Guide
Recovery is measured from the heart. Fatigue is not only in the body.
Nearly every readiness score on the market is HRV with a wrapper. That measures one real thing well and leaves out the part of fatigue that lives above the neck.
Guide
Nearly every readiness score on the market is HRV with a wrapper. That measures one real thing well and leaves out the part of fatigue that lives above the neck.
Open any recovery-focused wearable and the number on the front is, underneath, heart rate variability measured overnight, compared against your own rolling baseline, and adjusted for sleep and recent load. That is a defensible construction and HRV is a real signal — the honest account of what it does and does not mean is its own guide.
What follows is what it leaves out, and how much of the alternative is actually established.
Exercise physiology distinguishes peripheral fatigue — happening in the muscle itself, in its contractile machinery and metabolic environment — from central fatigue, a reduction in the drive the nervous system sends to the muscle.
The distinction is not theoretical. It is demonstrated by superimposed electrical stimulation: stimulate a muscle during a maximal voluntary contraction, and if extra force appears, the muscle had capacity the nervous system was not using. That gap is central fatigue, and it is measurable in a laboratory.
HRV speaks to autonomic state, which is related to but not the same as central drive. Neither is directly measured by a wrist strap.
Sleep composition — well supported. Slow-wave sleep is the stage most consistently associated with physical restoration, and it is defined by EEG. A device that measures slow waves is measuring the thing itself rather than estimating it from pulse. This is the strongest case for EEG in a recovery context.
Slow-wave enhancement — promising, narrow. Acoustic stimulation timed to the rising phase of a slow wave can increase slow-wave activity, and this has been replicated in controlled studies. It requires detecting the wave and responding within tens of milliseconds, so it needs EEG by definition. Whether it produces durable benefits outside the laboratory is much less settled than the phenomenon itself.
Sleep pressure — established in principle. Slow-wave activity in early sleep tracks how long you have been awake, which is one of the better-quantified relationships in sleep science. Turning that into a per-person daily readout is a further step.
Mental fatigue — real effect, not a product. Prolonged demanding cognitive work reduces subsequent physical endurance performance, which is a genuine and replicated finding. There is no validated EEG readout that quantifies it for an individual on a given day.
Central fatigue as a daily number — not established. This is where a recovery product would most like to go and where the evidence does not currently support going. Anyone offering it is ahead of the literature.
Recovery is dominated by things no wearable senses: what you ate, how much you drank, whether you are ill, what else is going on in your life. A readiness score is a partial measurement of a system where the unmeasured inputs are frequently larger than the measured ones.
Which argues for reading these numbers as one input among several rather than as a verdict — and for treating any device that sounds certain with proportionate suspicion.
Sleep duration and regularity are the highest-value, best-measured, most actionable things any of these devices give you. If a recovery number moves you toward sleeping longer or more consistently, it has done its job, and the mechanism by which it convinced you matters less than the outcome.
If it moves you toward anxiety about a number that varies for reasons no one can see, it has done the opposite.
Further reading