Six Rules for Effective Forecasting

We know now that the future is inherently unpredictable because not all variables can be known and even the smallest error imaginable in our data very quickly throws off our predictions. The best we can do is estimate the future by generating realistic, useful probabilities.

The challenge is not forecasting but understanding the past and present. Out of the myriad shifts and adjustments that occur each year, some are clues to the presence of a substantial wave of change and, once assembled into a pattern, point to the fundamental forces at work. The evidence lies in plain sight, waiting for you to read its deeper meani

There are two ways to handle such a world: try to predict, or try to prepare. Prediction is tempting. For all of human history, seers and soothsayers have turned a comfortable trade. The problem is that nearly all studies of “expert” predictions in such complex real-world realms as the stock market, geopolitics, and global finance have proven again

For those determined to master the future, the only recourse is to force-fit a standardised model of a predictable reality onto a world that remains complex and rich but unpredictable in all its individuality.

But the time horizon for accurate forecasts is dauntingly small: the Good Judgment Project found that, while many forecasters were accurate within only about 150 days, its own super-forecasters weren’t confident beyond 400 days.