Source-linked AI summary
Restricted connections among distinguished players support cooperation
Matjaz Perc, Attila Szolnoki, Gyorgy Szabo
TL;DR
The paper asks how cooperation can persist when defection is favored, and studies this in a spatial prisoner’s dilemma with heterogeneous teaching capabilities and temporary long-range links among influential players. It finds that sparse, weakly interconnected influential players can sustain cooperation across the full temptation range, with the optimal influential fraction decreasing as temptation rises.
Problem
The paper addresses the challenge of sustaining cooperation among selfish individuals when defectors can avoid cooperation costs and spread through imitation.
Method
The study simulates a spatial prisoner’s dilemma on a square lattice where a fraction μ has wx = 1 and distinguished players temporarily transfer strategies to distant teachers with probability p.
Results
Temporary long-range connections sustain cooperation across the full temptation range, while optimal μ decreases with increasing b and best cooperation requires very small p.
Takeaways & Limitations
Cooperation thrives when influential players are few and sparsely connected, enabling cooperative strategies to reach distant players without producing mean-field-like defection.
Abstract
from arXiv · showhide
We study the evolution of cooperation within the spatial prisoner's dilemma game on a square lattice where a fraction of players $μ$ can spread their strategy more easily than the rest due to a predetermined larger teaching capability. In addition, players characterized with the larger teaching capability are allowed to temporarily link with distant opponents of the same kind with probability $p$, thus introducing shortcut connections among the distinguished. We show that these additional temporary connections are able to sustain cooperation throughout the whole range of the temptation to defect. Remarkably, we observe that as the temptation to defect increases the optimal $μ$ decreases, and moreover, only minute values of $p$ warrant the best promotion of cooperation. Our study thus indicates that influential individuals must be few and sparsely connected in order for cooperation to thrive in a defection prone environment.
I. INTRODUCTION
The paper frames cooperation as vulnerable to defection in evolutionary games, while spatial structure and player heterogeneity can protect cooperative behavior. It proposes combining heterogeneous teaching capabilities with temporary same-rank connections to strengthen cooperation across temptation levels.
- Cooperation is difficult to sustain because defectors avoid cooperation costs and spread by imitating more successful strategies.
- Spatial prisoner’s dilemma games allow cooperators to survive by clustering and mutually resisting exploitation by invading defectors.
- Player heterogeneity in degree, teaching capability, or social rank reflects unequal opportunities to become influential in real-life societies.
- The model combines differing teaching capabilities with temporary connections between distant distinguished players of the same rank.
- For optimal μ and p, defectors remain outnumbered across the full temptation range, while optimal μ decreases continuously as b increases.
II. GAME DEFINITION AND SETUP
The study uses a spatial prisoner’s dilemma on a periodic square lattice, assigning fixed teaching capabilities to players and allowing distinguished players occasional long-range strategy transfers. Cooperation levels are estimated through Monte Carlo simulations, with finite-size stationary-state caveats.
- The game uses a periodic L × L square grid with nearest-neighbor interactions, equal-probability initial strategies, and temptation 1 < b ≤ 2.
- A fraction μ of players receives teaching capability wx = 1, while the remaining 1 − μ receives wx = 0.01.
- Distinguished players can transfer strategies to randomly selected distant teachers with probability p instead of nearest neighbors with probability 1 − p.
- Payoffs remain determined by four nearest-neighbor prisoner’s dilemma interactions even when strategy transfer occurs between distant teachers.
- Simulations use populations from 300 × 300 to 800 × 800 individuals and estimate cooperator density after discarding transients.
III. RESULTS
Cooperation is maximized by jointly tuning the fraction of distinguished players and their temporary long-range connection probability. The optimal fraction decreases as defection becomes more tempting, while sparse shortcuts support cooperation across the full temptation range.
- Jointly setting μ = 0.12 and p = 0.03 keeps ρC > 0.5 throughout the whole range of b.This combination outperforms varying either parameter alone.
- At μ = 0.3, cooperation is promoted best at an optimal p ≈0.05, and fine-tuned shortcuts extend cooperation from extinction at b = 1.3 to survival at b = 2.0.The long-range links also markedly enlarge the region of complete cooperative dominance.
- The ρC landscape exhibits a double resonance, requiring fine tuning of both p and μ for optimal cooperation.The resonance peaks at small values of both parameters.
- The optimal μ decreases continuously as the temptation b increases.This trend is reported across results obtained at b = 1.1, 1.2, 1.5, and 2.
- For μ below the critical coverage μc = 0.13965(5), a two-stage process converts weakened distinguished defectors and sustains cooperative colonies.At high μ, distinguished cooperators cannot succeed because their neighbors may be exploited by other distinguished defectors.
- At μ = 0.12 and b = 2, the optimal shortcut probability is very small, p ≃0.03.Lower p allows influential cooperators to disappear stochastically, whereas higher p drives the system toward mean-field behavior favoring defection.
IV. SUMMARY
Temporary long-range connections among distinguished players substantially promote cooperation on a regular lattice. Cooperation persists across the full considered temptation range, while defection-prone environments favor few, sparsely interconnected influential players.
- Temporary long-range connections among distinguished players substantially promote cooperation in the evolutionary prisoner’s dilemma on a regular lattice.The model combines heterogeneous player types with randomly evolving shortcuts among distinguished players.
- Cooperation is maintained across the whole temptation range 1 < b ≤2 considered for iterated prisoner’s dilemma games.
- Defection-prone environments require modest densities of influential players that are not strongly interconnected.