Practice scheduling for motor skills — varied vs blocked, spaced vs massed

Verdict

The principle is one of the most valuable ideas in skill acquisition: performance during practice and durable learning are different things, and they frequently point in opposite directions. The founding demonstration is clean — blocked practice was ~22% faster during acquisition, then 32% slower at delayed retention than random practice. Any coach or teacher who judges a schedule by in-session performance systematically selects the schedule that harms learning. That insight is robust and worth teaching.

The effect as a coaching prescription is much weaker than its textbook status implies, which is why the verdict is mixed rather than supportive. It is carried almost entirely by laboratory tasks, and it is null in exactly the population schools care about.

What the evidence shows

Three deflations the adversarial pass insisted on. First, Czyż's own sensitivity analyses cut the headline hard: after outlier removal the overall retention effect falls 0.63 → 0.43, applied settings → −0.01 (p=.94), and under-18s → −0.39. Second, the age moderation and the setting moderation are one confound, not two findings — in this corpus all older-adult experiments were laboratory studies and nearly all youth studies were applied, and the authors themselves ask whether setting is the real driver. Third, the two research teams are in open published dispute (Ammar & Schöllhorn replied in 2025), so these estimates are not settled.

Spacing is the more dependable of the two levers but is heavily conditional: large for simple discrete psychomotor tasks, near-nil for complex ones, and inflated roughly threefold in low-methodological-rigor studies.

Hereditarian-lens assessment

Risk: low — these are randomised within- or between-subject schedule manipulations, so selection and heritability cannot produce the contrasts. Worth noting what the topic does not claim: scheduling moves the rate and durability of skill acquisition, not underlying motor capacity. No practice schedule has been shown to move general athleticism (see talent & trainability).

Boundaries & what critics say

• Null in children — the single most important caveat for a school. Practitioners over-apply varied practice to the population it fits worst.
• Lab-to-field collapse — the pattern this database sees everywhere; here it is severe enough to reach zero.
• Complexity ceiling on spacing — real skills are complex, which is where spacing does least.
• Variability of practice has no clean modern meta-analysis; the differential-learning proxy is self-described exploratory, high bias risk, no transfer data.

Practical guidance

• Measure learning at a delay, never by end-of-session performance. This is the durable lesson.
• Use spaced practice for discrete, simple skills (fact-like motor elements); expect less on complex whole-skill work.
• For children, do not force random/interleaved schedules on the strength of the adult literature — the evidence there is null to negative.
• Prefer whole practice; part-practice is not superior at retention.

Open questions

• Whether the age moderation is genuinely developmental or an artifact of setting is unresolved by the authors themselves.
• No adequately powered field trial in children exists.