The Real Reason You Plateau (It's Usually a Tracking Problem)

"I've hit a plateau" is one of the most common sentences in music, and one of the least examined. We say it the way we'd report the weather — a flat, settled fact about the world. But listen to what's actually buried inside it: a claim that nothing is improving. And that is not a feeling. It's a measurement.

Here's the uncomfortable part. Most musicians take that measurement with the bluntest instrument they own — a whole-piece, all-at-once gut check: can I play it yet? does it sound good? That instrument is far too coarse to detect the gains that actually accumulate from one week to the next, which means it returns "no change" even when plenty is changing. Swap it for a sharper one and a surprising share of plateaus simply evaporate, because they were never stalls. They were progress you couldn't see.

"Plateau" is a measurement word

Think about what your usual progress gauge can and can't pick up. Sitting down, playing the piece, and asking "does this feel better than last week?" is a single, low-resolution reading taken at the level of the whole piece. It cannot register that the hard passage in bar 34 went from 88 to 96 on the metronome, or that you stumbled twice in your run-through instead of five times, or that your left hand finally stopped tensing up in the leap. Those are real improvements. They're just smaller than the smallest unit your gauge can read, so they round down to "same."

This is why a plateau feels so airtight from the inside. Week after week the needle reads "no change," and the obvious conclusion is that you've stopped getting better. But the needle reading "no change" and you actually not changing are two different things, and you've quietly assumed they're the same. The feeling of being stuck is real and worth taking seriously. The conclusion you draw from it — that nothing is happening — is usually a resolution error, not a fact about your playing. We made a related point about effort that looks like work but isn't in why your practice feels productive but isn't; this is the same gap viewed from the other side, where the work is real but the progress reads as zero.

"‘Nothing is improving’ is not an observation. It's a measurement — and most of us take it with the bluntest instrument we own."

Most plateaus are invisible progress, not stalls

The reason this matters so much is that the science of skill learning predicts long stretches where real improvement is genuinely invisible. Learning curves are not straight lines. Across a huge range of skills they show the same shape: a fast early climb, then a long, gradually flattening stretch where surface performance barely moves — the shape Fitts and Posner described decades ago and that countless studies have confirmed since. The flat part isn't the end of learning. It's a phase of it.

What's happening underneath is the interesting bit. During those flat stretches the brain is consolidating — automating the lower-level pieces of the skill and restructuring them into higher-order patterns — and that work happens below the level of conscious feedback. Researchers studying motor learning find that the relevant neural circuits keep remodelling well after measured performance appears to stop improving, which means the absence of a visible gain is not the absence of change. Progress on hard skills tends to arrive in bursts after exactly these quiet plateaus, as the consolidated sub-skills suddenly express themselves. You don't climb a smooth ramp; you climb a staircase, and the plateau is the flat part of a step.

The cruel design flaw is that this productive, invisible-progress phase feels identical to a genuine dead stop. Same flat needle, same frustration, same creeping suspicion that you've peaked. So this is precisely the moment musicians quit a piece, abandon a routine, or start thrashing — switching methods every few days in a panic — often right before the gain they'd already earned was about to surface. The plateau didn't beat them. Misreading it did.

But some plateaus are real

None of this means every plateau is a comforting illusion. Some are genuine, and the most common real one has a name: the "OK plateau," a term Joshua Foer drew from Anders Ericsson's research on expertise. It works like this. When you first learn something you're in a highly conscious, effortful mode — Fitts and Posner's cognitive and associative stages. With enough repetition the skill becomes automatic and you slip into the autonomous stage, where you can run the piece without really thinking about it. That automation is wonderful for things you've decided are "good enough," and it is poison for things you still want to improve, because a skill on autopilot stops developing. You're no longer practising; you're just repeating.

The tell for a real plateau is behavioural, not emotional. You're playing pieces through rather than attacking the parts that are hard. You've stopped setting the metronome higher. You reach for the sections that feel good and skim the ones that don't. Ericsson's finding is that the way back out is to break the autopilot deliberately — keep conscious control of the thing, raise the difficulty, and, in his phrase, practise failing rather than practising what you've already mastered. The fix for a real plateau is a change in how you practise, not simply more of the same hours.

The instrument that tells them apart

Now the two threads meet, and the practical conclusion falls out. By feel, the false plateau and the real one are indistinguishable — both are a flat needle and a frustrated player. You cannot tell from the inside which one you're in, which is exactly why "I've plateaued" is such an unreliable self-diagnosis. The only thing that separates them is a measurement taken at a finer grain than your gut check can manage. Get that, and the ambiguity collapses into a clear signal.

So pick a metric at the level where progress actually happens — not the whole piece, but the part of it that's hard. The tempo you can play bars 34 to 41 cleanly. The number of stumble-free run-throughs out of five. How many error spots are left in the worst passage. How many days it's been since a particular section last got better. These are small, boring, specific numbers, and that's the point: they move on a timescale your "does it feel better?" check can't perceive. Track one for two weeks and the plateau tells you what it actually is.

This is the unglamorous case for tracking your practice at the level of passages, not just sessions — and it's exactly the gap a tool like our own app, Practis, is built to close, so weigh it accordingly. The point isn't to gamify anything or to bury you in charts. It's to keep a record fine-grained enough that a 6 bpm gain or two fewer error spots actually shows up, so the false plateau outs itself as steady progress and the real one outs itself as a flat line you can act on. If you want the broader mechanics of watching progress accumulate over time, we get into them in the best way to track practice progress over time — the short version is that the right grain of measurement is what turns a vague "am I getting better?" into an answer.

"A real plateau tells you to change the practice. A false one only tells you your ruler is too short — and most are the second kind."

What to actually do this week

Before you conclude anything, run the two-week test. Pick the one piece that feels most stuck, choose a single fine-grained number from the list above, and write it down each session — nothing more elaborate than that. Resist every urge to declare the plateau real until you have two weeks of readings, because the whole problem is that your felt sense is the unreliable instrument here. Let the number, not the mood, deliver the verdict.

If after two weeks the number is climbing, you have your answer: keep going, you were never stuck. If it's genuinely flat, you also have your answer, and now it's a useful one — the plateau is real, and the move is to change how you practise rather than to grind more hours into the same groove. For the deeper toolkit on breaking a genuine stall, what to do when you hit a plateau and nothing seems to improve picks up exactly where this leaves off.

Common questions

"Plateau" is a verdict most musicians reach far too early, on far too little evidence — a strong feeling dressed up as a fact. Before you accept it, check whether you've simply outgrown the instrument you're measuring with. More often than not, you haven't stopped improving at all. You've just stopped being able to feel it, which is a very different problem, and a much easier one to fix.