The Law of the Times: Visualisation · Signals from the Periphery

THE LAW OF THE TIMES
"One close approach in fourteen is seen." Priced.
Sighting clocks: 1975 type-I (n≈2,000) · NUFORC US (64,627)
both peak at 21h, fifty years apart · the law's evening half replicates
awake × dark reproduces it: r = 0.88, nothing fitted
the ×14 swings ×5-×19 with the divisor's last two percentage points
switch the file · choose the divisor · drag the night floor

Evidence

fact ⇄ what kind of claim · live

The law's wording and the ×14 claim, verbatim from AIAA 75-42 (1975)

Poher & Vallée · the phrase "law of the times" is the literature's, not the paper's

REPORTED

The 1975 hourly curve: digitized trace sums 2,090 vs stated ~2,000

Figure 8, hand-drawn · all four landmarks reproduce · night cells at pen resolution

REPORTED

The modern clock: 64,627 US reports, peak 21h (14.9% of the day)

NUFORC public file · recomputed twice for this bench · no 3 a.m. bump

REPORTED

The awake fraction, hour by hour: 4.9% at 3 a.m., 79.0% at 9 p.m.

BLS American Time Use Survey, Table A-3 (2023) · the checkable divisor

MEASURED

The darkness term: sun below -6°, fraction of the year, latitude on a dial

pure geometry, zero constants · solar vs civil time is a stated ±1h liberty

MODELLED

The correction Nc = N ÷ P, run live for all four dataset × divisor pairs

their arithmetic exactly · Poisson whiskers added · checked by law-of-times-tune.mjs

MODELLED

The night-floor envelope: ×12.7 at 3.0% to ×18.7 at 1.6%

the divisor's own error bar, wider than the sampling error · the finding

MODELLED

July 4th: 1,206 reports vs a 192 median day; New Year's repeats it

same NUFORC file · 1st/15th rounding artifacts disclosed and excluded

REPORTED

…

…