Insights What Moves You

What Moves You — your own n-of-1 insights

Most apps tell you what a population does — "alcohol typically lowers recovery." NOOP works out the effect on your own data, on your own device. "What Moves You" ranks the things you log by how much they actually move your recovery, learns a personal dose-response for alcohol and late caffeine, and — in the evening — tells you what one more of something tends to cost you tomorrow.

Everything here is described as patterns in your own data — an association, never a cause, never advice.

---

What you need

• A bit of journal history. NOOP correlates the behaviours you log (alcohol, late caffeine, and the rest of your journal questions) against your daily outcomes (Charge, HRV, Rest, resting HR).
• The more nights you log, the sharper it gets — and it tells you honestly when it's still thin.

---

Where it lives

Open What Moves You — the wand-and-sparkles entry in the macOS sidebar, or reached through Insights on iPhone/Android.

---

1. "What moves your Charge"

The main feed. Each card is one behaviour you log, ranked by its real effect on the outcome you pick:

• A plain-English sentence with a with / without comparison ("recovery was 12% lower on days you logged alcohol — 61 vs 69").
• A lead/lag chip: "next morning" or "same day", because today's drink often shows up in tomorrow's recovery, not today's.
• A confidence pill — Solid / Building / Calibrating — so a behaviour you've only logged a few times reads honestly instead of shouting.

Use the outcome chooser at the top to switch between Charge / HRV / Rest / Resting HR.

If there isn't enough overlap between a behaviour and your outcomes yet, the card says so ("Calibrating — too thin to read yet") rather than inventing a number.

---

2. Personal dose-response (alcohol & late caffeine)

Binary "did you drink? yes/no" only goes so far. When you log alcohol or late caffeine, add an amount:

1. In the journal, on those questions, set the dose (e.g. 0 / 1 / 2 / 3+ drinks; for caffeine, how late you had it).
2. Open the Dose-response card in What Moves You. It plots your own points and fits a personal curve — "each extra drink lines up with about −N for you."
3. Until you've logged enough nights, the curve leans on typical patterns and says so ("based mostly on typical patterns — not yet yours"). Once you have your own nights, it becomes truly yours — and if your data contradicts the typical pattern (your drink-nights show no dip), NOOP says "in your data so far, this doesn't move your Charge." Your data always wins over the population once it exists.

Caffeine "dose" is a timing proxy (later = stronger), not milligrams — the card says so.

---

3. The evening "damage forecast"

In the evening, the Today screen shows a forecast card so you can decide before you act:

"A 2nd drink tonight tends to line up with about −7 on tomorrow's Charge for you (range −3 to −11). Based on 9 of your drink-nights."

There's a small stepper to preview 1 vs 2 vs 3, and the projected Charge band widens honestly when NOOP is less sure. In the morning, it shows the realised effect instead ("last night you logged 2 drinks; this morning's Charge came in at the low end of your range").

---

How to read it honestly

• Everything is an association, not a cause. The copy uses "tends to line up with," "in your data so far," "worth noticing" — never "causes" or "you should."
• Typical patterns are labelled as priors and are always overridden by your own data once you have enough.
• The damage forecast is a what-if on your own numbers — never a recommendation to drink or abstain, and never a medical warning.
• A wide band means low confidence. That's the honest signal, not a bug.

Privacy

All of this is computed on your device from data you already logged. Nothing is uploaded, and the priors NOOP shrinks toward are fixed, documented constants — never learned from other users.