The gap closes ——

Two lines, one of them faster

The upper line is how many episodes exist. It goes up by one a week and nothing you do changes it. The lower line is how many you have seen. Where they touch is the whole question, and after that they travel together.

—

—

The first extra episode is worth all the others

Every row keeps your backlog and the run length where you left them and changes only the pace. The right-hand column is what one more episode a week would buy you from that row: it is at its biggest in the very first row and it shrinks fast, because what you divide by is the gain and not the pace.

—

—

A week off is not a week off

Skip a week and you have not lost a week. While you were away the show put out another episode, so you come back owing more than you left owing, and that extra episode has to be cleared on top of everything else. Every row below is the same finish line, counted from today.

—

—

Catching up is the short part

The chase feels like the whole job because it is the part that takes effort. It is not the part that takes time. Drawn below is everything between now and the last episode, split at the week you draw level.

—

—

What each stretch costs in evenings

The same journey again, priced. An episode is twenty-four minutes of programme, so the hours column is just the episode column in another currency — but it is the currency you have to find, and the two stretches ask for it at very different rates.

—

—

Where these numbers come from

1. One episode a week is the assumption underneath everything. Broadcast anime runs to a weekly slot, so a show that is still on the air gains exactly one episode in the time you gain however many you watch. Every figure here is that subtraction. A show releasing a batch at a time is a different problem and this page will get it wrong.
2. Net gain is your pace minus one, and that is the whole model. Watch four a week against a show putting out one and you close the gap by three, not four. The week you draw level is the backlog divided by that net gain, rounded up to a whole week, because you cannot draw level in the middle of a week you have not finished.
3. One a week never draws level, and that is not an edge case. Net gain is zero, so the distance between you and the newest episode is frozen at whatever it is today, for as long as the show keeps airing. Most people watching a long-running show are in exactly this state and do not think of it as being behind.
4. Three different things can happen and the page says which. You catch the show while it is still airing; or the show ends before you reach it and you finish the rest at your own pace with nothing chasing you; or it has already finished, in which case the backlog is simply divided by your pace. The reading above changes wording for each.
5. Twenty-four minutes is the programme, not the slot. A half-hour television slot in Japan carries about twenty-four minutes of episode; the rest is advertising you will never see on a streaming service. Roughly three of those minutes are the opening and the credits, so a twelve-episode season is close to forty minutes of the same song. Nothing on this page subtracts them.
6. The cost of starting later is counted from today, not from then. Waiting a week means the finish line moves by that week plus however long it takes to clear the episode that aired while you waited. At a low net gain the second part can be as long as the first, which is why the column often reads two weeks for one week off.
7. Nothing here knows which show you mean. There is no title, no schedule and no list being read — three numbers go in and the arithmetic comes out. If the show is on a break, airing twice a week, or counts its recap episodes, you are the one who has to put that in the episode count.
8. Whole weeks throughout, and always rounded against you. Every week figure is rounded up. A chase that finishes on a Tuesday is reported as the week it finished in, because a week is the unit the show itself is using and half a week of being level is not a thing anybody experiences.