Formation of Dominance Relationships via Strategy Updating in an Asymmetric Hawk-Dove Game

TL;DR

This paper models how social animals develop dominance hierarchies through strategy updates based on past interactions within an asymmetric Hawk-Dove game framework, highlighting the impact of incomplete information on social dynamics.

Contribution

It introduces a dynamic game-theoretic model incorporating incomplete information to explain dominance hierarchy formation in social animals, aligning with observed behaviors.

Findings

Increased interactions accelerate asymmetry assessment.

Incomplete information leads to different outcomes than static perfect-information models.

Aggression levels can decrease over time with more interactions.

Abstract

We develop a model to describe the development of dominance relations between social animals as they use past experiences to inform future interactions. Using the game-theoretic framework of a Hawk-Dove game with asymmetric resource-holding potentials (RHPs), we derive a simple model that describes the social interactions of animals that compete for resources. We then consider a game-playing strategy in which animals acquire information about their RHP asymmetry from the results of their previous contests and subsequently use their asymmetry assessment to inform their behavior in future contests. We examine how directly incorporating the fact that animals have incomplete information in their interactions can lead to outcomes that differ from what would be expected if one considers the situation as a static game in which the animals have perfect information about the asymmetry size. We thereby obtain results that are consistent with observations of dominance-hierarchy formation in social animals. We also discuss how increased interactions between animals can speed up the asymmetry assessment process and how this can explain why aggression can sometimes decrease as the number of interactions between animals increases over time.