“Am I actually answering the question?”
“No, because actually answering it requires knowledge of what the President plans to do in the next six months, how the media is likely to react to what he does in the next six months, some idea of the probability of outside events (terror attack, victory in war, natural disaster) that would influence his popularity, and knowledge of when the elections take place in America, none of which I actually have.”
So what question was I answering?
“What have you heard about Obama recently, with extra weight on any sources you remember as American.”
Good point. As well as “how do I feel about the president?” being the wrong question, historical trends for presidential popularity when entering an election year, etc., are probably reasonably good predictors. Presidential popularity isn’t anti-inductive like a market.
I was going to point out that in this case substituting the question was actually rational behavior. There is no way of knowing what is going to happen in the next six months to influence it, it could be positive or negative, but whatever does will adjust his popularity from whatever it is right now; so whatever his popularity is right now will probably be a better prediction of what it will be in six months than whatever fantasy either a supporter or opponent might come up with.
How popular will the President be in six months?
Not very.
“Am I actually answering the question?” “No, because actually answering it requires knowledge of what the President plans to do in the next six months, how the media is likely to react to what he does in the next six months, some idea of the probability of outside events (terror attack, victory in war, natural disaster) that would influence his popularity, and knowledge of when the elections take place in America, none of which I actually have.”
So what question was I answering?
“What have you heard about Obama recently, with extra weight on any sources you remember as American.”
This is a bad example because a major error is people taking too much of an inside view.
Good point. As well as “how do I feel about the president?” being the wrong question, historical trends for presidential popularity when entering an election year, etc., are probably reasonably good predictors. Presidential popularity isn’t anti-inductive like a market.
I was going to point out that in this case substituting the question was actually rational behavior. There is no way of knowing what is going to happen in the next six months to influence it, it could be positive or negative, but whatever does will adjust his popularity from whatever it is right now; so whatever his popularity is right now will probably be a better prediction of what it will be in six months than whatever fantasy either a supporter or opponent might come up with.