Tuesday, October 07, 2008


The Rasmussen Reports daily Presidential Tracking Poll for Tuesday shows
Barack Obama attracting 52% of the vote while John McCain earns 44%.
***This is the second straight day that Obama has led by eight
percentage points, his largest lead of the year. For the past twelve
days, Obama’s support has ranged from 50% to 52% while McCain has been
at 44% of 45% every day***


Obama is viewed favorably by 58% of voters, McCain by 52% (see trends).
Obama is viewed Very Favorably by 39% of voters and Very Unfavorably by
28%. For McCain, those numbers are 24% Very Favorable and 26% Very
Unfavorable.

New statewide poll results from Colorado, Florida, Missouri, Ohio, and
Virginia confirm the national trend favoring Obama.

Electoral College projections now show Obama leading 217-163. However,
when “leaners” are included, Obama leads 300-174. A total of 270
Electoral Votes are needed to win the White House.

Voters now trust Democrats more than Republicans on all ten key issues
tracked regularly by Rasmussen Reports. Other polling shows that voters
overwhelmingly believe the United States is the best nation in the
world. At the same time, 50% believe that other nation’s tend to dislike
the USA.


---

I'm hopeful but not confident.

On Oct 7, 9:56 am, Boater wrote:
Tuesday, October 07, 2008

The Rasmussen Reports daily Presidential Tracking Poll for Tuesday shows
Barack Obama attracting 52% of the vote while John McCain earns 44%.
***This is the second straight day that Obama has led by eight
percentage points, his largest lead of the year. For the past twelve
days, Obama’s support has ranged from 50% to 52% while McCain has been
at 44% of 45% every day***

Obama is viewed favorably by 58% of voters, McCain by 52% (see trends).
Obama is viewed Very Favorably by 39% of voters and Very Unfavorably by
28%. For McCain, those numbers are 24% Very Favorable and 26% Very
Unfavorable.

New statewide poll results from Colorado, Florida, Missouri, Ohio, and
Virginia confirm the national trend favoring Obama.

Electoral College projections now show Obama leading 217-163. However,
when “leaners” are included, Obama leads 300-174. A total of 270
Electoral Votes are needed to win the White House.

Voters now trust Democrats more than Republicans on all ten key issues
tracked regularly by Rasmussen Reports. Other polling shows that voters
overwhelmingly believe the United States is the best nation in the
world. At the same time, 50% believe that other nation’s tend to dislike
the USA.

---

I'm hopeful but not confident.

On Sept 17, 2004, Kerry led Bush by 54-41% among registered
independents. It is independents who will determine who wins in this
election.

On Oct 7, 2:20 pm, wrote:
On Oct 7, 9:56 am, Boater wrote:



Tuesday, October 07, 2008

The Rasmussen Reports daily Presidential Tracking Poll for Tuesday shows
Barack Obama attracting 52% of the vote while John McCain earns 44%.
***This is the second straight day that Obama has led by eight
percentage points, his largest lead of the year. For the past twelve
days, Obama’s support has ranged from 50% to 52% while McCain has been
at 44% of 45% every day***

Obama is viewed favorably by 58% of voters, McCain by 52% (see trends).
Obama is viewed Very Favorably by 39% of voters and Very Unfavorably by
28%. For McCain, those numbers are 24% Very Favorable and 26% Very
Unfavorable.

New statewide poll results from Colorado, Florida, Missouri, Ohio, and
Virginia confirm the national trend favoring Obama.

Electoral College projections now show Obama leading 217-163. However,
when “leaners” are included, Obama leads 300-174. A total of 270
Electoral Votes are needed to win the White House.

Voters now trust Democrats more than Republicans on all ten key issues
tracked regularly by Rasmussen Reports. Other polling shows that voters
overwhelmingly believe the United States is the best nation in the
world. At the same time, 50% believe that other nation’s tend to dislike
the USA.

---

I'm hopeful but not confident.

On Sept 17, 2004, Kerry led Bush by 54-41% among registered
independents. It is independents who will determine who wins in this
election.

Harry's original post got me thinking about polling uncertainty.
Apparently, the day of the 2004 election, the Zogby poll showed Kerry
winning the electoral vote by a wide margin. So , how is uncertainty
determined, amazingly, simply Poisson statistics; it is the ratio of
the square root of the number of counts to the total number of
counts. I deal with this type of uncertainty every day when I measure
x-ray spectra of things. You think that the uncertainty should simply
be this simple ratio but you do the experiment again and find its well
outside of this uncertainty, a sure sign something is amiss. This
happens to me ALL THE TIME and my numbers of counts are in the
hundreds of thousands while political polls deal with only a couple
thousands of counts, no wonder they have problems.
Often, if I can identify the source of error, I can greatly increase
the number of counts so that this source of error is smaller than the
statistical uncertainty but there is no way this can be done in
political polling.
What if, many McCain supporters are like me, curmudgeons who hang up
as soon as they realize it is a sales call or pollster. There is no
way that increasing the poll size would improve accuracy.
What if, McCain supporters are more affluent and tend to more often
use cell phones instead of having a land line and pollsters are
forbidden to call cell phones?
These two things are not uncertainties, but errors in the design of
the poll.
Actual uncertainty might be something like "How often do people say
they will vote for one candidate and then actually vote for another"?
We cannot even estimate this uncertainty and it is probably larger
than the 3% statistical uncertainty of a 1000 person poll.