Measured comfort, never measured judgment
2026-08-26
Today the app learned to say "running like a usual night" during a night waking — and the rule I had to write to allow it is more interesting than the feature. The claim is backed by the child's own record: the median wakings per night over her recent evidenced nights, at least five of them, tonight excluded from its own baseline. But the comparison is only allowed to speak in one direction. At or under the usual, the copy says so, with the data behind it. Above the usual, the app does not say "more than most nights" — it falls back to the plain reassurance it already had. The same measurement, deliberately half-used.
The symmetric version is what the industry ships, and it is worth naming as the adversary here: trackers love a deviation. A red night, a below-average score, a "worse than last week" arrow — the data is symmetric, so the reporting is symmetric, and the 3 a.m. parent gets graded by their own logs. But the two directions of this comparison are not worth the same. "Tonight is inside the usual" dissolves a fear that was actively forming; "tonight is unusual" confirms one, attaches a number to it, and offers nothing to do about it at 3 a.m. Symmetric data, asymmetric value. The honest design follows the value, not the symmetry.
Is the one-sided version still honest? I think yes, and the distinction matters: the app never says something false, it chooses which true thing is worth saying. Every sentence it prints is backed; not every backed sentence gets printed. That is the same editorial stance every calm bedside doctor already takes, and nobody calls it lying.
Two numbers guard the claim so the comfort itself stays true. Five evidenced nights minimum before "usual" may be uttered at all — thin history stays silent rather than wrong. And a child whose usual night has zero wakings never gets told a waking is usual for her, because for her it isn't. The admission: the waking counts themselves are only as good as our hand-logging at 3 a.m., and a night where we slept through the logging reads as a calmer night than it was. The gate means one sloppy week degrades the feature to silence, not to a false claim. Silence is the failure mode I chose on purpose.
The counts are never displayed, in either direction. A number on a night is a score, and a score on a night is a grade on a parent, whatever the label says. The app knows the count; the parent hears only "this is like most nights." Nothing is wrong.