What your sleep tracker can and can’t tell you

You wear an Oura ring, an Apple Watch, or a Whoop band. Every morning you check your sleep score. You see numbers for deep sleep, REM sleep, sleep efficiency, and total sleep time. Some mornings the score is 85 and you feel terrible. Other mornings it’s 62 and you feel fine.

The inconsistency is a measurement problem. Your wearable is good at measuring certain things and unreliable at measuring others, and most people don’t know which is which. That gap produces a lot of misplaced anxiety about numbers that don’t mean what they look like they mean, and not enough attention to the numbers that actually matter.

Here’s what your sleep tracker gets right, what it gets wrong, and how to use the data without letting it make your sleep worse.

What your tracker gets right

Sleep timing

Your wearable knows when you fell asleep and when you woke up, usually within a few minutes of the actual time. The accelerometer and heart rate sensor are reasonably good at detecting the transition from wakefulness to sleep. Sleep onset, sleep offset, total time in bed: those numbers you can mostly trust.

This matters more than most people realize. Sleep timing, the consistency of it specifically, is one of the strongest predictors of sleep quality and metabolic health. Your sleep midpoint is the halfway point between when you fall asleep and when you wake up. If you fall asleep at 11 PM and wake at 7 AM, your sleep midpoint is 3 AM. If the next night you fall asleep at 1 AM and wake at 9 AM, your sleep midpoint is 5 AM.

That 2-hour shift has measurable downstream effects on your circadian rhythm, on insulin sensitivity the next day, on cortisol. Research tracking sleep midpoint variability over time has found consistent links to worse metabolic health, higher BMI, and poorer cardiovascular outcomes, independent of how many total hours people are sleeping.

Your tracker captures this reliably. The standard deviation of your sleep midpoint over the past 30 days is one of the most useful numbers it produces. In Protocol’s Sleep Health protocol, the target is a sleep midpoint SD under 30 minutes. We start paying close attention above 45 minutes, and above 60 minutes it gets addressed before almost anything else.

For more on why consistency outweighs duration, read Sleep Consistency Matters More Than Sleep Duration.

Resting heart rate trend

Your overnight resting heart rate, the minimum heart rate recorded while you’re asleep, is a well-validated metric. Your tracker measures this with near-clinical accuracy.

Trend is what matters, not any single night. A resting heart rate that gradually declines over weeks or months usually reflects improving cardiovascular fitness, better recovery, or both. One that’s running above your personal baseline for several consecutive nights is worth paying attention to. It can indicate illness, overtraining, stress, or alcohol.

A single elevated night is noise. Five consecutive elevated nights is a signal worth investigating, and your tracker is reliable enough to surface that pattern.

HRV trend direction

Heart rate variability (HRV) measures the variation in time between heartbeats. Higher HRV generally reflects better autonomic nervous system balance and greater physiological resilience. Lower HRV is associated with stress, fatigue, and poor recovery.

Your wearable measures HRV with reasonable directional reliability over weeks. The absolute number is less important than which direction it’s heading. A month-long rise usually means your body is handling load well. A sustained decline means something is taxing your system: poor sleep, overtraining, chronic stress, or some combination.

The catch is that “over weeks” part. Night-to-night HRV variation is large and normal, and a single bad reading means nothing. Look at your 7-day and 30-day lines, not last night’s number.

For a detailed guide to reading your HRV data without overreacting, read HRV: How to Actually Read Your Heart Rate Variability Data.

What your tracker gets wrong

Sleep efficiency

Sleep efficiency is the percentage of time in bed that you spend actually asleep. It’s clinically meaningful, and sleep researchers and therapists use it extensively. The problem is that your wearable calculates it poorly.

Determining sleep efficiency requires knowing exactly when you fell asleep and exactly when (and for how long) you woke up during the night. Your wearable detects the large transitions (falling asleep, waking up for the day) within minutes. But it struggles with the small ones. Wearables detect fewer than two-thirds of brief awakenings in head-to-head comparisons with polysomnography, and the precise moment of sleep onset is often misclassified by several minutes.

The accumulated error across a night produces an efficiency number that can be off by 10-15 percentage points or more. Your tracker says 82% efficiency. The real number might be 70% or 92%. That range is too wide to base decisions on.

Protocol uses a sleep diary for efficiency, not a wearable. It’s a structured daily log where you record your estimated sleep onset time, number of awakenings, and final wake time. It sounds low-tech because it is. But diary-based sleep efficiency has been validated against polysomnography (the clinical gold standard) and is more reliable for clinical decisions than consumer wearable estimates.

Deep sleep and REM percentages

This is where most people waste the most attention. Your Oura ring says you got 45 minutes of deep sleep. Your Whoop says you got 1 hour and 20 minutes. Same night. Both are wrong, or at least both are well within their error bars and not precise enough to act on.

Consumer wearables infer sleep stages from heart rate, movement, and sometimes blood oxygen. Clinical sleep staging uses an EEG, electrodes on the scalp measuring brain wave patterns directly. The accuracy gap between these two approaches is substantial, and no amount of better algorithms on the wearable side has closed it.

Validation studies comparing consumer wearables to EEG-based polysomnography consistently show:

• Total sleep time: reasonably accurate (within 20-30 minutes)
• Sleep onset and offset: reasonably accurate
• Deep sleep (N3): accuracy varies widely by device, with error bars large enough that night-to-night comparisons are meaningless
• REM sleep: same problem, with large error bars and inconsistent agreement with EEG
• Light sleep (N1/N2): typically the catch-all category that absorbs classification errors from the other stages

A deep sleep reading of 38 minutes versus 52 minutes on consecutive nights might reflect a genuine difference, or it might reflect the sensor’s inability to distinguish N2 from N3 during a particular transition. You can’t tell which.

Ignore the stage percentages. They will fluctuate, and the fluctuation tells you almost nothing. If your Oura ring says you got 30 minutes of deep sleep, the right response isn’t “my deep sleep is bad.” It’s “my ring estimated a number with wide error bars and I have no idea what the real figure is.”

Diagnosing sleep disorders

Your wearable cannot diagnose sleep apnea, restless legs syndrome, periodic limb movement disorder, or any other clinical sleep condition. It may produce data that hints at a problem (elevated heart rate, high movement, low blood oxygen), but the diagnostic threshold requires clinical testing.

If you suspect a sleep disorder, the pathway is a validated screening questionnaire (STOP-BANG for sleep apnea, for example) followed by a home sleep test or in-lab polysomnography. Your wearable data can support the conversation with your doctor. It cannot replace the diagnostic test.

What Protocol tells every member

During the first session of the Sleep Health protocol, every member hears this:

“Your ring or watch is excellent at tracking WHEN you sleep and how CONSISTENT you are. It is not accurate enough to tell you HOW you slept in terms of sleep stages. We use your sleep diary for that. Ignore the deep sleep percentage. It will vary widely and is not actionable.”

That single reframe eliminates most of the tracker-related anxiety members bring to their first session.

Orthosomnia: when tracking makes sleep worse

There’s a documented phenomenon called orthosomnia, a preoccupation with achieving “perfect” sleep tracker data that ends up worsening sleep. The loop runs like this: you check your score first thing in the morning, a low number triggers frustration, you go to bed that night worried about fixing it, the worry makes it harder to fall asleep, and the next score is worse. Then you check it again.

The warning signs are fairly recognizable. You check your sleep score the moment you wake up and feel distressed by the number. You notice anxiety before bed about being tracked. You make disproportionate life adjustments to optimize scores (canceling plans, refusing to travel). Or you catch yourself thinking “I can’t sleep because I’m worried about not sleeping.”

If this describes you, the intervention is counterintuitive but effective: take the wearable off before bed. Remove it 1 hour before your target lights-out time. Continue wearing it during the day for activity tracking, but stop wearing it for sleep. Simplify your sleep diary to three items: bed time, wake time, and a 1-5 quality rating.

Do this for at least 2 weeks. Most people find that their sleep improves once the performance pressure is removed. The tracker can be reintroduced later, once the anxiety cycle is broken.

What to actually track (and how often to check it)

Check weekly, not daily:

• Sleep midpoint SD (consistency), the most actionable number your wearable produces
• Average total sleep time
• Resting heart rate 7-day trend
• HRV 7-day trend

Check monthly:

• 30-day sleep midpoint SD
• 30-day resting heart rate trend
• 30-day HRV trend
• Whether your average bedtime and wake time have drifted

Stop checking entirely:

• Deep sleep percentage on any given night
• REM percentage on any given night
• Your sleep “score.” It’s a proprietary composite that blends reliable and unreliable inputs in ways you can’t audit.

Checking daily is mostly noise. A 7-day average of your sleep midpoint smooths out the Friday you stayed up late and shows you whether your overall pattern is stable or drifting. That’s the signal. You can’t see it day by day.

How Protocol uses wearable data

In the Sleep Health protocol, the wearable is one of two data sources. It handles timing, consistency, resting heart rate, and HRV. The sleep diary handles efficiency, onset latency, and subjective quality. Each fills in what the other misses.

The wearable data also connects across protocols. If you’ve done the Metabolic Health protocol, your coach can pull your CGM data alongside your sleep data: “Here’s your fasting glucose on mornings after you slept under 6 hours. Here’s your fasting glucose after 7-plus hours.” Seeing your own numbers side by side tends to land differently than any study statistic would.

Chronotype gets assessed using the MEQ-5, a five-question questionnaire that puts you in morning, intermediate, or evening type. That classification shapes when we schedule light exposure, when we set your caffeine curfew, and how we time your training recommendations around your natural rhythm.

The wearable supports all of this. It doesn’t replace any of it. Used well, it’s a useful input. Used poorly (obsessing over stage percentages, checking the score every morning like a grade), it can make things worse.

---

Ready to find out where you stand? Protocol’s Foundation Assessment measures what your annual physical misses (ApoB, HOMA-IR, DEXA body composition, VO2 max) and builds a specific action plan from the data.

Book a Discovery Call →