Evolutionary game theory

Evolutionary game theory (EGT) applies game theory principles to biological populations undergoing evolution. This theoretical framework models Darwinian competition through the analysis of contests and strategies. Its inception in 1973 involved John Maynard Smith and George R. Price, who formalized contests as strategic interactions and established mathematical criteria for predicting outcomes among competing strategies.

Evolutionary game theory distinguishes itself from classical game theory by emphasizing the dynamic evolution of strategies. These dynamics are significantly influenced by the prevalence of various competing strategies within a given population.

Evolutionary game theory has provided crucial insights into the evolutionary underpinnings of altruistic behaviors within a Darwinian context. Consequently, it has garnered considerable attention from scholars across diverse disciplines, including economics, sociology, anthropology, and philosophy.

History

Classical game theory

Classical non-cooperative game theory was developed by John von Neumann with the objective of identifying optimal strategies in adversarial competitions. Such contests involve participants, each possessing a range of available actions. These games may consist of a single interaction or unfold through multiple, repetitive rounds. A player's chosen sequence of actions defines their strategy. Outcomes are determined by established rules governing player actions, yielding specific payoffs; these rules and their associated payoffs are typically represented using decision trees or payoff matrices. A fundamental premise of classical theory is that players act rationally. Consequently, each player must anticipate and account for the strategic considerations of their opponents when formulating their own decisions.

The problem of ritualized behaviour

Evolutionary game theory originated from the challenge of elucidating ritualized animal behaviors during conflict, specifically addressing the apparent 'gentlemanly' conduct observed in resource contests. Prominent ethologists Niko Tinbergen and Konrad Lorenz posited that these behaviors served the collective benefit of the species. However, John Maynard Smith deemed this explanation inconsistent with Darwinian principles, which emphasize individual-level selection, rewarding self-interest over actions benefiting the common good. Maynard Smith, a mathematical biologist, subsequently adopted game theory, following a suggestion from George Price, despite prior unsuccessful attempts by Richard Lewontin to apply the theory.

Adapting game theory to evolutionary games

Maynard Smith recognized that an evolutionary adaptation of game theory obviated the requirement for rational players, instead merely necessitating the existence of a strategy. Game outcomes reveal the efficacy of a given strategy, mirroring how evolution evaluates alternative strategies based on their capacity for survival and reproduction. Within a biological context, strategies are conceptualized as genetically inherited traits that govern an individual's behavior, akin to computer programs. A strategy's success is contingent upon its performance against competing strategies (including instances of itself) and the frequency of their occurrence. Maynard Smith detailed his research in his publication, Evolution and the Theory of Games.

Participants strive to maximize their reproductive output, with payoffs quantified in units of fitness, representing their relative capacity for reproduction. This framework invariably involves a multi-player game with numerous competitors. The governing principles incorporate replicator dynamics, which describe how individuals with higher fitness proliferate more extensively within the population, while less fit individuals are progressively eliminated, as formalized in a replicator equation. For simplicity, replicator dynamics model heredity but exclude mutation, and typically assume asexual reproduction. These games are simulated iteratively without predefined termination criteria. Outcomes encompass population dynamics, strategic success, and the identification of any emergent equilibrium states. In contrast to classical game theory, players do not actively select or modify their strategies; instead, they are endowed with a strategy at birth, which their progeny subsequently inherit.

Evolutionary games

Models

Evolutionary game theory integrates core tenets of Darwinian evolution, specifically competition (conceptualized as the game), natural selection (manifested through replicator dynamics), and heredity. It has significantly advanced the comprehension of phenomena such as group selection, sexual selection, altruism, parental care, co-evolution, and ecological dynamics. Through the application of these models, numerous counter-intuitive scenarios within these domains have been rigorously substantiated mathematically.

Evolutionary dynamics in games are typically investigated using replicator equations. These equations illustrate the growth rate of a specific strategy's proportion within a population, equating this rate to the disparity between that strategy's average payoff and the overall population's average payoff. Continuous replicator equations operate under the assumptions of infinite populations, continuous temporal progression, complete population mixing, and the faithful transmission of strategies across generations. Within these equations, certain attractors, specifically all global asymptotically stable fixed points, represent evolutionarily stable states. An evolutionarily stable strategy is defined as one capable of persisting against all "mutant" strategies. In the realm of animal behavior, this typically implies that such strategies are genetically encoded and profoundly influenced by hereditary factors, thereby determining an organism's or player's strategic choices through these biological determinants.

Evolutionary games constitute mathematical constructs characterized by distinct rules, payoff structures, and resultant mathematical behaviors. Each specific "game" models various challenges organisms encounter, along with the potential strategies they might employ for survival and reproduction. These games are frequently assigned evocative names and narrative contexts that delineate the general scenario of a given game. Illustrative examples of such games encompass the Hawk-Dove game, the war of attrition, the stag hunt, the producer-scrounger model, the tragedy of the commons, and the prisoner's dilemma. Corresponding strategies for these games include, but are not limited to, hawk, dove, bourgeois, prober, defector, assessor, and retaliator. Under the specific rules of each game, these diverse strategies engage in competition, with mathematical analysis employed to ascertain the outcomes and emergent behaviors.

The Hawk-Dove Game

The seminal game analyzed by Maynard Smith is the classic Hawk-Dove game. This model was devised to address Lorenz and Tinbergen's problem concerning contests over a divisible resource. Participants in this game adopt one of two strategic roles: hawk or dove. These roles represent distinct subtypes or morphs within a single species, each employing a different strategy. A hawk initially exhibits aggression, subsequently escalating into a physical confrontation that persists until either victory or injury (defeat). Conversely, a dove initially displays aggression but retreats to safety if confronted with significant escalation. Should no such escalation occur, the dove endeavors to share the resource.

Assuming the resource possesses a value V, and the cost incurred from losing a confrontation is C:

• Should a hawk encounter a dove, the hawk acquires the entire resource V.
• When two hawks meet, they each win approximately half the time and lose half the time, resulting in an average outcome of V/2 - C/2.
• Upon encountering a hawk, a dove will withdraw, receiving no payoff (0).
• When two doves meet, they share the resource, each obtaining V/2.

The actual payoff, however, is contingent upon the probability of encountering either a hawk or a dove, which reflects the proportion of each strategy within the population at the time of a specific contest. This proportion, in turn, is shaped by the outcomes of all preceding contests. When the cost of losing (C) exceeds the value of winning (V)—a common scenario in natural environments—the mathematical analysis converges on an Evolutionarily Stable Strategy (ESS). This ESS represents a mixed strategy where the proportion of hawks in the population stabilizes at V/C. Any temporary perturbation in the population, such as the introduction of new hawks or doves, will result in the population reverting to this equilibrium point. The resolution of the Hawk-Dove game elucidates why the majority of animal conflicts involve ritualized fighting behaviors rather than direct, injurious battles. Crucially, this outcome is not predicated on "good of the species" behaviors, as proposed by Lorenz, but rather exclusively on the implications of actions driven by what are termed "selfish genes."

The War of Attrition

In the Hawk-Dove game, a shareable resource yields payoffs for both doves engaged in a pairwise contest. However, when a resource is unshareable, and an alternative might be accessible by disengaging and seeking it elsewhere, pure Hawk or Dove strategies prove less efficacious. The combination of an unshareable resource with a substantial cost of contest loss (e.g., injury or mortality) further reduces the payoffs for both Hawk and Dove strategies. Consequently, a safer, lower-cost strategy involving display, bluffing, and prolonged waiting emerges as viable, termed the "bluffer strategy." This transforms the game into an accumulation of costs, whether from displaying or from extended, unresolved engagement. It functions as an auction where the victor is the contestant willing to incur the greater cost, while the loser bears an equivalent cost but gains no resource. The mathematical models of evolutionary game theory consequently indicate that timed bluffing constitutes an optimal strategy.

In a war of attrition, any unwavering and predictable strategy is inherently unstable, as it would eventually be superseded by a mutant strategy that outcompetes it by investing a marginally greater 'waiting resource' to secure victory. Consequently, only a random, unpredictable strategy can persist within a population of bluffers. Contestants effectively select an acceptable cost, proportional to the value of the desired resource, thereby making a random bid as part of a mixed strategy—a strategy encompassing multiple potential actions. This process establishes a distribution of bids for a resource of a specific value V, from which the bid for any given contest is randomly selected. This distribution, representing an Evolutionarily Stable Strategy (ESS), can be calculated using the Bishop-Cannings theorem, which applies to all mixed-strategy ESSs. Parker and Thompson determined the distribution function for these contests as follows:

${\displaystyle p(x)={\frac {e^{-x/V}}{V}}.}$

Consequently, the cumulative population of individuals who withdraw (quitters) at any specific cost 'm' within this mixed-strategy solution is defined as:

p ( m ) = §1617§ − e − m / V , {\displaystyle p(m)=1-e^{-m/V},}

This relationship is visually represented in the adjacent graph. The intuitive premise that higher resource values correlate with extended waiting times is empirically supported. Such phenomena are observable in natural contexts, for instance, among male dung flies competing for mating sites, where the timing of contest disengagement aligns with predictions derived from evolutionary game theory.

The Emergence of Novel Strategies Through Asymmetries

Within a war of attrition, the absence of cues signaling bid size to an opponent is crucial; otherwise, such information could be exploited for an effective counter-strategy. Nevertheless, a 'mutant' strategy, known as the bourgeois strategy, can outperform a bluffer in the war of attrition if an appropriate asymmetry is present. The bourgeois strategy leverages an existing asymmetry to resolve stalemates. In natural environments, resource possession frequently serves as such an asymmetry. This strategy dictates playing a Hawk when possessing the resource, but displaying and subsequently retreating when not in possession. While demanding greater cognitive capacity than a pure Hawk strategy, the bourgeois strategy is prevalent in numerous animal contests, exemplified by interactions among mantis shrimps and speckled wood butterflies.

Social Behavior

Games such as Hawk-Dove and War of Attrition exemplify pure inter-individual competition, devoid of social components. When social influences are present, competitors can engage in strategic interactions through four distinct alternatives. These alternatives are illustrated in the accompanying figure, where a plus sign denotes a benefit and a minus sign indicates a cost.

• Within a cooperative or mutualistic relationship, the roles of "donor" and "recipient" become nearly indistinguishable, as both parties derive a benefit from their collaboration within the game. This implies a strategic scenario where joint action yields mutual gains, or where overarching constraints necessitate concerted behavior, effectively placing both entities in a shared predicament.
• An altruistic relationship involves a donor incurring a cost to confer a benefit upon a recipient. Typically, the recipient shares a kin relationship with the donor, and the benefit transfer is unidirectional. While behaviors involving reciprocal benefit exchange, each incurring a cost, are frequently labeled "altruistic," closer examination often reveals that such "altruism" stems from optimized "selfish" strategies.
• Spite fundamentally represents a "reversed" form of cooperation, wherein neither participant obtains a direct, tangible benefit. Generally, the ally is kin-related, and the advantage conferred is a less challenging competitive environment for that ally. George Price, a pioneering mathematical modeler of both altruism and spite, reportedly found this conceptual equivalence emotionally unsettling.
• From a game theory perspective, selfishness constitutes the fundamental criterion for all strategic decisions; strategies that do not prioritize self-preservation and self-replication are unsustainable in any game. Crucially, this dynamic is complicated by the multi-level nature of competition, encompassing genetic, individual, and group levels.

The Competition of Selfish Genes

Initially, it might seem that the participants in evolutionary games are the individuals of each generation directly involved in the game. However, individuals persist for only a single game cycle; instead, it is the strategies themselves that truly compete across these multi-generational games. Consequently, genes ultimately execute the complete contest—specifically, the selfish genes embodying these strategies. These competing genes reside within an individual and, to some extent, within all of that individual's kin. This genetic presence can profoundly influence the survival of particular strategies, especially concerning cooperation and defection. W. D. Hamilton, renowned for his theory of kin selection, investigated numerous such scenarios using game-theoretic models. A kin-based approach to game contests elucidates various aspects of social insect behavior, altruistic parent-offspring interactions, mutual protection mechanisms, and cooperative offspring care. For these types of games, Hamilton introduced an expanded concept of fitness: inclusive fitness, which accounts for an individual's direct offspring alongside any equivalent offspring found among their kin.

Beyond kin relatedness, Hamilton collaborated with Robert Axelrod to analyze cooperative games in contexts where kinship was absent, and reciprocal altruism became a significant factor.

Eusociality and Kin Selection

Eusocial insect workers relinquish their reproductive rights to the queen. It has been posited that kin selection, influenced by the genetic composition of these workers, might predispose them to altruistic behaviors. The majority of eusocial insect societies exhibit haplodiploid sexual determination, resulting in an exceptionally high degree of relatedness among workers.

However, this explanation for insect eusociality has faced challenges from several prominent evolutionary game theorists, notably Nowak and Wilson. They have published a controversial alternative game-theoretic explanation, which posits a sequential developmental process and group selection effects as drivers for these insect species.

The Prisoner's Dilemma

The phenomenon of altruism presented a significant challenge to the theory of evolution, a difficulty acknowledged by Darwin himself. While individual-level selection renders altruistic behaviors seemingly illogical, universal group-level selection, posited for the benefit of the species rather than the individual, fails to withstand the mathematical scrutiny of game theory and is not generally observed in nature. Nevertheless, altruistic behavior is prevalent among many social animals. The resolution to this paradox is found in applying evolutionary game theory to the Prisoner's Dilemma, a widely studied game that evaluates the payoffs associated with cooperation versus defection.

The Prisoner's Dilemma is frequently analyzed as a repetitive game, which allows participants the opportunity to retaliate for prior defections. Numerous strategies have been investigated, with the most effective competitive approaches generally involving initial cooperation coupled with a conditional retaliatory response. Among these, the tit-for-tat strategy, characterized by a straightforward algorithm, stands out as particularly renowned and successful.

The payoff for any single round of the game is determined by a specific payoff matrix. In multi-round scenarios, individual choices—to cooperate or defect—are made in each round, yielding a particular round payoff. However, it is the cumulative payoffs accrued over multiple rounds that are critical for evaluating and shaping the overall efficacy of various multi-round strategies, such as tit-for-tat.

Example 1 illustrates the fundamental single-round Prisoner's Dilemma. In this classic formulation, a player achieves the maximum payoff by defecting while their partner cooperates; this outcome is termed temptation. Conversely, if a player cooperates and their partner defects, the player receives the worst possible outcome, known as the sucker's payoff. Under these payoff conditions, the optimal choice, representing a Nash equilibrium, is to defect.

Example 2 examines the Prisoner's Dilemma when played repeatedly. The strategy employed here is tit-for-tat, which modifies behavior based on the partner's action in the preceding round, effectively rewarding cooperation and punishing defection. Over numerous rounds, this strategy leads to higher accumulated payoffs for both players through mutual cooperation and reduced payoffs for defection. Consequently, the incentive to defect is diminished, and the sucker's payoff also decreases, although the complete elimination of "invasion" by a purely defecting strategy is not guaranteed.

Evolutionary Pathways to Altruism

Altruism is defined as an individual undertaking a strategy that incurs a cost (C) to itself while providing a benefit (B) to another individual. This cost may manifest as a reduction in capability or resources vital for survival and reproduction, or as an increased risk to the altruist's own survival. Altruistic strategies can emerge through several mechanisms, including:

The Evolutionarily Stable Strategy

The Evolutionarily Stable Strategy (ESS) shares conceptual similarities with the Nash equilibrium in classical game theory but incorporates mathematically extended criteria. A Nash equilibrium represents a game state where no player can rationally improve their outcome by unilaterally altering their strategy, assuming others maintain theirs. An ESS, however, describes a dynamic equilibrium within a very large population of competitors, where no alternative mutant strategy can successfully infiltrate and disrupt the existing population dynamic, which itself is contingent on the population's strategic composition. Therefore, a successful ESS must be effective both when it is rare—enabling its initial entry into a competing population—and when it is prevalent—allowing it to defend itself against other strategies. This implies that an ESS must be robust even when contending with identical strategies.

It is important to clarify that an ESS is not:

• An optimal strategy: An ESS does not necessarily maximize fitness, and many ESS states exist significantly below the theoretical maximum fitness achievable within a given fitness landscape.
• A singular solution: Competitive scenarios often permit the existence of multiple ESS conditions. A particular contest might stabilize into any one of these possibilities, but substantial environmental perturbations can subsequently shift the system into an alternative ESS state.
• It is not always the case that an ESS exists; some evolutionary games lack an ESS. A notable example of an evolutionary game without an ESS is "rock-scissors-paper," which is observed in species like the side-blotched lizard (Uta stansburiana).
• While often perceived as an unbeatable strategy, an ESS is more precisely defined as an uninvadable strategy.

Determining an ESS state involves two primary approaches: analyzing the dynamics of population change or solving equations that define the stable stationary point conditions. For instance, in the hawk-dove game, one can identify a static population equilibrium where the fitness of doves precisely matches that of hawks, indicating equivalent growth rates and a stable point.

Let 'p' represent the probability of encountering a hawk, which implies that the probability of encountering a dove is (1-p).

Let Whawk denote the payoff for a hawk.

The payoff for a hawk (Whawk) is calculated as the sum of its payoff when encountering a dove and its payoff when encountering another hawk.

By substituting the values from the payoff matrix into the aforementioned equation:

Whawk= V·(1-p)+(V/2-C/2)·p

A similar calculation applies to a dove:

Wdove= V/2·(1-p)+0·(p)

Consequently,

Wdove= V/2·(1-p)

To find the equilibrium, the fitnesses of the hawk and dove strategies are equated:

V·(1-p)+(V/2-C/2)·p= V/2·(1-p)

Solving this equation for 'p' yields:

p= V/C

Therefore, at this static equilibrium point, the ESS for the population percentage of hawks is calculated as ESS_{(percent Hawk)}=V/C.

Furthermore, through the application of inequalities, it can be demonstrated that the introduction of an additional hawk or dove mutant into this ESS state ultimately diminishes the fitness of their respective types, confirming both a true Nash equilibrium and an ESS. This illustration highlights that when the costs of conflict, such as injury or mortality (C), substantially outweigh the potential benefits (V), the stable population will comprise a mixture of aggressors and doves, with doves forming a larger proportion. This finding provides an explanation for various behaviors observed in natural environments.

Unstable Games and Cyclic Dynamics

Rock-Paper-Scissors Game

The Rock-Paper-Scissors (RPS) game, when integrated into an evolutionary game framework, has been employed to model natural processes within ecological studies. Researchers have also utilized RPS games in experimental economics to investigate human social evolutionary dynamics in laboratory settings. These experiments have consistently revealed the social cyclic behaviors predicted by evolutionary game theory.

The Side-Blotched Lizard and Cyclic Games

The initial natural observation of an RPS-like dynamic was identified in the behaviors and throat coloration of a small lizard native to western North America. The side-blotched lizard (Uta stansburiana) exhibits polymorphism, featuring three distinct throat-color morphs, each employing a unique mating strategy:

• The orange-throated morph displays high aggression, controls a large territory, and seeks to mate with multiple females.
• The unaggressive yellow-throated morph mimics the appearance and behavior of female lizards, enabling it to covertly infiltrate the orange-throated males' territories to mate with females, thereby gaining a reproductive advantage.
• The blue-throated morph mates with and diligently guards a single female, effectively preventing the "sneaker" yellow-throated males from succeeding and consequently displacing them in the population.

Nevertheless, the blue-throated males are unable to overpower the more aggressive orange-throated males. Subsequent research has revealed that blue males exhibit altruistic behavior towards other blue males, characterized by three primary traits: they communicate using their blue coloration, they identify and establish proximity to other (unrelated) blue males, and they will even defend their partners against orange-throated males, sometimes to the point of death. This phenomenon is indicative of a distinct cooperative game involving a green-beard effect.

Within a given population, females exhibit consistent throat coloration, which influences both their reproductive output and offspring size. This dynamic, in turn, generates density-dependent cycles, constituting another form of interaction known as the r-K game. In this context, r represents the Malthusian parameter, which dictates exponential population growth, while K denotes the environmental carrying capacity. Orange-throated females typically produce larger clutches of smaller offspring, a strategy advantageous in low-density environments. Conversely, yellow- and blue-throated females yield smaller clutches of larger offspring, a tactic more successful under high-density conditions. This interplay consequently establishes perpetual cycles intrinsically linked to population density fluctuations. The concept of cycles driven by the density-dependent regulation of two distinct strategies was initially proposed by rodent researcher Dennis Chitty, leading to the designation of these interactions as "Chitty cycles." Natural populations frequently exhibit hierarchical layers of embedded evolutionary games. These complex interactions drive Rock-Paper-Scissors (RPS) cycles in males, characterized by a four-year periodicity, and r-K cycles in females, occurring over a two-year period.

The overarching ecological scenario mirrors the dynamics of the Rock-Paper-Scissors game, thereby establishing a four-year population cycle. While the RPS game among male side-blotched lizards lacks an Evolutionarily Stable Strategy (ESS), it does possess a Nash Equilibrium (NE), characterized by perpetual orbital dynamics around the NE attractor. Subsequent to this research on side-blotched lizards, numerous other three-strategy polymorphisms have been identified in various lizard species. Some of these exhibit RPS dynamics, integrating both the male-specific game and density regulation within a single sex (males). Recent investigations have revealed that mammals also host analogous RPS games in males and r-K games in females, featuring coat-color polymorphisms and associated behaviors that instigate cyclical patterns. Furthermore, this game is implicated in the evolution of male parental care and monogamy in rodents, and it significantly influences speciation rates. The r-K strategy games are demonstrably connected to population cycles observed in both rodents and lizards.

Upon learning that these lizards fundamentally engaged in a Rock-Paper-Scissors structured game, John Maynard Smith reportedly exclaimed, "They have read my book!"

Signaling, Sexual Selection, and the Handicap Principle

Beyond the challenge of elucidating the prevalence of altruism across numerous evolved organisms, Darwin grappled with a secondary enigma: the existence of phenotypic attributes in a substantial number of species that are overtly detrimental to their survival and, consequently, should be selected against by natural processes. A prime example is the cumbersome and extensive feather structure comprising a peacock's tail. Pertaining to this matter, Darwin communicated to a colleague, stating, "The sight of a feather in a peacock's tail, whenever I gaze at it, makes me sick." The mathematical framework of evolutionary game theory has provided explanations not only for the existence of altruism but also for the seemingly counterintuitive presence of structures like the peacock's tail and other similar biological encumbrances.

The challenges inherent in biological existence, such as resource acquisition, survival, and reproduction, exhibit striking parallels with fundamental economic principles like resource management, competitive strategy, and investment risk-return analysis. Game theory, initially developed for economic analysis, has consequently proven highly effective in elucidating numerous biological behaviors. A significant refinement in evolutionary game theory, particularly relevant to economic concepts, involves the sophisticated analysis of costs. While rudimentary cost models posit uniform penalties for all competitors, this assumption is often inaccurate. More successful participants typically possess or accumulate greater "wealth reserves" or "affordability" compared to their less successful counterparts. In evolutionary game theory, this "wealth effect" is mathematically represented by "resource holding potential (RHP)," demonstrating that the effective cost incurred by a competitor with higher RHP is less substantial than for one with lower RHP. Given that individuals with higher RHP are more desirable mates, capable of producing potentially successful offspring, it is logical that sexual selection would drive the evolution of RHP signaling among rivals. For such signaling to be effective, it must be conveyed honestly. Amotz Zahavi formalized this concept as the "handicap principle," wherein superior competitors overtly display their dominance through costly exhibitions. Since individuals with higher RHP can genuinely bear the expense of such displays, this signaling is intrinsically honest and reliably interpreted by receivers. A prominent natural illustration of this phenomenon is the elaborate and costly plumage of the peacock. Alan Grafen subsequently provided mathematical validation for the handicap principle through evolutionary game-theoretic modeling.

Coevolutionary Dynamics

Evolutionary game theory typically identifies two primary categories of dynamic outcomes:

• Evolutionary games that converge towards a stable equilibrium or a point of stasis for competing strategies, ultimately leading to an evolutionarily stable strategy.
• Evolutionary games that demonstrate cyclical behavior, such as the Rock-Paper-Scissors (RPS) game, where the proportions of rival strategies within the overall population fluctuate continuously over time.

A third dynamic, termed coevolutionary, integrates both intra-specific and inter-specific competitive interactions. Illustrative examples encompass predator-prey dynamics, host-parasite co-evolution, and mutualistic relationships. Evolutionary game models have been developed to analyze both pairwise and multi-species coevolutionary systems, with the overarching dynamics varying significantly between competitive and mutualistic contexts.

Within competitive (non-mutualistic) inter-species coevolutionary systems, species engage in an evolutionary arms race, wherein adaptations enhancing competitive advantage against other species are preferentially selected. This phenomenon is reflected in both game payoffs and replicator dynamics, culminating in a Red Queen dynamic. In this scenario, interacting species must continuously evolve merely to maintain their relative fitness or ecological position.

Numerous evolutionary game theory models have been formulated to address coevolutionary scenarios. A critical element in these systems is the incessant strategic adaptation characteristic of evolutionary arms races. Consequently, coevolutionary modeling frequently incorporates genetic algorithms to represent mutational effects, while computational simulations are employed to analyze the overall coevolutionary game dynamics. The resultant dynamics are then investigated by systematically altering various parameters. Given the simultaneous interaction of multiple variables, identifying optimal solutions necessitates multi-variable optimization techniques. The mathematical benchmarks for establishing stable points in such multivariable systems are Pareto efficiency and Pareto dominance, which serve as measures of solution optimality peaks.

Carl Bergstrom and Michael Lachmann have applied evolutionary game theory to investigate the distribution of benefits within mutualistic interactions among organisms. Their models, which incorporate Darwinian assumptions regarding fitness and utilize replicator dynamics, demonstrate that the organism evolving at a comparatively slower rate within a mutualistic partnership tends to acquire a disproportionately larger share of the accrued benefits or payoffs.

Model Extensions

Initially, a mathematical model designed to analyze system behavior should maintain maximal simplicity to facilitate a foundational comprehension of its core principles, often termed "first-order effects." Once this fundamental understanding is established, it becomes appropriate to investigate whether more nuanced parameters, or "second-order effects," exert further influence on primary behaviors or contribute to the emergence of additional systemic behaviors. Subsequent to Maynard Smith's foundational contributions to evolutionary game theory, the field has undergone substantial expansion, significantly enhancing the understanding of evolutionary dynamics, particularly concerning altruistic behaviors. Key extensions to evolutionary game theory include:

Spatial Games

Evolutionary processes are influenced by geographic factors such as gene flow and horizontal gene transfer. Spatial game models incorporate geometric considerations by positioning interacting agents within a cellular lattice, where interactions are restricted to immediate neighbors. Successful strategies propagate within these immediate vicinities and subsequently engage with adjacent areas. This model effectively demonstrates how localized groups of cooperators can infiltrate and establish altruism within the Prisoner's Dilemma game, even though Tit for Tat (TFT) constitutes a Nash Equilibrium but not an Evolutionarily Stable Strategy (ESS). Occasionally, spatial structure is conceptualized as a generalized network of interactions, forming the basis of evolutionary graph theory.

Impact of Information Availability

Within both evolutionary and conventional game theory, the influence of signaling, defined as the acquisition of information, holds paramount importance. This is exemplified by indirect reciprocity in the Prisoner's Dilemma, particularly in scenarios where interactions between identical paired individuals are not recurrent. Such models reflect the dynamics of most typical non-kin social interactions. In the absence of a probabilistic measure of reputation within the Prisoner's Dilemma, only direct reciprocity is attainable; however, the availability of such information facilitates indirect reciprocity.

Alternatively, agents may possess an arbitrary signal that initially lacks correlation with strategy but subsequently becomes correlated through evolutionary dynamics. This phenomenon is known as the green-beard effect, and is also implicated in the evolution of ethnocentrism in human populations. The specific game context determines whether this mechanism fosters the evolution of cooperation or irrational hostility.

A signaling game model, characterized by information asymmetry between sender and receiver, may be applicable across various biological scales, from the molecular to the multicellular level. Examples include its relevance to mate attraction or the evolutionary development of translational machinery from RNA sequences.

Finite Populations

Numerous evolutionary games have been modeled within finite populations to ascertain the potential impact of population size, particularly concerning the efficacy of mixed strategies.

Notes

Notes

References

Davis, Morton; "Game Theory – A Nontechnical Introduction," Dover Books, ISBN 0-486-29672-5.

• Davis, Morton; "Game Theory – A Nontechnical Introduction", Dover Books, ISBN 0-486-29672-5
• Dawkins, Richard (2006). The Selfish Gene (30th anniversary ed.). Oxford: Oxford University Press. ISBN 978-0-19-929115-1.0-19-513790-6.
• Hofbauer, J. and Sigmund, K.; "Evolutionary Games and Population Dynamics," Cambridge University Press, ISBN 0-521-62570-X.
• Kohn, Marek; "A Reason for Everything," Faber and Faber, ISBN 0-571-22393-1.
• Li Richter and Lehtonen (Eds.); "Half a century of evolutionary games: a synthesis of theory, application and future directions," Philosophical Transactions of the Royal Society B, Volume 378, Issue 1876.
• Sandholm, William H.; "Population Games and Evolutionary Dynamics," The MIT Press, ISBN 0262195879.
• Segerstrale, Ullica; "Nature's Oracle – The Life and Work of W.D. Hamilton," Oxford University Press, 2013, ISBN 978-0-19-860727-4.
• Sigmund, Karl; "Games of Life," Penguin Books, also Oxford University Press, 1993, ISBN 0198547838.
• Vincent and Brown; "Evolutionary Game Theory, Natural Selection and Darwinian Dynamics," Cambridge University Press, ISBN 0-521-84170-4.

• Theme issue 'Half a century of evolutionary games: a synthesis of theory, application and future directions' (2023)
• Evolutionary game theory at the Stanford Encyclopedia of Philosophy
• Research on evolving artificial moral ecologies is conducted at the Centre for Applied Ethics, University of British Columbia.
• "The Life and Work of John Maynard Smith, Interviewed by Richard Dawkins." Web of Stories, 1997, accessed via YouTube.Source: TORIma Academy Archive