How to Get Past a Plateau After Months of Practice

Five months in, the hours are still going in. You have not missed much, you still open the thing on the days you said you would, and by any fair standard you have done what the plan asked of you. That is the hard part and it is already behind you. Most people are not five months into anything.

And the work looks the way it looked in month three.

This is the point where a lot of otherwise determined people quietly decide they have found their ceiling. They have not. A plateau is almost never talent running out. It is a load problem. The practice became comfortable, and comfortable practice maintains a skill beautifully without moving it. The fix is not more hours of the same thing. It is a change in what the hours ask of you, made one variable at a time so you can see which change did the work.

Why am I not improving even though I practice every day?

Because you got good enough at the current version of the task that it stopped being demanding. A skill improves when a session asks for something you cannot quite do yet, and after a few months of repetition most practice quietly stops asking.

The clearest place to watch this happen is a gym, where the load is a number. Do ten push ups a day and your body adapts to ten push ups and then stops, and no amount of consistency changes that until the demand changes. Coaches call the fix progressive overload, and it is the same mechanism whether the load is weight, tempo, complexity or precision.

The research points the same way from a different direction. When a meta-analysis pooled every domain where deliberate practice has been measured, the authors concluded that practice is important but not as important as has been claimed, which is another way of saying that hours alone do not decide the result. The hours are still the price of entry. What the hours ask of you is the part you can still change today.

So read the plateau as feedback rather than as a verdict. Your practice is working exactly as designed. The design is what needs the edit, and everything below is more demanding than what you are doing now, not less.

Which variable should I change first?

Change one, not five, and pick the one that makes the current task hardest for you specifically. The usual five are difficulty, tempo, target size, constraint and attention, and changing several at once guarantees that when something moves you will not know what moved it.

• Raise the difficulty. Take the next version up: heavier, faster, longer, more complex. The smallest increment you can actually feel is the right one.
• Slow the tempo down. Counterintuitive and often the fastest fix. Take it slowly enough that it is genuinely clean, then rebuild the speed. Speed practiced over sloppiness only makes the sloppiness permanent.
• Shrink the target. Stop practicing the whole piece and practice the four bars that fall apart. Most plateaus are three bad seconds sitting inside two good minutes.
• Add a constraint. One take only. No notes. Half the time. Constraints force a different solution, and a different solution is exactly what a plateau is short of.
• Change what you attend to. Same repetition, different thing to watch: your breathing, your timing, the sound rather than the hands.

Give the changed version its own name in your notes and treat it as a new exercise. You are not doing the old thing harder. You are doing a different thing, and it deserves its own line.

Give it a fair trial before you rule it out. Almost every one of these feels worse for the first week, because you have just taken away the version of the task you were fluent at and replaced it with one you are not. That dip is the change working, not the change failing, and mistaking one for the other is why people cycle through five fixes in a month and keep none of them.

How do I get a useful outside eye on my work?

Ask one better person for one specific fix rather than for feedback in general. A vague request produces a vague answer, and the person you asked will spend their good will being encouraging instead of useful.

The request that works has three parts: what you were trying to do, what you think is going wrong, and the single thing you want them to look at. Here is bar nine, my hand tenses on the shift, tell me what my wrist is doing. That costs them ninety seconds and gives you a month of direction. You will often get a better answer from somebody two years ahead of you than from somebody twenty years ahead, because they still remember the fix.

If you do not have anybody, buy an hour. One paid hour with a good teacher, aimed at a single question you wrote down in advance, is worth more than a month of guessing and costs less than most people spend on gear. Send the question ahead of time so the hour starts at the answer instead of arriving there at minute fifty.

Who should I be giving a specific note to?

Somebody about a year behind you, and this month rather than eventually. Giving a precise note is the same skill as receiving one, it costs roughly twenty minutes, and it is the fastest way to find out whether you actually understand the fix you have been prescribing to yourself.

Here is the mechanism, stated plainly rather than warmly. When you watch somebody else's attempt and try to name exactly what to change, you have to convert a feeling into an instruction. That conversion is where you discover which parts of your own technique you can explain and which parts you have only been getting away with. The sentence you stall on while explaining is a hole in your own understanding, and for a person sitting on a plateau that is information you cannot buy any other way.

Do it the way you would want it done. One specific thing, named, with something to try instead. Then say what they did well by name, out loud, in front of somebody whose opinion matters to them, because a precise public commendation costs you one sentence and lands on them like a raise. They get a shortcut you had to earn slowly. You get an honest map of your own gaps. Both of those are real, and neither of them waits until you have arrived somewhere.

How long do I hold the new version before judging it?

Three weeks, or roughly twenty sessions, before you decide whether the change worked. Skills adapt on a slower clock than motivation does, and judging a new version after four days is how people churn through five fixes in a month and conclude that none of them work.

This is the calm part of the method and it is also the demanding part. Waiting here is not patience instead of progress, it is what lets you tell signal from noise, and it stops you re-deciding your entire approach every night. Write down the date you started the new version and the date three weeks out. Do not reopen the question in between.

A plateau is a load problem

Nobody who is genuinely good at anything got there without a stretch of months where nothing visible happened. The difference between the people who came out the other side and the people who stopped is almost never talent and almost never hours. One group treated the flat stretch as information and changed the load. The other treated it as a verdict.

So raise the difficulty. Hold it for three weeks. Get one specific note, and give one. You are not at your ceiling. You are at the top of the current version of the exercise, and there is always a next version.

Sources

• Cleveland Clinic, Your Simple Guide to Progressive Overload Training
• ACSM's Health & Fitness Journal, Ten Ways to Implement the Principle of Progressive Overload
• PubMed, Deliberate practice and performance in music, games, sports, education, and professions: a meta-analysis