5x Plus distance accuracy

I did a little distance check this morning with my 5x Plus. MTB ride with low speeds (4-5 mph tight and twisty). Fenix 5x plus (mounted to handlebars), Garmin FR 305 (mounted to handlebars), and Strava Andriod app on my phone (in pocket). All devices allowed ample time to get solid GPS signal/sync. I split the ride up into 3 separate 'rides'. About a 1 1/2 hour total. This is in a wooded city park w/ good cell service, no buildings, just trees. First 2 'rides' had Fenix in GPS+Galileo. On 3rd 'ride' switched to GSP only.

The 305 and the phone actually were very close in total in all three 'rides' created. They almost exactly aligned. Very impressive. The Fenix, on the other hand, consistently lost close to a 1/10th of a mile every 1 mile. I mean really consistently. It did it in both GPS+Galileo and GPS only. I also noticed a big difference in distance when I first got the watch (about 2 weeks ago) compared to my phone in GPS+Glonass mode, but didn't pay too much attention to it at the time.

What is the deal with this? I don't see a issue with the Fenix tracks (had a couple weird ones w/ Glonass), but this distance thing is unacceptable to me. Could it be that the Fenix is the actual accurate one? I doubt it, but I could be wrong.

***Edit: Need more rides before coming to conclusion. Rode again (3.5 miles) and all devices were pretty close.****
  • I have a question guys. Most of you must sync your activites with Strava. Did you try to fix the distance with the option there, and see if it looks better ? If you export your Fenix GPX file and import it anywhere, it will show a longer distance. Sometimes the difference is very small (mostly for clean tracks, no buildings or trees), sometimes it's significant. Even on Garmin Connect, reimport your GPX, and you will see a longer distance. So the pure GPS distance is always longer than what the watch is showing.

    I suspect that Garmin is maybe doing what Suunto is doing, they are mixing the GPS data with the accelerometer data (the Fenix has one, right ?), and smoothing the GPS results. The GPS is not perfect, when there are obstacles, it's waving too much left and right. The accelerometer data must be different, and Garmin could be using that to say "ok, the GPS reception isn't that great right now, let's calculate the distance a bit differently". That's why the watch is showing X and the actual GPS data is showing X+something. The idea is not bad, but it seems to be a bit heavy on correction. On my activities, it doesn't seem too far from the reality, since I'm often preparing my runs on a map and following the route. So I know what should be the perfect distance, and often the Fenix is not far. Others devices may show more, but if it takes into account all the little errors of the GPS, it's not actual distance I'm running. So I prefer that it underestimates a bit, rather than the opposite. I still think it's too heavy on the correction, there are cases when it's too low, for sure. I'm not mad at Garmin for that, in my situation it doesn't seem like a huge issue at all, the watch is doing a terrific job in many other domains.

    Did others of you try what I did with Strava (or any other solution to measure the distance of a GPX) ?