Sleep Tracking and Better Sleep: What Wearables Really Measure, What the Studies Say, and Every Sleep Article in One Place

Transparency: The author of this blog uses only WHOOP as a sleep tracker. He has never owned or tested Oura, the Samsung Galaxy Ring, Ultrahuman, RingConn or other rings. Study results come from the original publications (retrieved via PubMed on 16 Oct 2026); our own WHOOP figures are taken from the self-experiments linked below. This article is not medical advice and contains no dosage recommendations and no affiliate links. All sources are listed at the end. More on methodology

Key points

• Sleep duration and timing are measured fairly reliably by current wearables compared with the sleep lab.
• Time awake during the night is poorly detected. In two independent studies only 18 to 54% of wake periods were classified correctly.
• Sleep stages (light, deep, REM) are an estimate. Agreement with the sleep lab is moderate; only individual studies, mostly funded by the manufacturer, show clearly better results.
• Trends over weeks are more meaningful than single nights. But algorithm updates can shift trends too.
• The basics stay unspectacular: 7 or more hours for adults, and for chronic sleep problems behavioural therapy (CBT-I) comes first.
• Guide: below you will find every sleep article on this site, sorted by the question you have.

What a sleep tracker actually measures

In a sleep lab (polysomnography, PSG), brain waves, eye movements and muscle tone are recorded. From these, specialists decide for every 30-second segment whether you are awake or in light, deep or REM sleep. That is the reference all trackers are tested against.

A ring, band or watch measures none of that. They measure:

• movement via an accelerometer,
• heart rate and heart rate variability via an optical sensor,
• depending on the device, skin temperature and blood oxygen.

An algorithm estimates sleep and sleep stages from those signals. A study by Oura staff shows how much the extra signals add: using movement data alone, the model classified the four stages (wake, light, deep, REM) correctly 57% of the time; with heart rate, HRV, temperature and time-of-day features it reached 79% (Altini & Kinnunen 2021, 106 people, 440 nights).

What the validation studies show

Reading the numbers: kappa measures agreement beyond chance (0 = chance, 1 = perfect). Schyvens et al. describe their values of 0.21 to 0.53 as “fair to moderate”. “Sleep detected” and “wake detected” are the shares of 30-second segments the tracker classified the same way as the sleep lab.

Who paid for the studies: the Gen3 study (Svensson) received financial support from Oura according to its competing interests statement; the Altini study comes from an Oura employee and an Oura advisor. The lab behind the Miller study receives research support from WHOOP, though the authors say WHOOP was not involved in the study. Most authors of the Lee study work for the sleep tech company Asleep and, according to their disclosure, hold shares in it. That doesn’t make the results wrong, but independent replications are often missing.

A 2024 review requested by the Sleep Research Society sums up the state of play: consumer wearables outperform traditional actigraphy but have clear limits. It names misclassified wake periods, problems outside the main nighttime sleep period, artefacts, and unclear performance in people with certain characteristics such as darker skin or obesity (de Zambotti et al. 2024).

What you can trust your tracker on

1. Sleep duration, falling asleep and waking up: largely yes. That is what every device does best in every study.
2. How often and how long you were awake at night: mostly no. The low wake detection rates mean trackers often count awake minutes as sleep.
3. Deep and REM sleep on a single night: with caution. Schyvens et al. consider devices with higher kappa usable for tracking prolonged, significant changes in sleep architecture.
4. Trends over weeks: yes, as long as the algorithm doesn’t change.

That last point caught us out in our own data. In the sleepmaxxing self-experiment, the WHOOP deep sleep average rose from 47 minutes (Q4 2024) to 105 minutes (Q1 2026). The methylene blue self-experiment shows that part of this coincides with an update to WHOOP’s sleep stage algorithm between Q4 2024 and Q1 2025: 46 minutes of deep sleep on average before, over 90 after. If you compare your data over years, you need to know about breaks like that.

Evidence-based basics, no hype

How much sleep? The American Academy of Sleep Medicine and the Sleep Research Society recommend that adults regularly sleep 7 or more hours per night. The AASM stresses that individual needs vary.

Chronic sleep problems: the European Insomnia Guideline (Riemann et al. 2023) recommends cognitive behavioural therapy for insomnia (CBT-I) as the first-line treatment for chronic insomnia in adults of any age, delivered in person or digitally. Medication is the next step according to the guideline when CBT-I is not effective enough. For diagnosis, the guideline relies on a clinical interview, questionnaires and sleep diaries. It does not recommend actigraphy for routine evaluation, and recommends a sleep lab when other sleep disorders such as sleep-related breathing disorders are suspected.

Orthosomnia: in 2017 Baron et al. described patients who sought treatment because of unusual tracker data, such as short sleep or periods of light, restless sleep, and who developed a perfectionistic quest for “ideal” sleep. To these patients, tracker data often felt more consistent with their experience than validated measurements. If the morning number stresses you more than how you actually feel when you wake up, take that as a signal.

Supplements: the evidence differs widely from substance to substance, and we give no dosage recommendations here. What happened in our own data, and what didn’t, is in the self-experiments below. They are N=1 and not evidence for anyone else. Before taking any supplement, especially alongside medication, talk to a doctor or pharmacist. Methylene blue is an example of how serious interactions can be.

When to ask a doctor rather than your tracker: if you sleep badly for weeks, are seriously impaired during the day, or someone notices pauses in your breathing at night. A wearable does not make a diagnosis; Samsung, for example, states this explicitly for its Energy Score.

Guide: every sleep article on this site

You want to buy a sleep tracker

You already have a tracker and want to understand the data

You want to know what happened in our own data (N=1, WHOOP)

You want to start without expensive gadgets

Biohacking starter kit: what you actually need

Verdict

Wearables make good sleep diaries: they reliably capture when and how long you sleep. For time awake during the night and for the split into deep and REM sleep they are estimates, and their quality varies a lot between devices and studies. Use the trends, not the single night, and expect breaks caused by software updates. For sleep itself, the basics remain boring and well supported: enough time in bed, and for real sleep problems a medical assessment rather than tracker optimisation.

Sources

• Chinoy ED et al. Performance of seven consumer sleep-tracking devices compared with polysomnography. Sleep 2021;44(5):zsaa291: PubMed 33378539
• Miller DJ, Sargent C, Roach GD. A Validation of Six Wearable Devices for Estimating Sleep, Heart Rate and Heart Rate Variability in Healthy Adults. Sensors 2022;22(16):6317: PubMed 36016077
• Lee T et al. Accuracy of 11 Wearable, Nearable, and Airable Consumer Sleep Trackers. JMIR mHealth uHealth 2023;11:e50983: PubMed 37917155
• Svensson T et al. Validity and reliability of the Oura Ring Generation 3 (Gen3) with Oura sleep staging algorithm 2.0 (OSSA 2.0) when compared to multi-night ambulatory polysomnography. Sleep Med 2024;115:251-263: PubMed 38382312
• Schyvens AM et al. A performance validation of six commercial wrist-worn wearable sleep-tracking devices for sleep stage scoring compared to polysomnography. Sleep Adv 2025;6(2):zpaf021: PubMed 40303381
• Altini M, Kinnunen H. The Promise of Sleep: A Multi-Sensor Approach for Accurate Sleep Stage Detection Using the Oura Ring. Sensors 2021;21(13):4302: PubMed 34201861
• de Zambotti M et al. State of the science and recommendations for using wearable technology in sleep and circadian research. Sleep 2024;47(4):zsad325: PubMed 38149978
• Watson NF et al. Recommended Amount of Sleep for a Healthy Adult: A Joint Consensus Statement of the AASM and SRS. Sleep 2015;38(6):843-4: PubMed 26039963
• American Academy of Sleep Medicine, Adult Sleep Duration Health Advisory, accessed 16 Oct 2026: aasm.org
• Riemann D et al. The European Insomnia Guideline: An update on the diagnosis and treatment of insomnia 2023. J Sleep Res 2023;32(6):e14035: PubMed 38016484
• Baron KG et al. Orthosomnia: Are Some Patients Taking the Quantified Self Too Far? J Clin Sleep Med 2017;13(2):351-354: PubMed 27855740
• Samsung Germany, Galaxy Ring product page (note on the Energy Score; German), accessed 16 Oct 2026: samsung.com/de
• Our own WHOOP data: taken from sleepmaxxing and the methylene blue self-experiment

Not medical advice. As of 16 October 2026.