Manuscript abstract: We designed the board game Savanna Life to engage communities in the Greater Serengeti-Mara Ecosystem in Tanzania and Kenya in sustainable development planning. The game simulates increasing livelihoods constraints and provides a safe space for exploring alternative livelihood and investment strategies. Here we explore differences in game field preferences, collaboration and game performance between player using moral foundation, Schwarts value orientation and structure-agency theories as a framework. The game facilitated playing different strategies and promoted learning and adaptation throughout games. Using partial models, we found preferences attributable to gender, tribe, and nationality following these theories. Equally important, some preferences contradicted predictions perhaps indicating that players used the game to explore strategies deviating from what they would normally do and would be too risky to try in the real world. Women's preferences and performance, following the logic of the game, were better aligned with development objectives and food security than that of men. However, women were also less likely to collaborate, indicating that targeting development investments at women as suggested by the results should also increase the possibility of interactions and collaboration between women. Tanzanian nationals and the Maasai showed a higher preference for healthcare and education, indicating a need for relevant investment, in Maasai land. Women had a higher preference for poaching in the game. Therefore, more attention may be required to address women's concerns about household finances and food security in efforts to reduce poaching. Virtually all players reported learning insights leading to planned real-life changes in livelihood and investments strategies. Although we cannot claim that playing Savanna Life leads to positive real-life conservation and development outcomes, we believe that these statements reflect real-world intentions and aspirations.
Steps to reproduce
This is a simplified STATA do file with much of the data management omitted for ease of reproduction.