Source-linked AI summary
Competition and cooperation among different punishing strategies in the spatial public goods game
Xiaojie Chen, Attila Szolnoki, Matjaz Perc
TL;DR
The paper asks how different propensities to punish affect the evolution of cooperation and whether selection favors individually strong punishers. It models spatial public goods with multiple probabilistic punishing cooperator classes, finding that invasion velocities can produce both cooperation and competition among punishers.
Problem
The study addresses how selection operates when cooperators differ in their willingness to punish defectors, including whether individually weak punishers can persist alongside stronger classes.
Method
A spatial public goods game pits defectors against multiple cooperator classes that punish with class-specific probabilities and share sanctioning costs.
Results
At α = 0.42, C5 and C4 outperform defectors individually, whereas C3 disappears alone but survives when all punishing classes compete together.
Takeaways & Limitations
Punishing strategies can cooperate when slower strategies delay defectors, while under other conditions invasion-velocity differences produce a single victorious class.
Abstract
from arXiv · showhide
Inspired by the fact that people have diverse propensities to punish wrongdoers, we study a spatial public goods game with defectors and different types of punishing cooperators. During the game, cooperators punish defectors with class-specific probabilities and subsequently share the associated costs of sanctioning. We show that in the presence of different punishing cooperators the highest level of public cooperation is always attainable through a selection mechanism. Interestingly, the selection not necessarily favors the evolution of punishers who would be able to prevail on their own against the defectors, nor does it always hinder the evolution of punishers who would be unable to prevail on their own. Instead, the evolutionary success of punishing strategies depends sensitively on their invasion velocities, which in turn reveals fascinating examples of both competition and cooperation among them. Furthermore, we show that under favorable conditions, when punishment is not strictly necessary for the maintenance of public cooperation, the less aggressive, mild form of sanctioning is the sole victor of selection process. Our work reveals that natural strategy selection can not only promote, but sometimes also hinder competition among prosocial strategies.
I. INTRODUCTION
The paper studies how diverse willingness to punish shapes cooperation in spatial public goods games. It asks whether selection favors one punishing class or instead produces competition and synergy among prosocial strategies.
- Punishment can promote cooperation but is costly, and cooperators may abstain from sanctioning to avoid costs or retaliation.
- The model replaces uniform punishment with distinct cooperator classes that punish defectors at different probabilities.This variation reflects observed diversity in individuals’ willingness to punish.
- The study examines whether natural selection favors a particular punishing class or generates synergistic effects despite competition among punishers.
- Evolution is governed by selection mechanisms whose outcomes depend on the synergistic effects of cooperative behavior.
II. SPATIAL PUBLIC GOODS GAME WITH DIVERSE PUNISHMENT
The spatial game places multiple probabilistic punishing cooperator classes and defectors on a periodic square lattice. Contributions are pooled and shared, while punishment fines and costs are allocated according to each class’s realized sanctioning behavior.
- Individuals occupy a periodic L × L square lattice and participate in a public goods game contested by T cooperator classes and defectors.Each cooperator contributes c, defectors contribute nothing, and the multiplied pool is shared equally among group members.
- Cooperator class C_i punishes present defectors with probability i/(T −1), ranging from never-punishing C_0 to always-punishing C_T −1.C_0 acts as a second-order free-rider, whereas C_T −1 punishes whenever defectors are present.
- Each punished defector receives fine α, while participating punishers equally share sanctioning costs; each punisher bears (n −n_C)α/n_P.
- The noise amplitude is fixed at κ = 0.5, allowing better-performing strategies to spread more readily while retaining occasional imitation of less successful neighbors.
III. RESULTS
With r = 3.5, heterogeneous punishing cooperators can achieve dominant cooperation through selection, but surviving classes depend on punishment strength and invasion dynamics. At r = 4.0, simultaneous competition can also sustain cooperation where pure cooperators coexist with defectors without punishment.
- Competition with defectors: Cooperation becomes dominant when α ≥ 0.42, although some weak cooperator classes go extinct before defectors disappear.The total cooperator fraction increases monotonously with α in the heterogeneous model.
- Selection among punishing strategies: At α = 0.42, C5 and C4 outperform defectors individually, whereas C3 disappears alone but survives when all punishing strategies compete simultaneously.C0 can also attain a considerable fraction at high α despite dying out immediately against defectors alone at r = 3.5.
- Cooperation among punishers: C3 and C5 can jointly dominate the population even though C3 fails alone, while replacing C3 with C2 leaves C5 as the sole surviving cooperative strategy.The C3–C5 cooperative outcome is therefore fragile and depends on which mildly punishing strategy is present.
- Invasion velocities: Differences in invasion velocities explain why C3 can assist C5 against defectors longer than C2, despite both C2 and C3 ultimately losing individually.Figure 4 compares the time-dependent sign of ρCi −ρD for C2, C3, and C5.
- Competition at higher synergy: For r = 4.0, simultaneous competition produces complete cooperative dominance above α > 0.25, while pure cooperators can coexist with defectors without punishment because r exceeds 3.74.Individually, excessively large α is detrimental to C3, C4, and C5 because sanctioning costs outweigh the additional harm imposed on defectors.
- Selection at favorable synergy: At α ≈ 0.2 and r = 4.0, C3 alone survives alongside defectors while other punishing strategies die out, even though C3 cannot eliminate defectors individually at these α values.At larger α, C3 can become too effective against defectors, allowing other punishing strategies to persist; excessive α can also harm C3, C4, and C5 individually.
IV. CONCLUSION
The evolutionary outcome among punishing cooperators depends on the multiplication factor and on invasion velocities, allowing either cooperation or competition among strategies. Selection can support punishers that lose individually, while higher multiplication factors can produce a single victor.
- The outcome depends sensitively on the multiplication parameter r, with low-r games helping punishers unable to survive individual competition against defectors.
- A losing punishing strategy can delay defector victory long enough for a more successful punisher to eliminate defectors, producing cooperation between punishing strategies.
- At higher multiplication factors, competition can dominate and yield a single victor, even when the fittest punisher cannot completely defeat defectors alone.