I think the definition of crying wolf is announcing a wolf attack when you don’t really believe there is one. It makes you untrustworthy, rather than the general notion of a wolf attack.
So long as you accurately report your beliefs regarding an upcoming wolf attack, you don’t need to worry about crying wolf. People will see your predictions, assess how often you are correct, and use your predictions to adjust their own to the degree that they consider optimal. Trying to rig this system by over- or under- reporting just throws everybody off; nobody is starting from the assumption that you are infallible.
It seems like if you want to get people to believe you according to this, you should make predictions where it is hard to prove that you are incorrect (for instance, by giving no timetable). Ideally, people should discount your predictions by exactly as much as the vagueness in the prediction justifies, but real humans don’t behave ideally.
I don’t see how that follows. If someone’s predictions are all nebulous, unfalsifiable claims, then we start out viewing them as a useless predictor, and never update our beliefs about them.
Plenty of people have said things akin to “there will eventually be a wolf attack”, and while they’re quick to laud themselves when one inevitably happens, they don’t generally get rewarded for this. People who make unlikely predictions and win? They get a spotlight.
Come to think of it, I’m surprised we don’t see a bit more clout being awarded to high-scoring polymarket predictors. A guy with a public record of making millions of dollars a year by predicting military outcomes that others bet against seems like he should get a slew of prestigious consultant positions, but I haven’t noticed it happening.
I think the definition of crying wolf is announcing a wolf attack when you don’t really believe there is one. It makes you untrustworthy, rather than the general notion of a wolf attack.
So long as you accurately report your beliefs regarding an upcoming wolf attack, you don’t need to worry about crying wolf. People will see your predictions, assess how often you are correct, and use your predictions to adjust their own to the degree that they consider optimal. Trying to rig this system by over- or under- reporting just throws everybody off; nobody is starting from the assumption that you are infallible.
It seems like if you want to get people to believe you according to this, you should make predictions where it is hard to prove that you are incorrect (for instance, by giving no timetable). Ideally, people should discount your predictions by exactly as much as the vagueness in the prediction justifies, but real humans don’t behave ideally.
I don’t see how that follows. If someone’s predictions are all nebulous, unfalsifiable claims, then we start out viewing them as a useless predictor, and never update our beliefs about them.
Plenty of people have said things akin to “there will eventually be a wolf attack”, and while they’re quick to laud themselves when one inevitably happens, they don’t generally get rewarded for this. People who make unlikely predictions and win? They get a spotlight.
Come to think of it, I’m surprised we don’t see a bit more clout being awarded to high-scoring polymarket predictors. A guy with a public record of making millions of dollars a year by predicting military outcomes that others bet against seems like he should get a slew of prestigious consultant positions, but I haven’t noticed it happening.