Statistikhengst
Diamond Member
- Banned
- #1
The Rabbi made some outlandish claims on this thread:
2016 Election Prediction Thread Page 9 US Message Board - Political Discussion Forum
-and-
2016 Election Prediction Thread Page 9 US Message Board - Political Discussion Forum
-and-
namely, that Sarah Palin's VP candidacy in 2008 did not sink John McCain's ship.
I gave The Rabbi an opportunity to rethink this viewpoint THREE times, but he refused.
Unfortunately for him, The Rabbi has no idea how closely I followed the 2008 presidential election, including every single presidential poll (national and state level) for that year, just as I did in 2012. What he also does not know is that I did nightly battleground reports starting on 9/23/2008 and running every night until election night. Those battleground reports ran for 42 nights straight.
This is critical, for the second Palin set of interviews that was so disastrous happened on 9/24 and we can see the ripple effect of this interview very clearly.
So, some key dates to remember:
8/29/2008: Sarah Palin's candidacy announced by McCain at Wright State University, Fairborn (Dayton), OH, in the middle of the DEMOCRATIC National Convention, breaking with decency and convention.
9/11/2008: Sarah Palin's interview with Charlie Gibson is aired.
9/16/2008: The Stock Market melts-down.
9/24/2008: Sarah Palin's interviews with Katie Couric are aired.
Two more things to note:
1.) I did 2 week polling intervals to average polls, so the effects of any one event are likely to be seen STARTING circa 10-14 days after an event. And the full effect should be reflected in the averages between 14 and 21 days after an event. That's how polling aggregates work.
2.) Poll release dates are not the same as the polling dates themselves. A poll dated from 10/02 may have been conducted from 9/27-28, for instance. But rarely is a presidential poll released more than 7 days after the actual polling time frame. Most are released within 1-2 days after the polling actually took place.
This was the projected EV map as of September 1, 2008:
Statistikhengst s ELECTORAL POLITICS - 2013 and beyond Projected EV map - Obama vs. McCain based on Poll Convergence 7
Obama 275 / McCain 221 / Tossup 42
Compared to 9/22:
Statistikhengst s ELECTORAL POLITICS - 2013 and beyond Based on PC8 Obama 269 McCain 265 Tossup 4
Obama 269 / McCain 265 / Tossup 4
In all of the composites for 2008, this is the closest that McCain ever got to Obama: 265.
This was the projected EV map as of September 22, 2008, for the first three weeks of September. This map was created 6 days after the stock market crash, but two weeks after the first polling results from the announcement of Sarah Palin's VP candidacy started coming in. Remember, she was announced on 8/29 and the first polling results came in around 9/6, so, the two week time frame I mentioned is important here. And, polling based on the stock market crash, which most people did not really register yet, had yet to come in:
This seems counterintuitive, but remember, good polling averages are usually over two-week spans; this reduces a lot of statistical bumps. This high water mark for McCain represents his convention bounce, the pick of Palin, but before the public started to give input on the stock market crash and Palin's 2nd interview, which had not yet happened.
Now, for reference, here is the first battleground report, for 9/23, two days after the last EV map I reported (269/265):
Statistikhengst s ELECTORAL POLITICS - 2013 and beyond BATTLEGROUND STATISTICAL REPORT 09 23
On 9/23, IN, NV, FL, VA and OH were leaning slightly to McCain.
On, 9/28:
Statistikhengst s ELECTORAL POLITICS - 2013 and beyond Battleground State REPORT 09 28 compared to PC8 from 09 21.
the composite for Virginia moved from McCain to Obama and never looked back. That was 4 days after the disastrous Palin interview, but no real polling results had come out yet.
On 10/1/2008, 7 days after the Palin interview and after a lot of weekend polls came through, the battleground picture looked like this:
Statistikhengst s ELECTORAL POLITICS - 2013 and beyond BATTLEGROUND STATE REPORT 10 01
And the EV projection,which had been Obama 269/McCain 265/tossup 4 just nine days before, became:
Obama 311 / McCain 202 /tossups 25
And the tossups were OH and NV. In between 9/28 and 10/01, in the days following the Palin interview and also TWO weeks after the stock market crash of 9/16, OH, NH, CO and FL moved from McCain to Obama (though OH was so lean, I classified it as tossup).
So, 10/01 is turning point no. 1, exactly TWO WEEKS after the stock market crash, and by this time, all polls from before 9/16 (stock market crash) had been cycled out of the averages.
Two days later, on 10/03:
Statistikhengst s ELECTORAL POLITICS - 2013 and beyond BATTLEGROUND STATE REPORT 10 03
NV and NC moved from the McCain column to the Obama column and there was a major uptick for Obama in NH.
On 10/05, I put out a posting explaining my weighting method for all states, a method that I copied from a CONSERVATIVE electoral blogger with a very successful website.
10/07 was almost exactly TWO WEEKS after the Palin/Couric interviews were aired, and this battleground report therefore only included polls released AFTER the interviews:
Statistikhengst s ELECTORAL POLITICS - 2013 and beyond BATTLEGROUND STATE REPORT 10 07
EV projection as of 10/7:
Obama 353 / McCain 185 / tossup 11
(the tossup was Missouri)
On that day, the only state of all the states that Obama won in 2008 that was not yet in his column was: Indiana.
353 + 11 = 364, plus NE-02 (which was only polled once or twice as an individual congressional district) = 365.
On 10/11, I noted that the campaign was going so well for Obama that he opened more offices in Omaha, Nebraska (NE-02):
Statistikhengst s ELECTORAL POLITICS - 2013 and beyond WOW Obama really IS fighting for NE-02
On 10/16, I posted my "Poll Convergence" no. 10 - a complete analysis of all fifty states, not just the battlegrounds:
Statistikhengst s ELECTORAL POLITICS - 2013 and beyond Distillation of PC 10 Obama 349 McCain 174 Tossup 15
EV projection as of 10/16:
Obama 349 / McCain 185 / tossup 15
(the tossup was North Carolina
You can read every single battleground report and distillation of a poll convergence, it is all there for all to read.
There is no doubt that there were two major turning points in aggregate polling: October 1st, and October 7th/8th.
There were also specific polls about the Palin candidacy, specifically, about her competence:
1.) Skepticism of Palin Growing Poll Finds
And:
2.) Inside Obama s Sweeping Victory Pew Research Center
and:
3.) http://www.nytimes.com/2008/10/31/us/politics/31poll.html
and:
4.) Newsweek Poll, from 10/22-23:
Campaign 2008
Notice the date of 10/08 as well. Two weeks after the Couric interview.
Notice the value from 9/10. Quite a difference...
FOUR independent-from-each-other-pollsters, 3 weeks apart from each other: same result vis-a-vis Sarah Palin. The public decided that she was a real dud, the worst VP candidate since Eagleton 1972, just like JoeB131 said.
BTW, for those who want to see every single 2008 poll:
Statistikhengst s ELECTORAL POLITICS - 2013 and beyond FINAL POLL CONVERGENCE No. 12
Statistikhengst s ELECTORAL POLITICS - 2013 and beyond PC 12 - Distillation Obama 356 McCain 182 - however with adjustment Obama 375 McCain 163
--------------------------------------------------------------------------------------------------------------
So, The Rabbi made a claim and questioned my supposition that the Palin VP candidacy sank McCain. I have now presented a mound of data, all of which I collected and sorted through every single day of the Fall 2008 GE campaign. The numbers, and most importantly, the two dates (10/1, 10/7-8) clearly point to two major shifts in the polling averages, and the timing, including the two week polling average window, clearly points to:
a.) the stock market crash of 9/16
b.) Sarah Palin's second interview, from 9/24.
Words have consequences, and so did Sarah Palin's, or her lack of them, to be more specific. McCain probably would have lost, with or without Palin on the ticket, but no doubt her presence caused the losing margin to be larger and probably flipped IN and NC to Obama.
Not only that, polling specifically on the issue of her competence as a possible POTUS should anything have happened to John McCain, had he been elected, was unequivocally clear on this: 3/5 of Americans did not trust her to be able to do the job. There is no way your team can win an election with those kind of numbers.
2016 Election Prediction Thread Page 9 US Message Board - Political Discussion Forum
-and-
2016 Election Prediction Thread Page 9 US Message Board - Political Discussion Forum
-and-
namely, that Sarah Palin's VP candidacy in 2008 did not sink John McCain's ship.
I gave The Rabbi an opportunity to rethink this viewpoint THREE times, but he refused.
Unfortunately for him, The Rabbi has no idea how closely I followed the 2008 presidential election, including every single presidential poll (national and state level) for that year, just as I did in 2012. What he also does not know is that I did nightly battleground reports starting on 9/23/2008 and running every night until election night. Those battleground reports ran for 42 nights straight.
This is critical, for the second Palin set of interviews that was so disastrous happened on 9/24 and we can see the ripple effect of this interview very clearly.
So, some key dates to remember:
8/29/2008: Sarah Palin's candidacy announced by McCain at Wright State University, Fairborn (Dayton), OH, in the middle of the DEMOCRATIC National Convention, breaking with decency and convention.
9/11/2008: Sarah Palin's interview with Charlie Gibson is aired.
9/16/2008: The Stock Market melts-down.
9/24/2008: Sarah Palin's interviews with Katie Couric are aired.
Two more things to note:
1.) I did 2 week polling intervals to average polls, so the effects of any one event are likely to be seen STARTING circa 10-14 days after an event. And the full effect should be reflected in the averages between 14 and 21 days after an event. That's how polling aggregates work.
2.) Poll release dates are not the same as the polling dates themselves. A poll dated from 10/02 may have been conducted from 9/27-28, for instance. But rarely is a presidential poll released more than 7 days after the actual polling time frame. Most are released within 1-2 days after the polling actually took place.
This was the projected EV map as of September 1, 2008:
Statistikhengst s ELECTORAL POLITICS - 2013 and beyond Projected EV map - Obama vs. McCain based on Poll Convergence 7
Obama 275 / McCain 221 / Tossup 42
Compared to 9/22:
Statistikhengst s ELECTORAL POLITICS - 2013 and beyond Based on PC8 Obama 269 McCain 265 Tossup 4
Obama 269 / McCain 265 / Tossup 4
In all of the composites for 2008, this is the closest that McCain ever got to Obama: 265.
This was the projected EV map as of September 22, 2008, for the first three weeks of September. This map was created 6 days after the stock market crash, but two weeks after the first polling results from the announcement of Sarah Palin's VP candidacy started coming in. Remember, she was announced on 8/29 and the first polling results came in around 9/6, so, the two week time frame I mentioned is important here. And, polling based on the stock market crash, which most people did not really register yet, had yet to come in:
This seems counterintuitive, but remember, good polling averages are usually over two-week spans; this reduces a lot of statistical bumps. This high water mark for McCain represents his convention bounce, the pick of Palin, but before the public started to give input on the stock market crash and Palin's 2nd interview, which had not yet happened.
Now, for reference, here is the first battleground report, for 9/23, two days after the last EV map I reported (269/265):
Statistikhengst s ELECTORAL POLITICS - 2013 and beyond BATTLEGROUND STATISTICAL REPORT 09 23
On 9/23, IN, NV, FL, VA and OH were leaning slightly to McCain.
On, 9/28:
Statistikhengst s ELECTORAL POLITICS - 2013 and beyond Battleground State REPORT 09 28 compared to PC8 from 09 21.
the composite for Virginia moved from McCain to Obama and never looked back. That was 4 days after the disastrous Palin interview, but no real polling results had come out yet.
On 10/1/2008, 7 days after the Palin interview and after a lot of weekend polls came through, the battleground picture looked like this:
Statistikhengst s ELECTORAL POLITICS - 2013 and beyond BATTLEGROUND STATE REPORT 10 01
And the EV projection,which had been Obama 269/McCain 265/tossup 4 just nine days before, became:
Obama 311 / McCain 202 /tossups 25
And the tossups were OH and NV. In between 9/28 and 10/01, in the days following the Palin interview and also TWO weeks after the stock market crash of 9/16, OH, NH, CO and FL moved from McCain to Obama (though OH was so lean, I classified it as tossup).
So, 10/01 is turning point no. 1, exactly TWO WEEKS after the stock market crash, and by this time, all polls from before 9/16 (stock market crash) had been cycled out of the averages.
Two days later, on 10/03:
Statistikhengst s ELECTORAL POLITICS - 2013 and beyond BATTLEGROUND STATE REPORT 10 03
NV and NC moved from the McCain column to the Obama column and there was a major uptick for Obama in NH.
On 10/05, I put out a posting explaining my weighting method for all states, a method that I copied from a CONSERVATIVE electoral blogger with a very successful website.
10/07 was almost exactly TWO WEEKS after the Palin/Couric interviews were aired, and this battleground report therefore only included polls released AFTER the interviews:
Statistikhengst s ELECTORAL POLITICS - 2013 and beyond BATTLEGROUND STATE REPORT 10 07
EV projection as of 10/7:
Obama 353 / McCain 185 / tossup 11
(the tossup was Missouri)
On that day, the only state of all the states that Obama won in 2008 that was not yet in his column was: Indiana.
353 + 11 = 364, plus NE-02 (which was only polled once or twice as an individual congressional district) = 365.
On 10/11, I noted that the campaign was going so well for Obama that he opened more offices in Omaha, Nebraska (NE-02):
Statistikhengst s ELECTORAL POLITICS - 2013 and beyond WOW Obama really IS fighting for NE-02
On 10/16, I posted my "Poll Convergence" no. 10 - a complete analysis of all fifty states, not just the battlegrounds:
Statistikhengst s ELECTORAL POLITICS - 2013 and beyond Distillation of PC 10 Obama 349 McCain 174 Tossup 15
EV projection as of 10/16:
Obama 349 / McCain 185 / tossup 15
(the tossup was North Carolina
You can read every single battleground report and distillation of a poll convergence, it is all there for all to read.
There is no doubt that there were two major turning points in aggregate polling: October 1st, and October 7th/8th.
There were also specific polls about the Palin candidacy, specifically, about her competence:
1.) Skepticism of Palin Growing Poll Finds
And:
2.) Inside Obama s Sweeping Victory Pew Research Center
and:
3.) http://www.nytimes.com/2008/10/31/us/politics/31poll.html
and:
4.) Newsweek Poll, from 10/22-23:
Campaign 2008
Notice the date of 10/08 as well. Two weeks after the Couric interview.
Notice the value from 9/10. Quite a difference...
FOUR independent-from-each-other-pollsters, 3 weeks apart from each other: same result vis-a-vis Sarah Palin. The public decided that she was a real dud, the worst VP candidate since Eagleton 1972, just like JoeB131 said.
BTW, for those who want to see every single 2008 poll:
Statistikhengst s ELECTORAL POLITICS - 2013 and beyond FINAL POLL CONVERGENCE No. 12
Statistikhengst s ELECTORAL POLITICS - 2013 and beyond PC 12 - Distillation Obama 356 McCain 182 - however with adjustment Obama 375 McCain 163
--------------------------------------------------------------------------------------------------------------
So, The Rabbi made a claim and questioned my supposition that the Palin VP candidacy sank McCain. I have now presented a mound of data, all of which I collected and sorted through every single day of the Fall 2008 GE campaign. The numbers, and most importantly, the two dates (10/1, 10/7-8) clearly point to two major shifts in the polling averages, and the timing, including the two week polling average window, clearly points to:
a.) the stock market crash of 9/16
b.) Sarah Palin's second interview, from 9/24.
Words have consequences, and so did Sarah Palin's, or her lack of them, to be more specific. McCain probably would have lost, with or without Palin on the ticket, but no doubt her presence caused the losing margin to be larger and probably flipped IN and NC to Obama.
Not only that, polling specifically on the issue of her competence as a possible POTUS should anything have happened to John McCain, had he been elected, was unequivocally clear on this: 3/5 of Americans did not trust her to be able to do the job. There is no way your team can win an election with those kind of numbers.
Last edited:
