« previous | TPM CAFÉ READER POSTS HOME | next »
OBAMA's Amazing Bounce COntinues - Now Leads in VIIRGINIA!!
Obama continues to pile up the numbers as tonight's pre-Gore endorsement polls released by Rasmussen have Obama now AHEAD of McCain in VIRGINIA - and 12 points closer in KANSAS than a month ago.
The ramifications of these numbers is astounding - and speak to a potential blowout if they hold and spread as widely as they have in the past 2 weeks.
Pablano now has Obama at a 65.7% probability of victory in November - the highest for any candidate since he began his projections for 2008.
http://www.fivethirtyeight.com/
Advertisement
Comments (2)
Well do keep in mind the actual perspective of Poblano's numbers.
They are only statistical exercises, involving timelines etc. ie they're pretty esoteric. Hugely fun, but that's about it.
Quoting Poblano:
"I want to provide you, at any given moment in time, with the best possible projection of what's going to happen in the November election. This is inherently a forward-looking exercise. If what you're interested in instead is simply a summation of what the polls are telling you now, there are plenty of other websites that can provide that for you."
Thus his projections don't - they can't - account for what's going to happen between now and the election. eg They're formulated before the 527's start saturating the media. They're the sort of exercise that statisticians can get positively orgasmic about but they really tell you nothing about what's going to happen: they aren't predictive.
June 16, 2008 10:46 PM | Reply | Permalink
The Rasmussen Poll is not projective or predictive. As with all other polls, it provides an objective snapshot of the voting public at the time the poll is conducted - and also comparative information when the same sample is polled over time.
These substantial changes in which candidate received what percentage of the vote are hardly projective. They would appear to measure actual, tangible movement in the direction of the voting public over the past few weeks - since each shift is well outside the margin of error for the sample.
And Rasmussen is/was one of the 3 most accurate polling organizations in the primary season.
Now - the extrapolation of poll data into different outcome scenarios *IS* projective - and more open to potential error.
BUt in locating obvious trends, the data seem to indicate a definite and substantial movement to Obama in the past few weeks - and the same applies to nearly every poll taken since he became the definitive nominee.
But of course - as always - there is still a ways to go - though in 2004 the running poll averages did not deviate more than a point or 2 throughout the Summer and early Fall.
June 16, 2008 11:00 PM | Reply | Permalink
Post a Comment