Re:Fake news was illegal in 17th century colonial Massachusetts
by jonbarker on Tuesday May 08, 2018 @04:45PM
(#17024866)
Disclosure: I am a Nate Silver fan. Nate's expert opinion is only valuable if you accept the premise, which is that prediction science involves a range of possible outcomes. If you accept that premise, which I do, his opinion on the 2016 election, in which he stated several times that the odds of a Trump presidency were actually higher than most of the polls suggested, was arguably a more valuable opinion than either a coin toss, or most of the individual polls.
[ Parent
| on HN ]
Re:Fake news was illegal in 17th century colonial Massachusetts
by null0pointer on Wednesday May 09, 2018 @03:30AM
(#17028012)
Yeah this seems to be the disconnect. A lot of people will intuitively think (myself included) something like this. "You said Hillary had 90% chance to win, but Trump won. Therefore your prediction was WRONG". It actually takes a little more thought to understand that a Trump victory was a possibility in the model all along. And that you cannot say whether or not a prediction model is accurate based on a SINGLE result.