mxx1's avatar

Information entropy

Predicting the future is a mug’s game, as we all know.

You can reliably predict that the sun will come up tomorrow, but you can’t reliably nominate when the next car accident will occur in your street.

However a well informed expert can tell you that you are due one accident per year. She just can’t tell you when it will occur. And you might in fact get none, one, two or even three. But the average over ten years will in fact be one, until it’s not.

With great effort, your French technocrat will delve into why accidents happen on your street. They will measure your asphalt, scan your camber, record your pedestrians, breathylyse your drivers, etc. Then they will forensically deconstruct every prior accident and correlate them to these and a million other factors.

Eventually they’ll have a model which might predict under what conditions you are likely to have an accident.

Then you will have something like this: tomorrow there is a 0.000005% chance of an accident between 6pm and 7pm, but it depends on what time Larry gets RSA’ed, which itself depends on how many mates he’s drinking with, which itself depends on how many of them have cash, which depends on which of them won a jackpot the day before.

Then the whole palaver starts again because jackpots are reliable in terms of percentages but inconveniently unreliable in terms of “c’mon give it to me, now”.

Ironically, if it’s raining the probability of an accident goes up 10x.

Eventually you realise that you are fighting a form of information entropy. So much energy goes into modelling all the factors that are needed to vaguely foresee the future, that any beneficial returns from doing so are negated.

Which leads us to the Second law of information thermodynamics:

More enterprise is required to predict the timing of future events than any benefit that can possibly be derived from doing so.

And yet a good slab of western society remains enamoured to the task. What keeps them going is luck. You just have to be luckier than half of your competitors and you’re in front.

So, so many successful seers think that they are geniuses. But they can’t even crunch the simple numbers to see that statistically they are a one in two chance of being more successful than the next Nostradamus.