Profiles of Achievement, Affiliation and Power Motivation

Abstract

In this chapter, three profiles of achievement , affiliation and power motivation are introduced for two different games: roulette and the prisoners’ dilemma game. These profiles are tuned for agents playing roulette using a genetic algorithm, and then further examined in agents playing the prisoners’ dilemma game . We again demonstrate the similarities that can be observed between motivated game-playing agents and humans playing the same games. We show that we can create a diverse group of agents that compete or cooperate in different ways when playing games, and that models tuned on one game can be used in agents playing another game.