What are the chances of winning? Exploring the ecology and psychology

Emre Soyer (Ozyegin University) will talk about "What are the chances of winning? Exploring the ecology and psychology" (j/w Robin M. Hogarth -- see abstract below)
Where: BELSS -- our experimental lab on the 3rd floor of the Roentgen building
When: Wednesday March 19 at 17.30-18.20+
Afterwards: we go for a post-seminar aperitivo

In competitions, n competitors are ranked on a criterion to determine k winners (k < n). We explore the ecology of competitions by modeling the probability of winning, taking account of the numbers of winners, competitors, and relative skill levels. We find that across different situations and skill profiles, probabilities cluster at the extremes (close to 0 and 1). Moreover, this holds when we vary n, introduce luck in outcomes, and individuals are  overconfident. Using the multiple-cue probability learning paradigm to test whether people can learn to make appropriate judgments, we find that judgments reflect linear aggregation of informational cues but are biased:  estimates do not cluster at 0 and 1; and there is little learning. Moreover, choices to enter competitions reflect this bias. Since the appropriate normative calculations are demanding, we investigate the use of heuristic rules and  demonstrate their efficacy. Finally, we discuss variations of our model and suggest extensions of our paradigm.

We hope many will attend. If you have questions, contact Martin Dufwenberg (martin.dufwenberg@unibocconi.it), Joshua Miller (joshua.miller@unibocconi.it) or Fabrizio Iozzi (fabrizio.iozzi@unibocconi.it).