A set of twelve curls and a heavy single take the same minute. They don't cost the same.
Every training log treats them identically. Session length times how hard it felt — that's how training load has been measured for decades, and it's genuinely useful. But it can't tell the difference between an hour of moderate sets and an hour of near-maximal singles, and anyone who's done both knows they don't leave you in the same place.
Tapertrophy weights each session by what kind of training it was, before anything else happens. That's the whole difference, and everything else in the app follows from it.
session cost = how long × how hard it felt × what kind of training
× what you worked
The first two are standard. The third is the part nobody else does.
Max-effort work carries the highest weighting. Explosive work — Olympic lifts, plyometrics, sprints — is close behind. Heavy strength is next. Muscle-building work sits at the baseline. Conditioning is slightly above it. Easy cardio is well below. Mobility barely registers.
Lower body and full body carry slightly more than upper body. Arms alone carry slightly less.
Being straight about these numbers: they're informed estimates, not measurements. They're built on the idea that harder training costs more, which is well supported, and calibrated by judgment on the specifics, which is not. They'll be tuned as real data comes in. When they change, the app records which version scored each day, and days already logged keep the numbers they were scored with.
Your last 7 days against your last 28. A rough way of noticing whether you've ramped up quickly.
Being straight about this one too: this measure is disputed. Researchers have argued its statistical properties make it unreliable, and that it shouldn't be used to predict injury. The app uses it as a rough indicator of a workload spike and nothing more — and it doesn't use it at all until you have enough history for it to mean anything, because with thin data the number is noise.
Your four morning sliders, scored against your own recent average rather than against other people.
This matters more than it sounds. Someone who rates themselves 2 out of 5 on a good day isn't broken — that's their normal, and the app treats it as such. What counts is the change from your own baseline, not the absolute number.
Sleep debt against a 7.5-hour reference, over three days. How recently you did something demanding. And a rough nutrition signal from the three-state food tag.
Everything starts from a neutral figure and moves up or down from there. What you see is the total.
And it shows you every line. Tap the number and the whole calculation opens up. Nothing is hidden, nothing is proprietary, and you can check the arithmetic. No recovery app or wearable in this category discloses how its score is calculated. This one does, because a number you can't interrogate is a number you have to take on faith.
It can't measure anything about your body. No heart rate, no HRV, no sleep stages, no hormones. It reads what you type in and does arithmetic on it. That's the whole mechanism.
It doesn't measure your nervous system. Lifters use "CNS fatigue" for the wrung-out feeling after heavy work, and the app is built on the idea that harder training costs more — which is well supported. The specific mechanism people mean by that phrase is contested in the research, and this app doesn't claim to measure it.
It isn't a doctor. It doesn't diagnose anything, doesn't know your medical history, and nothing in it is medical advice.
It won't tell you what to do. It reports what your log says. The verdict is phrased as an opinion — if it were me — specifically so you can disagree with it. You know things about your own body that the app never will.
The app compares you to yourself. Until it has enough of your history, it's comparing you to assumptions.
Day 1 — your load is tracked immediately. The read works, but it's using defaults.
Around day 14 — it starts comparing your check-ins against your own average instead.
Around day 21 — the workload comparison comes in, once there's enough history for it to be stable.
Around day 28 — the baseline is full. Everything is running on your data.
You can shortcut most of this by logging your last two weeks when you install, which the app offers on day one.
The app tells you where it's up to on the Today screen. That's deliberate. An app that showed you a confident number on day two would be making it up.
The app exists to tell you when to back off. An app that punished you for resting would be arguing with itself.
So there's nothing to break, no chain to maintain, no guilt for a day you missed and no reward for a day you trained. Log a rest day and it counts as real information, because it is.
The load calculation builds on session-RPE, a long-established method for estimating training load. The acute-to-chronic comparison is a standard workload monitoring approach, used here with the limitations noted above.
The reaction test exists because cognitive slowing shows up consistently in research on under-recovered athletes — reaction times measurably slower than their own normal. It's the only objective measurement in an otherwise self-reported app, and it needs a couple of weeks of your own tests before it means anything.
The recovery guidance in the app is tiered by how strong the evidence actually is. Sleep extension is the only entry marked as strongly supported, because it's the only one that is.
This app estimates your recovery from what you tell it. It's a structured record of your own training and how you felt, with arithmetic applied to it and every step shown.
That's less than a wearable claims and more than a training log gives you. It's also the truthful description, which matters more than an impressive one.