Overview
“Sleep debt” is a useful mental model and a poor metric. Many apps present it as a running balance (hours owed, hours repaid), which implies an accounting relationship that the underlying evidence does not support.
Grio does not keep a sleep-debt counter, and it does not publish a readiness score. This page explains what it does instead, because the question behind both (am I under-recovered right now?) is a good one that deserves a real answer.
How it’s calculated
The three things Grio actually tracks
| What you might call it | What Grio has | Window |
|---|---|---|
| Last night’s sleep | The sleep contributor to Condition Score | One night, absolute curve plus your baseline |
| Accumulated shortfall | The Sustained Recovery Trend | Last 7 local dates vs the preceding 60 |
| Am I ready to train? | The Target Strain Zone | Your own recent strain history |
Each night is scored on its own. The sleep contributor grades that night’s duration on an absolute curve (with the plateau at roughly seven and a half to nine hours), adjusted by your personal baseline, then refined slightly by efficiency, regularity, and stage mix. There is no carried-forward balance.
Accumulation shows up as a trend, not a total. Because your 30-day baseline adapts, a genuinely worse fortnight will eventually be absorbed into “usual” and stop looking unusual. That is exactly the failure a rolling baseline has, and it is why Grio runs a second, non-overlapping comparison: the last seven local dates against the preceding sixty. When HRV or resting heart rate has stayed meaningfully different across that gap, it is reported as a concern, and Condition takes a small bounded deduction for it.
Readiness is answered by the training surface, not the recovery one. The Target Strain Zone sets today’s load range from your own recent strain history, with your Check-In as evidence; a Check-In reporting fatigue lowers it. Condition Score feeds that target as one input. It is never the target itself and never a permission gate on exercise.
Why no debt counter
Three reasons, all of them about honesty:
- The exchange rate is unknown: How much of one lost hour a later long night repays is not established well enough to publish as a number.
- It would rewrite history: a debt balance either updates past days or contradicts them. Grio’s dated readings do not move after you have seen them.
- The trend already answers the question, using measured physiology rather than an arithmetic assumption about hours.
Edge cases and fallbacks
- A missed night: the Check-In’s reported sleep fills the contributor.
- A very long night lands on the curve’s plateau. It is not penalised for being unusual and it is not extra credit.
- The sustained comparison is not mature yet: Grio shows a Recovery Context Stage describing what has been observed, without assigning a normal/abnormal direction.
How to read your score
- A single bad night shows up in one day’s Condition Score and then leaves. That is expected; it is only one night.
- A run of bad nights shows up twice: in each day’s score, and eventually in the sustained trend. The second one is the signal worth acting on.
- A trend concern is not a diagnosis: it reports an observed direction over a stated timeframe, and nothing more.
The misreading to avoid is treating a recovered Condition Score as evidence the deficit is repaid. The score describes today’s observable context. It is not an audit of the week.
What to do with it
- Protect the floor, not the average: Consistently avoiding very short nights moves the score far more than occasionally sleeping ten hours.
- When the sustained trend flags a concern, change something structural (bedtime, training volume, or alcohol), rather than adding a recovery ritual on top of the same week.
- Let the Target Strain Zone size your training: it already reads your recent load, and it already lowers itself when your Check-In reports fatigue.
Grio will not silently rewrite a committed Daily Plan or a durable goal on the strength of any of these signals. They inform; you decide.
Troubleshooting & FAQs
- Can I catch up on sleep at the weekend?
- Partially, and Grio will show you the part it can see. A long weekend night lands on the sleep curve's plateau and lifts that day's contributor; it does not retroactively repair the scores of the short nights before it. Those days keep the reading they had.
- Where is my readiness score?
- There isn't one, deliberately. Condition Score is whole-body context and is explicitly not a judgment of training readiness. The surface that speaks about today's training is the Target Strain Zone, which reads your own recent load, and Condition is one input to it rather than a gate on it.
- Three short nights in a row barely moved my Condition Score. Why?
- The rolling baseline adapts, which is exactly what makes a personal comparison useful and also what hides a slow decline. That is why the separate Sustained Recovery Trend compares your last seven days against the preceding sixty, which is why a concern there applies a bounded deduction to the score.