Mean Field Game

A game with a continuum of players where each player's influence on aggregate is negligible, simplifying equilibrium to a PDE system.

Ask the Game theory assistant 1 min read · Updated September 9, 2026

Definition

In a mean-field game, the state of the game is the distribution $m_t$ of agents over states. Each agent solves a Hamilton-Jacobi-Bellman equation $\partial_t u + H(x,\nabla u, m) = 0$ , and the population evolves via the Fokker-Planck equation $\partial_t m - \mathrm{div}(m\,\partial_p H) = 0$ . The system $(u,m)$ forms the MFG equilibrium.

Lasry-Lions (2006) and Huang-Caines-Malhamé (2006) independently developed the theory.

Intuition

With millions of agents, no individual can affect aggregate — but all are affected by it. The equilibrium couples optimal control (forward by agents) and population dynamics (forward in time).

MFG reduces a massive interactive system to two coupled PDEs, making analysis tractable.

Worked example

Taxi fleets: each driver chooses a location; the distribution of all drivers determines earnings. The MFG equilibrium captures spatial equilibrium.

Emission models: each firm decides abatement; aggregate emissions determine climate damage.

The math

A stationary MFG: $\lambda + H(x,\nabla u, m) = 0$ and $\mathrm{div}(m\,\partial_p H)=0$ , solved together with boundary conditions.

Existence and uniqueness of classical solutions require monotonicity of the Hamiltonian in $m$ (Lasry-Lions).

Where it is used

Economics (labor markets, wealth distribution), finance (systemic risk), engineering (autonomous driving), and epidemiology.

Active frontier in applied mathematics and algorithmic game theory.

Go deeper

More in Game theory

Assembled from the ReLU.chat curated knowledge base. These explanations are concise on purpose; check the sources for anything important.