Why does my calorie target change during the day?

Because your expenditure does. Activity is counted as it happens rather than predicted, so a long ride adds several hundred calories to the day it occurred on. A fixed target cannot know that Saturday had a two-hour session in it, which means the food you need after a hard session would have to come out of a guess made days earlier.

Why is resting energy projected but activity is not?

Because a basal rate is genuinely predictable and behaviour is not. Extrapolating the observed resting rate across the full day is a measurement; predicting your activity would be promising calories for a walk you have not taken, and the number would sag every time the forecast failed to come true.

What does my target start at in the morning?

At your projected full-day expenditure plus your surplus or deficit — not at a fraction of it. The platform reports resting energy as a running total, so a target built on the raw figure would read around 750 calories at breakfast. Projecting resting to twenty-four hours is what stops the morning number being nonsense.

Does the target match my phone's fitness app by the end of the day?

The expenditure component does, deliberately. Both use the same definition and the same moment, so by midnight the projection has converged on the measurement and the two agree. An app that computed its own private version of a number the operating system also reports would be one you had no way to audit.

What happens if I have no health data connected?

The target falls back to your committed maintenance plus your surplus, and says so. That is a fixed number for the day rather than one that moves with measured activity, which works but loses the responsiveness that makes the target useful for a hybrid athlete.

How Your Daily Calorie Target Is Calculated

The formula

restingProjected = basal so far × 1440 ÷ minutes elapsed
                   (clamped to between 1× and 2× the Mifflin-St Jeor estimate)

expenditure      = restingProjected + active so far + manual entries

target           = expenditure + surplus

Three decisions are embedded there and each one is doing work.

Resting is projected

Health platforms report basal energy as a running total. At 8am it has accumulated roughly a third of the day.

A target built on that raw figure would read about 750 calories at breakfast — not merely wrong but actively harmful, since that is the number someone sees when deciding what to eat. It would climb all day and mean nothing until bedtime.

A basal rate is close to constant, so extrapolating the observed rate across 1,440 minutes is a measurement rather than a guess.

Activity is never projected

Walking and workouts count as they happen.

Projecting them would mean promising calories for a walk not yet taken, and the number would sag every time the forecast failed. Worse, it would make the target fall during a rest day, which is exactly backwards from how it should feel.

The consequence is the useful one: the target rises the moment you move.

The clamp

Ten minutes after midnight, dividing by minutes elapsed produces nonsense. Clamping the projection to between one and two times the Mifflin-St Jeor estimate keeps it sane while the day accumulates.

It converges

By midnight the projection has converged on the measurement, and the expenditure figure equals what your platform's own fitness app reports.

That is deliberate. It is a number you can check — hold the two side by side and confirm they agree. An app computing its own private version of a figure the operating system also publishes is an app you have no way to audit.

Where the surplus comes from

target = expenditure + surplus

The surplus is what you configured against your goal and rate. The Food Log header spells the whole thing out rather than presenting a bare number:

Target 3,240 = 2,940 maintenance + 300 surplus · 812 to go

Without that line the header reads as a countdown to an unexplained number — and on a bulk, a countdown looks like a deficit you are closing rather than a surplus you are climbing to.

Eat-back

When active energy exceeds the sedentary baseline by more than 150 calories, Base adds back 90% of the excess rather than all of it.

The 0.9 is deliberate conservatism. Wrist-based active energy over-estimates, and eating back 100% of a biased estimate compounds the error every single day.

With no health data

The target falls back to committed maintenance plus surplus — a fixed number for the day — and says so.

It works. What it loses is the responsiveness, which for a hybrid athlete is most of the value: the whole problem the app exists to solve is that a week containing eight hours of endurance work does not cost the same as a week that does not.

Target versus maintenance

Two different questions, and it is worth keeping them apart:

Number	Question
Expenditure	What did today cost?
Maintenance	What does an average day cost?
Target	What should I eat today?

Expenditure is today's measurement. Maintenance is a slower figure back-calculated from your trend weight against your intake, and it is what the surplus is set against.

Common mistakes

Reading the morning target as final. It rises as you move, by design.

Assuming a rising target is a bug. It means you moved.

Eating back activity twice. The target already includes measured activity. Adding a manual entry for the same session double-counts it.

Trusting wrist calories precisely. The 90% eat-back factor exists because they are biased high.

Comparing the target against a formula from another app. Most produce a fixed daily figure from a multiplier. This one is measured and moves.

Panicking at a low target on a rest day. A rest day costs less. That is what a rest day is.

What to do next

Open the energy bar on your dashboard and expand it. Check that today's training appears in the exercise zone — if it does not, the session never reached your health store, and every number downstream of it is short by that session.