Reinforcement learning for board games

This project focuses on learning strategies that help humans to improve their performance. Additionally, we will consider the problem of detecting non-human players. An initial focus will be the card game 6 nimmt!.

  • Can we translate strategies learnt by machines into rules simple enough to be understood and used by human players?
  • Can we distinguish ‘real’ players from ‘bots’ based on their play style?

Further details are available on request.


  • R. Bettker, P. Minini, G. Pereira and J. V. C. Assunção, “Towards playing AIs for 7 Wonders: main patterns and strategies for 3-player games”, 2021 20th Brazilian Symposium on Computer Games and Digital Entertainment (SBGames), 2021, pp. 172-181, doi: 10.1109/SBGames54170.2021.00029.
Sebastian Vollmer
Sebastian Vollmer
Professor for Applications of Machine Learning

My research interests lie at the interface of applied probability, statistical inference and machine learning.