Related Experiment Videos
The evolution of inefficiency in a simulated stag hunt
1Department of Psychology, University of North Carolina, Chapel Hill, NC 27514, USA. neil@unc.edu
Behavior Research Methods, Instruments, & Computers : a Journal of the Psychonomic Society, Inc
|July 13, 2001
Summary
Evolutionary simulations show that risk influences cooperation. High-risk conditions favor inefficient outcomes, while low-risk conditions promote efficient cooperation with increased forgiveness and future discounting.
Area of Science:
- Evolutionary Psychology
- Game Theory
- Computational Neuroscience
Background:
- Coordination games are fundamental to understanding social behavior.
- Risk and reward influence decision-making and evolutionary outcomes.
- Reinforcement learning models provide insights into adaptive strategies.
Purpose of the Study:
- To investigate how risk conditions affect the evolution of cooperation in simulated agents.
- To explore the emergence of specific strategies like forgiveness and temporal discounting under varying risk levels.
- To demonstrate the application of agent-based simulation in evolutionary psychology.
Main Methods:
- Utilized genetic algorithms to evolve populations of reinforcement learning (Q-learning) agents.
- Simulated a repeated two-player symmetric coordination game.
- Implemented varying risk conditions to observe their impact on agent behavior.
Main Results:
- Populations evolved towards Pareto inefficient equilibria under high-risk conditions.
- Populations evolved towards Pareto efficient equilibria under low-risk conditions.
- Increased forgiveness and temporal discounting of future returns were observed in low-risk environments.
Conclusions:
- Risk is a critical factor shaping the evolution of cooperative strategies.
- Simulations effectively model the interplay between risk, strategy evolution, and social outcomes.
- Findings support the utility of computational simulations for advancing evolutionary psychology research.