What the data actually shows
Models of skill acquisition describe a movement from slow, deliberate, rule-following performance to fluent, largely automatic performance. Early on you are consciously thinking through every step; with practice the basics become automatic, which frees attention for higher-level aspects of the skill. This shift is part of why beginners improve visibly fast and then appear to stall — the easy automaticity gains come first.
Performance tends to follow a curve of diminishing returns: large early improvements, then progressively smaller ones for the same amount of practice. Plateaus — periods where measured performance flattens despite continued effort — are common and well documented. Often a plateau reflects consolidation, or the point where comfortable repetition stops producing gains and a harder kind of practice is needed to move forward.
Anders Ericsson's research on expertise argued that what distinguishes top performers is not just time spent but 'deliberate practice': focused, effortful work on specific weaknesses, with feedback, usually pushing just beyond current ability. This kind of practice is demanding and cannot be sustained for long stretches, which is part of why real improvement is slow — the effective kind of practice is precisely the kind that is hard to do for hours.
Why this feels different from how it actually is
It feels like it is taking too long because the early curve sets a false baseline. Your first weeks of obvious, satisfying progress create the expectation that improvement is roughly linear, so the long, flat stretch that follows feels like failure rather than the normal next stage.
Plateaus are also demoralising because effort and visible results come apart. You are still practising, perhaps harder than before, but the scoreboard isn't moving — sometimes called 'the dip.' The natural conclusion is that you have reached your ceiling, when more often you have reached the point where comfortable practice stops working and a different, harder approach is needed.
And we mostly see other people's finished competence, not the years of unglamorous practice behind it. The visible product of expertise hides the slow process that produced it, so your own slow progress feels uniquely sluggish by comparison with results you never saw being built.
The plateaus and the grind are not a malfunction; they are the normal middle of the process.
What the research says to do about it
Practice deliberately rather than just logging hours. The research suggests that targeted, effortful work on specific weaknesses — with feedback, at the edge of your current ability — drives improvement far more than comfortable repetition of things you can already do. This is more tiring per minute, which is why shorter, focused sessions often beat long, autopilot ones.
Expect plateaus and treat them as signals rather than verdicts. A flat stretch frequently means the current method has stopped challenging you and it is time to change the drill, raise the difficulty, or get outside feedback. Pushing through with the same comfortable practice is often what keeps the plateau in place.
Track inputs and small markers, not just outcomes, especially during slow stretches. Because results lag effort, the felt sense of progress can vanish even while you are improving. Concrete records of what you practised and small sub-skills that have become easier give a truer picture than the headline result, and help sustain effort through the dip.
What the research says does not help
Logging large amounts of comfortable, repetitive practice does surprisingly little once you are past the beginner stage. Mindlessly redoing what you can already do builds time-on-task but not skill; the research points to effortful practice at the edge of ability as the part that moves you forward.
Reading the 10,000-hour rule as a guarantee misleads more than it helps. The original research emphasised the quality and structure of practice, not a fixed number of hours, and the specific figure has been widely overstated. Hours alone, without deliberate, feedback-rich practice, do not reliably produce expertise.
Quitting at the first plateau, on the assumption that flat progress means you have hit your limit, cuts the process off exactly where the slow real gains tend to live. Most plateaus are a sign to change how you are practising, not proof that you cannot improve further.
The effective kind of practice is precisely the kind that is hard to do for hours — which is part of why real improvement is slow.
What this looks like in real life
The fast start that sets a false baseline
Your first weeks are full of obvious, satisfying progress as the basics become automatic. That early curve quietly teaches you to expect improvement to keep feeling that fast. When the long, flat stretch arrives, it reads as failure rather than the normal next stage — even though that flat-feeling stretch is usually where the slow, real improvement is happening.
Practising harder, scoreboard not moving
You're putting in the hours, maybe more than before, but measured performance has flattened — the dip. The natural conclusion is that you've reached your limit. More often you've reached the point where comfortable repetition has been exhausted, and a harder kind of practice is needed: raise the difficulty, target a specific weakness, or get outside feedback.
You only saw the finished version
You mostly encounter other people's finished competence, never the years of unglamorous practice behind it. With the slow process hidden, your own progress feels uniquely sluggish by comparison with results you never watched being built.
Real numbers in context
The honest shape to internalise is the curve, not a timeline. Skill tends to rise steeply at first and then flatten, so the same hour of practice buys a lot early and much less later. There is no universal number of hours to 'get good' — it varies enormously by domain, starting point, and the quality of practice — and the popular '10,000 hours' figure was an average from one set of studies that the original researchers say has been misread as a rule.
Plateaus are normal rather than exceptional. Across skills from music to sport to languages, periods of flat measured progress despite continued effort are a routine feature of the learning curve, often marking the point where comfortable practice has been exhausted. The useful reframe is that slow, plateaued stretches are usually the middle of getting good, not the end of it.