According to Deming, management is not playing games; it is all about prediction. A manager must be able to predict with a fair degree of certainty the likelihood of certain outcomes. How can this ever be achieved? It stands to reason that the first principle of prediction is knowledge of current state and this is why establishing control limits is vital.
The best way to grasp this idea is perceive it from the point of sports. Suppose you were hired as a coach for, say, a tennis player. How in the world would you guide the player into sustained victory?
As the title says, you must first establish what the capabilities of the player are. It may be that they have a salient strength in one aspect of the game but may be totally lacking or even incapable in an essential skill. In either case, your goal as the coach is to find out where their limits lie. These are what control limits are.
The difficulty with control limits is that they only make sense under conditions of stability. Exposing the player to a very wide variety of drills such as sparring games will mean that they are never stable and may be sampling a very high dimension space. The resulting data will be useless. Rather, confine the exploration to only one - not even a small number - task. For example, play forehand for a long time and measure various aspects of forehand. The goal is to know what type of forehand they can confidently respond with.