Although this is a fenix thread, I can report the same from a vivoactive 3. Values are really high even in non stressy situations (as today, Sunday, doing nothing). In addition, I face large gaps in the stress curve - when I am sleeping or lying relaxed on the couch. Sensor placement is perfect, because Heart rate is monitored without any problems (please see attached curve). So I assume that HRV can be calculated. So I have the strong feeling, that the algo has an issue