There's a decision that most people who exercise face every morning: how hard should I go today? And for most people, the answer is a gut feel — a vague sense of tiredness or energy that may or may not reflect what's actually happening in their physiology.
Your Apple Watch has been collecting data that could answer this question objectively every single night. The problem is that the raw numbers — HRV of 54ms, resting HR of 58 bpm, 6h42m of sleep with 14% deep — don't automatically become a decision. You'd have to mentally combine four metrics against your own personal baseline to figure out if today is a green day or a red day.
A recovery score does that math for you.
What Goes Into a Recovery Score
Metrya's Recovery Score is a 0–100 composite built from five signals, each weighted by how strongly research links it to actual recovery state:
The critical detail: every signal is compared against your own 30-day rolling baseline, not against population averages. If your normal resting HR is 62 bpm and today it's 68, that's flagged. If your normal resting HR is 55 bpm and today it's 58, that barely registers. Same number, completely different meaning — and only a personal baseline can make that distinction.
What Each Score Range Means
Why Your Personal Baseline Matters More Than Any Number
The most common mistake in interpreting recovery data is treating absolute numbers as meaningful. They're not — at least not without context.
HRV of 38ms. Is that good? For a 60-year-old sedentary adult, it might be perfectly normal. For a 25-year-old competitive athlete, it might represent deep fatigue. For that same athlete on a normal day, it might just be a Tuesday.
Population-average benchmarks are interesting academically but nearly useless for day-to-day decisions. Your baseline is everything. And building that baseline requires consistent measurement over time — which is why Apple Watch (or any HealthKit-compatible wearable) wearing it to bed most nights is the single most valuable thing you can do to make recovery scores meaningful.
Metrya builds its baseline using a rolling 30-day window. This means the baseline updates continuously — capturing your genuine current normal rather than locking in a stale snapshot from months ago. If your fitness improves significantly, your baseline rises to match. If you go through a high-stress period, the baseline adjusts.
What a Low Recovery Score Is Telling You
A low score doesn't mean something is wrong. It means your body is currently under load — and that the marginal cost of adding more stress (training) is higher than usual. Understanding why it's low is where the real value lies.
Common causes of a low recovery score:
High training load. Expected and normal. After a hard training block, your score should be suppressed. That's the adaptation signal working. The question is whether it's recovering on rest days.
Sleep debt. A single short night has surprisingly large effects on HRV and resting HR. Back-to-back poor nights are cumulative. Your score may stay suppressed for several days after a rough sleep period.
Alcohol. One of the most reliable suppressors. Even 2–3 drinks reliably drops next-morning recovery scores for most people. The effect is visible in the data within hours.
Illness onset. A sudden unexplained drop in recovery score — especially when accompanied by elevated resting HR — is often one of the earliest signs that you're fighting something. Rest and monitor closely.
Stress load. Psychological stress registers in the autonomic nervous system and shows up in HRV. Overwork, anxiety, or emotional strain all suppress recovery scores in ways that are physiologically indistinguishable from physical fatigue.
How to Actually Use It: The Practical Framework
Beyond the Number: Asking Why
The recovery score tells you what. An AI health advisor helps you understand why.
Tapping into Metrya's AI Advisor and asking "my recovery score has been under 50 for the past five days — what's going on?" gives the AI access to the same underlying data the score is built from. It can identify patterns you might miss: that your deep sleep percentage has been consistently low, that your resting HR has been elevated since a particular date, or that your HRV trend has been declining for two weeks.
This is the combination that makes the system genuinely useful: an automatic signal that tells you something is off, paired with an AI that can reason over your full data history to explain it.
No AI Key Required
The Recovery Score in Metrya runs entirely on your device. No API key, no subscription, no data sent anywhere. It computes from your local HealthKit data using your own 30-day rolling baseline. You get a daily readiness number and a full breakdown of every component — completely free, completely private.
The AI Advisor layer — the "why" behind the number — is where BYOK comes in. But the score itself is always available.
Know if today is a go or a no.
Free, on-device recovery scoring built from your Apple Health data. No AI key needed. No subscription required for the score.
Download Metrya — Free