A pinboard by
Ran Snitkovsky

PhD Student and teaching assistant, Tel Aviv University


Games played by agents who operate in a service system, according to an economic choice model.

Avoiding congestion is a task we are faced with on a daily basis, e.g., in traffic, cellular and Wi-Fi networks, waiting lines and shared resource accessing in general. Among the reasons we wish to avoid congestion are saving time, improving the quality of service and reducing expenses. In this particular research area we wish to investigate processes of interaction between customers and service providers in a system affected by congestion, when participants are granted the freedom of choice. A natural way to model problems of congestion is using the tools of Queueing Theory, which provides the mathematical framework of analyzing congested systems. This however is not enough, since we assume customers utilize their freedom to benefit their own interest. The combination of a Game Theoretic approach together with the tools of Queueing Theory, a framework we call "Rational Queueing", allows us to analyze congested systems where customers and servers behave strategically.

The models that we analyze draws inspiration from the behavior of users of sharing-economy platforms, such as Uber, Airbnb or TaskRabbit, as well as from other more traditional platforms such as stock exchanges and call centers. In this kind of interactions customers and service provider ought to make several decisions such whether to join the platform or use some outside option, how much money to offer/ask for the service, and more, depending on the choice model.

Of particular interest in the Game Theoretic approach is the notion of Nash Equilibrium, i.e., an economic situation where no single participant can increase his/her own benefit by changing his/her own decision. Smith's famous principle of "invisible hand" maintains that individuals' efforts to pursue their own interest may frequently benefit society more than if their actions were directly intending to benefit society. Nonetheless, in Rational Queueing models, equilibrium is typically not optimal in the sense of social welfare. Among the questions we attempt to resolve in the research are what kind of participants' behavior induces equilibrium in the market, and is the equilibrium behavior indeed consolidates with the socially-optimal behavior? If not, than what managerial actions can be taken to reduce the discrepancy? Studying these matters is our main goal, but also addressing other issues of importance that arise upon exploration.