This theory of human movement implies that the human body is always compensating for the random variations in its movement, (movement variability) due to many different factors- fatigue for example (a pitcher is not the same physiologically in the first inning as he is by the 5th inning).  These tiny movement compensations are the body’s way of adapting to the inherent movement variability).  During a swing, if a golfer's hips are not ‘loaded’ enough, she may speed up the rotation of her upper body.  If her elbow is too stiff, she may loosen her wrist. These minute changes are detectable only with high-speed sensors.
 

Small low cost sensors in motion capture technology have resulted in a surge in data collection of biomechanical movement variability. Analysis of the data suggests that seemingly repetitive motion (identical movements as those used in running, biking, pitching) have inherent subtle variations in the relative positions, angles, speeds and muscle tension of body parts, without a change in result.

For example, a golfer may appear to have the ‘same’ swing from the naked eye, but when examined on a micro scale, very small biomechanical changes are taking place within each swing, without a change the club's head speed, arc of the ball or, result of the effort.  How is this possible?*

This theory of human movement implies that the human body is always compensating for the random variations in its movement, (movement variability) due to many different factors- fatigue for example (a pitcher is not the same physiologically in the first inning as he is by the 5th inning.)  These tiny movement compensations are the body’s way of adapting to the inherent movement variability. During a swing, if a golfer's hips are not ‘loaded’ enough, she may speed up the rotation of her upper body.  If her elbow is too stiff, she may loosen her wrist. These minute changes are detectable only with high-speed sensors.

Neuroplasticity and compensation to change are the benchmarks of motor learning that are revealed in elite performance. A highly skilled quarterback can throw an accurate pass while the ‘usual’ mechanics are not available to him, for example while he is being tackled, or having his arm grabbed. And an elite tennis player can perform very well while playing with a frying pan instead of a racquet. High performance in these examples are demonstrations of advanced motor control strategies-aka adaptability.

It could be related to Entropy: the scientific principle that all organized systems eventually become disorganized. But the better question perhaps?  Is movement variability beneficial to human performance?  Some researchers think it helps to avoid repetitive use injury, aids in conservation of resources (muscle fuel), or helps distribute the physical demands of the tasks across body parts in changing environments.
 

Many pedagogical techniques, especially in sport specificity training, occlude necessary sensory input during the execution a skill. This technique trains the athlete to ‘find a different way’ to get the outcome they want. Researchers sometimes call these ‘task constraints’.  Trainers and coaches sometimes call it compensation training, in accordance with the constraint led approach to motor skill instruction.

There are different physical and temporal (time-related) demands for closed/discreet skills, those with a beginning and end (pitching, basketball, free throws, American football field goal kicks) when compared to open/fluid skills (running route to tackle an opponent in rugby, or a crossover-dribble-head-fake- jump-shot in basketball. These more complex skills are often multiple discreet skills sequenced together in a longer movement. 

Additionally, different situations in all sports demand alternating attention between internal and external cues. Both offense and defense require predicting, and then reacting to external cues. While other continuous repetitive skills (cycling, running, rowing) require sensing more internal movement cues from one’s own body. In general, the role of movement variability for different types of sports, positions, and unique skills is still being explored.  The most useful studies in a human performance would reveal an optimal level of movement variability and adaptation strategies correlated with performance measures and injury rates.

The most robust experiment is the real-world environment  including real world competition, which generally is comprised of practice-then-performance. Instead of training an athlete to achieve automaticity, I suggest training for ‘elite automaticity’  - train to achieve a wider bandwidth in which elite performance occurs, to adapt to environmental changes both internal factors and external factors by training kinesthetic awareness AND adaptation strategies:
 
- Dr. Jo Shattuck