How does FitBit Charge 4 actually calculate distance in Run and similar modes?

[Dr1v3n]

Per https://help.fitbit.com/articles/en_US/Help_article/1141.htm#distance :

"When you track an activity with GPS, your device calculates your distance using GPS data rather than steps. If you begin moving before you get a GPS signal, your device calculates distance using your steps until GPS connects."

However, I've found this to not be the case. Recently, I accidentally set FitBit Charge 4 to "Run" when I went on a bike ride. My bike ride was about 5 miles, but after I completed the workout, FitBit Charge 4 logged a 1.99 mile workout instead, but it had a GPS map which showed about 5 miles and was very accurate.

Why on earth does this happen? Why would it not just use its own GPS map, which has the distance travelled, to calculate my distance like the official answer above says? Something must be wrong here. Either this is a bug, or that answer needs to be fixed because I've conducted multiple tests (such as also going on a hike with my friend) and seen large inaccuracies which clearly show that GPS is NOT being used to calculate my distance travelled.

Please provide a more thorough answer of how distance is calculated, or else I am going to have to extract .TCX files every time I do workouts and just get the raw data from GPS.

[Rich_Laue]

Hi @Dr1v3n us fitbit users may not be able to speak for fitbit. However with some trackers we are told that both the stride/step count and GPS is used. 

[JimWalks]

I agree with Rich_Laue.  It makes sense that in cycling the 'stride' of peddling is only about .398 of your normal stride (.398 X 1.99 = 5).  Check in your Personal Info Advanced Settings if your stride length is 'Set Automatically' or 'Set Your Own'.  If 'Set Your Own', I would bet that the Fitbit probably ignores GPS distance and only used your stride length.

[Dr1v3n]

This is a great point, thank you for pointing this out. I have set it to Automatic (I have no idea why it was set to manual because I've never consciously measured or set my stride length before...) and will report back here my findings if this changes.

But to the FitBit staff, that observation above means that the page I cited in my original post[1] has incorrect information on it, because according to that page, GPS is used to calculate the distance when it's enabled. In fact, I've seen other official instructions elsewhere which say that if e.g. one is pushing a baby stroller while walking, it's better to use the Exercise app to start a "Walk" exercise so that GPS is used rather than steps/stride length, but clearly that's not the case as seen above.

1. https://help.fitbit.com/articles/en_US/Help_article/1141.htm#distance

[Rich_Laue]

A bicycle does not have a stride, your tracker on the arm can not count pedal strokes. (at best the tracker will count the bumps in the road that the bike rides over)

How far the bike travels per pedal will be affected by the gear ratio. Then we would have to add the distance when coasting. 

The gps, for distance on a bike is handled different then during a walk or run. 

Auto detected exercises will not display a distance. 

[Dr1v3n]

---

@Rich_Laue wrote:

A bicycle does not have a stride, your tracker on the arm can not count pedal strokes. (at best the tracker will count the bumps in the road that the bike rides over)

How far the bike travels per pedal will be affected by the gear ratio. Then we would have to add the distance when coasting. 

The gps, for distance on a bike is handled different then during a walk or run. 

Auto detected exercises will not display a distance. 

---

Right. I follow what you are saying, and appreciate the conversation. Assuming the assumptions made here are correct, I have 2 main points:

1. The information posted on the aforementioned page under the question "How does FitBit calculate my distance traveled?" says "When you track an activity with GPS, your device calculates your distance using GPS data rather than steps. If you begin moving before you get a GPS signal, your device calculates distance using your steps until GPS connects." This indicates that any time GPS is enabled for any tracked activities (via the Exercise app), GPS is the primary source for distance traveled. This is not the case.

2. Regardless of the above, I would expect that the FitBit would not be totally dumb and notice that "Hmm, something doesn't add up because we have 2,000 steps registered but the GPS track shows the user went 15 miles." There is no human that will go 15 miles in 2,000 steps.

Point #1 should not be understated. That guidance, which I have seen across multiple Interfaces from FitBit, suggests that GPS is the primary determination of distance when it is activated, and again, everything discussed here and on the following 3 feature suggestions:

1. https://community.fitbit.com/t5/Feature-Suggestions/Use-GPS-First-and-Steps-Second-to-Calculate-Runn...

2. https://community.fitbit.com/t5/Feature-Suggestions/Make-run-distance-based-only-on-GPS-data-at-leas...

3. https://community.fitbit.com/t5/Feature-Suggestions/Edit-distance-activity-based-on-GPS-or-steps/idi...

The purpose of this topic was to clarify exactly how this is calculated and also point out that the guidance on the "How does FItBit calculate my distance travelled" seems to be blatantly incorrect. Of course, I should not have done a bike ride while in "Run" mode, but it should be noted that this is what made me realize that GPS is NOT being used to calculate my distance travelled, because GPS doesn't care whether I am striding or riding or what.

Also, it's a bit odd that the FitBit App actually lets me change my exercise type after the fact, so I changed it to a Bike workout, but it still says 1.99 miles... So there must be either a bug in the software or the information about how distance travelled is computed is incorrect or could at least use some clarification. I think @JimWalks had some great insight here into what was going on specifically in my example.

[Rich_Laue]

With android only auto detected exercises and manually started Workout mode have an editable exercise type. 

Yes I have not been looking at the help docs. Why? Because real world testing by users and you have pointed out - fitbit uses more than gps to determine distance. Plus the moderators do agree. 

This is the reason for the requests you have found. 

It may be that gps is the primary but fitbit also factors in steps. I find with my stride set properly, I get a better distance report them the Active 2, which does only use GPS. 

Since this is not about one specific hardware tracker, but want a discussion on how gps is being implemented, I have moved this thread to a discussion board. 

[4gary]

You said it yourself... "but it had a GPS map which showed about 5 miles and was very accurate.". Therefore  use your gps data like it was intended.