Variable rewards

Where the idea comes from

In the 1950s, B. F. Skinner and Charles Ferster ran thousands of hours of experiments on how the timing of rewards shapes behavior, published as Schedules of Reinforcement (1957). One finding stuck. Rewards that come after an unpredictable number of actions, a variable ratio schedule, produce a rapid, steady rate of responding that is the most resistant to extinction. In plain English: when you never know which try will pay off, you keep trying, and you keep going after the rewards dry up.

The textbook example is the slot machine. It pays out on average every so many pulls, but never on a predictable pull.

Tech borrowed the idea openly. In a 2012 post and his 2014 book Hooked, product designer Nir Eyal described variable rewards as a core ingredient of habit-forming products and split them into three kinds: social (“tribe”), information and things (“hunt”), and mastery (“self”).

Why it works on you

Dopamine neurons respond less to rewards themselves than to surprises. A landmark 1997 paper showed they signal the difference between what you expected and what you got, a prediction error that drives learning. A later study found these neurons also respond to uncertainty itself, with sustained activity strongest when a reward is about as likely as not.

A feed is built from that uncertainty. Pull to refresh: maybe something, maybe nothing. A buzz: maybe the message you’re waiting for, maybe a promo. Likes trickle in at random. Infinite scroll means you can pull the lever forever.

This isn’t proof that phones are “like drugs”. It’s a well-understood learning principle, applied at scale.

How to tell it’s working on you

• You check the same app several times in ten minutes, knowing nothing’s changed.
• You refresh after posting to see who’s liked it.
• One great post justifies, in your head, the 40 boring ones before it.
• Getting a notification feels better than reading it.

What helps

Make rewards predictable. Batch your checks: email at 10 and 4, social after dinner. On recent iOS versions you can also group non-urgent notifications into a scheduled summary.

Remove the lever. Turn off likes and “someone posted” alerts, and move the apps out of thumb’s reach. That’s stimulus control.

Make each pull cost something. Unpredictable rewards rely on checks being free. In Bricked, once your daily limit is used up, each extra check is only a 5-minute peek. With Bricked+ you unlock it with a deliberate hold, not a tap, and the hold gets longer each time that day. On the free plan it’s a wait that grows the same way.

Give your brain better surprises. Unscripted conversations, sport and new places are also unpredictable, and they leave you with more than a feed does.