The goal of the paradigm shift in Air Traffic Management (ATM) is to increase its overall performance by means of redesigning processes, evolving to a more automated, autonomous and predictable system. Nevertheless, when dealing with automation, it is important to determine until what extent can any part of the system or process be automated, that is, to determine the right player. It is also likely to be tempted to make the system as autonomous as possible, avoiding the stiffness introduced by centralised processes, which means to choose the right place to drive it. Finally, it is also commonly accepted that the sooner an activity is planned, the more predictable the system will be, i.e., to determine the right time. However, reality creates constraints that make it impossible to reach the ideal status: fully automated, completely autonomous and totally anticipated. Considering the ATM system as a set of tasks and functions, its allocation can be defined as their placement in time, place and player. This paper presents operational research methodologies to estimate the best time, the best place and the best player for optimal performance of the ATM system.