Sensitivity of Electrophysiological Patterns in Level-K States as Function of Individual Coordination Ability

Abstract

Tacit coordination games are games in which players get rewarded by choosing the same alternatives as an unknown player when communication between the players is not allowed or not possible. Classical game theory fails to correctly predict the probability of successful coordination due to its inability to prioritize salient Nash equilibrium points (in this setting, also known as “focal points”). To bridge this gap a variety of theories have been proposed. A prominent theory is the level-k theory which assumes that players’ reasoning depth relies on their subjective level of reasoning. Previous studies have shown that there is an inherent difference in the coordination ability of individual players, and that there is a correlation between the individual coordination ability and electrophysiological measurements. The goal of this study is to measure and quantify the electrophysiological sensitivity patterns in level-k states in relation to the individual coordination abilities of players. We showed that the combined model capabilities (precision and recall) improve linearly depending on the player’s individual coordination ability. That is, player with higher coordination abilities exhibit a more pronounced and prominent electrophysiological patterns. This result enables the detection of capable coordinators based solely on their brain patterns. These results were obtained by constructing a machine-learning classification model which predicting one of the two cognitive states, picking (level-k = 0) or coordination (level-k > 0), based on electrophysiological recordings. The results of the classification process were analyzed for each participant individually in order to assess the sensitivity of the electrophysiological patterns according to his individual coordination ability.