Consumer wearables now report sleep duration, sleep stages, HRV, resting heart rate, and derived "readiness" or "recovery" scores with a precision that can feel authoritative. The numbers look clinical. The reality is that they are estimates — useful directional signals built on hardware and algorithms with meaningful limitations. Understanding those limitations prevents the data from becoming a source of anxiety rather than information.
What wearables actually measure and how
Most consumer wearables measure sleep and recovery through two primary sensors: an accelerometer (movement) and a photoplethysmography sensor (PPG, which uses light to detect blood flow through the skin at the wrist or finger).
Sleep staging — distinguishing between light sleep, deep sleep, and REM — is inferred from combinations of movement patterns and heart rate data. Studies comparing consumer wearables to polysomnography (PSG), the clinical gold standard for sleep measurement, consistently find that consumer devices achieve 60–75% accuracy in sleep stage classification. They are reliably accurate at detecting whether you're asleep or awake (typically >90% accuracy), but substantially less reliable at which specific stage you're in at any given time.
HRV measured through PPG is less accurate than ECG-based measurement. Motion artifact, sensor position, and skin characteristics all introduce noise. What wearables report as HRV is an estimate with device-specific measurement approaches that differ from each other and from clinical standards.
What this means for interpreting the numbers
Sleep staging numbers (2 hours of deep sleep, 1.5 hours of REM) should be treated as directional estimates rather than precise measurements. Changes in these numbers from night to night reflect a mix of actual physiological variation and measurement noise that cannot be reliably separated.
Resting heart rate is the most reliable metric wearables measure. It requires minimal inference and tracks well against ECG reference standards. Trends in resting heart rate — particularly a sustained elevation of 5+ bpm above your normal baseline — are a meaningful recovery signal.
HRV is individually specific to a degree that makes absolute numbers almost meaningless. Someone with an HRV of 20 ms and someone with an HRV of 80 ms can both be in excellent health — these are simply different baselines. What matters is how your HRV moves relative to your own normal range. A sustained downward trend from your baseline is meaningful; a single low reading is noise.
Readiness and recovery scores are proprietary calculations that combine these underlying signals with their own weightings and algorithms. They can be directionally useful but should not be treated as instructions.
How to use these signals well
Think of recovery metrics as one input among several, not as a directive. A low readiness score on a day when you feel well, have slept a reasonable number of hours, and have no unusual fatigue is not a strong reason to skip a planned session. A low score combined with subjective tiredness, elevated resting heart rate, and poor sleep quality across several consecutive nights is a more meaningful pattern.
The most useful framing: wearable recovery data is most valuable over weeks and months, revealing patterns in how your body responds to different training loads, sleep habits, alcohol, stress, and illness. Night-to-night variation is largely noise. Trends across weeks are signal.
Intentra surfaces sleep and recovery data with this framing in mind — showing trends rather than flagging individual nights, and presenting the data as context rather than instruction.