Arguing from experience using multiple groups of agents
@en