There is No Perfect Technique [How We Learn to Move, Chapter 1 Companion]
There is No Perfect Technique [How We Learn to Move, Chapter 1 Companion]
And there's no "correct" move, either--there's always multiple ways to get the same effect. Here's why we should think in terms of "functional" moves instead.
This is post #1 in the Combat Learning Companion series to Rob Gray’s How We Learn to Move.
It’s the air we breathe from our first day of school to our last day in the workforce and beyond…
Repetition, repetition, repetition. Practice makes perfect. Perfect practice makes better.
Everywhere we’re washed over with some form of messaging about how to learn that involves drudgery and drilling. But that repetition is not just to get it “into muscle memory.” The repetition is important because you need to get it right.
Martial artists are unilaterally obsessed with “perfecting technique.” Karate nerds and jiu jitsu potheads alike will pontificate on the exact minutiae of an expansive and never ending compendium of techniques.
“Technique conquers all,” they say.
Both traditionalists and Gracie jiu jitsu types speak often of ideal body mechanics when they talk about technique. At times, it sounds like a sort of Platonic form: a perfect abstract object that exists somewhere, rarely instantiated in the movements of the humans who seek to replicate it.
If you just do it properly—if you just do it textbook—then it will get you the outcome you desire and then some. We are told the most elite martial artists are capable of executing with perfect technique; and each situation has just a few, perhaps even just one, ideal technique.
…but is this really the case?
The Significance of Nikolai Bernstein’s Blacksmith
Because they’ve already decided what works best, martial arts coaches design training that revolves around a sort of rote repetition of those techniques in a sterile and compliant environment.
The purpose of these repetitions is to iron out variability in the visual quality of the movement pattern, because perfect or ideal technique looks the same every time it’s performed. This repeatability is a precondition of improved performance and future mastery.
Too much variance—perhaps any, with the case of kata—is bad. It means you’re unskilled with that technique and are less likely to pull it off in the performance environment (“the real thing”).
In 1922, a legendary Soviet scientist by the name of Nikolai Bernstein put this notion to the test.
The Hammering Experiment
Bernstein tasked novice and expert blacksmiths to hammer sheets of metal. They wore lights on their joints so that their trajectory paths could be tracked by the antiquated cameras of the time.
Dr. Gray writes (pgs. 4-5):
What did Bernstein find? […] The novice’s results seemed to confirm the conventional repetition story: they had highly variable outcomes (they did not hit the head of the chisel on the same spot) because they had highly variable movements (the path of the hammer followed on the way to the chisel was different on every strike). Novices were not as skillful because their movement was not repeatable. Case closed.
Except, not exactly. The expert’s results were unexpected (emphasis mine):
[T]he experienced blacksmith hit the same spot on the chisel but not by repeating the same movement every time. Bernstein coined this surprising result “repetition without repetition.” We repeat an action outcome but not by repeating the movement that produced it.
Bernstein’s experiment busted the most pervasive myth: that of the ideal, repeatable technique.
More importantly, he proved experts do not produce more consistent movement pattern quality than beginners. He proved, instead, that they produce a different kind of variability than beginners—functional.
Functional Variability is Expert Motor Control
According to Dr. Gray, this is how skilled movement—like blacksmith hammering—is produced in the traditional view (p. 7):
Plan the movement beforehand (pack a cannonball into a cannon).
Initiate it (light the wick).
Let it run (fire and wait until the cannonball makes contact with the target).
You’ll notice that by time #2 ends and #3 starts, there’s no way to change the movement or adjust during the movement. This is very different from the Ecological Dynamics view of skilled motor (movement) control, which Bernstein’s observations influenced.
After careful analysis of the elbow and shoulder joints relative to the hammer, Bernstein noticed that “the experienced blacksmith’s body movements were more variable than the movements of the hammer” (p. 7).
He understood this to mean that expert blacksmiths were in fact adjusting their hammer swings mid-swing, correcting any initial trajectory errors on the fly, in stark contrast to the old view. You can control a movement even after initiating it after all.
All that variability isn’t just a factoid or a tidbit of expert performance. It serves a very important function. And thus expert motor control is marked by a high degree of functional variability.
Old heads will no doubt recognize what might be going on here: a feature of self-organization. We’ll explore this more with Dr. Gray in later chapters.
Key Principle: Repetition Without Repetition
Hammering is a much simpler task than combat sports. With hammering, a coach could simply allow the hammerer to experiment with strokes until he is able to get consistent outcomes from inconsistent movements.
But with martial arts, there’s a dynamic opponent to deal with, more rules, and messier task goals. Therefore, some additional care is necessary to apply repetition without repetition (RWR) to martial arts training.
A three step process of my own making is outlined below.
Identify a movement problem
Translate problem into a task
Ensure sufficient time on, and repeat exposure to, the task
1. Identify a Movement Problem
For this, it’s useful to think in terms of what movement problem is presented by an activity.
According to The Constraints-Led Approach by Renshaw, Davids, and Newcombe, RWR is applied to more complex sports by creating environments in which players can solve the same sport-specific problems but in different ways (p. 81).
With martial arts, there are nearly infinite movement problems, but there are some major top-level situations that all learners will often find themselves navigating, regardless of the sport. When thinking of a problem, it's easiest to start with these common situations.
In sport taekwondo, for example, the cut kick is an ever-present problem for players who want to move into the proper range to score or score without obstruction. Players must circumvent the cut kick in order to score and avoid being scored upon.
(A similar problem exists in kickboxing/Muay Thai with the teep, but the hands potentially make the problem a bit more complex than the taekwondo version.)
There are many ways to deal with the cut kick: cancelling, footwork, counter kicking, blocking—usually in combination. With this in mind, we know we've got a candidate for a movement problem that is free enough to be translated into a good practice task.
2. Translate the Movement Problem into a Task
We have a problem: dealing with the cut kick. Now, how do we take this and build useful opportunities for repetition without repetition?
Let’s consider first the different ways you might deal with the cut kick in service of scoring or facilitating a better chance at scoring:
Establishing your own cut kick (usually involves a “cancel” as well)
Cancelling without a cut kick/cancelling to facilitate a score (e.g., cancel to “flop kick” or Brazilian kick)
Invasive footwork (less common; typically involves blocking and a quick follow-up kick)
Example of “cancels” are in the video embed below, for those not familiar with Olympic taekwondo terms.
With brand new students, more specific tasks are preferable to less specific task environments. So the different ways of getting around the cut kick are all useful landmarks to build games out of to encourage RWR.
Let’s say we want our athletes to explore the “establish your own cut” outcome as a potential (and broad-banner tactical) solution to the cut kick problem. We don’t want to only play this game, of course, but as part of a broader session it will be useful to beginners and in specific cases for others as well.
Therefore, we can create a game where the desired outcome of the player is to make strong contact with a cut kick as many times as possible during the game. You could also reset when a cut kick is successfully achieved, but this will likely limit the amount of “repetitions” of solving the “establish the cut kick” problem can be solved within a sub 5 minute gameplay window.
Why is this likely to encourage repetition without repetition?
Students have a specific and useful sport specific outcome to reach for (“establish the cut”/land a strong cut kick). However, even though the outcome is much more specific than, say, “get around the cut kick,” it still allows players to continuously organize their legs and arms in different ways in an attempt enact that outcome.
What we don’t want to do is stipulate exactly what manner of cancelling tactic (under, over, etc.) is necessary to accomplish the “establish the cut kick” outcome/task goal. This almost certainly precludes any meaningful level of RWR and, as we’ll discover in later chapters, it ruins the “representativeness” (i.e., the skill transferability) of the exercise design.
Below, you’ll find example of cancelling drills that seem like they might encourage RWR since each one is different. This is deceptive, however: there’s no opponent acting and reacting to the cut kick problem as she would in a match, no problem is actually solved and no outcome reached. It’s also likely that these drills, in addition to each being detailed prescriptions in their own right, are practiced in a predictable sequence. Neither allow for RWR.
To understand if a task is specific enough without being too restrictive, ask yourself these questions:
Are there still multiple ways to organize one’s limbs to complete this task/gain the designated outcome?
Are both players able to adapt to each other’s behaviors throughout the task and/or from trial-to-trial?
Are students merely cycling through a sequenced or randomized series of preplanned solutions?
So if…
a) they do not have multiple ways to organize to fulfill the task and/or
b) are not able to adapt to each other’s behaviors inside the game and/or between trials and/or
c) students are merely cycling through a sequence or randomized group of prescribed solutions…
then the task is very likely too restrictive to encourage repetition without repetition.
3. Ensure Time on Task and Repeat Task Exposure
This is more of a housekeeping item, but it’s still important. Learners need more than just a couple minutes to explore these environments, but they don’t need to be forced to engage with the same task for more than 10 minutes at a time.
This isn’t an exact science, but 5-10 minutes on important sport tasks is probably the sweet spot. You can toss in some other games to play and come back to it again in the same class, or you can program it to show up again in the next couple of weeks.
This leads me to my next point, which is probably more important: repetition is still critical in the constraints-led approach.
The difference is we don’t repeat solutions; but we do repeat our engagement with task environments. We can’t get good at managing them unless we’re in them long enough and often enough.
There's no good reason not to gain another exposure to a common movement problem for another 2, 3, 6 months.
Don’t get lost in the average
Talk about the “perfect” technique implicitly assumes that everyone’s body is the same. Some might retort that it’s based more on an average, but that’s even worse because it’s a consideration that explicitly excludes your specific body.
Of course, we know everyone has different bodies. Long-legged players find triangles far easier than short-legged players, who have to engage in increasingly minute adjustments to even lock a triangle or finish it without exploding their knees.
Instructors will often justify their preferred set of special details about finishing a given submission hold by saying, “this is the version that works for everyone.” It works for the most people. In a sense, it’s an averaged technique.
There isn’t just this singular way to finish a triangle choke, though. You don’t have to cut a perfect angle and get all your ducks in a pristine row, provided your legs are long enough relative to your opponent. If they aren’t, then you have to scale to that situation. But if you’re unusually tall, it might never matter, even at higher levels of competition.
And you know what? Let’s get really spicy.
Why do you even need to master a leg triangle at all? It seems plenty of jiu jitsu players get along fine without it. Throughout Marcelo Garcia’s illustrious fight record, BJJ Heroes only records one win by triangle.
The point here is that technical averages sound enticing but they are meaningless. They don’t account for your body. Every individual elite player in any sport moves differently than the other while solving the same problems.
Is there a way in which we can still think of a technique as “perfect?”
My immediate response is this: perfect how and for what?
(We shouldn’t think in terms of techniques at all, but that’s another discussion.)
At worst, "perfect" is marketing language. It's romantic talk. At best, it's philosophical language; but in that context, it most likely means maximal. As in: maxed out. It can’t get better.
Let’s assume the best about our friends on the other side of the debate. They aren’t trying to sell something (they usually are, but we’re pretending right now).
Instead, they mean that “perfect” corresponds to some maximal optimization to the movement mechanics. Say that we’re talking about the finishing mechanics of a guillotine choke. Now let’s say that, to them, the “perfect” guillotine choke hinges on one or two key details…
Readers, how could a single detail be maximal across all contexts?
It cannot. There are different neck sizes, different defenses, different arm sizes (of the choker), different shoulder structures. Things vary so much body to body and roll to roll in ways that meaningfully change the ease or difficulty of finishing a guillotine choke. Because they change, they require online—instant, in-the-moment—adaptation.
Traditional martial artists tend to classify the same punch mechanics aimed at different targets as separate techniques in one sense. Boxers differentiate between long hooks, shovel hooks, and uppercuts, all of which are gradients between similar movements controlled by range, angle, positioning, and available target surface.
It’s impossible to dictate the perfect range and the perfect target for a strike at all times. Consequently, it’s also impossible to dictate the perfect technique biomechanics, perfect target area for that technique, and perfect distance to throw it.
Instead, you take what you can get. You throw the most useful strikes you can within the chaos, from wherever you happen to find yourself, to achieve your goal of scoring and/or knockout.
That’s why we should think in terms of functional movement skills. Movement solutions that work: they accomplish the goal, regardless of the particulars.
Yes, these solutions can be optimized…but fulfilling the task goal is the most important thing.
Key Takeaways
Expert movers exhibit a lot of variability in the way they execute movement patterns, but the outcome is not variable. In other words, they’re consistently successful at achieving their tasks without repeating the same exact movement over and over again. The idea that skillful technique must be repeatable, with minimal or no variability in execution, is debunked.
Expertise is marked by the above defined “functional variability.” Therefore, there is no single ideal or correct technique to accomplish a movement-based task goal.
Coaches should understand Bernstein’s learning principle of repetition without repetition (RWR). Coaches can help create expert movers by encouraging learners to solve the same sport-specific problems in different ways and providing ample opportunities to do so. RWR is one of the four main practice design principles of the constraints-led approach to coaching.
Part 2 of this Combat Learning Companion series, “Errors are Necessary for Perception AND Skill,” is available now.
Amazon book listings are affiliate links. Purchasing a book through one of these links supports Combat Learning at no additional cost to you. Thank you!