My Sleep Record according to my (not so) Smart Samsung Watch

[ScottZZZ]

Hum, not sure about that explanation

That's not my explanation -- that is the medical textbook explanation.

There's NREM, classed as stages 1-4, and then there's REM.  "Depth" is a function of mental activity, and mental activity is far higher in REM than in Stages 1-4.  The fitness trackers are organizing their charts accordingly, with REM at the top (highest mental activity, other than awake) and stage 4 at the bottom (lowest mental activity, and most "restorative" phase). 

That organization isn't wrong.  The X axis (time) gives you the sequence, while the Y axis (depth) gives you the level of mental activity.

[ScottZZZ]

Scientific description of sleep staging, as described by all sleep science bodies is as follows;
......."Each cycle moves through each stage of sleep, sequentially : Awake, light sleep, deep sleep, REM, and repeat.

That is exactly correct -- for the SEQUENCE.  

But you seem to be misunderstanding that order to also imply the "depth" -- and it does not.  Deep sleep (phase 4) is the "deepest" (lowest mental activity), and REM is the highest (higher mental activity than phase 1).  That's why the various sleep trackers present the data with REM at the top and Deep Sleep at the bottom.

Here's a good psychology text on sleep phases from Washington State University.  It focuses on dreaming, but the first half has a great explanation of phases and mental activity (also note their hypnogram places REM at the top, just as the sleep tracker apps do):

   4.3 Stages of Sleep – Introductory Psychology (wsu.edu)

Here's the hypnogram from that article, which uses similar Y-axis as the Samsung and Fitbit (and other) trackers commonly use:
   

The "accuracy" of detecting sleep stages without brain wave activity is, in any case, still dubious.  I expect these tracker devices will get better at "estimating" the phase of sleep using movement, pulse, and body temp -- and they may in fact be helpful as they are -- but they simply don't have the inputs they need to accurately and reliable identify sleep stages.

[Dormeo]

Scott, I think you're right about what the OP was asking about -- my apologies to the OP for misunderstanding.

I just looked at the hypnogram including in my sleep study report. Several stages are collapsed into "light sleep." From top to bottom, the labels are: Wake, REM, L Sleep, D Sleep. And what an ugly night it was: reading from left to right: W, L, W, L, D, L, W, L, W, L, D, L, R, L, W, L, R, W, L, W, L, R, L, W, L, W, L, W, L, D, L, R, W.

And Scott, thanks for the link to the textbook. It has a lot of good information all in one place.

[Geer1]

This is the generally accepted and most widely used form of a hypnogram...

Almost all watches use what is called actigraphy, it is helpful for giving you a general idea of your sleep architecture but not accurate enough to compare to breathing data from OSCAR etc.

[ScottZZZ]

Almost all watches use what is called actigraphy, it is helpful for giving you a general idea of your sleep architecture but not accurate enough to compare to breathing data from OSCAR etc.

Agree.

I will say that I find movement and pulse data incredibly useful when combined with PAP data in OSCAR.  For example, movements immediately before or after PAP-reported apnea events can often lead to discarding those events as movement-related (if movement before/during reported event) or as an arousal (if movement immediately after event).  

Unfortunately, most of these smart watches and fitness trackers don’t provide a way to export data, or at least not in a way that can be used by OSCAR.

The pulse/movement data by itself - and algorithmic extrapolation to sleep phases - is not entirely useless, but the device manufacturers greatly embellish its value.

[quiescence at last]

also this is the study I refer to regarding many tricks to divine the sleep stage, showing about 65% accuracy.
https://journals.sagepub.com/doi/10.1177...8219827349

QAL

[ScottZZZ]

also this is the study I refer to regarding many tricks to divine the sleep stage, showing about 65% accuracy.  
https://journals.sagepub.com/doi/10.1177...8219827349

QAL

Excellent link, thank you, I hadn’t previously read that study.

There are a large number of similar studies. They prove that using ECG / pulse and movement data has some value (50-80% depending on stage), but ultimately get it wrong 30% or more of the time.

More data is always better, so a smart watch / fitness tracker is better than nothing since consumer EEG-based devices are not readily available. But users have to recognize the data they present are estimates at best.

[S. Manz]

Thank you for your thorough explanations of sleep "stages" Vs. "Depth and I think it would be safe to say that we all understand the difference now.
However back to my OP I am only interested to know the rough estimated time that my sleep progressed throughout the night in "order" of its stages which is why I find it easy to understand how this is represented by my Samsung Gear Fit-2 Pro as opposed to how it is done by my Samsung Galaxy smart watch...
My Gear Fit orders my sleep as per sleep order, while my other samsung does it based on brain activity, for which it has no input for, while all I want to see is:
1-How much time I spent in bed awake.
2-How much time I spent in light sleep
3-How much time I spent in deep sleep.
4-How much time I spent in Rem.
I want the graph to represent the "order", just like how my Samsung Gear fit-2 pro does.
   

[ScottZZZ]

Thank you for your thorough explanations of sleep "stages" Vs. "Depth and I think it would be safe to say that we all understand the difference now.
However back to my OP I am only interested to know the rough estimated time that my sleep progressed throughout the night in "order" of its stages which is why I find it easy to understand how this is represented by my Samsung Gear Fit-2 Pro as opposed to how it is done by my Samsung Galaxy smart watch...
My Gear Fit orders my sleep as per sleep order, while my other samsung does it based on brain activity, for which it has no input for, while all I want to see is:
1-How much time I spent in bed awake.
2-How much time I spent in light sleep
3-How much time I spent in deep sleep.
4-How much time I spent in Rem.
I want the graph to represent the "order", just like how my Samsung Gear fit-2 pro does.

Looking at your OP, it seems the Galaxy *does* report items 1-4.  Correct?

I understand you don't like having REM at the top.  But the Gear Fit doesn't really support that argument, since it didn't depict REM at all.  The obvious assumption would be that it falls into "Motionless", but that would mean REM and Phase4 are combined -- which any sleep medicine practitioner would point out as inappropriate, as those phases have very different mental and physical activity levels.

As for sequence on the Y-axis, any 2D continuous line graph is going to show the trace going "through" other phases, even if no time is spent in that phase.  For example, if REM were placed at the bottom, then when a REM phase transitioned directly to Phase-1, the trace would go through Phases 4, 3, and 2 to get there -- even though no time was spent in those Phases.  How is that depiction any less misleading than having the initial sleep phase trace going "through" REM (at the top) to get to Phase 1?

As I see it, Samsung is just following industry standards, and in fact sleep medicine standards, for how sleep phases are depicted on a hypnogram.  You are asking them to "Fix" something that Samsung, other sleep tracking device makers, and the sleep medicine community all agree is correct.  Given all that, I think it unlikely Samsung will respond to you with the changes you seek.