Data & method

Muscle attributions

Every exercise names one or more prime movers and a list of synergists and stabilisers. The prime mover is the muscle producing the joint action against the resistance; a synergist assists that action; a stabiliser holds a joint still so the action can happen. These follow standard kinesiology — the muscle is assigned from the joint action the exercise demands, not from where people report feeling it.

That distinction is the reason this site exists. "Works the chest" is not programming information. "The pectoralis major is the prime mover in shoulder horizontal adduction, with the anterior deltoid and triceps assisting" tells you why a close-grip press fatigues the triceps first, and what you can swap it for.

Gym folklore is deliberately absent. There is no inner or outer chest, no lower biceps, and no upper and lower abs that can be trained separately — the rectus abdominis is one muscle with one nerve supply. Grip width and bench angle shift emphasis between the heads of a muscle that has heads; they do not isolate regions of a muscle that does not.

Substitutes and progressions are computed, not written

The "if you cannot do this one" list on every exercise page is generated at build time from the movement pattern and the muscle data: candidates must share the pattern and at least one prime mover, and are ranked by how much of the full muscle profile they share and how close their difficulty rating is. The easier and harder progressions use the same rule with the difficulty rating pointing down or up.

The advantage of computing them is that they cannot contradict the page they sit on, and they cannot link to something that does not exist. The disadvantage is that a human coach would occasionally pick differently. Where the algorithm's pick is genuinely poor, the fix is better muscle data, not a hand-written exception.

Difficulty ratings

Every exercise carries a 1 to 5 rating for technical demand — how much practice and awareness it takes to do safely — not for how hard the set feels. A wall sit is exhausting and rates 1. A snatch-grip deadlift rates 4 whether the bar is loaded or empty.

MET values and calorie figures

The 2024 Adult Compendium of Physical Activities is the standard reference for the energy cost of physical activity, and it does not publish a MET value for individual gym exercises. It publishes values for categories: vigorous resistance training, moderate resistance training, squats, vigorous calisthenics, circuit training and so on.

So every exercise here is mapped to the closest published category, and the category is named in full on the page next to the number. That is honest in a way that "the barbell curl burns 4.2 METs" is not, because no such measurement exists. The arithmetic is the standard ACSM relationship: kcal per minute = MET × 3.5 × body mass in kg ÷ 200, shown for a 70 kg adult. The full activity dataset is published on a sister site, BurnLab.

One more caveat that matters: the figures are per minute of work, not per minute in the gym. A set of eight reps takes about twenty-five seconds. An hour-long session might contain twelve minutes of actual working time.

Rep and set guidance

The sets, reps and rest table on each page is derived from the movement's mechanics — compound, isolation, isometric or explosive — and follows the general consensus of resistance-training research: heavy loads with long rests for maximal strength, moderate loads and moderate rep ranges for hypertrophy, lighter loads and shorter rests for muscular endurance. These are starting points for a healthy adult, not individual prescriptions.

Illustrations

The figures on this site are simple line drawings of movement patterns, drawn as SVG code for this site. No photographs, GIFs or diagrams are taken from any other exercise database, and none are hotlinked from anywhere. Where an illustration would not add anything true, there is text instead.

What this site will not tell you

• That any exercise burns fat from the area it trains. Spot reduction has been tested directly and does not happen. Crunches build abdominal muscle; they do not remove abdominal fat.
• That "toning" is a separate process. There is muscle growth, there is fat loss, and there is strength. What people call toning is some combination of the three.
• That any exercise is dangerous or forbidden. Load, range and progression decide that far more than exercise selection does. Where a position genuinely deserves caution, the page says so and explains why.
• That one exercise is "the best". The best exercise for a muscle is usually the one you can load progressively, recover from and repeat for months.

Corrections

If a muscle attribution here is wrong, that is worth fixing and worth fixing quickly. Send the page URL and the correction.

General information only — not medical or coaching advice. Load, form and progression should be adjusted to your own experience and any injury history. When in doubt, get coached in person.