
I've had three watches over the last 4.5 years and I've logged my daily step counts from the watches as well as from my phone (as tracked by Google Fit). I excluded about a dozen days due to anomalies, mainly days where I'd go for a run and leavey phone at home. The stats were run on the daily data but the chart is weekly to get rid of some of the daily noise to see the relationships better. I forced the linear regression to have an intercept of zero for easier comparison. All watches counted more steps than my phone which largely makes sense since there still might be short periods of time where I put my phone down while I'm in the house. Fitbit was the most off, counting almost 20% more steps than my phone.
Posted by dang_rat_bandit
![Watches vs phone step counts [OC] Watches vs phone step counts [OC]](https://www.byteseu.com/wp-content/uploads/2026/06/3q8omzaygq4h1-1536x682.jpeg)
8 Comments
You also need a baseline count done by hand.
🥺
Well what happens if you wear all of the watches at once?
Omg how many miles is that a day?
Dear god, man. Just use fast travel.
If anyone is interested in the data: https://docs.google.com/spreadsheets/d/1fFGhdC3BbgpE5Mw55j_A2ErVC6RjZ3sH/edit?usp=drivesdk&ouid=103223588908820622864&rtpof=true&sd=true
Very interesting, I just got a Fitbit air, to replace my aging Withings HR from 2018
Wearing both currently and the Fitbit is getting close to double the steps…
Halfway through my day, Withings at 4200, Phone (Samsung) at 5000 and the Fitbit at 7400.
Crazy difference
A more statistically relevant tool than a regression is a paired t-test which looks at the error between the measurements each week to determine if the error is statistically different than zero (assuming the error is normally distributed).
[ Removed by Reddit ]