JointPPO: Diving Deeper into the Effectiveness of PPO in Multi-Agent Reinforcement Learning | Synapse