Tuesday, 21 October 2008

Patterns are comforting, even when imaginary, says Steve Jones .

From here;

On the opposite side of my rather dingy London square are several blocks of council flats; solid and comfortable, but less attractive than the Victorian houses facing them. They were built in the 1950s to replace a whole row destroyed during the war. The square is within a mile of three major railway stations, and no doubt the bombs were aimed there, but missed. The locals paid the price in blood and bricks and the East End, famously, paid an even higher one, as explosives rained on its streets while the effete inhabitants of South Kensington slept peacefully at home.

Except, that is, that a map of where the missiles actually landed shows that their overall distribution in inner London was more or less random. Yet the drama of the time led many people to a frantic search for regularities in an attempt to predict their own risk of being hit. Many Londoners reported that the sound of a V-1 flying bomb made them feel that it was aimed straight at them, which was unlikely at best.

Man is the pattern-seeking animal, so much so that he finds order even when it is not there. Books that claim that pets bark just before their owners come home, or that thinking about a friend makes that friend more likely to call, sell millions. The world has more astrologers than astronomers - and devotees of the former are much more likely than average to see shapes, figures and messages when given a piece of paper covered in random dots.

Such self-delusion is shared by their mentors, for Aries, Libra, Virgo and the other constellations are arbitrary groups of points named after non-existent relationships among stars that are in truth billions of light years apart.

People often turn to mystics in hard times (as after the First World War, when spiritualists did brisk business with bereaved parents). A skydiver about to jump might perceive an image of a parachute in a scatter of random spots, although when safe on the ground he can identify nothing. People who have not eaten for hours are more likely to find hints of sandwiches or forks in a meaningless jumble of dots than those who have just risen from the table. In the same way, poor children overestimate the size of a pile of coins more than their richer fellows, and students asked to estimate the height of professors compared with mere lecturers put mythical inches on the Man in Charge (which for me, at 5ft 7in, is welcome news.

It's all a matter of being in control of events. If one is being bombed, starved, or bullied, seeing a pattern in one's daily horrors, real or not, brings some internal comfort by providing the illusion of order, and even of personal authority, non-existent though it is. As a result, people tend to see more false regularities in unpleasant, rather than pleasant, times: when faced with bullies and bankruptcy rather than friends and wealth. Even sceptics cross their fingers or knock on wood more often when they try to stave off bad news than when they hope for good. In the financial world, that really matters, for to notice a trend in the market before anyone else does may be the road to fortune. Unfortunately, self-delusion in hard times is the rule there, too. Panels of investors given good or bad forecasts about a company's prospects behave rationally when times are calm, but once stocks gets turbulent and the graphs leap up and down, they detect pointers to further decline even when they do not exist. The financial panjandrum Warren Buffett has noticed - as he says, "Be fearful when others are greedy, and be greedy when others are fearful". If, at the bottom of the market, people imagine that they can see more depressing patterns than cheerful ones, spurious as both might be, they will sell when they should buy (or do nothing). The astute take advantage of such self-delusion and make a killing.

Brokers employ (or did until a few weeks ago) thousands of mathematicians who try to identify, and exploit, such irrational behaviour. Even so, from the Great Crash until today, a random model of the stock market's peaks and troughs - one that turns on the assumption of an overall lack of pattern over 60 years - fits the figures almost as well as do the most sophisticated analyses. As the financial bombs fall upon Lombard and Wall Streets, it is worth remembering in these uncertain times that your stocks are not necessarily the ones that are going to be hit.


  1. Provocative post, Psi.

    Given the ability to make patterns is crucial to our evolution. There are many cases where over-seeing patterns may be useful. Said more statistically, if the null hypothesis is noise and the value of accepting the null hypothesis is business as usual, then one might accept an alternative hypothesis if it means avoiding being bombed, losing all your money, gaining a lot of money, avoiding death, etc. It may even be just to add a little spice to life.

    When the lottery here is over $100 million dollars, if I'm at a store that sells lottery tickets, I'll buy one. I really don't expect to win, but the sacrifice on my part of $1 gives me the 5-minute dream of what I could do with those millions and it changes my perspective to where I believe, it may come true, just this one time, because I am destined for it.

  2. Hi Tom,

    The lottery isn't that big over here.

    My wife (Maths teacher) has a genuine tip to increase your chances of enjoying a lottery win.

    Buy your ticket just a few hours before the draw.

    You are then statistically more likely to enjoy your win.

    The reason is simple - depending on your age - if you buy the ticket a couple of days before the draw then you are more likely to die before the draw and so miss out on enjoying a win, than you are to win.

    Good luck.

    - - -

    BTW I will be pointing out Cliff's lottery fallacy to him (again) in due course.




    I am registered to start a science degree next year!!

  3. Good point about the timing of the lottery ticket purchase. I'd never considered the mortality odds!

    My four-year old was talking about the dangers of lightning the other day. I was trying to tell her that even if she tried to get hit, she never would, but all she seemed tuned to was that a boy at school said a dog got hit and killed by lightning.

    Good luck on the science education. Is that to be a researcher or an educator, or both?

  4. Hi Tom,


    It is a part time course so might take five or six years. BSc Natural Sciences at the Open University.

    Got a fulltime career for the next ten years or so though first. But when the kids have left home and the mortgage has gone. I will look to do the teacher training and have a "second career".

    Course doesn't start till next Feb - just doing ore-Reading until then.

    Regards Psi