Deliberate Practice: What It Is and How Students Use It

Deliberate practice is practice designed to improve one specific part of your performance: you work just beyond what you can currently do, get immediate feedback, and repeat with corrections. The term comes from psychologist K. Anders Ericsson and colleagues, whose 1993 study of violinists became one of the most cited papers on expertise.

For students, deliberate practice is the difference between doing another full practice exam and spending that hour on the three question types you keep getting wrong. It works. But the research since 1993 also shows it explains less of the gap between people than the original paper claimed, and the famous "10,000-hour rule" isn't what Ericsson found.

This guide covers what deliberate practice is, what the original study and its replication found, how much practice really matters, and a practical way to use it in your own studying.

What is deliberate practice?

In their 1993 paper, Ericsson, Ralf Krampe, and Clemens Tesch-Römer defined deliberate practice as activities specially designed to improve the current level of performance. Drawing on a century of learning research, they argued that practice improves performance most when you:

• attend to the task and make an effort to improve,
• get immediate, informative feedback on how you did,
• repeat the same or similar tasks, correcting errors each time,
• and ideally work with a teacher who diagnoses your errors and chooses the next task.

Ericsson and Robert Pool later used three labels in their 2016 book Peak that make the distinction clear:

Most students can't get a coach for every subject, so in practice you'll be doing purposeful practice, borrowing feedback from answer keys, mark schemes, classmates, and teachers' office hours. That's fine. The ingredients that matter most are the specific target, the feedback, and the corrected repetition.

The 1993 violin study

Ericsson's team recruited violin students at the Music Academy of West Berlin in three groups of ten: the "best" (nominated by faculty as potential international soloists), "good" violinists from the same department, and students training to be music teachers. Each estimated how many hours a week they had practiced alone at every age since starting.

By age 18, the best violinists reported about 7,400 hours of practice alone, the good violinists about 5,300, and the future teachers about 3,400. The authors concluded there was "complete correspondence" between skill level and accumulated practice, and that individual differences in ultimate performance could largely be accounted for by practice.

That conclusion went far beyond music. It became the scientific basis for popular books arguing that talent is overrated and that enough of the right practice can make anyone an expert.

Did the violin study replicate?

In 2019, Brooke Macnamara and Megha Maitra repeated the study with 39 violinists at the Cleveland Institute of Music and an affiliated university. They preregistered their plan and kept the interviewers blind to which group each violinist was in.

Practice still mattered: both top groups had practiced far more than the less accomplished group. But the best violinists had not practiced more than the good ones; the difference ran slightly the other way and wasn't statistically significant. In the replication, practice alone explained 26% of the variance in performance. The comparable figure from the original data, with the best and good groups combined against the future teachers, was 48%.

As the replication's authors put it, 26% is not an inconsequential amount. It just doesn't support the claim that practice largely accounts for who becomes an expert.

How much does deliberate practice explain?

In 2014, Macnamara, David Hambrick, and Frederick Oswald pooled 88 studies that measured both practice and performance. After a 2018 correction, deliberate practice explained 14% of the differences in performance overall.

The same team's 2016 meta-analysis of sports, summarized in their 2020 review, found that among elite athletes (national level and above), practice explained only about 1% of the differences between them. Part of the reason is that elite athletes all practice a lot, so their hours vary less.

Ericsson disputed these numbers. In a 2019 reply, he and Kyle Harwell argued that the meta-analysis counted many kinds of "structured practice" that didn't meet his original definition. Re-analyzing only the studies they judged to qualify, they estimated that practice explained 29% of performance differences, or 61% after correcting for measurement error. Hambrick and colleagues answered in 2020 that the definition of deliberate practice had shifted over the years, which makes the claim hard to test.

The fair summary for a student:

• Practice clearly helps. Across these studies, more practice generally goes with better performance.
• It's not the whole story. Other factors, such as prior knowledge, general ability, personality, and the quality of teaching, matter too.
• Hours alone are a weak predictor, especially in education. What you do in the hours is the part you control.

Is the 10,000-hour rule true?

No. Malcolm Gladwell popularized the "10,000-hour rule" in his 2008 book Outliers, citing Ericsson's violin study, where the best group had accumulated around 10,000 hours of practice alone by about age 20.

Ericsson rejected the rule. In his 2019 paper with Harwell, he wrote that although the research showed long periods of training were needed for international-level performance, there was no evidence for a magic number, and he repeated his earlier estimate that winning international piano competitions takes closer to 25,000 hours. He also pointed out that Gladwell counted all kinds of practice, including public performances and work, not deliberate practice specifically.

The replication adds a twist: by age 20, both the best and the good violinists had averaged more than 10,000 hours of practice alone. Crossing the line didn't separate them.

How students can use deliberate practice

Translate the four ingredients into study habits.

1. Pick one specific weakness

"Get better at calculus" is not a practice target. "Stop losing marks on chain rule problems with trig functions" is. The best source of targets is your own mistakes. Keep an error log (the method is covered in mistake-driven learning) and look for patterns: the same question type, the same misconception, the same careless step.

2. Work just beyond your current level

Problems you always get right teach you nothing new. Problems far beyond you teach you frustration. Aim for tasks you get right with effort, roughly the band psychologists call the zone of proximal development.

3. Get feedback fast

Check each answer before you start the next one, so a mistake doesn't repeat ten times. Sources of feedback:

• worked solutions and mark schemes from past papers
• a teacher, TA, or tutor reviewing one piece of work
• a classmate marking your essay plan against the rubric
• running your code, or checking your answer by substituting back

4. Repeat with corrections, then again later

Redo every missed problem without looking at the solution. Then redo it two or three days later. Getting it right just after seeing the answer is short-term memory; getting it right next week is learning. Space these repeats as described in distributed practice vs massed practice, and mix question types once you're past the basics, as in interleaving.

It won't feel like normal studying

In a 2011 study of competitors in the Scripps National Spelling Bee, solitary study of words and their origins predicted how far spellers got better than being quizzed by others or reading for pleasure. Spellers also rated it the most effortful and least enjoyable of the three. That matches most students' experience: working on weaknesses is uncomfortable, a pattern researchers call desirable difficulties.

Ericsson's 1993 paper argued that even experts can sustain only a few hours a day of this kind of practice, around four, which is a good reason to keep sessions short and focused.

Deliberate practice examples: a 45-minute session

Here's a 45-minute template you can run today, shown for a calculus student.

Swap in your subject:

• Essays: one paragraph at a time, judged against the rubric's "analysis" criterion, rewritten after feedback.
• Languages: one grammar pattern you keep getting wrong, 20 sentences, each checked immediately.
• Medical exams: one system where your question bank accuracy is lowest, untimed blocks with every explanation read.
• Coding: one type of bug you keep writing, small exercises with tests that fail until you fix it. Our guide to learning programming with active recall builds a weekly routine around this.