What the data actually shows
Ericsson's research on expert performers introduced the idea of deliberate practice: structured, goal-directed practice aimed just beyond current ability, with feedback and repetition on the specific things you cannot yet do. He repeatedly emphasised that this kind of practice is effortful and not inherently enjoyable, which is part of why most people do not sustain large amounts of it. The effort is the point; the discomfort is a by-product of working at your edge rather than a target in itself.
A related body of work by Robert Bjork on what he called "desirable difficulties" found that certain conditions which make practice feel harder in the moment — spacing it out, mixing topics, testing yourself, struggling to retrieve an answer — tend to produce stronger long-term learning than easier, smoother practice that feels more fluent. The discomfort there is informative: it reflects the brain doing the work that makes memories durable. But Bjork's qualifier is built into the name. The difficulties have to be desirable — appropriately matched to the learner — to help.
The same line of research is clear that difficulty is not automatically good. A difficulty that is too far beyond your current ability, or that produces confusion rather than productive struggle, tends to undermine learning rather than strengthen it. So the data does not support a simple "more pain, more gain" rule. It supports something narrower: the right kind of difficulty, at the right level, tends to help.
Why this feels different from how it actually is
The belief that practice has to hurt is reinforced by a powerful cultural story — the lone prodigy grinding in misery, the "no pain, no gain" slogan from athletics, the montage of suffering before triumph. These narratives are memorable and emotionally satisfying, and they quietly equate the discomfort with the cause of the improvement, when the research points to focused effort as the cause and discomfort as a frequent side effect.
It also feels true because the easy version of practice so visibly does not work. Going through the motions, repeating what you can already do, practising on autopilot — these feel comfortable and produce little progress. So it is natural to conclude that the comfort is the problem and that maximising discomfort must be the solution. The more accurate reading is that the comfortable version fails because it lacks effort and challenge, not because it lacks suffering.
And there is a confound: people who improve a lot often did tolerate a lot of discomfort, so the two travel together in the stories we hear. What is harder to see is the large number of people who suffered through punishing, poorly designed practice and improved little, or who burned out and quit. Survivorship makes pain look more necessary than the evidence says it is.
The goal is the effort, not the pain.
What the research says to do about it
Aim for the edge of your ability, not the edge of your endurance. The most consistent finding is that practice helps most when it targets specific things you cannot yet do reliably, with clear goals and prompt feedback. That is where the productive effort and the useful discomfort live — in working on the hard parts deliberately, rather than in making the whole session as gruelling as possible.
Build in the research-backed difficulties that feel harder but help: space practice out over time rather than cramming it, test yourself instead of only re-reviewing, and mix related skills rather than drilling one in isolation. These tend to feel less smooth and more effortful, and that very feeling is part of why they work — the struggle to retrieve and apply is doing the consolidation.
Protect sustainability, because the limiting factor for most learners is not intensity but whether they keep going. Research on deliberate practice notes that experts manage it in focused blocks with real recovery, not in endless punishing marathons. Effort you can repeat tomorrow beats heroic suffering you avoid for a week afterward.
What the research says does not help
Equating pain with progress does not help, and can actively mislead. Maximising how bad a session feels — longer, harder, more punishing — is not supported by the research as a route to faster improvement, and it raises the risk of burnout, injury in physical domains, and simply quitting. The evidence rewards the right kind of effort, not the maximum amount of suffering.
Mindless repetition of what you already do well is comfortable but largely wasted, which is the genuine kernel of truth behind the "it should be hard" instinct. The problem with easy practice is the absence of challenge and feedback, though — not the absence of misery. Adding difficulty that targets your weak points helps; adding difficulty that just makes you feel worse does not.
Pushing through difficulty that is far beyond your current level tends to backfire, producing confusion and discouragement rather than learning. The research on desirable difficulties is explicit that the difficulty has to be appropriately matched; struggle that is unproductive is just struggle. White-knuckling through material you are not ready for is not a shortcut.
Effort you can repeat tomorrow beats heroic suffering you avoid for a week afterward.
What this looks like in real life
The edge of your ability, not the edge of your endurance
Productive practice targets the specific things you can't yet do reliably, with clear goals and prompt feedback — that's where the useful discomfort lives. Making the whole session as gruelling as possible is a different thing entirely, and the research doesn't reward it. The effort is the active ingredient; the pain is a by-product of working at your edge, not a target.
The survivorship story we don't see
The people who improved a lot often did tolerate a lot of discomfort, so pain and progress travel together in the stories we hear. Harder to see are the many who suffered through punishing, poorly designed practice and improved little, or who burned out and quit. Survivorship makes pain look more necessary than the evidence says it is.
Real numbers in context
There is no clean number for how much practice "hurts," and any precise figure here should be treated with caution. The often-quoted "10,000 hours" popularisation of Ericsson's research is a notable example of how these findings get flattened: Ericsson himself pushed back on it, noting that the amount of practice needed varies enormously by domain and individual, that quality matters more than a fixed hour count, and that the figure was never meant as a universal threshold. The honest takeaway is about the nature of the practice, not a magic quantity.
What the research does support is a consistent qualitative pattern across studies of skill learning and memory: practice that is effortful, focused on weaknesses, spaced over time, and accompanied by feedback tends to outperform practice that is smooth, comfortable, and massed — even though the harder version usually feels less productive while you are doing it. The discomfort is often a sign the practice is working, but it is the effort and structure doing the work, not the pain.