Simultaneous games with purchase of randomly supplied perfect information: Oracle Games

We study the role of costly information in non-cooperative two-player games when an extrinsic third party information broker is introduced asymmetrically, allowing one player to obtain information about the other player's action. This broker or "oracle" is defined by a probability of response, supplying correct information randomly; the informed player can pay more for a higher probability of response. We determine the necessary and sufficient conditions for strategy profiles to be equilibria, in terms of how both players change their strategies in response to the existence of the oracle, as determined by its cost of information function. For mixed strategy equilibria, there is a continuous change as information becomes cheaper, with clear transitions occuring at critical {\it nodes} at which pure strategies become dominated (or undominated). These nodes separate distinct responses to the information for sale, alternating between regions where the paying player increases the amount of information purchased, and regions where the other player moves away from riskier strategies, in favor of safer bets that minimize losses. We derive conditions for these responses by defining a value of information.

[1]  Arthur J. Robson,et al.  Reinterpreting mixed strategy equilibria: a unification of the classical and Bayesian views , 2004, Games Econ. Behav..

[2]  César Martinelli,et al.  Rational ignorance and voting behavior , 2007, Int. J. Game Theory.

[3]  Joseph Y. Halpern,et al.  Game theory with translucent players , 2013, TARK.

[4]  S. Morris,et al.  Social Value of Public Information , 2002 .

[5]  David Delgado-Gómez,et al.  Cournot–Stackelberg games in competitive delocation , 2017, Ann. Oper. Res..

[6]  J. Kagel,et al.  The Effects of Pre-Play Communication in a Coordination Game with Incomplete Information , 2018, Journal of Economic Behavior & Organization.

[7]  Leslie B. Vosshall,et al.  The Survival Advantage of Olfaction in a Competitive Environment , 2008, Current Biology.

[8]  Alexandros Rigos,et al.  Flexible Information Acquisition in Large Coordination Games , 2018 .

[9]  Marco Tomassini,et al.  Know Thy Neighbor: Costly Information Can Hurt Cooperation in Dynamic Networks , 2014, PloS one.

[10]  John C. Harsanyi,et al.  Games with Incomplete Information Played by "Bayesian" Players, I-III: Part I. The Basic Model& , 2004, Manag. Sci..

[11]  Elchanan Ben-Porath,et al.  Communication in repeated games with costly monitoring , 2003, Games Econ. Behav..

[12]  David P. Myatt,et al.  Cournot competition and the social value of information , 2015 .

[13]  David P. Myatt,et al.  Endogenous Information Acquisition in Coordination Games , 2009 .

[14]  Steven D. Levitt,et al.  Testing Mixed-Strategy Equilibria When Players Are Heterogeneous: The Case of Penalty Kicks in Soccer , 2002 .

[15]  Kfir Eliaz,et al.  Paying for Confidence: An Experimental Study of the Demand for Non-Instrumental Information , 2009, Games Econ. Behav..

[16]  Heiner Schumacher,et al.  The Value of Recommendations , 2012, Games Econ. Behav..

[17]  John Dennis McDonald Strategy in Poker, Business and War , 1989 .

[18]  Michal Szkup,et al.  Information acquisition in global games of regime change , 2015, J. Econ. Theory.

[19]  Christian Hellwig,et al.  Knowing What Others Know: Coordination Motives in Information Acquisition (March 2007, with Laura Veldkamp) , 2005 .

[20]  Ming Yang,et al.  Coordination with Flexible Information Acquisition , 2014 .

[21]  M. Gloria Fiestras-Janeiro,et al.  An Introductory Course on Mathematical Game Theory , 2010 .

[22]  David I. Laibson,et al.  Costly Information Acquisition: Experimental Analysis of a Boundedly Rational Model , 2006 .

[23]  Yasuhiro Sakai Cournot and bertrand equilibria under imperfect information , 1986 .

[24]  Leeat Yariv,et al.  Games with espionage , 2004, Games Econ. Behav..

[25]  Andrés Perea,et al.  Repeated games with voluntary information purchase , 2009, Games Econ. Behav..