Learning in games with risky payoffs

Aric Shafran

Games and Economic Behavior, 2012, vol. 75, issue 1, 354-371

Abstract: This paper analyzes learning in multi-player noncooperative games with risky payoffs. The goal of the paper is to assess the relative importance of stochastic payoffs and expected payoffs in the learning process. A general learning model which nests several variations of reinforcement learning, belief-based learning, and experience-weighted attraction learning is used to analyze behavior in coordination game and prisonerʼs dilemma experiments with probabilistic payoffs. In all experiments, some subjects learn from past lottery outcomes, though the importance of these stochastic payoffs relative to expected payoffs depends on the game. Stochastic payoffs are less important when posted probabilities are equal to expected payoffs and more important when subjects are informed how much they would have earned from foregone strategies.

Keywords: Learning in games; Reinforcement learning; Fictitious play; Experiments; Stochastic payoffs (search for similar items in EconPapers)
JEL-codes: C72 C91 D81 (search for similar items in EconPapers)
Date: 2012
References: View references in EconPapers View complete reference list from CitEc
Citations: View citations in EconPapers (3)

Downloads: (external link)
http://www.sciencedirect.com/science/article/pii/S0899825611001485
Full text for ScienceDirect subscribers only

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:eee:gamebe:v:75:y:2012:i:1:p:354-371

DOI: 10.1016/j.geb.2011.09.001

Access Statistics for this article

Games and Economic Behavior is currently edited by E. Kalai

More articles in Games and Economic Behavior from Elsevier
Bibliographic data for series maintained by Catherine Liu ().