The Cooperation Ceiling: Extrinsic Population Dynamics and the Intrinsic Escape

Harry Foster, Vince Knight, Sebastian Krapohl · 2026

Read The Cooperation Ceiling: Extrinsic Population Dynamics and the Intrinsic Escape by Harry Foster, Vince Knight, Sebastian Krapohl (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2606-31740.txt.

Introduction

Foster, Knight and Krapohl study how cooperation evolves when players in a population differ from one another. Evolutionary game theory tracks strategies that spread or fade under an update rule. Many rules assume identical players, yet real groups vary in wealth, capacity and costs.

Four update rules are sorted into two families. Extrinsic rules have players copy what pays off for others, while intrinsic rules judge by their own traits or potential payoffs. The authors prove that extrinsic dynamics carry a ceiling on the cooperation rate. Intrinsic dynamics can break through that ceiling. A public goods game with unequal contributions demonstrates the gap between the two families.

Catalog record

CREATORHarry Foster, Vince Knight, Sebastian Krapohl
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE67 KB
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2606.31740

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

The Cooperation Ceiling: Extrinsic Population Dynamics and the Intrinsic Escape Harry Foster, Vince Knight, Sebastian Krapohl Abstract Evolutionary game theory provides a framework by which to study the emergence of cooperation in a population of self-interested actors. In such a framework, players’ decisions on whether or not to cooperate evolve according to decision rules called population dynamics. However, often games are studied under the assumption that all individuals play under the same conditions, and many common choices of update rule are not well suited for a heterogeneous population. In this paper, we categorise and compare four different population dynamics in such a population as “extrinsic”, where players learn by looking outward at the payoffs of other players, and “intrinsic”, where players look inwardly at their own attributes or potential payoffs. We show that extrinsic population dynamics admit a ceiling on the rate of cooperation which can be exceeded by intrinsic population dynamics, and demonstrate this using the public goods game with heterogeneous contributions. 1 Introduction The conditions under which cooperation can emerge among self-interested actors is a central question across biology, economics, and the social sciences. Many of the situations in which this question arises are collective action problems, in which individuals share an interest in a common good yet each would prefer that others bear the cost of providing it. The structure of such a dilemma maps onto a public goods game [ 2 , 19 , 22 , 21 , 45 ] : each actor decides whether to incur a private cost to contribute to a shared resource whose total value is multiplied by some factor r and distributed equally amongst all players. This creates the temptation to free-ride, and so purely selfish reasoning leads to a tragedy of the commons [ 18 , 54 ] , even when collective cooperation would benefit everyone. Both theory [ 43 ] and controlled experiments [ 32 , 33 ] show that framing cooperation as a collective-risk problem, where failure to meet a cooperation threshold incurs a shared loss, can substantially raise cooperation rates. Evolutionary game theory offers a framework for studying how cooperation can emerge among self-interested actors without assuming rational foresight [ 23 , 37 ] . In this approach, strategies that perform well spread through a population via one of several update rules, while poorly performing strategies decline. The conditions under which cooperative strategies can invade and stabilise have been extensively studied in two-player settings such as the iterated prisoner’s dilemma [ 4 ] and the snowdrift game [ 20 ] . The public goods game generalises these pairwise dilemmas to N players [ 19 ] , making it well suited to studying multi-lateral cooperation. In finite populations, evolutionary outcomes are stochastic: a rare mutant strategy spreads or vanishes with a probability that depends on the population size, payoff structure, and update rule [ 36 , 52 ] . The conditions under which one strategy invades a population have been studied in depth for particular evolutionary processes. The \frac{1}{N} fixation probability of a single mutant under neutral drift [ 37 ] serves as a benchmark for when selection favours a strategy, and results such as the \frac{1}{3} rule [ 36 , 38 ] hold even when a population updates with both the Moran process and Fermi imitation dynamics [ 29 ] . For symmetric 2\times 2 games, the condition a(N-2)+bN>cN+d(N-2) [ 25 , 50 ] on the payoff entries leads to the abundance of a given strategy in a wide range of evolutionary processes and for arbitrary mutation and selection intensity [ 1 ] . These results, however, concern a symmetric game played pairwise between identical members of a population. We instead consider all N -player games in which a defector outperforms a cooperator in a given state, amongst a fully heterogeneous population. A feature of real-world collective action that standard models typically set aside is that actors are not identical. Players differ in wealth, productive capacity, and the costs they face, and this inequality can reshape the dynamics of cooperation [ 22 ] . One commonly studied heterogeneous population is one in a spatial setting, where players are placed on a network and participate simultaneously in multiple games with their neighbours [ 35 , 39 , 48 ] . In [ 44 ] it is shown that social diversity in endowments promotes the emergence of cooperation in the public goods game: heterogeneous populations outperform their homogeneous counterparts over a wide range of parameters. This echoes earlier work showing that scale-free network topology, a form of structural heterogeneity, provides a unifying framework for the emergence of cooperation [ 42 ] . When contributions depend on dynamically accumulated reputation [ 31 ] , high-reputation defectors receive reduced transfers, weakening defection’s advantage. When heterogeneity arises from differences in network degree so that players allocate contributions according to the difference in the number of connections [ 27 ] , the payoff structure either penalises medium-degree defectors, or creates small clusters of strong cooperators. The characteristic spatial mechanism is the formation of cooperating clusters that are more profitable than their defecting neighbours; provided r is sufficiently large, these clusters resist invasion and eventually spread. Heterogeneity does not, however, unconditionally favour cooperation. Flores et al. [ 9 ] show that when the contribution level is itself a strategic variable rather than a fixed attribute, low-value contributions act as a form of defection against the most collectively beneficial strategies, so that the emergence of cooperation does not guarantee the emergence of high-value cooperation. How the choice of update rule interacts with population heterogeneity is also poorly understood. The Moran process [ 26 , 34 ] , Fermi imitation dynamics [ 49 ] , aspiration dynamics [ 8 ] , and introspection dynamics [ 5 , 6 ] all model strategy revision in distinct ways, and each may respond differently to player-specific payoffs. In particular, intrinsic dynamics, in which a player evaluates a prospective strategy change based on their payoff in comparison with their own counterfactual payoff or aspiration level, are better suited to heterogeneous populations than extrinsic dynamics, in which a player may copy a high-fitness neighbour without considering whether the same strategy would serve them as well. We make the following contributions. First, we build on an established general framework for evolutionary dynamics on fully heterogeneous, well-mixed populations of N ordered individuals, in which each player may have a distinct payoff function and contribution level. Second, we formalise the notion of a purely extrinsic population dynamic, a definition encompassing both the Moran process [ 26 , 34 ] and Fermi imitation dynamics [ 49 ] , and show that such dynamics are inherently constrained by an upper bound on the probability of cooperation. Third, we consider two purely intrinsic population dynamics, aspiration dynamics [ 8 ] and introspection dynamics [ 5 , 6 ] , showing that these processes can overcome the ceiling of purely extrinsic dynamics. For each population dynamic, we use the public goods game [ 2 , 19 , 21 , 22 , 45 ] to demonstrate our results. The remainder of the paper is organised as follows. Section 2 introduces the population model and the public goods game. Section

Read the complete text