Tracker calories wildly off
A Stanford study highlighted on X shows fitness devices can overestimate calorie burn by anywhere from 27% to 93%, so trusting device-read calories to ‘earn’ food is risky. That gap comes from generic formulas that ignore a person’s efficiency and fitness level — the takeaway is to track food intake rather than eat back device numbers. (x.com)
Your watch can be almost dead-on for heart rate and still be wildly wrong about calories. In a Stanford study of 7 wrist devices, 6 measured heart rate within 5%, but none measured energy expenditure accurately. (med.stanford.edu) Calories burned is a harder number because the device is trying to guess your body’s engine output, not just count beats. Stanford tested that by putting 60 volunteers on treadmills and stationary bikes and comparing the watches to lab-grade reference methods. (med.stanford.edu) The gap was huge. The most accurate device was still off by an average of 27%, and the least accurate missed by 93%. (med.stanford.edu) Stanford’s team looked at Apple Watch, Basis Peak, Fitbit Surge, Microsoft Band, Mio Alpha 2, PulseOn, and Samsung Gear S2. Apple’s device performed best on both heart rate and calorie estimates, while Samsung’s had the highest error rates in that test. (cnbc.com) The reason is simple: two people can do the same workout and burn different amounts of energy. Euan Ashley at Stanford said one person may walk smoothly while another “waddle[s] along,” and that difference changes the real calorie cost even when the watch sees similar motion and pulse data. (cnbc.com) That means the calorie number on your wrist is usually a formula, not a measurement. It is built from signals like heart rate and movement, then adjusted with generic inputs such as sex, age, height, and weight. (med.stanford.edu) Stanford also found the errors were not evenly spread across everyone. Mistakes were more common in men, in people with higher body mass index, and in people with darker skin tone, which shows how a one-size-fits-all algorithm can miss real human variation. (cnbc.com) The researchers set a practical target of keeping error under 10% for everyday use. None of the devices hit that bar for calories, which is why Ashley said people should not use “calories burned” to decide what to eat afterward. (med.stanford.edu; cnbc.com) If your watch says you burned 600 calories, a 27% error means the real number could be closer to 438. A 93% error means the estimate is not a guide so much as a rough guess with a very confident-looking screen. (med.stanford.edu) The useful part of a tracker is the trend line: steps, time moving, workout frequency, and heart rate during the session. The risky part is “earning” food from the calorie readout, because the number can be wrong before the first bite. (med.stanford.edu; cnbc.com)
