The wisdom of crowds has no use for predicting the lottery

Michael Giberson

English illusionist Derren Brown hosted a live television show last week during which he appeared to have predicted winning lottery numbers.  He subsequently claimed (among other things) that he used the “wisdom of crowds” to generate the prediction.

In a follow-up show last night, watched by 3 million people, Brown said he used “a powerful, beautiful secret that can only be achieved when we all put our heads together.”

He went on to say that he had gathered a panel of 24 people who wrote down their predictions after studying the last year’s worth of numbers.

The guesses for each ball were then added up and divided by 24 to get the average guess.

Brown said it took a while to perfect the “deep maths” technique.

According to him, the predictions were correct because of the “wisdom of the crowd” theory which suggests that a large group of people making average guesses will come up with the correct figure as an average of all their attempts.

The explanation is, of course, complete bunk and an editor at the Research Digest blog of the British Psychological Society will have none of it:

… Brown [has] committed a disservice to the public understanding of psychology. He invoked a real, fascinating phenomenon in social psychology – the so-called “wisdom of crowds” – distorted it, and half-baked it with flim flam about “automatic writing” and “deep maths”.

The wisdom of crowds is the consistent finding that the averaged judgements of a diverse group of independent people will nearly always be more accurate than any single person’s judgement, no matter how expert that individual is…. Judgements biased in one direction will be cancelled out by judgements biased in the other direction, as the group’s combined verdict homes in on the truth.

… There’s also a fascinating literature on why crowds often work badly, rather than fulfilling their potential for wisdom. In group meetings, for example, research shows that people have an unfortunate tendency to talk about the information that they share, thereby undermining the diversity of knowledge in the group. Similarly, social dynamics can lead to diseases of the crowd such as “group think“….

Returning to Derren Brown’s lottery explanation, we can see that the wisdom of crowds has no use for predicting the lottery. His group of 24 individuals did not have diverse insight into what numbers will come next.

In addition, the “wisdom of crowds” is of no advantage in predicting random draws from a known distribution, as various math and risk experts have explained.

How did Brown do it? Clever camera work is the consensus answer. Chris Masse at Midas Oracle has collected several video clips – Brown’s prediction, Brown’s explanation, and several clips describing the camera tricks involved – along with supporting explanations.  Midas Oracle is, of course, the group blog devoted to prediction markets and other wisdom-of-crowd topics.

It is a kind of compliment to the math and science behind the wisdom-of-crowds phenomena that Brown would think to invoke it in his post-prediction patter.  Brown believed that enough of his audience would have some awareness of the concept that his references might seem plausible.  He also believed that his audience wouldn’t understand it well enough to see through his claims.

Yet pseudo-scientific claims do water down public understanding of what real science is. One magician’s claims about one of his stunts will not undermine science, of course, but it worthwhile to expose such pseudo-scientific quackery when spotted.  The stunt may have been sort of entertaining, but the explanation is all bunk.