C16 Practice explains only part
narrowed 4 attacks: 3 survived, 1 narrowed · serves Q1 · last change 27 October 2026
Deliberate practice, as it is usually measured, explains only a minority of the differences in how well people perform, in every domain studied. When only studies that use the strict original definition are kept, one reanalysis by the theory's author found about 29%, and estimated 61% after correcting for measurement error (Ericsson & Harwell 2019).
In practice
Plan for practice, but do not tell learners that hours alone decide who becomes good. Starting point, feedback, the kind of task and other factors also matter, especially in school and work.
Go deeper
Where it holds
Main limits
- About how much of performance differences practice accounts for, not whether practice helps (it does).
- Practice is mostly measured by recalled hours, so the shares are estimates (Macnamara et al. 2014).
- The original theory claimed practice explains most differences (Ericsson et al. 1993).
- Counting only strict deliberate practice and correcting for measurement error, one reanalysis by the theory's author estimated 61% (Ericsson & Harwell 2019); uncorrected it was 29%.
- Kill-condition tests survived: Macnamara & Maitra 2019 (violin replication, 26%); Macnamara, Moreau & Hambrick 2016 (sports meta, 18% overall, 1% elite); Platz et al. 2014 (music meta, corrected r = 0.61 ≈ 37% variance, still below half).
All limits (8)
Inside the circle: the core of the claim. Outside: every bound and carve-out in the claim file, numbered, with the attack that forced or added it where the record shows one.
forced by a narrowing attackin the original wordingwritten at creation, or added without a narrowing
Warm cards in the list were added by later attacks; plain cards were in the original draft or written at creation.
Share of individual differences in performance accounted for by accumulated deliberate practice, across domains; not whether practice improves performance within a person, and not specific practice schedules (C11).
Written when the claim was created (16 October 2026); not forced by an attack
Evidence on file: Macnamara, Hambrick & Oswald 2014 (Psychological Science; meta-analysis; 26% games, 21% music, 18% sports, 4% education, under 1% professions); Ericsson, Krampe & Tesch-Römer 1993 (Psychological Review; the theory under test).
Written when the claim was created (16 October 2026); not forced by an attack
The minority share holds for practice as usually measured and defined; for strict-definition studies corrected for unreliable measures, one reanalysis of 14 effects (games, music, sports, one education study, none from professions; reliabilities assumed by the authors) estimated 61% (Ericsson & Harwell 2019).
Forced by attack #1 · Ericsson & Harwell 2019 · 16 October 2026
Kill-condition test survived 2026-10-27: Macnamara, Moreau & Hambrick 2016 (Perspectives on Psychological Science 11(3), 333–350, DOI 10.1177/1745691616635591; full text stored; 52 independent samples, 63 effect sizes, N = 2,765): deliberate practice explained 18% of variance in sports performance (95% CI 12–25%); skill-level moderator significant, Q(2) = 7.04 — subelite 19% (r = .44), mixed 29% (r = .54), elite only 1% (r = .11, p = .50).
Added with attack #3 (survived; it did not narrow the claim) · Macnamara, Moreau & Hambrick 2016 · 27 October 2026
Internally paced sports reached 41% (r = .64), still below half.
Added with attack #3 (survived; it did not narrow the claim) · Macnamara, Moreau & Hambrick 2016 · 27 October 2026
Kill fails.
Added with attack #4 (survived; it did not narrow the claim) · Platz, Kopiez, Lehmann & Wolf 2014 · 27 October 2026
Kill-condition test survived 2026-10-27 extra6: Platz, Kopiez, Lehmann & Wolf 2014 (Frontiers in Psychology 5:646, DOI 10.3389/fpsyg.2014.00646; OA full text stored; 13 studies, N = 788): task-specific practice ↔ musical achievement rc = 0.61 (95% CI [0.54, 0.67]); under the majority-of-variance kill, ~37% (upper CI ~45%) still below half.
Added with attack #4 (survived; it did not narrow the claim) · Platz, Kopiez, Lehmann & Wolf 2014 · 27 October 2026
Authors caution against the r² reading; numbers still fail the kill.
Added with attack #4 (survived; it did not narrow the claim) · Platz, Kopiez, Lehmann & Wolf 2014 · 27 October 2026
1 of 8 bounds and carve-outs were forced by a specific attack. 3 attacks hit the boundary and did not move it: #2, #3, #4.
How it changed
The ledger records 4 attacks, but no earlier wording of C16 is on record, so the text cannot be replayed step by step. Only the current wording is shown.
Deliberate practice, as it is usually measured, explains only a minority of the differences in how well people perform, in every domain studied. When only studies that use the strict original definition are kept, one reanalysis by the theory's author found about 29%, and estimated 61% after correcting for measurement error (Ericsson & Harwell 2019).
What would change this
What would kill it: A meta-analysis or large study finds that deliberate practice explains the majority (more than half) of performance differences in a domain. Added as a claim on 16 October 2026, promoted from a candidate belief in the evidence library.
Attack record
4 attacks: 3 survived, 1 narrowed. Counts of attacks, not a score.
Full attack list (4)
Restricting Macnamara et al.'s meta-analysis to studies that measured practice under the original strict definition of deliberate (or purposeful) practice, Ericsson & Harwell 2019 found practice explained about 29% of performance differences (r = 0.54, k = 14 effects), and 61% after correcting for unreliable practice and performance measures, more than half, which is C16's kill-condition.
Strongest reading and verdict reasoning
Steelmans to: C16 dies on a meta-analysis that finds deliberate practice explains the majority of performance differences in a domain; this reanalysis claims exactly that for the strict definition across games, music and sports.
Why: Broadening attack run 2026-10-16 (broad, cycle 1) Europe/Stockholm. Full text read (open access; not stored). Random-effects r = 0.54 (95% CI 0.44–0.63), about 29% of variance vs 14% in Macnamara et al. 2014; by domain games r = 0.50 (k = 5), music r = 0.71 (k = 3), sports r = 0.58 (k = 5), education one study (r = 0.31, spelling bee), professions none. The 61% depends on assumed reliabilities (0.6 for practice estimates, 0.8 for performance) chosen by the authors. Kill withheld: the uncorrected share is still a minority; the corrected figure rests on assumed reliabilities and few studies; it is a reanalysis by the theory's originator, not an independent meta-analysis. Narrowed instead: the minority share is bound to practice as usually measured and defined. Not a restatement: C16's Bounds named Macnamara et al. 2014 and the 1993 theory, not a strict-definition reanalysis. Not a preprint. Claim and Bounds rewritten same turn; brief and the practical rule Limits updated.
The preregistered replication of the study that founded the claim that practice "largely" accounts for expert performance (Ericsson et al. 1993) would meet C16's kill-condition if accumulated practice explained more than half of the performance differences between violinists.
Strongest reading and verdict reasoning
Steelmans to: C16 dies on a large study where deliberate practice explains the majority of performance differences; re-running the canonical violin study under preregistration is the most direct test of that.
Why: Week run 2026-10-16, cycle 4 Europe/Stockholm. Full text read (PMC6731745; not stored). Preregistered on OSF, results-blind accepted, double-blind interviews, 39 violinists in three skill groups. Practice alone to age 18 explained 26% of variance (η² = 0.26, 95% CI 0.03–0.44) vs 48% in the original; teacher-designed practice 23%; best and good violinists did not differ in accumulated practice (8,224 vs 9,844 hours, p = 0.364). Kill-condition failed. Limits: 39 participants, one domain, retrospective estimates; the upper confidence bound (0.44) is still below half. Not a preprint.
A sports-domain meta-analysis would meet C16's kill-condition if deliberate practice explained the majority of performance differences in sports; Macnamara, Moreau & Hambrick 2016 found 18% overall and only 1% among elite athletes.
Strongest reading and verdict reasoning
Steelmans to: C16 dies on a meta-analysis that finds deliberate practice explains the majority of performance differences in a domain; sports is a core domain of the theory, and the elite split is the strongest place the majority claim should hold if true.
Why: Extra attack pass 2026-10-27 Europe/Stockholm (Jonatan: keep implementing the improve plan). widened-search slot for C16 (3 fruitless; least recently targeted among 3e qualifiers). Full text read (author-posted published PDF, now stored). 52 samples, 63 effect sizes, N = 2,765; 18% overall (95% CI 12–25%); elite r = .11, p = .50 (1%); skill-level Q(2) = 7.04. Internally paced sports 41% still below half. Kill-condition failed. Not a restatement of Macnamara et al. 2014 in Bounds: that entry had no elite split and was not logged as a kill-test. Not a preprint. Claim sentence unchanged; Bounds gain the kill-test. Status unchanged (weakened). Fruitless streak reset.
A music-domain meta-analysis would meet C16's kill-condition if deliberate practice explained the majority of performance differences in music; Platz, Kopiez, Lehmann & Wolf 2014 found a corrected correlation rc = 0.61 (95% CI [0.54, 0.67]) between task-specific practice and objectively measured musical achievement (13 studies, N = 788).
Strongest reading and verdict reasoning
Steelmans to: C16 dies on a meta-analysis that finds deliberate practice explains the majority of performance differences in a domain; music is a core domain of the theory, and this OA meta is independent of the Macnamara sports meta already survived.
Why: Extra6 attack pass 2026-10-27 Europe/Stockholm (parent: C19 OA kill-survive hunt exhausted; pivot to below-floor C16). Full text read and stored (Frontiers OA PDF). Mean corrected rc = 0.61 [0.54, 0.67]; ρ = 0.61; ro = 0.54. Under C16's majority-of-variance kill (>50%), r = 0.61 implies about 37% of variance and the upper CI r = 0.67 implies about 45% — still below half. Authors warn against the r² reading and treat 0.61 as a lower bound on the association; that does not push the share over 50% on the numbers reported. Kill fails. Not a restatement of Macnamara et al. 2014 music 21% in Bounds (different meta, higher estimate, now logged as kill-test). Not a preprint. Claim sentence and Status unchanged (weakened). Bounds and the practical rule updated same turn. C16 now 3 of 3 survivals → floor.
Evidence
8 linked sources: 6 supports, 1 limits, 1 challenges (first-pass draft, pending human review). supports: Macnamara et al. 2014, Macnamara & Maitra 2019, Mosing et al. 2014, Hambrick et al. 2014, Macnamara et al. 2016, Platz et al. 2014 · limits: Ericsson & Harwell 2019 · challenges: Ericsson et al. 1993