Blaze Distance Measurement is Inaccurate

[MuzDownUnder]

I have been testing the Blaze vs the Surge on my last two walks. The Blaze is not reading distance correctly using Connected GPS. Here are the detailed results from the last quite slow walk.

Unit              Blaze    Surge           Notes

Time             41:50    41:42.22      I pressed the two starts and stops within 0.5 sec

Dist               2.59km 2.20km        The 2.20 is accurate - on both walks the Flaze was well over

Avg Pace      16'07"   18'55"          The surge is closer

Hrt Avg          86         80

Steps             2953     2920            The surge step counting is a little erratic

Elevation       36.57    36.57

My phone is a Samsung S5 running Android 5.0

[KrisTheRock]

I will need to test the accuracy of distance and pace tomorrow morning. Is it inacurate for everybody? The main reason I got the blaze was to see my current pace while running. If it's inacurate, I'll return it.

[tennistoo]

I have had my Blaze for two days and have tested it with a walk, a run, a treadmill walk and run, around the house, grocery etc.  I have used it along with my One, my surge and Runkeeper.  My One, Surge and Runkeeper numbers are all very close.  The Blaze steps are off by 30%, my distance is off by almost 50 %.  Heart Rate is way off.  It can not be blamed on the Surge being wrong.  This will be going back.

[KrisTheRock]

I did a very short test this morning. It's snowing a lot over here so I'll wait before doing a longer test. So far so good. Exact same distance calculated by Endomondo and fitbit. They were both using my Nexus 5 gps. Heart rate was also pretty close. Endomondo was a basis peak for bpm.

[Cracker65]

Not wanting to labor on the point of Blaze distance calculations, but I started recording these in a spreadsheet so that I might be able to calibrate the Blaze distance calulations using the stride settings. I am still maintining an error factor of around 11% between Blaze & FitBit even when adjusting my stride by 15cm (6"). Needless to say that the 'Connected GPS' as advertised is only good for showing on a map where you have been, the Blaze's info is useless at the moment except for recording steps and HR.

Manually drawn map from Goole Maps shows distance ~4.85km

[Cracker65]

@KrisTheRockYou might try doing a longer walk than 5mins, see how the accuracy deviates then

[KrisTheRock]

Here's another test. Still short but more than twice the distance and still accurate. It should be warm over here tomorrow so I'll go for a 5k run to really test it out.

I also wanted to test your theory about the gps being ignored for pace and distance so for one minute, I started jogging at about 180 steps per minute but without going fast at all. I was going slower than when I was walking and the pace was reflecting this. If it was only based on steps, since I was doing about 180 steps per minute, I would have seen my current pace go much faster, at about 10 min per mile. Instead, it went slower, to about 25 min per mile. This, and the fact that the total distance is very close, indicates that the gps IS being used for distance and pace calculation.

Thoughts?

[KiloLima64]

Here is my take on all of this. From what I can see you're talking about 10% difference. Most civilian GPS unit are only accurate to about 10-15%. Cell phone GPS is even less accurate. If you are looking for something super accurate then you might need to look at a more expensive product. Most trackers are within 10% which is an acceptable variation. Just my 2 cents.

[11butterscotch]

That's ridiculous to accept a 10% variance, most trackers including Fitbit advert a 5% variance, and I've seen gps accuracy for device and phone listed as 95% accuracy.

[DavS]

I just did a quick comparison between some data I took with my Charge HR using connected GPS via the fitbit app, and data from my Blaze using connected GPS but initiated as a "walk" via the exercise screen on the Blaze itself. I went into google maps and used the "measure distance" function and found that the data from my Charge HR was almost dead on, while the distance given using my Blaze was about 25% too short.  I am wondering if there is a difference between the connected GPS between initiating it from the Blaze or from the fitbit app?  I will try another walk later this afternoon with my Blaze, but initiated from my phone instead and see if it is more accurate.  For those of you having trouble with the distance, are you initiating the exercise from the Blaze or from within the fitbit app?  Could the Blaze be modifying the distance somehow?

[ChaneRZ]

Blaze is indeed not accurate. Made two comparison and used two phones to take out the errors. Iphone 6s plus with nike app running and Iphone 6 with Fitbit Blaze connected GPS. Blaze is under counting and it would be big if I ran farther. Nike 3.31 miles VS Blaze 3.0 miles.

[SunriseGirl]

When I was shorted distance, I had initiated it from the Blaze. I took another walk yesterday that was indicated from the blaze....seemed much closer to the distances I always got from my charge HR indicated from the Fitbit app. Tomorrow I am going to take the exact same walk I did yesterday but will initiate the connected GPS Fromm the app and not my Blaze and see what happens.

[SunriseGirl]

I meant to type initiated, not indicated.

[Cracker65]

In all my cases the FitBit measurements have been initiated from the Exercise 'Walk' from the Blaze. And I purposely wait until I see three bars in the top corner indicating the connection strenght before starting.

[DavS]

I just went out and did the same walk as I did earlier, except I used the fitbit app to initiate the walk. It's pretty cloudy and there is a little bit of rain, so I'm not sure how accurate the GPS in my phone is right now (Galaxy S6), but I got a distance of 0.9 miles this time versus 0.77 miles when I initiated it from the Blaze. When I trace out the walk on Google maps, I get 0.98 miles (5197.57 feet), so it looks as though I still got shorted by a bit but it is much closer to the actual distance. Actually, looking at the route, it had my end point about 140 feet closer than actual, so if I add that in I get around 0.93 so it is off by only 0.05 miles or so.  So it does look as though there is a distance discrepancy between using connected GPS initiated by the Blaze and by the fitbit app.  Tomorrow I will try to edit my stride and see if it changes the Blaze distance.

 

[DavS]

Interestingly enough, the 0.9 miles walk initiated from the map only credited me with 964 steps in the "impact on my day" section. The 0.77 mile wrong distance walk gave me 1231 steps. Weird. Definitely some bugs to work out!

[DavS]

Sorry, "map" should have been "app".

[KiloLima64]

GPS 101. You will get 95%+ accuracy in GPS if you triangulate with at least 3 points (satellites). 3 points is not always possible and most of the time isn't due to obstructions and such. With a 2 point (2D) fix your accuracy will be about 10% off. See bolded note below, your talking about somewhere in the area of 3.5 meter accuracy. The GPS in most phones don't tell you how many satellites they are locked onto to to get your fix. Again if the hang up is the fact that the unit using your phones GPS is off then just use the tried and true method with advanced math. 

Calculate your stride length walking and running: 

CALCULATING YOUR STRIDE LENGTH

1. Go to a track or some place where you're sure of the distance.
2. Count your steps as you walk across that distance, making sure you travel at least 20 steps.
3. Divide the total distance (in feet) taken by the number of steps to get your stride length.

Your running stride can be calculated the same way, only by running a known distance rather than walking.

Now take your stride length and follow this formula:

 There are 5,280 feet in a mile. So, divide 5,280 by your stride length figure to get your number of steps in a mile. If you want the answer in kilometers, divide 3,280 by your stride length figure to get your number of steps in 1 km.

Since you know the distance you are walking just do the math and that will give you accurate step count to compare your unit to. 

Personally as long as I am in the ball park I don't care if there is 15% accuracy difference. I'm not training for a marathon or anything like that. If I was I would be worried more about HR and time to cover a distance and other factors. I'm just using my Fitbit to keep me motivated and give me some sort of idea of how far I walked and how many steps I took and if I reached a particular goal. Again as I stated if you don't find this unit accurate enough get one that you are happy with and keep it. The bottom line is we all get some activity and or workout and improve our health and fun doing it so that we can all be here a long time to kick the can.

Note: From the GPS.gov webpage 

The U.S. government is committed to providing GPS to the civilian community at the performance levels specified in the GPS Standard Positioning Service (SPS) Performance Standard. For example, the GPS signal in space will provide a "worst case" pseudorange accuracy of 7.8 meters at a 95% confidence level. (This is not the same as user accuracy; pseudorange is the distance from a GPS satellite to a receiver.)

The actual accuracy users attain depends on factors outside the government's control, including atmospheric effects, sky blockage, and receiver quality. Real-world data from the FAA show that their high-quality GPS SPS receivers provide better than 3.5 meter horizontal accuracy.

[MuzDownUnder]

@KiloLima64 Great theory and possibly has some merit. It is negated though by the fact that I and others are running two or more apps on the same walk using the same GPS data on the same phone and only the Blaze is wrong.

[Cracker65]

The issue @KiloLima64 is not the GPS accuracy as all the apps on my phone using my phones GPS should be reporting the same distance as this is the single source (with the same error factor in effect). The FitBit Blaze is using the phone GPS and the activity is started at the same time as my other phone based activity trackers. The phone based activity trackers all report the same distance and time as I would expect. The Blaze initiated Exercise 'Walk' connected to my phone has an error of around +11% on every GPS connected walk.

So why is just the Blaze distance wrong?

I have tried making adjustments using stride/run settings to improve the calibration but this appears to have no effect on distance as I have demonstrated. I dont need to have it perfect but would expect to have it at least the same as what the phone is and not reporting >10% for distance.

[FelixFive]

---

@KiloLima64 wrote:

GPS 101. You will get 95%+ accuracy in GPS if you triangulate with at least 3 points (satellites). 3 points is not always possible and most of the time isn't due to obstructions and such. With a 2 point (2D) fix your accuracy will be about 10% off. See bolded note below, your talking about somewhere in the area of 3.5 meter accuracy.

---

If you're going to educate with your GPS 101 knowledge, I'd suggest you revisit the fourth satellite needed for trilateration.  Otherwise, I'd suggest retaking the course.