Reply #171: you ought to back off

[OnTheOtherHand]

You are expressing opinions about exit polls, and you do not evince much knowledge of exit polls, nor much willingness to think about exit polls.

You made the rather strange claim that the VNS data "correctly predicted the missing votes in Broward, Duval and Palm Beach Counties in FL." You have never substantiated that claim, and I do not understand how you ever could.

It seems to rest on your recollection of what "the VNS guy on election night" said "an hour or so after FL was uncalled for Gore." Probably, what you heard someone explaining was (as in the report I linked to) that the vote count from Duval had been misentered -- as in the Konner et al. report, p. 13, which states, "(Later scrutiny determined that a VNS keypunch operator had entered incorrect vote-count data, which had the effect of making it appear that Gore had won 98 per cent of the Duval County vote tabulated up to that time.)" (My emphasis.) Obviously ballot spoilage in Duval would not have had that effect.

Regardless, the exit poll was too small reliably to detect spoilage of 50,000 votes in two or three individual counties. If someone from VNS really did go on TV around 11 or midnight to say any such thing, he was gibbering -- and, if so, I think more people would have noticed. So, do you have any other reason to believe your assertion?

(Hey, if you remember what network you were watching, maybe we can get a transcript and try to sort this out.)

Have you noticed, in fact, that you are pretty much ignoring the content of our arguments about exit polls? You have a lot to say about what you know (and believe) about the election, and what you think you heard some VNS guy say, and that is about it.

You are welcome to believe whatever you want about Gore's winning margin in Florida, and you may indeed be able to make the case. (NB: if I took your suggestion to "multiply by several thousand," I would end at least in the hundreds of millions, which perhaps we can agree to rule out.) But if you want to claim that the exit poll supports your view, you have to make a specific, substantiated argument to that effect. Without such an argument -- which you seem to have no clue how to formulate, much less defend -- you have only a false appeal to authority, plus an apparent propensity to cast unjustified aspersions on people who point out your mistakes. This is what I have tried to caution you against.

If you do not know how to argue intelligently about exit polls, and you are not willing to learn, why don't you just drop the exit polls and build your case about Gore's winning margin in Florida? That would be an intellectually justifiable approach. Statistical error 'works' in both directions; you are unlikely to choose a margin that is inconsistent with the exit poll interviews, although of course it may be inconsistent with other data.

False appeals to authority, coupled with unjustified aspersions, are in very poor taste in my opinion.