Your morning readiness, with the working shown

Each morning you see how last night compared with your own previous 60 days, and what moved it: HRV against your range, how much sleep you're short, anything worth a second look, and what tends to follow the tags you log. There's an estimate too, but the point is that you can see why, so a number never moves without a reason you can check.

What the Today card looks like

The card shows how last night compares with your usual, and names what moved it. Tap it and every input appears next to your own usual, with the effect it had and the number of nights behind it.

TodayEstimate

44Around your usual
Readiness estimate

Data confidence: high

• Lowered by: Sleep
• 7-night HRV: within your range
• Recent sleep shortfall: 1 h 40 min

Illustration with made-up numbers

How today is calculated

Signal	Last night	Your usual	Effect
HRV	44 ms SDNN	47 ms (60 nights)	About your usual
Resting heart rate	55 bpm	54 bpm (60 nights)	About your usual
Sleep	6 h 5 min	7 h 20 min (58 nights)	Lowered it
Breathing rate	14.6 /min	14.5 /min (55 nights)	About your usual
Wrist temperature	–	–	No reading last night

Model sensitivity: 40 to 50. That's the range you get by leaving out one signal at a time.

What you get each morning

• A readiness estimate with its range

  One number for how last night compares with your usual, where 50 means at your usual. It comes with the range you'd get leaving out any one signal, and how much history it rests on.

  How it's worked out

• HRV status against your own range

  Your last 7 nights of HRV against your 60-day range: within, below or above. A single low night doesn't flip it, because it needs at least 3 readings in the week.

  How it's worked out

• Recent sleep shortfall

  How far behind your own estimated need you are, fading over about a week. Your need is learned from your nights and kept between 7 and 10 hours.

  How it's worked out

• Changes worth a look

  A quiet note when two overnight signals have both moved well away from your typical range. It's built to stay silent on ordinary nights: at most 2 in any 7 days.

  How it's worked out

• Tag patterns

  Tap tags like alcohol, sauna or late meal, and see how the nights after them compare with the nights after days without them. Nothing is called a pattern before 10 nights with and 10 without.

  How it's worked out

• A fitness age range

  Your Apple Watch cardio fitness estimate against a large study of healthy adults, shown as an age range, because watch estimates have a wide margin. Never one exact age.

  How it's worked out

How to read the estimate

• 50 is your usual, not a pass mark. The estimate is 100 × Φ(Z) of a weighted average of how far each signal is from your usual, so 50 means at your usual and it doesn't mean "50% recovered". The card reads 40 to 60 as "around your usual".
• The range tells you how much to trust one number. Model sensitivity recalculates the estimate leaving out one signal at a time. If the range is wide, one signal is doing most of the work, and the "Lowered by" line names it.
• Data confidence is about history, not accuracy. It's low, medium or high depending on which signals you have and how many nights stand behind each, full at 42 nights.
• No data means no number. It needs last night's sleep, plus HRV or resting heart rate, and 14 nights of history. A missing night is never filled in, and the card says what's missing.
• HRV here is SDNN. Apple Watch records HRV as SDNN, which runs differently from the RMSSD some other apps show, so the numbers aren't comparable across apps. The app labels it and never mixes the two.

How other readiness scores explain themselves

Every score below compares you with your own recent nights. They differ in what their middle means and in how much of the method you're allowed to see. A 2025 review of 14 wearable composite scores found none disclosed its exact formula (Doherty et al.).

From each company's own help pages, read on 27 October 2026: Apple, Oura, WHOOP, Athlytic. Different scales mean a 70 in one app isn't a 70 in another.

If you have Apple's Readiness app

Apple's Readiness app runs on Apple Watch Series 12 and Ultra 4. It also weighs activity and training load and updates through the day. Use both if you have one: MetricBridge's estimate is a fixed morning look at overnight signals, and it shows each input next to your usual. On an earlier Apple Watch, it works from the HRV, resting heart rate and sleep your watch already records in Apple Health. The two numbers will differ, because they measure different things on different scales.

What it isn't

• It isn't a diagnosis. It describes patterns in your own wearable data as a deviation from your baseline. It doesn't detect illness and doesn't tell you whether exercise is safe. If you feel unwell, a healthcare professional can help.
• It isn't a battery. There's no energy gauge that drains through the day, because there's no published model for one.
• It isn't sent anywhere. Everything is computed on your iPhone. The app reads Apple Health and writes nothing back, and tags stay in the app's own file.
• It isn't fitted to anyone else. The weights are fixed design choices, published with a version number. If they change, the methods page changes first.
• It isn't a medical device. MetricBridge is a general-wellness app, not a regulated medical device. The estimate doesn't account for conditions or medications and shouldn't be the only thing you rely on to decide whether to train.

Common questions