Explaining the distance problem in detail

Version 21.10 arrives (it really bothers me to be a beta tester), I start the first tests, see the same behavior, and come up with my theories about why the failure occurs. The difference persists, although minimal, occurs in the same place, and I begin to rule out possible causes. I suspect it's my speed and the way I run those 3 km. During those kilometers, I run at a speed of 7 to 7:30 kph.

This morning, I make the plan as follows: I'm going to walk 3 km and run the other 3 km. It's all going to be a single activity (running).

I start walking for 3km, and oh my God, it's terrible! The distance loss is over 100m per kilometer. I have to start manually marking the kilometers, even losing 320m on one. When I reach kilometer 3, I start running, and when I reach kilometer 4, I finally understand the problem and understand why it's failing so much.

Garmins DON'T MEASURE DISTANCE, THEY CALCULATE IT INSTANTLY. Everything is based on the accelerometer (poorly) to calculate distances. That's why when you increase the speed in the Walk activity, it can't measure correctly, and when you decrease the speed in the Run activity, it can't measure that you're walking. The accelerometer is basically waiting for you to move at a certain speed and in a certain way to measure "correctly."

Reviewing the activity, I can see a lot of fluctuation in the graph. It's not until I start running that it normalizes.

Why do I say it's bad? Because any movement outside of your activity that doesn't meet the algorithm's expectations will affect your route and reduce your actual distance, for example: looking at your watch screen frequently, checking your phone, adjusting your headphones, adjusting your hat, drinking water, and practically anything that doesn't require you to stop and requires your hands (which is why looking at your watch makes your cadence disappear).

For me, it perfectly explains why the Suunto Vertical never fails and its distance is always the same. It measures it through the map and geolocation (it doesn't calculate). It cares little or nothing about the accelerometer (hence the step differences). For the Vertical, it's irrelevant whether you're running or walking; it measures distance (GPS) and time.

Fenix ​​7x

Firmware Version: 21.10

Distance: 6.18KM

Difference: 480M

Steps: NA

connect.garmin.com/.../19013097170

Hand position: Left

Suunto Vertical

Firmware version: NA

Distance: 6.66KM

Difference: 0

Steps: 5476

maps.suunto.com/.../6816215546e1bd4b419cb6aa

Hand position: Right

Well, you might ask: Mark, then what's the point of GPS in activities? And I'll tell you: to take a nice picture of the route.

That's the bad part. I doubt Garmin will change the way it measures routes, and I partially understand why (it's easier to put effort into the accelerometer and its measurement algorithm than to support all maps for all cities (or pay for it)).

What should be ideal?

Distance measurement should be based on the map. There could be fluctuations due to poor coverage, but in that case, it should use an accelerometer.

Now, why do I mention that all Garmin devices suffer from this? A few days ago, I was able to test the famous and innovative Fenix ​​8, and yes, it does exactly the same thing. Despite its brand-new software (I must confess, I loved it and it took me a long time to return it, but for the same bug, no thanks), it has the same behavior.

Fenix ​​8 51mm Solar

Firmware Version: 13.37

Distance: 7.54 km

Difference: -160 m

Steps: 9212

connect.garmin.com/.../18941563173

Hand position: Left

Suunto Vertical

Firmware version: NA

Distance: 7.70

Difference: 0

Steps: 8163

maps.suunto.com/.../680cf11c3d344622c3241fc6

Hand position: Right

Do you use a hr chest strap?

No 

Garmin-Laurie? Is that a known problem?

See thread in current beta area:

https://forums.garmin.com/beta-program/fenix-7-series/f/community-discussion/407556/distance-measurement-inaccurate