Dramatically Under counting steps!

[PastorPaul1957]

My wife has had problems with her Fibit Blaze under counting her steps by 30 in a 100.  We have tried every solution to no avail, and frankly these forum boards have not been very helpful.  However, we have discovered that by her wearing my Blaze and I hers, that the problem is with her gait.  Furthermore, we have narrowed it down to her arm swing.  If she reduces her arm swing, walks with ther hand in her pocket or tightly held to her body, the count is pretty accurate.  This is a serious problem as you can imagine.  Even more telling is that if I exaggerate my arm swing while I walk I get a similar 20% or more undercounting of steps.  This is very frustrating because that impacts distance and calorie measurements.  When you lay down $200 or more on a piece of equipment, you don't expct it to be quite so inaccurate or to work so inefficiently.  Did the manufacturer in trials seriously not consider this problem or encounter it?  Obviously random vigorous wrist activity can elevate step counts, nominally, but a consistent day in day out 30-35% under count is significant, espeically on targeted waking exercises, and it makes the user feel like giving up on the product.  This was a Christmas present, so both of us are feeling cheated, and the hope that by doing this together we might improve our commitment to a healthier way of life is dwindling.  I would like to hear from Fibit, with more than a few platiudes and suggestions about resetting.  No objections to hearing from others, but I don't need 10 more experimental suggestions which we have already tried, and which are not realistic permanent changes to lifestyle in order to accomodate a device (like wearing around an ankle or shoving it down a sock).  Done that, been there.  So far nothing works except an unnatural gait for her.  She has tried three different products with the same results. Thanks!

[Matt_sydney]

Have you tried changing the 'stride length' under Advanced settings in the app?

[PastorPaul1957]

Matt,



We have set it on automatic, where as it had been on a manually set value
previously, but it made no difference. I can't wrap my head around whether
to use manual or automatic, and if manual settings, whether to set it
accurately or try a sort of work-around by choosing a value that
compensates. If I do that, and if it could work, I am not sure whether to
set it bigger or smaller than the actual stride.



My impression is somehow the way my wife walks the momentum of her arm swing
counter acts other movement, and produces a false negative on the steps.



Thanks for the observation. we may experiment tonight with stride length. we
spent a good amount of time and carefully counted steps last night figuring
out the other stuff I referred to in my earlier post. So I am pretty
convinced this is about her gait, and not a "fault" in the Blaze per se. It
may be a design or function issue that it responds in this way. She
apparently is not the only one!



Thanks Much!



Paul

[vgasmen]

Paul,

If you stand still and swing your arm you're wearing your Fitbit on you will also get steps!!!! My son discovered this. I have complained multiple times but to no avail. It probably is countering the steps.

Also, I've been on the treadmill on an incline holding the top bar of the treadmill so that my arm is NOT swinging and I've gone 3-5 minutes without getting a single step...ON A TREADMILL!!!! It's crazy!!

Roni

[Matt_sydney]

Great detailed reply. I think it must be gait. We had the same issue with my Dad. He has a bit of an unusual gait from a illness 30 years ago. He also use both hands and is ambidextrous.

We were able to get his steps very close to mine by putting the watch on his right wrist and measuring his stride length on repetive drills. Took some averages and used the info to manually adjust stride input. It worked for him. Best of luck. 

[Colinm39]

@Matt_sydney@PastorPaul1957When I was experimenting with the missing steps I accepted there can be an issue.

Interestingly I found, when you have your arm at right angles across your body and not moving the arm, just walking, looking at the watch for the brief time it is displaying the steps, it registered for every step..

[PastorPaul1957]

Yes this is what we are finding. if the arm is still, or moves little, it
registers every step, but if your gait is such that you're an arm swinger,
it will under count quite dramatically. My wife's arm swing is not wild but
it is quite a bit more than mine, I would say and arc of 12 inches or a
little more, when she is in her stride.

[PastorPaul1957]

Yes, it seems to go in both directions... sometimes undercounting as in this case... but others are having an issue with over counting... when I wash my hands by rubbing them together under the water, I get steps... I even wake up with steps!  But mostly this is incidental no more than 250 in a day.  My wife is losing 30 steps in 100, or 30% consistently when she walks for exercise, and when she compares to a friend at work.

[PastorPaul1957]

Matt,

I am going to try this today.  Thanks!

[Fitmommagoals]

I'm having the same problem!  Nothing I've tried works.  I've tried all sorts of suggestions and even when walking side by side with friends and coworkers with the same device my steps are substantially lower than theirs!  I want my money back!!!!

[PastorPaul1957]

Matt

I tried the stride adjustment... It helped marginally but then that changed the distance calculation.  Still under counting.

[Matt_sydney]

Sorry to hear that. One step forward two steps back kinda thing. Been thinking about it and don't know what else to suggest. If it works fine on your wrist, but not on your wife's wrist I would then think it's not the watches issue, but your wife's gait, arm swing and stride. Return it for something other than a Fitbit. You will then know if it's a universal or isolated tracking issue. Best of luck 🙂

•sent from iPhone•

[Colinm39]

@PastorPaul1957Stride adjustment has nothing to the steps. It has with another issue I have posted here because I found a distance issue moving at domestic speeds..

Here is a test I did the other day for a Group I'm a member of and proved there is nothing wrong with the Blaze, if worn where it will register the steps 100%. This comparison was with my 2nd Blaze and my ever accurate Fitbit One.  It shows 7% less steps on the wrist.

Speed 5.2 kmh or 3.2 mph, about walking for exercise speed.

Fitbit One.. Waist... 1,823 steps....
Blaze 2 .. Ankle ..... 1,848 " "
Blaze 1 .. Wrist ...... 1,720 " "     -7.0%

[PastorPaul1957]

I think it is a bit of drawback to own a watch that you can't wear in your wrist!  🙂  My wife is actually carrying hers in her pocket today to see if it does better.  But that hardly seems to be the point.

[PastorPaul1957]

In my wife's case she is experincing dramatic under counts.  In my case I do notice occasional under counting, but more often over counting because of movements interpreted as steps.  This is not dramatic and the odd 200-300 steps in an entire day is not going to hurt much.  But a 7% drop or in her case 30% loss of steps has a marked effect on the overall stats at the end of the day.  That comes out to more than a mile distance in her case.  Most of us in other arenas would not accept a 7% margin of error, and certainly not 30%.  I don't know what to try next, or if just to give in.  

[SonzTwin]

I have a different but related problem of under-counting steps.  I've found all Fitbits (3 models I've used, incl. the current Blaze) to be pretty darn accurate counting steps when I'm walking; but they become wildly inaccurate when counting running steps - the faster you run, the more it undercounts.  The following are estimates of my pace, but the under-counting ranges from 5% at a 10-minute mile pace to 50% undercounting at a 6-minute/mile pace - MY idea of a sprint :-). I check it by manually counting my own steps and mapping it to what my Fitbit tells me; I've done this numerous times to eliminate any one-off counting errors, and the results are consistent.  It doesn't matter if you set it to "running" exercise mode - the under-counting persists, both within the "run" mode and on the grand counter.  Since I've found this in 3 out of 3 Fitbits, I'd think this is a design problem, which needs to be fixed, or label it ONLY as a step-counter for walking, and DON'T have a "run" mode!  That said, I'd love to see more evidence: could owners of Blaze do the same manual counting while running fast and compare that count to what the Fitbit tells you, and report back?  Thanks.

It does not bother me that when you swing your arms without moving that it counts your steps, and not even the non-counting when you're on a treadmill and your arms aren't moving cuz you're holding on to the bar.  The solution is simple, especially for Blaze, where the main unit pops off easily - just put it in your pocket, with or without the band.  Also, to confirm what someone says, length of stride should have nothing to do with counting steps.