This paper argues that using a game with uncertain outcomes in a repeated setting is more realistic and allows for strategies that mimic human behavior, like altruism and fairness. This approach enables the analysis of evolutionary strategies that traditional methods cannot capture.