Jerry · September 8, 2026

jtr gets back on the guitar with 25 to 40 minutes some evenings, not a conservatory schedule. Work, family, a real life. Somewhere along the way, almost everyone absorbs the same three pieces of practice wisdom: put in your hours, space your sessions out so your brain can consolidate them, and if a session felt sloppy, it probably was. All three sound obviously true. Two of them are, at best, oversold, and one has now been tested directly on an instrument — and lost.

The 10,000-hour theory has an asterisk its fans skip

The "put in your hours" idea traces to a single, specific 1993 paper by K. Anders Ericsson, Ralf Krampe, and Clemens Tesch-Römer, and it claimed more than the popular retelling admits. The authors didn't say practice was one important factor among many — they wrote that they "reject any important role for innate ability" and that accumulated deliberate practice gives "a sufficient account" of who becomes an expert. That's the theory in its strongest, least hedged form.

Twenty-one years later, a team led by Brooke Macnamara ran the numbers on the whole field's worth of studies — 157 effect sizes, more than 11,000 people, across music, games, sports, education, and professions — and found deliberate practice explains 26% of variance in games, 21% in music, 18% in sports, and less than 5% everywhere else. Put plainly, and in the words of one of the study's own co-authors: "in all major domains in which deliberate practice has been studied, most of the variance in performance is explained by factors other than deliberate practice." Even in the exact domain the original theory was built on — music — roughly four-fifths of why one performer outplays another has nothing to do with how many hours of practice they logged.

That should have been the end of it, except it wasn't. In 2019, Ericsson and a co-author, Kyle Harwell, went back into the same underlying data and argued the 2014 meta-analysis had quietly redefined "deliberate practice" into something looser than Ericsson's own 1993 definition — which requires individualized coaching, targeted correction of specific weaknesses, and real feedback, not just solo repetition. Apply the strict original definition, they argued, and accumulated practice explains 29% of variance before correcting for measurement error, and 61% after. That's not a minor footnote; it's the theory's own authors going back into their critics' numbers and finding something very different.

Nobody outside this argument has settled it. What matters for someone practicing alone, without a coach, is that both sides of the fight actually point the same direction: if the high numbers depend on individualized, corrected, coached practice — which is Ericsson's own original claim — then an adult practicing alone at home was probably never going to inherit them in the first place. A separate study by Megha Macnamara and Sonali Maitra in 2019 tested this directly, comparing teacher-designed practice to unaided solo practice among real musicians, and found teacher-designed practice was not more predictive of outcomes — and the musicians themselves didn't rate it as more useful. So the honest starting point isn't "does deliberate practice work" — nobody in this literature disputes that engaged time on an instrument accumulates into skill. It's "does the specific, coached, maximalist version of the theory apply to solo adult practice," and the answer, from inside the theory's own camp, leans no.

The one number that survived: an effort ceiling, not a motivation problem

Buried in that same 1993 paper, underneath the famous hours claim, is a much more useful and far less quoted finding. Ericsson's team reviewed evidence of "essentially no benefit from durations exceeding 4 hr per day and reduced benefits from practice exceeding 2 hr," estimating the truly "effective" duration closer to one hour daily. They framed this as an effort-and- recovery constraint, the same reason athletes don't train maximally every single day — not a discipline problem. Elite training programs, per the same paper, start children on sessions of 10 to 20 minutes and build up gradually. That is itself a quiet admission that "more hours, unconditionally better" was never really the theory, even before its critics arrived.

The catch: that number came from elite violin students training multiple hours a day, not from anyone practicing 25 to 40 minutes. No study in this literature tested whether an analogous, much smaller ceiling exists at low daily volumes. It's useful only as a direction — more clock time isn't free of diminishing returns — not as a specific number for a short adult session.

The advice everyone repeats about spacing practice out — tested directly on a piano, and it failed

This is the real find. "Space your practice out instead of cramming it, so your brain can consolidate it overnight or between sessions" is close to universal folk wisdom, and it does have real backing — in general. A 1999 meta-analysis by John Donovan and David Radosevich found spacing beats massing practice together with a solid effect size (d = 0.46) across the motor- and verbal-learning literature. But buried in that same analysis is a moderator that changes everything for a skill like guitar: for highly complex tasks, the spacing advantage shrinks to nearly nothing (d = 0.07). Almost all of the underlying research, though, was built on simple lab tasks — finger-tapping sequences, button-press patterns — not anything as complex as playing real music. So for years, that complexity discount stayed a plausible extrapolation nobody had actually tested on an instrument.

In 2017, a team led by Melody Wiseheart finally did. They taught untrained adults real piano material — a 17-note sequence, then phrases from five actual songs — across five different spacing conditions, from immediately repeating a phrase to waiting 15 minutes between attempts, then tested retention a few minutes later. The result: no spacing effect at all. Not smaller. Not a trend in the expected direction. Zero measurable difference in note accuracy, timing consistency, or how consistently they pressed the keys, across either the short sequence or the actual song phrases.

Nine years later, in 2026, a separate lab — Ishaan Gupta and Timothy Rickard, the same Rickard whose group had already been fighting the sleep-consolidation battle described below — ran two further experiments manipulating spacing and rest length in a different motor-sequence task, explicitly cited Wiseheart's piano result as corroboration, and found the same pattern by a different route: spaced practice looked genuinely better while people were training, but that apparent advantage disappeared on a later test given under equal conditions. Two labs, nine years apart, two different tasks — one of them an actual musical instrument — and the central piece of advice most people never question did not survive contact with a direct test.

Sleep is a real, live fight — and the newest evidence cuts against the popular version

The companion claim to spacing is that sleep consolidates what you practiced that day. This one is genuinely contested among the researchers who study it, not settled in either direction, and it's worth reporting as the live argument it actually is rather than picking a side to make the article feel tidier.

Shanna Pan and Timothy Rickard's 2015 meta-analysis pooled 1,296 subjects across 34 studies and found that once you control for known confounds — time-of-day testing effects, total training duration, and fatigue built up during the original session — the apparent overnight advantage for motor skills didn't hold up as evidence of true sleep-driven consolidation. Esther Adi-Japha and Avi Karni published a formal, named rebuttal in 2016, arguing the meta-analysis wrongly treated every study's consolidation mechanism as identical and, specifically, that pooling children and adults into one analysis could be masking a real effect that shows up more reliably in children. Rickard and Pan defended their original conclusion in 2017. Then in 2022, Rickard's team went back a third time — not to re-argue the same data, but to run a completely different statistical method built to detect publication bias, and found the sleep-consolidation literature showed "pronounced" bias toward publishing positive results, which independently strengthens the case for the null.

The newest piece closes the loop on the age question specifically. A 2024 study by Van Roy and colleagues was purpose-built to test exactly what Adi-Japha and Karni's critique had flagged — whether pooling children and adults obscures something real — and found that offline motor-consolidation gains genuinely are largest in children and decline into young adulthood, with no age difference in how fast people learn a skill in the first place. That doesn't end the argument. It does mean that whatever sleep-consolidation advantage exists in this literature is calibrated more toward children's data than an adult's, which is exactly the direction that should make an adult guitarist skeptical of "sleep on it and you'll consolidate the lick" as a confident claim rather than a live, unresolved hypothesis.

Don't trust how the session felt

If there's one finding in all of this that's both solid and immediately usable, it's this: a session that feels fatigued, sloppy, or unproductive near the end is not, by itself, evidence that nothing was learned. Susan Lee and Robert Genovese found back in 1988 that massed, unbroken practice depresses how good performance feels in the moment without a matching depression in what was actually retained. Gupta and Rickard's 2026 study found the identical pattern independently, decades later, in a different task: the spaced groups genuinely felt and performed better during training, and that advantage evaporated on a later test given under equal conditions. Twice, in two unrelated experiments, the thing that felt like learning and the thing that actually was learning came apart.

The "21 days" line isn't a rough estimate. It's wrong.

One more popular number, borrowed from a completely different field but routinely applied to practice habits, didn't survive contact with its own source either. The "21 days to build a habit" claim traces to a 1960s self-help book, not to research — and the actual study most often invoked to back it up, Phillippa Lally's 2010 research on habit formation, found a mean of 66 days to reach automaticity, with individual results ranging from 18 days to a projected 254, and roughly half of participants still hadn't locked in a stable habit after 84 days. More complex habits took meaningfully longer than simple ones — a second, unrelated line of evidence (alongside the spacing-effect complexity discount above) that complexity slows this kind of process down, from a completely different research tradition. If picking the guitar back up feels slower to become automatic than 21 days promised, that isn't failure. That's the actual data.

What actually held up when tested on someone like jtr

Only one study in this whole search targeted the exact population — an adult amateur picking up an instrument. Alexandra Ritchie and Gráinne Kearney's 2018 pilot gave adult beginner string players a structured self-regulation worksheet, the most obvious "be more deliberate about your practice" intervention available. It didn't move their self-efficacy, and the learners themselves didn't credit it with helping. That's a narrow result about one specific tool, not proof that structure never helps an adult beginner — but it's real, on-topic evidence, and it points toward humility rather than a tidy formula.

The honest bottom line

Time actually spent playing, engaged, is the one variable nobody in any of these disputes questions. Every camp in the deliberate-practice argument agrees on that much; they disagree only about how much extra a specific method adds on top of it. For a 25-to-40-minute session, the evidence argues for spending that time actually playing rather than architecting its internal structure around spacing or rest folklore that has now failed a direct instrument test. Consistency over months, not weeks, is the one piece of advice with real numbers behind it — Lally's 66-day mean, with complex behaviors running longer still. And the single most useful mental adjustment might be the least glamorous: don't judge a session by how it felt walking away from it.

What this doesn't offer is a confident, minute-by-minute practice template. Most of this research was built on children, conservatory-track musicians, or short single-session lab tasks with brand-new material — not a 25-to-40-minute return to an instrument someone hasn't played in years. Piano is not guitar; nobody has run Wiseheart's test on a fretted instrument. Whether sleep actually helps an adult guitarist consolidate a lick remains a genuinely open, actively argued question, not a settled "yes" or "no." The honest finding here isn't a formula. It's that some of the most confidently repeated advice about how to practice has now been tested directly and didn't survive — while the much simpler, less satisfying advice underneath it, just keep showing up and playing, has never actually been in question.