2016 Postmortem
Related: About this forumToday's 538 2016 Primary Forecasts: Iowa-Hillary-82%, Bernie-18% / NH-Hillary-57%, Bernie-43%
http://projects.fivethirtyeight.com/election-2016/primary-forecast/iowa-democratic/dsc
(52,166 posts)but I have to say those are more pro Hillary than I think is warranted. I still think Sanders is a favorite to win New Hampshire even if Hillary wins Iowa in anything short of a blow out which I think is rather unlikely. I do think she is a favorite to win Iowa but I don't think the odds are 82/18 more like 60/40.
Alfresco
(1,698 posts)reformist2
(9,841 posts)Godhumor
(6,437 posts)As their model sees Iowa as a Clinton won, that influence spreads to NH. If Clinton loses Iowa, the model will not be nearly as bullish on her in NH.
As someone who makes models, I can say with confidence that they will be tinkering with the model levers as information comes in that shows what was predictive and what wasn't. Best thing to do is consider their new primary prediction model as being in beta but visible to the public.
Number23
(24,544 posts)sonofspy777
(360 posts)2016 the year Nate Silver no longer had relevance.
wilsonbooks
(972 posts)Fearless
(18,421 posts)DFW
(54,435 posts)I mean, even the most ardent Hillary supporter should view those numbers with suspicion. Any Sanders supporter in his or her right mind would certainly do so if the forecasts instead had those numbers in Bernie's favor..
Recursion
(56,582 posts)A race polling at 52/48 is about 80% likely to be won by the candidate polling at 52.
HereSince1628
(36,063 posts)I suspect Nate is killing his reputation for relying on 'just the facts'.
John Poet
(2,510 posts)out of his ass.
Rosa Luxemburg
(28,627 posts)Rosa Luxemburg
(28,627 posts)jfern
(5,204 posts)Also, how are Jindal's odds at winning Iowa? And obvious Trump crash and burned 4 months ago.