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

Sleep Tracking errors on 955

The sleep tracking function and therefore all of the training advisories are woefully inaccurate! To the point I'm considering returning the watch. The sleep durations are miles out, the sleep stages are obviously wrong. My watch and connect are showing different times/data and despite setting normal sleep start/finish times the watch thinks lounging on the sofa is 'sleeping'.

How can I sort this functionality / bug out foreword i give up and get my money back?

Ian

  • Concerning your point 1: Any modifications and changes you make in Garmin Connect will be updated to your watch as soon as you synchronize for the next time. Garmin Connect is not in its own universe, it is a tool to analyze and edit the data on your watch - and vice versa.

  • I can categorically say GC didn’t update the watch sleep screen 

  • That's right.

    But I don't understand how some people were saying Fitbit are good at the sleep tracking.

    The watch is using the optical HR sensor to detect pulses and accelerometer sensor to detect motion.

    Assuming your pulse is low and no movement while sleeping or lying on the couch with eyes open, how does the watch differentiate the 2 actions ???

  • I don’t think it can differentiate and relies on sleep time settings to ignore resting/naps on a time of day basis. That is simpler than using sensors and variability across a population.

  • I have the settings for sleep from 23.00 to 08.00, but the watch has created two sleep periods in the afternoons starting just before 13.00 when I was eating. Editing on GC was partially successful in that some of the information used in the training readiness updated, but not all. So, for example, the sleep duration updated, but the score for sleep and the rating did not and the result went from 84 to below 50. The score for sleep then continued affecting results for a few days. My recovery time went from 8 hours in the morning as a result of good sleep,  to 8 hours in the evening as a result of the afternoon "sleep" which the 955 now said was poor. I contacted support on 12th July but no response yet.

  • Garmin definitely have a problem with the 955 tracking sleep properly! I returned a 955 solar with this problem and the replacement 955 non solar has the same issue - recording rest outside of the sleep window as sleep. This error then affects the physio metrics!

  • When this happened to me, I used the phone app to edit Sleep start/stop time for that day back to what it should have been.

  • I know that but it doesn't update the watch where the physio scores (training readiness) are kept

  • Replacing the device or buying another model from Garmin will not lead to anything.

    Sleep Monitoring algorithms are the same for all Garmin devices. They use algorithm from Firstbeat Analytics. If we carefully read the principles of its work, it seems that these algorithms are not ideal: "These include heart rate, heart rate variability (HRV), respiration rate derived from heartbeat data, wrist or body movement captured with an accelerometer, and time of day" - it completely corresponds to my state when I read a book or watch a video before going to bed lol. The advice that Garmin gives for a more correct setting of sleep tracking (Setting the Sleep Window) does not seem to fix the problem (but still worth using). Many real Garmin user feedback at forum and my own experience with previous Garmin devices indicates that, yes, the sleep tracking algorithms are not perfect.

    Sometimes we need to edit sleep beginning time - it's not the end of the world, that's why Garmin made the editing function. I have now purchased a new 955 and I was ready for what I wrote above.

    But NOW this comes with a problem: edited sleep doesn't update Training Readiness metrics. During the day, the TR indicator is updated on the watch depending on the change in the growth of Acute Load and the decrease in Recovery Time, but the edited quality of sleep is no longer affected. And this seems to affect the result and the next three days.

    Dear Garmin, we are not asking you to change sleep tracking algorithms (this is a separate big topic for discussion, I think), but at least fix the effect of editable sleep and history on TR, because two bugs for a new function at once is too much. And also, it is one of the functions that influenced my purchase of the 955. Thank you and best wishes!

  • Thanks for that information and I come to the same conclusion! My first 995 (solar) recorded rest during the day as sleep and was returned the second 955 (non solar) is much the same. Garmin have asked for a data dump from the watch and they have‘opened an investigation’ so I’ll see what that reveals. 

    it seems to me that the software could be coded to prevent sleep recording outside the sleep window and, as you say, allow correction edits and recalculation.