Practice explains less of expert performance than the famous number implied

Where the number came from

A widely cited study of musicians compared groups at different levels of accomplishment and found that the most accomplished had accumulated substantially more hours of solitary, effortful practice by a given age. It was a careful piece of work and its authors made a specific argument: that a particular kind of practice, not mere repetition, was what distinguished levels of performance.

A popular book later rendered this as a round number of hours required for mastery, and that version travelled far beyond anything the research supported. The original figure was an approximate average for one group in one domain, not a threshold, not a requirement, and not a promise. The authors of the underlying work have said as much.

What deliberate practice actually specifies

The concept is narrower and more useful than the slogan. Deliberate practice means working at the edge of current ability, on a specific component rather than the whole activity, with immediate feedback and repeated correction, usually under guidance. It is effortful, it isn’t enjoyable in the way performing is, and most people do less of it than they think.

The distinction from ordinary repetition matters. Doing an activity many times in the usual way produces improvement early and then a long plateau, because the performance settles into something automatic and nothing forces it past its current level. This part of the picture is well supported and is the practical takeaway from the whole literature.

What the replications and meta-analyses found

When later researchers pooled studies across many domains, the relationship between accumulated practice and performance turned out to be real but far more modest than the popular account, and highly variable by field. Practice accounted for a substantial share of the differences in some structured activities such as games and music, and much less in domains with unpredictable conditions, such as professions where each case differs.

The disagreements about how to interpret this have been sharp and are not settled. Critics of the pooled analyses argue that they include studies with poor measures of practice quality, lumping ordinary repetition together with the deliberate kind. Defenders point out that the original claims were made broadly and should be tested broadly. What is no longer defensible is the strong version: that practice quantity is the dominant explanation of expert performance in general.

It is worth noting explicitly that no figure for how much practice explains what is quoted here, because the estimates differ across analyses and reporting one as though it were settled would misrepresent an active dispute.

What else is in the picture

Differences in starting points, in the age at which someone began, in physical characteristics where relevant, in access to coaching, and in how quickly a person improves per hour all contribute. The last of those is interesting and under-discussed: two people doing identical practice don’t improve at identical rates, and that variation is part of what the practice-quantity account leaves out.

None of this is an argument for fatalism, which is the usual overcorrection. Practice remains the largest factor an individual actually controls, and in almost every domain the amount of well-structured practice most people do is far below what would be productive. The honest framing is that practice is necessary, powerful and not sufficient.

The part that survives and is worth using

Strip away the disputed magnitudes and a robust core remains. Improvement comes from working on specific weaknesses rather than repeating the whole activity. Feedback has to be tight enough to identify what went wrong. Difficulty has to be adjusted upward as capability grows, or the practice stops being practice. Guidance shortens the process considerably, mostly by identifying the weakness you cannot see.

These claims are supported across many settings and they do not depend on any particular number being correct. They also happen to describe what good coaching has always done, which is a reasonable sign that the underlying observation was sound before it was quantified.

Why the myth was so appealing

A fixed hour count is comforting in both directions. It promises that mastery is available to anyone willing to put in the time, which is democratic and partly true, and it excuses those who have not by pointing at the size of the number. Neither reading requires thinking about the quality of the hours, which is the part that actually varies.

That is the general lesson worth keeping. When a research finding turns into a single memorable figure, the figure has usually absorbed the conditions and dropped them. Asking what the number was measured on, and in whom, restores most of what was lost.