Skip to content

Complete mixing

An evolutionary-game modeling assumption that every population member encounters every other member with equal probability, removing spatial or network assortativity from expected payoffs and replicator dynamics.

Version
v1 · 2026-09-28 · History
Domain-specific #
8596
Domain group
Natural Sciences
Origin domain
Biology & Ecology
Subdomains
Evolutionary Game Theory, Population Biology → Biology & Ecology

Core Idea

In evolutionary game theory, complete mixing removes interaction structure by giving each individual equal opportunity to encounter every other. Expected payoff is computed from global type frequencies, an assumption built into standard replicator dynamics. The assumption is analytical, not a claim that organisms literally meet everyone. The assumption is analytical, not a claim that organisms literally meet everyone.

How would you explain it like I'm…

Everybody Meets Everybody

Imagine a big party where everyone is equally likely to bump into everyone else, no matter where they stand or who their friends are. Scientists sometimes pretend animals meet like that to make the math easier, even though real animals mostly meet their neighbors and family.

Pretend Everyone Meets Equally

Scientists who study how behaviors spread in groups of animals sometimes use a shortcut called complete mixing. They pretend every animal has the same chance of meeting every other animal. Then they can figure out how well a behavior pays off just by knowing how common each behavior is in the whole group. It is a pretend rule for the math, not a claim that animals really meet everyone. Real animals often meet neighbors, family, and the same partners again and again, and when that matters, the shortcut can give the wrong answer.

Well-Mixed Population Assumption

In evolutionary game theory, the payoff an individual gets depends on whom it interacts with. Complete mixing, or a well-mixed population, is the assumption that each individual has an equal chance of encountering every other, so there is no interaction structure. Under that assumption, an individual's expected payoff can be computed just from the overall frequencies of each strategy type in the population. Standard replicator dynamics, which models how strategy frequencies change according to payoff, builds this assumption in. It is an analytical simplification, not a claim that organisms literally meet everyone. In reality, spatial closeness, social networks, kinship and repeated contact make encounters nonuniform, and when those affect payoffs a well-mixed model can miss effects like clustering and assortment of similar types.

 

Complete mixing, in evolutionary game theory, removes interaction structure by giving every individual equal probability of encountering every other. Expected payoffs are then functions of global type frequencies alone, which is the assumption built into standard replicator dynamics. It is an analytical device rather than an empirical claim that organisms meet all others. Real populations usually have nonuniform encounters driven by spatial proximity, network ties, kinship, and repeated contact. When those mechanisms affect payoffs, the well-mixed model can miss clustering and assortment, meaning like types interacting preferentially, which can change which strategies succeed.

Scope of Application

Use complete mixing as an explicit baseline, with the interaction population and sampling rule declared. Use complete mixing as an explicit baseline, with the interaction population and sampling rule declared.

  • Replicator dynamics. Averages payoff over global frequencies.
  • Evolutionary games. Provides a structure-free comparison.
  • Simulation. Defines all-to-all or uniform random encounters.
  • Spatial models. Serves as a contrast to local interaction.
  • Network games. Tests effects of degree and assortment.

Clarity

Large population and random movement do not prove uniform partner choice. The positive test is the probability kernel used for payoff-bearing encounters, not a verbal claim that the system is mixed. The closest near miss sets the boundary: Random mixing on a finite network is closest: contacts may be randomized among existing edges while edge constraints still make partner probabilities unequal.

Manages Complexity

Removing interaction topology simplifies dynamics to type frequencies, but that compression discards clustering, correlations, and local competition. Comparison with structured variants reveals which results depend on the assumption. The central tractability–interaction realism tradeoff is this: Global averaging simplifies analysis while erasing spatial assortment. A second randomness–uniformity tension matters because A stochastic encounter process can still be nonuniform.

Abstract Reasoning

Use three linked moves: define the population and payoff-bearing encounter; write the partner-sampling probability for each pair; check whether space, network, or type changes that probability. As a collapse test, the case exits when space, kinship, network degree, repeated partners, or assortative matching changes encounter probabilities. A fourth check is to compute global expected payoffs only after uniformity holds. A final check is to compare a structured model when locality is biologically relevant.

Knowledge Transfer

Uniform-sampling logic transfers to epidemiological and social interaction baselines, but the payoff and update equations do not. The stopping boundary is unequal encounter probability, even if the system is colloquially called random. The nearest stopping boundary is explicit: Random mixing on a finite network is closest: contacts may be randomized among existing edges while edge constraints still make partner probabilities unequal. The inclusion test remains: A model qualifies when interaction partners are sampled uniformly from the whole population for the relevant payoff evaluation. The structure no longer applies when the case exits when space, kinship, network degree, repeated partners, or assortative matching changes encounter probabilities. No canonical parent prime is currently asserted; broader structural comparisons remain related-prime analogies until separately adjudicated in the DAG. Partners are drawn from a declared distribution. Local structure is replaced by a global average.

Relationships to Other Abstractions

Local relationship map for Complete mixingParents appear above the current abstraction, mutual partners to the right, and children below. Node labels state whether each abstraction is prime or domain-specific; colors identify relation types.Complete mixingDOMAINPrime abstraction: Assumption — is a kind ofAssumptionPRIME

Current abstraction Complete mixing Domain-specific

Parents (1) — more general patterns this builds on

  • Complete mixing is a kind of Assumption Prime

    Complete mixing is a strict kind of Assumption: its frozen identity entails the parent's defining structure while adding domain-specific restrictions.

Hierarchy path (1) — routes to 1 parentless root

Neighborhood in Abstraction Space

Complete mixing sits in a moderately populated region (55th percentile for distinctiveness): it has near-neighbors but no dense thicket of look-alikes.

Family — Selection, Speciation & Experimental Evolution (22 abstractions)

Nearest neighbors

Computed from structural-signature embeddings · 2026-10-08