I assume you will write in detail about scoring in later posts. Could you please explain how to evaluate forecasts that were made at different points of time im that context? E.g. your own changing invasion probabilities.
That's a great question. I should probably do a whole post on revising forecasts. But broadly speaking I revise my forecasts for some combination of three reasons: 1) the facts have changed; 2) I've learned new information; and 3) I've changed my mind about what conclusion to draw from the facts. As a rule of thumb, if your initial forecast was good you shouldn't have to revise your probabilities too much. But sometimes something major does change, or you learn something new, or you just come to believe you made a serious mistake in your reasoning. In forecasting the war in Ukraine, for example, all three things have been true: the nature of the conflict has changed dramatically, I learned a lot about both sides' capabilities, and I came to believe I was wrong about some things that I maybe should have known early in the war.
I assume you will write in detail about scoring in later posts. Could you please explain how to evaluate forecasts that were made at different points of time im that context? E.g. your own changing invasion probabilities.
That's a great question. I should probably do a whole post on revising forecasts. But broadly speaking I revise my forecasts for some combination of three reasons: 1) the facts have changed; 2) I've learned new information; and 3) I've changed my mind about what conclusion to draw from the facts. As a rule of thumb, if your initial forecast was good you shouldn't have to revise your probabilities too much. But sometimes something major does change, or you learn something new, or you just come to believe you made a serious mistake in your reasoning. In forecasting the war in Ukraine, for example, all three things have been true: the nature of the conflict has changed dramatically, I learned a lot about both sides' capabilities, and I came to believe I was wrong about some things that I maybe should have known early in the war.