1 month in Epix2

Regarding 1 and 2, this is how Garmin explains it: Health Science | Garmin Technology

My 245 does 1 & 2 better. Even though it uses the old sleep tracking. I haven't had Epix long enough for it to comment on HRV, so maybe that will fix things. Currently it thinks I am more awake during night than the case is. I hope all estimates will improve with time. I have been very impressed with my 245's prediction, though I initially just bought it for buttons (Apple watch owner) and daily training suggestions.

I train 5+ times a week on my bike and do some strength training for overall fitness as well. In my case the experience has been completely the opposite. I had poor, fair and good sleep scores and days starting with 17 as body battery (stress from work and 2 hours of sleep the night before my summer vacation) as well as others with 100 (after a day of good training on the bike and good sleep). I'm just getting rid of Covid and I could clearly see how my body reacted to it thanks to the epix. Actually, I knew before I took the test that something was quiet wrong with my body and that I wasn't just a cold.

So the watch is working but maybe it's not the right tool for you

1) Sleep score - when you drill into into it where is the issue? For example mine last night was 72% which is about as good as I get due to duration - so 5h40 min (correct - still getting used to a CPAP so body still requires very little sleep - really need close to 8hrs if not more for a good/excellent score) gives me fair; Stress (very NB) - was good - ideally this should be excellent, but had a coffee less than 2hrs prior to going to sleep and have paid the price with a bout of high stress (makes sense); ideally deep should be good/excellent to get a good sleep score, and Awake/Restlessness is excellent (pretty common for me).These are the main ones to have good/excellent to get a good/excellent score. Light and REM correctly don't have a major factor on sleep score.

2) Body Battery is based of your stress levels, so take a look at sleep and make sure no bouts of stress while sleeping. And I agree that my BB doesn't always tie up with how I feel when I wake, but when I look back at the end of the day it makes sense. If I am less than 50% on BB and do an activity that is higher than base, while I feel fine in the morning, I will be dragging by evening - and hey my BB is very low at that point. If under 75% and do Threshold or Anaerobic activity, then same issue - drag in the evening - and body battery matches that. So for me its more like its giving me a overall day impression of how I will be (i.e. similar to the new training status metric that is coming/already out with Beta - just not as advanced).

3. Trending data - things like RHR - but can't just take it in isolation - more a case of I am usually around 55-60 as a norm - so if I suddenly see 65 pop up I know I have an issue coming. For habit/lifestyle change - you see these more in Sleep and HRV (Stress). Injury and Illness you see more in Stress and then see if filter down. For example if my stress tanks during the evening, sleep out, BBT out. Take a look at RHR - if that is out it probably means I am coming down with something (immediate jumps in RHR usually point to illness); whereas if no jump its probably pointing to over training potential injury (or have changed habit/lifestyle) - and if it slowly worsens definitely pointing to potentially injury. But RHR going higher and no issues with Stress probably means you are reducing your training load. Its really working out looking at trends as to how they work for your body more than anything else. But the trend data in the app is more than enough - not sure why you feel it isn't good enough as you get day view, 7 day view, 4 week view and annual view and can see list of your data. Certain data trend is pointless i.e. sleep score, but RHR , Stress is much more useful.BB I consider is more point in time as end result rather than a driver. ultimate trend data is of course 7DL and 7DL optimal band which is great at showing whether your training is maintaining, improving or declining. Intensity trend, steps, calories etc are more useful from a perspective of achieving targets

Will agree that sleep tracking trend data is pretty useless, but then again its debateable how accurate any watches trend data is, but if you want more on this would suggest using Habit to get better trend data (also covers all the other health data too - tend to look at it once a month): https://habitdash.com/overview/

1)It's pretty accurate to me...and it's clearly visible between 2 different periods(first one i was in vacation most of the time and the actual one is with work, a lot more training and less sleep overall).

Another clear example is yesterday's night.

During the day i had my longest run until now. 2:14h of run with a quiet hot day. 

Obviously, stress levels were high during the whole day and higher than usual during night too.

While the slept time was pretty long...for my standards at least, the overall score was not so high.

In normal conditions i have the stress levels a lot lower during night(average around 12 or lower) and more than 8h of sleep would give me a score in 90s.

As jaamgans said, what's the problem with your score? Which area? Just saying it's pointless without specify...it's a bit pointless.

especially as there might be very good reasons as to why you are getting a poor score i.e. high stress levels due to drinking alcohol for example - always kills my sleep score.

My REM sleep is always low.  Last night is a perfect example.  It's understood that dreaming takes place in REM.  I woke on two different occasions from dreams - 11:00pm and 1:30 am.  On both occasions I used the rest room and went back to bed.  You can see both awake times.  However, there was no REM around that time.  95% of the time my only REM is during the last 2 hours of my sleep.  This of course messes with my sleep score which in return messes with other metrics.  And, I don't drink alcohol at all, and never drink caffein at night.  I think those are just easy targets for Garmin to point at.

I didn't buy the watch for sleep information so it isn't too big a deal for me.  I'm more interested in the heart/oxygen/HRV info.  I do pay attention to the oxygen stats while sleeping though.  I find this information accurate so overall I'm pleased with the watch.

REM/Light Garmin/First beat really struggles to differentiate between the two. Deep it tends to be pretty good at, and as that is much more important, probably just as well. Personally I focus much more on duration, stress, restlessness and awake and deep - I ignore the rest. You will notice after a time that the score is much more weighted on those 4 metrics, especially stress and duration. Stress, restlessness/awake and deep are also much more impacted by lifestyle/habit changes and by illness and injury.

Sleep data is dodgy at best. I find the watch undercounts my Awake time and I've seen Deep sleep recorded while I was awake. Body battery seems to be best correlated to Stress, and I find is quite accurate. Other measurements of use are Recovery from exercise and the new Training Readiness, which is greatly impacted by sleep quality. Apart from Sleep, I find the physiological measurements fairly representative of how I feel and a decent guide to what I can take on. BTW, my wife's AW7 has, if anything, even worse sleep data.

Sleep score doesn't really tell you anything you don't already know.  Slept 4 hours last night, while rolling around in bed, yep, bad night.  I think the best score I've hit was 83.  I rarely get anywhere near 8 hours.  It's sort of a vague, not quite useless kind of thing.

Body battery, yeah, that's kind of weird.  It sort of correlates, confirms that I'm pretty wiped by the end of a day's work, but then what?  I gotta work out or not, it doesn't really effect anything.