A working tool · v0.1

An atlas of prediction

Every field has its own way of guessing what happens next. Statisticians extrapolate, engineers simulate, foresight practitioners build scenarios, designers make futures you can argue with. I kept meeting these methods one at a time, in separate rooms. This is an attempt to put them in the same one: 52 methods, sortable by what you have and what you need.

52 methods 8 families

Find a method — two questions

About

No field owns prediction. Real problems — climate, health, supply chains, industrial transitions — cut across disciplines, but most of us only ever learn the forecasting habits of our own.

I started this list while researching what happens to ships at the end of their lives, where questions about the future arrive from engineering, economics, policy and communities all at once. It is a pilot: around fifty methods, curated rather than complete. Each entry says what a method predicts, where it comes from, how much data it wants, and when it tends to work. The two questions at the top will narrow the list, but most real problems need several methods together, not a single winner.

Spotted an error, or a method I have missed? Tell me. An atlas of prediction is itself a prediction, and it will need revising.