On open-loop and closed-loop nash strategies