How wrong are watches and trackers about calories burned

The watch reads 612 calories after an hour in the gym, and the figure turns into a plan on the spot: right, that's an extra plate at dinner. What rarely gets asked is how much that number actually tells you.

Short version: the heart rate your watch measures is usually solid, and the calorie figure it builds out of that heart rate is not. Two different things, shown on the same screen.

What is reliably known

The evidence is consistent here: wrist-based optical sensors track heart rate well when effort is steady. Walking, running at an even pace, a stationary bike — error against an ECG usually sits under 5%. Step counts hold up reasonably too, at least during ordinary walking.

Equally consistent is the finding that calorie estimates do not inherit that accuracy. Set against laboratory methods — gas exchange measurement, doubly labelled water — energy expenditure figures from consumer devices typically land 20-30% off, and further off for certain activities and certain people. No mainstream device has shown small, dependable error across all conditions.

The reason is structural. A watch does not measure energy. It measures pulse and movement, then pushes both through a model that converts them into calories. That model assumes a relationship between heart rate and oxygen uptake which shifts from person to person, depending on fitness, muscle mass, room temperature, hydration, sleep, stress and caffeine. Two people of the same weight at the same heart rate are not burning the same amount.

The worst cases are activities where that link breaks entirely. Lifting is the textbook example. Your heart rate hits 150 in a set of squats, but oxygen uptake doesn't follow, because the effort is short and partly anaerobic. The watch reads it as sustained cardiovascular work and inflates the total. Intervals, circuits and stop-start sports do the same.

One more firm point: most of what you burn in a day has nothing to do with training. Resting metabolism is usually 60-70% of daily expenditure, and no watch measures it — it is estimated from your height, weight, age and sex. If that estimate is 200 calories out, the whole day is 200 calories out before you leave the house.

What is likely but not settled

Devices are probably least wrong for flat running and walking, where GPS pace and stride length give the model something physical to work with instead of heart rate alone. The evidence points that way; exactly how much better it is has not been pinned down.

A tracker is also probably better at catching change within one person than at producing an absolute figure. If the same session reads 400 one week and 460 the next, the second one was likely harder. Useful — but the evidence for this is thinner than the evidence about the size of the error.

Personalisation — calibration from an estimated VO2 max, your logged weight, months of wear — should in principle narrow the gap, and every manufacturer leans on it. How much it actually delivers is poorly researched outside the companies selling it.

Skin tone and tattoos: motion artefacts are a clear problem, since a loose strap, a swinging arm or a cold wrist all degrade the signal. Findings on pigmentation are mixed, and it would be overreaching to claim more than that.

What gets passed on wrongly

"The heart rate is accurate, so the calories must be." It doesn't follow. An accurate pulse is only the input to a model that remains an estimate. A chest strap reads heart rate better than a wrist, but it doesn't fix the step where pulse becomes energy.

"It said 700, so I can eat 700 more." Two errors stack here: the device's, and your own food logging, which typically under-records intake by 20-30%. The easy outcome is eating more than you spent.

"The fat-burning zone." At lower intensity a larger share of energy does come from fat. But total expenditure and long-term fat loss don't hinge on that share. A screen telling you you're "in the fat zone" says nothing about the result.

"Afterburn." Expenditure does stay slightly raised after training, but it's an add-on on the order of a few per cent of the session itself, not hundreds of calories. The afterburn numbers apps display are not well supported.

"Active calories are what I burned." Depends on the device. On some it means expenditure above resting; on others it's total expenditure during the activity. For an hour's training the difference can be 60-80 calories purely from the definition.

What this means for your training and eating

Don't treat the calorie figure as a budget. Treat it as a record that you trained, and as a rough check on whether this week was harder than last.

For losing or gaining, start with the scales and the calendar, not the device:

• Weigh yourself in the morning under the same conditions and work from a weekly average — daily swings of 1-2 kg are water and food in transit.
• Watch that average for 2-3 weeks. If it isn't moving, shift intake by 200-300 calories a day in the direction you want.
• If you insist on eating back what you burned, eat back half of what the watch says. It's crude cover against an overestimate.

Don't judge strength work by calories, because that's where the watch is weakest. The useful markers are the load on the bar, the reps you got and how close to failure you took the set.

Use what the watch does well: resting heart rate tracked across weeks, session count and duration, steps as a proxy for daily movement. Those numbers you can actually act on.

And don't expect your figure to match anyone else's. The response to identical training varies between people, and so does the pulse-to-expenditure relationship. Caffeine before a session lifts your heart rate and therefore the calorie estimate without you spending anything extra, which is one reason to know how much you're actually taking.

The logic mirrors body fat measurement: the device hands you a number with error attached, so the direction over time is worth more than any single reading.

When to see a doctor

A watch is not a diagnostic tool, but it sometimes notices something. See a doctor if you get chest pain or pressure and breathlessness on exertion, if you feel dizzy or have fainted, if you notice pronounced skipped or racing beats, if your resting heart rate jumps for no reason and stays 15-20 beats higher for several weeks, or if you're losing weight without trying. Don't diagnose from a screen — and don't dismiss a symptom because the watch said everything was fine.

In short

• Wrist heart rate is dependable during steady effort, usually within 5%.
• Calories are a model's estimate, typically 20-30% off, worse for lifting and intervals.
• Resting metabolism is 60-70% of your daily burn and the watch only assumes it.
• Set intake from your weekly average weight, not from the number on the watch.
• Use the device for consistency, resting heart rate and trend — not for meal arithmetic.