How TrainForge calculates your calories

2026-05-22 · 6 min read

A plain-English walkthrough of how we turn your weight, activity, and goal into a daily calorie target — and why the number adapts every week once you've logged data.

Every calorie calculator on the internet asks for your weight, height, age, and activity level, then spits out a number. Most of them are wrong — not because the formula is broken, but because nobody actually burns calories at the rate a formula predicts. Two people with identical stats can have maintenance calories that differ by 400+. TrainForge handles this in two stages: an estimate when you first sign up, then a measured target once we have real data to work with.

Stage 1: the initial estimate

When you finish onboarding, we don't have any food logs or weight history yet, so we use a standard activity-multiplier method:

Maintenance calories ≈ body weight (lbs) × activity multiplier
Sedentary = 13 · Light = 14 · Moderate = 15 · Active = 16 · Very active = 17

A 180 lb member who selected "moderate" activity gets an initial maintenance estimate of 2,700 cal/day (180 × 15). This is rough, but it's good enough to start with — better than waiting two weeks for data before giving you any target at all.

Stage 2: goal adjustment

Maintenance is the number that keeps your weight stable. To change body composition, we add or subtract calories based on your goal. The amounts aren't arbitrary — they come from one rule:

3,500 kcal ≈ 1 lb of body fat. Spread over 7 days, a 500/day deficit creates roughly 1 lb of fat loss per week.

Here's how each goal translates that into a daily target:

Lose fat: −500/day → ~1 lb/week loss. Big enough to see results, small enough to preserve muscle if you train hard and eat enough protein.
Build muscle: +300/day on training days, near maintenance on rest days → ~0.5 lb/week gain. Lean muscle gain is slower than fat loss; bigger surpluses just add more fat.
Get stronger: Near maintenance with a small training-day boost. Focus is on lifting heavier, not changing the scale.
Recomp: +250 on training days, −200 on rest days. Net is a slight weekly deficit while training-day surplus fuels muscle. The hardest goal — see why recomp is hard.
Maintain: Maintenance every day. Useful between cuts and bulks.

Protein is calculated separately as grams per pound of body weight: 1.2 g/lb for fat loss (high — protects muscle in a deficit), 1.1 g/lb for recomp, 1.0 g/lb for muscle/strength, 0.8 g/lb for general health.

Stage 3: measured TDEE

After you've logged food for 5+ days and weighed in at least twice, we switch from the estimate to a measured TDEE (Total Daily Energy Expenditure). This is the part most apps don't do, and it's why your real maintenance can be 400 cal off from any formula's prediction.

Measured TDEE = average daily intake − (weekly weight change × 3,500 / 7)

If you ate 2,500 cal/day on average and lost 1 lb that week, your real TDEE is about 3,000 cal/day (you burned 500/day more than you ate). If you ate 2,500 and stayed the same weight, your maintenance is 2,500 — much lower than the 2,700 estimate. We then re-apply your goal adjustment to the measured number, so your target gets more accurate over time.

Confidence levels

Your TDEE estimate has a confidence rating you'll see on the Nutrition page:

Low: <5 days logged or <2 weigh-ins. Still on the formula estimate.
Medium: 5+ days, 2+ weigh-ins. First real measurement.
High: 10+ days, 3+ weigh-ins. Confident enough to make weekly adjustments.

Why we don't use Mifflin-St Jeor

The standard formula (Mifflin-St Jeor) uses height, age, sex, and body weight to predict basal metabolic rate, then multiplies by an activity factor. It's the most accurate formula on paper, but the predictions still vary ±15% from real measurements, and the activity multipliers are just educated guesses about how much you move.

We use a simpler weight-times-multiplier as the starting point because once you've logged for a week, the measured TDEE replaces the estimate anyway. A more elaborate formula upfront gives you a slightly better starting guess but no advantage long-term.

What to do with this

Log food consistently for the first two weeks — that's what unlocks the adaptive system.
Weigh yourself 3–4 mornings a week. We use the average, not single days (single weigh-ins swing ±3 lbs from water).
Don't manually override your target chasing a faster result. Bigger deficits don't mean faster fat loss; they mean more muscle loss and a higher chance of bingeing.
Trust the trend over the day. If your weekly average is moving in the right direction, the math is working.