2016 Postmortem
Related: About this forumFinal results from Princeton Election Consortium, 10/30/12
EV
Obama 305
Romney 233
Meta-margin
O+2.36
Odds of re-election
Random Drift 93%
Bayesian Prediction 98%
http://election.princeton.edu/
charlyvi
(6,537 posts)Are they reliable? Please be honest.
Alekei_Firebird
(320 posts)He's incredibly reliable.
charlyvi
(6,537 posts)But it seems too good to be true.
BlueDemKev
(3,003 posts)...even if we can't get 305 EV's, I'll be more than happy to settle for 280-290 EV's!
RandySF
(59,414 posts)AaronMayorga
(128 posts)Tutonic
(2,522 posts)n/t
Maraya1969
(22,507 posts)cthulu2016
(10,960 posts)regnaD kciN
(26,045 posts)...Sam called 2004 exactly. It's mentioned on his website. I think the person giving Kerry 75% was Zogby.
ProudToBeBlueInRhody
(16,399 posts)I know how he feels.
AaronMayorga
(128 posts)trend at the time (which he called an "error" . He takes about it in the site's FAQ.
http://election.princeton.edu/faq/
AaronMayorga
(128 posts)And, in '04, his model correctly predicted the outcome (Bush win / exact margin). He's been on NPR's Science Friday along with 538's Nate Silver, so yes, he's reliable (arguably more so than Mr. Silver).
doc03
(35,389 posts)regnaD kciN
(26,045 posts)...hence my putting the date in the subject line.
PEC updates six times per day, at 8 A.M., 10 A.M., noon, 3 P.M., 5 P.M., and 8 P.M. (all times Eastern).
doc03
(35,389 posts)elleng
(131,197 posts)2roos
(26 posts)A better approach than Nate Silver - Professor Wang doesn't bother with any data other than polls.
Don't worry about the math - just check his % numbers - and the graphs are easy to decipher.
StrongBad
(2,100 posts)What's interesting is that his model is much more simplistic than Nates but turns out just as good (and in some cases better) predictions.
freshwest
(53,661 posts)ProudProgressiveNow
(6,129 posts)TroyD
(4,551 posts)Does that refer to by how much Obama will win the popular vote?