Predicting and Understanding Initial Play

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.

[1]  TWO-WEEK Loan COpy,et al.  University of California , 1886, The American journal of dental science.

[2]  Jstor,et al.  Invention in the Industrial Research Laboratory , 1963, Journal of Political Economy.

[3]  R. Thaler,et al.  THE BEHAVIORAL LIFE‐CYCLE HYPOTHESIS , 1988 .

[4]  J. R. Quinlan Learning With Continuous Classes , 1992 .

[5]  D. Stahl,et al.  Experimental evidence on players' models of other players , 1994 .

[6]  Stephen Morris,et al.  P-dominance and belief potential , 2010 .

[7]  D. Stahl,et al.  On Players' Models of Other Players: Theory and Experimental Evidence , 1995 .

[8]  V. Crawford Adaptive dynamics in coordination games , 1995 .

[9]  Daniel Friedman,et al.  Individual Learning in Normal Form Games: Some Laboratory Results☆☆☆ , 1997 .

[10]  A. Roth,et al.  Predicting How People Play Games: Reinforcement Learning in Experimental Games with Unique, Mixed Strategy Equilibria , 1998 .

[11]  D. Fudenberg,et al.  The Theory of Learning in Games , 1998 .

[12]  Miguel A. Costa-Gomes,et al.  Cognition and Behavior in Normal-Form Games: An Experimental Study , 1998 .

[13]  Colin Camerer,et al.  Experience‐weighted Attraction Learning in Normal Form Games , 1999 .

[14]  Dale O. Stahl,et al.  Rule Learning in Symmetric Normal-Form Games: Theory and Evidence , 2000, Games Econ. Behav..

[15]  M. Rabin Risk Aversion and Expected Utility Theory: A Calibration Theorem , 2000 .

[16]  D. Stahl,et al.  Modeling and Testing for Heterogeneity in Observed Strategic Behavior , 2001, Review of Economics and Statistics.

[17]  A. Schotter,et al.  An Experimental Study of Belief Learning Using Elicited Beliefs , 2002 .

[18]  Charles A. Holt,et al.  Risk Aversion and Incentive Effects , 2002 .

[19]  D. Stahl,et al.  Level-n bounded rationality and dominated strategies in normal-form games , 2008 .

[20]  Eric R. Ziegel,et al.  The Elements of Statistical Learning , 2003, Technometrics.

[21]  Colin Camerer,et al.  A Cognitive Hierarchy Model of Games , 2004 .

[22]  Miguel A. Costa-Gomes,et al.  Stated Beliefs and Play in Normal-Form Games , 2007 .

[23]  Leo Breiman,et al.  Bagging Predictors , 1996, Machine Learning.

[24]  Eibe Frank,et al.  Logistic Model Trees , 2003, Machine Learning.

[25]  A. Timmermann Forecast Combinations , 2005 .

[26]  C. Granger,et al.  Handbook of Economic Forecasting , 2006 .

[27]  D. Fudenberg,et al.  A Dual Self Model of Impulse Control , 2004, The American economic review.

[28]  D. Stahl,et al.  Equilibrium selection and bounded rationality in symmetric normal-form games , 2007 .

[29]  R. Aumann,et al.  Unraveling in Guessing Games : An Experimental Study , 2007 .

[30]  Colin Camerer,et al.  Experienced-Weighted Attraction Learning in Normal Form Games , 2007 .

[31]  V. Crawford,et al.  Level-k Auctions: Can a Non-Equilibrium Model of Strategic Thinking Explain the Winner's Curse and Overbidding in Private-Value Auctions? , 2007 .

[32]  Robert L. Slonim,et al.  Learning and equilibrium as useful approximations: Accuracy of prediction on randomly selected constant sum games , 2007 .

[33]  D. Sgroi,et al.  Learning to play 3×3 games: Neural networks as bounded-rational players , 2009 .

[34]  Thomas R. Palfrey,et al.  Heterogeneous quantal response equilibrium and cognitive hierarchies , 2006, J. Econ. Theory.

[35]  Panagiotis G. Ipeirotis,et al.  Running Experiments on Amazon Mechanical Turk , 2010, Judgment and Decision Making.

[36]  B. D. Bruin,et al.  Normal Form Games , 2010 .

[37]  D. Fudenberg,et al.  Risk, Delay, and Convex Self-Control Costs , 2010 .

[38]  Alvin E. Roth,et al.  A Choice Prediction Competition for Social Preferences in Simple Extensive Form Games: An Introduction , 2011, Games.

[39]  R. Raghavan Is it strategic , 2011 .

[40]  Asen Ivanov,et al.  Attitudes to ambiguity in one-shot normal-form games: An experimental study , 2011, Games Econ. Behav..

[41]  J. Doug Tygar,et al.  Adversarial machine learning , 2019, AISec '11.

[42]  Reza Ebrahimpour,et al.  Mixture of experts: a literature survey , 2014, Artificial Intelligence Review.

[43]  David G. Rand,et al.  The promise of Mechanical Turk: how online labor markets can help theorists run behavioral experiments. , 2012, Journal of theoretical biology.

[44]  Miguel A. Costa-Gomes,et al.  Structural Models of Nonequilibrium Strategic Thinking: Theory, Evidence, and Applications , 2013 .

[45]  Kevin Leyton-Brown,et al.  Level-0 meta-models for predicting human behavior in games , 2014, EC.

[46]  Marco Del Negro,et al.  Dynamic Prediction Pools: An Investigation of Financial Frictions and Forecasting Performance , 2014 .

[47]  Thomas J. Leeper,et al.  The Generalizability of Survey Experiments* , 2015, Journal of Experimental Political Science.

[48]  Using Methods from Machine Learning to Evaluate Behavioral Models of Choice Under Risk and Ambiguity , 2015 .

[49]  Kevin Leyton-Brown,et al.  Deep Learning for Predicting Human Strategic Behavior , 2016, NIPS.

[50]  Teck-Hua Ho,et al.  A generalized cognitive hierarchy model of games , 2016, Games Econ. Behav..

[51]  Scott Clifford,et al.  Validity and Mechanical Turk: An assessment of exclusion methods and interactive experiments , 2017, Comput. Hum. Behav..

[52]  Annie Liang,et al.  The Theory is Predictive, but is it Complete?: An Application to Human Perception of Randomness , 2017, EC.

[53]  Mel W. Khaw,et al.  Temporal Discounting and Search Habits: Evidence for a Task-Dependent Relationship , 2018, Front. Psychol..

[54]  E. Snowberg,et al.  Testing the Waters: Behavior Across Participant Pools , 2018, American Economic Review.

[55]  Stefano DellaVigna,et al.  Predicting Experimental Results: Who Knows What? , 2016, Journal of Political Economy.

[56]  Alec Smith,et al.  Dynamic Unstructured Bargaining with Private Information: Theory, Experiment, and Outcome Prediction via Machine Learning , 2019, Manag. Sci..

[57]  EbrahimpourReza,et al.  Mixture-of-Experts , 2022, Encyclopedia of Big Data.