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?

[DavidHermann]

... well, wait for the Software problems down the road with the GARMINs. I
gave up on them two years ago, my wife just now.
I understand that nothing is perfect in this world, ok. But I just hate
wasting time with electronic stuff that, for inexplicable reasons, quits
working after a few months.
BTW, my CHARGE3 worked fine yesterday after running in a different area. It
seems that the GPS signal is disturbed at my usual circuit. Will try run
somewhere else now and keep an eye on the thing ...
Rgds from Asia!
H.

[Sandles]

Is it a Software Problem? Fixing such a problem, that is causing a
fundamental capability to fail, for a device that has been out for this
long useless, should be fixed in a fifth of the time they've spent denying
and others have defended or shifted blame to phones, or that we should our
expectations.

Feel it's more a combination of incompetence, poor customer service and
misunderstanding of common sense and best practice.

Why should I run in a different area if all somewhat basic sports/smart
watches i've had for the past 5-10 years have been accurate and consistent,
with near identical distance over the same runs with both the apps they
came with and automatic or .gpx to Strava? All paired with the Samsung
Galaxy released at the given time.
- Garmin VivoActive
- Garmin VivoSmart #X
- Samsung Gear Fit 2 Pro
- Samsung Gear S3
- Probably another device between those i forget.

I even use a Garmin Approach s2 for Golf. Funny that i considered a Garmin
Ionic because it may be the only one that does Golf, but here distance is
even more important than running and the two Samsung's listed above remain
as accurate as the Garmin. Actually within 5 metre accuracy when compared
to mates using a lazer range finder! Guess the Fitbit would give me
another excuse for being 50-100m short of target.

Fun times

[ukAdamR]

What does seem to improve the Charge 3 accuracy is inputting your stride length manually. To determine this you need to accurately divide your route distance by steps taken, then input it into your Fitbit personal information profile. (This may have already been explained in the thread previously.)

First take a look at your activity history and find all of the items that are of the same regular exercise. These should all be of very similar step counts and time taken but, as per the GPS issue described in this thread, have wildly fluctuating distance travelled. -- Calculate an average step count figure, though you should be able to get a very good approximation just by looking at the list of numbers. It's important that you have a broad quantity of exercises logged to get a good step count approximation average. Try gathering at least 2 weeks worth of regular rhythm exercise logs.

Next you will need to have a highly accurate figure for your regular exercise. Either obtain this information from your Google Maps timeline, or for the highest accuracy manually draw your route as a path using the downloadable Google Earth Pro. (This application is now free for all Google users without subscription.) -- Depending on whether you're working in imperial or metric you will need to ensure your distance figure is in inches or centimetres instead of miles or kilometres.

• Inches = Miles x 63,360
• Centimetres = Kilometres x 100,000
• Or use Google to do this for you, e.g. search for "3.25 miles in inches".

So for my consistent daily walk route of 3.25 miles over an average of 5,400 steps my stride length is 205,920 divided by 5,400 = 38.133333. (Fitbit will only store 1 digit of fractional precision so this got rounded to 38.1.)

Now go and put that figure into your Fitbit personal information under Stride Length. (Change from Set Automatically to Set Your Own.) Make sure you use the correct walking or running input box according to whether your exercise is typically a walk or run, and just leave the other input as is. (You can come back to this later if you do both exercise types.)

After doing this I'm finding that the distance Fitbit claims my walk is within a much tighter range, from about 3.11 miles to 3.33 miles. This isn't as consistent as Google's measurements but it's a radical improvement, and definitely within the acceptable 5% margin of error for civilian grade GPS. What this suggests to me is that Fitbit isn't really using GPS properly at all and heavily relies on an accurate stride length being defined, which I'm sure others in this thread have already raised. It also shows that Fitbit clearly can't determine stride length from GPS accurately either, which is why our distances fluctuate so much.

This workaround does not vindicate Fitbit. Their software still has a fault with only a very small minority of customers finding a manual resolution.

• The automatic stride length feature does not work.
• Tracking your exercise distance by GPS only does not work.
• You should not have to do the above to tell Fitbit how to do its job. -- It claims to be self-calibrating but does not function fit for purpose.

I'm not saying this will definitely work for you, but hopefully this at least improves the issue. 🙂

[leftyx]

Hi @ukAdamR, thanks for the detailed explanation!

I found the same that setting the stride length manually produces a more consistent result - this aligns to the formula published on the fitbit site - Steps x Stride Length = Distance Traveled. However you are also absolutely right that the distance calculation using GPS is way off - according to the fitbit site "When you track an activity with GPS, your device calculates your distance using GPS data rather than...". At best, it seems that GPS tracking updates the stride length, which is then used for distance calculation. Fitbit really needs to be more accurate with distance calculation, and setting a manual stride length is a workaround - as long as we run consistently. This is where I have a challenge, sometimes I run by myself and sometimes with my kids (a bit more stop/start), so there is a ~15% variation in stride length (based on GPS track editor distance / fitbit steps)..

Also I received my Charge 4 last week and went for 2 runs at the same place this weekend. The build-in GPS absolutely screwed up the map measurement where it shows 1.65km (of 5.82) and 3.45 (of 6.15)km for the two runs. That is a different issue, but it does mean there is no reliable comparison with the GPS data (using GPS track editor). Comparing with strava however there is a 3% (5.68 instead of 5.82) to 10% (6.15 instead of 5.6) difference. As the runs were are the same place and route I'm inclined to say the Strava results are more accurate... 

[Springy.BitFit]

@ukAdamR that is pretty interesting way to do it and get what you trying to do. The problem is peoples easy/tempo/intervals runs (ok this is for running) stride length is completely different.

For me I just sync it to Strava and then let strava correct the distance as I know the GPS data that fitbit has is accurate, just they use the stupid stride length and any other secret crazy sauce they have to try and "make it more accurate".

[ukAdamR]

---

@Springy.BitFit wrote:

@ukAdamR that is pretty interesting way to do it and get what you trying to do. The problem is peoples easy/tempo/intervals runs (ok this is for running) stride length is completely different.

---

That would be a fundamental flaw with Fitbit anyway.

Unless it updated your stride lengths dynamically and constantly during every exercise it has no way to store/use different stride lengths based on different running speeds. We already know Fitbit does a terrible job at determining stride length based on GPS so I'd never trust it to determine a further breakdown based on your current speed.

It would have been a more sensible designed to use GPS for distance whenever possible instead of only for determining stride length, with a stride length fallback should GPS become unavailable. Both Android and iOS have location APIs they could use to determine this constantly so it's not like Fitbit need to have a custom implementation.

[samanchek]

Just wanted to chime in that I'm having the same issue. Just realized that it has been overestimating all my runs about about 8-10%. Doesn't seem like much but when you're training with specific targets in mind it is really disappointing to fund out you were running 30 seconds slower than you thought.

I haven't done as extensive testing as some others on this forum but if I tell Strava to "Correct" the distance, it consistently improves the accuracy significantly. This tells it to ignore the distance reported by Fitbit and calculate its own distance based on the raw GPS data. It doesn't really help when running because you can't get up to date pacing.

At this point I just want a refund and to find a new tracker. Has anybody had any luck getting a refund? I bought it 3 months ago from Best Buy so I don't think they will refund it.

Anybody have a good replacement?

[Dpcc]

Yep. Same problem as everyone here. I have a charge 3. I wanted to do a better job with splits so I tried map my run with the voice feedback option. (I am horrible at pacing and start much too fast and tire too quickly.) the Map My Run App was about 10-12% reduced

distance compared to Fitbit. I then checked it with my car. Map My Run is correct,

I have been using GPS with Fitbit by the way. Really a bummer to realize I was running 10% shorter than I thought as well as much slower. Horrible product.

[leftyx]

I moved to a Charge 4 earlier this month, first results were horrible but since firmware 1.96.29 (around 19 July) the results seem much more accurate. This firmware introduces the option to use connected GPS which I have disabled (due to inaccurate results with Charge 3 and it kinda defeats the purpose of having built-in GPS in the first place).

Two most recent runs using Charge 4 built-in GPS, following my usual route of just over 5kms;

Date	FitBit (distance)	Strava (distance)	GPS track editor (distance)
28.07	5.19	5.15	4.87
29.07	5.21	5.21	4.94

Both FitBit and Strava report 1-2% difference between the runs, and the difference between FitBit (built-in GPS) and Strava (phone GPS) is in that same range. Not sure why GPS track editor (using the TCX file from the FitBit activity) reports a 5-6% lower distance, but at least its also consistent with that. 

This is the accuracy we can expect from FitBit (or any product that does activity GPS tracking). 

[DavidHermann]

Since I am running more in open country (few trees) the GPS works fine ...
not sure tree obstruction was the problem ...

[ukAdamR]

Just coming back with an update from my 11th July calibrations... Looks like this is going to resolve the issue for the foreseeable future. I'm getting very consistent figures as of that.

(Ignore the red marked distances. They are walking to/from work rather than my typical 3.25 mile walking route.)

[dbdb1111]

I have exactly the same issue when using the biking activity.  Even if I wait for the the Fitbit to vibrate, telling me the app has found a GPS signal, ever since the last update, the distance is very inaccurate.  The last time I rode, I went 24 miles, but the Fitbit said I had gone 38.  The route shown in the Fitbit app is accurate, however.  Same issue - route is correct, distance is wrong.  I did not have this issue prior to the last update.  I have a Charge 3.

[PRCstuart]

So now I have a new problem.  I just got a brand new treadmill and when I put my hands on the detector, I get a completely different heart rate than on my Fitbit.  

I was running and my Fitbit said heart rate was in the 150's but my treadmill said in the 90's.  

Anybody else have this discrepancy? 

[PRCstuart]

Thanks for this!  I just looked up Strava and I'll give it a try. 

[GiacomoP83]

Same issue here: Charge 3. Always worked well, until last update. I wrote to Support and they confirmed that they know there is an issue and are working to solve it. I asked them if I can downgrade to previous version but they don't answered to the question (really unacceptable). It is absolutely frustrating to see a discrepancy of 1~1.5 km on a 10km run. I'm really evaluating to move to another tracker.... 

How is it possible that so many users have this problem and FitBit not even worthy of giving an answer? I don't understand....

[Procter101]

Can only say keep telling them not fit for purpose or lemon law.
If you can speak to support they are able to authorise replacement
My versa 2 replacement is fine.
God luck

[Nuuts]

That's exactly why I've given up on Fitbit. Always inaccurate and there's no support from the "support". 

I'm now a very happy user of Polar Vantage M. Very accurate. 

I won't go for Fitbit anymore; it really is a useless piece of **ahem**e to be completely honest.

[Pineconn]

I'm in the same boat. My Charge 3 has been super accurate with GPS distancing up until the recent firmware release. Now, my 4-mile runs are being logged anywhere between 3.7 and 4.3 miles, and changes in run speed are not being picked up. (In fact, when I drop the hammer on the last mile of my run, it thinks I'm going slower.) Clearly there's a GPS desync issue, and it's trying to use stride length to fill in the [massive] gaps.

This problem is frustrating since I'm a data geek and want my runs to be logged accurately! I've tried turning WiFi off during runs, re-pairing my Charge 3 to my phone (Bluetooth), reinstalling the Fitbit app, restarting my Charge 3, and clearing user data. Power saver is off; Fitbit app has all location permissions.

[Dpcc]

Yeah it is really frustrating. I am blown away that this is an issue for fitness tracker! I am furthest thing from an elite athlete but it is important for me to see how I am doing with my pace. I want to see if I am improving or not. Bummer to have trust in the Fitbit.

[geodepa]

Similar direction and magnitude of error here as reported by dbdb1111. Just replaced my recently dead charge 2 with new Charge 3. The Charge 2, while it lasted, was reliable and consistent in measuring distance and pace for GPS runs. On first GPS run with Charge 3 (with same iphone and app as before with charge 2), the Charge 3 had accurate GPS route data from my phone (i.e. the map was correct) but converted the actual 2.48 mile run into a 3.6 mile run, reporting a pace during and after the run that was about 2 minutes faster than my usual pace. I'm not particularly proud of my 12 minute pace, but at least with Charge 2 I knew what it was. Having read this and other threads on this topic, and seeing persistent similar reports with little hope of any solution, I will probably return the Charge 3 and switch away from fitbit family, which is a shame, as I really liked my Charge 2 and was looking forward to similar functionality and reliability, plus the waterproof feature of Charge 3, which as a swimmer could have been very useful to me. Users shouldn't be forced into elaborate and inaccurate manual workarounds because some wunderkind in development figured out a roundabout way to convert GPS data rather than just use that data directly for what it is. Writing this mainly for fitbit's benefit, on the off chance that they check here now and then for customer feedback.