“Something will happen in the region this year” is always right and tells you nothing. “X will do Y at place Z inside window W” is usually wrong and, when right, tells you everything. The difference is conjunction — and it's the hidden axis along which a forecast carries information.
A forecaster who says 'something will happen in the region this year' is always right and has told you nothing. One who says 'this actor will do this thing, at this place, inside this window' is usually wrong and, on the rare occasion they're right, has told you everything. The difference between them is conjunction — the stacking of specifics — and it is the hidden axis along which a forecast carries real information.
The mechanics are just probability. Every specific you add to a claim multiplies the ways it can fail: the chance of A-and-B-and-C can only be lower than the chance of A alone. 'An attack' is likely somewhere, sometime. 'An attack, in Moscow, in March or April, on a public gathering' is far less likely to happen by chance — which is exactly why, when it does, it counts for so much more. Specificity is self-imposed difficulty, and difficulty is where information lives.
There's a famous reasoning error hiding underneath this, and it cuts in the forecaster's favor if they're not careful. Kahneman and Tversky's conjunction fallacy showed that people rate a detailed, vivid scenario as more probable than its own components — 'Linda is a bank teller and a feminist' feels likelier than 'Linda is a bank teller,' though it can't be. Pundits exploit the same instinct in reverse: a richly specific narrative feels insightful precisely because the brain misreads detail as likelihood. The discipline is to know exactly how improbable your own conjunction is, and to own that number rather than hide behind the vividness.