MASSIVE FRAUD! MY ANALYSIS VINDICATED BY A FORMER MIT MATH PROFESSOR

[TruthIsAll]

http://www.bluelemur.com/index.php?p=405

[leftchick]

1. I Choose Possible cause Number 4.....

<snip>

Anick reasons that there are four possible causes of the “Bush gains.” (1) Significantly greater lying or refusal to speak to pollsters in Bush voters versus Kerry voters; (2) Consistent/systematic errors in weighting demographic groups; (3) A surge of Bush voters after 4 p.m., in all states; (4) Systematic tampering/hacking of reported vote totals, in Bush’s favor.

In no state did Bush have a loss. Bush’s support in the reported vote tallies went up in every single state compared with the exit polling.

[TruthIsAll]

10. Compare results: Prof odds 1 in 50,000. Mine: 1 in 2,500 (but we agree)

Here is my 18 exit poll analysis:
http://www.democraticunderground.com/discuss/duboard.ph...

1. The MIT professor gives Bush a 1 in 50,000 (.002%) of gaining 4.15% in the vote over the 4pm exit polls.

2. My analysis gives Kerry a 99.96% probability of getting over 50% in the final vote (based on the 18 state average). That is equivalent to saying that there is a .04% probability (1 in 2,500) that Bush would get over 50% of the 2-party vote vs. Kerry.

There is virtually NO discrepancy in the results, since the professor calculated the probability of Bush gaining a higher percentage than I do in my calculation.

Therefore the professor's probability is lower than mine by a factor of 1/20.

However you slice it, the odds are minuscule that Bush won fairly.

The fact that the discrepancies were near zero in 10 states, and extremely high in IMPORTANT states like FL (optiscan, BBV) and OH (absentee/provisional ballots, spoiled punch cards and BBV) tells the story.

It's not JUST BBV. There are many ways to rig the election.

[TalkingDog]

2. I have NO maths

well, okay statistics....

'splain this to me like you would a 12 year old, so I can 'splain it to people on my mailing list.

Statistically significant means squat. Give me a day to day analogy. The chances of Bush gaining 4% in all the polls is like:

-my toaster suddenly producing money instead of toast.

-my cat bringing home 50 mice in one day rather than just the usual one.

Trust me, you help me out with this and someday I'll repay you when you desperately need the difference between Bauhausian Asbstraction and Neo-Expressionist Abstraction explained to you.

[ima_sinnic]

3. yes, you have the basic idea:

"statistically significant" expresses the chances of something happening or not happening by random chance. It is based on probabilities.


the election was a fraud

[Goldmund]

4. It's like...

...if you randomly threw an open bucket of paint up in the air, and the paint spilled on the sidewalk in just such a way to spell out "FUCK BUSH" without any intervention by you.

[TalkingDog]

7. That would be nice

>...if you randomly threw an open bucket of paint up in the air, and the paint spilled on the sidewalk in just such a way to spell out "FUCK BUSH" without any intervention by you.

But is that truly analogus? Because it would take billions of tries for anything approximating "Fuck Bush" to come out of a randomly thrown paint splash. Are you saying that the chances of the votes being skewed are literally billions to 1?

When these articles talk about statistical variance, they don't take us poor publically educated schlubs into account and give us the Powerball Lottery version of how likely it is to happen.

Thanks for the assist.

[Goldmund]

8. Oh, I didn't actually sit down and calculate it

And since the electoral data they're working with is not yet complete, they don't know yet either. When we have figured out the exact probability of exit polling being wrong so drastically in just such a way, then we'll be able to sit down and make comparisons with real-life situations like the one I gave -- and I think it's important that we do so, because as you point out, using statistics lingo to do it would be far less effective.

I just basically pulled the comparison out of my ass, and I'm not claiming that the respective probability of the two situations is numerically comparable. But once we figure everything out, we will need to make such comparisons, and point out that to believe all this is a mere coincidence is equivalent to believing that the "FUCK BUSH" grafitti you see on the sidewalk is just a result of a random paint splash -- or something similar to that.

[anamandujano]

9. The reality

is that the paint came down a spelled "f*ck Kerry and the USA".

[Goldmund]

5. In his second and third paragraph,

he says that he can't find any significant variance in the error between EV and analog machines. I thought there was a very significant variance.

[peacetalksforall]

6. There are now three threads for this article.