Mixing Time and Stationary Expected Social Welfare of Logit Dynamics

@article{Auletta2013MixingTA,
  title={Mixing Time and Stationary Expected Social Welfare of Logit Dynamics},
  author={Vincenzo Auletta and Diodato Ferraioli and Francesco Pasquale and Giuseppe Persiano},
  journal={Theory of Computing Systems},
  year={2013},
  volume={53},
  pages={3-40}
}
We study logit dynamics (Blume in Games Econ. Behav. 5:387–424, 1993) for strategic games. This dynamics works as follows: at every stage of the game a player is selected uniformly at random and she plays according to a noisy best-response where the noise level is tuned by a parameter β. Such a dynamics defines a family of ergodic Markov chains, indexed by β, over the set of strategy profiles. We believe that the stationary distribution of these Markov chains gives a meaningful description of… 

Stability and Metastability of the Logit Dynamics of Strategic Games

It is shown that for ptential games, the mixing time is related to properties of the potential landscape and thus it is proposed to look at metastable distributions for such systems.

Metastability of Logit Dynamics for Coordination Games

A quantitative definition of “metastable probability distributions” for a Markov chain is given and the metastability of the logit dynamics for some classes of coordination games is studied.

Metastability of Logit Dynamics for Coordination Games (full version)

Logit Dynamics [Blume, Games and Economic Behavior, 1993] is a randomized best response dynamics for strategic games: at every time step a player is selected uniformly at random and she chooses a new

Reversibility and Mixing Time for Logit Dynamics with Concurrent Updates

The dynamics where at every time step every player simultaneously updates her strategy according to the logit choice function is called the “all-logit”, as opposed to the classical “one-logIt” dynamics, which is a subclass of potential games.

Logit Dynamics with Concurrent Updates for Local Interaction Potential Games

It is proved that local interaction potential games characterize the class of games for which the all-logit dynamics is reversible, and it is shown that the mixing time of the all"-logit" dynamics has the same twofold behavior that has been highlighted in the case of the one- logit.

Metastability of the Logit Dynamics for Asymptotically Well-Behaved Potential Games

It is proved that every potential game, for which the behavior of the logit dynamics is not chaotic as n increases, admits distributions stable for a super-polynomial number of steps in n no matter the players’ irrationality and the starting profile of the dynamics.

Convergence to Equilibrium of Logit Dynamics for Strategic Games

The first general bounds on the mixing time of the Markov chain associated to the logit dynamics for wide classes of strategic games are presented, and nearly tight bounds for potential games and games with dominant strategies are proved.

Metastability of Asymptotically Well-Behaved Potential Games

It is proved that any such game admits distributions which are metastable no matter the level of noise present in the system, and the starting profile of the dynamics, which can be quickly reached if the rationality level is not too big when compared to the inverse of the maximum difference in potential.

Decentralized dynamics for finite opinion games

Metastability of Asymptotically Well-Behaved Potential Games - (Extended Abstract)

It is proved that any such game admits distributions which are metastable no matter the level of noise present in the system, and the starting profile of the dynamics, which can be quickly reached if the rationality level is not too big when compared to the inverse of the maximum difference in potential.

References

SHOWING 1-10 OF 24 REFERENCES

Convergence to Equilibrium of Logit Dynamics for Strategic Games

The first general bounds on the mixing time of the Markov chain associated to the logit dynamics for wide classes of strategic games are presented, and nearly tight bounds for potential games and games with dominant strategies are proved.

On the Inefficiency Ratio of Stable Equilibria in Congestion Games

Observations show that in the game environments where a small noise is present, the system as a whole works better than what a pessimist may predict, and suggest that in congestion games, introducing aSmall noise in the payoff of the agents may improve the social welfare.

The Statistical Mechanics of Strategic Interaction

I study strategic interaction among players who live on a lattice. Each player interacts directly with only a finite set of neighbors, but any two players indirectly interact through a finite chain

Population Games And Evolutionary Dynamics

  • W. Sandholm
  • Economics
    Economic learning and social evolution
  • 2010
The text first considers population games, which provide a simple, powerful model for studying strategic interactions among large numbers of anonymous agents, and studies the dynamics of behavior in these games, providing foundations for two distinct approaches to aggregate behavior dynamics.

Sink equilibria and convergence

It is argued that there is a natural convergence process to sink equilibria in games where agents use pure strategies, and the price of sinking is an alternative measure of the social cost of a lack of coordination, which measures the worst case ratio between thevalue of a sink equilibrium and the value of the socially optimal solution.

Convergence to Equilibrium in Local Interaction Games

The present paper characterizes the rate of convergence as a function of graph structure and derives a dichotomy between well-connected (e.g. random) graphs that show slow convergence and poorly connected, low dimensional graphs that shows fast convergence.

The complexity of game dynamics: BGP oscillations, sink equilibria, and beyond

It is established that it is PSPACE-complete to tell whether a system of path preferences in the BGP protocol can lead to oscillatory behavior and a new equilibrium concept inspired by game dynamics, unit recall equilibria, is proposed, which is shown to be close to universal and universal and both tractable and universal.

Calibrated Learning and Correlated Equilibrium

For each correlated equilibrium there is some calibrated learning rule that the players can use which results in their playing this correlated equilibrium in the limit, and the statistical concept of a calibration is strongly related to the game theoretic concept of correlated equilibrium.

The Theory of Learning in Games

The Diffusion of Innovations in Social Networks

We consider processes in which new technologies and forms of behavior are transmitted through social or geographic networks. Agents adopt behaviors based on a combination of their inherent payoff and