Open the app in the morning and there it is, rendered like a stock chart: a stacked bar of last night's sleep, so many minutes of deep, so many of REM, one glowing score out of a hundred. I check mine before I am fully awake. I know better, and I check it anyway. That reflex is the whole problem.
Here is what the ring is actually good at. In a 2024 study in the journal Sensors, Rebecca Robbins and colleagues at Brigham and Women's Hospital put an Oura Ring Gen3, a Fitbit Sense 2, and an Apple Watch Series 8 on 35 healthy adults and compared a night of readings against polysomnography, the electrodes-and-wires setup in a sleep lab that counts as the truth. For the single question of asleep or awake, all three devices hit at least 95 percent sensitivity, meaning that of the moments the lab scored as sleep, the device caught at least 95 percent of them, and they matched the lab on more than 90 percent of the night overall. As a coarse instrument, telling sleep from wake, the thing on your finger works.
Now the part it sells you and cannot deliver. The much harder task is sorting that sleep into stages, light, deep, and REM, and on that four-way question the same devices dropped to a sensitivity range of 50 to 86 percent. The Apple Watch swung from 50.5 to 86.1 percent depending on the stage, the Oura sat around 76 to 79, the Fitbit landed between 61.7 and 78. Every device showed poor agreement on deep sleep and REM specifically, the two numbers people fixate on. The bar chart you are grading yourself against is, stage by stage, closer to a guess than a measurement.
There is a reason for this that no software update will patch, and remember we are now on the four-stage breakdown, not the reliable asleep-or-awake call. When a wrist device is unsure, it tends to fall back on light sleep, which sits in the middle on both heart rate and movement. Anne-Marie Schyvens and colleagues documented that default across six devices in SLEEP Advances in 2025. Worse, for anyone chasing a perfect score: even the lab is fuzzy. In the American Academy of Sleep Medicine's inter-scorer program, Rosenberg and Van Hout found that trained human technicians scoring the same night agreed only about 82.6 percent of the time, and on deep sleep just 67 percent. Oura reports its own algorithm reaching roughly 79 percent agreement with the lab and frames that as a win, which, against a human ceiling of about 83 percent, it more or less is. The honest reading is that four-stage sleep is a number nobody, wired up or not, pins down with the precision the tidy chart implies.
Sold a precise number that is not precise, people do what people do. They worry about it. In 2017 the sleep psychologist Kelly Glazer Baron and colleagues named this, in a Journal of Clinical Sleep Medicine case series of three patients who trusted their trackers over their own doctors: orthosomnia, from ortho for correct and somnia for sleep, built on the same root as orthorexia, the pursuit of perfect eating that makes eating worse. The perfectionistic quest for ideal sleep, they wrote, becomes the thing keeping you awake.
How common is it now. At the SLEEP 2026 meeting, Rebecca Robbins and colleagues surveyed 1,280 US adults and found that among people who track their sleep, 30.9 percent scored positive for orthosomnia risk, feeling nervous or on edge about their tracker data, which the authors put at roughly double earlier estimates. One receipt to keep in view: Robbins first-authored the validation study above, and this survey documenting the anxiety was funded by Oura Ring Ltd, a company that sells the rings. That does not make the finding wrong, but it is worth naming, alongside a sample skewed white and middle-aged and a conference abstract not yet through full peer review.
So use the thing, sanely. The trustworthy output is the coarse one and the slow one: roughly when you fell asleep, roughly how long you were down, and the shape of that over a week or a month. A trend line rewards the boring inputs that actually work, a consistent bedtime, less alcohol, a dark room. The nightly stage breakdown is the part to stop grading. When your REM number is down and you feel fine, believe your body, not the chart. I wear the ring. I have also learned to leave it in a drawer on the nights I catch myself negotiating with a bar graph before breakfast, which is the most honest sleep advice I own.
