Surge Step Count Inaccuracy

[HappyFunBear]

I've had my Surge for a while and I did some initial callibration testing (versus counted steps and iPhone fitness app). 

This morning I retested it and it appears to be counting only about 2/3 of my actual steps.  In the morning I put my iPhone in my pocket not long after I put on my Surge.  Comparing after my ride to work:

fitbit   iphone  %
2389      3534  67%

Sanity tested by counting steps to 100.  iPhone got almost exactly 100, fitbit got about 70 (I was still walking so I didn't stop to do the exact math).

Got to work and restarted the fitbit.  Then walked around the building a bit:

∆fitbit   ∆iphone  %
296       446   66%

Seems to be consistently counting 2/3 of the actual count.  What can cause this and what is the fix for it?

NB: I only ever use GPS for biking and haven't in months, this is pure step count.

Thanks

[RobertoME]

Hi there @HappyFunBear, welcome aboard to our Community.  I found very interesting this comparison with your iPhone and your Surge. I'm going quite techy right now, but it will help to illustrate my explanation.

Note that your Surge uses accelerometer to understand your motions or movements. This information is turn into acceleration data, The tracker will provide detailed information about frequency, duration, intensity, and patterns of movement to determine your steps taken, distance traveled, and so on. All this is processed by a finely tuned algorithm or a software in your tracker for step counting. Here is the difference; If a motion pattern along with subsequent acceleration measurement data, meet the threshold established by the algorithm, the motion will be counted as a step. Otherwise If that threshold is not met, the algorithm won't count the motion as a step.

To make more clear my point, the iPhone uses a different technology integrating the sensors of the same phone to process this data. So indeed, the numbers between each device will be different at the moment to compare.

To improve the accuracy on your Surge you can specify whether you wear the tracker on your dominant or non-dominant wrist:

• The dominant wrist setting decreases the sensitivity of step counting and should reduce any over counting of steps when your body is not moving.
• The non-dominant wrist setting increases the sensitivity of step counting and should reduce any under counting of steps. Non-dominant is the default.

To change your settings, in the Fitbit app go to Account tab and choose your tracker at the top of the screen.

Hope this helps and let me know how it goes. 

[HappyFunBear]

Sorry for the slow reply.

After dealing with Support and getting a replacement unit I retested with that one and found almost exactly the same results.  Still seeing roughly 70% of manually counted steps.  So it had to be something idiosyncratic I was doing/not doing.

The answer I have come up with it that it's related to how much I swing my arm while I walk.  I retested while letting my left (non-dom) arm just sort of hang rather than swing and the results were almost 100% accurate.  I haven't retested the old one yet, but I am content that this is the explanation. Sadly though, it's not really a solution, since walking with my arm sort of limp is not normal for me (or most people).

Considering the size of effect I saw I'm surprised I didn't see anything about htis in the FAQ/standard support help.  It was a night/day difference in accuracy from just changing a minor behavior.

[shipo]

@HappyFunBear, the question is, which is correct?  I've compared my Surge to a good old fashioned pedometer and the results were always within a few percent; could well be your iPhone is overcounting steps.

[HappyFunBear]

Yeah, I wouldn't trust the iPhone numbers either if I hadn't done some manual counts to test it.  Most of the "tests" are just me counting steps for 100-200 and then comparing.  

After I had done a couple manual counts I took a longer walk (~3k steps) and checked against the iPhone. They were almost step for step the same.

So if I don't swing my arm much I get very good results, but the most naturally I walk the worse it gets.  Very strange, and I'm boggled that I don't see this problem referenced all over.  I doubt there's something really weird I'm doing just in swinging my arm.

Barron

[JimboHiker]

I had the same problem with my new Blaze - it was only counting 70 out of 100 steps.  I switched it to my dominant hand and the step count is MUCH better - between 95 and 100 steps out of 100.  I also tried it on my belt loop and got several readings in the 90-95 range.

[RobertoME]

Hi there @HappyFunBear and thank you for joining us @shipo. Also I have to offer a warm welcome to our Community @JimboHiker. 

It would never crossed my mind that swing the arm while you walk will affect the accuracy of the Surge. In fact I have never pay attention if this has an impact on my numbers. I think I will have to test it on my own to see why is happening this. 

In the meantime the recommendation I can provide is experimenting with how high you place the tracker on your wrist. This most of the times is recommended for heart rate accuracy, but I found that it helps too with my step accuracy. So moving the tracker up a couple inches can be helpful.

I'm going to test this theory and I will come back with my findings. Keep me updated how it goes and see you soon friends.

[Redmoonstar]

I had the same problem with my surge only it wasn't tracking correctly whether I swung my arm or not (I do swing it naturally when I walk ) I did several resets in a row and it did improve the step readings however whenever I take my dog for a walk it misses an average of 3000 steps for 7-8 klms. This happens whether I am holding the lead with either arm.. I bought this for hr and steps I also have a fitbit ultra which has been great and this + when I don't take the dog is how I know the surge count is out. This is a lot of steps to miss especially in a challenge!