2014arXiv (Cornell University)Open access

How much is convenient to defect? A method to estimate the cooperation probability in Prisoner's Dilemma and other games

Cesco Reale

Open full text 0 citations

Abstract

In many cases the Nash equilibria are not predictive of the experimental players' behaviour. For some games of Game Theory it is proposed here a method to estimate the probabilities with which the different options will be actually chosen by balanced players, i.e. players that are neither too competitive, nor too cooperative. This will allow to measure the intrinsec cooperativeness degree of a game, only in function of its payoffs. The method is shaped on the Prisoner's Dilemma, then generalized for asymmetric tables, N players and N options. It is adapted to other conditions like Chicken Game, Battle of the Sexes, Stag Hunt and Translators (a new name proposed for a particular condition). Then the method is applied to other games like Diner's Dilemma, Public Goods Game, Traveler's Dilemma and War of Attrition. These games are so analyzed in a probabilistic way that is consistent to what we could expect intuitively, overcoming some known paradoxes of the Game Theory.

Open-access reader

About this research paper

What this paper is about

In many cases the Nash equilibria are not predictive of the experimental players' behaviour. For some games of Game Theory it is proposed here a method to estimate the probabilities with which the different options will be actually chosen by balanced players, i.e. players that are neither too competitive, nor too cooperative. This will allow to measure the intrinsec cooperativeness degree of a game, only in function of its payoffs. The method is shaped on the Prisoner's Dilemma, then generalized for asymmetric tables, N players and N options. It is adapted to other conditions like Chicken Game, Battle of the Sexes, Stag Hunt and Translators (a new name proposed for a particular condition). Then the method is applied to other games like Diner's Dilemma, Public Goods Game, Traveler's Dilemma and War of Attrition. These games are so analyzed in a probabilistic way that is consistent to what we could expect intuitively, overcoming some known paradoxes of the Game Theory.

Why it matters

A significance statement is not available in the OpenAlex record.

Key contribution

A contribution statement is not available in the OpenAlex record.

Method / approach

Method details are not available in the OpenAlex metadata.

Main findings

Findings are not separately available in the OpenAlex metadata.

Limitations

Limitations are not available in the OpenAlex metadata.

Applications

Application details are not available in the OpenAlex metadata.

Available abstract

In many cases the Nash equilibria are not predictive of the experimental players' behaviour. For some games of Game Theory it is proposed here a method to estimate the probabilities with which the different options will be actually chosen by balanced players, i.e. players that are neither too competitive, nor too cooperative. This will allow to measure the intrinsec cooperativeness degree of a game, only in function of its payoffs. The method is shaped on the Prisoner's Dilemma, then generalized for asymmetric tables, N players and N options. It is adapted to other conditions like Chicken Game, Battle of the Sexes, Stag Hunt and Translators (a new name proposed for a particular condition). Then the method is applied to other games like Diner's Dilemma, Public Goods Game, Traveler's Dilemma and War of Attrition. These games are so analyzed in a probabilistic way that is consistent to what we could expect intuitively, overcoming some known paradoxes of the Game Theory.

Key concepts: Mathematical economics, Prisoner's dilemma, Dilemma, Game theory, Superrationality, Traveler's dilemma, Cooperativeness, Repeated game

Related papers

Back to paper searchBrowse research topicsOriginal source
How much is convenient to defect? A method to estimate the cooperation probability in Prisoner's Dilemma and other games — Research Paper | ScholarLens