Beyond the guide · a refinement

When the strain isn't physical: what the watch catches, and what it misses

Wiggers is honest that a step-counter misses the non-physical draws on your energy: “too much mental activity… often goes undetected in your Garmin,” though she expects its autonomic aftermath to surface as an HRV drop “that night or the following night.” She lumps it together as “mental activity.” Sorting my own days by what the demand actually was (physical, cognitive, or emotional) splits that lump apart. The kinds of load are read from the day's own notes, never from the watch, so the mental ones are a real cross-check, not the watch grading itself.


a refinement of the guide · least-settled layer

What it adds to the guide

The guide says the watch can't see non-physical load, but its aftermath may surface overnight. My data sharpens that: the aftermath shows up for emotional load (in stress and body-battery, not in heart rate, peaking the following night just as she times it) and is absent for cognitive load, which stays invisible even overnight. The watch sees your emotional load, just not in your heart rate; it doesn't see your cognitive load at all.

An exploratory extension: maybe two kinds of crash

Pushing further (and this is a tendency, not a finding), the crashes themselves seem to sort into two shapes: quiet ones (body battery not depleted, heart rate not raised) whose run-up was emotional or interpersonal, and loud ones (the stress-and-heart-rate storm) that lean toward illness. Both are equally real crashes, the same depth and duration; they differ only in whether the watch registers them. The quiet ones are the crashes the watch can't see.

The honest floor: requiring the trigger, the watch, and the notes to all agree labels only 5 of the 29 crashes cleanly; the other 24 stay unclear. So it's a real tendency, not a classifier: “crashes seem to come in a quiet type and a loud type,” never “we can tell you what set yours off.” And whether a virus set one off simply can't be read from this record (my own symptoms can't tell a real virus from severe PEM), so that one waits on a forward-looking fix: noting when others in the house are ill.

How to hold this

  • Newer, less settled than anything in the field guide: descriptive, n=1, self-reported load, wide confidence intervals shown not hidden.
  • What the watch sees, not what triggers a crash. Whether emotional load also helps cause crashes is a separate, still-suggestive lead held back for a proper pre-registered test, deliberately not claimed here.
  • Non-circular for the mental kinds. Emotional and cognitive load are tagged from the notes, not the watch, so their agreement with the channels is a real cross-check; physical-vs-activity is partly circular and the least novel of the three.
  • Descriptive, no cause. “Emotional load co-occurred with an overnight dip in my body,” never “it caused it.”

The full assembly, in the research repo: the autonomic fingerprints of load (R4) ↗.

↑ Back to beyond the guide

Open full page ↗