Your smartwatch information relies equally on software algorithms as it does on hardware, and when the calculations shift, it can be quite disorienting.
Owners of Galaxy Watches have reported a surge in frequent high-stress notifications since upgrading to One UI 8 Watch, even though their lifestyles remain unchanged. Whether this is due to a bug or an overly sensitive new algorithm monitoring stress levels, Samsung isn’t the only smartwatch manufacturer to perplex and upset its users with abrupt data modifications, nor will it be the last.
Android Authority initially highlighted Reddit threads from Galaxy Watch 7 users who were expressing dissatisfaction over receiving multiple high-stress alerts each day, a situation they hadn’t encountered before the Wear OS 6 update. The problem was also duplicated on a Galaxy Watch 8.
I failed to observe any stress data discrepancies during my review of the Galaxy Watch 8 Classic. However, putting it on today, I’ve seen the watch react quickly from Relaxed at my desk to Moderate or High at the slightest stimulus. Some responses seem valid, like feeling exhilarated while watching a basketball match. Yet, at other times, I’m registered as having Moderate stress just from standing, or High while moving around inside the house.
I’m certainly not under stress, and I like to think I’m fit enough that walking isn’t taxing on my heart. Something behind the scenes has changed in how Samsung assesses stress. However, due to the abruptness of these changes, individuals may be more inclined to turn off alerts and overlook the Stress widget rather than assume they’ve been under stress all along.
SamMobile reported a similar (but opposite) phenomenon earlier this month: Galaxy Watch users unexpectedly transitioned from moderate Sleep Scores to achieving 99 or 100 ratings after typical nights of sleep. If your smartwatch data seems implausibly positive, why take it seriously?
We’re at a juncture with smartwatches and smart rings where companies aspire for them to function as medical devices, detecting indicators of hypertension in Apple Watches, heart failure in Galaxy Watches, or pulse loss in Pixel Watches. Nonetheless, moments like these serve as a reminder that smartwatches must build trust with their wearers, or else warnings will be overlooked.
Back-end smartwatch modifications, whether beneficial or detrimental, undermine user confidence.
In August, Fitbit declared that “your sleep tracking just became more accurate” in a forum post, clarifying that their updated tracking system could now “identify brief awakenings previously missed” on all Fitbit trackers and Pixel Watches.
Earlier that summer, Fitbit users on Reddit voiced concerns that their scores had unexpectedly soared above what was reasonable for their sleep quality. Therefore, this stricter algorithm transition disoriented users, causing their scores to drop dramatically while their awake scores skyrocketed.
Daily users of these devices invest in the algorithm, so an abrupt change can shake their confidence in the gadget and lead to feelings of embarrassment for having trusted the old data. Why address an issue when the device previously provided assurance that there wasn’t one?
Moreover, although the Quantified Scientist conducted an EEG test that revealed the new Fitbit algorithm is indeed more precise, it remains uncertain whether his findings are universally applicable or if devices perform better with certain physiological traits than others.
In a related instance, Google unveiled a new step counting algorithm in March 2025 aimed at recognizing more “challenging” steps such as walking with a stroller, shopping cart, trekking poles, or pushing a wheelchair. A month later, Google reverted to the previous algorithm.
I experimented with the algorithm and found it surprisingly effective for challenging scenarios, just as claimed. However, other users encountered significantly inflated step counts, leading Google to find it easier to switch back to the traditional system that effectively tracked average steps rather than attempt to rectify the issues in these “particular situations.”
Samsung’s updated stress algorithm may be completely inaccurate, or it might simply be an overly enthusiastic version that is closer to the truth than the previous system. Other devices assess my stress at much lower levels, but there’s no way to verify if they are indeed more accurate, even though I would prefer they were.
I can compare heart rates with a chest strap, GPS with my actual position, or step counts with a manual tally. Additionally, other metrics like passive AFib detection necessitate FDA verification for accuracy. Yet there will always exist “wellness” metrics where faith in a brand becomes the only source of reliance.
I would prefer these devices provide me with accurate information rather than dilute the reality; if I am under stress, I want to know. However, it may be essential for these brands to approach us more thoughtfully regarding changes, or to inform us in advance that results may fluctuate, along with reasons to trust the outcomes, instead of presuming we will unconditionally accept that our health has deteriorated.
