Comparing my 2-party exit poll/vote percentage data to Freeman's

[TruthIsAll]

My published vote totals are slightly more current than
Freeman's.

His average exit poll deviation is higher by 0.09% (1.67%
vs.1.578%).
His average absolute value deviation is higher by 0.164%
(2.248% vs. 2.084%)

His average Kerry 2-party exit poll is 47.630%
My average Kerry 2-party exit poll is 47.634% 
The roundoff effects appear to cancel each other out.

His average Kerry 2-party vote is 45.961%
My average Kerry 2-party vote is 46.056% 

Both data sets show that 43 states moved from the exit polls
to the vote in favor of Bush. Only MT and CA moved in opposite
directions. 

KP= Kerry 2-party poll%						
KV= Kerry 2-party vote%	
KDev= KV - KP					
						
Avg	46.056	45.961	47.634	47.630	-1.578 -1.670
	KV    KV      KP      KP    KDev	KDev
St	TIA   Frmn    TIA    Frmn     TIA	Frmn
AL	37.102	37.078	41.000	41.080	-3.90	-4.00
AK	36.774	36.170	40.500	40.141	-3.73	-3.97
AR	45.067	44.740	46.600	46.931	-1.53	-2.19
AZ	44.725	45.027	47.000	46.596	-2.28	-1.57
CA	55.041	55.206	54.000	55.732	1.04	-0.53

CO	47.634	47.349	49.100	49.072	-1.47	-1.72
CT	55.275	55.237	58.500	58.474	-3.23	-3.24
DE	53.832	90.632	58.500	91.633	-4.67	-1.00
DC	90.520	53.817	91.000	58.439	-0.48	-4.62
FL	47.476	47.469	50.505	49.930	-3.03	-2.46

GA	41.646	41.582	43.000	43.110	-1.35	-1.53
HI	54.404	54.373	53.300	53.320	1.10	1.05
ID	30.677	30.707	33.500	33.333	-2.82	-2.63
IL	55.205	54.993	57.000	57.131	-1.79	-2.14
IN	39.577	39.456	41.000	40.970	-1.42	-1.51

IA	49.663	49.544	50.649	50.673	-0.99	-1.13
KS	37.133	36.970	35.000	34.602	2.13	2.37
KY	39.992	39.994	41.000	40.755	-1.01	-0.76
LA	42.669	42.626	44.500	44.495	-1.83	-1.87
ME	54.583	54.479	54.745	54.834	-0.16	-0.35

MD	56.566	56.249	57.000	57.038	-0.43	-0.79
MA	62.743	62.701	66.000	66.463	-3.26	-3.76
MI	51.726	51.734	52.500	52.551	-0.77	-0.82
MN	51.761	51.764	54.500	54.612	-2.74	-2.85
MS	40.495	39.906	43.257	43.203	-2.76	-3.30

MO	46.380	46.327	47.500	47.475	-1.12	-1.15
MT	39.500	39.505	39.760	39.282	-0.26	0.22
NE	33.153	32.533	36.763	36.541	-3.61	-4.01
NV	48.682	48.666	49.351	50.660	-0.67	-1.99
NH	50.690	50.684	55.400	55.495	-4.71	-4.81

NJ	53.371	53.127	55.000	56.126	-1.63	-3.00
NM	49.600	49.416	51.300	51.343	-1.70	-1.93
NY	59.288	58.793	63.000	63.965	-3.71	-5.17
NC	43.758	43.715	48.000	47.312	-4.24	-3.60
ND	36.091	36.086	34.000	33.577	2.09	2.51

OH	48.939	48.749	52.100	52.060	-3.16	-3.31
OK	34.430	34.438	35.000	34.727	-0.57	-0.29
OR	52.109	51.969	51.200	51.223	0.91	0.75
PA	51.258	51.130	54.346	54.408	-3.09	-3.28
RI	60.578	60.482	64.000	64.237	-3.42	-3.75

SC	41.365	41.305	46.000	45.785	-4.64	-4.48
SD	39.087	39.088	37.762	37.417	1.32	1.67
TN	42.815	42.780	41.500	41.153	1.31	1.63
TX	38.490	38.490	37.000	36.838	1.49	1.65
UT	26.654	27.061	30.500	29.930	-3.85	-2.87

VT	60.301	60.336	65.000	65.686	-4.70	-5.35
VA	45.866	45.649	47.959	47.959	-2.09	-2.31
WA	53.647	53.598	54.945	55.066	-1.30	-1.47
WV	43.520	43.477	45.255	45.194	-1.73	-1.72
WI	50.192	50.199	52.500	50.214	-2.31	-0.02
WY	29.687	29.704	30.900	32.072	-1.21	-2.37

Avg	46.056	45.961	47.634	47.630	-1.578 -1.670
						

[bleever]

1. Thanks for the update. 43 states.

Since the "response rate" theory of exit poll skew shows up stillborn in Mitofsky's report, based on their own data, what will they fall back on?

Thanks again.

[TruthIsAll]

16. A probability analysis based on the Freeman numbers.

Here is a probability analysis of vote deviations, based on
the exit poll margin of error, using Freeman's exit poll and
vote data. Exit poll and vote percentages are adjusted to the
2-party equivalent.

Freeman's data:
42 states deviated to Bush, 16 exceeded the margin of error.
Prob = 1 - NORMDIST(15,50,.025,TRUE)
Odds = 1/Prob = 1 in 19 trillion.

My data: 
42 states deviated to Bush, 15 exceeded the margin of error.
Prob = 1 - NORMDIST(14,50,.025,TRUE)
Odds = 1/Prob = 1 in 1 trillion.

N    = exit poll sample size
MOE  = margin of error = 1/sqrt(N)
Std  = standard deviation = MOE/1.96
Exit = Kerry 2-party exit poll%
Vote = Kerry 2-party vote%
Dev  = Vote - Exit
Prob = probability of deviation

										
	N	MOE	StD	Exit	Vote	Dev	Prob  	 1 in	Favor	>MoE?
AK	910	3.31%	1.69%	40.141	36.17	-3.97%	0.9%	106	Bush	yes
AL	730	3.70%	1.89%	41.08	37.078	-4.00%	1.7%	59	Bush	yes
AR	1402	2.67%	1.36%	46.931	44.74	-2.19%	5.4%	19	Bush	
AZ	1859	2.32%	1.18%	46.596	45.027	-1.57%	9.2%	11	Bush	
CA	1919	2.28%	1.16%	55.732	55.206	-0.53%	32.6%	3	Bush	

CO	2515	1.99%	1.02%	49.072	47.349	-1.72%	4.5%	22	Bush	
CT	872	3.39%	1.73%	58.474	55.237	-3.24%	3.0%	33	Bush	
DC	795	3.55%	1.81%	91.633	90.632	-1.00%	29.0%	3	Bush	
DE	770	3.60%	1.84%	58.439	53.817	-4.62%	0.6%	167	Bush	yes
FL	2846	1.87%	0.96%	49.93	47.469	-2.46%	0.5%	199	Bush	yes

GA	1536	2.55%	1.30%	43.11	41.582	-1.53%	12.0%	8	Bush	
HI	499	4.48%	2.28%	53.32	54.373	1.05%	32.2%	3	Kerry	
IA	2502	2.00%	1.02%	50.673	49.544	-1.13%	13.4%	7	Bush	
ID	559	4.23%	2.16%	33.333	30.707	-2.63%	11.2%	9	Bush	
IL	1392	2.68%	1.37%	57.131	54.993	-2.14%	5.9%	17	Bush	

IN	926	3.29%	1.68%	40.97	39.456	-1.51%	18.3%	5	Bush	
KS	654	3.91%	2.00%	34.602	36.97	2.37%	11.8%	9	Kerry	
KY	1034	3.11%	1.59%	40.755	39.994	-0.76%	31.6%	3	Bush	
LA	1669	2.45%	1.25%	44.495	42.626	-1.87%	6.7%	15	Bush	
MA	889	3.35%	1.71%	66.463	62.701	-3.76%	1.4%	72	Bush	yes

MD	1000	3.16%	1.61%	57.038	56.249	-0.79%	31.2%	3	Bush	
ME	1968	2.25%	1.15%	54.834	54.479	-0.36%	37.9%	3	Bush	
MI	2452	2.02%	1.03%	52.551	51.734	-0.82%	21.4%	5	Bush	
MN	2178	2.14%	1.09%	54.612	51.764	-2.85%	0.5%	218	Bush	yes
MO	2158	2.15%	1.10%	47.475	46.327	-1.15%	14.8%	7	Bush	

MS	798	3.54%	1.81%	43.203	39.906	-3.30%	3.4%	29	Bush	
MT	640	3.95%	2.02%	39.282	39.505	0.22%	45.6%	2	Kerry	
NC	2167	2.15%	1.10%	47.312	43.715	-3.60%	0.1%	1939	Bush	yes
ND	649	3.93%	2.00%	33.577	36.086	2.51%	10.5%	10	Kerry	
NE	785	3.57%	1.82%	36.541	32.533	-4.01%	1.4%	72	Bush	yes

NH	1849	2.33%	1.19%	55.495	50.684	-4.81%	0.0%	39823	Bush	yes
NJ	1520	2.56%	1.31%	56.126	53.127	-3.00%	1.1%	91	Bush	yes
NM	1951	2.26%	1.16%	51.343	49.416	-1.93%	4.8%	21	Bush	
NV	2116	2.17%	1.11%	50.66	48.666	-1.99%	3.6%	28	Bush	
NY	1452	2.62%	1.34%	63.965	58.793	-5.17%	0.0%	17833	Bush	yes

OH	1963	2.26%	1.15%	52.06	48.749	-3.31%	0.2%	495	Bush	yes
OK	1539	2.55%	1.30%	34.727	34.438	-0.29%	41.2%	2	Bush	
OR	1064	3.07%	1.56%	51.223	51.969	0.75%	31.7%	3	Kerry	
PA	1930	2.28%	1.16%	54.408	51.13	-3.28%	0.2%	420	Bush	yes
RI	809	3.52%	1.79%	64.237	60.482	-3.76%	1.8%	55	Bush	yes

SC	1735	2.40%	1.22%	45.785	41.305	-4.48%	0.0%	7851	Bush	yes
SD	1495	2.59%	1.32%	37.417	39.088	1.67%	10.3%	10	Kerry	
TN	1774	2.37%	1.21%	41.153	42.78	1.63%	9.0%	11	Kerry	
TX	1671	2.45%	1.25%	36.838	38.49	1.65%	9.3%	11	Kerry	
UT	798	3.54%	1.81%	29.93	27.061	-2.87%	5.6%	18	Bush	

VA	1431	2.64%	1.35%	47.959	45.649	-2.31%	4.3%	23	Bush	
VT	685	3.82%	1.95%	65.686	60.336	-5.35%	0.3%	330	Bush	yes
WA	2123	2.17%	1.11%	55.066	53.598	-1.47%	9.2%	11	Bush	
WI	2223	2.12%	1.08%	50.214	50.199	-0.02%	49.4%	2	Bush	
WV	1722	2.41%	1.23%	45.194	43.477	-1.72%	8.1%	12	Bush	
WY	684	3.82%	1.95%	32.072	29.704	-2.37%	11.2%	9	Bush	
										
Avg	1,443	2.85%	1.46%	48.840	47.002	-1.84%	10.34%	10		16

[TruthIsAll]

17. Correction to the averages

KV	  KV	     KP    KP
TIA	  Frmn    TIA    Frmn
47.093 47.002	48.818 48.840

[TruthIsAll]

2. Slight fix: DC and DE numbers switched. Averages not affected.

DE 53.832 53.817 58.500 58.439 -4.67 -4.62
DC 90.520 90.632 91.000 91.633 -0.48 -1.00

[garybeck]

13. do I need to fix it on the article I published:

http://www.solarbus.org/stealyourelection/articles/0130-exitpolls.html

also wondering if you've figured out why your numbers are sligthly different than freemans? Is it just because you used the official certified count?

thanks

[TruthIsAll]

15. No, you have the latest - 15 states over the MOE, 43 to Bush n/t

You can check the latest count.
My vote totals were slightly higher.

Virtually no difference, but Kerry's % improved slightly with the later count.

[Wilms]

3. kick n/t

[indianaleft]

4. undersampling exurbs

TruthisAll, as the resident exit poll guru of the forum, if not the world, what do you think of the theory that the Mitofsky folks simply didn't believe that there would be 4-5 million new evangelical voters and that they therefore didn't sample enough precincts in the "exurbs" and similar "megachurch" areas where it appears that Bush won millions of new votes?

If that theory is true, it would account for a virtually nationwide miscount in the exit polls, right? If you can debunk it, then I'm afraid we have nothing to fall back on but fraud.

[libertypirate]

5. You need some proof to make that statement...

Although the corporate liberal media oxymoron has entered that as conjecture they don't offer any proof, neither has anyone else.

There has been not one proof driven explanation of why Bush won, find someone’s statement that can be backed up by actual data findings and I'll give you a buffalo nickel.

lp

[NationalEnquirer]

6. Thats a crock.

The whole premise of the evangelical vote is a crock.
Something made up by Rove, and put into the media, where it got legs.
Even the exit polling about "moral values" is a crock, and has been debunked BY CONSERVATIVES no less.

[FogerRox]

7. Evangelical vote here in NJ? Really? LOL LOL

[indianaleft]

8. saying it's a crock doesn't prove it,

nor does flaming, shouting, capitalizing, etc. I doubt the evangelical theory myself, and think TruthIsAll and others have conclusively proven that something is very awry with these exit polls. I'm just say ing that until we disprove the theories on which the GOP and the MSM are depending, we haven't won the battle.

Perhaps a fine-level examination of voting turnout in the exurbs and Bush % in 2000 there vs. 2004 would do the trick; perhaps this has been done somewhere? Has it?

By the way, if there are leftists in Indiana (and there are), I'm betting that there are evangelicals in New Jersey. They may not be of the good ol' snake handling, Southern variety, but there's a whole bunch of these northern megachurch types around most metropolitan areas, and they're probably in NJ, too.

[Joe Chi Minh]

9. No, the onus is on you

to prove that what you contend is not mere conjecture.

There are very strong arguments for suggesting the reverse, i.e. that there are many more true Evangelical Christians who would have recoiled in horror at the very thought of voting for Bush, since the Christian Gospels and Epistles are so fiercely hostile to the priorities of the rich man and his acolytes.

The Republicans routinely dismiss reasoned hypotheses as not amounting to hard evidence, and here are you suggesting that the case for Republican fraud (not just the Democrat case for Republican fraud), is dead in the water unless your conjecture based solely on hearsay is refuted - no doubt with "hard evidence". You people have got some nerve...!

We had a troll here the other day who could have been the good Kenneth, himself, the way he was insinuating that the scandalous lack of voting machines in some heavily-Democratic polling places was due to no more than poor logistics, rather than malicious partisan inertia; and some other equally cockameeny evasion - which escapes me now, it was so outlandish.

[indianaleft]

10. reasoned hypothesis=proof?

I'm not suggesting that the disparity between the exit polls and the tabulation doesn't demonstrate pretty strongly to me, and to most on DU, that cheating happened.

But look: it's a reasoned hypothesis that cheating happened; isn't an equally reasoned hypothesis that the exit polls were off due to an error by Mitofsky in choosing which precincts to sample? (If he'd sampled only farmers, for example, the EP's would have shown Bush winning with 75% of the vote instead of 50.5%. So it's critical that the sampling accurately reflect the national voting profile.

Having said that, I realize that the election day voting sample was very large, and that Mitofsky et al are supposed to be experts at this, and that they should have been ready, both for increased youth turnout for Kerry as well as for increased evangelical turnout for Bush -- since those were the announced strategies by both camps. So you are probably right -- they didn't make a sampling mistake like the one I proposed. But it would be nice to find out. The best clue that they didn't make such a mistake is that surely Mitofsky would have said so if he had. But on the other hand, he would look ridiculous and never work in this or any town again, so he is better off not explaining.

As to burden of proof, until they inaugurate Kerry, I'm afraid that, fair or not, the onus is on us.

Not everyone who disagrees with you is a troll. I am new at posting, but have been reading DU since last summer.

I just want to know whether our democracy is certainly stolen, or only possibly so.

[TruthIsAll]

11. So, by your logic, in order not to look silly, Mitofsky ....

would rather claim there was a Reluctant Bush Responder bias.
In fact, the data in the report shows just the opposite.

The onus is on them to explain not only the discrepancy's, but why they hypothesized a theory which was not only bogus, but totally at variance with the data in the report.

[Name removed]

14. Deleted message

Message removed by moderator. Click here to review the message board rules.

[NationalEnquirer]

12. Its a crock because it is based on BS invented by ROVE!

Last election cycle, he made up the story about the evangelicals staying home.
Well him or some other republican.
But probably rove was involved anyways.

[indianaleft]

18. You're right about the precinct selection -- I'm wrong

Had a chance to look at the Mitofsky report more closely and realize that Mitof still claims that his precinct sampling was flawless. Since this was obviously the moment to admit that he undersampled the evangelical vote -- which clearly surprised everyone (if it was there) -- I'm thinking that you're right and I'm wrong. Given that Mitofsky claims his sampling was right and tries to pin it on untrained exit pollers and overenthusiastic Kerry voters -- which is just a lie -- it's clear that the rational explanations have all fallen by the wayside.

What I meant about burden of proof is that, unfortunately, 80% of the people and 95% of the US Congress seem to believe the election wasn't stolen; thus, Bush is President, legally or not. Somehow, though we lack the economic and political clout, we have to do things that will turn those percentages around. perhaps it's not so much a burden of proof as a burden of persuasion, but the task is there before us.

Everyone keep digging. The rats will be found.

[FogerRox]

19. Recent poll says 30% think it was stolen-20% is an older poll