Choosing a Good Toolkit, I: Formulation, Heuristics, and Asymptotic Properties

Abstract

The problem of choosing an optimal toolkit day after day, when the distribution of values of different toolkits is uncertain and can only be observed by carrying different toolkits, is a multi-armed bandit problem with non-independent arms. Accordingly, except for very simple specifications, this problem cannot (practically) be solved. Decision makers facing this problem presumably resort to “sensible” decision heuristics, employing past experience and, perhaps, what they know about the problem. In this paper, Part I of a two-part study of this situation, we provide a general formulation the problem, describe a variety of heuristics, and provide some theoretical results about how the heuristics perform asymptotically. (In the second paper, we compare the performance of these heuristics through simulations.)

Keywords: Heuristics, multi-armed bandits, behavioral decision making

JEL Classification: C63, D03, D83, D90