Wearable metrics cheat sheet

One page on what each number is, how much weight it deserves, and what to do when it moves. Written for the coach reading someone else's data, not the athlete reading their own.

The metrics

Confidence is how much a single reading is worth on its own. Everything here should be read against the athlete's own baseline rather than any population range.

MetricWhat it reflectsConfidence in one readingWhen it moves
Sleep durationTime actually asleepHigh — hard to measure wrongThe most actionable metric there is. Ask about it before anything else.
Sleep consistencyRegularity of sleep and wake timesHighSlow to change and worth coaching directly. Often explains a noisy HRV series.
HRVAutonomic state overnightLow — very noisy day to dayOnly act on a multi-day trend. One low morning is usually nothing.
Resting heart rateBaseline cardiovascular stateMedium-high — more stable than HRVA sustained rise across days is one of the more reliable signals available.
Readiness / recovery scoreA vendor composite of the aboveMedium — but it hides which input movedTreat as a prompt to look underneath, never as a verdict.
Steps / daily activityNon-training movementHighUseful context. A collapse in general activity often precedes an athlete telling you anything.
Self-reported sorenessPerceived tissue stateMediumRead against the session that caused it. Soreness with no session to explain it is different.
Self-reported motivationPsychological stateMediumA slow drift down across a block is one of the earliest warnings a coach gets.

The short version

If you remember nothing else from this page.

Copy and paste — no signup

1. Compare an athlete to THEMSELVES, never to the squad.

2. Trend beats any single reading. Especially for HRV.

3. Sleep duration is the most actionable number you have.

4. Resting HR rising over several days is worth a conversation.

5. A composite score tells you SOMETHING moved,
   not WHICH thing moved. Look underneath.

6. The athlete knows why. The sensor does not.

7. Not enough data is a valid answer.
   Say so instead of assuming they are fine.

8. None of this is diagnostic. Anything that looks
   like a health concern goes to a medical professional.

The limits of all of it

These are training-context signals produced by consumer hardware. They do not detect illness, injury or any medical condition, and they should not be used to overrule how an athlete says they feel or what a clinician has told them. Their value is in prompting the right conversation earlier.

Frequently asked

Which single metric should I watch if I only have time for one?
Sleep duration. It is the most reliably measured, the most actionable, and it explains a large share of movement in everything else on the list.
How big a change is worth acting on?
There is no universal threshold and any number given would be invented. Significance depends on how variable that athlete normally is — a five percent move matters for someone very stable and is noise for someone who swings widely.
Can I use this with athletes who have no wearable?
The bottom half applies unchanged. Self-reported soreness, motivation, energy and mood support a baseline of their own without any hardware.

Have this applied automatically

Loadwell reads these signals against each athlete's own baseline and tells you which ones moved, so you are not doing the comparison in your head.

14-day free trial. No card required to start.

Keep reading