When your fitness tracker or phone says you've hit your steps, burned enough calories, or had excellent sleep, treat the number as a claim — not proof. A shopper checks how their legs actually feel and whether they remember doing the movement. A builder adds one independent check before letting the data drive recommendations or close any loop.
A wrist wearing a fitness tracker showing an inflated step count over a parent pushing a stroller
The day the watch lied about the walk
You spend Saturday running errands. You push a stroller for two hours, carry bags in and out of stores, and gesture while chatting with a friend over coffee. Your wrist feels tired from all the motion. At the end of the day, your watch shows 13,800 steps and says you've burned more than 600 extra calories. You feel accomplished. You have a relaxed dinner and skip the evening walk you had planned.
The next morning, you check the scale. Nothing moved in the direction you expected. Your legs don't feel like they did 13,000 steps of real walking. You feel the same as any other weekend.
What the watch counted was mostly your arms swinging the stroller and bags, not the distance travelled by your legs. The algorithm saw movement patterns it had been trained to call "steps" and reported success. You acted on the report.
This is tracker theatre: the performance of recorded activity without the actual movement that the number implies.
Person checking watch after errands next to stroller
How a tracker decides you've "moved"
Most wrist-worn devices and phone health apps use accelerometers to detect repetitive motion. The software looks for patterns that resemble a step — a certain up-and-down or swinging motion at a certain frequency. It does not know whether the motion came from your legs walking down the street or from your arms pushing a stroller, steering a shopping cart, or even the vibrations while driving.
In controlled lab walking, many devices are reasonably close to accurate. In real daily life, the gap opens up. One analysis of multiple activity bands found they overcounted steps by 25 percent. Another study comparing wrist-worn monitors to a reliable hip-worn pedometer found an 11.4% overestimation overall in real-world conditions, driven by a 22% overestimation during everyday tasks.
The device has no way to see the difference between "I walked the distance" and "my arms moved a lot." It simply reports the pattern it detected as a success.
Diagram showing the mismatch between the wrist tracker arm movement and actual walking
Make it trustworthy
For everyday readers: Treat the number as interesting data, not a fact about your body. At the end of the day, ask yourself two questions before you act on it: Did I actually feel my legs working for most of those steps? Can I point to a stretch of time where I know I was walking the distance the watch claims? If the answer is no to either, the number is probably inflated.
A one-minute check you can run today: Pick a normal activity you do often (pushing a stroller, carrying bags, working at a standing desk). Count your actual steps for one minute while doing it the normal way. Compare that to what the watch shows for the same minute. The gap is often obvious.
For builders: When your system receives step, calorie, or sleep data from a wearable, store the raw claim separately from anything you treat as verified. Do not award streaks, adjust calorie targets, or send "great job" messages based only on the device's success signal. Require a second, independent signal — user confirmation, a different sensor, or an explicit review step — before the data influences user-facing outcomes.
One check you can run today: After the next time you integrate wearable data, look at the last 100 "active day" or "good sleep" events. How many of them have any record of the user confirming the number or another system cross-checking it? If the answer is close to zero, your system is currently running on tracker theatre.
Person verifying the tracker by checking their own movement
Check the actual movement
Watch with the check next to the person feeling their actual movement
A shopper now looks at how their legs and body actually feel before deciding the day was active enough. A builder adds one external verification step after every wearable success signal. The simple list of quick checks that separates what the device claims from what actually happened lives in one keepable sheet.
Keep the companion sheet: The Tracker Reality Check (PDF) → https://resources.theifstatement.dev/the-tracker-reality-check.pdf
The If Statement — the logic everything rides on.