The 10,000-Hour Rule

What it is

The rule says that expertise has a price, and that the price is about 10,000 hours of practice. It is attractive because it is specific, it turns talent into a matter of time, and it gives a learner a number to work towards.

The rule's source is a study of violinists. In 1993 Anders Ericsson and two colleagues compared three groups at the Music Academy of West Berlin, each of ten players, rated best, good and less accomplished. By age 18 the best group had accumulated an average of 7,410 hours of practice alone, the good group 5,301 and the less accomplished group 3,420. These are the figures reported in the 2019 replication by Brooke Macnamara and a colleague, who also report that the original authors summed practice only up to age 18. The 1993 authors called the kind of practice that tracked the gap deliberate practice, and they argued that accumulated practice largely accounted for the differences between the groups.

The popular rule arrived later. The replication's authors write that Gladwell's Outliers described what they call the now famous 10,000-hour rule, and they treat the book as one of the popular works inspired by the deliberate-practice view. The rule therefore sits one step away from the study. It is a round number attached to a research result, and the research was about a particular kind of practice in a particular group of musicians, not about a threshold for mastery.

The distinction matters because the two things can come apart. A claim that practice matters is a claim about a direction. A claim that 10,000 hours produces mastery is a claim about a quantity, and a quantity can be tested. The replication below found that the quantity did not separate the groups it was tested on.

In effect

In research: the Berlin and Cleveland violinists

The replication is the cleanest test of the quantity claim. Macnamara and a colleague repeated the 1993 design in 2019 with 39 violin students in three groups of 13, using a double-blind procedure. The best and good groups were nominated by faculty at the Cleveland Institute of Music, and the less accomplished group came from the music department at Case Western Reserve University.

Practice alone by age 18 averaged 8,224 hours in the best group, 9,844 in the good group and 4,558 in the less accomplished group. The gap between the top two groups was not statistically significant. By age 20, both top groups had averaged more than 10,000 hours, and the figure did not separate them. Current weekly practice was significantly lower in the best group than in the good group, at 22.7 hours against 30.7.

The original authors had written that it is impossible for an individual with less accumulated practice at some age to catch up. The replication's authors say their result contradicts that claim. Across the three groups, accumulated practice alone accounted for 26 per cent of the variance in performance, against 48 per cent in the original study. The authors describe 26 per cent as not inconsequential and say it does not support the original claim that practice largely accounts for performance level.

In everyday learning: a boy and a ball

Matthew Syed's Black Box Thinking (2015) uses the rule as shorthand. After describing experts who perform well after years of practice, it says these findings have led to the conclusion that expertise is, at least in part, about practice, the so-called 10,000-hour rule (p. 49). A footnote on the same page states a narrower point, that extended practice seems to be a prerequisite for expertise in predictable environments.

The book's central case is David Beckham, who as a boy managed five or six keep-me-ups at first, reached 50 after six months and 200 six months after that, and by nine could keep a ball in the air 2,003 times (pp. 270 to 271). The case shows long, patient practice. It cannot show that practice produced the result, because it has no comparison with players who practised as hard and did not reach the top.

In publishing: how a number becomes a rule

A rule needs a carrier, and here the carrier was a best-selling book. Outliers carried a result about accumulated violin practice into popular use as a slogan about mastery, and Syed's phrase "so-called" treats it as already familiar. The number travelled further than the qualifications around it. The study measured practice alone by age 18 in three small groups. The rule applies one figure to every field, and the 2014 meta-analysis found that practice explained about 26 per cent of performance variance in games, 21 in music, 18 in sport, 4 in education and under 1 in the professions, with about 12 per cent overall.

What it does not say

It does not say that practice is unimportant. The meta-analysis authors describe deliberate practice as unquestionably important, and in the replication the less accomplished group had logged roughly half the practice of the other two.

It does not say that 10,000 is a threshold. In the replication the best and good groups both passed 10,000 hours by age 20, and the best group had logged fewer hours than the good group by age 18 and practised less each week.

It does not settle the dispute over definitions. Ericsson argued in 2016, in a commentary on a sports meta-analysis by Macnamara and colleagues, that deliberate practice had been stretched to cover group activities, watching games on television, play and competition. In 2019 he and a co-author reanalysed the data with three criteria for reproducible performance and purposeful practice. Practice then explained 29 per cent of the variance, and 61 per cent after correction for measurement unreliability. Both sides have a case, and the rule belongs to neither of them.

It does not rest on large samples. The replication used 39 players, a third more than the original, and its authors list the small sample, the tendency of retrospective estimates to run high and a possible difference in expertise between their groups and the Berlin groups as limits.

It does not claim to apply to every domain. The share of variance explained by practice runs from about 26 per cent in games to under 1 per cent in the professions, so a single number of hours cannot describe them all.

Who developed it

The label comes from Malcolm Gladwell and Outliers (2008). The data it rests on come from Anders Ericsson and colleagues (1993). The most direct tests of it are the work of Brooke Macnamara and colleagues, in a 2014 meta-analysis and a 2019 replication.

---