Coping with complex political dilemmas characterizing the management of common resources requires developing stakeholder learning networks to create a cooperative decision-making environment in which trust, understanding, and mutual trust develop over time. Involved in such social learning processes, local actors are strengthened, their empowerment being a consequence of their involvement rather than a starting point. Fostering the dynamic of involvement of local actors in the processes of adaptive co-management should then allow the creation of a virtuous circle. Researchers have proposed several tools and methods to foster social learning among small rural farmers, role-playing being one of them. An RPG session takes place as a succession of game turns, which encourages participants to build together iteratively a mutual understanding of the interdependencies connecting the entities of the simulated system. On this basis, participants seek agreements to define and implement management measures. In this sense, an RPG session constitutes a “shared experience” to which it may be interesting to refer in order to introduce local populations to concepts linked to the commons such as “commoning”. Here we present how RPGs are used with Masai pastoralists living around Amboseli National Park (Kenya). We first conducted three sessions of a generic RPG called ReHab (Resources and Habitat), initially intended to make students aware of the importance of communicative rationality in contested landscapes. An inhabitant was trained to facilitate the sessions. The participants had no difficulty in “entering the game “. Certain situations that arose during the game clearly echoed existing practices in reality (for example, collective decision to enter the protected area in the face of a critical situation, collective decision to create a reconstitution area). Now that the local players have understood the advantage of working with such stylized tools, we are going to develop a contextual game to specifically address their current problems.