Algorithmic game theory and econometrics

Abstract

The traditional econometrics approach for inferring properties of strategic interactions that are not fully observable in the data, heavily relies on the assumption that the observed strategic behavior has settled at an equilibrium. This assumption is not robust in complex economic environments such as online markets where players are typically unaware of all the parameters of the game in which they are participating, but rather only learn their utility after taking an action. Behavioral models from online learning theory have recently emerged as an attractive alternative to the equilibrium assumption and have been extensively analyzed from a theoretical standpoint in the algorithmic game theory literature over the past decade. In this letter we survey two recent works, , Hoy et al. 2015, in which we take a learning agent approach to econometrics, i.e. infer properties of the game, such as private valuations or efficiency of observed allocation, by only assuming that the observed repeated behavior is the outcome of a no-regret learning algorithm, rather than a static equilibrium. In both works we apply our methods to datasets from Microsoft's sponsored search auction system.

Routing Games in the Wild: Efficiency, Equilibration and Regret

“…We refer to Table 2 for the number of datapoints collected for the classes of users taking either public or private transportation. [Bajari et al, 2013;Syrgkanis, 2015]. These works employ a data-driven approach to analyzing the economic behavior of real world systems and agent interactions.…”

“…Algorithmic Game Theory and Econometrics. Recently there has been a surge of interest in combining techniques from algorithmic game theory with the traditional goals of econometrics [Bajari et al, 2013;Syrgkanis, 2015]. These works employ a data-driven approach to analyzing the economic behavior of real world systems and agent interactions.…”

“…PoA= 1, which are typically considered optimal, might actually reflect traffic flows which are deadlocked in severe traffic jams. 4 Finally, PoA calculations can be invalidated if we move into theoretical models that allow for risk averse agents [Angelidakis et al, 2013;Hota et al, 2013;Meir and Parkes, 2015;Nikolova and Stier-Moses, 2014;2015;Piliouras et al, 2016]. So, summing up, we have that a high PoA might be artificially large, whereas a low PoA might actually reflect socially undesirable (in fact catastrophic) conditions.…”

How bad is selfish routing in practice?

“…Algorithmic Game Theory and Econometrics. Recently there has been a surge of interest in combining techniques from algorithmic game theory with the traditional goals of econometrics [Syrgkanis, 2015]. In [Nekipelov et al, 2015] the authors developed theoretical tools for inferring agent valuations from observed data in the generalized second price auction without relying on the Nash equilibrium assumption, using behavioral models from online learning theory such as regret-minimization.…”