Febble's case explodes with a simple scenario....

[Rex_Goodheart]

Let's suppose an election consists of two districts, Richland and Poorland, and two candidates, Bush and Kerry.

Richland has 100 voters, Poorland has 100 voters.

An exit poll was taken. In Richland, the pollsters interviewed ten voters, and the exit poll showed 6 in favor of Bush, and 4 in favor of Kerry. In Poorland, the pollsters interviewed ten voters, and the exit poll showed 3 in favor of Bush and 7 in favor of Kerry. So, the total exit poll indicates a 110-90 win for Kerry.

In the meantime, Mr. Tabulator is rigging the voting machines. In Richland, he gives Mr. Bush a 61-39 win, and in Poorland he gives Mr. Bush a 40-60 loss. Mr. Bush officially wins the election, 101-99.

However, the voters actually cast their votes as follows: 59-41 for Bush in Richland, and 67-33 for Kerry in Poorland, in what would have amounted to a 108-92 Kerry victory, just as the exit poll suggested.

But Febble doesn't know about that last part. Instead, she sees something significant in the fact that Bush's vote was understated by only 1% in Richland, but by a whopping 10% in Poorland! So, she goes about searching for explanations, and comes up with "we'd expect greater WPE's where the precincts are poorer and more chaotic, and that's just what we see" and other such drivel. Never mind that there is no historical basis, whatsoever, that supports a contention that Bush voters become more reluctant under poorer conditions.

And, sadly, never mind that the true reason for the variances (note: not "errors" as Mitofksy and Febble have taken to calling them in Orwellian fashion) are much simpler: the election was corrupted. And though Febble claims innocence, and perhaps she IS only fallacious in an unintentional manner, she is nevertheless fallacious, and her specious argumentation only gives encouragement to more cheating in the future.

[autorank]

1. Words of wisdom. Words of caution. Welcome to DU!!!

"her specious argumentation only gives encouragement to more cheating in the future."

Elegant, in the scientific sense, and accurate to a tee! Thank you for sharing this.

[liam_laddie]

2. UK exit polls

Let's hope that the UK results track very closely to vote results.
If they do, it will vindicate all of us who believe the polling here
in Nov. told the truth. Unless * and Rove interfere there...

[progressivejazz]

3. The problem is...

Febble has no way of knowing that last part in the real election.

And neither do you.

The scenario is just something you made up. It hardly explodes her case any more than a different made up scenario reinforces it.

[Land Shark]

4. progressivejazz, you are the one that hasn't "exploded" anything

for purposes of a hypothetical (which is, by definition, made up of facts that are not required to be knowable), one certainly can assume facts that can not be truly nailed down under most circumstances in "real life".

Hypotheticals like the one above can show that a particular hypothesis (such as that forwarded by Febble) "prove too much", that is, they tend to apply with equal force both where they ought to and where they ought not to....

By positing knowledge of the "actual" vote count, the hypothetical proves that Febble will be highly likely to dismiss out of hand a set of facts that establishes ACTUAL FRAUD. And that's the key. Febble's argument provides cover for actual fraud (provided it is described accurately in the above hypothetical, which I have not tried to verify, nor have I verified the internal validity of the hypothetical, rather i am saying that one most certainly can legitimately "make up" facts for the purpose of thinking experiments.)

[Rex_Goodheart]

5. No, progressivejazz...

That's not a problem at all. Febble's fallacy doesn't rest or fall on anybody's actual knowledge of the true intent of the voter. Even if nobody knows the true votes she will have claimed post hoc significance for something that has no probative value because it has multiple possible meanings, including corruption, as I just demonstrated. My scenario doesn't have to be an actual case; it only has to be a potential case... and there was great potential for corruption.