Showing posts with label coincidence. Show all posts
Showing posts with label coincidence. Show all posts

Tuesday, 29 January 2008

Question 5: Good luck or good management? Part One.

What are the chances that something as amazing as you or me could randomly appear in the universe?


What are the chances that a planet capable of supporting life should happen to exist in the universe? What are the chances that the unlikely combination of events that science asserts caused life to errupt should happen spontanteously on such a planet? What are the chances that random mutation and natural selection should lead from that initial simple life to something as sophisticated as the human race? Is it not true that the scientific model of how mankind came to be is just incredibly complicated and convoluted? Is it not much more likely that there is a simple and clean explanation involving a conscious creator?


Is simple always better?


Albert Einstein had a strong conviction that a good model should be a simple and elegant model: if an explanation for something is too complicated then you have probably got it wrong. But there is no scientific justification for this, it is a personal preference or prejudice. Models in physics that Einstein disliked because they were not elegant have turned out to be highly useful and long-lived models. Our preference for simple explanations may be a property of our brains, not a requirement on the universe in general.


We also need to be aware of things that appear very complex but are actually very simple. Anybody who has had a computer with a screen saver is likely to have come across Conway's Game of Life. A relatively complicated and seemingly unpredictable pattern of moving shapes based upon four very simple rules. And this is not a contrived mathematical curiosity; this is a common feature of scientific models ranging from the movement of minute particles to the growing patterns of plants to the logic in the brains of simple creatures and beyond. Simple rules lead to complicated results.


Just how unlikely is unlikely?

In one of the diversions I mentioned the double-edged sword of intuition. One area where intuition can be extremely unhelpful is probability and coincidence.

Experiment has shown that people are generally very bad at judging the relative likelihood of things. For example, if I were to toss a normal coin, which of the two following sequences is more unlikely?

  1) heads tails tails heads tails heads tails
2) heads heads heads heads heads heads heads
Or if I were to pick letters entirely at random, which of the following two sequences is more unlikely?

  a) yrqwburix
b) hellodave
Most people without a basic training in probability will say that sequence (2) and sequence (b) are more unlikely. Hitting a string of seven heads in a row feels unlikely. Randomly generating a recognisable phrase feels unlikely. In fact, in both examples the sequences have exactly the same probability of occuring. If you don't believe this, that is intuition messing with you. It's powerful stuff.

So ... just because a thing seems intuitively unlikely, that does not make it unlikely. You have to take a step back and examine the thing objectively using science.

"What are the chances that a planet capable of supporting life should happen to exist in the universe?"

Likelihood depends broadly on two things: how uncommon the thing is, and how big the sample is. As an everyday example: if one person buys a lottery ticket they are unlikely to win. If several million people buy lottery tickets, one of them is quite likely to win. Most lotteries eventually have a winner.

Lemma: The universe is big.

Very very very big. Our sun is part of the milky way (the vague band of light that you can see in the sky if you live somewhere with a low level of man-made light). The current best estimate for the number of other stars in the milky way is around two trillion (2,000,000,000,000). And the milky way is one of (again at a best estimate) around half a trillion (500,000,000,000) galaxies. A ballpark estimate for the total number of stars in the universe is a septillion (1,000,000,000,000,000,000,000,000). The sun is one of a septillion stars.

A septillion is a great word but not really a number that a human brain can grasp intuitively so let us examine it another way. There are roughly six billion people on the planet. There are an average of 100 thousand hairs (or hair follicles if they're bald) on each of their heads. Now imagine cutting every single one of those hairs into a thousand pieces. There are still around one and a half million stars for every tiny chopped piece of every single hair on every human head on earth. Speaking for myself, I struggle to intuitively grasp six billion people, let alone the hairs on their heads. A septillion is a staggeringly big number.

The random appearance of human beings seems intuitively unlikely. But the universe is even more unintuitively gigantic. So again ... you have to put intuition to one side and attempt to grapple with this subject objectively with science.