The current trainings of deadly force for police officers are standard cardboard targets at a shooting range. Each case the officer is expected judiciously before shooting but he may not have much time. Besides, the decision occurs only in his mind. However, the trainings about how to make the decision to justifying in using deadly force exists a gap. This study proposes a decision-making process may help fill the training gap. First, for understanding the mental behaviors, some literatures of artificial intelligence and superfast sports have reviewed. Second, it defined 3 stages of deadly force use: 1) adduce evidence, 2) judgment and 3) response. Then the training goal is help officers in gaining mass decision-making experiences by scenario-based learning.