Re:House panel looking into Reddit post about Clinton's emai...
by skylan_q on Monday September 19, 2016 @09:12PM
(#12536207)
Don't use that site for any insight. Nate Silver had Trump at a "1%" chance of winning the GOP primaries. He then admitted that he got the call wrong because he really didn't have any models and was just spitballing with friends. His accuracy in the previous general election was based on the fact that he had visibility into private polling (WAY more accurate) and other forecasters didn't.
[ Parent
| on HN ]
Re:House panel looking into Reddit post about Clinton's emai...
by jpfed on Monday September 19, 2016 @10:24PM
(#12536549)
He did have models, and they correctly predicted Trump's victory. He chose to disregard them. Then he wrote a mea culpa article explaining his mistake, which he is unlikely to repeat.
Re:House panel looking into Reddit post about Clinton's emai...
by jedberg on Monday September 19, 2016 @09:23PM
(#12536259)
> His accuracy in the previous general election was based on the fact that he had visibility into private polling (WAY more accurate) and other forecasters didn't.
Where did you hear that? I'm not doubting you I'm just interested because I've never heard this before. My understanding was that he was so accurate last time around because the election followed normal patterns and his math was just better than everyone else's gut.
Re:House panel looking into Reddit post about Clinton's emai...
by tomjakubowski on Monday September 19, 2016 @09:26PM
(#12536283)
> he had visibility into private polling
Could you elaborate? Is "private polling" the same as "internal polling," which is done by the campaigns? If so, my understanding was that the Romney camp's internal polls in 2012 turned out to be extremely inaccurate.