Become a Movement Degenerate

Practice makes perfect, perfect practice makes perfect, 10,000 hours of training are needed, repetition, repetition, repetition etc….

Truism’s abound when it comes to discussion surrounding skill development, but what actually works? 

In his book How We Learn to Move: A Revolution in the Way We Coach and Practice Sports Skills author Rob Gray introduces the concept of degeneracy. 

“Degeneracy occurs in a system when structurally dissimilar components can perform similar functions – they are effectively interchangeable.”

We as humans experience internal AND external variability. Imagine the boxer, throwing a jab – a basic strike that most are taught through a series of with rules and top down instruction (keep your chin tucked, etc) in a decontextualized drill. However, when you watch a a fight in the ring the boxer’s feet, hips, head, and neck may be in 100’s of different arrangements during each jab of a fight. The opponent presents a high amount of external variability (they don’t want to be hit!) while the boxer’s body presents internal variability  – the angle of the shoulder, elbow, and trunk will be slightly different with every jab he or she throws in order to reproduce the knock out punch each time. 

Alas, just as man cannot step into the same river twice (for he is not the same man and it is not the same river!), so too an athlete cannot complete the same movement – for she is not the same athlete and it is not the same movement environment!

Repetiton Without Repetition & the Cyclography Revolution

Novice Blacksmith Cyclography collected by Bernstein

     Nikolai Bernstein, the father of modern Kinesiology, and the man who coined the term Biomechanics conducted his famous hammer strike study using cyclography. Cyclography is a method of measuring movement. By putting tracers on a test subject, a scientist can analyze joint movement moment by moment. Now, this is something that we can now do with machines, but in the early 1900’s it took painstaking hours of going second by second to trace lines of movement. In his famous motor control study Bernstein and colleagues analyzed expert and novice blacksmiths striking a hammer on an object.

     At the time, the general expectation was that skilled movers would complete all elements of a movement accurately over and over again. It is from assumption that identical repetitions of a desired action are thought to be the surefire path to skillful execution. However, in his study where he compared expert and novice black smiths, Bernstein found something novel. As expected, those who were less skilled both struck the chisel less and had inconsistent hammer strikes paths. However, in defiance of expectation, those who struck the hammer precisely on the same spot still did not have identical swings of the hammer! Rather, their joint angles varied greatly swing to swing, but instead of being disorganized as the novices were, the experts leveraged their capacity for variability for greater accuracy and less variability in their ability to hit the target. Their bodies were organizing themselves through micro responses and adjustments to make sure the hammer hit the target. Here, the skilled movers were separated not by being able to repeat all elements of the task identically each time but by being able to self organize anew to accomplish the same outcome through immense variability. 

     The human body has many degrees of freedom through which it can accomplish a task, and it is through perception and response that the degrees of freedom are harnessed to elicit that response. We could not possibly control all of these degrees of freedom through top down conscious thought. Instead, those degrees of freedom, like the flock of birds who do not speak but move in unison, are conducted by the perception and response network of the whole body in chorus. We may intentionally limit DoF as coaches for learning matters, or the body may create clusters of or boundaries around DoF to accomplish tasks. 

In Latash’s article The Bliss of Motor Abundance the author describes their own version of Gray’s degeneracy and Bernstein’s variability through the concept of abundance where in the body has many many ways of accomplishing a task – this abundance of options affords high degrees of flexibility depending on the environment or constraints that may be present. 

As Bernstein put it, after witnessing the image of the expert blacksmith, skillful movers do something he called “repetition without repetition” a phrase that has become popularized in this area of interest.

The Expert Blacksmith of Bernstein's Study

Stochastic Resonance

In his book Antifragile, mathematical statistician Nassim Nicholas Taleb states that “Antifragility is beyond resilience or robustness. The resilient resists shocks and stays the same; the antifragile gets better.”

The abundance of different motor options creates a system that is not only robust to disruption but antifragile, if any given set of solutions is cut off it is no bother as the body has other avenues for achieving what it wants. If one set of options yields sub par results, it will preferentially cluster and leverage those which yield superior outcomes first in the future. 

According to Latash, having lots of variance that does not negatively affect performance but improves the emergence of solutions to a motor task. When the system is noisy or rich with workable solutions the phenomenon of stochastic resonance occurs. 

Stochastic resonance is a phenomenon where in if a signal is too weak to be clear, adding noise to the system acts as an amplifier and clarifies the weaker signal. Visualized, this looks something like this: 

Here, the addition of noise to the original imagine clarifies the archway;too much and the image is muddled, too little and the threshold of clarity is not satisfied. Quite similarly, noise during a sporting practice has a threshold. Imagine a sparring situation, inadequate feedback from a training partner and the fighter never receives adequately realistic inputs to hone their skills, however excessive attacks equally overwhelm and shut down learning. Therefore training exercises must be “noisy enough” but not “too noisy” to optimally perform. 

Drillers make killers, or do they?

Most “drills” in sports training are decontextualized to the extreme. Often, drills consist of an activity where the coach has inadvertently removed two critical sporting elements – perception and game type decision making. Sure, these drills may alone be skills which an athlete can display improvement on – but often times improving at the drill does not express what Yuri Verskoshansky described as “dynamic correspondence” where in a drill or exercise has desired “carry over” to a sporting task.  In other words the athlete may get better at dribbling the ball through the cones, but if this activity doesn’t improve their basketball performance (and it likely won’t) then it doesn’t satisfy the demands of dynamic correspondence. The gap between practice and play and improvement and non improvement but be addressed through coaching and superior methods of training, but what are our alternatives if most of our drilling tools fall into the category of “not noisy” enough to elicit learning or not realistic enough to elicit carryover. 

Consider another problem; the athlete who has learned all the steps by wrote rehearsal or by having a coach hammer verbal cues into them (step left, step right, stand tall, etc)  – this athlete may be able to express the skill in a sanitized gym session but will likely will demonstrate skill decay and default to un-refined and un-trained decision making and problem solving strategies under competitive stress – i.e. they choke. These athletes are intellectually knowledgeable and can often demonstrate impressive skills in the gym but display the classic folly of being unable to bring it on competition day. 

In light of all this information how should the coach facilitate skill development so that improvements are robust to competitive stress? 

In his Constraints based model, Newell describes three types of constraints:

Individual – these constraints are internal or embodied and can be separated into structural (the bones, muscles, etc) and functional (the mind)

Environmental – these constraints are external can be physical such as the height of an object or ideological such as social boundaries

Task – these constraints are external and are the constraints of the task itself such as the rule set of a game, the weight of the equipment, etc.

By manipulating and leveraging these constraints in different ways, the coach can facilitate development without incessant yelling or verbal cueing and replace those things with athlete exploration. 

A wonderful example of this is a style of coaching and improving footwork and distance management in striking that my martial arts coach uses; in order to improve our footwork we spar with a variety of  weapons and stances. These different tasks are similar but leverage different constraints (implement length, lead leg position, where hits are or are not permitted, etc). By gamifying our sparring we have more fun, get into a flow state, and focus on problem solving dynamically without chronic verbal input such as a repetitious “reach further” or “move your feet” – these things may occur organically but the coach doesn’t have to constantly yell it at the athlete to achieve the outcome. When we blend constraint manipulation we do less mental hoop jumping and more being present, having fun, and deeply learning than we would if we exclusively trained with a series of dissociated drills or solely relied on verbal cueing. 

Another option or alternative coaching method is the Differential Learning method. An example of how I have utilized this approach is with my client Donna who is a badminton player and who wanted to improve her serving skills. Instead of practicing every serve identically or perfectly we intentionally practiced over serving and underserving. An example of this in grappling practice is when my coach has me spin around and get dizzy before having to defend a series of attacks. By expanding the variables we expose ourselves to we increase the noise of the system, subvert attractors i.e. move the athlete away from the solutions they repeatedly gravitate towards in their typical environment or practice conditions. 

Indeed, drillers DO make killers but the type of drilling is important. Rather than decoupling the perception and action demands of the sport, the coach can better serve the athlete by simplifying demands but still requiring game-like perception and action demands. The athlete is also better served when they are given activities that permit them to map and understand their body, the task, and the space around them in a variety of ways such as over or undershooting a basketball to get better at making shots. 

A brief note on capacity building vs skill development

A strength coaches we often train divergent populations; those who participate in weightlifting sports such as powerlifting, strongman, and olympic lifting; those who participate in non-strength sports such as soccer, mma, tennis, etc.

For a weightlifting sporting athlete their strength development and physical conditioning overlaps with their sport conditioning but for the sporting athlete their strength and conditioning is aimed at improving the physiological qualities that support their sport. This addendum is not a treatise on how to train these indivudals but a reminder to the coach to not get too caught up in making their sporting athletes into powerlifters or strongmen but rather to keep the goal of improving their physiology to support their sporting skill development. It does your tennis all star no good to become excellent at power snatches if it does not help them express themselves in a superior fashion during a tennis match. 


Biryukova E, Sirotkina I. Forward to Bernstein: Movement Complexity as a New Frontier. Front Neurosci. 2020 Jun 3;14:553. doi: 10.3389/fnins.2020.00553. PMID: 32581691; PMCID: PMC7283918.

Chernavsky, A. V., & Talis, V. L. (2021). On the concept of activity in the last papers of Nikolai Bernstein. Human Movement Science, 80, 102886.

Gray, Rob. How We Learn to Move: A Revolution in the Way We Coach & Practice Sports Skills . Perception Action Consulting & Education LLC. Kindle Edition.. 

Moss, F. (2004). Stochastic resonance and Sensory Information Processing: A tutorial and review of application. Clinical Neurophysiology, 115(2), 267–281.

Prilutsky, B. I., & Zatsiorsky, V. M. (2020). Neural control principles. Bernstein’s Construction of Movements, 272–285.

Vázquez-Rodríguez, B., Avena-Koenigsberger, A., Sporns, O. et al. Stochastic resonance at criticality in a network model of the human cortex. Sci Rep 7, 13020 (2017).


Zatsiorsky VM, Gao F, Latash ML. Motor control goes beyond physics: differential effects of gravity and inertia on finger forces during manipulation of hand-held objects. Exp Brain Res. 2005 Apr;162(3):300-8. doi: 10.1007/s00221-004-2152-2. Epub 2004 Dec 4. PMID: 15580485; PMCID: PMC2845181.

%d bloggers like this: