How does my watch know that I am stress ?

In reply to your questions

Those with an issue:
- Are you wearing the watch to bed so that resting heart rate is calculated correctly - Yes the watch only leaves my wrist to charge up, on ce a week, and during the day.
- Is sleep time set correctly Yes it is, 2200 through until 0645.
- Do you have resting heart rate set to levels determined by the watch or are you using a manual value - I am using a watch determined value.
- Have you reviewed all values in your user profile and if yes, are any that are incomplete, inaccurate or set manually? I have been over most forums trying to self diagnose the issue as Garmin were not interested. Hence, i think i have covered every HR and sleep related setting that might help.

Thanks for the insight C966. I notice now that there is a problem with the resting heart rate calculations as well that seems to correspond to bad stress calculations. For example, I noticed that a couple of nights ago at 9:45 before I went to bed that my days resting heart rate had been calculated at 52 BPM. I went to bed and woke up at 4:00 AM and checked again. My new days RHR had been calculated at 55 BPM but the previous days RHR shot up to 72 BPM. If I look at the stress graphs, I am in extreme values of up to 90 between my bed time and wake up time, although my hear rate was a normal looking 65 that slowly lowered to about 48 or 50 before I woke up. I don't think its normal for the previous days resting heart rate to be recalculated based on the current day's sleeping pattern and reading (although I may be wrong). This has happened now three days in a row where my RHR moves from low 50s to high 60 or low 70s the next day when I wake up. I believe the whole heart rate, stress algorithms are wrong. What we need to know is why and how Garmin can fix? This is inconsistent, and I cannot think of a setting I changed that would have initiated the change. Has anyone else seen there RHR previous days numbers increase on waking in the morning?

Curiously, I woke up this morning and we had firmware version 8.0 on my watch. The odd thing is that I have gone to the Garmin website to get an idea of what bugs it was to fix, and there is nothing about it on Garmins website. Weird.

Having a sniff around Garmin Firmware updates and seeing your point about resting heart rate calculation, I saw this beta release does have a fix to improve th resting heart rate measurement. I assume that 8.0 is the approved version of this beta perhaps.

A new Beta software for the Forerunner 935 has been released. Full change log, instructions, and software for installing the beta can be found here.

Changes made from version 7.60 to 7.93:

• Added support for a Sunrise & Sunset widget. Settings > Widgets > Add Widgets > Sunrise & Sunset.
  
• Added support for an Alternate Time Zones Widget. Settings > Widgets > Add Widgets > Alt. time Zones.
  
• Added a setting to display a calibration prompt when using a power meter. When using a Vector, this will also allow the user to set the power meter crank length. Menu > Settings > Sensors & Accessories > Power Meter > Calibration Prompt.
  
• Improved resting heart rate measurement.
  
• Fixed Bluetooth connectivity issues with Huawei phones.
  
• Fixed a potential issue causing the backlight to not turn on when pressing keys.
  
• Fixed a potential issue preventing alerts from being added.
  
• Fixed an issue causing updates to the Lengths data field to be delayed during Pool Swim activities.
  
• Fixed an issue where the altitude on device wasn't accurate when using the Jumpmaster app.
  
• Fixed an issue where Broadcasting Heart Rate was not working as expected, if accessed from the power controls menu.
  
• Fixed an issue where Flashlight would turn off when using gestures.
  

Do we know how specifically determines the Garmin Stress score is calculated? I haven’t seen anything that’s very specific, more general in nature. If you comparitively are inactive with the accelerometer (motion) and have a higher heart rate (excitement/emotion/hormone/physical response) I would believe would dictate a higher stress score - taken further if your Heart Rate Variability (heart beat to heart beat timing) was more variable (not consistent) it would lead to greater stress, more pathological/physical. Bottom line what specially determines a stress score? I may have missed specifics on the how - does anyone know? Curious as well.

I believe its mainly based on heart rate variability though I supposed it may factor in other metrics such as RHR, however its a low variability that indicates higher physical stress as opposed to lower which indicates you are more rested.

Do we know how specifically determines the Garmin Stress score is calculated? I haven’t seen anything that’s very specific, more general in nature. If you comparitively are inactive with the accelerometer (motion) and have a higher heart rate (excitement/emotion/hormone/physical response) I would believe would dictate a higher stress score - taken further if your Heart Rate Variability (heart beat to heart beat timing) was more variable (not consistent) it would lead to greater stress, more pathological/physical. Bottom line what specially determines a stress score? I may have missed specifics on the how - does anyone know? Curious as well.

Firstbeat do a decent job of explaining how the All-Day stress score works in the link below.
https://www.firstbeat.com/en/consumer-feature/quick-stress-level-test/

At a high level it's entirely related to HRV - high variability strongly correlates to a healthy and rested physical condition, and a low variability indicates a stressed body that is trying to regulate itself.
It's not totally tied to heart rate, but a lower heart rate is more likely to show higher variability between beats, so even without the all-day stress score you can still get an indication of your physical state by checking the change in your resting heart rate day to day.

As to how useful it is - personally I find the information interesting, but it just confirms how I'm already feeling as opposed to giving me insights.
I'd find it way more interesting if they could use the data to give you "forecasts". Like for example if today is Monday and based on how I'm trending and my typical weekly patterns then I'm likely to be feeling drained by Saturday. If I had a big event planned for Saturday I could use this information to alter my lead-in plans for example.
Or if it was able to say something like "based on the past 3 days 80% of people following this same trend end up skipping their typical exercise routines, most likely due to fatigue and/or illness"
That would make it truly useful and actionable data.

Garmin do now seem to be owning up to this issue being a p[roblem. I have recently had an RMA of a 935, but the same all day stress issue keeps occurring. I asked them to reopen my case and got this response.

"Thank you for the reply. I have reviewed the stress data for the last few days and can see where there seems to be a great difference in the stress level and rest periods with no apparent reason for the difference. I have added this information in to a new research request and we will continue to investigate and let you know when we have a fix. We do know that it has something to do with the resting heart rate value and this would explain why it seems to work properly intermittently when you reset the device."

No idea as to how high a priority it is, but fingers crossed they deep dive this p[roblem and release a bug free firmweare version to eradicate this problem.

Garmin do now seem to be owning up to this issue being a p[roblem. I have recently had an RMA of a 935, but the same all day stress issue keeps occurring. I asked them to reopen my case and got this response.

"Thank you for the reply. I have reviewed the stress data for the last few days and can see where there seems to be a great difference in the stress level and rest periods with no apparent reason for the difference. I have added this information in to a new research request and we will continue to investigate and let you know when we have a fix. We do know that it has something to do with the resting heart rate value and this would explain why it seems to work properly intermittently when you reset the device."

No idea as to how high a priority it is, but fingers crossed they deep dive this p[roblem and release a bug free firmweare version to eradicate this problem.

Yesterday I got a resting HR of 49 and the stress levels looked OK, today I got a resting HR of 58 (yeah, resting HR doesn't seem very reliable either) and the stress levels are all screwed up... so there may be indeed a connection. Right now I can't trust either RHR or HRV stress info on the Fenix5... they need to give this issue some priority cause these are pretty basic functions that it needs to do right.

I contacted customer support in English a couple of weeks ago and got diverted to Customer Support Spain still, and they said they weren't aware of other reports of such a problem and after trying to inform them of it and of other cases they pretty much didn't care about finding out about the problem or hearing about it and played possum.