Exit Polls MATCHED Pre-Election polls within 1/7th of ONE percent

[TruthIsAll]

The final state polls entered into the Election model had
Kerry leading by ONE percent in the national weighted average:
47.88-46.89 
This is equivalent to Kerry getting 50.5% of the two-party
vote

Exit Polls entered into the model had Kerry leading
50.60%-49.34%. 
This is equivalent to Kerry getting 50.63% of the two party
vote. 

The difference between the two sets of polls is .13%.
This is LESS THAN 1/7 of ONE percent.

The FINAL Pre-election state polls from Zogby, ARG, et al
almost EXACTLY AGREED with the Exit polls.

That is, until the polls abruptly turned to Bush at 1AM, Nov
3.

	Final State Polls	            Exit Polls			
	Kerry%	Bush%	K/K+B	Kerry% Bush% K/K+B	Diff
AL	39	57	40.6%	41.00	59.00	41.0%	-0.37%
AK	30	57	34.5%	40.50	59.50	40.5%	-6.02%
AZ	45	50	47.4%	47.00	53.00	47.0%	0.37%
AR	48	48	50.0%	46.60	53.40	46.6%	3.40%
CA	49	42	53.8%	54.00	46.00	54.0%	-0.15%
							
CO	47	48	49.5%	49.10	50.90	49.1%	0.37%
CT	52	42	55.3%	58.50	41.50	58.5%	-3.18%
DE	45	38	54.2%	58.50	41.50	58.5%	-4.28%
DC	78	11	87.6%	91.00	9.00	91.0%	-3.36%
FL	50	47	51.5%	50.00	49.00	50.5%	1.04%
							
GA	42	52	44.7%	43.00	57.00	43.0%	1.68%
HI	45	45	50.0%	53.30	46.70	53.3%	-3.30%
ID	30	59	33.7%	33.50	66.50	33.5%	0.21%
IL	54	42	56.3%	57.00	43.00	57.0%	-0.75%
IN	39	58	40.2%	41.00	59.00	41.0%	-0.79%
							
IA	50	44	53.2%	50.65	49.35	50.7%	2.54%
KS	37	60	38.1%	35.00	65.00	35.0%	3.14%
KY	39	56	41.1%	41.00	59.00	41.0%	0.05%
LA	40	48	45.5%	44.50	55.50	44.5%	0.95%
ME	50	39	56.2%	54.75	45.25	54.8%	1.43%
							
MD	54	43	55.7%	57.00	43.00	57.0%	-1.33%
MA	64	27	70.3%	66.00	34.00	66.0%	4.33%
MI	52	45	53.6%	52.50	47.50	52.5%	1.11%
MN	52	44	54.2%	54.50	45.50	54.5%	-0.33%
MS	42	51	45.2%	43.25	56.75	43.3%	1.91%
							
MO	44	49	47.3%	47.50	52.50	47.5%	-0.19%
MT	36	57	38.7%	39.75	60.25	39.8%	-1.04%
NE	32	61	34.4%	36.75	63.25	36.8%	-2.34%
NV	49	49	50.0%	49.35	50.65	49.4%	0.65%
NH	47	47	50.0%	55.40	44.60	55.4%	-5.40%
							
NJ	50	42	54.3%	52.80	46.20	53.3%	1.01%
NM	49	49	50.0%	51.30	48.70	51.3%	-1.30%
NY	57	39	59.4%	63.00	37.00	63.0%	-3.63%
NC	46.7	50.3	48.1%	48.00	52.00	48.0%	0.14%
ND	35	55	38.9%	34.00	66.00	34.0%	4.89%
							
OH	50	47	51.5%	52.10	47.90	52.1%	-0.55%
OK	28	61	31.5%	35.00	65.00	35.0%	-3.54%
OR	50	44	53.2%	51.20	48.80	51.2%	1.99%
PA	50	45	52.6%	54.35	45.65	54.4%	-1.72%
RI	56	36	60.9%	64.00	36.00	64.0%	-3.13%
							
SC	42	55	43.3%	46.00	54.00	46.0%	-2.70%
SD	42	52	44.7%	37.75	62.25	37.8%	6.93%
TN	47.8	50.3	48.7%	41.50	58.50	41.5%	7.23%
TX	37	59	38.5%	37.00	63.00	37.0%	1.54%
UT	24	69	25.8%	30.50	69.50	30.5%	-4.69%
							
VT	53	40	57.0%	65.00	35.00	65.0%	-8.01%
VA	47	51	48.0%	45.40	54.10	45.6%	2.33%
WA	52	44	54.2%	54.95	45.05	55.0%	-0.78%
WV	45.8	48.6	48.5%	45.25	54.75	45.3%	3.27%
WI	51	44	53.7%	52.50	47.50	52.5%	1.18%
WY	29	65	30.9%	29.00	65.50	31.0%	1.50%
							
Avg	47.88	46.89	50.50%	50.60	49.34	50.63%	-0.13%

[althecat]

1. That Zogby point is well made TIA

Meanwhile I am not giving up on getting the full real data...

NEW WORK TEAM TASK THREAD - TO GET THE FULL EXIT POLL DATA

http://www.democraticunderground.com/discuss/duboard.php?az=view_all&address=203x58415

[TruthIsAll]

4. Althecat, my data was actually a MIX of the 4pm and 12AM polls

to be precise. I should be more careful next time.

But the exit polling numbers between 4PM and 12:20 AM are virtually the same.

[althecat]

6. I guessed that in retrospect.. it makes the quest for the actual data

.. even more important.

al

[goclark]

2. Yep! The mighty Voodoo Machines and Rove

[mhr]

3. Hi Truth Is All

One wonders if there is enough angst left in the American electorate to care any more.

I personally invested six months in the Frost vs Sessions campaign.

I personally watched the ES&S iVotronic machines used in Dallas County switch democratic votes to republican votes.

I am still unemployed and am looking at homelessness in six months.

I personally feel let down by the democratic leadership and have no more fight left in me.

Sometimes the old truths are the hardest to swallow:

"you can't fight city hall!"

[TruthIsAll]

5. Got to keep fighting. I have felt the same pain you are experiencing.

We will get through this.

The truth cannot be denied.

I still believe Hope is on the way.

[Stop Shrub]

17. its not over

you are wrong, you do have plenty of fight left in you thats why you are on these boards. We will prevail I can feel the tides changing...

[mhr]

18. All Of My Elected Representatives Are Republican

Talking with these folks is useless.

All of the local media is right-wing.

Talking with these folks is useless.

The bulk of the local populace is Fundie.

Talking with these folks is useless.

Short of open revolt and revolution there is not much that can be done locally.

[idiosyncratic]

7. I'm crying again . . . n/t

[Surikat]

8. That ought to tell you something...

...considering how reliable polls in general turn out to be most times.

Seriously, though I think your 1/7 of 1% number is a red herring. Just eyeballing your data it looks to me like you are not averaging the absolute values of your deltas. When you don't average the absolute values of the deltas you get the difference between the means of the polls and the exit polls, not the average difference between the individual polls and their corresponding exit polls. Those aren't the same two numbers by a long, long shot.

You might want to check that.

[TruthIsAll]

13. You are wrong. The plusses and minuses cancel.

We look for the mean (average). Period. We look for an equal dispersion, positive and negative.

If you went to the race track three times, winning 100, 200 and losing 100, what is your average win?

It's 200/3= 67 win average.

Its NOT (100+200+100)/3=133
You get THAT bogus result by changing your 100 loss to a 100 win.
No can do.

Did you ever study the Standard Error of the Mean?
It's NOT called the Standard Error of the Absolue Value Mean.

If I ask you to take an average of a set, do you ask for the sign of the numbers?

You need to take a course in statistics.

[Surikat]

15. You want to know the average difference between a...

...pre-election poll and its corresponding exit poll, not the difference between the average of all pre-election polls and all post-election polls.

You could have a set of polls where half of all of the pre-election polls were 40% for Kerry and their corresponding exit polls were 60% while the other half were 60% for pre-election polls and their corresponding exit polls were 40%. The mean of the pre-election polls would be 50% as would the mean of the exit polls. The difference between the means would be 0%.

The average difference between individual polls, however, would be 20%. That is a very meaningful difference while the difference between the means is useless.

It's one thing to know a bunch of rote statistics definitions and formulae. It's quite another to be able to sift meaning out of a particular universe of measurements.

Did YOU have a course in statistics?

[Epictetus]

9. You can not rely on Zogby Polls

You need to more research on the pre election day polls. If you compare them to the other polls done by survey USA, mason dixon, Rassmussen, as well as other independent polls, you would come up with results that were close to the actually vote totals. Zogby never did a good job with state polls as proved in 2000 and 2002. Mason Dixon is the gold star of state polls.

[Surikat]

10. bingo... Zogby really screwed up this time around...

[TruthIsAll]

11. Did you read the post. Zogby was right. The Exit Polls matched.

And the exit polls were right.

Until they were manipulated AFTER midnight to match the fraudulent votes.

You just don't (or won't) get it.

[Epictetus]

14. Zogby's Polls

were based on the exit poll information i was not based on telephone polls. How effective do you think polls calling poople at home during work hours are going to be. In order for them to make sense he would have had to track down his samples at work, without access to their phone numbers. The polls you should be comparing to for accuracy are the polls done on Friday before and Monday samples. They were the accurate ones. Even some of Zogby's state polls done previos to election day are more accurate.
I do get it, becuase I did my research you don't get it.
Zogby stuck his finger in the air after looking at the raw data from the exit polls.

[Sideways]

20. Wow What An Impressive Debut On Du Epic

Not only did you come off as strident, churlish, arrogant, inarticulate, and wholly uniformed you insulted one of Du's premiere members. If you last another day here I will be shocked I tell you shocked.

If Zogby stuck his finger in the air blah blah blah I'm not even going to venture where you stuck your finger.

Bye.

[TruthIsAll]

12. Mason Dixon is the GOP Gold Star of Pollsters.

They were way off, except for a few Red states.

[DELUSIONAL]

16. Kick

[TruthIsAll]

19. am kick

tia

[Sideways]

21. Kick

[TruthIsAll]

22. Exit Polls matched the pre-electon polls: Worth quite a few kicks

Is the new spin going to be that the final state pre-election polls were off also, beyond the MoE?