Derren Brown is a very clever man, and he does what he does very well.
My issue with him is mostly that, while he purports to use misdirection and illusion, people still actually believe that he has some kind of psychic powers. Derren himself doesn't even believe in what he's telling other people to believe – but then it wouldn't work if he didn't manage it convincingly! Such a tricky situation, especially for the whole making money side of this malarkey.
Anyway, onto the show. Basically it was a lot of faff and bullshit, with some false complexities thrown in. I'm not a mathematician although I was watching with one; which is lucky, because if you didn't have someone there saying "I know this, it's a simple maths game" etc, then it must have been very easy for his "deep maths" blurb to wash over your head.
I'm not going to touch on his ability to talk people towards numbers and decisions – I haven't studied what he does as closely as others but I find it fascinating how it works. Luckily that has nothing to do with "predicting" lottery results.
Let's start with the coin toss game. Thankfully Charlie immediately recognised this game. I originally thought it would have been played with each person grouping their coin-toss results into sets of three. That would have been entirely fair. But because each coin-toss result was in an ongoing list, then the waters get muddied.
Afterwards, he kindly showed us the rule that you can use to trick your only friend who didn't watch the show or hasn't heard of this game. The reason – I think, I haven't had it explained – it works is that the inverted coin-face taken at the start of YOUR chosen grouping acts as a "reset" button effectively, and because it's in your grouping, you will always effectively be "one step ahead" of your opponent. The only exception is if they get their combination straight away. As Charlie says, there's a 1/8 chance of that happening, so each time you play it's a 7/8 chance you will win.
By calling this "deep maths" he pitches it straight over the heads of the majority of his audience who turn to themselves/partners/friend
Next he mentioned the concept of group consciousness. I made a lot of scoffing noises at this point, and got more and more obnoxious about it because I was so angry at him talking about all this absolute nonsense that he, as a skeptic, knows is not true! This also goes for automatic writing. And the lovely team-building exercises that were just there to give a nice video clip but served no purpose in "choosing the numbers".
Anyway, Francis Galton and the ox. I've just skimmed this story as I'm rushing to write this (while the RAGE continues!) but 800 people see an ox and guess its weight. He adds up each guess and divides by 800 and lo and behold, it's 1lb off! Magical? No.
These people looked at an ox and guessed what it might weigh. Some were butchers and farmers who might have an idea what an ox that size would weigh, and some were just ordinary people with maybe less of a clue. So slightly more informed people would be more likely to guess close to the weight, and less informed people are more likely to guess ridiculously high or ridiculously low. Now I don't remember a lot of A level maths, but that sounds possibly like it could be to do with a normal distribution, which draws a bell curve and you can calculate how likely it is that someone guessed a certain weight.
This story isn't analogous to guessing lottery numbers at all! As Derren transferred from this story into his Wisdom of Crowds (scoff scoff), I had one thing to say –
"There's no ox!"
Where is the ox? Where is what his group are trying to get as close to as possible, whilst guessing?
There isn't one! There's no aim or template or stencil, no facsimile they're trying to emulate. It just. Doesn't. Work. They're picking random numbers.
They are also doing it with a year of past lottery numbers in front of them, having studied them. He's not quite saying it, but again is insinuating a link between past numbers and ones that are more likely to come up next; a trap that people who may have previously self-confessed as not really understanding probabilities might fall into. It's a simple concept, and I think if he'd come outright and stated that the past numbers had a link to the future numbers then people would have called his bluff. But they were susceptible (as everyone he uses is) and it wasn't completely stated in that way.
Anyway. Again, I'm not a mathematician- A level was a long time and a lot of beer ago- even though I'm lucky enough to spend some of my time in the vicinity of one. ("lucky" refers to SOME of the time…only joking. Honest.)
But there is something fundamentally flawed with averaging out lottery numbers, because they are just numbers with no actual quality. Now bear with me here. We can look at oxen; four oxen, four different weights. Add their weights together and divide by four: we have the average weight of an ox in that group. Super!
Lottery numbers are not indicative of pounds, kilometres, sweets in jars or anything else. They are numbers just because that is the easiest way of denoting 49 different things. We could have had The National Lottery and each ball had a different letter on it! Can you add up letters and average them out? No. Or colours?
Imagine that. Let's cut down Lotto to a manageable size. There are 6 balls: Red, Orange, Yellow, Green, Blue and Purple. One ball is chosen at random each week (in a mini machine like Lancelot or Guinevere or whatever they call the machines nowadays?) and people who guessed correctly it would be that colour ball win some money. Lovely!
But the problem with that for Derren is that we can't play with these balls (oo-err) and find any hidden meanings with them as easily as we could if they were numbers. Imagine now a 49 ball Lotto, but still each of them are colours. That's tough! But entirely feasible. Just really difficult to run. I can imagine it now,
"This week's winning balls are, Burnt Sienna, Sky Blue, Azure, Eggshell, Lime and Orange! A good week for the red-based balls, and two citrus colours! What will happen next week?"
So the uses of the numbers are being completely mixed up here. There are digits being used to denote different balls, and there are mathematical results. By averaging out random guesses, you are mixing up the uses in a meaningless way.
However, he matched some! How did he do that?
I can't go into it deeply because I don't possess the depth of knowledge to explain it all in an unquestionable way, but in a basic way you can see it like this; Claire and Charlie guess some lottery numbers and average them out. For the first ball, Claire chooses 1. Charlie chooses 49. This gives them an mean of 25.
For the second ball, Claire chooses 10 and Charlie chooses 40. The average is 25.
I could carry on. The mean will go up and down but it'll tend to be around the middle of the range, because that's what it'll do. The numbers they got were all middle type numbers anyway. That's one thing.
Secondly, which is more important although slightly less about this maths – and Derren did give clues to this – he has spent about a year on this. There are Lotto draws twice a week. He's already demonstrated selective showing of filmed sequences before, in "The System". To me this clearly indicates what he's done; the same tedious exercise over and over again, filming people over and over – maybe the same people, maybe not, I don't know, I haven't thought too deeply about this, but doing this enough times until the right mean numbers come up. He has a few shots at it. The Lotto wins showed were earlier this month. What's he been doing before then?!
Right. So we've covered misleading statistics, and averaging out useless numbers. How did he do it? Well, he never showed the people in the group that "predicted" for him what the numbers were; there is no proof that he used those numbers that they had generated.
I'm sure they could have been held incommunicado for an amount of time – hey, they do it with Big Brother contestants! – to protect the numbers, if necessary. But they weren't.
There was trickery afoot, and I can only speculate as to what exactly he did, but I'm gutted that he convinced so many people with his silly number averaging thing.
PS I was actually quite interested to find out how he did, which is partly why I was angry enough to write this! But mostly because I hated what he did with the misleading maths.