Theories are one thing, applying them is another. The example of the windshield wiper (Anticipation, Part 2) serves as a basis to apply the model of anticipation. You will have a hard time following the details, if you haven’t read certain blog entries. So, if you feel unfamiliar with some of the presented things, do yourself a favor and roam the blog for getting used to the mentioned aspects (Anticipation, part 1, part 2, part 3). Following, a short revision of the model and the example:
Model of anticipation
1. Example of the windshield wiper (fixed speed and displacement)
The basic principle of the circular process has already been addressed. The missing elements involve response selection, comparison of stimuli and the response execution.
The response selection in this example involves the choice of the directional change. The internal model knows that there must be a directional change (causality is given). The partners’ hand moves e.g. from left to right. So, it is clear that the directional change involves a movement from right to left. Additionally, the speed and the maximum displacement are set for a few seconds. This helps the internal model to determine and scale the new movement. Basically, it is not a big deal to choose the right movement.
The comparison of the stimuli involves the target stimulus and the actual stimulus. The target stimulus is again derived from the internal model. The maximum displacement and the speed are known. Therefore, the target stimulus can be chosen. The internal model estimates when the actual stimulus arrives. Still, it is mostly derived by looking at the partners’ hand at low speed. Thus, a comparison of both stimuli is possible. This is not the case at high speed, which is covered in the following paragraph.
The response execution must be divided into two parts. This phase usually needs much more attention, but I want to keep it simple. At low speed, the stimuli arrive at a processible speed. Thus, the response execution can be perfectly scaled to the ongoing events. At very high speeds, a continuous perception of the ongoing events is no longer possible due to the limitation of the circular process (Anticipation, part 2, mentioned delay). Now, a predetermined and scaled movement, mostly derived from the internal model, is continuously executed. This means that your movement and your partner’s movement seem to go on in parallel, but they are not causally connected. This is only the impression of an external observer. If the reacting partner were to close his eyes, he could go on without even looking at his partner’s movement. Both modes complement each other at intermediary speeds.
The more interesting part is the appearance of errors as soon as speed and displacement are no longer fixed.
2. Example of the windshield wiper (variable speed and displacement)
Most of problems in this exercise arise from a wrong internal model. The internal model must be “fed” with information to have an idea what is happening. Furthermore, the information must be decent enough to project future events. The appearance of patterns and traits was consistent in the first part. Fixed speed and displacement helped a lot to reduce the number of possibilities. Basically, there was just one possibility. But now, at a certain speed, you miss information about when and how the directional change will occur. You can no longer rely on your internal model to solve the situation. The time to update the internal model is insufficient. Remember the tree diagram? Let’s compare the two exercises. This time, for convenience reasons, it’s a depiction as a circular process. The second part is shortened and only one element from the circular process, together with its alternatives, is depicted.
Fixed speed and displacement
Variable speed and displacement
As you can see, the number of possibilities to choose from started to grow after passing a certain speed. The circular process is overloaded. Even though the information is available (your partner is still moving, isn’t he?), identifying it and incorporating it into the internal model takes too much time. This also means that the drawn conclusion can no longer match the causality condition. Additionally, you can no longer count on the internal model to be an appropriate forecast of future events. Therefore, the different alternatives cannot be isolated. Now, based on these problems, how should your internal model choose which movement is appropriate to solve the given task? The missing stimuli make it even worse (please remember, the stimuli exist, but are unavailable). The exact same stimuli are needed for the comparison of the target to the actual stimuli. Both stimuli (!) are now missing. A chosen movement cannot be executed systematically.
I think that this little paragraph is enough to indicate some of the more pressing problems. There are certainly others, and this isn’t my intended format for a checklist. The upside is that such reflection clearly delivers a long and detailed list. The addressed problems can be examined separately. Afterwards, counter-measures can be taken. You have to face the fact that anticipation is not an option in such a situation if no counter-measures can be found.
A final note: Have you really tried the exercise yourself? If you slowly increased the speed, you may have noticed that there was a significant lag between your hand and your partners’ shortly before you completely failed. This lag indicates the incoming failure if the speed is increased just a little more. Beware of this effect. If you already have this effect in training, then it is very likely that the movement will fail in a less “training-friendly” environment.