We should try to get the full, state by state poll data:

[ewulf]

just getting the (pure, 12:00) numbers that were posted by the networks would be GREAT, but we really need to focus as locally as possible, preferably on the actual sample of precincts, to really get things working.

As althecat posted here:

http://www.democraticunderground.com/discuss/duboard.php?az=show_mesg&forum=203&topic_id=58415&mesg_id=58415

Edison and Mitofsky, the polling firms that collected exit poll data for the National Election Pool, are willing to sell their full, state-by-state exit polling reports for $750 dollars a piece, or 38,000 for the whole county (althecat thinks it can be worked down to $10,000). This is a fair amount of money, but I'm sure their are a number of groups out their that would be quite capable of raising such a sum, and a lot of curious people who would be willing to pitch in to see the full reports. A hell of a lot more was raised for the recount in NH.

If we had access to this data, it could end up saving this overall election auditing project a lot more money in the end: we could target our recount requests and other investigations much more effectively. It would be a good investment.

I would expect that the DNC would have already bough the full data for strategy analysis purposes (at least if they were worth their salt as a political organization they would have), but somehow I don't think they're in a sharing mode right now.

Could people with contacts see if there are any organizations out there that would be willing to take this up?

[truehawk]

1. I agree

The more data we have the better chance we have.

[Be Brave]

2. I think DU should also get exi t poll data from more than one source.

Edison and Mitofsky was the one used by the media, but were there other agencies conducting exit polls? The more sources the better -- this would make the statistical analysis more robust and accurate.

[ewulf]

4. More polls would be better,

thats why its such a shame that all of the major media companies put everything into one bag this time around. The parties did some exit polling, but I don't think we'll be getting those numbers.

As for other, independent groups, I'm not really sure who did what. votewatch.us probably did some sort of exit polling on 11/2, but I'm really not sure what exactly they have been doing since the election... They have a report on their exit-poll results from Bernalillo County, New Mexico, where they apparently set people to 40 polling places,

http://www.votewatch.us/Members/StevenHertzberg/report.2004-11-02.2293324889/view?searchterm=None

but the report has limited use: they didn't ask people party affiliation or who they voted for: we could probably use their gender distribution numbers: (52-48 Women-Men) and their provisional-regular ballot numbers, but overall its not to usefull...

As for other organizations, I'm sure someone else on DU knows?

[Carolab]

3. One proviso

It has to be the ORIGINAL data, not the Roved up stuff from later in the evening.

[ewulf]

5. I would hope that, for $700 a state

they would give us the full data without readjustment...

[Be Brave]

6. kick

[Be Brave]

7. kick n/t

[TennisGuy2004]

8. An Important Question about the Diebold machines

I just had a thought. What if certain of those manual voting machines (using the opti-scan technology) were actually calibrated so that you thought you were voting for John Kerry and you're actually voting for George Bush. It wouldn't show up in the hand recounts because it would register as a valid vote for George Bush, even though you thought you were voting for Kerry.

For instance, this type of fraud could be used in certain "close" counties, like Hamilton Country for instance, and some of New Hampshire.

For example, let's say the exit polls from Hamilton County (Cincinatti) show Kerry leading 53% to 47%. And then it turns out Bush "wins" the county 53% to 47% because of these "reverse" voting apparati. Is this even remotely possible?

[Be Brave]

9. ewulf, are there any updates to the efforts by DU to get

exit polls from several sources? Any progress on this?