The long argument

Why the formula is not enough.

The case at length: how the measurement works, what it cannot tell you, and where other apps beat mine.

A note from the person who made it

I built this for myself.

I wanted to know what I actually burn, rather than what a formula assumed I burn, and nothing I tried would tell me. So I worked it out from what I ate and what the scale did, and then I kept using it, because it turned out to be the thing I had been missing.

Everything else I do is already inside that number: the walks, the training, the days I barely move, none of it logged separately. Watching it move with my week made me want to move more, which is not something I expected from a tracker.

It helped me. That is the only reason it exists. There is no company behind it and nothing to sell, so leaving it free was not a strategy, it was just the obvious thing to do with something I had already built.

If you are where I was, wanting a number that is measured instead of assumed, it is yours.

I still use it every day
01 · The mechanism

A calculator predicts. This measures.

Ask three apps what you burn in a day and you get three confident numbers, all produced the same way: Mifflin-St Jeor BMR multiplied by an activity level you picked from a dropdown. The formula is a population average published in 1990. For any one person it can be off by 500 calories a day, and the multiplier is a guess you made about yourself.

TrueBurn starts with the same formula, labels it an estimate, and then throws it away. What you ate across a window, plus what your weight trend actually did, implies what you must have burned. That is the whole idea, and the arithmetic is short enough to print.

Worked in full

Measured
expenditure = intake + (weight lost × 3,500 ÷ days)
Average logged intake, 28 days 2,180 kcal/day
Weight lost across the window 3.2 lb
Energy per lb of body mass 3,500 kcal
Implied deficit (3.2 × 3,500 ÷ 28) 400 kcal/day
Measured expenditure 2,580 kcal/day

An example, not your data. Every figure in the app carries this same panel. If I cannot explain how I got a number, I do not display it.

02 · What it cannot tell you

TrueBurn cannot reliably detect metabolic adaptation in an individual, and neither can anyone else with a scale and a food log.

The effect people want measured is smaller than the error in the measurement that would have to detect it. So there is no adaptation score here, no recovery percentage, and no measured-over-predicted ratio presented as a diagnosis. Those are not missing features. They are decisions, and they are permanent.

You will not see a plus-or-minus figure next to the number either. I could work out how much the daily wobble on the scale moves it, but that is the small problem. The big one is that most people log a little less than they actually ate, in the same direction almost every day, and no plus-or-minus can account for that. A tidy range would make the number look more exact than it is.

The one
real ask

A stable 15% under-log is fine, because it cancels. What actually breaks the math is drift. Log the same way most days is achievable where log everything is not, and it happens to be the true requirement.

03 · The research

I read the studies first, and it cost me the main feature.

The original plan was to tell you whether years of dieting had slowed your metabolism down. That was going to be the reason to use this app.

Then I went and read what the research actually says, and it says nobody can answer that honestly from a scale and a food log. So the feature was cut before it was ever built. None of this makes the number more accurate. It is just how the decisions here get made.

What the reading changed
Cut The feature this app was going to be sold on.Cut A scored questionnaire meant to spot disordered eating. It would have been wrong about four people in five.Corrected Six mistakes in my own arithmetic, found before anyone else could find them.
04 · What it refuses to do

Nine things this app will not do, and the reason for each.

This is not a caveats section bolted to the end. Naming what the product cannot do is how it earns the right to be believed about the rest.

No metabolic reset, under any name
The term has no clinical definition, no MeSH entry, and no measurement standard. Reverse dieting specifically has no published full-length human trial. The only randomized data is a null conference abstract that trended against it. Read what the evidence actually says
No adaptation score or recovery percentage
The effect being scored is an order of magnitude smaller than the error in the number it would be scored against. Marked permanent, not deferred.
No adaptation detector at all
Modeling showed it would fire on a quarter to a third of perfectly healthy users from equation error alone.
No eating-disorder screener
At general-population prevalence, roughly four of every five people a scored screener flags would not have an eating disorder. Instead, referral wording drawn from fields already held, phrased as observation rather than diagnosis.
No red "you went over" states
Green for achieving something is fine. Red for failing to stay under a limit is a documented harm mechanism. What is built today is the asymmetry: a hard floor on recommended intake and no ceiling enforcement, so the app protects you from under-eating and declines to nag you about over-eating.
No exercise-compensation haircut
Despite a tempting 28% figure, because that number is a between-person regression slope whose own authors disclaim using it as a within-person correction.
No under-logging buffer in the arithmetic
Padding your intake to guess at what you missed would make the number less honest, not more accurate. The consistency ask does that work instead.
Never a silent adjustment
If you answer a question about your week, the number stays what the arithmetic says. The answers change the words around it, never the figure itself.
No social comparison, in any form, ever
Stated as a value during onboarding rather than being quietly absent, and permanent rather than deferred.

One thing worth naming. There is an entire content industry waiting at the moment a diet stalls, selling metabolic resets and reverse diets for a problem people usually do not have. That is the claim being refused here, not any particular app.

05 · The competition

I am not the first to do this.

Other apps measure expenditure from intake and weight trend too, and claiming otherwise would be the exact kind of thing this project exists to refuse.

MacroFactor is built on the same arithmetic and does it well, with a real team and years of refinement behind it. Carbon Diet Coach occupies adjacent ground. Adaptive-TDEE spreadsheets have circulated in lifting communities for years. The energy-balance equation is not proprietary and nobody should pretend it is.

Prices are approximate because they move. MacroFactor is around $12 a month and has said publicly there will never be a free tier. Carbon Diet Coach is around $14. Both require an account and hold your history on their servers.

Where they are ahead of me

Their expenditure model has had years of work behind it that mine has not. Their food database is checked by hand where mine is USDA plus a short curated list. They read active-calorie data from Apple Health, so exercise needs less logging by hand. If price is not an obstacle for you, paying for one of them is a perfectly sensible thing to do, and I am not going to pretend otherwise.

So who is
this for

The person who will not pay $70+ a year, or who does not want their weight history sitting in someone else's database, currently has no option in this category at all. That person is who TrueBurn is for.

If that sounds like you, it is free and there is nothing to sign up for.

About two weeks of logging before the measured number appears. Until then it shows the formula estimate and calls it an estimate.

Free. Private. No sign-up.