Predicting and Understanding Initial Play

39 Pages Posted: 28 Nov 2017 Last revised: 16 Jun 2019

See all articles by Drew Fudenberg

Drew Fudenberg

Massachusetts Institute of Technology (MIT)

Annie Liang

University of Pennsylvania - Department of Economics

Date Written: March 13, 2019

Abstract

We use machine learning to uncover regularities in the initial play of matrix games. We first train a prediction algorithm on data from past experiments. Examining the games where our algorithm predicts correctly, but existing economic models don’t, leads us to add a parameter to the best performing model that improves predictive accuracy. We obtain even better predictions with a hybrid model that uses a decision tree to decide game-by-game which of two economic models to use for prediction. Finally, we explore the usefulness of crowd-sourced predictions for making better predictions, and for discovering additional relevant game features.

Keywords: prediction, machine learning, behavioral game theory, cognitive hierarchy

JEL Classification: C72, D03, C51, C91

Suggested Citation

Fudenberg, Drew and Liang, Annie, Predicting and Understanding Initial Play (March 13, 2019). PIER Working Paper No. 18-009. Available at SSRN: https://ssrn.com/abstract=3076682 or http://dx.doi.org/10.2139/ssrn.3076682

Drew Fudenberg

Massachusetts Institute of Technology (MIT) ( email )

77 Massachusetts Avenue
50 Memorial Drive
Cambridge, MA 02139-4307
United States

Annie Liang (Contact Author)

University of Pennsylvania - Department of Economics ( email )

Ronald O. Perelman Center for Political Science
133 South 36th Street
Philadelphia, PA 19104-6297
United States

Register to save articles to
your library

Register

Paper statistics

Downloads
267
Abstract Views
984
rank
112,785
PlumX Metrics