How Noise Creates Diversity express

Original: "Emergence of polymorphism in stochastic evolutionary games"
· Bill Nunn, Marcel Ortgiese, Tim Rogers
arXiv:2607.14950 · 2026-07-16 · CC BY 4.0 · 1 min
Random noise can reinforce, not erase, diversity in a population.

📄 Showing the "Simple" version — "Popular" is not ready yet. Add it to favorites to help prioritize it.

For a long time, biologists argued: if a herd of animals stably maintains a mix of aggressive and peaceful individuals, does each one alternately change behavior, or do some always fight while others always yield? In classical equations, these options are indistinguishable. The 'Hawks and Doves' game was invented to describe the struggle for prey: hawks fight to the end, doves retreat, avoiding injury.

Physicist Ludwig Boltzmann discovered that disorder (random noise) in any system usually averages everything out. It seemed that in a population, random fluctuations in numbers should gradually erase differences, turning everyone into 'middling' individuals — this idea was dubbed the 'trembling hand' hypothesis. However, a new study showed the opposite: in small groups, noise doesn't blend, it sorts, separating individuals into pure 'hawks' and 'doves' for a long time — like a invisible framework holding the structure. Instead of a homogeneous crowd, a stable diversity emerges.

Randomness, once seen as the destroyer of order, turns out to be its architect. Mathematician John von Neumann, who laid the foundations of game theory, would probably be surprised by this twist: noise not only creates diversity but also guards it.

In the 'Hawks and Doves' game, the strategies are named after birds, but it's just an abstraction; in reality, hawks and doves don't play such a game.

Scientists
Ludwig BoltzmannRudolf ClausiusJacob BekensteinStephen HawkingJames Clerk MaxwellEdward Witten
Concepts
game theory stochastic processes Cope's rule stochastic gene expression evolution Price equation epidemic random forest Baldwin effect Fisher's fundamental theorem of natural selection Mendel's law of independent assortment species Wright–Fisher model emergent collective learning random variable chaotic attractor genetic algorithm probabilistic model biological stochastic dynamics Hamilton's rule