Flawed algorithms

[LF12345]

I have a brand new Luxe and the data recording and scoring algorithm in your software is flawed. I noticed this in the past with old Fitbits and especially the scale. Bioimpedance scales are not  particularly accurate but the body fat percentage was 10% higher than it should be.

In terms of my Luxe, I got 3 hours of sleep yesterday and Fitbit gave me an overall score in the same range as other nights. In terms of sleep time, 33 out of 50. What is this based on? Time asleep vs  time awake? Time asleep (in this case 3) vs. the goal (8)?

Also, I question the time awake score, as I also wear an Apple Watch and it doesn’t record this time awake. I certainly do not remember waking up. Probably Fitbit records every time I move or rollover as time awake.

Obviously you have to take the recorded data with a big grain of salt. But to score a night of three hours as 64/fair and time asleep as 33/50 (fair) is illogical.

Fitbit, if you are reading this, please work on your algorithms because your interpretation of the data misses the mark.

I mean, in terms of managing expectations, when you take an online personality test you are (I would hope) expecting something frivolous…as opposed to, say, seeing a psychologist.

The thing is, FitBit, you are recording data that is somewhere in the ballpark of accurate, and the meaning you are ascribing to it is somewhere between mood ring and online personality test. Which makes me think that your algorithms need some work. If you are reading this, please look into it. Thanks!

[DramaQueenDiva]

I'm actually having trouble with the sleep function myself.  Supposedly, the device starts tracking after not sensing any movement for 3 hours or more and is retroactive from the last moment it recognized movement.  Fitbit does NOT record any time you roll over as time awake.  

As far as the score, that's based on your heart rate, and if you have a device that records it, SPO2.