Comments on: Real Election Fraud

By: ” The Machine “ « YouViewed/Editorial

” The Machine “ « YouViewed/Editorial — Fri, 02 Nov 2012 10:52:16 +0000

[...] Real Election Fraud (politicalirony.com) [...]

By: Arthanyel

Arthanyel — Thu, 01 Nov 2012 18:25:23 +0000

More disturbing news – Republicans in Ohio installing “emergency software patches” on voting machines, refuse to allow independent examination or explain why they are needed:

http://www.freepress.org/departments/display/19/2012/4766

By: Dan

Dan — Thu, 01 Nov 2012 00:17:51 +0000

Analyzing the digits of numbers, given certain independence assumptions, is a well-defined statistic. I didn’t see anything that well defined in this document.

As I said, the main problem is that it’s hard to critique because it’s so poorly laid out. If this kind of scientific argument came to me to review I would reject it on the basis of incoherence.

By: Peter

Peter — Wed, 31 Oct 2012 22:51:03 +0000

Nah. Donald Duck would be silly. You need to pick a candidate that’s actually on the ballot. Go with the Green party or something like that. I think they’re probably in all 50 states.

By: Dan

Dan — Wed, 31 Oct 2012 20:06:12 +0000

I always insist on leaving a paper trail.

By: Iron Knee

Iron Knee — Wed, 31 Oct 2012 17:40:07 +0000

I agree with Arthanyel. This is like when the IRS analyzes dollar amounts on a tax return. For real numbers, the first digit of each number is much more likely to be “1″ than other numbers. But in made up numbers, all digits are equally likely.

Having smooth distributions is a good sign that there is something wrong with the votes.

So even if this is not “proof” of election fraud, it is a damn good reason to insist on a paper trail so the votes can be verified. How can anyone disagree with that?

By: Arthanyel

Arthanyel — Wed, 31 Oct 2012 17:08:15 +0000

Dr Dan – I haven’t had a chance to re-do their analysis, although I plan to do so. Here is the issue that concerned me – one of my primary jobs is consulting on patterns in unstructured data. While your point about adjusting for ethnicity etc. sounds reasonable, in reality there are few voting precincts that are homogeneous.

And since that is the case, what I would expect to see is that there is some statistical variance, precinct to precinct, and that while an overall trend might fit a curve, there would be substantial outliers. For example, some densely populated precincts will be predominantly college students, some will be minority poor, and some will be affluent white – and they should show VARYING results.

But the data looks like they don’t – it looks like a fairly linear trend with no significant outliers. And that just shouldn’t be what we see.

Not to mention the (now deceased) Republican operative Michael Connell in Ohio that had said he had gimmicked the software to do something just like this, and shortly thereafter (and just before his testimony) died in a plane crash.

I agree the analysis doesn’t look professional – but the evenness and lack of outliers in the pattern disturbs me.

By: Michael

Michael — Wed, 31 Oct 2012 04:19:23 +0000

I think the arguments are offered in the way of brib–ahem–campaign contributions to the secretaries of state involved.