- RNP
- Motor Learning
- Error, uncertainty, variability
- Prediction–reafference mismatch: why your system only learns when its forecast is wrong
SeriesError, uncertainty, variability7 of 13
Prediction–reafference mismatch: why your system only learns when its forecast is wrong
The prediction–reafference mismatch is the gap between what your system predicted and the sensory feedback it actually gets. That gap is what triggers motor learning.
In short. The prediction–reafference mismatch is the functional difference between what your motor system predicts and the sensory feedback it actually receives. When that gap appears, learning kicks in. Without it, nothing changes.
The exact moment it kicks in
Watch an athlete launch into a jump they thought they had nailed. Before they even leave the ground, their system has already predicted what they're going to feel: the pressure underfoot, the tension through the posterior chain, the moment of landing. Then the feedback arrives. And it doesn't match. The landing is harder, shorter, a beat off from what was expected.
That moment has a name. It's the prediction–reafference mismatch. It's the exact second your system discovers its forecast was wrong. And that is precisely where, and nowhere else, motor learning begins.
Reafference is the sensory flow that comes back during and after the movement (see reafference for details on sensory feedback). The prediction is what the system expected to receive. The mismatch is the distance between the two.
Not a flaw, a signal
This gap is often mistaken for a mistake to be erased. It's the opposite. The mismatch is neither good nor bad: it's neutral, functional, necessary. It belongs to the same family as error, understood as a gap between prediction and reafference, not as a failing.
As long as the system's prediction matches the sensory feedback, it has absolutely no reason to reorganize. It stays with its current solution, optimizes it, automates it. But it learns nothing new.
It's the gap that forces the update. When sensory information contradicts the forecast, the system has no choice but to look for a new solution. The mismatch is the trigger. Not a symptom to suppress, but the fuel for recalibration. This mechanism cuts across every task, and it's non-negotiable.
The RNP training space, free for life
Get access to RNP training, no strings attached.
Too little mismatch: inertia
Tasks that are too easy, feedback that's too frequent, constant guidance, an environment that's too stable. The outcome fits in one sentence: high performance, zero learning.
The system gets no surprises. No sensory contradiction comes along to shake up its prediction. It sits comfortably in its stable attractor, optimizes locally, and performs the same drill beautifully for months. And it transfers nothing.
You know the scene. A session that's too clean, where everything lands, where the coach's eye never leaves the athlete. It's reassuring. But a session with no prediction–reafference gap is a session in which the system had no reason to learn. This absence of a gap is a close cousin of the absence of uncertainty: no surprise, no exploration.
Too much mismatch: panic
At the other end, excess. A task that's too hard, stacked constraints, sensory overload, emotional pressure. The gap between what the system predicts and what it perceives becomes so wide that it can't do anything with it.
The result is freezing. Crude compensation. Disengagement. The system is overwhelmed, has no solution available, panics and regresses. It tries to force its way back to a known solution, even a dysfunctional one, just to survive. You can see it in the athlete's body: shoulders creeping up, breath shortening, gaze locking. There's no learning here either.
A mismatch that's too large doesn't trigger search, it triggers protection.
The optimal zone: dose the gap, don't remove it
Between inertia and panic lies a critical zone. Roughly 70 to 80% success. Errors that are frequent but tolerable. Manageable uncertainty. Sustained engagement.
That's where the mismatch is just big enough to force reorganization without breaking the system. That's where plasticity peaks. It's the target of any serious learning design.
So your job isn't to eliminate the gap. It's to calibrate it. To keep it in that window where the system is wrong enough to search, but not so wrong that it collapses. Too easy, you're wasting your time. Too hard, you break the machine. In the zone, you create the conditions for things to reorganize on their own. It's exactly the same logic that separates structured motor variability from chaos: constrain differently to provoke recalibration, without drowning the information.
That's also why immediate corrective feedback is a false friend: by removing the visible gap too quickly, it deprives the system of the sensory contradiction it needs. The distinction between internal and external feedback plays out right here.
Mismatch and self-organization
When the mismatch persists within the right window, the system doesn't correct a shape. It shifts a solution. The old attractor becomes costly, a new one becomes more stable, and the system slides into it by itself. That's self-organization: you don't force the movement, you change the constraint landscape so the prediction–reafference gap does the work for you.
In the RNP framework, this gap is exactly what the sensorimotor loop reads and recalibrates: predict, perceive, compare, update.
No error, no mismatch, no learning. Motor learning isn't something to protect from error; it's something to organize around it.
In one sentence: the prediction–reafference mismatch is the gap between the system's forecast and its actual sensory feedback, and it's this gap, dosed within the right window, that triggers learning.
Field test. In your next session, spot the moment the athlete is surprised by what they feel: a harder contact, an off-beat tempo, an unexpected load. If that moment never comes, your task is too easy or too guided. If it overwhelms them, it's too hard. Aim for the window where surprise exists without spilling over.
Scope: a functional, non-medical approach. It complements healthcare follow-up; it does not replace it.