Home / Concept of the Week / Variable ratio reinforcement
Concept of the Week · Episode 6
A schedule of reinforcement in which a behaviour is rewarded after an unpredictable number of responses, so the next reward could always be one try away. It produces the most persistent behaviour psychology has managed to make: it is how a fruit machine works, how you keep a fox coming to your garden, and how a social feed is designed.
Explained by Professor Peter Kinderman, clinical psychologist · First published 25 August 2026 · Updated 21 October 2026
B.F. Skinner's work on schedules of reinforcement in the 1950s. Reward every time and the reward stops registering; stop rewarding and the behaviour dies out; reward at random and it keeps going almost indefinitely.
Variable ratio reinforcement. Reward a behaviour every single time and the reward stops registering. Stop rewarding it and the behaviour dies out. Reward it unpredictably, on average about one time in three, and you get the most persistent behaviour psychology has managed to produce. It is how you keep a fox coming to your garden, how a fruit machine works, and, Peter argues, how a feed is designed. The uncomfortable part is that knowing all this confers no protection whatsoever.
Twenty nine US states took Meta to court in Oakland over exactly this mechanism, and Peter opened episode six by explaining it before the news had even started. The same schedule turned up twice more in the episode: in why the Met Office lets the public name storms, and in why the two day weekend took a century to win. Peter is careful to say that his own screen time is not great, which is why the Triple T that week was set limits, and set them yourself rather than letting the app decide.
Hear it in context
Meta on trial, named storms and the weekend at 100
The psychology that keeps a fox in your garden is the psychology on trial in Oakland, with named storms and a hundred years of the weekend in between.
A fruit machine is the classic one. Peter's example on the show is an arcade machine that pays out about one time in three, but never on a count you can predict. Scratch cards work the same way, as does a salesperson on commission who closes a deal after an unpredictable number of calls, and checking your phone, where some unlocks bring a message worth having and most do not.
A fixed ratio schedule rewards a set number of responses every time, like a coffee card that gives you the tenth cup free. People work steadily towards the reward and tend to pause after it. A variable ratio schedule moves the target around an average, so there is no natural place to pause: responding stays high and steady, and it takes far longer to die out once the rewards stop.
Fixed ratio (a reward after a set number of responses), variable ratio (after an unpredictable number), fixed interval (for the first response after a set time) and variable interval (for the first response after an unpredictable time). Charles Ferster and B.F. Skinner set them out in Schedules of Reinforcement in 1957. Variable ratio produces the highest, steadiest rate of responding of the four.
Because the next pull might be the one. With a predictable reward you learn when to stop; with an unpredictable one there is never a clear signal that it is over, so the behaviour continues through long dry spells. That is why it resists extinction better than any other schedule.
The lawsuit argues that it is, and the psychologists giving evidence in Oakland describe the mechanism in Skinner's terms. Peter's view on the show is that whether or not it was designed with the textbook open, the effect is the same: a feed that pays out unpredictably will keep you scrolling.
The primary literature behind the idea, for anyone who wants to check it. Also on Wikipedia.
Kinderman, P. (2026). Variable ratio reinforcement. Concept of the Week, Experience: Analysing Newsmakers, episode 6, with Hilary Jeffs. https://www.experiencepodco.com/concepts/variable-ratio-reinforcement.html
Mirror neurons · Authenticity · Confirmation bias · Attribution theory · INUS conditions · The illusory truth effect · Role exit · Motivated reasoning · The bystander effect · The halo effect · The spotlight effect · Attentional blindness · Conspicuous consumption
Concept of the Week is a regular feature of Experience: Analysing Newsmakers, the weekly podcast in which clinical psychologist Professor Peter Kinderman and communications expert Hilary Jeffs take the news apart to see what is driving it. Peter picks a psychological idea most of us have half-remembered wrong and explains what it actually shows. New episodes every Tuesday. Where to listen.