Explanations for gambling addiction: learning theory as applied to…

The idea

Gambling is the awkward case for learning theory: gamblers lose far more often than they win, yet the behaviour is unusually hard to extinguish. The answer is the reinforcement schedule. Wins arrive on a variable ratio schedule, unpredictably after a varying number of bets, and Skinner showed that this schedule produces high, steady responding and the greatest resistance to extinction of all, because no run of losses tells the gambler the reward has stopped coming. Near misses reinforce too, being processed almost like wins. Cognitive theory explains the beliefs that keep the schedule working: the gambler’s fallacy that a loss makes a win due, an illusion of control from choosing numbers, and a recall bias that remembers wins vividly and losses barely. Griffiths (1994) had 60 fruit-machine players think aloud; regular gamblers produced far more irrational verbalisations.

What earns the marks

• Operant conditioning: a win delivers money and excitement immediately after the bet, positively reinforcing gambling; for some, gambling also removes boredom or low mood, negatively reinforcing it.
• Partial reinforcement: only some responses are rewarded, and partially reinforced behaviour is far more resistant to extinction than continuously reinforced behaviour.
• Near misses: two jackpot symbols and a third just off the payline are processed as near-wins and reinforce continued play despite being losses.
• Cognitive bias, gambler’s fallacy: believing independent events are linked, so a run of losses means a win is "due".
• Cognitive bias, illusion of control: believing skill, ritual or choosing one’s own numbers influences a chance outcome.
• Cognitive bias, recall and attributional bias: wins are remembered and attributed to skill, losses are minimised, forgotten or blamed on bad luck, so experience never corrects the belief.

Common mistakes and how to put them right

• Describing partial reinforcement as "sometimes being rewarded" without saying why that matters.

  Put it right: The creditworthy point is resistance to extinction, so a losing streak does not stop the behaviour.

• Confusing variable ratio with variable interval.

  Put it right: Gambling pays after an unpredictable number of bets, not after an unpredictable amount of time.

• Explaining the gambler’s fallacy backwards, saying a winner expects to keep winning.

  Put it right: The fallacy is that past losses make a future win more likely when the events are independent.

• Presenting Griffiths (1994) as showing regular gamblers are more skilful.

  Put it right: Objectively the two groups performed similarly; the difference was in how they talked about the machine.

• Treating think-aloud verbalisations as proven beliefs, when they may be commentary produced for the researcher rather than the thinking that drives play.

Exam tips

• Use the word "unpredictable" explicitly when explaining variable reinforcement; examiners look for it as the feature that resists extinction.
• A scenario naming someone who "feels a win is due" or "picks their own numbers" is handing you the gambler’s fallacy and the illusion of control. Name the bias, then apply it to that person’s detail.
• If asked to compare the two explanations, make the point that they operate at different levels, the schedule maintaining the behaviour and the biases maintaining the belief, rather than declaring one simply better.