Source-linked AI summary
Collective behavior and evolutionary games - An introduction
Matjaz Perc, Paolo Grigolini
TL;DR
The paper introduces a special issue seeking unifying principles of collective behavior across nature and society, including but not limited to evolutionary games. It surveys game structures, simulation practices, empirical tensions, and future directions spanning networks, criticality, cognition, and consciousness.
Problem
Collective behavior arises across natural and social systems, but understanding its common principles requires connecting diverse phenomena and reconciling theoretical evolutionary-game models with empirical human behavior.
Method
The paper provides an introductory synthesis of evolutionary-game formulations, strategic extensions, simulation methods, empirical findings, and interdisciplinary research directions.
Results
The introduction identifies structured populations, strategic complexity, interdependent networks, and links to criticality as central contexts for studying collective behavior.
Takeaways & Limitations
The special issue welcomes interdisciplinary work on collective behavior, including synchronization, human cognition, social networks, and evolutionary games.
Abstract
from arXiv · showhide
This is an introduction to the special issue titled "Collective behavior and evolutionary games" that is in the making at Chaos, Solitons & Fractals. The term collective behavior covers many different phenomena in nature and society. From bird flocks and fish swarms to social movements and herding effects, it is the lack of a central planner that makes the spontaneous emergence of sometimes beautifully ordered and seemingly meticulously designed behavior all the more sensational and intriguing. The goal of the special issue is to attract submissions that identify unifying principles that describe the essential aspects of collective behavior, and which thus allow for a better interpretation and foster the understanding of the complexity arising in such systems. As the title of the special issue suggests, the later may come from the realm of evolutionary games, but this is certainly not a necessity, neither for this special issue, and certainly not in general. Interdisciplinary work on all aspects of collective behavior, regardless of background and motivation, and including synchronization and human cognition, is very welcome.
1. Evolutionary games
Evolutionary games model cooperation and defection through pairwise or group interactions, and structured populations can produce collective behavior. The section introduces core payoff conditions and the public-goods framework.
- Evolutionary games are especially likely to display collective behavior when played on structured populations.
- In the prisoner’s dilemma, defectors dominate cooperators in well-mixed populations, whereas snowdrift games can support coexistence.
- Cooperative clusters on a square lattice enabled cooperators to survive alongside defectors in the prisoner’s dilemma.
- Pairwise games assign payoffs to two interacting players, while public-goods games distribute multiplied group contributions among all group members.
- The public-goods setup places N = L2 players in overlapping groups, with cooperators contributing a and defectors contributing nothing.
2. Strategic complexity and more games
Adding strategies and game types substantially expands the possible collective dynamics. These models can produce intricate phase diagrams and cyclic dominance under different interaction and strategy-space assumptions.
- Additional strategies include loners, volunteers, rewarders, punishers, conditional cooperators, and conditional punishers.
- These strategic extensions generate phase diagrams ranging from single- and two-strategy stationary states to rock-paper-scissors-type cyclic dominance.
- Studied game classes include collective-risk dilemmas, stag-hunt dilemmas, and ultimatum games alongside prisoner’s dilemma, snowdrift, and public-goods games.
- Behavior depends notably on whether interactions are well-mixed or structured and whether strategies are discrete or continuous.
3. Simulations versus reality
Monte Carlo simulations dominate the analysis of evolutionary games on structured populations, but empirical and experimental findings do not yet align fully with theoretical explanations. Recent human studies both challenge and support network-based accounts of cooperation.
- Monte Carlo simulations are the predominant analysis method for evolutionary games on structured populations.
- Each elementary simulation step selects a player and neighbor, compares payoffs, and attempts strategy adoption under uncertainty modeled by the Fermi function.
- Repeating the elementary step N times gives every player one average opportunity to update and defines one full Monte Carlo step.
- Large-scale human experiments motivate efforts to reconcile simulations with reality because network reciprocity does not explain all socially responsible choices in prisoner’s dilemma settings.
- Evidence also supports cooperation in human social networks and finds that dynamic social networks promote cooperation in experiments.
4. Future research
The section proposes future research on interdependent networks and connections among sociology, neurophysiology, criticality, cognition, and collective behavior. It highlights cooperation effects, cross-disciplinary parallels, and continuously updated publication.
- 4. Future research: Interdependent or multiplex networks are proposed as a priority because changes in one network can produce catastrophic consequences in another.
- 4. Future research: Existing studies find that interdependence can promote cooperation through network reciprocity or information sharing, but excessive interdependence is detrimental.
- 4. Future research: The special issue invites research connecting sociological game theory with neurophysiological criticality and brain dynamics.
- 4. Future research: Figure 1 presents cyclic dominance across prisoner’s dilemma, public-goods, ultimatum, and peer-punishment settings, including survival of the weakest and interface monolayers.
- 4. Future research: A dynamical model sharing brain-dynamics criticality properties has been used to explain Arab Spring events, motivating links among criticality, swarm intelligence, cognition, and consciousness.
- 4. Future research: The special issue will be updated continuously so new papers can be published immediately after acceptance into the issue.