Re:You cannot police misinformation in a fair, reproducible,...
by thaumasiotes on Tuesday February 08, 2022 @10:25PM
(#30268382)
> When polls predicted a 99% chance of Hillary Clinton winning, you and I and most of our peers know that the polls weren't wrong. Sometimes chance comes out that way.
And it sounds like you're confusing them here. "Sometimes chance comes out that way" is something that applies to aleatoric probabilities. But major elections are not uncertain in an aleatoric sense; they are uncertain in an epistemic sense, and the polls were correct to worry about why they were wrong.
Andrew Gelman just wrote about the difference between aleatoric and epistemic uncertainty: https://statmodeling.stat.columbia.edu/2022/02/03/epistemic-...
And it sounds like you're confusing them here. "Sometimes chance comes out that way" is something that applies to aleatoric probabilities. But major elections are not uncertain in an aleatoric sense; they are uncertain in an epistemic sense, and the polls were correct to worry about why they were wrong.