JOVANA
Explore Library Glossary Getting Started Three Levels Fields How it works Mission
Join the mission
Back to the library
Biology 1973

The Logic of Animal Conflict

John Maynard Smith & George R. Price

Why animals usually bluff instead of fighting to the death — and the one strategy evolution cannot beat.

Choose your version
In depth · the introduction

Two deer lock antlers and shove — but rarely gore each other. Two rattlesnakes wrestle without ever using their fangs. Why do animals pull their punches?

The big idea

For a long time biologists answered: animals hold back “for the good of the species” — if everyone fought to the death, the species would suffer. Maynard Smith and Price showed you don’t need that story. Treat fighting as a game. Each animal follows a strategy — a rule for when to bluff and when to escalate — and the strategies that leave the most offspring spread. The winning rules turn out to be the restrained ones.

Their key invention was a test for a strategy that lasts. A strategy is evolutionarily stable if, once nearly everyone in the population uses it, no rare newcomer playing anything else can do better and take over. A population of pure fighters fails this test — they injure each other so often that a more cautious mutant out-breeds them. Restraint wins not because it’s noble, but because it pays.

How it came about

John Maynard Smith was a leading British evolutionary biologist (and a former aircraft engineer) with a taste for turning fuzzy arguments into clean models. The other half of the paper, George Price, was an American chemist and engineer who had taught himself evolution and sent Nature a long, strange manuscript about why animals don’t use their deadliest weapons on each other. Maynard Smith was asked to judge it for the journal. Instead of rejecting it, he saw the gold inside, tracked Price down, and the two compressed the idea into four pages. Maynard Smith spent the rest of his career crediting Price — whose own life ended tragically in 1975 — for the central concept they had hit upon.

Why it mattered

It gave biology a whole new tool: game theory, but with “survival of the fittest” doing the choosing instead of clever players. That single move explained a pile of puzzles — why fights are ritualized, why animals respect ownership, why males and females are born in roughly equal numbers — all as strategies no individual can profitably break away from. And because the logic doesn’t care whether the “players” are deer or companies or computer programs, it spread into economics, politics and artificial intelligence.

A way to picture it

Think of drivers at a four-way junction. If everyone barrels through (all “Hawk”), there are constant crashes and everyone loses. If everyone always yields (all “Dove”), a single pushy driver who never stops gets through every time — so “always yield” can’t hold either. The stable habit is a mix: most people take turns, but the system tolerates a certain number of chancers. Crank up the cost of a crash — make the cars faster and the impacts deadlier — and people drive far more cautiously. That’s exactly what the model says: more dangerous weapons, fewer Hawks.

An interactive chart with two payoff lines, one for a Hawk and one for a Dove, drawn against how many in the population play Hawk. The lines cross at the stable mix; a bar below shows the resulting proportion of Hawks to Doves. A slider sets how costly an injury is, and raising it shifts the mix toward fewer Hawks.

Where it sits

Darwin (1859, in this Library) gave the principle — selection favours what reproduces — but not the mathematics of strategies that depend on what everyone else is doing. John Nash (1950, also here) supplied the equilibrium idea for rational players; Maynard Smith and Price translated it for blind natural selection. Set beside Hamilton’s kin selection (1964, also here), which explains cooperation among relatives, the ESS explains restraint among strangers — together they map much of why animals do not simply tear each other apart.

The original document
Original source text
J. Maynard Smith & G. R. Price · Nature, Vol. 246, No. 5427 (2 November 1973), pp. 15–18.
The puzzle (from the abstract)
Conflicts between animals of the same species usually are of “limited war” type, not causing serious injury.
This is often explained as due to group or species selection for behaviour benefiting the species rather than individuals.
Game theory and computer simulation analyses show, however, that a “limited war” strategy benefits individual animals as well as the species.
The contest
The authors imagine repeated pairwise contests over a resource. In each move an animal can use a “conventional” tactic (display, ritual, no real harm) or a “dangerous” one (escalate, risk injury). A strategy is a rule for choosing — including how to respond to what the opponent just did. Payoffs are added up across contests as a proxy for Darwinian fitness, and computer simulation lets the strategies play many rounds against one another.
The five strategies
Five rules are pitted against each other: Mouse (never escalates — always conventional, retreats if attacked); Hawk (always escalates, fights until injured or the rival flees); Bully (escalates only until someone fights back, then yields); Retaliator (conventional by default, but escalates in kind if attacked); and Prober-Retaliator (a Retaliator that occasionally probes with an escalation to test the opponent).
The result
Pure Hawk is not stable — a population of Hawks pays so much in injury that a more cautious mutant does better. The strategies that resist invasion are the “limited-war” ones: Retaliator (and Prober-Retaliator) are evolutionarily stable, or nearly so. Restraint, in other words, is what individual selection settles on. The paper names this property the evolutionarily stable strategy: a strategy that, if adopted by almost all of a population, no rare alternative strategy can do better against.
[ … ]
J. Maynard Smith & G. R. Price · Nature · 2 November 1973