Learning and the mind
Deliberate practice
Expert skill is built by structured, effortful practice aimed at specific weaknesses with feedback, not by hours spent doing the activity.
Anders Ericsson, Ralf Krampe and Clemens Tesch-Römer, Psychological Review 100(3) (1993).
Ericsson and colleagues defined deliberate practice as a highly structured activity whose explicit goal is to improve performance: tasks are invented to attack specific weaknesses, performance is monitored to find the next thing to fix, and the work takes effort and is, in their words, not inherently enjoyable. Playing the piece you already play well is not deliberate practice; neither is a match or a concert. Their 1993 paper argued that the gap between elite performers and the merely good is largely explained by how much of this practice they have accumulated, and backed it with violinists and pianists at the Music Academy of West Berlin whose practice histories lined up with their rated skill.
The mechanism is a feedback loop at the edge of ability. You work on something you cannot yet do, get fast and specific information about what went wrong, adjust, and repeat, with a teacher choosing the next task. Ericsson's violinists rated practice alone as the most relevant activity for improving and among the most effortful, and the best of them organised their days around it. In a working life the equivalent is code review by someone better than you, a recording of your own talk, or a test harness that reports exactly which cases failed rather than a green or red run.
The popular version has drifted a long way from the paper. Malcolm Gladwell's 10,000-hour rule was taken from the best violinists' average by age 20; Ericsson wrote in 2016 that it was only an average, that half of those ten had not reached it, and that nothing about the number is special. The size of the effect is also contested. Macnamara, Hambrick and Oswald pooled 88 studies in 2014 and found deliberate practice explained 26 percent of the variance in performance in games, 21 in music, 18 in sports, 4 in education and under 1 in professions. A 2019 attempt to repeat the Berlin study found the best and good violinists had practised about the same amount. Practice is necessary. It is not the whole story.
Example
In the study published in Psychological Review in 1993, Ericsson, Krampe and Tesch-Römer asked 30 young violinists at the Music Academy of West Berlin, 10 each rated best, good and future music teachers, to reconstruct their practice histories. By age 18 the best group had accumulated an average of 7,410 hours of practice alone, the good group 5,301 and the teachers 3,420; 10 middle-aged violinists from the Berlin Philharmonic and the Radio Symphony Orchestra reported 7,336 hours by the same age. The figure of about 10,000 hours by age 20 for the best group, later made famous, was an average, and Ericsson wrote in 2016 that half of those ten had not reached it.
Where it applies
For a working engineer the deliberate part is choosing the task that exposes the gap: implement the algorithm you always import, debug without the profiler, write the proof before reading it. An evaluation set with per-case failures is a coach; a dashboard that says 94 percent is not. Pair it with the forgetting curve, since a skill not revisited decays like a fact.
Think about it
In the last month, what did you practise that you could not already do, and who or what told you what went wrong?
Applied it to something real? Log it.
Test yourselfWhat makes practice deliberate in Ericsson's sense, and what did the 2014 meta-analysis find about how much it explains?
Structured work on specific weaknesses, with feedback, effortful rather than enjoyable. Macnamara and colleagues found it explained 26 percent of performance variance in games, 21 in music, 18 in sports, 4 in education, under 1 in professions.
Related models
- Desirable difficultiesConditions that make learning feel slower and harder often make it last longer; the ease you feel while studying is a poor guide to what you keep.
- Feedback loopsA system's output feeds back into its input: balancing loops hold a level steady, reinforcing loops make change grow on itself.
- Circle of competenceKnow where your real understanding ends, and keep your big bets inside that line.
- ChunkingWorking memory holds only a few items, so experts get around the limit by grouping details into larger units they recognise at a glance.
Sources and further reading
- Anders Ericsson, Ralf Krampe and Clemens Tesch-Römer, The Role of Deliberate Practice in the Acquisition of Expert Performance, Psychological Review 100(3) (1993)
- Brooke Macnamara, David Hambrick and Frederick Oswald, Deliberate Practice and Performance in Music, Games, Sports, Education, and Professions: A Meta-Analysis, Psychological Science 25(8) (2014)
- Anders Ericsson and Robert Pool, Peak (2016), excerpt on the 10,000-hour rule, Salon