Sixty seconds on one foot, eyes closed, on a foam cushion. The pelvis stays stable, the ankle works finely, the athlete could hold for another minute. Ten minutes later, during a change of direction with a ball and an opponent in contact, the same support slips and the same athlete falls to the ground. Same ankle, same person, same session.
The RNP training space, free for life
Get access to RNP training, no strings attached.
Hello to you, the movement professional,
The objection comes quickly on this subject, and it is healthy: saying that holding balance work does not prepare for the field is the kind of statement mainly used to sell something else. The field is full of people who disqualify their neighbor's tool to promote their own. So let's set boundaries right away. What follows does not say that holding a posture is useless, and it is not based on any promise of quantified performance. What follows separates two tasks that the nervous system handles differently and looks at what can be measured when trying to pass one off as the other.
It remains to be seen what holding a posture really demands, and especially what it never demands.
Intuition ranks these two things on the same scale of difficulty. Standing on two feet, then on one foot, then on one foot with eyes closed, then on one foot with eyes closed on an unstable surface. A clean, readable progression, easy to note on a session sheet. The problem is that the last step of this scale leads to nothing resembling the field.
Postural control in a held posture solves a specific problem: keeping the projection of the center of mass inside the base of support while the body constantly oscillates. It's a maintenance task, where any deviation from the frame is an error to correct. The system works in a closed loop, it detects a deviation, corrects it, and repeats it several dozen times per minute. What limits performance is the precision of the correction.
A change of direction poses the opposite problem. The athlete deliberately moves their center of mass out of the base of support, lets it fall outward, and places a support further to catch it. The deviation from the frame is no longer an error; it is the gesture itself. What limits performance is the accuracy of placing the next support, decided before the imbalance is felt.
One task rewards correction, the other rewards anticipation. The nervous system does not solve them with the same circuit, and there is no mechanical reason why progressing on the first would make progress on the second.
What happens before the movement is the most counter-intuitive part of the file, and it is the one that decides the rest.
When a standing person quickly raises an arm, the activity of the trunk and supporting leg muscles changes before the activation of the deltoid. Not after, before. These anticipatory postural adjustments precede the voluntary movement by a few dozen milliseconds, and they are proportionate to the disturbance this movement will produce. The system does not undergo imbalance to correct it afterward; it prepares for it because it already knows what it is commanding.
This knowledge has a known support. The efference copy, this duplicate of the motor command sent in parallel with the movement, allows the system to predict the state the body will be in and prepare the corresponding postural response. This mechanism is why we do not unbalance ourselves by raising an arm, whereas a push of the same intensity from the outside does cause us to sway.
On the field, this anticipation is linked to something other than the athlete's own movement. It is linked to reading the situation, the trajectory of the ball, the orientation of the opponent's hips, the moment when the support will have to leave. The motor decision and the accompanying postural adjustment are prepared together, in the same time window.
A held posture contains none of these ingredients. There is no large voluntary command to anticipate, no trajectory to read, no moment to choose. The athlete trains their correction loop while the field will demand their prediction loop.
Looking at what the static task removes, the list becomes quite brutal.
It removes speed. Corrections in a held posture occur over comfortable durations, whereas a change of direction support is played in a short contact window, where there is no time to feel and then correct.
It removes head movement. In held balance, the gaze fixes on a point, and the head locks, which is the most economical strategy to succeed in the task. On the field, the head turns, tilts, absorbs contacts, while the gaze must remain fixed on a moving target. The semicircular canals then constantly measure rotational accelerations, and the vestibulo-ocular reflex works at full capacity to stabilize the image. A held posture with a fixed head leaves this part of the vestibular system at rest.
It removes the change of context. Conditions remain constant from start to finish of the series, so the weighting between inputs settles and no longer needs to move. The sensory conflict and its resolution, which are the real arbitration work in play, do not occur.
It removes the decision-making load. The athlete knows in advance what they must do, for how long, and on which foot. Nothing is to be chosen.
What remains after these subtractions is a slow maintenance task, with an immobile head, in a stable environment and without decision. The body is in the dark on many things permanently, and this task is precisely the one that lights up the fewest lamps at once.
The point is verifiable, and it has been verified several times.
When people are trained on a given balance task and then tested on other balance tasks, the gain remains largely confined to the trained task. A study published on this question has a title that summarizes the conclusion: "Task-specificity of balance training". The progress on the training device is clear, and the transfer to neighboring balance tasks remains weak.
The correlations between measurements go in the same direction. Scores obtained on a postural oscillation platform correlate poorly with scores obtained on moving balance tests and poorly with actual sports performances. The term balance covers a family of tasks that do not predict each other, which explains why one can be excellent at one and ordinary at others.
A reading trap lurks just after. Athletes who stand best on one foot with eyes closed are often also those who change direction most cleanly, and this observation is real. It still does not say that the first produces the second. A sports practice rich in varied supports builds both skills in parallel, which is more than enough to create the correlation without either causing the other. Sorting athletes on a static test remains possible; building them through this test is another assertion.
Task specificity is not a methodological precaution; it is the result.
Excluding transfer to performance does not mean discarding the tool, and the domain where it holds deserves to be named precisely.
Balance work on reduced support has data on its side regarding the recurrence of ankle injuries in people with a history. The effect is documented and modest; it concerns a recurrence risk and not a quantified performance, and it is part of a broader support. LabO RNP is neither medical nor paramedical; it is an addition to a care pathway and never a replacement.
The tool also retains entry value. For a person with very poor ankle and hip availability, a slow and simple task gives time to feel, and this time is useful before loading speed. The sensor foot then provides exploitable information, provided the task remains readable.
Its most solid function remains reading. Comparing a person to themselves, from week to week, on a standardized and slow task, gives a clean signal because precisely nothing else moves. This stability that makes the task poor in training makes it valuable in evaluation. Testing and building are two uses of the same tool, and confusing the two is the most common slip.
Reading through archaic reflexes adds a piece that the held posture does not show because it precisely immobilizes the concerned region.
The tonic labyrinthine reflex, the TLR, is triggered by the position of the head relative to gravity and redistributes tone in flexion or extension. The asymmetric tonic neck reflex, the ATNR, and the symmetric tonic neck reflex, the STNR, respond to the position of the head on the trunk, thus to deep cervical receptors, knowing that the same rotation simultaneously solicits the labyrinth. The door remains open to both inputs.
What these responses make visible is that a moving head modifies the tonic availability of the limbs. A held posture with a fixed head shows none of these influences, which gives a clean and incomplete image of what the athlete has available when playing.
A methodological reminder avoids the most common mistake here. The meaning of a reflex is its triggering stimulus, not the set of systems activated when it occurs. The Galant is the clearest example, triggered by a simple tactile stroke along the lumbar spine while its response is postural and massive. Classifying a reflex based on what activates simultaneously remains the quickest way to make the wrong intervention.
The practical consequence fits in one sentence: keep the constraint, change the task that carries it.
The constraints approach provides the direct lever. Rather than instructing a posture, you create a situation where the support becomes difficult to place, and you let the system find its solution. A support to place in a reduced area during a run, a change of direction triggered by a late visual signal, a light contact imposed at the moment of passing on one foot. In each of these cases, the balance requirement is real, and it arrives in a task that resembles the game.
Representative design poses the control question: is the information triggering the gesture in training the same as that which will trigger it in a match? An athlete who knows in advance on which foot they will turn is not training for the same problem as one who must read it on their opponent.
Attentional focus changes the result without changing the exercise. An instruction focused on the effect of the movement outside the body produces less postural oscillation and a more automatic gesture than an instruction focused on the segment itself. On a balance task, saying to keep the gaze on a moving target and push the ground gives a different organization than saying to contract the ankle.
The dosage is adjusted to speed and unpredictability, never to the apparent instability of the support. Adding foam under the feet of an already slow exercise makes it more spectacular without bringing it closer to the field by a centimeter.
Three registers stand out here, and confusing them is what makes the subject unreadable.
Anticipatory postural adjustments that precede voluntary movement, task specificity of balance gains, low correlation between postural oscillation measurements and moving balance tests, and the effect of an external focus instruction on oscillation are established facts, derived from reproduced protocols.
The reading that consists of classifying a training situation according to what it demands from the system, correction or anticipation, is a design framework consistent with these data. It remains a working grid, not an experimental result.
The third register is the one most sold. The idea that a balance constraint reintroduced into a representative task improves a quantified match performance, a percentage of duels won, or a change of direction time, falls under field reading. Transfer data are thin and heterogeneous, and no one can honestly promise you that.
A falsifiable benchmark, to remain tenable. If task specificity is indeed the commanding variable, then two matched groups in volume, one working on held postures on reduced support, the other on the same balance requirement within a representative task with head movement and unpredictable triggering, should separate on a field measure and not on the platform. If both progress equally on the representative task, the argument falls, and the held posture was sufficient.
An athlete does not lose their support because they balance poorly; they lose it because they placed it too late.
👉 I want to know what my training situations really demand from my athletes' system: labo-rnp.com/fr/pros
Static balance involves keeping the center of mass over a non-moving base of support, correcting oscillations. Dynamic balance involves deliberately moving the center of mass out of this base and catching it with a support placed further away. The former rewards correction, the latter anticipation.
The gains obtained from a balance task remain largely confined to that task. Transfer to field performance is not demonstrated, and correlations between postural sway scores and sports performance are weak.
There is data supporting its effect on the recurrence of ankle injuries in people with a history, with a modest effect. Its strongest value remains in evaluation, comparing a person to themselves on a standardized task. Our complete position on instability training, on what it costs in force production and attention, is developed separately.
Because the head carries the labyrinth, and its position relative to gravity and the trunk redistributes limb tone through tonic reflexes RTL, RTAC, and RTSC. A posture held with a fixed head leaves this dimension out of the work.
By maintaining the balance constraint but placing it in a task that resembles play: support to be placed under time constraint, triggered by a late signal, head movement under target requirement, imposed contact. The dosage is adjusted based on speed and unpredictability rather than the instability of the support.
By the LabO RNP team
Discover how the vestibulo-ocular reflex maintains gaze stability in motion. An essential for movement professionals.
Read the article Sensory Sensors - VestibularLearn about the latency of the vestibulo-ocular reflex (VOR) and its crucial role in compensating for head movements. Stay informed about the standards and implications of abnormal latency!
Read the article Sensory Sensors - VestibularDiscover how to test the vestibular function! Combine clinical and instrumental tests for accurate assessment. Learn about the methods and their interpretation.
Read the articleThe RNP system is meant to be read, targeted and tested from your next assessment. Start with the member space, for free.