The Nash Equilibrium with Inertia in Population Games

TL;DR

This paper introduces inertial Nash equilibria in population games, accounting for switching costs, and proposes a new algorithm with convergence guarantees, demonstrated through ride-hailing platform analysis.

Contribution

It defines inertial Nash equilibria incorporating switching costs and develops a convergent better-response algorithm for their computation.

Findings

Inertial Nash equilibria include all classical Nash equilibria.

Classical algorithms are ineffective with switching costs.

The proposed algorithm converges to inertial Nash equilibria.

Abstract

In the traditional game-theoretic set up, where agents select actions and experience corresponding utilities, an equilibrium is a configuration where no agent can improve their utility by unilaterally switching to a different action. In this work, we introduce the novel notion of inertial Nash equilibrium to account for the fact that, in many practical situations, action changes do not come for free. Specifically, we consider a population game and introduce the coefficients $c_{ij}$ describing the cost an agent incurs by switching from action $i$ to action $j$. We define an inertial Nash equilibrium as a distribution over the action space where no agent benefits in moving to a different action, while taking into account the cost of this change. First, we show that the set of inertial Nash equilibria contains all the Nash equilibria, but is in general not convex. Second, we argue that classical algorithms for computing Nash equilibria cannot be used in the presence of switching costs. We then propose a natural better-response dynamics and prove its convergence to an inertial Nash equilibrium. We apply our results to predict the drivers' distribution of an on-demand ride-hailing platform.