Learning and Communication in Sender-Reciever Games

An Economic Investigation

A. Blume, D.V. DeJong, G. Neumann, N.E. Savin

Research output: Working paperDiscussion paperOther research output

224 Downloads (Pure)

Abstract

This paper compares the performance of stimulus response (SR) and belief-based learning (BBL) using data from game theory experiments. The environment, extensive form games played in a population setting, is novel in the empirical literature on learning in games. Both the SR and BBL models fit the data reasonably well in common interest games with history while the test results accept SR and reject BBL in games with no history and in all but one of the divergent interest games. Estimation is challenging since the likelihood function is not globally concave and the results may be subject to convergence bias.
Original languageEnglish
Place of PublicationTilburg
PublisherMicroeconomics
Number of pages48
Volume2000-09
Publication statusPublished - 2000

Publication series

NameCentER Discussion Paper
Volume2000-09

Fingerprint

History
Economics
Communication
Game theory
Experiments

Keywords

  • econometrics
  • game theory and experiments

Cite this

Blume, A., DeJong, D. V., Neumann, G., & Savin, N. E. (2000). Learning and Communication in Sender-Reciever Games: An Economic Investigation. (CentER Discussion Paper; Vol. 2000-09). Tilburg: Microeconomics.
Blume, A. ; DeJong, D.V. ; Neumann, G. ; Savin, N.E. / Learning and Communication in Sender-Reciever Games : An Economic Investigation. Tilburg : Microeconomics, 2000. (CentER Discussion Paper).
@techreport{138dc36b5269421a9e79bd854c5a9baf,
title = "Learning and Communication in Sender-Reciever Games: An Economic Investigation",
abstract = "This paper compares the performance of stimulus response (SR) and belief-based learning (BBL) using data from game theory experiments. The environment, extensive form games played in a population setting, is novel in the empirical literature on learning in games. Both the SR and BBL models fit the data reasonably well in common interest games with history while the test results accept SR and reject BBL in games with no history and in all but one of the divergent interest games. Estimation is challenging since the likelihood function is not globally concave and the results may be subject to convergence bias.",
keywords = "econometrics, game theory and experiments",
author = "A. Blume and D.V. DeJong and G. Neumann and N.E. Savin",
note = "Pagination: 48",
year = "2000",
language = "English",
volume = "2000-09",
series = "CentER Discussion Paper",
publisher = "Microeconomics",
type = "WorkingPaper",
institution = "Microeconomics",

}

Blume, A, DeJong, DV, Neumann, G & Savin, NE 2000 'Learning and Communication in Sender-Reciever Games: An Economic Investigation' CentER Discussion Paper, vol. 2000-09, Microeconomics, Tilburg.

Learning and Communication in Sender-Reciever Games : An Economic Investigation. / Blume, A.; DeJong, D.V.; Neumann, G.; Savin, N.E.

Tilburg : Microeconomics, 2000. (CentER Discussion Paper; Vol. 2000-09).

Research output: Working paperDiscussion paperOther research output

TY - UNPB

T1 - Learning and Communication in Sender-Reciever Games

T2 - An Economic Investigation

AU - Blume, A.

AU - DeJong, D.V.

AU - Neumann, G.

AU - Savin, N.E.

N1 - Pagination: 48

PY - 2000

Y1 - 2000

N2 - This paper compares the performance of stimulus response (SR) and belief-based learning (BBL) using data from game theory experiments. The environment, extensive form games played in a population setting, is novel in the empirical literature on learning in games. Both the SR and BBL models fit the data reasonably well in common interest games with history while the test results accept SR and reject BBL in games with no history and in all but one of the divergent interest games. Estimation is challenging since the likelihood function is not globally concave and the results may be subject to convergence bias.

AB - This paper compares the performance of stimulus response (SR) and belief-based learning (BBL) using data from game theory experiments. The environment, extensive form games played in a population setting, is novel in the empirical literature on learning in games. Both the SR and BBL models fit the data reasonably well in common interest games with history while the test results accept SR and reject BBL in games with no history and in all but one of the divergent interest games. Estimation is challenging since the likelihood function is not globally concave and the results may be subject to convergence bias.

KW - econometrics

KW - game theory and experiments

M3 - Discussion paper

VL - 2000-09

T3 - CentER Discussion Paper

BT - Learning and Communication in Sender-Reciever Games

PB - Microeconomics

CY - Tilburg

ER -

Blume A, DeJong DV, Neumann G, Savin NE. Learning and Communication in Sender-Reciever Games: An Economic Investigation. Tilburg: Microeconomics. 2000. (CentER Discussion Paper).