Your ring said rest. You felt fine. Or the score was green and you felt wrecked. That argument is not a bug in you. It is what happens when a derived number is allowed to outvote a body that can still talk.
Wearables are excellent at movement, overnight heart rate, and how long you were down. Those are sensors. A composite that tells you to stop moving is an opinion. The missing channel in most wearable apps is you: stress, soreness, energy, food noise, the night that felt terrible while the app printed a 92.
A system that cannot hear how you feel will overfit a noisy estimate. That is not anti-data. It is anti-oracle.
Two Mismatches Everyone Knows
Green score, wrecked body. You slept in a hotel, ate late, woke up thick-headed, and the app still called the night a win because duration looked fine and staging guessed well. You train through it because the number said you could, then wonder why the session was garbage.
Red score, ready body. Staging called REM low. Readiness dropped. You felt recovered, had protein, and had a lift planned. You skipped it because a pie chart vetoed the calendar. A week of those vetoes is how people on a GLP-1 quietly lose the lifting habit that protects muscle.
Both failures come from the same hierarchy: the derived score sits above lived experience instead of next to it.
Why The Score Drifts From The Body
- Staging is an inference. Light, deep, and REM from a finger are estimates. A 2025 sleep-lab comparison put four-stage accuracy near a coin flip even when sleep-versus-wake was much stronger. Details in why that layer should stay beta.
- One night is noisy. Alcohol, travel, a late meal, a sick kid, a new mattress. HRV and resting heart rate get useful as a week, not as a morning verdict.
- Illness and hormones show up in you first. Sometimes the score lags. Sometimes it fires early. Either way, a check-in is data, not a diary entry you feel guilty for.
Treat The Check-In As A Sensor
Subjective ratings get a bad reputation because people use them instead of measurement. Used next to heart rate, sleep duration, and training load, they catch the nights the algorithm cannot see.
The useful version is boring and specific. Not a paragraph in Notes. A 1-to-5 on stress, energy, soreness, and, if you are on a GLP-1, food noise. Three days of that beats one dramatic journal entry.
| Today's call | What to trust | What not to do |
|---|---|---|
| Session intensity | How you feel + this week's training load + resting HR trend | Cancel because REM was “low” |
| Protein floor | The target you already set. Appetite is not the target | Eat less because Sleep Score dipped |
| Recovery story | 7-day HRV and resting HR, plus check-ins | Rewrite the week off one morning tile |
The Simple Rule
Heart rate variability and resting heart rate: trend them over a week. Sleep duration: protect the bedtime. Staging pie: curiosity. How you feel plus the training you actually did: decide the session.
If you are 45 to 64 and on tirzepatide or semaglutide, that rule is not a biohacker nicety. A readiness veto that steals two lifting days a month is a muscle problem wearing a recovery costume. Pair it with the muscle-preservation playbook and reading HRV, sleep, glucose, and labs as one system.
What Mallet Does With The Vote
Mallet readiness is a body-state gauge, not a diligence score. Sleep still counts. So do HRV, resting heart rate, and training load. How you feel is not flavor text. Daily check-in stress is a scored factor, so a week of “I am not okay” can move the number even when the ring is cheerful.
The number still does not get to write protein, labs, or the lifting plan by itself. That is the point of putting the check-in in the same system as the wearable instead of under it.
Log how you feel the same way you log a set. Mallet scores the check-in next to sleep and heart data so a red tile cannot quietly delete your training week. Get early access →
This article is for education, not medical advice. Do not ignore persistent fatigue, pain, or illness. Those belong with a clinician, not a readiness tile.