We’ve all heard about positive reinforcement and negative reinforcement; the idea of rewarding good behavior and punishing bad behavior. This is operant conditioning, wherein the strength of a behavior is modified by reward or punishment.
There’s another type of reinforcement that is particularly manipulative and addictive, called intermittent, or random, reinforcement. This is based on the finding that if you want to condition someone to do something, rewarding them consistently may not be the best way to do it. What’s more effective is rewarding at random intervals. This technique influences casinos, advertisers, romantic partners, video games, investors and you, the reader.
As psychoanalyst Roberta Satow says: “Many of us have been on the wrong side of intermittent reinforcement — hungering for the crumbs that we sometimes get and sometimes don’t — hoping that this time we will get it.”
Psychologists are discovering why things like slot machines can induce financially crippling addiction. A key ingredient is random reinforcement.
Slot machines do not payout every 5th, every 10th, or any other fixed amount of plays. It pays out randomly. This means that it is conceivably possible to win several times in a row. As a result, players are able to rationalize their losses by hoping for a streak of good luck.
“Few people will ever realize a large jackpot win, and thus the impact of this feature of [slot machines] is largely psychological rather than actual. In other words, people are influenced by the possibilities of a win rather than the actuality of winning. Second, for a lucky few, the actual experience of a jackpot win can influence their subsequent gambling behavior. In fact, the presence of a big win – but not necessarily a jackpot win – is often noted as a motivating factor in the gambling careers of problem players.
A frequent observation is that many problem gamblers experience a “big win” early in their gambling careers that is subjectively felt to be a motivating factor in their continuing gambling involvement. In particular, it may create the impression that reasonably large wins are fairly common, and thus gambling could represent a net economic gain over time. Therefore, anecdotally, winning jackpots seems to have the potential to exacerbate, or even cause, gambling problems” .
Some relationships are based on intermittent reinforcement. In these kinds of relationships, the things we need like love, connection, belonging, appreciation, affection and commitment, are only granted inconsistently, unpredictably and occasionally. People in intermittent reinforcement relationships give or withhold certain needs from their partner, granting them either randomly enough to develop an addiction, or giving them in response to the exact behavior they want to see in their partner. For example: the boyfriend that is sweet and kind, buys you stuff, but then ignores you for days on a whim. Or alternatively, the girlfriend that cooks, cleans, makes lunch for you, but then flirts with other men in front of you.
The fact that needs are only granted occasionally, in an intermittent reinforcement pattern, keeps their partners hooked. These “hot and cold” relationships are notoriously addictive, and may partly explain why some people don’t leave their abusive partners.
Self proclaimed pickup artists have operationalized intermittent reinforcement as a technique, known as push/pull. People suffering from borderline personality disorder may also unintentionally push/pull in their relationships.
Traders, especially day-traders, have puzzled psychologists for decades. On one hand, there is general agreement among researchers that trading is more like rolling dice than like playing poker. That is, trying to actively predict stocks is more of a gamble than a game of skill . As such, as little as 13% of the day trader population earns money. Despite the numbers overwhelmingly being stacked against traders, they still continue their endeavors. Everybody thinks they can beat the market, yet science says nearly no one can.
The reason they continue trading is random reinforcement. Investopedia describes the phenomenon as such: “Random reinforcement occurs when traders are rewarded (at random) for behaviors that are not actually of their own doing. Because of random reinforcement, traders may grow overconfident in their ability even though it may be objectively unfounded” .
Trading is similar in some ways to gambling addicts who keep playing because they win just enough to keep them there, but of course they are losing money over the long run.
The ‘Like’—a popular feature on social media—shares features with both monetary and social rewards as a means of feedback that shapes reinforcement learning. The Like is a new kind of reward: as with money, it is a secondary reinforcer, and it is represented by discrete values .
When you check your social media you sometimes have new likes and notifications, but you rarely know for sure when a new like will come, so the habit of checking is reinforced. This follows the pattern of random reinforcement, and is why social media addiction is so prevalent.
“You pull down and refresh [say your newsfeed, or email], and there is going to be something new at the top,” says Tristan Harris, formerly a design ethicist at Google. “It operates just like the slot machines in Vegas.”
It’s the same for dating apps like Tinder. “We’re playing a slot machine to see if we got a match,” Harris says. “When we tap the number of red notifications, we’re playing a slot machine to what’s underneath” .
Games routinely implement intermittent reinforcement patterns to make them more addictive. Catching Pokémon is an example, wherein a successful catch is the reward, and the catch-rate causes intermittent and irregular patterns. Another example is the famous Mario Question Mark Block, whereby players rarely know what type of item will come from them. As a result, ? Blocks function similarly to lottery tickets, and players become conditioned to interact with them.
In a study of video games, researchers found that video games utilizing intermittent reinforcement are significantly more enjoyable than video games where actions are reinforced every time, known as continuous reinforcement . Another study found that when playing games with intermittent reinforcement patterns, players are less likely to quit gameplay when compared to playing games with predictable rewards . In fact, it is assumed that these intermittent reward schedules play a critical role in video game addiction .
Moreover, the similarity of intermittent reinforcement in video games and gambling, may suggest that gaming is predictive of gambling problems. One paper found that high frequency video-game players gamble more than low frequency video-game players, and take greater risks on the blackjack gambling task although no overall differences in success were found .