Practice Routines That Actually Work

Why long unstructured practice underperforms short focused sessions, and a concrete structure you can use from tomorrow.

Practice Routines That Actually Work

The most common practice mistake is not too little time โ€” it is unfocused time. An hour of playing through a piece repeatedly feels productive and produces much less improvement than twenty minutes of targeted work.

Why playing through does not work

Playing a piece start to finish mostly rehearses the parts you can already play. The difficult bar gets played once, poorly, and then you move on. Repetition without attention builds in mistakes rather than removing them.

Worse, speed comes from the sections you know well, so playing through reinforces an uneven performance: fast where it is easy, hesitant where it is not.

A structure that works

1. Warm-up (3โ€“5 minutes). Scales, arpeggios, or easy familiar material. Not technique for its own sake โ€” just getting the hands moving and the attention engaged.

2. Targeted work (10โ€“15 minutes). This is the actual practice. Identify one or two specific problems and work only on those.

3. Consolidation (5โ€“10 minutes). Play larger sections, or the whole piece, applying what you just fixed.

4. Review (2โ€“3 minutes). Play something you already know well. Ending on success matters more than it sounds.

Twenty-five focused minutes beats an hour of drift.

The targeted work itself

For a passage that will not come right:

Isolate it. Two bars, sometimes one. Not the whole page.

Slow down โ€” properly slow. Slow enough that you can play it perfectly. If you make mistakes at this tempo, it is still too fast. Most people practise just above the tempo where they can actually play it cleanly, which is why the same mistakes persist for weeks.

Hands separately. If one hand is the problem, fix that hand alone.

Find the actual cause. "This bar is hard" is not a diagnosis. Is it a fingering problem? A rhythm you have not internalised? A jump you cannot see early enough? Each has a different fix.

Use a metronome, incrementally. Get it clean at 60, then 66, then 72. Jumping from 60 to 90 reintroduces the errors you just removed.

Reconnect. Play across the seam โ€” the last bar before the problem plus the problem plus the bar after. Problems often live at the joins, not in the passage itself.

Spacing beats massing

Practising one thing for forty minutes is less effective than three short sessions across a day. Consolidation happens between sessions, including while you sleep. This is well established in motor learning research and it is the simplest change most people can make.

Recording yourself

Recording is uncomfortable and extremely informative. Things that feel fine while playing often sound uneven, rushed, or unbalanced on playback. One recording a week tells you what your ears are missing in the moment.

What "practised" should mean

A passage is practised when you can:

  • Play it correctly three times in a row, not once
  • Play it from a bar before, not only from the start
  • Play it at tempo after not having played it for a day

If it only works on the fourth attempt, it is not learned yet.

A realistic note

Progress is not linear. There are days when things feel worse than last week โ€” usually a sign that you are attempting something harder, not that you are regressing. Consistency over months matters far more than any single session.