Why Your Practice Time Is Lying to You (The Distribution Problem)
You practised the piece for 45 minutes. The timer agrees, and so does your memory. But "45 minutes on the piece" hides the only thing that matters — where in the piece those minutes actually went. And they almost always went to the wrong place.
Pract.is Editorial
Research-based practice guidance for musicians from the Pract.is editorial team.
You sat down and practised the piece for forty-five minutes. The timer says so. Your memory agrees — you were there, you did the work. But "forty-five minutes on the piece" is one of the most quietly misleading numbers in all of music, because it aggregates away the single thing that decided whether those minutes helped: where in the piece they went. And for almost everyone, almost always, they went to the wrong place.
This is the distribution problem. Total practice time is a lie of aggregation — it sums up minutes that were not equal and hands you a tidy figure that feels like progress. Meanwhile the actual variable that matters, the way your attention was distributed across the bars of the piece, stays invisible to the timer and, worse, to your own recollection. Here's why your minutes cluster where they do, why you can't see it happening, and how to fix a problem you can't currently even observe.
The minutes were never equal
Start with the finding that should end the "how long did you practise?" conversation for good. Robert Duke, Amy Simmons, and Carla Davis Cash had seventeen advanced pianists learn a genuinely difficult three-bar passage and tested their retention the next day. When they looked for what predicted who played it well, they found no significant relationship between retention and practice time, total repetitions, or complete run-throughs. None. What separated the best players was entirely how they spent the time — they targeted the errors, corrected them immediately, and returned to the exact trouble spot until it held. The title of their paper is the whole lesson: it's not how much, it's how.
Sit with what that means for a timer. If time practised has no reliable relationship to what you retain, then a number measuring time practised is measuring the wrong quantity. It's not a slightly-imperfect proxy for improvement; it's answering a different question than the one you care about. The currency of getting better isn't minutes. It's targeted attention on the specific things that don't work yet — and that currency is spent, or wasted, in the distribution.
"You didn't practise the piece for 45 minutes. You practised the parts you can already play for 40 — and the part that beats you for 5."
Where your minutes actually go
Left to its own devices, practice flows downhill, and the downhill direction is always toward the parts that already work. Three habits push it there. You start from the top, so the opening gets replayed on every single restart until it's the most-rehearsed music in the piece. You gravitate to the sections that sound good, because playing them is rewarding and playing the broken bit is not. And you skim or silently skip the hard passage — a bar of it, a wince, and back to something that feels like music. The end of the piece, meanwhile, barely gets visited at all, because you rarely make it that far before restarting.
Add it up across a real session and the shape is stark: the comfortable seventy per cent of the piece absorbs the overwhelming majority of the minutes, while the hard ten per cent — the ten per cent that actually determines whether the piece works — gets a rounding error. This isn't laziness or weak character. It's gravity, the same pull toward the rewarding and away from the frustrating that shows up whenever anyone practises anything unsupervised. Here is what that distribution typically looks like laid over a single piece.
Where 45 minutes on one piece actually went
Same 45 minutes. Nearly all of it on what already works — a rounding error on the passage that needs it most, and the ending you never reach.
Why you can't see it happening
The maldistribution would be easy to fix if you could see it, and the cruel part is that you can't — because both instruments you'd use to check are blind to it. The timer aggregates by design: it gives you one number for the whole piece and has no idea the number is lumpy. And your own memory is worse than the timer, not better. Attention isn't stored as a tidy map of where it went; the sections you replayed most feel like "the practice" and dominate the recollection, so you systematically overestimate how much you covered the hard bits. You walk away genuinely believing you worked on the difficult passage, because you touched it, and touching it is indistinguishable in memory from working on it.
So the piece gets reported as "practised" by two witnesses who both only saw the aggregate. Nobody — not the clock, not you — recorded the one fact that mattered: the passage that decides the piece got ninety seconds. This is the same reason effort can feel like progress when it isn't, a cousin problem we pull apart in why your practice feels productive but isn't — except here the illusion isn't about effort quality, it's specifically about location: where on the page the time landed.
Fixing a problem you can finally see
Every real fix for the distribution problem is the same move: force or measure the distribution instead of leaving it to gravity. Four of them, roughly in order of impact.
Done consistently, these invert the whole profile. A well-distributed session spends its first and freshest minutes on the bars that don't work, gives the transitions in and out of them a few careful passes, and treats the comfortable stretches as maintenance rather than the main event — a quick confirmation, not a loop. The graph flips: the tall bars move over the hard part, and the easy opening gets whatever's left. It feels worse in the moment, because you spend the session inside the part that fights you instead of the part that flatters you. It is, reliably, the session that actually moves the piece forward.
That last fix is the reason we built practice tracking that lives on the score itself rather than on a stopwatch. As you work, Practis score practice builds a heatmap right on the page: the sections you pour time into glow warm, the passages you keep avoiding stay cold, and the bars you've marked as trouble spots light up so the neglected-but-critical passage is impossible to overlook. It doesn't nag or judge — it just makes the distribution visible, which turns out to be most of the fix, because almost nobody knowingly avoids the hard passage once they can see, in colour, that they've been doing exactly that. (For the wider question of which practice numbers are worth watching at all, we sort signal from vanity in the practice data worth collecting.)
"The question was never how long you practised. It was where the practice went — and until you can see that on the page, the timer will keep flattering you."
Common questions
Isn't playing through the whole piece important too?
Yes — but as a small, deliberate slice, and for a different purpose: continuity, stamina, hearing the shape of the whole. A run-through is a test, not the study. The trouble starts when the test quietly becomes the practice, and the daily run-through — which by definition spends the least time on the hardest bars — turns into the main event. Run it occasionally to check; fix it in isolation the rest of the time.
How do I know which passage is "the hard one"?
Usually it's the one you just skimmed. Avoidance is the tell — the bar your hands hurry past, the transition you brace for, the spot where you'd rather go back to the top than continue. If it's genuinely unclear, one slow, honest pass with a pencil in hand exposes it fast: wherever you slow down, stumble, or wince is where the minutes should have been going.
Does this apply to memorised or non-classical music?
Completely — the distribution problem is about attention, not about sheet music. A memorised sonata, a jazz standard, a song you're arranging: the same gravity pulls you toward the sections that already groove and away from the shaky bridge or the awkward key change. Any piece you can divide into parts has a distribution, and left to itself, yours is almost certainly lumpy in the usual direction.
The timer isn't wrong about the forty-five minutes; it's just answering a question you never needed the answer to. The one you needed — where did the practice actually go? — it can't answer, and neither can your memory, which is why the hard passage can stay exactly as hard as it was through weeks of honest, well-intentioned work. Stop asking your practice how long it was. Start asking it where it went. The moment you can see the distribution, the piece starts getting better in the one place it never was.
More on practice tips.
Practice Tracking for Returning Musicians: Restarting Without the Guilt
The comeback rarely fails on the music — your hands recover faster than you think. It fails on the comparison: present-you measured against the best you ever were. Change the yardstick, and the guilt that makes returners quit again simply has nowhere to stand.
Pract.is Editorial
How to Use Practice Notes So They Actually Help Next Session
Most musicians don't have a note-writing problem. They have a note-reading problem — they journal diligently and never open it again. The fix is to write each note as an order to your next self, and to build the one trigger that makes you read it.
Pract.is Editorial
What a Good Week of Practice Actually Looks Like (With Real Numbers)
The imaginary good week is seven identical hours at the instrument. We pulled the real practice logs of Practis users to see what the good weeks actually look like — and they are shorter, lumpier, and far more forgiving than the fantasy.
Pract.is Editorial