GPS distance totally wrong on Charge 3

[BertilDator]

Was doing cross country skiing, and set activity to "run" (as there is very limited activity list). Had phone connected, and got perfect GPS route on Fitbit activity. However, all the splits are way off. Total distance in Fitbit shows 4.8km while real distance was 7.5 km (I know the route, and even double checked via ontegomap.com).

How can Fitbit calculate distance so wrongly even though the GPS route is perfect on the activity? I repeat, the tracked route is 100% correct on the Fitbit activity in question, so there is no disconnections, or trees, or anything between the phone and GPS satellite, nor between phone and Charge 3.

Is Fitbit using GPS route for distance calculation when assisted GPS is enabled during "Run" activity, or does it STILL use stride length?

Should I use "Biking" activity for XC skiing to get accurate distance based on the GPS route?

[NAuld]

Yes. I’m having the same problem. FitBit Charge 3 isn’t accurate with my runs. I’m having to use Strava or RunKeeper for accurate distance.

[Nuuts]

I've the same issue as well; C3 showing completely inaccurate distances.

Today I went for a 5.2 km run. I know the distance is 5.2 km and it was correctly tracked by my sisters' Polar and Garmin trackers (as well as Google Maps). 

Fitbit is showing 6.97 km, ie. adding 1.77 km to my run. 

This happens all the time and I've reported this to support who are not being helpful at all.

Looks like I'll be going for Garmin and throwing Fitbit into the bin where it belongs. So disappointed. Would never recommend Fitbit after this experience. Its performance is definitely not what I expected and what was paid for.

[SunsetRunner]

OK, went for walk today of around 9k, tracked with Charge 3, Fitbit app and Samsung Health (I'd normally use Runkeeper,  but that's been playing up since last update).  Three different readings of 8.83k from Charge 3, 9.12k from Fitbit app and 9.35k from Samsung Health.  I can understand that there might be some difference between Fitbit and Samsung Health, but I would have thought that Charge 3 and the app would have been a bit closer, given that they use the same GPS. 

I changed the stride length in Charge 3 to 1.5m so, given a step count of 11,116 steps from the Charge 3, I would have expected a distance of 16.7k.  It would seem that stride length and step count have no bearing on the outcome. 

Definitely would not buy another Fitbit device, certainly wouldn't recommend them to anyone else and would even be very wary about buying a similar device of any brand in the future. 

[leftyx]

For my "daily" run I run 5km and walk back home, I've changed the approach as the FitBit distance reading has been inaccurate, now using a fixed endpoint (a police station on the route).

Also note FitBit app has been updated to 3.20, which is the latest available for Android. 

I've been tracking the run using FitBit charge 3 connected GPS (waiting for it to be connected before starting the run) and strava, and comparing the FitBit .tcx output in GPSTrackEditor.

Date	distance	time	pace	steps	distance	time	pace	time	no# points	distance	no# of points	distance
12.05	5.54	26	4'42	4250	4.88	26:11:00	5'22	0:26:09	1575	4.81	1573	4.81
13.05	5.06	26	5'13	4283	4.88	26:19:00	5'23	0:26:32	1574	4.79	1567	4.79

So for the exact same run, Strava records exactly the same distance where GPSTrackEditor (using FitBit GPS data!) shows a >1% difference. However, FitBit shows a more than 8% difference in distance (which also causes a ~11% distance in pace calculation). 

The FitBit duration and number of steps are pretty similar, as are the Strava duration and GPSTrackEditor duration and number of GPS trackpoints. This all indicates a very similar run and the FitBit distance calculation is incorrect (resulting in an incorrect pace calculation).

[t.parker]

@leftyx this is because the distance is not based on the GPS data only but on steps and stride. This is how Fitbit designed that and since it's been working like that for many years I don't expect it's going to be changed. True, it makes live stats useless for runners but Fitbit devices were never designed as sports watches.

Simple test done with Charge 2 (also using Connected GPS):

Just stepping in place increases the distance. GPS is recording, too but doesn't seem to contribute to the distance. Steps and stride, on the other hand, do contribute. I don't believe this will ever be changed to work differently.

[Procter101]

Hi, can you explain how my fitbit records 11000 steps and 5.22 miles of activity I. E no choosing of activity on the fitbit when actually I probably only walked 100 yards but rode 32 miles at 11.7 mph. 

[t.parker]

@Procter101 if you rode (bike, motorbike) then it picks up bumps in the road as steps. Even when tracking Bike activity still the steps detection is not blocks and any bumps or uneven road will cause adding extra steps. It's just a side-effect (not a feature, nor bug).

[SgtPepper]

---

@t.parker wrote:

@leftyx this is because the distance is not based on the GPS data only but on steps and stride. This is how Fitbit designed that and since it's been working like that for many years I don't expect it's going to be changed. True, it makes live stats useless for runners but Fitbit devices were never designed as sports watches.

I agree that this is the behavior we observe - although as someone pointed out earlier, Fitbit support does state the following:

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, this statement appears to be incorrect and, as you said, stride length is being used. Having said that, stride length should be being updated automatically during every GPS tracked walk and runs (Advanced Settings->Stride Length->"Set automatically" is enabled in the Android app), and so perhaps this is what they mean by saying distance is calculated by using GPS data (albeit indirectly)

Either way, it is inaccurate. 

On the plus side I'm definitely getting more accurate readings than I was a week ago - but now it is overestimating my distance travelled by between 5-10% rather than underestimating by 15-20%. 

[t.parker]

@SgtPepper the problem with Fitbit support is that they don't really know how it works, nobody knows and very often they give contradicting answers. That's why I tested it myself with many runs while I had Charge 2 and Ionic (one with Connected GPS, the other built-in GPS). The captured route usually (if GPS worked all time during the run) was accurate enough to provide the correct distance. However, the distance reported after the run (and during as well) was always wrong. The thing is before I was into running and started serious training I didn't even notice there was something wrong. Back then, I just tried to lose weight and get more active. After reaching a certain level of fitness when things "got serious" I noticed something was off. I started running with Strava, Endomondo, and Fitbit running at the same time, then I got another sports tracker and everything matched except Fitbit. I think Fitbit may be relying on such users who simply are not so much into data and sports that some inaccuracies are acceptable for them (or even invisible). The simplicity helps improving battery life which is something that the majority of users want so it's a trade-off. This is my guess based on years of using Fitbit and interacting with the support. After moving on and getting a proper sports watch I see how different the audience is for both products. My sports watch software doesn't support logging food, sleep tracking is very simple and probably it never occurred to the developers to provide things like women's health tracking or something from Ionic - Bitmoji. The users would die laughing. It's a matter of audience. That's just my theory, nobody has to agree 😄

[Procter101]

Thanks for this bumps info
I still cannot get my head round 11.000 steps or bumps and 5.18 miles vs 32
miles actual.
Thanks again

[HansimGlueck]

@t.parker 

The problem (in my opinion) is, that Fitbits support is implying, that the device can be used as a proper sports watch because GPS is used to track runs.

And that just isn't the case.

And noone from Fitbit support cares....

[Procter101]

How true and they do not care!!!

[t.parker]

@HansimGlueck not sure how Fitbit is nowadays advertising their devices but GPS doesn't make a proper sports watch. There are a lot more features needed like being able to use external sensors, more stats (cadence, running power), power management that allows GPS run longer, in-watch navigation, uploading routes on the watch, custom sports, customizing activity tracking screens, better support for intervals, multisport support, triathlon support, on-watch training plans, better swimming features (still, there is no open water swimming nor being able to track HR when swimming), on the software side better analysis tools, etc. This is just a small subset of features that I find useful and I use it almost daily. The majority of popular sports watches offer a lot more. But if you look at Galaxy Active or Apple Watch series - these are not sports watches at all, yet they both have built-in GPS. I tried to use Galaxy Active and found it alright but not for me (major dealbreaker is not being able to connect external sensors). If you look at the features of C4, it's no so different from C3 or even C2. Over the years, Fitbit developed lots of new features, some even improved but sports tracking got stuck years behind to treat Fitbit trackers as serious sports watches - this part of the market is already lost for Fitbit unless they will come up with something incredibly advanced at this field. C4 is just an activity tracker with built-in GPS. Sports tracking is very poor. At least, this is how I see it (but I may be into sports too much ;)). That's why I don't think Fitbit even has a reason to fix such issues 😞

[t.parker]

Ok. Just to update something. I totally don't believe the GPS contributes to the distance. I took Charge 2 for a run together with another GPS sports watch. I ran 10.5km. Charge 2 ended up with 8.32km. The map recorded with GPS shows the correct route of 10.4km (after uploading TCX to mygpsfiles the distance is calculated as such). it is an acceptable difference (10.4/10.5). However, Charge 2, with connected GPS was showing wrong data during the whole run. Here comes the best thing. I stopped for a while and took snapshots of both watches. Then I started running on the spot. Surprise, surprise - the better sports watch didn't budge - I don't move, distance doesn't change. However, Charge 2 started adding up the distance. This is enough proof that GPS doesn't contribute to the real-time data (distance, pace) at all. Then, the data from the watch is synchronized to the Fitbit account. So what the GPS is used for? I guess only to capture the route to display the map later, nothing more.

Here's a short video showing how it works (the same thing was going on with Ionic, despite built-in GPS):

https://www.youtube.com/watch?v=o0xnCfeWd-4

On the left hand, 2.51km stays 2.51km while on the right hand 1.91km turns into 1.95km. Looks like I ran 40m on the spot.

As I said, you may have the direct evidence and still, the only advice is going to be rebooting your watch 🙂 If one is a serious runner (or an athlete) then Fitbit is one of the worst choices.

[Sandles]

My Samsung Gear Fit 2 Pro, or Galaxy Watch both might not be a 'Sports
Watch' in your book, but they both cover a majority of the 'SportsWatch'
feature you claim - certainly enough for a Runner or Rider (that knows
where they are going) including Cadence with assisted GPS.

Not sure how on screen navigation makes it a 'sports watch', except for
example Golf or Mountain Biking, but they also have dedicated watches in
their own right.

The Gear Fit 2 Pro is similar in size, albeit a touch thicker, but it is
older and my purpose for bringing in the Galaxy

If the FitBit Charge 3 was marketed as just a pedominator and notification
tool i wouldn't have bought it, but isn't FitBit all about Sports and why
can't something as simple as a Charge 3 be more of a Sports Watch than a
Samsung Gear Fit 2? Wish Samsung's small GPS/sportish devices weren't
discontinued as the Charge isn't taking the opportunity to apply to a niche
market that doesn't want a bulky watch for exercise and could wear
discreetly 24/7 whilst not making expensive dress watches redundant.

I'm seeking to confirm whether the Charge 4 will focus on GPS to track
Distance (why else would it have it) as it isn't for navigation, but if
FitBit were logical, the Charge 3 would use Connected GPS to measure
distance and have consistent output with other apps.

[leftyx]

@t.parker @SgtPepper agree that the FitBit distance is not calculated from GPS data (regardless of FitBit support stating uses GPS data) which raises a question on the accuracy of the automatically calculated stride length.

Looking at the last two runs (same start/endpoint) there is <1 difference in number of steps (FitBit) and <2% difference in duration (FitBit rounds up to minutes so looked at .TCX data observed in GPSTrackEditor) which indicates the runs were pretty similar. Yet stride length has almost 10% difference  (1.30 and 1.18) between the two runs. 

I've now set stride length manually (average of (GPSTrackEditor distancde / FitBit steps)) and will check after my next run if this makes a difference.

Note I previously tried setting stride length to a very small value but it didn't seem to make a difference. I guess the set stride length wouldn't be considered if GPS is used to used to calculate stride length to (indirectly) calculate distance, but I don't understand why FitBit would calculate distance in that way. 

[t.parker]

@Sandles it could be that I just require a lot more but I must admit, I'm a data freak so I like a lot of data and excellent accuracy. That's why external sensors support is so important because wrist HR never works for me. Then, for example, connecting the foot POD so if it comes to run on the treadmill (not very often but foot PODs work great also outdoors) I collect accurate data, too. This helps me to plan trainings. The same for cycling - I use external sensors to obtain more data. I realize, however, that I may be a minority here in terms of requirements 🙂

Let alone smartwatches (I know with additional apps they can elevate sports tracking functionality), the Charge series is sports-watch-wannabe at best considering very poor sports tracking and lack of accuracy caused by design choices. What I'm trying to say, complaining about distance measurement by Fitbit devices will be just complaining as the method of measurement won't change. Fitbit kept itself to this design choice for a very long time and has no real reason to change it. I was battling it in the past and my Ionic died before anything useful happened. Of course, a small device like a Charge could be a great sports watch. Fitbit just chose not to make it so. Products are marketed in a way that will bring money to the company. As consumers, we should always do the full research and pick the one which meets the majority of our requirements ignoring the marketing. The availability of features that we need is more important than the name of the company on the product. Whatever we call it, sports watch, smartwatch, tracker - as long as it does what we expect it to, that's the right choice 🙂

About built-in GPS Charge 4 seems to be doing exactly the same - steps and stride rather than GPS. I don't own Charge 4 (and really, don't want to invest my money after the disappointment that Ionic was), but looking at the forum, there are threads complaining about the distance. As an ex-owner of Ionic, which also has builtin-GPS, the GPS worked the same way - just to store the route but not contributing to distance computation. It would be best if someone who owns Charge 4 did the same test I've done with connected GPS - see whether the distance goes up when skipping in place. The guy in this thread:

https://community.fitbit.com/t5/Charge-4/Exercise-Stops-Randomly/m-p/4242045/highlight/true#M5558

may be onto something (the exported TCX files show the same thing as the ones I exported using Ionic in the past - distance goes up regardless of GPS coordinates).

[GinPhysicist]

I have a charge 4 and have been getting the same issue. I got the charge 4 with the past week and haven't had the chance to test all the settings yet. However I have gone on long walks with my dad who has a far more expensive and sophisticated Garmin watch and by the end of the walk (we both start at the same time and end at the same time) my distance is 20% longer.

I haven't tried exporting the gps files yet but cant figure out why this issuing is happening.

I have been however, taking my phone with me so I am unsure if it is actually using the watches gps or the phones gps (I also have a Samsung galaxy s9+ as people have previously mentioned that they have been having issues with android phones).

[GinPhysicist]

PS. I did my stride length × steps taken and it did not equal the total distance travelled so the distance being displayed is being (incorrectly) calculated by the GPS.

11720 (steps) × 0.69 (stride length) = 8086.8 metres

Total distance travelled = 9840 metres

[leftyx]

Following up I did another test after setting stride length manually (average of (GPSTrackEditor distance / FitBit steps)) - this seems to provide a much more accurate (~4% difference to GPSTrackEditor calculated distance) result for FitBit calculated distance. As @SgtPepper mentioned, it seems to auto-calculate the stride length (assuming using GPS data) and then use that as input for the distance calculation. 

There is another issue there - the calculated stride length seems inaccurate. Between 3 of the same runs there is only a negligible difference in number of steps (<2%) and duration (<1%), but a significant (4-15%) difference in stride length (FitBit distance / FitBit steps). Not sure why (might be a small sample rate?).