The Science Behind the Sculpt Credit Score

Did you do the sessions you planned?

Where we draw the line: 80% of planned sessions scores well. You do not need 100%.

Across exercise programmes that report attendance, average adherence is 80% of prescribed sessions (range 60–98%), and 80% is the cutoff trials conventionally use to define “high adherence.” Where outcomes have been split by attendance, the line falls near 75%.

Why this is graded Convention: 80% is the field’s working definition and its observed average. It is not a proven cliff, and no study has established one. We use it because it is the best-supported convention available, not because it is a finding.

Training more than planned earns nothing extra here. Extra work raises fatigue, and fatigue is priced by Recovery below.

Are you moving at the rate your goal actually needs?

Where we draw the line: Full credit from the pace your goal needs, up to 1.0% of bodyweight per week when losing and 0.5% when gaining.

This is the factor where going too fast counts against you, and it is the best-evidenced number on this page.

Losing weight. Garthe and colleagues randomised 24 elite athletes to lose either 0.7% or 1.4% of bodyweight per week while strength training four times a week. The slower group gained 2.1% lean body mass; the faster group gained none (−0.2%). Same weight off the scale, very different bodies underneath.

Gaining weight, where the ceiling is much tighter. In a separate Garthe trial, 39 elite athletes went through a gaining phase. The faster group put on 0.38% of bodyweight per week and the slower group 0.16%. Muscle gain was similar between them, but the faster group gained roughly five times the fat (1.1 kg vs 0.2 kg). Past a slow rate, extra weight is mostly fat. The published recommendation for lifters in a gaining phase is 0.25 to 0.5% of bodyweight a week, so 0.5% is our ceiling, and anything faster counts as too fast.

How it is scored. A slow, deliberate goal on track for its date earns full credit; you are not marked down for being gentle. Moving slower than your goal needs costs credit, and moving the wrong way earns none. Moving faster than the ceiling costs credit gradually, reaching zero at twice the ceiling. That falloff is our convention.

Strength goals. A goal to lift more is judged on that lift's estimated one-rep max, taken from your best set in each session. The pace it needs is capped at 1% of your one-rep max per week, so an ambitious date cannot ask for more than that. Unlike bodyweight, climbing faster than planned never costs credit. There is no published safe ceiling for getting stronger, so this rule is our convention. Endurance goals are not scored yet.

Pace reads the trend across your weigh-ins or your sessions of the lift, so each goal needs at least five over the last four weeks, spread across a week or more. Until then that goal is left out of your score rather than counted as zero.

Are you training inside what your body can absorb?

Where we draw the line: Starting a session on a muscle group you train directly, while it is below 33% freshness, costs you. Training often does not.

Every session leaves fatigue in the muscles it worked, and that fatigue fades. We model it as a decaying curve rather than a fixed rest period.

How long fatigue lasts depends on how hard you trained. Morán-Navarro and colleagues found that sets stopped well short of failure were recovered within about six hours, while sets taken to failure needed 48 to 72 hours. So after each exercise we ask how many more reps you had left, and fatigue from work taken to failure clears several times slower than fatigue from easy work.

Training a muscle group often is never penalised. When total weekly volume is the same, training a muscle more often grows and strengthens it at least as well as training it once a week. Only starting a session on a muscle that is already fatigued counts against you, and only for the muscles the session trains directly. Muscles that assist, like your core bracing during a squat, add half their share of fatigue but never make a session count as a miss on their own.

Why this is graded Moderate: the recovery times come from controlled trials, but the exact speeds we clear fatigue at, the 33% line and the half share for assisting muscles are our conventions. They are starting values, and we will calibrate them against how people actually perform in their next session. Research also shows recovery speeds up as you get fitter; we do not model that yet.

Are you eating for the goal you set?

Where we draw the line: Judged on weekly averages, not single days. Three good weeks out of four scores well.

We score weeks, not days, and there are two reasons.

The first is that a weekly average is what actually drives the outcome. Eating 1,800 one day and 2,800 the next averages out to exactly your target, and penalising that would be penalising nothing.

The second is measurement. Under FDA rules a pack can legally contain up to 20% more calories than its label says, and when checked against laboratory methods people under-report their calories by around 12 to 20% and their protein by 11 to 15%. Scoring a single day against a tight target would be scoring noise. Averaging a week shrinks that error instead of compounding it.

Calories are judged against the same target and the same on-track range your food screen shows, including any calories you add back for activity. The score is never stricter than your own food screen.

Why three weeks out of four, and not four. Conlin and colleagues randomised resistance-trained lifters to a flexible or a rigid diet through a ten-week deficit and found equivalent fat loss. Rigidity bought nothing. Separately, Westenhoefer and Stewart found that rigid dietary restraint tracks with more frequent and more severe binge eating and higher disinhibition. Demanding a perfect month would buy no better result and would reward the eating style associated with the worse relationship with food. We are not going to build that.

Protein. Protein is judged on your 7-day average in grams per kilogram of bodyweight. Across meta-analyses of people who lift, muscle and strength gains slow sharply between about 1.3 and 1.6 g/kg a day. So full credit starts at your target or 1.6 g/kg, whichever is lower (1.8 g/kg in a calorie deficit, where needs rise), and never below 1.3 g/kg. Because protein is under-reported, a week within 15% of that level still earns full credit. Eating more protein is never marked down: studies in healthy adults show no harm to kidney function, and extra calories already count in the calorie half.

We score calories and protein only, and they count equally toward each week. That split is our choice; no study sets it. Carbs and fat are what is left once those two are set, so scoring them separately would count the same thing twice and punish a trade-off that is entirely legitimate.

Is the work spread across what your plan needs?

Where we draw the line: 10–20 hard sets per muscle group per week.

Schoenfeld, Ogborn and Krieger pooled 34 treatment groups from 15 studies and found a graded relationship between weekly sets and muscle growth. Each additional set was worth roughly 0.37% more growth, with a threshold near 10 weekly sets per muscle group for close to maximal results.

Under 10 sets, a muscle group is below the range where it grows well. The upper bound of 20 to 25 sets, past which returns clearly diminish, is a practical convention rather than a meta-analytic number, and we treat it that way. Going over is flagged gently, not punished.

“Hard sets” means working sets you rate as two reps or fewer left in the tank. Easier sets do not count, or the window means nothing. That cut at two is our convention. An exercise that works several parts of one muscle group, like a row across your lats and upper back, counts once for that group.

A week here means the last seven days, not the calendar week, so Balance does not drop back to zero when a new week starts.

How long have you been consistent?

Where we draw the line: Six weeks to register, 66 days to matter, six months to max out.

Lally and colleagues followed 96 people forming new habits and found a median of 66 days to reach 95% of automaticity (the point where the behaviour stops taking conscious effort). The range was wide, 18 to 254 days, and complex behaviours like a full workout sat at the longer end, 60 to 150 days.

So our curve starts registering at six weeks, crosses its midpoint at 66 days, and saturates around six months. Past that, more time predicts nothing further.

The same study found that missing a single day barely matters. Automaticity dropped by less than half a point and recovered quickly. That finding shows up all over this score: one bad day is never treated as a setback.

This counts the days you actually logged a workout or a meal, across your whole time on Sculpt, not how long ago you signed up. It does not reset when you set a new goal, and a dormant account stops earning here. Like a thin credit file, it starts at zero, and at 5% of the score it can never outweigh real progress on your goal.