Smartwatch walk distance looking funny? You’re not alone

Wearables like smartwatches are utterly packed with sensors, and in using them we gather all sorts of data. But just how accurate are all of those numbers? With some, like calories burned during an exercise, it makes sense that the figure’s just an estimate, and we shouldn’t expect it to be precise. Other numbers, like how many steps you’ve taken, we’d assume were exact figures. Which of those camps would you think distance measurements fall into?

Since smartwatches are generally equipped with GPS receivers, it would make sense that any distance measurements they report are as accurate as the GPS position data they’re computing. But over on Reddit’s Pixel Watch sub, user almosttan has been sharing a very confusing development that happens on their Pixel Watch 4 when comparing the distance clocked on a walk to the distance measured while running or cycling.

First noticing a discrepancy a few weeks back, almosttan initially suspected that their Watch 4 was computing walk distance by multiplying the number of steps taken by their entered stride length, rather than actually using GPS measurements. And after testing that theory across a few other modes of locomotion, it’s looking increasingly plausible.

Check out these two journeys, back and forth from the same points. Granted, the route taken is a little different between them, but that on its own doesn’t seem to explain the massive disconnect in distance reported.

Just to double-check, we measured both of those in Google Maps. The bicycle ride is just about right on the money:

But when we’re looking at the walk, the actual distance traveled turns out to be over four times as far as what the watch reports:

Now, this makes us very curious what almosttan has entered for their stride length (and what it actually might be), but there’s clearly a problem here with computing walk distances.

Those screenshots from before are of the Strava app, and at least if you’re using Strava on the web, you might have some luck there. There’s clearly GPS data available (since we can see the route), and Strava offers a “Correct Distance” tool in your browser that should allow it to recompute your walk’s true length.

But is there a better approach to all this? Where is this bad data actually originating? We know that Fitbit supports automatic stride length correction, using GPS data and your step count to get a more accurate measurement — but Google’s support docs mention that’s for runs, so we wonder if walking is a blind spot there.

Right now, we’re reaching out to Google to see if it can shine any light on what’s going so very wrong here — and we’ll update you with anything we hear back. In the meantime, just make sure you don’t have a wildly inaccurate stride length saved, and maybe think about logging your walks as just very slow runs.