The man who wrote the book on hooking you
later wrote the book on resisting it.

A compulsion loop is a repeating cycle: take an action, get a reward, feel a fresh pull toward doing it again. It's built directly on operant conditioning research, and the version of it running through your apps was engineered by people who studied exactly how to make that cycle hard to walk away from.

The term has an unusually well-documented history for something this widely used, a specific origin in game design, a bestselling business book that formalized it, and a public falling-out over whether the people who build it or the people who use it bear the responsibility for resisting it. All three parts matter.

John Hopson, a researcher at the game studio Bungie, introduced the compulsion loop to game design in a 2001 Gamasutra article on behavioral game design. His argument was direct: a set of conclusions from B.F. Skinner's operant conditioning research, developed on rats, pigeons, and chimpanzees, could be applied deliberately to condition player behavior in games.

The idea went quiet for a few years, then came back hard. Between 2010 and 2012, social gaming companies, Zynga chief among them, rebuilt the compulsion loop into the backbone of the free-to-play social game boom, structured around a simple cycle: take an action, get a reward that triggers a dopamine or serotonin response, then build fresh anticipation for the next one.

Nir Eyal formalized the same underlying idea into a widely taught business framework. He first taught it as a course at Stanford's Graduate School of Business in the fall of 2012, refined it the next year at the Hasso Plattner Institute of Design, and published it in 2014 as Hooked: How to Build Habit-Forming Products.

The Hook Model breaks the loop into four named stages: Trigger, Action, Variable Reward, and Investment. An external trigger, a notification, an email, gets things started. The user takes an action. They receive a variable reward. Then they invest something back into the product, a like, data, content, effort, which sets up the next trigger and makes the next cycle a little stickier than the last. Eyal's book became one of the most cited design frameworks in the industry, taught in product teams well beyond the startups it was originally aimed at.

In 2019, Eyal published Indistractable: How to Control Your Attention and Choose Your Life, reframing the fight against habit-forming products as something the individual needs to win through better internal skills, not something the product needs to stop doing. His position, stated plainly elsewhere: tech isn't the problem, being "distractable" is.

That framing drew real, specific pushback, not just general skepticism. Author James Williams, a leading voice on attention economics, argued the logic was flawed and that asking people to develop superpowers of self-control is an irresponsible substitute for changing the design itself. Other critics have been blunter, noting the parallel to tobacco industry messaging, which for decades emphasized personal responsibility and self-control specifically to keep the conversation away from the product's design. Tristan Harris, working the other side of this same debate, has described the broader dynamic as "a race to the bottom of the brain stem."

Eyal hasn't disavowed Hooked. He maintains that building habit-forming products isn't inherently wrong, and that the responsibility for managing the resulting pull sits with the user. Whether that's a reasonable division of responsibility, or a convenient one for whoever built the loop in the first place, is the actual disagreement underneath the two books.

Loot boxes are the sharpest real-world edge of this argument, because they've actually reached regulators, not just critics. Belgium became the first country to ban paid loot boxes, in 2018, with its Gaming Commission ruling that the mechanic qualifies as gambling under existing law.

The ban has been a mixed success at best. Research specifically documenting how players, including adolescents, found ways to circumvent the restriction shows that a legal ban on one specific implementation doesn't dismantle the underlying loop, it just pushes players toward workarounds. Separately, the World Health Organization added gaming disorder to the ICD-11 in 2019, a formal acknowledgment that compulsive engagement with these loops can rise to the level of a diagnosable condition, not just an aggressive design choice.

Strip away the branding and the compulsion loop is a repackaging of a much older, well-tested finding. Action, reward, anticipation is the same shape as variable ratio reinforcement, the reward schedule Skinner identified as producing the most persistent behavior of any pattern he tested. The Hook Model gave it a friendlier vocabulary, and product teams a repeatable process for building it into a product roadmap. Neither of those things changed the underlying mechanism.

The Indistractable critique lands on a real, practical question: is it reasonable to ask someone to out-resist a loop that a product team spent months tuning? Williams's objection wasn't abstract. Demanding superpowers of self-control from users, while the trigger keeps firing exactly as designed, puts the entire burden on the side of the equation with the least engineering behind it.

Fella takes the other side of that argument. It doesn't ask you to win the moment the trigger fires. It removes the trigger's availability by default, keeping the app blocked so the action-reward-anticipation cycle never gets a chance to start on its own. The one 5-minute unlock a day is a deliberate, scheduled exception, not an open invitation to step back into a loop built by people who study exactly how to keep it running.