Learning to make predictions

Why Prediction?

One of the things we are constantly doing is predicting the future, whether it is the micropredictions of movement or forecasts of what might happen next year.

To feel in control

A basic need we have is for a sense of control. If we can predict what will happen, this gives us a lot better chance to control things.

If we do not know what will happen next then we cannot relax and must constantly be on the lookout for danger.

To decide

In our ruminating and decision-making we are constantly looking forward, trying to decide the best course of action to achieve our goals and avoid potential discomforts.

If we can predict accurately, then we will make good decisions and be successful in meeting our goals and objectives.

Cause and effect

Being able to predict is about connecting cause and effect. If we can connect the cause of today to the effect of tomorrow, we can predict. And if we can create a chain of these, we can predict what will happen next week.

Being able to explain cause and effect meets yet another need and allows us to appear rational to other people, thus appearing predictable (and hence meeting their needs for prediction and control).

When we are right and wrong

We make many predictions, often based on a lot of opinion and hope, and on very little fact. Unsurprisingly, we are wrong quite often.

Being wrong can be surprising, but we are used to it and generally cope by making various excuses, typically blaming contextual factors. When you are making a public prediction, it helps to predict much like others are predicting, as when you all are wrong, you can hide in the crowd.

When you are right, you can trumpet your predictive powers. As we hide our predictive failures, the few successes can sound quite amazing.

The professionals

If you are unable to predict, there are many industries based on predicting the future, from market analysts who tell you what shares to buy and sell, to weather forecasters who model winds and clouds.

Even well-qualified people are bad at predicting, and studies of Wall Street analysts and football experts have been shown to be little better than average in their predictions.

Pundits are often careful to lace their words with possibility rather than certainty, saying how shares could hit an all-time high or that there is a 30% chance of rain this afternoon. But they say it with such conviction, we believe the stronger emotions over the weaker words.

And of course there are the more esoteric seers and prophets who will turn cards or peek at tea leaves to divine your personal future. Science scorns such methods but the industry thrives nevertheless. Our need to know the future leads us to blindly accept the pronouncements of anyone who will point the way, and perhaps especially when they take our money for the pleasure.

Professionals know the traps and often choose between two strategies. The linear strategy is based on the principle that patterns repeat, and so tomorrow will likely be like next year. But such forecasters are blind to the big changes that occur (Karl Marx said that when train of history hits a curve, the academics fall off.)

Non-linear forecasters know that things sometimes change radically, and seek to become known as the person who predicted the crash or the boom. They hence make frequent alarming predictions, safe in the knowledge that their failures can be safely forgotten whilst their one or two great successes can bring them fame and fortune. Studies of economic forecasters have shown that those who get the big ones right are worse than average at predicting 'normal' events.