Résumé

People working across disciplines compete for resources and recognition. The competition arenas are often systematized as contests, in which a group of individuals (i.e. a committee) stand on behalf of a larger population and decide which candidates will be selected. We simulate committee decisions in contests using datasets where many people have evaluated the same candidates. We draw raters to stand in the committee and determine the winners by aggregating their ratings. Our approach reveals that contests are lotteries—winning depends on the draw of committee members who will evaluate the candidates and their tastes. As winning depends on a fortuitous draw, the evaluations of the winning candidates are largely inflated when compared to the average perception of their quality in the population. Further, the chances of winning are skewed in favor of candidates with more extreme positive ratings as compared to candidates with high average ratings. Our results can help candidates assess their chances of winning and manage their expectations when participating in contests, and can inform the design of effective committees.

Référence

Pantelis Analytis (Danish Institute for Advanced Studies), « The committee curse: why winners in contests are largely overrated », IAST General Seminar, Toulouse : IAST, 12 novembre 2024, 11h30–12h30, salle Auditorium 4 (First floor - TSE building).