Identifying patterns of human behavior: an analysis on experimental data of the Public Goods Game

Abstract

[en] The cooperation capacity among different people or countries is nowadays essential to meet certain collective objectives such as undergo a climate action, solve the pollution of the seas, or in general initiate a collective action where everybody has to contribute. Public Goods and Collective Risk Dilemmas (CRD) games have been widely used to analyse these problematics, trying to understand which is the kind of contribution of all actors engaged. In this report we add new perspectives to the debate by analysing one of the largests samples that has been used. This analysis is done by means of Machine Learning (ML) techniques which suppose a new way of addressing the identification of patterns in CRD data. In terms of the game theoretical analysis we found that an unequal distribution of resources at the beginning of the game ends up causing imbalances, mainly in the distribution of costs, which has important implications in terms of climate justice.