How People Actually Learn Skills

Researchers have been studying how people learn movement skills for the better part of a hundred years. Most of what gets passed around gyms hasn't caught up.

There are journals full of this research, decades of experiments, whole theories that have risen and been set aside. And yet most of what we hear — “just get more reps,” “groove the perfect form,” “it's muscle memory” — is either a watered-down version of a model the field has largely moved past, or it ignores the research altogether.

I don't say that to talk down to anyone. For most of my career I coached on instinct and inherited wisdom too. But when I went back to formally study skill acquisition, the thing that struck me wasn't that coaches are wrong — it's that we're often working from an old map while the territory has been redrawn. This is my attempt to hand you a current one, without the jargon wall that usually keeps this locked inside academia.

Three families of ideas

Broadly, the science of skill learning has organized itself into three families: Information Processing, Cognitive models, and Ecological models. These aren't just labels for academics to argue over — each one quietly tells you how to run your practice.

Information Processing: the brain as a computer

The oldest of the three treats the brain like a computer: information comes in through the senses, gets processed against a stored “motor program,” and produces a movement. Learning, in this view, is building and refining that internal program — the landmark being Schmidt's schema theory (Schmidt, 1975).

Its roots matter, though: it grew out of studying simple, repeatable laboratory tasks — reaction-time tests, lever pulls — where one stable, correct response makes sense. The trouble came when we exported that model wholesale into sport, where no two shots are ever identical and the environment refuses to hold still. If the goal is a single stored program run the same way every time, then a defender, a bad pass, or tired legs in the fourth quarter all become “noise” to eliminate — instead of the very thing the skill exists to handle. That mismatch is a big reason the field has moved away from pure Information Processing as an explanation for open, dynamic skills like shooting in a game.

Cognitive models: still very much alive

The broader cognitive family kept the mind at the center but got less rigidly mechanical. Fitts & Posner's stages of learning (Fitts & Posner, 1967) — cognitive → associative → autonomous — is probably the most useful idea most coaches have never heard named: a beginner has to think their way through every piece; with practice, attention frees up; eventually it runs largely on its own.

I want to be clear here, because this is where people overreach: cognitive approaches are not “debunked” or dead. There's excellent, current research in this tradition, and some of the sharpest minds in the field lean this way. Attention, memory, decision-making — they genuinely matter. Anyone who tells you the cognitive side has been disproven is selling you something.

A great example of the cognitive tradition's practical value is the Challenge Point Framework (Guadagnoli & Lee, 2004). The core idea: learning is optimized when the difficulty of the task is matched to the learner — hard enough to demand real problem-solving, but not so hard that they can't process what's happening. Too easy and there's nothing to learn; too hard and it's just noise. Every good coach feels this intuitively; the framework gives it a name and a logic, and it's one of the most useful ideas you can carry into any practice — no matter which camp you lean toward.

Ecological models: where the momentum has shifted

The newer wave — and where the research momentum has clearly moved — flips the starting point. Instead of beginning inside the head with a program, ecological models begin with the relationship between the athlete and their environment. They trace back to the perceptual psychologist James Gibson (Gibson, 1979) and the movement scientist Nikolai Bernstein (Bernstein, 1967).

The core idea: perception and action are coupled — we perceive in order to act, and act in order to perceive — and skilled movement is less a stored recipe than the body self-organizing a solution to the specific problem in front of it. Newell's constraints model (Newell, 1986) frames every movement as emerging from three interacting constraints: the performer, the task, and the environment. Change any one and the movement changes. That's a fundamentally different premise than “run the program.”

And to be fair to the whole field: this is a shift in momentum, not a knockout. More researchers and coaches move toward ecological models every year, but serious work continues across all three traditions. The honest position isn't tribal — it's knowing what each family gets right.

A closer look at the ecological toolkit

Under the ecological umbrella sit a few approaches you'll hear named. Each is a tool, with strengths and blind spots:

• Constraints-Led Approach (CLA) — (Davids, Button & Bennett, 2008). Manipulate constraints — rules, space, equipment, a single instruction — so the athlete is guided to discover a functional solution rather than being handed the “correct” technique. Strength: builds adaptable, game-transferable skill; the environment does the teaching. Hole: with a true beginner, or a habit so ingrained it needs isolating, pure discovery can be slow — or quietly let an inefficient pattern survive.
• Differential Learning — (Schöllhorn and colleagues). Deliberately inject variability, even “errors,” and avoid repeating the same movement twice, so the system explores a wide landscape of solutions. Strength: robustness; it fights rigid over-grooving. Hole: it can feel chaotic, and dose and timing matter — it's rarely where you start.
• Nonlinear Pedagogy — (Chow, Davids, Renshaw & Button, 2016). The teaching framework that ties much of this together: representative tasks, manipulating constraints, encouraging exploration, and being comfortable with the messy, non-straight-line path real learning takes.
• Representative Learning Design — (Pinder, Davids, Renshaw & Araújo, 2011). Practice should represent the performance environment: the information available in the drill should match the information available in the game, so what you build actually transfers.

BernsteinHammer.jpeg

Principles over camps

Here's the part I care about most. A high-level coach's job isn't to pick a camp and wave the flag. It's to understand the principles deeply enough to use the tools well — and to know each tool's strengths, weaknesses, and blind spots. The CLA is powerful and it has holes. Differential learning is useful and not for every moment. Even the old motor-program idea captures something real about early, isolated skill-building. Dogma is the enemy. The best coaches I know are relentlessly practical: principle-driven, tool-agnostic, matching the method to the athlete and the moment.

It's why a cognitive-rooted idea like the Challenge Point sits comfortably in an ecological coach's practice — good principles don't care which camp discovered them.

A quick glossary

A few terms worth knowing, in plain language:

• Affordances — the action possibilities the environment offers a specific athlete (a lane to attack, a closeout to shoot over).
• Constraints — the boundaries that shape a movement: performer, task, environment.
• Degrees of freedom — the many independently moving body parts that must be coordinated (Bernstein's problem). Beginners often “freeze” some; skill “frees” them.
• Self-organization — functional movement emerging from those interacting constraints, without a step-by-step internal command.
• Degeneracy — many different movement solutions can reach the same outcome. It's why there is no single “perfect” form.
• Perception–action coupling — perceiving and moving as one continuous loop, not separate steps.
• Representativeness — how faithfully a drill preserves the real information and demands of the game.

Bringing it together

You don't need a PhD to coach well. But you should know which map you're using — because the map decides how you run every drill, what you call a mistake, and what you think “improvement” even means.

The shift toward ecological models isn't a trend chasing novelty; it's the science slowly catching up to what the best coaches have always sensed: skill lives in the relationship between a player and the game, not in a frozen ideal of form.

Understand the principles. Respect the tools. Stay humble about the holes. And match the method to the shooter in front of you. That's the whole job.

References

The foundational works behind this piece:

1. Bernstein, N. (1967). The Co-ordination and Regulation of Movements. Pergamon Press.
2. Chow, J. Y., Davids, K., Renshaw, I., & Button, C. (2016). Nonlinear Pedagogy in Skill Acquisition: An Introduction. Routledge.
3. Davids, K., Button, C., & Bennett, S. (2008). Dynamics of Skill Acquisition: A Constraints-Led Approach. Human Kinetics.
4. Fitts, P. M., & Posner, M. I. (1967). Human Performance. Brooks/Cole.
5. Gibson, J. J. (1979). The Ecological Approach to Visual Perception. Houghton Mifflin.
6. Guadagnoli, M. A., & Lee, T. D. (2004). Challenge point: A framework for conceptualizing the effects of various practice conditions in motor learning. Journal of Motor Behavior, 36(2), 212–224.
7. Newell, K. M. (1986). Constraints on the development of coordination. In M. G. Wade & H. T. A. Whiting (Eds.), Motor Development in Children: Aspects of Coordination and Control.
8. Pinder, R. A., Davids, K., Renshaw, I., & Araújo, D. (2011). Representative learning design and functionality of research and practice in sport. Journal of Sport & Exercise Psychology, 33(1), 146–155.
9. Schmidt, R. A. (1975). A schema theory of discrete motor skill learning. Psychological Review, 82(4), 225–260.
10. Schöllhorn, W. I. (2000). Applications of systems dynamic principles to technique and strength training. Acta Academiae Olympiquae Estoniae, 8, 67–85.