Your wearable gives you a sleep score every morning. But what is it actually measuring, and how much should you trust it? Here's what's behind the number.
Sleep Scores Are Composite Indexes
No sleep score measures a single thing. Every major device synthesizes multiple physiological signals into a single number — but each device chooses different signals and different weightings. This is why comparing sleep scores across devices is meaningless. An Oura 78 and a Fitbit 78 are built from different inputs.
The signals most devices use:
Sleep Duration
How long you were actually asleep (not just in bed). Almost every algorithm weights this heavily. Sleeping 6 hours structurally caps your score regardless of other factors. The target duration for most adults is 7–9 hours.
Sleep Architecture (Stages)
Modern wearables estimate how much time you spent in each sleep stage: N1 (transitional light sleep), N2 (consolidated light sleep), N3 (deep/slow-wave sleep), and REM. They do this by combining wrist movement patterns with heart rate variability patterns — not EEG, which is the clinical standard. This means consumer stage estimates have error margins of 15–30%.
Deep sleep (N3): Physical recovery, immune function, growth hormone release, brain waste clearance. Target: 15–20% of total sleep, or roughly 90 minutes in a 7.5-hour night.
REM sleep: Memory consolidation, emotional regulation, creativity. Target: 20–25% of total sleep, or roughly 90–105 minutes in a 7.5-hour night.
Light sleep (N2): Motor memory, sensory consolidation. Makes up the bulk of sleep (50–60%) and isn't a problem in normal amounts.
Heart Rate Variability (HRV)
HRV measures the variation between consecutive heartbeats. Higher HRV generally indicates a well-recovered, low-stress autonomic nervous system state. Oura, WHOOP, and Garmin all weight HRV heavily because it reflects systemic recovery, not just what happened last night. HRV is suppressed by alcohol, overtraining, illness, and psychological stress.
Resting Heart Rate (RHR)
Your overnight resting heart rate reflects cardiovascular recovery. Lower RHR is associated with better fitness and recovery. Alcohol consistently elevates overnight RHR — this is one reason drinking shows up clearly in sleep scores.
Sleep Efficiency
Time asleep divided by time in bed. Below 85% suggests either fragmented sleep (many awakenings) or prolonged time lying awake. Fragmentation can come from sleep apnea, environmental disruption, or full arousal disorders.
Restlessness
Movement during sleep is a proxy for arousals and transitions between sleep stages. High restlessness indicates fragmented, lower-quality sleep.
What a High Score Actually Means
A consistently high sleep score (85+ on most devices) indicates:
- You're sleeping long enough
- Your sleep is efficient and consolidated (not fragmented)
- Your deep and REM sleep are within healthy ranges
- Your HRV is strong relative to your personal baseline
- Your resting heart rate is normal or low
A high score doesn't mean you had perfect sleep — all consumer devices have error margins. But a consistently high score strongly correlates with good sleep health.
What a Low Score Tells You
A low score is a diagnostic signal, not a verdict. The useful question isn't "how do I make this number go up" — it's "which component is low, and why?"
Open your app's sleep breakdown:
- Low duration: You're not spending enough time asleep. Go to bed earlier.
- Low deep sleep: Most commonly alcohol, late caffeine, warm room, or irregular schedule.
- Low REM: Alcohol (strongest effect), insufficient total sleep, inconsistent schedule.
- High overnight stress / low HRV: Alcohol, overtraining, illness, or psychological stress.
- Low efficiency: Fragmented sleep — look for apnea, noise, or light.
Should You Trust Your Sleep Score?
Consumer sleep tracking is accurate enough to track trends and measure the effect of behavior changes. It's not accurate enough for clinical diagnosis.
The most useful application: test one habit change for 10–14 days and compare your scores before and after. The trend is reliable even if the absolute numbers have error margins.
Tracking with SleepBetter.ai alongside your wearable data gives you both — the behavioral log that explains the score, and the score that confirms the behavior change is working.