The meaningful predictions equation

I have written much on predictions before. The following equation appeared in a different format in The arrogance of mathematics and Chaos Theory. I haven’t changed the content of it; only the layout. I’ll explain in a while.

There are major problems with predictions. The first big one is that all predictions depend on ‘the future’ – no surprise there. I’ve always argued – and will continue to argue – that the future does not exist. It is not a place or time. But who am I to say anything if ‘everybody’ believes it exists at a point in time beyond now? In Is the future real – I said “The future is not a place and/or time that ‘we’ can know about, in advance of it happening. One cannot just go into the future, check it out and report back what happened in that future.” But still people want to know what will happen in the future.

Useless predictions

The whole point of The Meaningful Predictions Equation is understanding where value can be added in some predictive statement.

Here are some useless predictions:

1. All human beings will die.
2. The sun will rise in the East tomorrow morning.
3. You will consume a drink containing water within the next 4 days.
4. You will break wind at some point in the next 3 days.
5. You will breathe in and out several times within the next hour.
6. All dogs on earth today will be dead in the next 50 years.

All of the above are about what will happen in ‘the future’ but there are no surprises there. Anybody who predicts the above will be rightly looked upon as a joker or a nut case. They are not valuable or meaningful statements of what will happen because ‘everybody’ knows those things. Why they know, is because the so-called predictions are based on patterns that are commonly known.

Meaningful predictions

Asteroid impact

In my world, meaningful predictions are probabilistic statements that are considered to be of value and importance. In my analysis of Asteroid impact the issue of how risk is perceived became important to predictions about it. Some people may perceive a small risk as unimportant but others higher up in a risk hierarchy may consider a small risk as very important. Hence the potential practical value and importance of a prediction depends partly on the status of who perceives a risk.

But value necessarily depends on other things such as how robust was the analysis, which led to the prediction. I wouldn’t expect there to be much value attached to a prediction from some geezer sitting in a remote cornfield near Manchester; scribbling a few notes on the back of an envelope. But I would expect value to be attached to a scientist who has been studying asteroid movements for the last 10 years, and after careful statistical analysis of real data has come up with some conclusions.

I like the ‘asteroid impact’ thing because there is no regular pattern of asteroid impacts to the earth available to everybody. So – a well reasoned study based on hard data which tells us something worrying, is likely to be of value.

COVID related

Throughout the pandemic – and it isn’t over – everybody wanted to know what would happen next. Our political masters did some lockdowns of the population that proved to be useless. In their minds they must have thought, ‘If we don’t do lockdowns, things could get seriously worse.‘ The latter sort of statement is predictive in nature.

They were correct in their thinking. So – why did lockdowns not work?  I’ve used the analogy of a fast moving ball  and the bat that has to strike it for success. You may have gotten the path of the ball correctly worked out in your mind, but your swing was too late or too early – in which case your middle stumps went flying!

But wait – that your stumps went flying does not mean that your probabilistic estimate of the path of the ball was wrong. You got it wrong because your timing was wrong. Getting it wrong in taking adaptive action does not mean the prediction was wrong.

So with COVID, probabilistic estimates (predictions) were correct, valuable and important. What was wrong was the delayed response. Getting it wrong in the response, does not change the validity of the prediction.

Crossing a busy road

You do not need to be a road-traffic scientist to cross a busy road safely. Ordinary adults do this every day. Only a tiny minority are run over. Crossing safely means observing the flow of traffic in either direction, timing the traffic and spotting the low risk gap in which to cross. Most people will successfully find that gap of about 5 seconds.

Closing your eyes and crossing a busy road means that you’re taking a blind chance. Any fool could predict that your chances of being run over for being so stupid, is pretty high.

Probability

Most people know this is about chance. The chances of being run over when crossing a busy road with eyes closed, is much higher compared to looking with eyes open and taking appropriate actions.

Let’s say for argument’s sake that probability of being run over with eyes closed was 90%. It is still possible to cross a busy with eyes closed and not get run over. Some will say ‘That’s luck‘. Well, it isn’t. There was a 10% probability in this hypothetical (100 – 90 = 10). Mathematically you happened to be one of the minority who would not have been run over.

Take away points

1. Meaningful predictions are probabilistic statements that are considered to be of value and importance. They do not own the future.
2. The value and importance of meaningful predictions is determined by the integrity of data, evidence and analysis that takes place.
3. Where there is a paucity of hard data or evidence the human mind can make reasonable estimates of probability.
4. The validity of a meaningful prediction is not negated by outcomes.