Sunday's Daily Trackers
Mark Blumenthal | October 19, 2008
Topics: Daily Trackers
A quick update on today's national tracking poll releases. First, the national trend now clearly shows a slight narrowing over the last week. The now smoothed trend line indicates that the Obama margin peaked at 7.8 points (50.1% to 42.3%) a week ago on October 12. As of this writing, it stands at 5.7 points (49.3% to 43.6%).
However, if we look just at the daily tracking polls since Thursday, the patterns across the six polls that report over the weekend are inconsistent. On today's releases, two (Gallup and Rasmussen) show a slightly wider margin for Obama, two (Reuters/Zogby and IBD/TIPP) show slightly smaller margins and two (Kos/Research200 and Diageo/Hotline) have unchanged margins. If we compare today's releases to Thursdays, the pattern is the same: three slightly improved for Obama, three slightly improved for McCain.
By Mark Blumenthal | October 19, 2008 9:04 PM | Permalink | Comments (4) | TrackBacks (0)
Comments
I'm just noting again that we seem to have hit a dry period for non-tracker national polls (if I am not mistaken, all twelve of the last polls in the national chart are trackers). That may or may not be making a difference, but I am personally interested to see what happens when (presumably) we get a new round of non-trackers.
Posted on October 19, 2008 9:19 PM
"The now smoothed trend line indicates that the Obama margin peaked at 7.8 points (50.1% to 42.3%) a week ago on October 12. As of this writing, it stands at 5.7 points (49.3% to 43.6%)."
That's not enough for McCain. He needs to accelerate this tightening, or he will end up just short on election day.
Posted on October 19, 2008 11:29 PM
There seems to be something wrong with the numbers reported by Gallup. I think at least the sample sizes are inconsistent.
They're reporting three sets of numbers, RV, LV traditional, and LV extended. Sample sizes should - for every single day and/or rolling average - be largest in RV, followed be LV extended; for LV traditional the sample sizes should be smallest. On the first few days this was the case.
But take a look at these numbers taken from the Gallup site:
10/14-16/08 (reported October 17, 2008)
RV: 2805
LV ext.: 2314
LV trad.: 2155
10/15-17/08 (reported October 18, 2008)
RV: 2796
LV ext.: 2263
LV trad.: 2572
10/16-18/08 (reported October 19, 2008)
RV: 2798
LV ext.: 2277
LV trad.: 2590
Suddenly the least inclusive LV model has a bigger size. This can't be right.
Posted on October 19, 2008 11:51 PM
"Suddenly the least inclusive LV model has a bigger size. This can't be right."
I would assume it's just a matter of getting those two numbers mixed up. That would explain everything with the least amount of change.
Posted on October 20, 2008 1:32 AM
Post a comment