What doesn't workand why it's still taught.
Sixteen ideas that circulate as though they were science. Each with the source that refutes it. No site that lives off its traffic can publish this list: these are its own pages.
How to read this list
Not everything here is refuted the same way, and the difference matters.
- Measured: studies put it to the test and it didn't hold. The strongest case.
- No evidence: it never entered the literature. That isn't the same as “refuted” — nobody ever measured it, and the burden of proof sits with whoever claims it.
- Misapplied: the phenomenon is real, but the practical conclusion hung on it doesn't follow.
Presenting the second case as though it were the first would repeat the very error this list calls out.
- No evidence
Somatotypes: ectomorph, mesomorph, endomorph
Your body type determines how you should train and eat.
The system comes from 1940s constitutional psychology, which held that body shape predicts temperament — same family as phrenology, with an explicit eugenic root. The modern anthropometric somatotype is a real descriptive measure, but it describes the shape you have today: it doesn't predict how you'll respond to a diet or a programme. The leap from “you're an endomorph” to “so train like this” was never validated.
- Measured
Your maximum muscular potential
Your height and wrist circumference determine how much muscle you can ever carry.
The formulas come from the measurements of some 300 bodybuilding champions — the tail of the distribution — published in a self-published book. The famous “natural limit” of 25 fat-free mass index comes from a 1995 study of steroid users and non-users, and it's falsified: between 21 and 26 % of college football players exceed it. Telling someone their ceiling is 78 kg of lean mass doesn't inform them: it hands them a causal excuse for a plateau that almost always comes from the programme, the sleep or the calories.
- Measured
Your body fat percentage from a tape measure
Measuring your waist and neck gives your body fat percentage.
The Navy method has a standard error of 3 to 4 points, and against DXA it underestimates in men and overestimates in women. For an individual, the 95 % limits of agreement are around ±8 points: the number could be off by the width of an entire category. And worse for what you'd actually want it for: circumference measures were shown not to capture body composition change across 8 weeks of training. It works neither for the value nor for the trend. What does work is charting raw circumferences and watching where they go.
- No evidence
Ideal measurements
There are ideal arm, chest and waist proportions for your frame.
The coefficients come from the measurements of one 1950s bodybuilder and from a Greek sculptural convention. They never entered the scientific literature, so there's nothing to refute: there was never a claim to test. What they do produce is a target shaped like a goal and filled with dissatisfaction.
- Measured
Your max heart rate is 220 minus your age
Subtract your age from 220 and that's your max; training zones follow from it.
That equation didn't come from research. Its origin was traced: a 1971 observation over about eleven mixed references, with no formal regression, propagated by textbooks without citation. Its inherent error is 7 to 11 beats. The best available is 208 minus 0.7 times age, from a meta-analysis of 351 studies and 18,712 people. But even that carries a residual deviation of around 10 beats — the width of a whole zone. For a strength app, offering no zones is more honest than offering them.
- Measured
The calories you burned lifting
Your watch knows how many calories you spent in a lifting session.
It was measured against indirect calorimetry during resistance exercise: correlations run from 0.10 to 0.34, with concordance coefficients below 0.45. Apple, Galaxy and Garmin all underestimate. And there's a physiological reason: in lifting, heart rate rises from intrathoracic pressure and localised activation, not from proportional oxygen demand — so every heart-rate-based model is estimating the wrong variable.
- Misapplied
The fat-burning zone
Training at low intensity burns more fat.
There is an intensity at which the rate of fat oxidation peaks, around 60-65 % of maximal oxygen uptake; that part is real. What doesn't follow is the conclusion: what determines fat loss is cumulative energy balance, not which substrate you oxidised during that hour. A real phenomenon with a recommendation that isn't deduced from it.
- Measured
Your fibre type from how many reps you can do
If you do more than 20 reps at 70 % you're slow-twitch and should train differently.
Here the verdict isn't “no evidence”: there's weak group-level signal and nothing at the individual level. With actual biopsy, the correlation between fibre proportion and reps to failure came in below 0.2 — under 4 % of variance explained. What did correlate strongly was capillary density: the test measures capillarisation, not fibre type. Between the extreme groups there are thirteen points of difference, which makes it useless for classifying a person.
- Measured
Your metabolic age
The scale says your body is 34 even though you're 45.
Each manufacturer's algorithm is proprietary and unpublished, which makes it literally unfalsifiable. The only peer-reviewed work characterising it found it redundant: it tracks chronological age and the fat and muscle mass the scale already measured, and not what you'd expect it to predict. It's a repackaging with a layer of alarm.
- Measured
8 to 12 reps for size, 1 to 5 for strength
Each rep range produces a different adaptation.
A meta-analysis of 21 studies found hypertrophy occurs equally across the load spectrum when sets are taken near failure; the advantage of heavy load is specific to maximal strength. A within-subject study with volume equated measured growth of 20.5 %, 20.4 % and 19.5 % for 40, 60 and 80 % of max — indistinguishable from each other; only 20 % fell short. The variable that governs is proximity to failure, not the rep count.
- Measured
Confusing the muscle by switching exercises
You have to vary exercises often so the muscle doesn't adapt.
It was tested with 19 trained men over 8 weeks, comparing randomised against fixed exercises with sets and proximity to failure equated: no differences in muscle thickness, body composition or maximal strength, and the varied group came out slightly worse at the rectus femoris. The only real effect was higher motivation — an adherence argument, not a physiological one.
- Measured
The afterburn effect
After training you keep burning calories for hours.
It was measured in a review of 16 studies: the extra expenditure over the hour after training runs from 14 to 19 calories. Compared isocalorically, it's about 27 calories after lifting against 40 after cardio. A 20-calorie afterburn is half a banana.
- Measured
Exercises to lose belly fat
Working a region burns the fat over it.
Six weeks of abdominal work measured with DXA: no reduction in abdominal or total fat. And a design using the contralateral leg as its own control over 12 weeks: no change in local fat mass. Fat is mobilised systemically, not from the muscle underneath.
- No evidence
Your MEV, MAV and MRV
Each person has a minimum effective and a maximum recoverable volume in weekly sets.
In eleven years not a single peer-reviewed study has measured those quantities in a person. It's not that the evidence is weak: there isn't any. And there's experimental evidence against the ceiling: 31 trained people were taken to around 52 weekly quad sets for 12 weeks and the collapse the framework predicts didn't happen. What does exist is a measured dose-response curve, with diminishing returns and no identifiable ceiling.
- Measured
Resting 30 to 60 seconds for hypertrophy
Short rests maximise growth through the hormonal response.
It's the most repeated obsolete advice in the field. The chain of evidence that dismantled it ended up measuring the opposite: 3-minute rests produced more hypertrophy and more strength than 1-minute rests. The acute hormonal response it rested on turned out not to predict growth.
- No evidence
INOL and Prilepin's table
There are indices that tell you whether your weekly load is optimal.
They were searched for in the biomedical literature databases and there's nothing: zero studies. INOL comes from a self-published 2005 ebook and its bands are the author's own. Prilepin's table is cited as though it came from Soviet research on thousands of lifters, but the original source isn't verifiable, and the only peer-reviewed work touching it is a conference abstract with nine participants and no control group. They're folklore shaped like an equation.
Why we publish this
Because we can. Myosin doesn't sell supplements, doesn't run ads and doesn't bill against the traffic of a somatotype calculator.
A site that lives off its traffic can't publish this list, because every entry is one of its own pages. That asymmetry is the one advantage budget doesn't buy.
