Contents

  1. Core principles
  2. Which signals Arry uses
  3. Why personal baseline matters
  4. What the system does not claim
  5. How this connects to privacy

Core principles

Recovery Score in Arry is not a medical conclusion and does not try to “predict the body” from a single number. The goal is narrower and more practical: help you judge whether today is a reasonable day to push, hold steady, or recover, and show which factors influenced that decision most.

  • we use multiple signals at once instead of one metric;
  • we compare you primarily against your own baseline;
  • we avoid absolute claims such as “fully ready” or “not ready at all”;
  • we show context, not only an output score.

Which signals Arry uses

The exact data available depends on what your device writes into Apple Health and which permissions you grant. The core estimate is built from signals such as:

HRV
Heart rate variability as one marker of nervous-system strain and adaptation.
Sleep
Sleep duration, quality, regularity, and stages when available.
RHR
Resting heart rate as a practical signal of accumulated fatigue and general state.
Load
Recent training and daily load so the last few days are not ignored.

Depending on the scenario, these layers can be supplemented by activity, stress context, and workout history. The point is not to guess perfectly, but to gather enough evidence for a careful daily recommendation.

Why personal baseline matters

Two different people can have very different “normal” HRV, resting heart rate, and responses to the same amount of sleep. That is why Arry tries to interpret your signals against your own range and how it changes over time, rather than against a single idealized standard.

A more useful question is not “Is my HRV normal today?” but “Has my HRV moved away from my own baseline, and do other signals agree with that move?”

This is also why the first few days after connecting HealthKit matter: the system needs an initial frame for personalization and then keeps updating context as data accumulates.

What to keep in mind

  • the same HRV value can mean different things for different people;
  • one poor night does not always equal a red day if the rest of the picture is stable;
  • a strong night does not erase accumulated fatigue from a hard block;
  • decisions are stronger when several signals point in the same direction.

What the system does not claim

Arry is not a medical device, does not diagnose illness, and does not replace a clinician. Recovery Score and other in-app recommendations are support tools for day-to-day load management and self-observation.

  • we do not claim to predict performance perfectly from one score;
  • we do not claim that a low score automatically means a health problem;
  • we do not treat any single system as a full replacement for subjective state and judgment.

How this connects to privacy

The Arry methodology is designed to use only the signals required for the feature and keep core processing as close to the device as possible. That matters not only for privacy, but also for trust: people should understand which signals influence the score and why.

More on data handling is on the privacy page. More on who builds Arry and how the site materials are prepared is on the about page.