This discussion has been locked.
You can no longer post new replies to this discussion. If you have a question you can start a new discussion

Deep sleep accuracy?

I am curious to hear what other people's experience is with Deep Sleep stats.  According to my 945, I get about 10-15 mins of Deep Sleep a night.  First, I do not know if this is normal for the average person.  Second, is the 945 sleep stats accurate?

  • Well, your eyeballs move when you're awake, aren't they? The watch is correct. Just kidding ;-)

  • Former Member
    0 Former Member over 4 years ago in reply to Ganfas
    My problem with this watch and the sleep tracking is not with the Deep Sleep but with the REM.

    I wouldn't be too concerned with REM (or any of the sleep readings) unless you were tested by a professional in a lab.  In order to be in REM, you must have these three things.  1.  The obvious, rapid eye movement.  2.  Increased brain activity. 3.  Slight twitching of the muscles.  Your Garmin, Fitbit, Oura ring, etc., at best may detect the muscle twitching, but are unable to detect the other two.  They do nothing but "guess" when you're in REM based heart rate using clinical studies that show when the average person enters REM and the average length of time.  If you look at the disclaimers on all their websites, it even says the readings are estimates.  The sleep function with any fitness/activity device should be taken with a grain of salt.  Do any of us really need to have a device to tell us if we had a good or bad night of sleep anyway? 

  • But the thing with REM is, not The REM per si, but it detecting REM when i'm awake, for example in the morning i'm awake but still in bed, or if i wake in the middle of the nigth and stay like that for let's say 10 minutes, it detects REM, On the fitbit charge 2, i don't know how but in the same conditions it said i was awake

  • I have noticed this: by downloading the sleep data as soon as I wake up and consulting GC within the first 12 hours, the data is correct (Example = 4 hours deep 2 hours light and 10 minutes awake, and this is all right). If I consult the same data after 24 hours, these are completely changed and the deep phase becomes all phase rem! I think that GC re-elaborates this data but there is a programming error.

  • Just got my Forerunner 945 yesterday. First night sleep tracking - 1h38min

  • I'm seeing the exact same thing on both my former Fenix 5X+ and now on my Marq Athlete. I've contacted firstbeat with screenshots showing this thinking they were the ones to blame. This is the response I got:

    So apparently, Garmin needs to fix this.

  • Twist! Garmin Connect has not been reworking my sleep for two days and the data remains correct !! This night 4h10 'of deep sleep + 2H14' of light sleep! Last night 4h04 'deep and 2h22' light! Gc has solved the bug!

  • Noooo !! Garmin Connect has reviewed the sleep of the last few nights, transforming all the phases of deep sleep into Rem sleep! We are back to the fact that if I analyze sleep in the first hours after waking up I would say that it is perfect. The next day Garmin Connect recalibrates sleep and generates the error. I therefore got into the habit of transforming the graphs generated by just waking up into .pdf files and made a folder out of all my nights of sleep !!

  • I think it is (also) a question which watch you use. I had my 945 replaced by Garmin and used my old fenix 3HR for a week and with that I had much longer deep sleep periods (no REM, because fenix 3HR doesn't record it). I think the measurements from the fenix are "better".

  • Bumping this to let them know this has not been improved at all