Thoughts about Garmin device relative GPS accuracy

Interested in getting others’ thoughts on this. I have both a Garmin Fenix 6S sapphire watch and a Garmin Edge 830 bike computer. My tendency is to use both to record rides, both road and MTB. Initially I used only the 830 but would sometimes find that the watch would yell at me for “high heart rate” if it wasn’t actually recording an activity, so I started recording rides with it, too.

On road rides (meaning rides with fewer overhead impediments to the GPS satellites) I find the two devices are remarkably consistent – like nearly identical total mileage readings even after long rides. However, on MTB rides, which I do in pretty thick area forests, the mileage disparity is much larger – it usually works out to around 2% (roughly 0.2 miles on a ~10-mile ride), and the Fenix always displays the longer distance. I should note that the software/maps in each are fully up to date and I always use the “GPS + Galileo” setting for both devices.

I find this interesting. I can only attribute the distance disparity to what I guess are GPS signal strength differences between the devices. Which do you think is closer to the truth? Is one unit known to have a better receiver than the other when under thick tree canopies? Logic might suggest the larger distance number is more accurate (the longer distance implying fewer dropouts between datapoints – or something?) but maybe that’s just wishful thinking. If I look at a completed ride’s map traces on either Garmin Connect or Strava, they look identical between the two devices.

This does not keep me up at night; I just find it curious. Would love to hear others’ ideas or knowledge about the relative accuracy of these gadgets. Thanks in advance!

I haven’t done the Fenix vs. Edge head to head, but one thing I did notice is that speed and distance accuracy improves when my Edge is paired to a Garmin Speed sensor. That esp. helps when the GPS coverage might be more sketchy. This could likely work with a Fenix as well. I always have my Edge with me when riding, so I don’t worry about connecting Speed or Cadence to my Fenix.

Responsiveness to change is also much better with the Speed sensor. Without it, even the Edge would take a bit to catch up on speed changes.

Do you have autopause set to on on the edge? Off road you are often travelling at a speed that’s low enough that the edge thinks you are paused.

As you have surmised, triangulating off of satellites isn’t the most accurate way to measure speed and distance. Of course, most of us find it good enough. As @MrBirchling said above, a speed sensor that is measuring wheel speed and thus the distance the wheel turns is the more accurate way of doing things if that matters to you.

Could just be implementation differences between the devices. For example capture frequency. Or maybe one of the devices is using accelerometers to fine-tune the distances. Or one is just doing straight-line distance between the capture points while the other is fitting some fancy bezier curve and measuring the lengths of those. This seems very plausible on an MTB ride where there tend to be lots of twists and turns. But if you want to go full conspiracy, check out the Coastline paradox