2016 Postmortem
Related: About this forumCNN poll predicted Romney and Obama would be tied in 2012
Warning about the CNN poll. Their last three election poll in 2012 had Romney at 49% and Obama at 49% http://www.realclearpolitics.com/epolls/2012/president/us/general_election_romney_vs_obama-1171.html
Obama ended up winning by 4%.
saltpoint
(50,986 posts)presidential race was that Romney, up to the last few hours of election night, was absolutely convinced he would win.
That may be why Karl Rove was spouting psychotic babble about missing Ohio votes all over cable, that is, even more virulently than he usually spouts psychotic babble.
It might be interesting to see what Romney's internals missed and see if the same variable applies to the 2012 CNN poll.
Dem2
(8,178 posts)and I won't give right-wing RCP any clicks - so what did this post inform me of?
Mass
(27,315 posts)Reality is that the Ohio race is way closer than it should be. The good news is that, with VA and PA way in Clinton camp, she does not need Ohio as much (though it would be good if she could win it).
It is close - just checked Huffpo average is Clinton +2.2 including this likely outlier poll.
Stand and Fight
(7,520 posts)Loki Liesmith
(4,602 posts)and you'll be fine.

molova
(543 posts)Are they both fine?
Loki Liesmith
(4,602 posts)They both capture a decent estimate of the center of the distribution.
And HuffPo imposes a smoothing on the data that RCP doesn't use, which probably makes it a bit better. But both still fine.
(and if you want to do even better, take the median of your polls)
Doctor Jack
(3,072 posts)Via his twitter
MFM008
(20,042 posts)A CNN poll tries to save his ass. Same thing happened after her convention spike.