The Activity Screen: Where Your Energy Actually Went

What feeds the target, and what does not

restingProjected = basalSoFar * 1440 / minutesElapsed   // clamped
expenditure      = restingProjected + activeSoFar + manualKcal
target           = expenditure + surplusKcal

Active energy counts as it is measured, with no threshold and no discount. Walking and workouts raise the target the moment they happen, which for a hybrid athlete is the point: Saturday's long run appears in Saturday's allowance rather than being paid for from a guess made on Monday.

Resting is projected to twenty-four hours because the platform reports it as a running total. At 8am it has accumulated a third of the day, and a target built on the raw sum reads about 750 kcal — which is exactly the number someone would see at breakfast.

Activity is never projected. Doing so would assume you will keep moving at the current rate, which is wrong in both directions and worst first thing in the morning.

The deliberate refusal

A lifting session contributes nothing to expenditure beyond what the health store already attributed to it.

Two reasons, and the second is the important one.

It would be double-counted. Active energy already contains whatever the platform assigned to that workout. Adding a separate MET-based estimate on top counts the same session twice.

Lifting is genuinely cheap. A working set is thirty seconds; the rest is three minutes. Averaged across a session including the rest, resistance training is roughly 5 METs — much lower than people expect, and much lower than the recovery cost would suggest.

That gap between energy cost and recovery cost is the thing worth internalising. A heavy lower-body session costs very little energy and a great deal of recovery. An easy hour of running costs a great deal of energy and very little recovery. Treating either currency as a proxy for the other is how people mis-plan both their eating and their training.

Steps, which matter more than they look

Non-exercise activity thermogenesis is the most variable component of daily energy expenditure — between people and within one person across a week. A desk day at 3,000 steps and a day on your feet at 15,000 differ by several hundred calories for the same person.

It also adapts. Eat less and you move less, without noticing. That adaptation is the single largest reason a formula-based calorie target drifts out of true over a block, and it is invisible to any activity multiplier.

Measured maintenance absorbs it automatically because it measures the outcome rather than predicting the inputs. The step count is shown here so you can see it happening. The method.

The eat-back rule that was retired

Worth documenting because the reasoning generalises.

Base used to credit 90% of the excess over a thirty-day median of active energy, above a 150 kcal threshold. It almost never fired, and the reason is instructive: that "sedentary baseline" was the median of all days, training days included — so it was inflated by exactly the exercise it existed to pay back.

A baseline computed over a population that includes the thing you are trying to detect cannot detect it.

Active calories now count toward the target as they are measured, with no threshold and no discount. Crediting a bonus on top would count the same walk twice.

Manual entries

For anything the device missed: a hike without a watch, a machine that does not sync, a session where the tracking failed.

Manual entries store the bodyweight they were logged at. Editing your profile six months later does not rewrite the energy cost of a past session, because the cost of that session was a function of the body that did it. A log that silently rewrites its own history is a log you cannot reason about.

Energy for a manual entry is computed from a MET value for the activity times bodyweight times duration — the same scale, in reverse, that underlies almost every published activity calorie figure.

How much to trust the numbers

Honestly:

Wrist-based estimates rest on assumptions about your economy that are not measured and are not yours. Use them as a trend and let measured maintenance carry the decisions, because a 300-calorie error is well inside normal and larger than most people's entire surplus.

When there is no data

The screen says so. "Move with the phone or watch on, and share Active Energy" — not a zero, and not a dash.

If the health store has nothing for the day at all, the daily target falls back to committed maintenance plus your surplus, and says that it has. One fabricated zero teaches you that every other number in the app might also be decorative, and there is no way to earn that back.

What to do next

The dashboard is where these numbers become today's target, and how many calories running actually costs covers why the watch's figure and the scale's answer frequently disagree.