General Discussion
Related: Editorials & Other Articles, Issue Forums, Alliance Forums, Region ForumsReminder: Sam Wang is more accurate than Nate.
He is the most accurate, period.
http://www.businessinsider.com/meet-sam-wang-the-neuroscientist-who-beat-even-nate-silver-in-his-election-prediction-2012-11
Will never understand why everyone is so hung up on Nate.
aaaaaa5a
(4,667 posts)Accurate. There are no wild fluctuations. Almost all presidential races are steady with little change week to week and month to month.Wang understands this. His models are great.
gto
(24 posts)Wounded Bear
(58,656 posts)now, I think 538 worries more about sports than politics. He made some tweaks to his modeling this year IIRC, and the results seem to have made his output entirely too variable to be really acceptable IMHO. Perhaps long term he's still good, but he's had numerous times that the numbers seemed to change rapidly based on minimal changes IRL.
aaaaaa5a
(4,667 posts)There is no way HRC chances of winning dropped 20 points in a week. That many opinions don't change that fast to have that kind of impact on a presidential race.
Presidential races are steady. Almost all of the numbers and results are baked in and known before the race even begins. Nate's model is great entertainment, but it is to sensitive to be considered the best in the business.
cbdo2007
(9,213 posts)RAFisher
(466 posts)Silver had a perfect map in 2012. Wang had Florida going to Romney. Wang did correctly pick Obama's vote share however he had Obama winning by only 2.2%. Silver had Obama winning by 2.5%. Obama actually won by 3.9%. Wang predicted the mode and median EV but both Silver's predidction was closer to the actual result. I don't even know where that article got the figure that Wang predicted 332 EVs for Obama. His final prediction doesn't show that.
This is stupid to pretend that Wang always more accurate than Silver.
http://election.princeton.edu/2012/11/06/presidential-prediction-2012-final/
http://fivethirtyeight.blogs.nytimes.com/fivethirtyeights-2012-forecast/?_r=0