Prior Distributions for the Bradley-Terry Model of Paired Comparisons

John T. Whelan
Page number: 
Publication date: 
December 14, 2017
E-print ID: 

The Bradley-Terry model assigns probabilities for the outcome of paired comparison experiments based on strength parameters associated with the objects being compared. We consider different proposed choices of prior parameter distributions for Bayesian inference of the strength parameters based on the paired comparison results. We evaluate them according to four desiderata motivated by the use of inferred Bradley-Terry parameters to rate teams on the basis of outcomes of a set of games: invariance under interchange of teams, invariance under interchange of winning and losing, normalizability and invariance under elimination of teams. We consider various proposals which fail to satisfy one or more of these desiderata, and illustrate two proposals which satisfy them. Both are one-parameter independent distributions for the logarithms of the team strengths: 1) Gaussian and 2) Type III generalized logistic.

Home | Links | Credits | Search | Terms of Use | Disclaimer | My Account 

Center for Computational Relativity and Gravitation, School of Mathematical Sciences,
Rochester Institute of Technology,
One Lomb Memorial Drive, Rochester, New York 14623, USA
Phone: (585) 475-7752
Fax: (585) 475-7340
Education - This is a contributing Drupal Theme
Design by WeebPal.