Why a balance score, and not another checklist.

The loop you're actually fighting

Your reward system runs on a baseline. Something better than expected nudges it up, something worse nudges it down, and over time the whole scale re-centres on whatever you feed it most. This is ordinary, useful biology: it is how anything gets learned.

The trouble is that a phone can deliver a reward faster, more reliably and more often than anything in the environment that system evolved in. Not a stronger reward — a more frequent one, available in under a second, with no effort in between. Repeat that for long enough and the baseline re-centres around it. Everything slower then feels flat by comparison: reading, conversation, work that takes an afternoon before it pays anything back.

Which is also why "just try harder" is aimed at the wrong thing. Willpower is a conscious, expensive, exhaustible resource, and it is being asked to out-argue an automatic process that never gets tired. It wins some days. It cannot win every day, and a system that needs you to win every day isn't a system.

What the balance score actually measures

If the baseline is the thing that moved, then the useful question each day is simple: did today drain that system or restore it? That's all the balance score is — a running tally of the things you logged, weighted by which direction they push.

• Restoring inputs — completed habits, focus sessions, sleep, exercise, time spent on something that took effort before it paid.
• Draining inputs — logged slips, urges acted on, and the categories of screen use you told the app you're trying to reduce.
• Context — the day of the week, the hour, and what preceded it, used to find patterns rather than to change the score.

Three deliberate constraints on that number, all of which cost us features and all of which we'd make again:

It is a fixed rule set, not a model. The weights are written down. The same inputs always produce the same score. Nothing is learned from you in the background and nothing drifts between releases without a changelog. Machine learning would produce a smoother number that no one — including us — could explain, and an unexplainable number is not something you can act on.

Every point is traceable. Tap any day and the app lists what earned it, item by item. If you disagree with the score, you can see precisely which entry you disagree with and correct it.

It is not a grade. There's no target, no streak attached to it, and no notification when it drops. It's an instrument reading. Instruments that shout at you get taped over.

Why the streak works the way it does

The standard streak mechanic is loss aversion, deployed on purpose. It works — for a while — because the fear of breaking a chain is a stronger motivator than the appeal of extending one. Then you break it, and the same mechanic that was holding you there becomes the reason to quit entirely. Most people don't restart at day zero. They just stop opening the app.

That is precisely the dynamic NeuroPath exists to help you escape, so reproducing it inside the app would be incoherent. Instead:

• Your personal best is never erased. A slip resets the current streak and touches nothing else. Fifty-two days happened; deleting the record wouldn't make it un-happen.
• Nothing counts down at you. No warnings about what you're about to lose, no colour designed to sting, no notification the evening you miss one.
• No loss-framed language anywhere. Not "failed", not "broken". You logged an event; the app recorded it.
• Negative events are never inferred. The app will offer to record a slip if a signal suggests one. It will never write one on your behalf. Being wrongly told you relapsed is worse than missing a log.

Patterns, found by rules you can read

Log an urge or a trigger and you get two taps of work and a timestamp. Do that for a few weeks and the shape shows up on its own: the hours they cluster in, the days they follow, what tends to precede them.

The detection behind that is fixed lookup and aggregation logic — counting occurrences in time buckets and surfacing the ones that stand out. It is deliberately simple, because a pattern you can verify against your own memory is worth more than a confident one you can't. When the app says urges cluster after short sleep, you can check that claim against your own week. That's the point.

What we don't claim

Everything above describes how the app is built and why. None of it is a clinical claim, and we'd rather say so here than let the omission imply one.

• NeuroPath is not a medical device and is not a substitute for therapy, diagnosis or crisis support.
• We publish no outcome statistics. No success rates, no percentages, no "clinically proven". We haven't run the trials that would justify any of those, so we don't cite them.
• The Academy is educational, not prescriptive. It explains established ideas about habit formation and reward in plain language. It does not diagnose you and does not tell you what your results will be.
• Tracking is not treatment. Seeing a pattern clearly helps. For a genuine addiction it is a companion to real support, not a replacement for it.

If you're in crisis, please contact your local emergency services or a crisis line. That is not a disclaimer we're required to print — it's the honest answer to a question this app cannot help with.