Steps Count

[Sranjan]

I find that the steps count that the tracker shows is not accurate, when compared against a manual count of steps.

Moreover,the error is not even consistent over different measurements, with a variation of 0.5 times to 1.5 times the number of steps counted manually.

Has anyone faced this problem? Is there some other variable that get applied as a correction factor?

[Rich_Laue]

How are you testing this? 

Taking a walk with arms swinging on a natural way is the best. 

Do several years of 100 200 and 300 steps. 

[Sranjan]

Hi Rich.

Thanks for your intervention.

I have done these measurements both with my arms swinging and while keeping my left arm still, with the tracker facing me.

I have done these measurements over short counts (100, 200 steps) and also over long counts (thousands of steps over an extended period). Results have been similar.

I had posted my comment only after doing this exercise completely. Have you also done this experiment yourself?What has been your experience?

[Aclobos]

@Sranjan, lt took me a few months to adjust my stride length. I found myself walking with a different stride whenever I took my walks. You’ll have to play with your stride length settings. Here is a link to Fitbit website, it lists some articles about inaccurate readings and ways to correct them.

http://help.fitbit.com/articles/en_US/Help_article/1136

[Rich_Laue]

My experiance was until the second update my step count on with the Blaze was off by 15 steps.. Walk 100 it counted 85, walk 500 it counted 485. Etc. 

After the update a count of 100 gives me 98-100 steps, 200 gives 198-200, and the same for 500 and 1000. 

I start the walk with my right leg and count every left step. 50 count gives me 100, 100 count gives 200, etc. 

I find if the arm swings naturally or is stationary to my body, the steps reported are accurate. 

Independent tests of tracker a in general show that the pants pocket can be off by as much as 15%. 

Have you tried a restart, or better yet a complete shutdown of the Blaze? 

[Sranjan]

Thanks again for the detailed guidance Rich. Lemme try theses and report back on the findings.
regards

[Sranjan]

Thanks for the advice Aclobos.

I had actually not figured that steps count could have a relation with the stride length. I assume that it is some sort of vibration sensor that looks for footfalls and counts the same. Some people have also reported that the device under-reports the steps when they walk on a softer surface.

I have seen that my stride length is not constant. When walk is faster the stride length increases, and vice versa. Do you experience the same?

[Sranjan]

I think I failed to keep a track of when the update in question came.

But i repeated the experiment last evening and results are much better. The errors now were negligible.