Animal communication, astrology, tarot and the like are perfectly harmless when taken as a bit of fun. The trouble starts when they are dressed up in the language of "science", "quantum mechanics" and "the subconscious" — at which point they become outright charlatanry. From what I have observed, most articles that cite scientific theory to prop up fortune-telling are little more than far-fetched leaps of logic and idle chatter. There is even, believe it or not, an animal communicator who explains the practice through quantum entanglement (Quantum Entanglement) — which tells you just how low the scientific bar is set. So why do all these forms of "fortune-telling" hold such a grip on the public? In my view, the reasons are these:
- (1) "Fortune-telling" doesn't actually need to predict the future — a "prophecy" that fits a certain pattern can, in itself, increase the chances of the predicted outcome coming true.
- (2) Even if (1) doesn't hold, so long as there are enough fortune-tellers, some of them will inevitably turn out to be extremely accurate — and those fortune-tellers will not only impress the public, they may even astonish themselves with their own predictions, and so come to believe in them utterly.
The argument of this piece runs as follows: first assume that every form of fortune-telling in the world is bogus, then use a thought experiment to observe what seemingly accurate results (p) a purely fictitious art of prediction can still produce — and thereby establish that observing (p) cannot prove the truth of the prediction. Because whether the art of prediction is genuine or false, (p) can be observed either way.
Prediction versus fortune-telling: what's the difference?
Before we begin, let's settle a few definitions and starting assumptions: what exactly are "prediction" and "fortune-telling"?
In this article, "prediction" means making an inference about the future — for example:
- (a) I'm confident about the English exam, so I should get an A or above.
- (b) I rolled a die three times and got 6, 1 and 5, so I reckon the next Chief Executive will be elected with 615 votes. Both (a) and (b) count as "prediction" in this article.
"Fortune-telling" refers to a method of inferring the future where the result that method produces has no commonsense, generally acceptable connection to its prediction. So (a) is not fortune-telling in method, whereas (b) — along with astrology, tarot and so on — all qualify as fortune-telling in method.
Moreover, this article deals only with predictions that can be judged true or false, and that are not true the moment they are made.
For example: The Hang Seng Index will reach 30,000 points next January.
The so-called "accuracy" of fortune-telling rests in large part on whether its predictions are vague enough to be neither confirmable nor falsifiable — but those predictions are not the focus of this piece. They are, of course, a stock trick of the trade, and interested readers can look up the Barnum Effect (Barnum Effect). Naturally, the fortune-tellers down the lane do occasionally make precise predictions — and it is precisely these kinds of predictions that this article sets out to examine, the aim being to strip away the vague and ambiguous factors and get straight to the heart of what fortune-telling really is.
(1) "Fortune-telling" doesn't need to actually predict the future, "fortune-telling" can create the future
On to the main argument — the reason fortune-telling is so easily believed is that self-fulfilling prophecies (Self-Fulfilling Prophecies) are real enough to give any method of fortune-telling — genuine or false — the power to influence the future through a large proportion of the predictions it produces. Before I unpack this in detail, let me share an anecdote from my secondary school days. Back then I had a close friend who had one notable feature: he was fat — really fat. One day, bored out of my mind and feeling rather childish, I told him point-blank that he fancied a certain girl.
In fact, the fat lad and the girl were barely acquainted — you could hardly say they knew each other at all, let alone that he harboured any secret feelings for her. And yet, after I kept needling him relentlessly, several weeks later he genuinely did fall for that girl. This is what a self-fulfilling prophecy is: because of your own belief in a prediction, you change your own behaviour so as to bring about the outcome the prediction described.
The mechanism behind the art of prediction is really one and the same, and the effect ought to be even stronger than my idle teasing of that fat lad, because I'd guess most people who seek out a fortune-teller's help already have a degree of belief in fortune-telling. Look at the hot topics of fortune-telling — "asking about love", "asking about a relationship" and the like are bound to be near the top of the list — and it is precisely on these topics that the attitude of the person being read can strongly influence the outcome. Take, for instance, a couple on the brink of breaking up who go to have their tarot read, and the fortune-teller tells them: "You two are destined for each other; there's no way you'll break up." Naturally, that one line may well lower the odds of their breaking up. Suppose the couple were originally going to split, but after hearing the fortune-teller's words they stay together. In that case, the prediction has not predicted the outcome — it has manufactured it. And the effect of that prediction (i.e. manufacturing the outcome) has nothing to do with the method of fortune-telling (the tarot). The fortune-teller could just as well use random.org to do the reading, read by the direction the splashes go when he takes a leak, or do no reading at all — so long as he says the same thing, the outcome will be the same.
In short, so long as:
- (a) any method of fortune-telling predicts that X will happen, and;
- (b) at least one person believes that prediction, and;
- (c) that person's belief in the prediction raises the probability of X happening.
- Then — the probability of X happening will be higher than the probability of X happening without any prediction having been made.
In reality, a great many predictions satisfy (c), such as love (where one person's belief is enough to change the outcome) or the stock market (where collective belief can change the outcome).
Traditional fortune-tellers do sometimes invoke so-called "statistics" to back up their results. As far as I've heard — and I may be poorly informed — I have yet to come across a single article that uses statistics correctly to analyse the effectiveness of fortune-telling. But let's grant, for argument's sake, that they would in fact apply statistics correctly.
Suppose we had the following study:
100 men are about to attend a speed-dating event. Before they go, a fortune-teller uses tarot to make a prediction and tells every man the result; the result can only be (a) or (b):
- (a) "You will find a partner at the event", or
- (b) "You will not find a partner at the event".
- 50 men are predicted (a); 50 men are predicted (b).
Two years later, it turns out that of the people predicted (a), 40 found a partner at the event, while of those predicted (b), only 15 found a partner. A result like this is significant (Significant) on a Chi-Square Test (Chi-Square Test) (p = .000001).
That is to say, the group a person was predicted into and whether they found a partner are indeed related. But is a result like this enough to prove that tarot fortune-telling is reliable? The answer is no — because you cannot confirm whether the fortune-telling predicted the outcome or caused it: believing they could find a partner gave these men a surge of confidence and self-assurance, which naturally made them more attractive to women.
Of course, the degree to which a client's belief in the fortune-telling affects reality varies. If a particular fortune-teller could, under carefully controlled experimental conditions, perfectly predict whether several hundred cancer patients would live or die, then there really would be a strong chance that he did possess the power to foresee the future through fortune-telling. Because whether a client believes their cancer can or cannot be cured does, to a greater or lesser extent, affect the course of the illness. But the factors influencing the progression of cancer are many, and it is highly unlikely that the outcome could hinge on belief alone. As for the fortune-telling topics that personal belief can influence — career, love and so on — establishing convincingly that an accurate result is due to the fortune-teller's foresight rather than the client's personal belief would require so many "accurate" results that it becomes an extremely thorny problem. Taking the speed-dating example above, if that fortune-teller could predict every single person's outcome without a single error, then this degree of accuracy would not be so easy to dismiss either.
Let's bend over backwards to be fair: if a fortune-teller claimed to have accurately predicted the romantic fates of several hundred people, and the clients did not even know the results of the readings — so they could not have been influenced by self-fulfilling prophecies — then what? Let's first assume that what he says is no lie — does that mean fortune-telling really can foresee the future?
(2) So long as there are enough fortune-tellers, some will inevitably be extremely accurate – the "fortune-telling" law of large numbers
The first thing to look at is how many readings that fortune-teller performed. If the number of people he read for runs into the thousands, then accurately predicting the outcomes of several hundred of them is hardly surprising. The right standard for judging whether fortune-telling is effective should be: if the fortune-teller were simply guessing at random, how improbable (how low the probability) would it be to arrive at results this "accurate"?
For a fortune-teller guessing at random, a certain number of accurate results is only to be expected. It's like guessing the outcome of a "coin toss": if you can guess wrong (or right) time after time, hitting the mark every single time, then perhaps you do have the power of foresight. After all, to guess wrong (or right) on 20 coin tosses in a row, the probability is only 0.00009% (P = 1/2²⁰); random guessing is very hard to get right. By the standard above, then, if you could guess the outcomes of 20 coin tosses correctly, you might just have fortune-telling powers… wait, why only "might"? It can't be easy to get this many right (or wrong), surely.
In fact the standard set out above runs into two problems when you try to apply it in practice. First, just how low must the probability be before we can confirm that a person has fortune-telling powers? This is, I fear, extremely hard to reach consensus on. Second, there is survivorship bias (Survivorship Bias) — one of today's main subjects.
What exactly is survivorship bias? Let me first share with you a secret recipe for making a fortune, and explain it slowly afterwards:
First, you need a list of 10,000 email addresses. Split the list in half and email each half: tell one half you forecast the stock market will rise next week, and tell the other half it will fall. A week later, if the market has risen, take the list of 5,000 people who received the "the market will rise" message, split it in half again, and repeat the same step. If the market falls, do the same with the people who received the "the market will fall" message. Repeat every week. Five weeks on, roughly 300 people will have watched you perfectly predict the market's movements over the past five weeks with their own eyes. By then they ought to start admiring your "market intuition", and you can step forward as a "stock-market guru", talk these people into investing their money with you, and you're laughing all the way to the bank.
Of course, for the people you don't immediately fool, you can keep sending them emails — with half of them you'll succeed, while the other half will increasingly come to revere you as a deity. Just imagine: if someone had correctly predicted the market's movements ten weeks running, you too would probably be tempted to let them invest on your behalf. The crux of this scam is this: so long as the number of predictions is large enough, some of them are bound to come true and survive. And ordinary people tend to be more deeply impressed by the predictions that do come true — just imagine that if a fortune-teller's prophecy fell flat, most people would simply shrug it off with a laugh; but if it came true, they would bow down in awe.
Ignoring a fortune-teller's failed prophecies while treating the ones that come true as ironclad proof of their abilities — this phenomenon shows up not only in how we view a single fortune-teller, but even more so in how we view fortune-tellers as a group. I have never done a formal count and don't know how many fortune-tellers there are in Hong Kong. But when you always have a few friends who call themselves fortune-tellers, you know the number isn't small. And where there are many fortune-tellers, there will naturally be a grand master among them. Because even if fortune-telling can't truly predict the future, so long as there are enough fortune-tellers, a small handful of them will, by sheer luck, produce more accurate results than everyone else. It is very hard for a single fortune-teller to be right every single time, but finding one fortune-teller who is right every single time within a huge crowd of fortune-tellers is nowhere near as hard. The situation is rather like a large crowd gambling in Macau (let's say at pure games of chance like the slots): most people lose their shirts, while one lone individual wins big — which is hardly grounds for claiming that person has superpowers.
Or, to put it another way, if I claimed that a Mark Six lottery winner has the power to foresee the future — because the odds of winning the Mark Six are so vanishingly low — would you accept that argument? Seen from this angle, those prophets who leave everyone gasping in amazement may be nothing more than people who happen to be as lucky as a Mark Six winner.
Comments
No comments yet — share your thoughts.