Why this site does not print calorie numbers

What people are told

That a given activity for a given time consumes a specific number of calories. The treadmill says so. The watch says so. The article says so, usually to three significant figures.

What is actually known

Every one of those numbers is an estimate produced by a formula, not a measurement.

Actually measuring energy expenditure requires indirect calorimetry — analysing inhaled and exhaled gas — or doubly labelled water over days. Neither is happening on a treadmill or a wrist.

What is happening instead is a model: bodyweight, movement type, duration, sometimes heart rate, fed into an equation derived from population averages. Individual expenditure varies with body composition, movement efficiency, fitness, temperature and a good deal else the model does not know about you.

Validation studies of consumer activity trackers have generally found step counts and heart rate to be reasonable, and energy expenditure to be the weakest output — with errors substantial enough to matter to anyone using them for decisions.

There is a second problem the numbers hide. Total daily energy expenditure does not simply rise by whatever an activity "costs"; there is evidence of compensation, with people moving less at other times or physiological adjustments partly offsetting the addition. How much this happens is debated, and it means the arithmetic of "burn this, subtract that" is not the clean sum it appears.

What follows from that

A number carrying an unstated error of that size is worse than no number, because a figure to three significant figures communicates a precision that does not exist, and people make decisions on the precision rather than the estimate.

This site prints no calorie figures anywhere, and there is an automated test that fails the build if one appears in any record. Along with the absence of a browse-by-goal facet, it is one of the refusals built into the code rather than the copy.

Effort is not measured by an invented number. What you did, for how many, through what range, and how hard it was are all real and all available.

None of this means energy balance is not real. It is. It means the specific figures you are shown are not accurate enough to manage it with, and this site does not do nutrition — a deliberate scope decision, not an oversight.

What we still cannot say

How wrong any particular device is for any particular person. Error varies by device, by activity and by individual, and that variation is the point.

How much compensation occurs. It is observed, it is debated, and its size is not settled.

Whether a wrong number is still useful as a relative measure — tracking whether today was more than yesterday, rather than what either was. That is a more defensible use, and it is not how the numbers are presented or understood.