Skip to content
Deterministic scoring — no model guessing at your mood

See which way you are heading — while it still matters.

Mood trackers tell you where you have been. NoeticCog turns thirty seconds of daily self-report into four scores, then runs the same arithmetic forward to show you where today’s pattern ends up if you hold it.

  • No card required
  • Export or delete everything
  • Not a medical device

Try the model

The real scoring engine, running here in your browser.

Live
2 / 5
4 / 5
6 h
2 / 5
43
Strained
Stability today

If every day looked like this

Next 60 days

050100Critical
Burnout51Spiral10
Burnout risk settles at
51High
Spiral risk settles at
10Low

Held for a month, risk levels off around 51 — high, but short of critical.

Move a slider. Those numbers come from the production engine.


The premise

By the time a bad stretch is obvious, you have usually been in it for weeks. The signals were there the whole time — they were just spread across days you never compared.

The point of measuring daily is not the diary. It is that a trend becomes visible long before a crisis does.

How it works

Three steps, once a day.

01

Answer five questions

Energy, stress, sleep, social contact, and whether you were stuck in your own head. Thirty seconds. You can correct today’s entry any time before midnight.

02

The engine scores the day

Four numbers, computed in the database by fixed arithmetic. The same inputs always produce the same outputs — there is no model deciding how you are doing.

03

It projects that forward

Hold today’s pattern and the engine tells you where the accumulating risks settle, and how many days until they cross the line. That is the part a diary cannot do.


The model

Four numbers that behave differently on purpose.

One measures today. One measures change. Two accumulate. Collapsing them into a single “wellness score” would throw away the only thing that makes them useful.

Stability

Higher is better

How much capacity today actually had.

Stateless — it depends only on today, so a good day counts as a good day no matter what preceded it.

Volatility

Lower is steadier

How far today sits from yesterday.

The distance between consecutive days. Two people can average the same score while one is level and the other is swinging.

Burnout risk

Lower is better

Strain you have not paid back yet.

Accumulates. Strain rises quadratically when stress outruns energy; recovery pays it down linearly; the balance decays about 2% a day.

Spiral risk

Lower is better

The self-reinforcing pattern: ruminating, withdrawing, sleeping badly.

Carries roughly 60% of itself into the next day, so it is slow to build and slow to clear.

Every score is explainable. Open any day and the app shows the terms that produced it — which input contributed what, and how much yesterday carried over. Nothing is a black box, because nothing is a model.


Self-experiments

Stop guessing whether it helped.

“I think the earlier bedtime is working” is a hypothesis, and you already have the data to test it. NoeticCog runs a proper n-of-1 trial: you commit to the change and the metric before you start, it takes a baseline window, and then it compares.

  • Pre-registered

    You choose the lever and the primary metric up front, so you cannot go hunting for whichever number moved.

  • Effect size, not vibes

    It reports the change and Cohen’s d against your own baseline variance — and refuses a verdict when the window is too short or the shift too small to act on.

  • Honest about limits

    It deliberately does not print a p-value. Daily self-reports are autocorrelated, which breaks the independence those tests assume; quoting one would look rigorous and mean nothing.

Lights out by 23:00

14-day trial · primary metric: stability

Baseline
54
14 days before
Trial window
67
14 days on
Change+13 points
Effect sized = 1.12

Large effect, and the window was long enough to trust it. Worth keeping.

Illustrative figures. Your result is computed from your own history.


Privacy and safety

This is the most sensitive data you own.

So the guarantees are structural rather than promised. Access is enforced by the database itself, not by application code remembering to check.

Row-level security everywhere

Every table carries a policy tying rows to their owner. A query that forgot to filter returns nothing rather than someone else’s day.

Export whenever you want

Everything you have entered, on every plan, in a format you can actually read.

Delete means delete

One click removes your history. Not a flag on a row that stays behind.

Clinician access is consented and logged

Nothing is shared until you invite someone and they accept. Every time they open your timeline it is written to a log you can read, and you can revoke access at any time.

Trusted contacts confirm first

Someone you name is only alerted after they confirm by email, and they can unsubscribe themselves. Nobody gets messages about you without agreeing to.

Crisis resources are never gated

They are available on every plan, including free, and they do not sit behind a paywall or a sign-up.

Works without a connection

Check in on a plane. The entry is queued on the device and sent when you are back — and sending it twice is harmless.

Medication, tracked alongside

Log doses and adherence next to the scores, so a change in regimen and a change in trend can be looked at together.


Plans

The daily instrument is free, permanently.

Paid tiers add the parts that look forward or look for causes. Nothing that helps you in a bad moment is ever behind them.

Free

The whole instrument, for one person.

Start free
  • Daily check-in and the four scores
  • Score explainer — why today came out that way
  • Medication and dose tracking
  • Note search across your history
  • Reminders
  • Trusted contacts
  • Import and full export
  • Forecast and equilibrium (not included)
  • Pattern insights (not included)
  • Self-experiments (not included)

Pro

Most complete

Everything that looks forward or looks for causes.

Start free, upgrade later
  • Everything in Free
  • Forecast, equilibrium and days-to-critical
  • Self-experiments with effect sizes
  • Medication impact analysis
  • Pattern insights and similar days
  • Wearable sync
  • Annotations and cohort comparison
  • Share a read-only timeline with a clinician
  • AI theme summaries of your notes

Clinic

For practitioners carrying a caseload.

Get in touch
  • Everything a patient needs, per patient
  • Clinician portal with a patient roster
  • Invite-and-accept consent for every relationship
  • Every access written to an audit log
  • Patients can revoke access at any time

Questions

The ones worth asking first.

Is this therapy, or a diagnosis?

No. NoeticCog is not a medical device. It does not diagnose or treat anything, and it is not a substitute for professional care. It measures patterns you report about yourself and does arithmetic on them. If you are in danger, contact your local emergency services.

Is an AI deciding how I am doing?

No. Every score, the forecast and the experiment statistics are fixed arithmetic running in the database — same inputs, same outputs, every time, and you can read the formula for any day on the score explainer. The only optional AI feature summarises themes in your own written notes, and it never touches a score.

What happens if I miss a day?

Nothing breaks. The engine is time-aware: it knows how many days elapsed and decays the accumulating risks accordingly, rather than pretending the gap did not happen. You can also backfill an earlier date.

Who can see my data?

You, and nobody else by default. Every table is protected by row-level security, so access is enforced by the database rather than by application code remembering to check. Sharing with a clinician requires you to invite them and them to accept, and you can revoke it at any time — every access they make is logged for you to read.

Can I get my data out, or delete it?

Yes, on every plan. Full export of everything you have entered, and a one-click delete that actually removes it.

Does it work offline?

Yes. A check-in made without a connection is queued on the device and submitted when you are back. Submitting the same day twice is safe — the day is upserted, not duplicated.

Your first reading is thirty seconds away.

The trend needs about two weeks of days to say anything useful. The sooner you start, the sooner it can.

Start free