Does deliberate practice explain expertise?

Where did ten thousand hours come from?

A book rather than a study. Anders Ericsson's research on violinists in the early 1990s found the best performers had accumulated more hours of focused practice than their peers. Malcolm Gladwell turned an average from that work into a threshold in Outliers, and Ericsson spent years afterwards objecting to the version that spread.

The original claim was about the importance of structured practice. The popular claim is that a specific quantity produces mastery, and no study established that number as a requirement.

How much does practice actually explain?

It depends enormously on the domain, which is the central finding. Macnamara, Hambrick and Oswald pooled the literature across five areas and reported the variance in performance that deliberate practice accounted for in each.

Games came out at 26 per cent. Music at 21 per cent. Sports at 18 per cent. Education dropped to 4 per cent. Professions came in under 1 per cent.

Read the top of that range first: a quarter of the difference between chess players is explained by how much they practised, which is substantial. Then read the bottom. Among professionals, practice explained almost none of the difference in performance at all.

Why does the domain matter so much?

The pattern tracks how structured the activity is. Chess and violin have fixed rules, a clear standard of correct, and feedback that arrives immediately. You can tell a good move from a bad one and repeat the correction.

A profession has none of that reliability. The problems change, success depends on other people, feedback arrives late or never, and there is often no agreed definition of what a better performance would be.

Deliberate practice needs a target to aim at. Where the target is stable, hours convert into skill. Where it moves, they convert into experience, which is a different thing.

What explains the rest?

The authors are careful here, and so should you be. Their conclusion is that deliberate practice is important and less important than had been argued, which stops short of naming what fills the remaining variance.

Candidates include starting age, genetic factors, working memory capacity, coaching quality, and the plain fact that people who improve quickly tend to keep going. Untangling those is genuinely unresolved.

One trap is worth naming. If talented people practise more because practice is rewarding for them, then practice and ability are entangled in the data, and the correlation cannot separate the two.

What is worth doing?

Build feedback into the domain before adding hours. Where the evidence for practice is strongest, feedback was already there. In unstructured work, creating it is the higher-leverage move.

Stop treating the hours as the goal. Twenty-six per cent is the ceiling of what practice explained anywhere in this analysis, so hitting a plateau says little about whether you practised wrong.

Use methods with their own evidence. Spaced repetition and the memory techniques that survived testing beat undirected repetition, and they apply to the hours you already spend.

And notice the pattern this shares with growth mindset. Both ideas promise that effort is decisive, and both shrink considerably when measured carefully.