I am not a number! I am a stupid prediction!

What is it with numbers? Why the need to generate them when it makes absolutely no sense? Why can't we cheerfully accept that the most accurate answers to many important questions are either "shed loads" or "bugger all"?

Exhibit A is yesterday's comically daft prediction by Japan's National Institute for Environment Studies in a report on floods compiled at the behest of the government. Its alarmist finding was dutifully reported in the Nikkei and a couple of other papers. By the end of the century, it shivered, the annual cost to Japan of dealing with flooding could hit 8.7 trillion yen. Not Y8 trillion. Not Y9 trillion. But Y8.7 trillion.

Yup. You read it correctly. This agency, bloated with state-funding and addled by the need to feel relevant, is offering a financial prediction based on a 91-year forward view and, astoundingly, uses a decimal point in the final figure. Which is lunacy. We are talking about 91 years in the future, for goodness sake. Anything could happen. Imagine running a government think tank in 1918 and predicting that a flight to Milan from London will cost less than the egg sandwich they serve on board, or that every word ever printed could be stored on a piece of processed sand the size of a peanut. Nobody would listen. Come up with a number - Wilbur Wright saying in 1901 that man would not fly for 50 years - and you somehow seem more credible.

Analysed more closely, that Y8.7 figure is truly ridiculous. There may not be a yen in 91 years' time and even if there were, 8.7 trillion of them might buy you a book of matches, a clip of ammunition for your plasma rifle or Italy. We may be living on the moon, or in boats or space stations. Money may not exist in the current sense of the word. The oceans could subside, rather than rise. Rain may be a yearned-for treat, rather than a dangerous threat to the economy.

This bizarre prognostication, remember, came out on a day that the government of Singapore downgraded its economic outlook for the third time in five months. That's right. A tiny city state, with a notoriously tight governmental rein on economic minutiae, has, in the space of five months, twice been forced to admit that it cannot predict where that economy will be in eight weeks' time. Singapore is hardly alone. This crisis has made a hilarious daily mockery of almost every numerical prediction out there, from Toyota to OPEC.

All very entertaining, but we need to be more realistic about things. It is not just the recent financial turmoil that has made a mockery of forecasting, but history itself. We are sensationally bad at prediction, and always have been. And forget the sort of howlers that feature in trite after-dinner speeches -  Tom Watson of IBM predicting a global market for just five computers, for example. Every day unleashes a relentless cascade of evidence that we have absolutely no idea what will happen tomorrow.

Comments

[yusoshi]

Come up with a number? Howabout zero - the probability that any of us will live long enough to find out and therefore the level that we care?