Poor Accuracy

Options
fyoung1111
fyoung1111 Posts: 109 Member
I've been a long-time user and fan of the Fitbit One. So much so that I have purchased several of them for other people. I did have a couple of problems with the One. It was forever attempting to escape from its little silicon holder and disappear into the wild. I finally resorted to carrying it in my pocket when I had one. The other thing was that it was a pain to put it in its little wrist wrap for sleep tracking every night and the reverse of that procedure in the morning.

Fitbit's Flex would have addressed these problems but I did not care for it's minimal display nor its lack of stair counting functionality. I was pretty excited when Fitbit announced the Force and pre-ordered mine directly from the company. It arrived just before Christmas so I have accumulated a little more than one month of experience with it.

I first noticed a problem when I was walking on an indoor track that is 1/16 of a mile long. If I counted my steps in my head, they always fell in a range of 100-110 steps per circuit. The Force was counting between 60 and 90 steps each time around. Bored and intrigued, I began to play with the variables under my control. First, I slowed down from my usual pace of 4 mph to 3. The error disappeared. Next I increased my pace to a slow jog of approximately 6 mph. Again, the error disappeared.

When I reported these experiments to Fitbit, I got quick, courteous responses none of which were helpful such as "wear your Force on your non-dominant wrist and make sure your settings reflect this" or "adjust your stride settings" or "step counting isn't accurate on a treadmill".

Yesterday, I walked three miles, outdoors, in Alameda, California at a Runkeeper reported average speed of 4.13 miles per hour. I wore both my Fitbit Force and my Polar Loop on my right (non-dominant) wrist and recorded the step counts reported on the devices themselves at the beginning and end of my three mile walk.

The Force reported 3,490 steps over this 3 mile route. That equates to an ABSURD 54.65" per step.

The Loop reported a much more believable 5,585 steps yielding a reasonable 34.15" per step.

We can argue about whether or not the Loop was accurate but its resulting step length was spot on. The Force's computed step length was not even physically possible confirming my previous suspicion that it is dramatically under-counting my steps.

BTW I ran the three miles back and the Force and Loop counted my steps within 1% of each other and both counts seemed accurate based on indicated step length.

I've reported this most recent experiment to Fitbit in addition to several others (on and off treadmills). Hopefully they will address the problem in a future firmware update.

I would just switch to the Loop but it does not integrate with MFP yet and I'n not giving up MFP!

I am encouraging other Force users to do some simple experiments and report their results here. Just count 100 steps in your head while walking slowly and see how many steps your Force counts. Then do it again at a brisk walking pace. Finally, try it one more time running at any speed you are comfortable with. Please let us know how it comes out.