Deliberate practice: why hours alone don't make experts

Evidence grade: STRONG for the core distinction, MIXED on the maximal claims. That structured, feedback-rich, weakness-targeted practice outperforms raw experience is robust across domains. Ericsson's stronger claim — that deliberate practice largely explains expert performance — met serious meta-analytic challenge; the honest reading keeps the mechanism and drops the totalizing arithmetic.

The most consequential finding in expertise research is also its most misquoted. Anders Ericsson's studies of elite performers (the 1993 violinist study above all) found that what separated the best wasn't hours logged but a particular kind of practice — and that ordinary experience, even decades of it, reliably plateaus. The finding escaped into culture as "10,000 hours makes anyone an expert", which inverts the actual lesson: the hours were never the ingredient; the structure of the hours was.

What makes practice deliberate

Ericsson's characterization, consistent across the research program: deliberate practice is (1) designed to improve a specific aspect of performance — not performance in general, and pointedly not the parts you're already good at; (2) conducted at the edge of current ability, where failure is frequent (desirable difficulty, by another road); (3) saturated with immediate, informative feedback (the feedback conditions); and (4) effortful and repeated — with repetitions targeted at the weakness until it moves, then re-targeted.

The contrast case is what most time-on-task actually is: performance, not practice. Playing the piece, playing the game, typing the essay — executing at your current level, comfortably, with the weaknesses politely avoided. Experience of this kind famously fails to improve professionals (the expertise literature's recurring scandal: years of experience barely predicting skill in many fields). The plateau isn't a talent ceiling; it's the signature of practice that stopped being deliberate.

What the debates settled

The maximal claim — practice-explains-(nearly)-everything — didn't survive scrutiny: Macnamara, Hambrick and Oswald's meta-analysis (2014) found deliberate practice explaining a meaningful-but-minority share of performance variance, with the rest owed to starting age, cognitive factors, and much else. Ericsson disputed their codings; the exchange was real science working. The stable residue matters for education: practice structure is the largest controllable factor — genes and starting ages aren't levers; the design of Tuesday's practice is. And a second Ericsson emphasis survives untouched: deliberate practice requires knowing what to work on, which for novices means the diagnosis usually must come from outside — a teacher, or a system that watches (why untutored practice plateaus is this principle in miniature).

What the evidence doesn't say

It doesn't say 10,000 hours of anything produces expertise — that's the folklore this page exists to correct.
It doesn't say all practice must be grim — deliberate practice is effortful by design, but for children especially, structure can wear game clothing so long as the targeting, edge, and feedback survive the costume.
It doesn't apply its full machinery to every goal — deliberate practice is the technology of excellence; fluency for ordinary life needs its lighter cousin (targeted practice with feedback), which is the version schools mostly need.

In the classroom

Audit time-on-task for performance-vs-practice: "do more typing" is performance; "this week: the number row, which your accuracy data says is your weakness" is practice.
Target the weakness, not the average — practice aimed at what a child already does well is rehearsal of comfort.
Keep the edge calibrated: frequent-but-recoverable failure is the setting (mastery structures automate the calibration).
Close every rep with feedback and re-attempt — the next-ten-minutes loop.

How Wiz Kids applies this

The system is built to keep practice deliberate without a tutor per child: per-skill accuracy data identifies weaknesses, the prerequisite graph holds the edge-of-ability position, every action gets immediate task-level feedback, par scoring targets efficiency once correctness is automatic (a new specific aspect to improve), and the spaced review queue re-targets whatever slips. The stars are the costume; the four conditions are the machine.

References

Ericsson, K. A., Krampe, R. T., & Tesch-Römer, C. (1993). The role of deliberate practice in the acquisition of expert performance. Psychological Review, 100(3).
Ericsson, K. A., & Pool, R. (2016). Peak: Secrets from the New Science of Expertise — the accessible statement, including the corrections to the folklore.
Macnamara, B. N., Hambrick, D. Z., & Oswald, F. L. (2014). Deliberate practice and performance: A meta-analysis. Psychological Science, 25(8).
Ericsson, K. A. (2016). Summing up hours of any type of practice versus identifying optimal practice activities. Perspectives on Psychological Science, 11(3) — the reply.

© Glu IO Pty. Ltd. — Wiz Kids (wiz.kids). Link freely; republication requires permission — see terms. Found an error in our reading of the research? We correct fast: tell any teacher piloting Wiz Kids.