How to practice piano effectively: a session that actually moves the needle

There is a kind of practicing that feels productive and changes nothing. You sit down, play the piece from the top, stumble in the same two bars you stumbled in yesterday, play to the end, feel vaguely unsatisfied, and do it again. Forty minutes later you have practised the easy parts forty times and the hard part twice, badly. Most of us learned to practise this way because nobody told us there was another way, and because playing through feels like music while the other way feels like work.

The other way is not complicated. It is a handful of habits that anyone can adopt in a week. What follows is the version I wish someone had handed me at fourteen.

Decide what the session is for

Before you play a note, name the thing you are trying to change. Not "practise the Chopin" — that is a location, not a goal. Something like: the left-hand jump in bars 9–12 at 80, or the second page from memory, or scales in E major, hands together, even. If you cannot say what would be different at the end of the session, you will not notice whether it happened.

This is why Chopinly's clock will not start until you tell it what you are working on. It is a small friction with a large effect: you have to say the goal out loud, and the minutes that follow belong to that goal and no other.

Two or three goals per session is plenty. A session with eight goals is a session with none.

Work in small units

The unit of practice is not the piece. It is not even the page. It is the smallest chunk that contains the problem: usually two to four bars, sometimes a single beat. Pick the unit, play it, and if it goes wrong, make the unit smaller until it goes right.

There is a piece of folk wisdom that says to practise the hard parts first. It is true, but incomplete. The real rule is: practise the boundaries. Difficult passages are rarely difficult all the way through; they are difficult at the joins, where one hand position changes into another or where the pattern breaks. Find the join, isolate it with a note of context on either side, and drill that. Then widen the unit by a bar and check that the join still holds inside the bigger phrase.

Overlapping is the trick that makes small units add up to a piece. If you drill bars 9–10 and then bars 11–12 separately, bar 10 into bar 11 is a join you never practised. Drill 9–10, then 10–11, then 11–12, then 9–12.

Slow enough to be right

Playing slowly is the most repeated and the least followed advice in piano teaching, so let me be specific about what it means. Slow enough is a tempo at which you can play the unit three times in a row without a wrong note, a hesitation, or a tension you notice. If you cannot, it is not slow enough, and the metronome dial goes down, however humiliating the number.

The reason is mechanical. Every repetition trains something. A repetition with a mistake in it trains the mistake with the same efficiency as a clean repetition trains the passage. The brain does not know which one you meant. When you play a passage too fast and stumble, you are not practising the passage; you are practising stumbling, and you are getting better at it.

Once the unit passes three-in-a-row, raise the tempo a few clicks — no more than four or five beats per minute at a time. Chopinly's metronome will stamp each tempo onto the goal so the goal's page shows the climb as a line over days. Watching that line is a surprisingly strong reason to keep going.

Hands separately, then together, then separately again

Beginners are told to practise hands separately and then stop doing it as soon as they can play hands together at all. That is backwards. Advanced players practise hands separately more, not less, because a hand on its own reveals problems the other hand was covering: a left hand that has never actually learned the fingering, a right hand that is relying on the bass to keep time.

A useful pattern for a difficult passage: right hand alone until secure; left hand alone until secure; together slowly; then left hand alone again, at the together tempo, and notice how much it has forgotten. Then together again.

Repetition with a purpose

Repetition is where practice actually happens, and it is also where attention goes to die. The tenth time through a passage is the one where your mind is on dinner. Two things keep repetitions alive.

The first is to change something each time: the dynamic, the articulation, the rhythm (dotted rhythms, then reversed dotted rhythms), the accent (on the second sixteenth instead of the first). Each variation forces the hands to solve the passage again rather than replay a recording.

The second is to count in a way that ends. "Play it until it's good" never ends. "Three clean in a row, then stop" ends, and it ends on a success. If you make a mistake on the third, the count resets to zero. This is the single best practice rule I know, and it makes the next rule possible.

Test, don't just repeat

Repetition is training. It is not evidence. The evidence that a passage is learned is that you can play it once, cold, in context, at tempo, after doing something else. So end each unit of work with a test: put the metronome on the target tempo, play the unit inside the surrounding phrase, once, and be honest about the result. If it fails, the unit goes back in the queue for tomorrow. If it passes, write that down.

Recording is the least forgiving test and the most useful. Record the passage on your phone, listen back, and you will hear the unevenness your hands were hiding from your ears. Chopinly's recorder keeps the take attached to the piece, so next month you can play this week's take against that one and hear what changed.

Write down where you stopped

The last two minutes of a session are worth more than the first ten, and almost everybody skips them. Before you get up, write one line per goal: what tempo you reached, which bar is still unreliable, what you want to try tomorrow. LH bars 9–12 clean at 76; join into 13 still rough; try dotted rhythms.

Tomorrow, that note is the first thing you see when you press play on the piece, and the first five minutes of the session — normally spent rediscovering where you were — are spent working. Over a month, the notes thread on a piece becomes the story of how you learned it, which is more useful than it sounds when you come back to the piece a year later.

Play through, but at the end

None of this means you may never play the piece from the top. You should; that is what the work is for, and the pleasure of it is what keeps you at the instrument. Just do it at the end of the session rather than the start, as a reward rather than a warm-up, and do it once with full attention rather than four times with half. Notice which bars wobbled. Those are tomorrow's units.

A session, concretely

Here is what forty minutes looks like when the pieces above are in place.

Forty minutes of this beats two hours of playing through, and it is not close. The log will show you why: after a month, the passage you drilled has forty minutes of practice on it and a tempo line from 60 to 100, while the passage you played through has a few seconds a day and no line at all.

What to do tomorrow

Start with the note you wrote. Test yesterday's passage cold before you drill it — if it holds, it was learned; if it doesn't, it was only rehearsed, and it needs another slow day. Then pick the next join.

That is the whole method: a goal, a small unit, slow and correct, repetition that ends, a test, a note. It is not a secret. It is just uncomfortable enough that most people do not do it, which is exactly why it works.