Mitofsky on Nightline:"26 of 30 deviations to Bush" Wrong. It was 30 of 30

[TruthIsAll]

Mitofsky said 26 states deviated to Bush, 4 to Kerry. Wrong. 
 
30 states deviated to Bush beyond 2.0%, 23 beyond the MOE.
Not ONE state deviated to Kerry beyond 2.0% or the MOE.

Probability: Less than .00000 00000 00000 00000 00000 00000

The odds: 
  ************* IMPOSSIBLE (div by zero) *************



Overall, 43 states deviated to Bush (incl. DC)
The odds: 
     *********** 1 in 2.9 million *****************


Mitofsky doesn't even know his own numbers.

He should look here:

ToBush:  Sample	Kerry  %Dev	VoteDev
NY	1,452	63.0%	-4.61%	-340,551
FL	2,846	50.5%	-3.38%	-256,646
PA	1,930	54.3%	-3.39%	-195,347
OH	1,963	52.1%	-3.39%	-190,916
NC	2,167	48.0%	-4.41%	-154,467

IL	1,392	57.0%	-2.18%	-114,922
MA	889	66.0%	-3.91%	-113,738
MN	2,178	54.5%	-3.30%	-93,256
WI	2,223	52.5%	-2.74%	-81,894
SC	1,735	46.0%	-5.04%	-81,480

AL	730	41.0%	-4.14%	-77,899
VA	1,431	48.0%	-2.40%	-76,544
NJ	1,520	55.0%	-2.05%	-74,027
CT	872	58.5%	-4.19%	-66,085
MI	2,452	52.5%	-1.25%	-60,673

WA	2,123	54.9%	-2.09%	-59,705
AZ	1,859	47.0%	-2.60%	-52,391
GA	1,536	43.0%	-1.59%	-52,331
CO	2,515	49.1%	-2.02%	-42,881
IN	926	41.0%	-1.74%	-42,843

LA	1,669	44.5%	-2.19%	-42,449
UT	798	30.5%	-4.47%	-41,384
MO	2,158	47.5%	-1.40%	-38,227
MS	798	43.3%	-3.04%	-34,593
NH	1,849	55.4%	-5.05%	-34,119

NE	785	36.8%	-4.04%	-31,386
MD	1,000	57.0%	-1.03%	-24,499
KY	1,034	41.0%	-1.31%	-23,570
AR	1,402	46.6%	-2.05%	-21,638
IA	2,502	50.6%	-1.36%	-20,509

ID	559	33.5%	-3.24%	-19,358
RI	809	64.0%	-4.43%	-19,303
DE	770	58.5%	-5.14%	-19,276
VT	685	65.0%	-5.78%	-17,981
NM	1,951	51.30%	-2.25%	-17,042

WV	1,722	45.3%	-2.05%	-15,528
AK	910	40.5%	-4.89%	-15,257
NV	2,116	49.4%	-1.26%	-10,397
ME	1,968	54.7%	-1.17%	-8,682
OK	1,539	35.0%	-0.57%	-8,349

MT	640	39.8%	-1.20%	-5,384
WY	684	30.9%	-1.77%	-4,295
DC	795	91.0%	-1.57%	-3,555

ToKerry:
SD	1,495	37.8%	0.68%	2,645
HI	499	53.3%	0.71%	3,044

ND	649	34.0%	1.50%	4,692
OR	1,064	51.2%	0.40%	7,316
KS	654	35.0%	1.62%	19,283
TN	1,774	41.5%	1.03%	24,992
CA	1,919	54.0%	0.43%	53,746

TX	1,671	37.0%	1.23%	90,930



 
  

[roenyc]

1. he has been lying so much

he cant keep them straight.

[TruthIsAll]

29. DISREGARD THIS POST: PLEASE LINK HERE FOR CORRECTION

http://www.democraticunderground.com/discuss/duboard.php?az=view_all&address=203x300746

[fasttense]

2. Now that Mitofsky admits his poll was in error

Do you think he will give the media conglomerate it's ten million back? I'm sure they didn't contract him for flawed data.

[PROGRESSIVE1]

3. kick

[skylarmae]

4. but will they set the record straight tonight

on the program. Seems to me that the truth is what will set us free here. Just keep pounding the truth at them until they really understand what happened here and before. IMO it has been going on since 1996. Does the phrase, 'stuck on stupid' ring true here? If so, just take a page from Ronald Reagan's play book. Say something enough times to the masses, and they will believe it. That's exactly how we have gotten into this mess. Now its time to set the record straight.

[TruthIsAll]

16. Nightline mentioned 1000 to 1 odds . Much too low.

Its ZERO probability for 20 states over MOE for Bush.

Its 1 in 3 million for 43 states to deviate by any amount to Bush.

[itzamirakul]

5. OK, then let's take some action

Copy and paste your opinion in an email and send it to the tv station and to the show, itself. We have to start showing some quiet outrage about these lies they are sending into our homes.

[TruthIsAll]

6. Go ahead. You can do it. n/t

.

[itzamirakul]

15. OK...I will...

I did not see the program but I will write in to complain today.

It would help if a lot of us would start to do that regularly.

[fooj]

18. I agree. Please make sure Koppel gets a copy!

You know what? I bet Randi Rhodes will give this story legs...should someone send her info, as well? Maybe a letter to the editor @ NYT and Washington Post? I think we should break this bullshit WIDE OPEN.......

[EFerrari]

7. Anyone ever think someone has threatened Mitofsky? n/t

[dandrhesse]

21. anyone and anything in the way of the greed machine has likely been

threatened. Given that fact, I would say yes there is a good chance that has occurred.

[qanda]

8. TIA, I follow your work

And trust me when I say that I generally have no clue about all of this. But I was wondering if the PA, FL, and OH percentages are extra weird. What are the chances that those particular states would deviate by the same amount or within .01% of each other?

[TruthIsAll]

11. I noticed that. The probability they would be so close? Very, very low.

.

[Blue in the face]

13. Hey TIA

Where did the exit poll numbers that you cited come from?

[Name removed]

9. Deleted message

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

[TruthIsAll]

14. There is a BIG difference between 26-4 and 30-0.

If you can find where my numbers are wrong, please point it out.

[dandrhesse]

22. right you are, think about that. you know tons more than me about the

numbers and their significance. I would say he is counting on folks like you to know the significance of picking 26.

[Name removed]

28. Deleted message

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

[Blue in the face]

10. Where are these numbers coming from again?

[TruthIsAll]

12. 1) Final state tallies and 2) Simon 11/03 12:22 am exit poll download

Both converted to 2-party percentages.

% Deviations = 1 - 2

[Bill Bored]

17. I like the ranking by vote count shift.

It's a nice compromise between the EV ranking and the percentage ranking. A lot of information in a small space.

Anyway, do you feel like looking at the party affiliations by state in the pre- and post- adjustment reports? Is this data available? It is for the regional results.

[TruthIsAll]

19. I don't know if its available, but if so, perhaps I will look at it. n/t

.

[Pendulum]

20. Picky correction

First, TIA, I really like what you do. Especially when you bring the numbers to life with pictures or coin flip analogies.

I have a suggestion for how to report .00000 00000 00000 00000 00000 00000

If you got this by multiplying a bunch of probabilities together, it doesn't mean that the probability is zero. It means that the number is smaller than can be represented with those many digits. So, instead of saying probability zero with "impossible" odds, you could say probability less than .00000 00000 00000 00000 00000 00001 with odds less than 1 in 1 000 000 000 000 000 000 000 000 000 000. These odds are less than 1 in a thousand million billion trillion.

According to the Powerball website at

http://www.powerball.com/powerball/pb_prizes.asp

it is far more likely that I win the powerball jackpot three days in a row, buying just one ticket each day.

Not impossible, but in a way it's even less possible than impossible!

[TruthIsAll]

23. You are ABSOLUTELY correct. But to me, 1.0E-30 is effectively ZERO

Of course it'S NOT be ABSOLUTE ZERO.

NOTHING IS.

But, since Excel gives an error when I compute the probability using in the BINOMDIST function, I have NO IDEA how small the number is.

What is 1.0E+30?

1.0E+6 = 1 MILLION
1.0E+9 = 1 BILLION
1.0E+12= 1 TRILLION
1.0E+24= 1 TRILLION TRILLON
X= 1.0E+30 = 1 BILLION TRILLION TRILLION

CAN'T WE JUST SAY 1/X = ZERO?

[Pobeka]

24. We could say pigs would have wings first.

Seriously. If pigs had wings they could escape captivity and the odds are they would evolve this capability before this many states would swing that far in one direction.

[yurbud]

25. link to source of this data?

I need it to do a graphic, and I want to make sure I have a respectable footnote.

[TruthIsAll]

27. CORRECTION: 15 states deviated to Bush beyond the MOE

Disregard this analysis. It is incorrect.

State Exit Poll Data:
Jonathan Simon download 12:22 am Nov.3 (google for link).

National vote data by state:
Again, google for link to final voting stats. Check CNN, MSNBC, etc.

I use two-party percentages only: Bush+Gore = 100%

Here is the link to the corrected results:
http://www.democraticunderground.com/discuss/duboard.php?az=view_all&address=203x300746

[sepia_steel]

26. Thanks TIA!

Your work always impresses me, esp. since I'm not a numbers girl.

[understandinglife]

30. a TBO;24/7 kick (nt)