As a concrete case, last night I charged the watch to “100%” at 21:00 and wore it during sleep as usual. By morning, the battery had dropped to 95%, representing a 5% overnight drain. Under identical conditions, the normal overnight consumption is typically around 1%.
This deviation strongly suggests that the battery was not truly at a full state of charge when the system reported 100%. A more realistic estimation is that the actual battery voltage was likely closer to 96%, while the software prematurely capped and displayed it as 100%.
This reinforces the hypothesis that the current SOC algorithm is overestimating charge level due to calibration thresholds or coulomb counter drift, resulting in a misleading full-charge indication and an abnormal discharge pattern immediately after.
In practical terms, the system is reporting “logical 100%”
but the battery is operating at “electrical 96%”.
This mismatch is not a matter of user perception, but a measurable inconsistency between displayed SOC and real battery behavior.