Source-linked AI summary
Spreading of cooperative behaviour across interdependent groups
Luo-Luo Jiang, Matjaz Perc
TL;DR
The paper examines how between-group links affect cooperation in interdependent groups playing the prisoner’s dilemma. Simulations across regular, random, and scale-free topologies find an intermediate linking probability that maximizes cooperation, while between-group links are more likely than in-group links to join cooperators.
Problem
The paper asks how between-group links contribute to cooperation across interdependent groups, including how their role compares with in-group links.
Source: Results · p. 3Method
The authors simulate prisoner’s dilemma dynamics on groups with regular, random, or scale-free topology, varying the probability p of forming between-group links.
Source: Methods · pp. 4–5Results
An intermediate probability of between-group links maximizes cooperation; these links are also more likely than in-group links to connect two cooperators.
Source: Results · pp. 2–3Takeaways & Limitations
Between-group links appear especially important for spreading cooperative behaviour across interdependent groups.
Source: Results · p. 3, Discussion · p. 4
Abstract
from arXiv · showhide
Recent empirical research has shown that links between groups reinforce individuals within groups to adopt cooperative behaviour. Moreover, links between networks may induce cascading failures, competitive percolation, or contribute to efficient transportation. Here we show that there in fact exists an intermediate fraction of links between groups that is optimal for the evolution of cooperation in the prisoner's dilemma game. We consider individual groups with regular, random, and scale-free topology, and study their different combinations to reveal that an intermediate interdependence optimally facilitates the spreading of cooperative behaviour between groups. Excessive between-group links simply unify the two groups and make them act as one, while too rare between-group links preclude a useful information flow between the two groups. Interestingly, we find that between-group links are more likely to connect two cooperators than in-group links, thus supporting the conclusion that they are of paramount importance.
Results
The results show that an intermediate probability of between-group links maximizes cooperation across group-A topologies. The response depends on which group’s links are diluted and on group A’s topology and the temptation to defect.
Results
- An intermediate between-group-link probability maximizes cooperation across group-A topologies, with the largest increase over independent groups for regular and the smallest for scale-free networks.The scale-free case improves only marginally because that topology already strongly promotes cooperation.Source: Results · p. 2
- The overall bell-shaped response comes mainly from group B’s link dilution; group A’s cooperation rises with p when regular or random and stays nearly flat when scale-free.Group A gains links while group B replaces in-group links with between-group links, so the latter’s outcome also depends on group A’s topology.Source: Results · pp. 2–3, Figure 3 · p. 2
- Between-group links are more likely than in-group links to join two cooperators, and linked players have higher next-round cooperation probability, especially with one to four cooperative neighbours.The comparisons are reported for random group A; the next-round comparison uses p = 0.15.Source: Results · p. 3, Figure 4 · p. 3
- The effect of p varies with temptation b: regular group A has an optimum only above b = 1.15, random group A has one across b, and scale-free group A responds little.For random topology, the optimal p shifts lower as b increases.Source: Results · pp. 3–4, Figure 5 · p. 4
Discussion
The study identifies an optimal level of interdependence for cooperation across group topologies, with the clearest gains in regular or random groups. Between-group links are associated with cooperation, but the groups must retain sufficient independence.
Discussion
- An optimal link-formation probability promotes cooperation across topologies, with the clearest gains in regular or random groups and little improvement for scale-free groups.For scale-free groups, the positive impact of their topology alone precludes significant interdependence-driven improvement.Source: Discussion · p. 4
- In regular or random groups, benefits stem from diluting links in one group and adding links to the other; dilution is generally stronger.Source: Discussion · p. 4
- Players with between-group links are more likely to cooperate, and between-group links more often connect two cooperators than in-group links.Source: Discussion · p. 4
- The findings align with Hadza camp observations and prior interdependent-network research on cooperation [6] [7].Those prior studies describe distinct mechanisms, including coupling dynamics and interdependent network reciprocity.Source: Discussion · p. 4
- Interdependence can help resolve social dilemmas, but enough independence must remain for each network to function if cooperation evolves poorly in the other.Source: Discussion · p. 4
Methods
The methods construct two groups with regular, random, or scale-free networks, then vary their interdependence by rewiring links from group B to group A. Monte Carlo simulations model strategy updates through payoff-based imitation.
Methods
- Groups use regular, random, or scale-free topologies, constructed systematically from simpler to more complex interaction structures.Regular and random groups connect each player to k others; scale-free groups follow the Barabási–Albert algorithm [71].Source: Methods · p. 4
- Each group-B player may sever one in-group link with probability p and connect to a randomly chosen group-A player, with at most one cross-group link per player.Source: Methods · p. 4
- For p = 1, the groups effectively act as one because all players from group B are linked with all players from group A.Source: Methods · p. 5
- Simulations begin with equal numbers of cooperators and defectors; players compare payoffs with a random neighbor and may adopt the neighbor’s strategy.Adoption is possible when the neighbor’s payoff is higher, with probability q = (πy − πx)/(kmaxb), avoiding unequal selection intensity across degrees.Source: Methods · p. 5