UNDERBARK

What a sleep score can and cannot explain

A sleep score can summarize available records. It cannot explain every part of the night, diagnose a problem, or decide what is safe for you.

A score is a summary of available records

A wearable records or estimates selected events. Underbark assembles the sleep records available to it and applies its own interpretation. You experience the night and the following day. Those are related kinds of information, but they are not the same thing.

The Sleep Score is software interpretation, not a direct sensor measurement or a complete explanation of the night. It can make the available record easier to scan while leaving room for details the record cannot observe, including how rested or unwell you feel.

What Underbark includes

Underbark combines five component categories into its Sleep Score. These are app interpretation categories, not clinical standards or population targets.

  • Duration describes how much sleep appears in the assembled night.
  • Recorded stages describe stage categories recorded in Apple Health by the device or app that supplied the sleep record. When Apple Watch supplied the record, its stages are estimates from Apple’s system for estimating sleep stages. A record from another device or app does not inherit that Apple Watch estimate. This tells you where the sleep record came from.
  • Time asleep compared with time in bed distinguishes recorded sleep from the wider time you had available for sleep.
  • Consistency compares bedtime timing with recent usable history only when at least three bedtime records exist. Before then, Underbark uses a neutral fallback while history is insufficient. That fallback is not a personal comparison or a preferred bedtime.
  • Recorded awake periods are the awake periods in the sleep record that interrupt the assembled sleep episode.

The combined score can describe how these available records compare across the categories Underbark uses. It cannot establish why a night looked that way, prove restoration, or turn an app setting into a biological rule.

The score also does not make each component a complete explanation. A duration component does not identify why sleep was shorter. A stage category is an estimate from the device or app that supplied the sleep record, not a laboratory reading of every moment. Recorded awake periods do not explain every awakening you remember.

Recorded sleep and felt sleep are different evidence

Recorded structure and your experience can point in a similar direction, or they can disagree. Research on objective measures and subjective ratings supports treating them as related but distinct evidence, rather than assuming either can replace the other.

A relatively strong score does not invalidate tiredness, fogginess, symptoms, or feeling unrested. A lower score does not mean you must feel bad. The score cannot identify the reason for a difference, and disagreement does not automatically prove user error or device failure.

Read the score alongside how the night and day actually feel. Repeated differences can be useful context without becoming proof of a cause.

This is especially important when the record seems precise. A compact number can be consistent about the information it received while still leaving out parts of your experience. Treating the score and your own observations as separate inputs keeps both in proportion.

When a score looks unexpected

A quick check of the available record can show how much of the night was recorded. It cannot reconstruct a part of the night that was never recorded.

  • Check whether the watch was worn during the relevant sleep period.
  • Check charge and how much of the night was recorded.
  • Check that the relevant Health permissions are available.
  • Check whether the device and metric support the relevant recording.
  • Look at which score component changed.
  • Compare the result with how you actually feel.

Gaps remain unknown. A missing stage does not show that the stage did not occur, an absent awake record does not prove uninterrupted sleep, and a partial night does not become complete because a score is displayed.

Availability can change from night to night. The checks above are a way to understand what reached Underbark, not a way to infer what happened during a gap.

Where the score stops

Keep the number in proportion by asking practical questions: Was the record complete? Which component changed? Does the pattern repeat? How do I actually feel? Those questions can help you notice context without asking the score to make a clinical or safety judgment.

A pattern can prompt attention without supplying an explanation. The available records may help you decide what to observe over time, but they do not replace your judgment about symptoms, injury, illness, circumstances, or professional advice.

Persistent symptoms or concerns deserve advice from an appropriately qualified health professional. For help understanding how Underbark handles its available records, visit the Underbark support page.

Sources