Repetition Without Repetition

Bernstein’s phrase: you do not repeat the same movement; you repeat the solution to a problem under slightly different conditions. That is how skill becomes robust instead of brittle.

Ladders and submission systems on this site are built that way. Armbar System is the same finish problem from four different starts. Each rung changes the organism–environment relationship; the invariant (isolate the arm, load the elbow) stays.

If an athlete only ever hits the same entry from the same start, they have memorised a script. If they can finish from mount, closed guard, and a broken posture, they have the problem.

The science. Nikolai Bernstein watched skilled workers repeat the same task and found that they never produced the same movement twice. He called it repetition without repetition: what gets repeated is the problem being solved, while the solution varies every time to fit the moment. The phrase carries most of what the ecological approach says about how to train.

The body has far more moving parts than any single movement needs. Bernstein’s question was how a person controls that redundancy — how the right configuration emerges from a near-infinite set. His answer was that coordination is the temporary, flexible mastering of those degrees of freedom for the task at hand, re-solved each time conditions shift. A grappler finishing a strangle adjusts grip, angle, and pressure to the partner in front of them; the next finish, against a different reaction, recruits a different configuration toward the same end.

Many different configurations can produce the same functional outcome — a property the research calls degeneracy. A back strangle finishes with the forearm deep, with a short choke, with a switch to a different grip; all compress the carotids. This is not noise to be trained out. It is the source of adaptability: a grappler with one rigid solution fails when the opponent removes it, while a grappler who owns the outcome finds another route. It is also why the positional games phrase their goals as outcomes — force the tap by compressing the carotids, not “hit the rear naked choke” — and why the invariants are stated as mechanical truths rather than techniques.

Drilling one identical motion many times grooves a single solution. It looks like progress in the room, where the partner feeds the same reaction, and it fails on contact with a resisting opponent who varies. Practice that varies the conditions — different starts, different counters, a live partner — trains the process of solving, which is the thing that generalises. The cost is that varied practice looks worse during the session and better afterward, which is exactly the trade representative practice asks a coach to accept.

This is not an argument against ever repeating. Grooving a shape a few times to get a foothold, or rehearsing a safety skill until it is automatic, has a place — drilling versus games marks where. The claim is narrower: the bulk of skill is variable problem-solving, so the bulk of practice should be too. The constraints-led approach is how a coach builds that variability on purpose rather than leaving it to chance.

Bernstein, N. A. (1967). The Co-ordination and Regulation of Movements. The degrees-of-freedom problem and “repetition without repetition.”

Edelman, G. M., & Gally, J. A. (2001). “Degeneracy and complexity in biological systems.” The biological basis of many-routes-to-one-outcome.

Davids, K., Button, C., & Bennett, S. (2008). Dynamics of Skill Acquisition: A Constraints-Led Approach; Gray, R. (2021). How We Learn to Move.

New to how the site trains? The method hub maps the whole approach.