We think a quantitative forecast can help people to take a step back from obsessing over the latest daily poll release. By averaging polls together and arriving at a probabilistic forecast, the aim is to help people take a more nuanced view of the campaign, one that is less driven by whatever the latest polling number happens to be, and instead summarizes all of the available evidence, with the context of history.
But we humans are bad at understanding uncertainty. Reactions to Mr. Trump’s victory suggest that, despite our efforts, we failed at explaining that an 85 percent chance is not a 100 percent chance. If we did it all again, we would probably emphasize uncertainty in a more visceral way, rather than using a simple statement of probability, an analogy to a real-world event like an N.F.L. field goal or a shaded margin of error.
By more effectively communicating the uncertainty in the world, we’d hope to give our readers a more useful indicator of the state of the race — one that is built on data rather than anecdote and conjecture.
via www.nytimes.com
A mathematician friend of mine wasn't happy when there was a 93% chance of winning. He wanted 100%. He understands probabilities.
