for action in gameState.getLegalActions(agentInd):

        if agentInd == gameState.getNumAgents()-1: #then consider max's moves
          if curDepth == self.depth: #leaf node
            return self.evaluationFunction(gameState)
          value = min(value, self.maxVal(gameState.generateSuccessor(agentInd, action), curDepth+1))
        else: #if agentInd < gameState.getNumAgents()-1: #accumulate all ghost moves
          value = min(value, self.minVal(gameState.generateSuccessor(agentInd, action), curDepth, agentInd+1))