How recovery is calculated
Recovery is a 0–100 reference score that compares today's HRV, resting heart rate, sleep, respiratory rate and wrist temperature against your own trailing 28-day baseline, then weights them — it measures you against your usual, not against other people.
Inputs and weights
- Heart rate variability — 35%, higher is better
- Resting heart rate — 25%, lower is better
- Sleep — 25% (duration blended with deep-sleep share)
- Respiratory rate — 10%, any deviation costs (both directions)
- Sleeping wrist temperature — 5%, also both directions
The method
Each vital is first expressed as its deviation from your personal baseline in units of that baseline's spread (a z-score), then direction-adjusted: high HRV is good, a high resting heart rate is not, while respiratory rate and wrist temperature cost points in either direction because any deviation is the signal.
The calibration: a vital exactly at baseline reads 70, and each unit of spread moves it ±15. So a score near 70 means "same as your usual", not "barely passing".
A missing vital is not counted as zero — its weight is redistributed across the rest. The score is withheld only when HRV, resting heart rate and sleep are all absent, because at that point no core signal remains.
How to read it
67 and above reads as high, 34–66 as medium, below 34 as low. Those bands drive a colour and one line of guidance — nothing more.
A low score does not mean something is wrong. Hard training, short sleep, jet lag, alcohol and the onset of a cold all push it down, which is the score doing its job. Several low days in a row is what deserves attention.
Nor does it need to stay high. Consistently high usually means recent training has not been hard enough to leave a mark.
Limitations
- This is a reference index, not a medical diagnosis, and not a substitute for professional advice.
- The baseline needs at least 7 days of data; with less, the score is less stable.
- If you wear more than one device, an overlapping metric takes its values from a single source for the whole window — mixing two vendors' HRV algorithms would poison the baseline (WHOOP reports rMSSD, Apple reports SDNN; on the same wrist those are different numbers).
- Occasional wearers see more volatility, because the baseline itself rests on fewer nights.
References
What it supports：Definitions, measurement windows and population norms for HRV metrics.
Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology. Heart rate variability: standards of measurement, physiological interpretation, and clinical use. Circulation. 1996;93(5):1043-65.
What it supports：The standard method for measuring and interpreting HRV, including when time- and frequency-domain metrics apply.
Plews DJ, Laursen PB, Stanley J, Kilding AE, Buchheit M. Training adaptation and heart rate variability in elite endurance athletes: opening the door to effective monitoring. Sports Medicine. 2013;43(9):773-81.
What it supports：HRV should be read against a personal rolling baseline rather than population norms — the direct basis for why we publish no absolute thresholds.
What it supports：The physiological link between HRV and autonomic regulatory capacity.
What it supports：Meta-analytic evidence linking psychological stress to reduced HRV.
Related metrics
Every metric on this page is a reference index for observing your own trends. It is not medical advice and cannot replace a professional diagnosis. If you have health concerns, consult a qualified clinician.
Source module: lib/vitality/score.ts · computeReadiness
Download Vita free — see your Body Age and Recovery in minutes
Download on the App Store