Fall asleep hour regression model is a chart that helps you decide when you should go to sleep. The aim of this chart is to show you how changes in your fall asleep hour will affect your rating and deep sleep %.
May be that when recording is auto only near the awake time, it record short / late sleeping time?
What exactly does this mean? Like I would be sleeping the whole day away if Follow what you are your Advice is trying to give me! What on Earth how can anybody sleep that long
@petr-urbandroid - about Irregularity:
I often wake up in the middle of night and stay awake for more than 15 minutes, sometimes hours.
I would like to count the mid sleep time from my whole sleep period, both before and after my awake period.
Now, I can not set a goal for Irregularity, because it will be very wrong. Sometimes the first period is longer and counts, sometimes the later one.
@Maria_Friberg many thanks for your feedback. Can you please explain, are you stopping the tracking in this case… normally we should detect hat you are awake from your phone usage, light in the room etc… so you do not need to stop tracking. Also you can manually pause tracking and any pause done this way would be considered as awake… Did that help?
I am sorry. English is not my first language.
I don’t stop tracking and the awake period is counted as awake (as it should).
The problem is how the app counts irregularity.
Mid sleep hour is counted “from” (this is where my English went wrong - I mean counted “out of”/on/by?) either my first sleep period or the last sleep period.
I would like my mid sleep hour to depend on the whole span. From my sleep hour to my wake up hour, even if I am awake a couple of hours in the middle.
My total sleep time should of course not include the awake gap.
Does this make sence? 
An example:
I sleep from 00:00 to 07:00 Monday and Tuesday.
On both nights I am awake for one hour.
This seems very regular to me.
But on Monday my awake period starts at 02:00 and on Tuesday it starts at 04:00.
So the app counts mid sleep hour as:
Monday: 05:00 (03:00-07:00)
Tuesday: 02:00 (00:00-04:00)
Very irregular.
In this case I would like mid sleep hour to be:
03:30 (00:00-07:00) on both days.
When I set a goal for improving irregularity the graph goes bananas.
Maybe I’m still missing the point and also let me double check that… but IMHO we calculate the midsleep hour from the entire tracking, so we just take from and to of the track ignoring awakes… this is a simplification, but there are many “bad” ways how to handle this better for someone’s patterns and worse for others…
This post made me believe that a wake period longer than 15 minutes would count as two different periods and that only the longest one would count as the main sleep.
But maybe it all counts as one sleep “session” if I don’t stop tracking during my awake time?
Hey, first time posting here. With sleep deficit i think it would be good idea if over time the deficit automatically would slowly get back to zero. At the moment say if I had a slot of sleep deficit over a month ago my overall sleep deficit remains high unless I start over sleeping… allot. So when I’m switching through different time periods it now reports in quarter year I have a deficit of 23 hours but in 14 days I’m in the positive by 2 hours. But I’m now hitting close to my targets each day but that would mean my overall deficit is never going to come down. Hopefully that makes sense. Technically over time all the stats should slowly creep back to neutral.
Hi, the deficit is shown for the whole displayed period.
The deficit shown on the Sleep Score card shows the current deficit for the past 14 days, as it is the period that influences you the most.
If you are currently in 2 hours surplus, the overall deficit will be balanced eventually.
This may be obvious to some, but I want to be sure I understand 100%. Mid-sleep means the hour that on average is the middle of your sleep cycle, right??
