Were undervotes defaulting as votes for Bush?

[Hamoth]

I was reading abotu the problems going on with the county that triad failed to properly rig where the machine and hand counts were failing to match. What I found curious was that, from whatI could tell, undervotes where no presidential choice was present, were being logged as votes for Bush. If the mahcines were setup, as we have seen in some other places, namely NM, where Bush was the default choice in the software until something else is selected, then wouldn't all undervotes go to Bush? If only a few machines were so setup, this could be hte mechanism by which thousands of expected Kerry votes would be switched through sloppy voting and a default Bush column, to the bush column...

Could the recent standoff in that ohio county be the result of discovering one such machien that was rigged to log undervotes as bush votes due to a default bush presidential choice? This would jive with other observations and explain quite a bit. The voter suppression and intentional use of crappy machines and rushed lines would cause more undervotes, and therefore, more bush votes!

Any opinions or people smarter than me that can substantiate / debunk this idea?

[k8conant]

1. We'd have to see the software....

to know if Bush was the default.

[Hamoth]

2. perhaps including undervotes in the "test stack" would tell.

I think there would be a lot of ways to analyze this statistically.

For example, I hear that there are roughly half the undervotes this time that there were in 2000. This is one of the things often pointed to in claims of a better election...but since ohio used the same machinery (punch cards mostly) for the election, what caused there to be less undervotes? The counting perhaps?

Well if 1/3rd of the opscans are rigged to default as bush votes...then we would see precincts with an amazing lack of undervotes with punch cards. This would likely show up in places that were troublesome and that would be even more suspicious.

Are there ways to get ahold of the relevant data broken down by precent and machine if possible, showing ratio of votes to undervotes? Are there any places with a suspiciously low number of undervotes relative to the surrounding precincts?

Did these missing undervotes then correspond to a surprise growth of bush votes?

[berniew1]

16. Where is documentation on Machines in N.M. set up default to Bush

[skids]

17. Hrm this is very interesting.

One of the growing memes is that a primary target of the rigging was to switch Republicans who vote for Kerry back to Bush, and further conjecturing this was done because it would be hard to find, due to these people having a history of voting Republican.

That could easily be extended to undervotes. I've been working with a hidden assumption that the correlation between undervotes and Democrat support in the Republican or 50/50 rural counties is a result of invalidating Dem votes with bad calibration or equipment. But that same correlation could also be caused by switching the true undervote (people who could not stomach voting for either Kerry or Bush) to Bush in republican leaning areas.

As far as the rest of your points, yes, yes, and yes, but not so strongly as to be too blatantly obvious. There are a few "problem precincts" but those have extremely large undervote, so it would be the original scenario above -- voiding Kerry votes by making sure the Dem's machines suck.

If you want to put some time in, you can get precinct data for counties on paper and enter them. We have a project going on now where we are trying to get as much of the data in as possible. As a side-benefit, you get to play around with some per-precinct data on your own. pm me if you are interested.

However, as a second answer to your question about the data, some counties give you the information on "undervotes", other counties do not, so you can't separate undervotes from overvotes, some clean up their data to make it look like there was 0 undervote, and still other counties lump the undervote and overvote in with all the writeins in a category called "other" (we think, they don't actually say what that category includes.) It's a total clusterf**k. One thing I need right now is a complete canvas including broken-out absentee votes for an ohio e-voting or opscan county -- but I can't find it, and I have 9-10 counties here plus what's out on the web.

[On the Road]

3. If So, There Should be No Undervotes

Or, if the machine counted them and then spit out the card, the number of physical ballots should not match the official vote.

[Hamoth]

4. Exactly And isnt that what we saw?

In this thread: http://www.democraticunderground.com/discuss/duboard.php?az=view_all&address=203x163172

if only one of three machines in a polling place were rigged to do this, then you would have a reduction of undervotes appearent, but an equivilant increase in bush votes. The machine used in the thread I refferenced, from what I could glean was doing just this!

Was this "error" state-wide?

[jj24]

5. very likely...

this sounds like a really viable possibility!!

[Hamoth]

6. I would like to check the data

But where is it? I see many analysis on here of various certified election data and undervote / overvote ratios, etc...

I bet that we will see that ocuntied using the EXACT same technology as 2000 had an equal reduction in undervotes as increase in bush votes. This should be easy as hell to determine...anyone?

[jj24]

7. I'm Sorry...

I don't know where people are getting their info. anyone - pls help!!!!

[newyawker99]

10. Hi jj24!!

Welcome to DU!! :toast:

[jj24]

12. Thank You!!!!

[Hamoth]

8. How about the famous DU data heads?

Can any of the statitisics heads here chime in with a rough go-over as to the credibility of this possibility? If I'm on to anything here, it could really bolster Arnebeck's case!

[merh]

9. I think your theory makes sense. Email Arneback with it.

Don't let the others tell you he gets too much email, I am sure he has someone reviewing it for him and passing it on to folks working on this with him.

Welcome! :hi:

[Hamoth]

11. Thanks!

Will do...is it on alliance's site?

[merh]

13. Yes, I believe it is

Hang on, let me see if I have it .....

Got it: [e-mail]

Great point and worth sharing with others.

[Hamoth]

15. Thanks! Here's what I emailed him:

I was reading about the problems going on with Fairfield County where the machine and hand counts were failing to match. What I found curious was that, from what I could tell, 4 undervotes where no presidential choice was present, were being logged as votes for Bush. If the machines were setup, as we have seen in some other places, namely NM, where Bush was the default choice in the software until something else is selected, then wouldn't all undervotes go to Bush? If only a few machines were so setup, this could be the mechanism by which thousands of expected Kerry votes would be switched through sloppy voting and a default Bush column, to the bush column...

Could the recent standoff in Fairfield County, Ohio be the result of discovering one such machine that was rigged to log undervotes as bush votes due to a default bush presidential choice? This would jive with other observations and explain quite a bit. The voter suppression and intentional use of crappy machines and rushed lines would cause more undervotes, and therefore, more bush votes!

We know that Blackwell and his friends often proclaim that there were nearly half as many undervotes in the election than previously in 2000. However we also know that so much of the state uses the same punch-card machinery that I wonder at how the reduction was achieved? The tabulation is the new component, and we may therefore ask how the tabulators reduced the number of undervotes in a clearly mechanical process. While claims of greater accuracy are issued, one may wonder by what standard the accuracy has been measured. It is known that most under and over voting occurs in heavily democratic precincts, and thus counting undervotes as Bush votes might well effectively switch expected Kerry votes, to the Bush column during tabulation. There might exist the specific mechanism for the anomalies you have observed. Is it possible that some machines were predisposed to log undervotes as votes for Bush?


Any opinions or people smarter than me that can substantiate / debunk this idea?

[BlueDog2u]

14. Very interesting theory

I like the elegance of it. It is simple but would be very difficult to detect, and perhaps account for just enough votes to swing states like Ohio or Florida.

Sounds very worth investigating to me.