John M. C. Hutchinson (Max Planck Institute for Human Development): Rules of thumb and optimality

We often invoke rules of thumb as a hand-waving explanation for why an optimality model has failed to make accurate predictions. But I will demonstrate that incorporating constraints into optimality models provides useful suggestions about what sorts of rules of thumb we expect to observe. As an example I consider patch-leaving rules for foragers. I will give some other examples of rules of thumb that have been derived empirically, many of them from social insects. In most case empirical evidence is restricted to what cues are involved, rather than very quantitative information on the form of the rules. Finally I may introduce the research of Gigerenzer's ABC group showing that simple heuristics can perform as well or better than much more complex procedures.