ABSTRACT
A forecast is just the set of probabilities attached to a set of future events. In order to understand a forecast, all one needs to do is to interpret those two bits of information. Unfortunately, there are pitfalls to communicating each element, so that the user of a forecast understands what its producer means. One source of potential problems is ambiguity regarding the event being predicted and what exactly is being said about it. Another is the difficulty of determining the relevance of the problem that the forecaster has solved for the problem that the user is facing. Problems can also arise out of epistemological and sociological issues of trust and context. A simple framework is offered for considering these communication problems and is then illustrated with a mixture of systematic data and anecdotal observation. The criticality of these different problems is considered, along with procedures that might reduce them.
