Twelve Is
A Calendar

The four shapes almost everything takes

Everything else on the list is a leftover: a six-episode short-form series, a seven-episode finale block, a series that lost a week to a sports broadcast. The four shapes above account for nearly all of it.

A cour is three months of Tuesdays

Japanese television is scheduled in quarters, and the industry word for one of them is a cour — usually explained as a borrowing from the French cours, a course of study. January to March, April to June, July to September, October to December. Each block is about thirteen weeks long, and a series that takes one weekly slot for one block gets thirteen airings.

Thirteen airings is not thirteen episodes. A national sports broadcast, a new year’s schedule, an election night or a special programme eats a week, and the series comes back the following Tuesday one episode short. That is where twelve comes from. Twelve and thirteen are the same decision, made in the same meeting, and the difference between them is usually something that happened to the calendar rather than to the story.

Once you know that, the shape of any season index stops looking like a list of creative choices and starts looking like what it is: a grid of slots, filled.

Why the slot decides the length

Four separate pressures point at the same number, which is why it has held for so long. None of them are about how long a story needs to be.

1. The slot is bought by the quarter. Late-night airtime is purchased in blocks that line up with the broadcast quarter. Booking two-thirds of a block is not cheaper than booking one and it leaves the station with a hole, so nobody does it. The unit of purchase becomes the unit of storytelling.
2. The money is raised one cour at a time. A production committee — a group of publishers, distributors, record labels and toy companies pooling the budget — commits to a defined number of episodes with a defined cost. Asking that group for a second cour before the first has aired is asking them to double an exposure with no new information. So the first order is one cour, and the second one waits.
3. The discs are packaged to fit. The home release schedule was built around a cour: a fixed number of volumes, released monthly, carrying two or three episodes each. Twelve divides cleanly into that and nineteen does not. Physical sales matter far less than they did, but the shape they imposed on the schedule stayed.
4. The source material has to last. Most series adapt a manga or a light novel that is still being written. One cour consumes a manageable slice of it. Two cours can catch up with the author, which is how a series ends up inventing an ending nobody wanted — the risk every adaptation committee is trying not to take.

“Season 2” means four different things

The phrase is borrowed from American television, where it has one meaning. Here it covers four arrangements that are made at different times, by different people, with completely different odds of arriving.

1. A split cour. Twenty-four episodes commissioned at once, aired as twelve, a season off, twelve more. The gap is a scheduling device, not a renewal — the whole thing was ordered together and much of it was finished before the first half went out. This is the one that actually arrives on time, because it was never in doubt.
2. A continuous two-cour run. Twenty-six weeks in the same slot with no break. Expensive, increasingly rare, and a strong signal about what the committee expects: nobody books half a year of airtime for a title they are unsure of.
3. A sequel ordered after the fact. The first cour aired, the discs and the streaming numbers came in, and the committee met again. Two to three years is normal, because the staff that made the first one have been booked onto other shows in the meantime. This is where most of the waiting happens.
4. A new adaptation of the same property. Different studio, different staff, sometimes a different continuity, marketed as a continuation. The episode count carries over from nothing.

An index that prints “Season 2” next to a title is telling you about the marketing, not about the production. The episode count is often the better clue: twenty-four in one block usually means it was all commissioned together.

Every count you will see on the list

Read down this and the airing list becomes legible. Almost every number in it is one of these, and each one is the fingerprint of a decision made before production started.

1. 1A film, an OVA or a special — not a season at all
2. 3–6Short-form: a three-minute slot, or a web release — minutes, not episodes, is the honest unit
3. 10–11One cour that lost weeks, or a deliberately tight adaptation
4. 12The default. One cour, one preemption — the single most common number in the medium
5. 13One cour, uninterrupted
6. 22–23Two cours that each lost a week
7. 24Two cours, usually split — commissioned as one order
8. 25–26Two cours, continuous, uninterrupted
9. 38–39Three cours — unusual, and almost always a title with a fixed ending to reach
10. 50–52A full broadcast year — now mostly children’s and family slots
11. 100+A long-runner that has stopped counting seasons and simply airs
12. BlankStill running, or the count was never published — left blank here rather than guessed

Why the airing list measures hours, not titles

This is the practical consequence, and it is the reason the season page on this site refuses to answer “how many shows are there” as if that were the question. A hundred and forty titles is not a workload. A hundred and forty titles where most are twelve episodes of about twenty-four minutes, several are films, and a handful are three-minute shorts is a workload — and the two numbers are not close to each other.

1. A title is not a unit of time. One film and one two-cour series both count as one entry. One of them is ninety minutes and the other is ten hours. Any count of titles treats those as equal, which is why the season page converts to hours before it says anything.
2. The slot is longer than the episode. A half-hour slot carries roughly twenty-three or twenty-four minutes of programme; the rest is advertising and the two songs. Multiply by twelve and the difference between the slot and the show is over an hour per series.
3. A blank count is not a zero. Series still airing, and series whose final count was never announced, arrive with the field empty. Filling it with an average would make the season’s total look tidier and would be an invented number. It is left blank, and the blanks are counted separately.

None of this makes the numbers on the index wrong. It makes them smaller than they look. A season is not a hundred and forty decisions about how long a story should be; it is a hundred and forty slots of a standard size, and the interesting question is which ones got a second block.