The Need for Control – Skinner & Superstitious Pigeons
We seem to be hard-wired to seek causality – finding correlations where there are none – and struggle to see events as independent. Early man had a very limited understanding of the mechanics of a harsh environment so desperately sought to infer cause or consequence of actions for events that were either random, or simply too complex to understand. Superstition and ritual emerged as ways to somehow gain a sense of control, and these are still very evident today.
Fascinating experiments illustrate this behaviour right across the animal kingdom, one of the most famous being that of Harvard psychologist B. F. Skinner.
In 1948 Skinner introduced food hoppers at timed intervals into a pigeon’s cage. The pigeon had no influence over the frequency of the feeding, but whatever behaviour they happened to be displaying as food was introduced – nodding head, turning – became conditioned as responsible for the appearance of the food. Each pigeon developed a unique ritual that it thought triggered the reward.
The same is true for humans, who have associated a sense of control with various rituals (e.g. avoiding black cats, never walking under ladders). These still occur today in less obvious places. Studies of City Traders, for example, have shown that they perceive a sense of control over random systems.
Skinner’s findings led him to define what he called Operant Conditioning – responses that reinforced, punished or had a neutral impact on behaviour. Personal choice plays a big part in reinforcement and casino games provide great examples of this.
The Influence of Personal Choice – Wanting to Throw the Dice
The majority of casino games are entirely random, yet bettors display an irrational belief in their ability to control outcomes where they can assert personal choice.
This tendency is illustrated by gamblers betting more when they are the ones throwing the dice in Craps, or spinning the ball into the roulette wheel. In both examples bettors demonstrate an irrational belief in their ability to control what is a random outcome because they exercise personal choice in the process of the game. This is also true by proxy, such as when bettors piggyback on the bets of someone on a lucky streak – someone perceived to be especially in control.
Lotteries provide another illustration of the illusion of control as players place greater confidence in potential success when playing their own numbers, rather than machine-generated entries. Given the outcome is entirely random, the chances of success are the same.
In fact, random lottery selections often increase potential winnings because of the common use of birth dates for numbers. The chances of winning are the same, but the likelihood of sharing the prize increases as selections are skewed towards numbers from 1 to 31, decreasing the potential payout.
In the casino examples, the behaviours that produced success – how the ball was thrown, the ritual involved – will be reinforced, whereas those producing failure are likely to generate a punishment response, decreasing our propensity to repeat them.
However, because life and gambling cannot simply be reduced to narrow definitions of success and failure, and the causes thereof, the potential for irrationally reinforcing behaviour is enormous.
Dangerous Information & Near Misses – Beware Back-Fitting
You may have heard the expression ‘a little information is a dangerous thing’. This is particularly true in gambling, where the sense of empowerment that a degree of knowledge about a team, sports or event, gives rise to an exaggerated sense of understanding and predictive ability.
Where success follows, but not necessarily as expected, bettors often back-fit the logic of their choices to match what happened, producing the reinforcement described above, and the illusion of control.
This is equally true of near misses, where bettors take enough encouragement from almost getting things right, to reinforce the behaviour. As a result they persevere with what they think is a valid approach, when in fact in all probability there is no valid correlation.
Some things to bear in mind to mitigate this effect:
- Accept that you have no control where outcomes are random
- Test narrow, clearly defined hypotheses, one at a time
- Don’t take betting tips simply at face value
- Try and differentiate between signal and noise
As with superstition, the pathological association of betting outcomes and particular behaviours is unfortunately hard to resist, because these are hard-wired behaviours. The key for bettors is to be as disciplined as possible when drawing conclusions from their betting history, and keep the “lucky pants” for clubbing.