The ledger
Every day gets one line in a ledger. The line is a simple energy balance:
BMR is your baseline burn. What your body spends just existing. Active burn is what you spent on top of it, usually from a watch. Intake is what you ate. When you spend more than you eat, the day banks a deficit; the sum of those deficits is the total you've banked toward your goal.
You can log any subset of the three on a given day, just a weigh-in, just intake, whatever you have. A day only earns a deficit when it has both intake and burn, and here's the part most trackers get wrong: a partial day is never scored as a zero. A day where you logged nothing simply isn't in the ledger. It can't drag your average down, and there's no streak it broke.
Two honest edge cases get their own treatment. A fast is a real day with a real 0 kcal intake. Confirm it as a fast and its full deficit counts, because you genuinely earned it. And a day you know you only half-tracked can be set aside: its data stays visible, but it yields no deficit at all, so a knowingly-incomplete day can't pollute your pace or your total.
The scale is noisy
A single weigh-in is mostly water. Yesterday's salt, this morning's glycogen, whether you drank a glass of water before stepping on. Day to day, the scale swings by amounts that have nothing to do with fat. So Fit Rocket never reasons from one weigh-in. It reasons from a smoothed trend.
The trend is an exponentially weighted average of your weigh-ins: each new reading moves the trend 20% of the way toward it (α = 0.2 per day), so a one-off spike barely dents it while a real shift shows up within days. The smoothing is gap-aware. A weigh-in after N days away moves the trend as far as N daily steps would have, so a missed week doesn't corrupt the line, it just picks up where it left off.
Every downstream number reads many weigh-ins, never one: the burn estimate and the goal date start from the trend, and the accuracy check fits a line through six weeks of them. Your raw weigh-ins are still shown; they're just not what the math believes.
A burn estimate that follows you down
The baseline burn comes from the Mifflin–St Jeor equation, the standard clinical estimate:
Notice the first term: 10 kcal per kilogram. As you lose weight, your body genuinely burns less, about 10 kcal/day less per kilogram lost. A tracker that computes your burn once at sign-up and never revisits it quietly overstates your deficit more with every kilogram, and the error compounds for months.
Fit Rocket recomputes the equation at your current trend weight every time a new day is logged, so the estimate follows you down. But it never rewrites the past: each day's ledger line keeps the burn estimate that was true when it was written, and editing an old day keeps its original snapshot. The ledger is a record, not a restatement.
The two bars
The app's core idea is two progress bars that measure the same journey from opposite ends. One shows the deficit you've banked, as a fraction of the total your goal requires. The other shows the weight you've actually lost, as a fraction of the total you set out to lose. They're tied together by the textbook convention that a kilogram of body fat holds roughly 7,700 kcal, the metric form of the old 3,500-calories-per-pound rule that's been in nutrition texts since the 1950s. It's an approximation, and Fit Rocket treats it as one: a trend signal, not a calorimeter.
Here's why the two bars matter: if your logging is accurate, they are algebraically the same number. Banked deficit ÷ 7,700 is kilograms lost, so both fractions cancel to the same ratio. Which means any gap between the bars is measurement error, and the direction of the gap tells you whose.
When the deficit bar runs ahead of the weight bar, you're losing slower than your numbers claim, usually food that was eaten but never logged. When the weight bar runs ahead, you're losing faster than logged. Over-counted intake, under-counted burn, or a transient water swing. Fit Rocket names the gap plainly instead of letting your logs grade their own homework.
The goal date
The goal date starts from the scale, not from the ledger. Your goal is a weight, so where you are is your weight trend, and what's left is the distance from the trend to your target:
Your pace is your daily deficit, weighted toward your recent days: yesterday counts twice as much as a day ten days earlier, and anything older than six weeks is left out. There's no window for a day to fall out of, so a good day doesn't vanish from your pace a week later and yank the date around. It's your actual recent effort, not your best week and not a plan's assumption.
What your logs miss comes from the scale, and it gets its own section below. Take it away from your pace and you have your real pace, the one your body is actually moving at. Divide, and you get the day you arrive if you keep doing what you're doing. It comes with a likely range, the middle 80% of where it could land, because a date to the day would claim more than the data knows. It's also capped at two years out. Beyond that, a date is theater.
Breaks don't break it. Four days or more without a counted day is a break: the scale's read splits there and skips the first few days back, while a trip's salt and water drain, so a holiday's regain isn't charged to your logging. Without a counted day in the last week, the scale alone can keep the date alive, but only with a recent weigh-in and only when it's clearly heading for your goal. Otherwise there's no date, rather than an old pace re-anchored to today. And reaching your goal means the trend at your target three mornings in a row, not one lucky weigh-in.
The consequence is a date that moves. Push harder and it pulls closer; ease off and it drifts out. That's not a bug to smooth over. It's the entire point. A goal date that answers to your effort is worth checking every morning. One that doesn't is a screensaver. You can try the projection yourself with your own numbers.
What your logs miss
Even careful logging misses things. A splash of oil, a bite while cooking, a watch that runs generous, a baseline burn that runs a little high for your body. The two bars show the gap, and the goal date allows for it. Fit Rocket fits a straight line through your weigh-ins over the last six weeks, converts its slope through the same 7,700 kcal/kg, and compares it with the deficit your logs claim over the same days. The difference, in kcal a day, is what your logs miss.
It won't jump to conclusions. Until your weigh-ins say otherwise, it assumes your logs are right, give or take 150 kcal a day, and a few noisy weeks only move it part of the way. Weigh-ins that run in streaks, water held for days, count for less than independent ones, and the first week of a new cut or build is left out, because it's mostly water. Slips are forgiven by physics, not by policy: one skipped dinner spread over six weeks is noise.
There's nothing to accept. The date already goes by your scale, so the gap never waits for your say-so, and nothing in your logs is rewritten: each day keeps the numbers it was logged with. When the gap is big enough to be worth knowing, Fit Rocket tells you once, in plain words.
What Fit Rocket refuses to do
The method is defined as much by what it leaves out:
- No fixed calorie budget. Fit Rocket measures what you did; it doesn't hand you a daily number to obey and then scold you against it.
- No streaks, no guilt. A day you didn't log isn't a failure to atone for. It just isn't in the ledger. There is nothing to keep alive.
- Workouts never feed the burn. A logged lift or run stays in your diary as a record. If a watch measures your active burn, it already counted that workout. Adding it again would count the same calories twice, and an inflated deficit is exactly the dishonesty the two bars exist to catch.
- History is never restated. New knowledge shapes new days; the days already written keep the numbers they were written with.
For the plain-language version of all this. What the app actually feels like to use. Read how it works. For the short answers, the FAQ.