In this paper, we propose a novel approach for verifying the compliance of turn-based multi-agent Reinforcement Learning (TMARL) agents with complex requirements in stochastic multiplayer games. Our method overcomes the limitations of existing →