After the exorcism removes their demons, they would be clear minded enough to thereafter vote conservative.

Although in Frankins case, we may need multi sessions.

Sounds like a plan. But don’t the courts find out. They may not belief in seperation of Satan and State.

I’ve been following the Star-Tribune’s running tally of the recount, and it’s not valid statistically to “project” partial adjustments to an entire county based on the percentage of votes already counted

Yeah, I know that. There are lots of bad assumptions in what I did, it was mostly something to kill time during lunch.

Assumptions made that I can think of:

– Each county is homogenous. That’s obviously not true, but hopefully there will be some offsetting deviations. But since I don’t the have the time or energy to overlay with census tract that has to stand.

– The errors will break proportionate to the overall vote in a county. Again, probably not right.

– The challenged ballots evaporate, or at least offset. I have no idea how this will play out.

– The overall conduct of the recount will honest and above board. Who are we kidding?

Again not scientific, but possibly a decent baseline to check on occasion.

I had a similar ballot today. Somebody had written in a person’s name (another candidate running for state house in a different district) in every available write in spot. This person chose, however to fill in the oval for different candidates in some races and for this write in candidate in others. When it came to voter intent, writing the name is not enough, you have to choose to write the name and fill in the oval. If you simply wrote in the name and didn’t mark the oval it was an undervote.

Here’s a quick mathematical projection if anybody’s interested as of Noon CST:

Coleman is leading by a little in counties that have completed the recount: Coleman net +4

Franken is leading in counties that are in the process of counting. By doing a county by county projection of the votes found for each county and projecting to a 100% completion: Franken net +286

I’ve been following the Star-Tribune’s running tally of the recount, and it’s not valid statistically to “project” partial adjustments to an entire county based on the percentage of votes already counted. The big counties around Minneapolis-Saint Paul are very heterogeneous, with the big cities favoring Franken and the suburbs favoring Coleman–you can see that if you click on the “city map” for that area. If most of the already-recounted votes come from a Franken-heavy or Coleman-heavy area, the initial trend may reverse itself as the recount progresses.

The other problem with this “projection” is that there can be relatively large vote shifts from one precinct, which throws the idea of proportionality out the window. For example, when Olmsted County was 98% counted, it showed a net 2-vote shift for Coleman. Now it is 99% counted with a net 19-vote shift for Coleman, probably a disproportionate shift from one precinct. The problem is, nobody knows what vote shifts might still lurk in the precincts yet to be recounted.

Another problem with “projection” is that the candidate ahead in the original vote tally doesn’t necessarily gain votes in the recount. Franken won Minneapolis by better than a 3-to-1 margin, but Coleman netted votes in the recount there (so far), and Franken has gained votes in some counties which Coleman won, which are completely recounted.

One major trend observed in the recounts is that in most counties, both candidates are LOSING votes–possibly votes counted by the machines but found to be defective by hand, or possibly ballots run through the machines multiple times by mistake. If previously-counted votes are being cancelled, the candidate ahead in that area will tend to lose them faster…

One good sign–Coleman’s lead was down to 130 votes earlier today, but has bounced back up to 156 votes now.

But with several large counties not scheduled to be counted until early December, and over 800 ballots already challenged (between the two campaigns), this is a long way from over.

What’s up in Renville County? According to the data, Franken challenged 476 ballots – out of 6181 cast. That’s over 7.7% of all ballots cast.

SouthernRoots on November 21, 2008 at 1:34 PM

If you’re looking at the Star-Tribune tally (as I am) it must have been an error. By now (2:37 PM) Franken is only challenging 2 ballots in Renville county. Possibly, someone at the Star-Tribune accidentally typed the statewide total of Franken challenges into the box for Renville County.

Probably a big relief to the Renville County canvassing board.

Coleman is leading by a little in counties that have completed the recount: Coleman net +4

Franken is leading in counties that are in the process of counting. By doing a county by county projection of the votes found for each county and projecting to a 100% completion: Franken net +286

The counties that haven’t started yet tend to lean to Coleman, especially the bigger ones. Looking at the vote differential and how things have broken in the completed counties, I see Coleman picking up 274 in these counties.

Overall Franken nets 7 votes. Coleman wins.

What’s up in Renville County? According to the data, Franken challenged 476 ballots – out of 6181 cast. That’s over 7.7% of all ballots cast.

SouthernRoots on November 21, 2008 at 1:34 PM

It’s Franken-stein’s chosen place to steal the election. Don’t want to make it too obvious by going for someplace big like Hennepin or Ramsey or St. Louis County. The hicks out in Renville County won’t know any better.

New Mark Twain story: The Jumping Votes of Renville County.

p.s. whos badass anyway?

Now, as for the Day 2 Rulings:

1. Count It – Norm Coleman
2. Count It – Al Franken
3. Count It – Norm Coleman
4. Count It – Norm Coleman