Short Bio
I am a Full Professor (Professeur des Universités) in Economics at the University Paris 1 Panthéon-Sorbonne and a Research Fellow at the Centre d’Economie de la Sorbonne (since 2019). I graduated in 2009 with a Ph.D in Economics from the University of Minnesota under the supervision of Aldo Rustichini. Since 2021, I am also a member in the Council of the Cyprus Agency of Quality Assurance & Acceditation in Higher Education.
I am an applied game theorist interested in modelling behavior. I am particularly intrigued by the study of behavior that deviates from perfect rationality. I thus employ experiments to collect data, which I analyze to better understand (and model) economic decision-making. Over the years, my research interests have spanned from an analysis and modelling of behavior in repeated games to that in prediction markets.
New Research
Non-Parametric Strategy Inference in Repeated Games coauthored with Laurent Mathevet, Julian Romero and Huanren Zhang develops two non-parametric, pattern-mining methods to investigate learning dynamics and behavioral regularities in infinitely-repeated games. Our first approach, the action-convergence criterion, draws on string-searching algorithms to identify recurring action profiles by minimizing mismatches over variable time horizons. Our second approach utilizes a modified k-means clustering algorithm, which transforms experimental sequences into numerical vectors through multi-temporal weighting, to endogenously reveal distinct behavioral 'storylines' across the short, medium, and long run. We apply this framework to a rich set of 2X2 games with high discount factors. Empirically, we establish a sharp counterpoint to the theoretical abundance of
equilibria: long-run play is highly concentrated, exhibiting significant entropy reduction, and stability substantially increases over time. We find that efficiency and egalitarianism act as powerful learned attractors, becoming more pronounced as interactions mature. Finally, using a Dynamic Programming approach, we statistically reject the null hypothesis of independent play, confirming that these patterns reflect genuine intertemporal correlation consistent with purposeful coordination.
The website was updated on February 1, 2026