Skip to main content
The Evolution of Altruism: Game Theory in Multilevel Selection and Inclusive Fitness
Journal of Theoretical Biology (2007)
  • Jeffrey Alan Fletcher, Portland State University
  • Martin Zwick, Portland State University

Although the prisoner's dilemma (PD) has been used extensively to study reciprocal altruism, here we show that the n-player prisoner's dilemma (NPD) is also central to two other prominent theories of the evolution of altruism: inclusive fitness and multilevel selection. An NPD model captures the essential factors for the evolution of altruism directly in its parameters and integrates important aspects of these two theories such as Hamilton's rule, Simpson's paradox, and the Price covariance equation. The model also suggests a simple interpretation of the Price selection decomposition and an alternative decomposition that is symmetrical and complementary to it. In some situations this alternative shows the temporal changes in within- and between-group selection more clearly than the Price equation. In addition, we provide a new perspective on strong vs. weak altruism by identifying their different underlying game structures (based on absolute fitness) and showing how their evolutionary dynamics are nevertheless similar under selection (based on relative fitness). In contrast to conventional wisdom, the model shows that both strong and weak altruism can evolve in periodically formed random groups of non-conditional strategies if groups are multigenerational. An integrative approach based on the NPD helps unify different perspectives on the evolution of altruism.

  • Altruism--Statistics,
  • Kin selection (Evolution)
Publication Date
March 7, 2007
Publisher Statement
Copyright (2007) Elsevier
Citation Information
Jeffrey Alan Fletcher and Martin Zwick. "The Evolution of Altruism: Game Theory in Multilevel Selection and Inclusive Fitness" Journal of Theoretical Biology Vol. 245 Iss. 1 (2007)
Available at: