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

Garmin can't make up its mind about sleep time?

After synchronizing this morning, Garmin Connect (on the web) showed a sleep time of 7:59 for last night. Just a few minutes later, it showed 7:27 instead. What's going on?
  • Garmin Connect sometimes reevaluates sleep data after more data becomes available.
  • Former Member
    0 Former Member
    Garmin had bugs in the sleep algorithms and Garmin connect. The new "advanced sleep monitoring" produced much less deep sleep hours. The deep sleep hours from Garmin can be changed from 3 hour to 10 minutes within ten minutes. The raw movement data is much better than current sleep summary. It is better using any good phone sleep tracking APPs before Garmin fix time counting bug in Garmin connect.
  • Garmin Connect sometimes reevaluates sleep data after more data becomes available.

    What do you mean by "more data"? The VA3 had already synced its sleep data with Garmin Connect before presenting the first number. Why should the number change just a few minutes later?
  • The VA3 had already synced its sleep data with Garmin Connect before presenting the first number.


    Not exactly. No sleep data is synced, because no sleep data exists on the watch. What the watch holds, and what is transfered to garmin connect, is wrist movement data. This is not the same as sleep data, and does not directly translate into sleep data. Garmin Connect looks at all the movement data available and tries to figure out the fall-asleep and wake-up time.

    Why should the number change just a few minutes later?


    Because more movement data becomes available, which can create some new context for the sleep detection algorithm, and garmin connect may reconsider based on this.

    Please note, that I'm not trying to argue whether this mechanism is good and/or efficient at properly detecting sleep.
  • Because more movement data becomes available, which can create some new context for the sleep detection algorithm, and garmin connect may reconsider based on this.

    Please note, that I'm not trying to argue whether this mechanism is good and/or efficient at properly detecting sleep.

    Many thanks for your patient and illuminating explanation!

  • What the watch holds, and what is transfered to garmin connect, is wrist movement data.

    Doesn't Garmin's sleep detection make any use of heart rate data? Wouldn't that help to improve the accuracy?

  • On some latest watches it does.
  • Apparently my VA3 (less than a year old) is not among them. Even though it measures heart rate, if I get up, take off the watch and put it on a shelf (without operating anything on it) and go about my business (e.g. take a shower), it thinks that I am still sleeping :-{
    P.S., Why should this depend on having the latest model? If the sleep data is actually evaluated on the servers, isn't it enough for the watch to supply the heart rate data?