1 of 187 mil: Odds 7 of 18 states deviate beyond exit poll MOE to Bush

[TruthIsAll]

THE CLINCHER. SPIN THIS, PETER JENNINGS!

Another probability calculation which confirms the odds of
today's earlier post of 1 out of 250 million.

Well, not quite. It's 1 out of 187 million. 

Here's the calculation.

Let's calculate the odds that 7 out of 18 battleground states
would deviate OUTSIDE the margin of error and ALL go to Bush.

These are the Kerry exit poll and actual vote margins. 

Exit Poll	Actual BushGain
MO 	-8	-7	-1
CO	-7	-6	-1
AZ	-10	-10	0
MI	3	3	0
AR	-9	-9	0
IL	11	11	0
LA	-14	-15	1
IA	0	-1	1
NM	2	0	2
NV	-1	-3	2
ME	11	8	3

WI	5	1	4
OH	2	-2	4
PA	7	2	5
FL	1	-5	6
MN	10	3	7
NC	-4	-13	9
NH	17	1	16
			
			
			
Battleground States: 18	
Prob that a given state is outside the MOE and deviates to
Bush:2.50%	

Total number of states outside the MOE which deviated to
Bush: 7	

The calculation requires the BINOMIAL distribution:			

 Probability (all 7 states outside MOE go to  Bush)
= 1-BINOMDIST(7,18,0.025,TRUE)

= 0.000000533%	

or 1 out of	187,453,436	
		
 
 

[Perky]

1. Funny things you can do with stats

in the two months leading up to the election I did analysis of all the polling data available on electoral vote.com

I was specifically looking tor momentum after the debates and in the last week as the undecideds made up their minds.


I looks specifically at the 18 states which had the possibility of swing one way or the other. the other 32 and DC were clearly red or blue.

For the last two weeks leading up to the election I looked at the most recent ten polls in each of the states... I did this every single day with out fail.... I tracked the trend

Truth be told no one had any momentum everyday it was the same thing both candidates were polling between 46 and 47 points in the aggregate Kerry was alway ahead but there was no decided break in the last 7 days of the polling.

I had things at 4 to5% undecided in several of these states and the difference between the candidate in any given state ranged from .1 % to about 4.5% (basically smaller the the MoE.

I like a lot of people though the undecided 4% would break 3-1 to Kerry and Kerry as a result would win OHIO and FL which were dead heats.


Guess what for the last 4 days of the campaign I had Bush in minuscule leads in every state he won not including the undecideds. I had Kerry ahead in all the state he wound up winning again by very small margin.

Only one state flipped from Blue to Red and that was IOWA.. Basically the major pollsters in the aggregate nailed the results.

The only thin that did not happen is that the undecided apparently went heavier for Bush than was normative.

Bottom line it the Pollster nailed the election results and the exit polling apparently did not.

a lot of the exit polls were reporting results by like noon... I wonder if Democrats vote early and that is why they were so off.

I can share the spreadsheet with any one who would like it....realiz that there were over 900 polls between September 1 and November 1. so it winds up being a 2 meg file

Email here on DU and I will be happy to send you the file.

[Awsi Dooger]

2. TruthIsAll, this is GARBAGE; you don't understand margin for error

Sorry to be so blunt, but it's accurate and therefore this mathematical exercise is worthless. Try finding a clincher that's not so easily spun into the toilet.

MOE is applied to EACH number, not just the MARGIN between the two candidates. You are making the most basic of mistakes, something network anchors and talking heads fall prey to repeatedly.

Let's look at this example, a hypothetical state poll, let's say Florida:

Kerry 50%
Bush 49%

Margin for error 4%

Now let's say Bush wins 52-47%. That is NOT outside the margin for error, despite the 6 point gap. The MOE is applied to EACH of the two numbers. Therefore, Kerry's support is expected to range from 46-54% and Bush from 45-53%. That is where the 95% confidence level kicks in. It is statistically more likely Kerry will be ahead by 1, and so on, than down 4 or 5 or 6, but all fall within the MOE.

Don't take my word for it. Look it up on the internet. Here is the first link in a search I just did: http://momentlinger.typepad.com/momentlingeron/2004/08/understanding_p.html

Besides using the most favorable exit polls for Kerry, like the absurd 57-41 edge in New Hampshire, and even if your initial use of MOE had been correct, you are also understating margin for error among state polls. You are obviously using 3%, based on where Ohio and Wisconsin fit on your list. The margins differed by 4%, yet are listed outside the margin for error on your list. The company that does nationwide exit polls states very clearly on their website that state exit polls include a 4% margin for error, not 3%: http://www.exit-poll.net/faq.html#a15

"What is the Margin of Error for an exit poll?
Every number estimated from a sample may depart from the official vote count. The difference between a sample result and the number one would get if everyone who cast a vote was interviewed in exactly the same way is called the sampling error. That does not mean the sample result is wrong. Instead, it refers to the potential error due to sampling. The margin of error for a 95% confidence interval is about +/- 3% for a typical characteristic from the national exit poll and +/-4% for a typical state exit poll. Characteristics that are more concentrated in a few polling places, such as race, have larger sampling errors. Other nonsampling factors may increase the total error."

[Zomby Woof]

3. BUSH IS TOAST!

No wait... he isn't.

And yes, I am saying "I told you so".

[Name removed]

4. Deleted message

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

[Tweed]

5. Yeah take a poll to see how many people think another DUer is a troll

That sounds like something that isn't against the rules. :eyes:

You might want to tone down your comments T&A. It's against the rules to suggest that someone else is a troll. If you think that someone is from the Free Republic, alert the moderators. Don't turn the DU into some kind of Salem, MA.

[Awsi Dooger]

7. He sometimes makes ludicruous mathematical assertions then goes off on me

For having the audacity to point some of them out. Believe me, I've restrained more times than I should have. Especially regarding the 2002 midterm.

I must be the worlds worst freeper. In my early 40s and voted straight ticket Democratic every time. But I'm not going to sit idle while misrepresentations are made. Right now we should be looking forward. TIA has considerable math skills that he abuses to pretend we won when we did not.

[TruthIsAll]

12. Go ahead. Just try to refute the Excel Binomial Distribution

BINOMDIST
See Also

Returns the individual term binomial distribution probability. Use BINOMDIST in problems with a fixed number of tests or trials, when the outcomes of any trial are only success or failure, when trials are independent, and when the probability of success is constant throughout the experiment. For example, BINOMDIST can calculate the probability that two of the next three babies born are male.

Syntax

BINOMDIST(number_s,trials,probability_s,cumulative)

Number_s is the number of successes in trials.

Trials is the number of independent trials.

Probability_s is the probability of success on each trial.

Cumulative is a logical value that determines the form of the function. If cumulative is TRUE, then BINOMDIST returns the cumulative distribution function, which is the probability that there are at most number_s successes; if FALSE, it returns the probability mass function, which is the probability that there are number_s successes.

[TruthIsAll]

15. You have a lot of nerve questioning my integrity.

You have an agenda. I have the truth.

[Heath.Hunnicutt]

6. My poll response: chill. go to bed.

TiA, Awsie had an OK point about regressing from those MoE numbers. Also, it won't hurt your final result that much. So wake up tomorrow and re-run them.

As for statistics, Awsie has 1000+ posts. If Awsie is trolling, I am confused as to why.

Peace,
Heath

[Awsi Dooger]

8. That's right, my beef was his claim they were outside the MOE

When that is not correct for most of them. I also don't like accepting the 57-41 number in New Hampshire as legit, when anyone who follows politics at all understands that state never hreatened to favor Kerry by even 1/3 that much.

Very true, Heath. An analysis that evaluates how far the numbers differed from base would provide a huge number TIA could post in headlines. Margin of error doesn't need to be mentioned, because he got it wrong.

[Donkeyboy75]

9. Also, if there is a fundamental error in the way the polls were conducted

that shifted 3 points too far for Kerry, the numbers aren't strange. First of all, the odd calculation is flawed. Second of all, these calculations assume that sampling of voters is truly random...which isn't necessarily true with exit polls.

[TruthIsAll]

10. The odds calculation is flawed? Prove it.You're an amateur talking to pros

Youi are so naive. These people voted. It's better than a pre-election poll. Fot he 100th time, exit polls are extremely acurate, despite what you hear on CNN.

I also question your knowledge of statistics. You are an amateur talking to professionals.

tia

[Donkeyboy75]

11. Really? You're a professional statistician?

The odds calculation has been dealt with elsewhere on this thread.

The exit polling results are NOT necessarily extremely accurate. If, for instance, the pollers oversampled in urban areas, or they oversampled at times democrats are more likely to vote, there will be a polling area that carries through all polls. Get it? These numbers, while intriguing, do NOT prove voter fraud.

[TruthIsAll]

13. Why is it Exit polls are only wrong when Bush is running?

Or when the Senate is at stake (2002)?
Go argue with the TWO professsors whose analyses agree with mine.

tia

[Donkeyboy75]

14. Strawman, buddy.

We're talking about this election, but I can answer that question for you. What if Democrats are simply more likely to talk to exit pollers? Hmm? That would certainly explain 2002.

I'm not saying something funny didn't happen with the vote totals. But these numbers are in no way, shape or form evidence that vote tampering occured.

I have a feeling that many of the DUers using this as evidence would argue the other way: if Kerry had won, but exit polls showed Bush had won, the same people would be whining that the exit polls were rigged to favor Bush and cause unrest.

[TruthIsAll]

16. Democrats more likely to talk? You're reaching, buddy.

That's about the worst argument out there. Total BS and you know it.

Jeez, you are really desperate, aren't you?

[Donkeyboy75]

17. Yeah...

good response. I'll take the myriad reasons things could go wrong with exit polling over junk science supporting a tin foil hat conspiracy. And I'm desperate. :eyes:

[ArkDem]

19. Pretty Funny!! Below is my suggestion for a vacation for you

after all this election 'work'.

http://www.hillmanwonders.com/nile_river_cruise/nile_river_cruise.htm

[Donkeyboy75]

18. By the way, where is the link showing the analysis of two

professors agrees with yours?

[TruthIsAll]

20. Here are BOTH of them. And there is much more out there.

THIS PROFESSOR AT THE UNIV. OF PENN. WRITES THAT THE ODDS ARE 250 MILLION TO ONE THAT THE ELECTION REEKS.

http://www.democraticunderground.com/discuss/duboard.php?az=view_all&address=203x36855

I did a similar calculation. In mine, the odds are ONE out of 187 million that exit polls in 7 out of 18 Battleground states would ALL shift beyond the MoE to Bush in the actuals.



And the other one from a former MIT professor(only 50,000 to 1)
http://www.bluelemur.com/index.php?p=405

[Donkeyboy75]

22. Your first three links didn't work.

But I did read the last one, and I found this:

The chance given in that article is 1 in 50000. Where do you get the other number?

David Anick gives multiple possible reasons for this anomaly...which I also have described elsewhere in this thread.

This variance is not at a level where the legitimacy would be challenged by international election observers.



But go ahead, pick and choose only those things you want to hear...and convince yourself the election was hacked. It's much easier than thinking of ways to actually win elections for our side.

[Nederland]

21. Doesn't this assume...

...that the exit poll sample is an accurate picture of the voting population?

[Nederland]

23. Here's why the exit polls were skewed:

http://www.msnbc.msn.com/id/5445086/

1) The exit pollsters asked voters to answer a questionnaire that had not one or two questions... but 30. So most people in a hurry were not going to stick around. (And you can imagine the challenge for somebody with "energetic" toddlers or little children.) This could overstate the turnout of "single" voters (a group that broke towards Kerry) and understate "married with children voters" (a group that broke towards President Bush.)

2) The exit pollsters were clearly identifiable, through logos plastered on their clipboards and logos on the questionnaire, as representing a consortium of the major broadcast news organizations: (NBC, CBS, ABC, Fox, and CNN.) The logos of the big three, NBC, CBS, and ABC are far more well known than the logo for Fox. My point is that voters who are suspicious of the major broadcast networks (whether those suspicions are justified or not) usually aren't interested in helping us do anything. Who are these voters? They tend to be Evangelical Christians and other conservatives. And if these voters are shying away from the network pollsters, the exit numbers are going to underestimate the president's support.

And I'll add my own:

3) The exit polls reportedly consisted of samples that were 59% women. This is obviously inaccurate and benefits Kerry significantly. Until we get the raw exit poll data however, we will never know if the samples were distorted or not.

[tommcintyre]

25. 59% women? More women vote in swing states?

The discrepancies were greatest against Kerry in the swing states.

You and others, keep setting up this "straw man" of needing the raw data. It would be nice, but "you don't need to be a weatherman to know which way the wind blows" (or the meteorological data)

[Merlin]

28. Not buyin' it. They've been doing exit polls for 60 years-- accurately.

What suddenly happened to change that in 2000-04? Did they forget? I don't think so.

Evangelicals distrust network news? I thought they loved FAUX.

Exit pollsters asked 30 questions, and that made people impatient. Well it's not like they're doing anything they haven't done before.

The ruse about oversampling women voters doesn't even pass the laugh test. The majority (51%) of white women voted for Bush.

[Merlin]

29. dupe

.

[tommcintyre]

24. kick

[Name removed]

26. Deleted message

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

[The Flaming Red Head]

27. kick