Posted on 10/27/2012 8:57:02 PM PDT by Jet Jaguar
The final Gallup 2008 pre-election poll -- based on Oct. 31-Nov. 2 Gallup Poll Daily tracking -- shows Barack Obama with a 53% to 42% advantage over John McCain among likely voters. When undecided voters are allocated proportionately to the two candidates to better approximate the actual vote, the estimate becomes 55% for Obama to 44% for McCain.
The trend data clearly show Obama ending the campaign with an upward movement in support, with eight to 11 percentage point leads among likely voters in Gallup's last four reports of data extending back to Oct. 28. Obama's final leads among both registered voters and likely voters are the largest of the campaign.
Gallup's final estimate is based on Gallup's traditional likely voter model, and assumes an estimated turnout of 64% of the voting age population, an increase over 2004. (Gallup estimates voter turnout from the results of key voter turnout questions, using a model that compares how respondents' answers to these questions have related, historically, to actual turnout.) This year's higher turnout estimate is fueled by a surge in early voting -- 28% of registered voters in the final poll indicated they had already voted -- and higher turnout among blacks than in any of the last four presidential elections.
Gallup has been calculating and reporting an expanded likely voter model alongside its traditional model over the past month. In the end, the candidate selection estimates of the two models converged, and both show the same unallocated 53% to 42% margin for Obama.
(Excerpt) Read more at gallup.com ...
We can nitpick over the percentages, but Gallup has correctly picked the winner in every Presidential election since FDR with the exceptions of Truman-Dewey and Ford-Carter.
I think they’ll have this one right too.
OK, what is your point? This is so 2012....
Which is not what usually happens -- usually last minute indies break heavily for the challenger to an incumbent.
At what date prior to the election?
It overestimated Obama’s final winning percentage by two points - it was 53% to 46% in 2008.
this means nothing without the actual final numbers from the actual vote.
My point is Obama is nowhere near 54% this time in any polling. He is floating or dropping from 46%.
All they say is it was their “final survey” - so I’m guessing a few days before? Yeah, it’s early but I can’t see Obama closing the gap Gallup has been predicting and taking the lead.
Here’s the actual data:
http://www.gallup.com/poll/9442/election-polls-accuracy-record-presidential-elections.aspx
Another thing - you’ll notice that the final Gallup estimate nearly always underestimates the Republican candidate and overestimates the Democrat, in those races where there is a sizable error. To me, this would seem to indicate a consistent oversampling of Democrats in their samples. But overall, the Republican candidate has done better than the final Gallup poll estimated he would do.
Gallup and Rasmussen appear to be the pollsters with the best reputations and best track records over time (Gallup dates back to 1936!) and thus have the most to lose by reporting bogus data.
Media controlled pollsters, such as the dying dinosaurs NYT and CBS, continue to do all they can to portray this as a close race that is still winnable by Obama but as pollsters, they just don't have any respect or credibility.
If history is a guide, the numbers that Gallup and Rasmussen are reporting next week at this time, will reflect within a point or two, the actual results of the Nov 6 election.
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.