Measured · No. 5 · 2026-10-30
The One Percent That Holds a Quarter
1,479 segments over twenty minutes are 1.1% of the record and hold 23.6% of all story airtime, while the fifth of segments under a minute holds 5%.
Count the segments and television looks like a stream of short hits. Count the minutes and it looks like something else entirely. Both counts are in the record, and they disagree by a factor of five.
There are 134,977 story segments in 99 days of the collection, holding 4,822 hours of story airtime. Sorted by length, the airtime piles up at the long end:
| Segment length | Segments | Share of segments | Airtime | Share of airtime |
|---|---|---|---|---|
| over 20 min | 1,479 | 1.1% | 1,138 h | 23.6% |
| 5–10 min | 10,491 | 7.8% | 1,206 h | 25.0% |
| 10–20 min | 3,675 | 2.7% | 827 h | 17.1% |
| 3–5 min | 11,124 | 8.2% | 726 h | 15.1% |
| 1–3 min | 21,523 | 15.9% | 684 h | 14.2% |
| under 1 min | 28,685 | 21.3% | 242 h | 5.0% |
One percent of the segments — 1,479 of them, the ones that run past twenty minutes — hold nearly a quarter of all the airtime. Meanwhile the fifth of the record made of sub-minute segments holds 5% of it. A channel can run thirty short hits while a single long block quietly eats the hour.
This is the arithmetic that makes coverage claims slippery. "Fox covered it for nine minutes" and "Fox mentioned it eleven times" describe the same nine minutes and mean different things; without the segment-length distribution, both sides of that argument are quoting the metric that favours them.
It also explains why absence is so common (No. 3): the twenty-minute blocks are scarce, and every minute one of them holds is a minute not spent on something else. Airtime is not a uniform resource — it comes in large, indivisible chunks, and the chunks are where the agenda lives.
What this does not say. It does not say long segments are bad, or that short ones are filler. A twenty-minute interview can be the most informative thing on air; a forty-five-second hit can be the only coverage a story gets. The piece is about the arithmetic of the schedule, not its quality.
Method. Source: stories where kind='story', window 2026-07-15 → 2026-10-23. Segment length = ts_end - ts_start; buckets are the boundaries shown. Airtime share = bucket hours over total 4,822.5 h. Figures pinned in data/measured/facts.json (shape_of_day.airtime_share); rerun with python -m parrot.measured.
How to check every number. Figures are generated by
python -m parrot.measured into data/measured/facts.json, each one
carrying the query that produced it. The recording window closed on 2026-10-23; the desk does
not add to it. Corrections are published against the facts file, not edited into the prose.