Small group forecasting using proportional-prize contests

Ronald Peeters, Fan Rao, Leonard Wolk*

*Corresponding author for this work

Research output: Contribution to JournalArticleAcademicpeer-review

Abstract

We consider a proportional-prize contest to forecast future events, and show that, in equilibrium, this mechanism possesses perfect forecasting ability for any group size when the contestants share common knowledge about the probabilities by which future events realize. Data gathered in a laboratory experiment confirm the performance invariance to group size. By contrast, when realization probabilities are not common knowledge, there are some differences across group sizes. The mechanism operates marginally better with three or four compared to two players. However, this effect is mainly driven by players’ behavior rather than by differences in the beliefs they form about the realization probabilities.
Original languageEnglish
Pages (from-to)293-317
Number of pages25
JournalTheory and Decision
Volume92
Issue number2
DOIs
Publication statusPublished - 15 Feb 2022

Bibliographical note

Publisher Copyright:
© 2021, The Author(s).

Keywords

  • Behavioral mechanism design
  • Contests
  • Experiment
  • Forecasting

Fingerprint

Dive into the research topics of 'Small group forecasting using proportional-prize contests'. Together they form a unique fingerprint.

Cite this