Doom Saver

Attention

Why Willpower Loses to Infinite Scroll

You are not weak willed. You are up against an interface with no stopping cues, and willpower is the wrong tool for that specific problem.

Updated First published By

The short answer

Willpower fails against infinite scroll because it is designed to resist a decision, and scrolling does not involve one. You never choose to spend ninety minutes. You choose one more swipe, forty times, and each choice is genuinely trivial.

The missing piece is a stopping cue. Books, films and meals all end on their own. A feed removes every natural endpoint, so the only thing that can produce one is something outside the feed.

The stopping problem

Almost everything people consume has an end built into it. Chapters finish, episodes roll credits, plates empty. Those endings do more work than we notice, because each one is a moment where continuing becomes an active choice rather than the default.

Infinite scroll removes all of them. There is no last post, no natural pause, and no moment where the interface asks anything of you. Continuing is not a decision you make, it is what happens when you do nothing.

People notice the difference. In a 2021 study of the YouTube app, users said autoplay and recommendations undermined their sense of control over their viewing, while playlists and search supported it.

This reframes the problem. If you are waiting for a moment of resolve, you are waiting for a moment the design has deliberately removed.

Why each swipe wins

The individual unit is tiny. One more swipe costs about a second, and one second is trivially affordable, so refusing it feels absurd. You are never once presented with the choice you would actually refuse, which is whether to spend the next forty minutes this way.

Add unpredictable rewards. Most posts are dull, some are good, and you cannot tell which is next. Behaviour that is rewarded only some of the time tends to carry on longer once the rewards dry up than behaviour rewarded every time. This partial reinforcement effect is well documented in laboratory studies, though it has been harder to find outside the lab, and it helps explain why the boring posts do not put you off.

Then add the fact that you are usually tired. For many people scrolling happens in the evening, when energy is low and the appeal of something requiring nothing is at its highest. Be wary of the popular explanation, though, that self control is a fuel tank that runs dry. A 2016 replication across 23 labs and more than 2,000 people found that this ego depletion effect was close to zero.

Time perception collapses

The strangest part is how badly you judge duration. A 2021 meta-analysis in Nature Human Behaviour found that people's own estimates of their media use were rarely an accurate reflection of what their devices logged.

Our sense of elapsed time leans on landmarks: things that happened, transitions, changes of activity. In a classic 1978 experiment by Richard Block and Marjorie Reed, a stretch of time filled with different kinds of task was remembered as longer than one filled with a single kind. A feed is one activity repeated, so it produces almost no landmarks, and a long stretch can collapse into something that feels short. This is not carelessness. It is how remembered time works when nothing distinguishable occurs.

It explains a common experience: you are not fighting an urge and losing, you look up and the time has simply gone.

What works instead

If the problem is missing endings and broken time perception, the fix is not more resolve. It is to reintroduce what the interface removed.

  • An external clock. Since you cannot feel elapsed time in a feed, it has to come from outside it. Visible, and while you are still there, not afterwards.
  • Artificial endings. Any transition works as a stopping point: standing up, the end of a track, arriving somewhere. In a 2023 field study of infinite scrolling, most of the reasons people gave for ending a session came from what was going on around them, not from the app. Attach the scroll to something that finishes on its own.
  • Fewer entrances. Notifications and home screen placement decide how often the loop starts. Starting less often beats stopping earlier, because stopping is the hard part.
  • Lower the target. Aiming for zero makes one slip into a failure. Aiming to notice at twenty minutes instead of ninety is a target you can actually hit repeatedly.

The reframe worth keeping

Feeling weak willed about this is close to universal, and it is misplaced. These interfaces are refined against enormous numbers of people, and finding them hard to leave is the intended outcome rather than a personal shortcoming.

The useful conclusion is not to try harder. It is to stop using a tool suited to resisting decisions on a problem where no decision is ever presented, and to put the ending back in yourself.

Common questions

Why can I not stop scrolling even when I want to?

Infinite feeds remove every natural stopping point, so continuing is the default rather than a choice. Combined with unpredictable rewards and the way uniform activity distorts your sense of time, there is no moment where willpower gets a chance to act.

Is infinite scroll designed to be addictive?

Infinite scroll and pull to refresh are deliberate design features, and both make continuing the default. Whether the result meets a clinical definition of addiction is still debated.

Why does scrolling feel like less time than it was?

Your sense of duration relies on memorable landmarks, and a feed is largely uniform, so it generates very few. With nothing to mark the passage of time, a long session compresses in memory.

Does knowing all this make it easier to stop?

Understanding alone changes less than you would hope. A 2006 meta-analysis of 47 experiments found that a medium to large change in what people intended to do led to only a small to medium change in what they did. Knowing all this helps mainly by redirecting effort away from willpower and towards changing the situation, which is where the leverage actually is.

About the author

Samina Jameel. Samina builds Doom Saver. She works in product at IBM in Dublin, studied at IIM Tiruchirappalli, and makes small, opinionated apps on the side, including MyMir and SubLeak. She has lost more evenings to a feed than she would like to admit, which is why Grim exists. LinkedIn.

Doom Saver puts the timer where you can see it

Grim, a hooded knight, sits on top of the apps you choose, counting. He does not block anything, and he does not stay calm. Free on Google Play.

Get it on Google Play

Read next