"Turning autoplay off is fiddly"

The method's first step is a setting, and the objection to it is that the setting is buried and the menu is annoying and you'll do it at the weekend.

It takes two minutes per service. Settings, playback, autoplay next episode, off. On some devices it's in the account settings online rather than on the television. Two minutes, once, and the countdown is gone from every episode you'll ever watch on that service, and the credits come back, and the door swings your way. Chapter 20 has the general shape.

The fiddliness is worth a second look, though, because it isn't an accident. Autoplay is on by default, and turning it off is a few menus deep, for the reason Johnson and Goldstein's paper describes. The side that requires an act is the side fewer people end up on. If autoplay were off by default and you had to dig through menus to turn it on, most people would never see a countdown, and would watch less, and the services know that, and the setting is where it is. The two minutes it takes are two minutes the design has made as unlikely as it can.

That's the argument for doing it now, tonight, before the next episode rather than at the weekend. The weekend is where settings go to not get changed. It's two minutes, and after them the whole rest of the method gets easier, because the stop is the default again and you're no longer pressing a button against a timer every night.

If there are several services and several devices, do the one you watch most on first. The rest can follow when you next use them. And if a service won't let you turn it off, which some don't on some devices, then chapter 21's number and chapter 22's end time do the work instead, and the remote lives on the far side of the room, so that the act of continuing costs at least a stand.