Does a readiness score predict performance? What the evidence says

What the score is made of

Oura's readiness is a weighted composite of things measured overnight: resting heart rate and how it compares to your baseline, HRV against your baseline, body temperature deviation, how quickly your heart rate settled during the night, recent sleep balance and recent activity balance. Every one of those inputs is backwards-looking. None of them observes tomorrow. The score is a summary of the night that just happened, dressed in the grammar of a forecast.

What 'predicts performance' would have to mean

For the claim to hold, a low score would have to reliably precede a measurably worse effort — slower times, lower power, fewer reps at the same perceived effort — and a high score a better one, across many people and many days. That is a demanding test, and the reason it is demanding is that day-to-day performance variation in trained people is small and dominated by other things: pacing, fuelling, weather, motivation, where you are in a training block. A signal has to be quite strong to show up over that.

What the evidence actually shows

Group-level relationships between HRV and training response are reasonably well established: HRV-guided training programmes have in several trials produced outcomes as good as or slightly better than fixed programmes, mainly by moving hard sessions away from suppressed days. That is genuine and useful. But that is not the same as a proprietary composite score predicting your individual performance tomorrow, and studies of consumer recovery scores against athletic output are few, small, and mixed. Validation of a device's sleep staging or heart rate accuracy — which manufacturers do publish — says nothing about the predictive validity of the score built on top. Treat any confident claim in either direction with suspicion, including confident dismissal.

The self-fulfilling problem

There is a further complication that is easy to miss: seeing the score changes the day. Told they slept badly, people report more fatigue and perform slightly worse — this has been shown experimentally with sham feedback, where the number was fabricated. So even a real association between a low score and a bad session is partly the score causing the bad session. This is not an argument for ignoring the data; it is an argument for looking at it after you have taken your own reading of how you feel.

Where the score genuinely earns its place

Read backwards instead of forwards, readiness is informative. A sharp unexplained drop with a temperature rise is a decent early flag for an oncoming infection, often a day before symptoms. A run of suppressed mornings after a heavy block is real evidence of accumulated load. A reliable dip after certain evenings tells you what those evenings cost. In each case the score is describing something that already happened to your physiology, which is what it actually measures — and over weeks, that description is trustworthy in a way a single morning is not.

Using it without letting it decide

Take your own rating of energy and soreness before you open the app; that keeps your judgement uncontaminated and gives you something to compare against. Use the trend, not today's value, to shape a training week. Reserve the single-day score for the extremes — a big drop with a fever-ish temperature deviation is worth respecting; a score of 71 versus your usual 78 is not information about your legs. And when the number and the body disagree persistently, that gap is itself the interesting data.

Vitra shows readiness and its underlying inputs as trends against your own rolling baseline on your Mac or PC, so you can see whether a low morning is a blip or a pattern, and compare it with how you actually felt and performed. Its guidance is research rules and rolling statistics computed on your own machine — not a verdict on your day. General information, not medical advice.

Frequently asked questions