Elicitation of Probabilities Using Competitive Scoring Rules

D. Marc Kilgour () and Yigal Gerchak ()
Additional contact information
D. Marc Kilgour: Department of Mathematics, Wilfrid Laurier University, Waterloo, Ontario, Canada N2L 3C5
Yigal Gerchak: Department of Industrial Engineering, Tel-Aviv University, Tel-Aviv 69978, Israel

Decision Analysis, 2004, vol. 1, issue 2, 108-113

Abstract: Several forecasters predict the probability of an event, and then make or receive payments contingent on their predictions and on whether the event actually occurs. The payment functions generalize the concept of scoring rule to a competitive setting. We allow for exogenously determined subsidies to each forecaster, and require that the scheme be anonymous, neutral, and truth-inducing. By centering each forecaster's payment at the average payment to all other forecasters, we construct competitive scoring rules that reward the better predictors. Applications include multiparty betting and fixed-budget surveys to determine subjects' truthful probability assessments. We discuss when forecasters would voluntarily participate in such a competition, and relate our results to the scoring rules proposed by De Finetti (1974) for eliciting probabilities.

Keywords: elicitation of probabilities; prediction of probabilities; competitive prediction; scoring rules; multiparty bets (search for similar items in EconPapers)
Date: 2004
References: View references in EconPapers View complete reference list from CitEc
Citations: View citations in EconPapers (30)

Downloads: (external link)
http://dx.doi.org/10.1287/deca.1030.0003 (application/pdf)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:inm:ordeca:v:1:y:2004:i:2:p:108-113

Access Statistics for this article

More articles in Decision Analysis from INFORMS Contact information at EDIRC.
Bibliographic data for series maintained by Chris Asher ().