be an antecedent to task creativity, as can mild depression, while mania can be an antecedent for behavioral stereotypy (antithesis of creativity), and manic-depressive types swing in behavioral variability on tasks depending on their state.

One interpretation is that dissatisfaction is a problem or environment exploration mechanism, and that behavioral stereotypy is the proper response for optimizing on a stable reward.

This has also been found in rats. As you lower tweak the probability x magnitude of reward, rat behaviors become more variable.

Different kinds of reward scheduling has also been tested with children suffering from disorders where behavioral variability or stereotypy is a feature, such as autism, with similar findings but no obvious application for treatment.