Simple Models of Discrete Choice and Their Performance in Bandit Experiments

  • Authors:
  • Noah Gans;George Knox;Rachel Croson

  • Affiliations:
  • OPIM Department, The Wharton School, University of Pennsylvania, 3730 Walnut Street, Suite 500, Philadelphia, Pennsylvania 19104;Department of Marketing, Tilburg University, Room K 1019, P.O. Box 90153, 5000 LE Tilburg, The Netherlands;OPIM Department, The Wharton School, University of Pennsylvania, 3730 Walnut Street, Suite 500, Philadelphia, Pennsylvania 19104

  • Venue:
  • Manufacturing & Service Operations Management
  • Year:
  • 2007

Quantified Score

Hi-index 0.00

Visualization

Abstract

Recent operations management papers model customers as solving multiarmed bandit problems, positing that consumers use a particular heuristic when choosing among suppliers. These papers then analyze the resulting competition among suppliers and mathematically characterize the equilibrium actions. There remains a question, however, as to whether the original customer models on which the analyses are built are reasonable representations of actual consumer choice. In this paper, we empirically investigate how well these choice rules match actual performance as people solve two-armed Bernoulli bandit problems. We find that some of the most analytically tractable models perform best in tests of model fit. We also find that the expected number of consecutive trials of a given supplier is increasing in its expected quality level, with increasing differences, a result consistent with the models' predictions as well as with loyalty effects described in the popular management literature.