MN Sen Recount (UPDATE: Stuart Smalley certified winner, lawsuit forthcoming)

[Sam Spade]

And you thought challenge *overkill* was bad before...  I'm hoping this is a misprint, otherwise I'll have as much *fun* as with Hennepin.

Sherburne
Vote Changes
Coleman -433, Franken -426

Challenges
Coleman 452, Franken 422

...help...

I just saw this as well. This is almost as many challenges as in Hennepin! I understand that the canvassing board has pleaded for help from both sides to reduce the challenge pile. Does that mean that the attys will agree swap challenged ballots and allow them either both in or out at the same time?

http://www.startribune.com/politics/state/35166869.html?elr=KArksLckD8EQDUoaEyqyP4O:DW3ckUiD3aPc:_Yyc:aULPQL7PQLanchO7DiUT

[Sam Spade]

I'm going and updating results this evening before I go out.

Basically, in addition to the challenge gap, there are actual changes in most of the counties that you can ascertain separate of the challenge gap and all that.

Most of these changes are minor and cancel each other out, both within the county and between the counties.  The only two I noted so far were changes in favor of Franken in Ramsey and St. Louis County of about 30-35 votes each, although that was a couple of days ago and this may have changed. (oddly enough, Hennepin shows very little).

We have another significant change tonight.  Coleman gained about 25 votes in Becker County distinct from challenges.  We'll see if any more appear.

[Sam Spade]

We have another significant change tonight.  Coleman gained about 25 votes in Becker County distinct from challenges.  We'll see if any more appear.

Of course, as soon as I say that, there are more.  Specifically, Coleman lost 25 votes in some Blue Earth County precinct (probably a mistabulation).  Easy come, easy go. 

[Sam Spade]

If any of the ballots that were initially ruled for one candidate or ruled invalid were, upon inspection, ruled for the opposing side, the other side is guaranteed to challenge.

Actually, that's incorrect.  A lot of gains and losses have occurred where no candidate has challenged. 

And in the challenge gap (which should be the number of votes which have *probably* been removed from each side due to challenges) and you get a number similar to the number Franken's camp tossed out.  (slightly higher, but then again, I presumed they might inflate a bit).

[Sam Spade]

If Coleman is challenging undervotes (and we don't know that for sure) it doesn't help Franken one bit. 

Both sides are challenging undervotes/overvotes aggressively, though Franken is probably slightly more aggressive.

[Sam Spade]

Are you a political fanatic and can't get enough of the Minnesota recount?

Are you looking for a way to waste away your Thanksgiving without talking to the rest of your family (or dread!) the mother-in-law?

Then play the Star-Tribune Ballot Challenge!  599 ballots for your perusal and voting choices.  Show your partisanship and stupidity now!!!

http://senaterecount.startribune.com/

[Sam Spade]

Some of these ballots are quite funny.  And of course, the reasons for some of the challenges are almost as funny.

[Sam Spade]

http://senaterecount.startribune.com/media/ballotPDFs/washington_WoodburyP7_challengedballot49.pdf

The Franken/Bachmann voter...

Clearly someone who needs to be locked in a mental institution or shot immediately...

[Sam Spade]

Haven't updated the topline results, choosing instead to finish the precinct-by-precinct reports.

According to my unofficial results, which is based on a certain number of assumptions whose error rate should even out between the two candidates, here is what I get, based on SOS.

This is not perfect, as I have not double-checked a few of these results, but it should be fairly good:

SOS
Coleman: 1,044,255 (1,046,101)
Franken: 1,040,285 (1,042,208)

Coleman: +292 (+215)

Actual Changes (these are changes I can ascertain from the numbers or from the extrapolation of assumptions regarding subtractions of votes or challenges)
Coleman: -5
Franken: +43
Total Change: Franken +48

The raw change that occurs if you remove my assumptions regarding subtractions of votes or challenges is Franken +70.

Challenge Gap
Coleman: -1842 (CH by Franken)
Franken: -1965 (CH by Coleman)
Total Gap: Franken +123

The raw change that occurs if you remove the same above-stated assumptions makes the challenge gap Franken +145.

Non-Subtracted Challenges
Coleman: 443
Franken: 395

We must take a moment to discuss these challenges.  To be plain, these are challenges that cannot be directly traced to ballots that have already been counted.  Granted the tracing is based on an assumption, but the assumption is already proving to be correct in the ballots the Star-Tribune has given us.  These ballots can be challenges of blank ballots, overvotes, votes for other candidates.  They can also be challenges of ballots that were not counted in the original count, but were counted in the second count and then challenged.  It is quite possible that Coleman has challenged a number of ballots that were not originally counted as Franken, but were counted as such the second time.  If so, then those numbers would need to be reduced from the Result.

Final Result: Coleman +169 (w/o assumptions: Coleman +147)

Counties to Be Completed
Beltrami (1 precinct) - Star Tribune has reported no change in this precinct already.
Blue Earth (9 precincts)
Brown (100%) - Star Tribune has already reported this.
Dodge (100%)
Hennepin (61 precincts) - Pretty much all precincts are located in urban Minneapolis.
Jackson (100%)
Le Sueur (100%)
Ramsey (34 precincts) - Pretty much outer ring St. Paul suburbs.  All of urban St. Paul has reported.
Rock (100%)
Scott (100%)
Sherburne (100%) - Star Tribune has already reported; SOS is probably just lazy, though I don't blame them.
Stevens (100%)
Winona (100%)
Wright (100%)

[Sam Spade]

I noticed that in my set of numbers Coleman's result should be 2 less than it is (in other words +167/+145).  I probably inputed one of the challenges backwards, so I'll correct it later.

Anyways, I wanted to point out the assumptions I'm building on so that folks have some type of model with which to compare.

At its simplest, I have divided challenges into two types - 1) those associated with subtraction of votes per precinct; 2) those not associated with subtraction of votes per precinct.
1.  But before we get to that, there are countless material changes upwards and downwards for Coleman and Franken per precinct that we can document without any assumptions.  In other words, if any votes are added to the candidate (e.g. 1 vote to Coleman) in the precinct, we can presume that this is a material change.  Similarly, if 1 vote is subtracted from a candidate (e.g. 1 vote subtracted from Coleman) in the precinct without any corresponding challenge, we can presume that this, too, is a material change.  Having analyzed that, these are the material changes:

Material Changes
Coleman +78
Franken +148
Change: Franken +70
2. For the two basic groups above, we must make some assumptions.  The assumption associated with challenges per precinct that result in subtracted votes is that we should assume Coleman is challenging Franken votes (or votes counted as Franken in the first count) and vice versa. 

As noted below, there are certain places where the assumption appears to be clearly reversed, but this only accounts for roughly 4%-5% of the challenged votes.  I keep this in mind, because Coleman challenging a Coleman vote that has been subtracted from the overall total stands 1) a slightly higher chance of success than Franken challenging a Coleman vote (based on what I've seen); 2) results in a net loss if unsuccessful.

Anyway, without using the above assumption, the number of net subtractions or the "Challenge Gap" is as such:

Challenge Gap
Coleman -1924 (challenged by Franken)
Franken - 2071 (challenged by Coleman)
Change: Franken +147
3.  The last group of challenges are those which are not connected to any subtraction of votes.  Once again, we base what these challenges are following the assumption that subtractions are either Franken challenging Coleman votes or vice versa. 

Visually, it reads this way (and this helps in understanding point two also):

You must be logged in to read this quote.

A few permutations from this variable are possible.  For example, Coleman could be challenging the subtraction of a Coleman vote.  Of course, this would then require Franken to challenge the subtraction of a Franken vote.  But considering the nature of challenges already and the tendencies of County Boards of Elections, this type of possibility seems fairly remote.  Henceforth, the best answer is to make the assumption that Coleman is challenging Franken votes and vice versa.

More trouble occurs from this type of pattern:

You must be logged in to read this quote.

In this case, two challenges connect to votes that have been subtracted (presumably, as stated above), but one challenge doesn't.  In trying to allocate these types of challenges, I have kept the original assumption.  Franken is challenging Coleman votes and vice versa.  Henceforth here, the one challenge that doesn't connect to the subtracted votes is Coleman's.

Now, as to the flavor of this challenge - I do not presuppose.  There are a number of possibilities - blank vote, overvote, some other candidate's vote.  Only one possibility changes the math greatly.  With the example above, if Coleman was challenging a vote that was not counted as Franken's in the first count but was counted as Franken's in the second count, then this number would change the challenge gap strongly.  In a certain way, it is the opposite of the earlier noted "Coleman challenging a subtracted Coleman ballot" because it is the candidate challenging the decision of the County Board of Elections which changed the result from the first count. 

Much as above, one would have to think that this type of challenge would be more successful than the type that doesn't reflect any decisions of the County Board of Elections, but maybe my presumption is wrong.

Anyways, here are the non-subtracted challenges:

Non-Subtracted Challenges
Coleman: 443
Franken: 395
4. There is one other curious type of number that results in the precinct counts that relates to the Challenge Gap.  It is this type of result.

You must be logged in to read this quote.

Clearly, this results in a material change of one vote, but the results provided are completely unclear as to which ballot Coleman is challenging.

As above, I have taken my assumption to this level as well.  Basing that Coleman is much more likely than not to challenge Franken's vote, the material change must be the subtraction of the Coleman vote.

There are 95 votes in the Challenge Gap that fall under these parameters and are divided as follows:

Coleman: -42
Franken: -53
Change: Coleman +11 (in the Challenge Gap/Material Change)
5.  Lastly, as mentioned above, there are certain places where we can determine (with not absolute, but fairly absolute certainty) that Coleman is challenging Coleman's ballot and vice versa...

You must be logged in to read this quote.

...or the more difficult example...

You must be logged in to read this quote.

Once again, all this is based on the assumption in challenges that Coleman is challenging Franken's subtracted ballot and vice versa.  There are 93 such ballots that we can determine using this method, which gives us:

Coleman - 41
Franken - 52
It is off of these assumptions and numbers that I get the Coleman +167/145 result.  I will continue to update this week and retool the numbers.

Lastly, I do want to point out that if the challenges break in a certain way there can be a difference - this is merely going down to the precinct level and trying to give some clarity...

[Sam Spade]

Star-Tribune says that Franken would have to prevail on 6% more of his challenges than Coleman to succeed. 

Of course, maybe this analysis is like Star-Tribune polling.  FWIW, if my numbers are correct, 6% sounds about right, actually, especially in the number of successful challenges is quite low.

http://www.startribune.com/politics/national/senate/35263049.html?elr=KArksLckD8EQDUoaEyqyP4O:DW3ckUiD3aPc:_Yyc:aULPQL7PQLanchO7DiUr

[Sam Spade]

Too bad the Star-Tribune is not smart enough to release that this isn't a Senate race challenge (another recount actually)

http://senaterecount.startribune.com/media/ballotPDFs/Morrison_LittleFallsW2_challengedballot1.pdf

[Sam Spade]

I'm really baffled by the Nate Silver hatred from Republicans. All he did was predict an election correctly.

The reason why Republican partisans don't like Silver is easy to recognize. 

First, his methods, while not infallible, are certainly quite useful, and should probably be fairly accurate most of the time provided he takes care of all the variables properly.  Success is often the great polarizer in politics.

Moreover, when he produces his models, he provides a classic partisan analysis (in other words - he sees the glass as half-Democrat, not half-Republican, heck even three-quarters Democrat).  That adds to the annoyance factor.

The third factor is, of course, that those who quote him tend to view him as a hallowed expert, or something similar.  This is not the greatest analogy, and certainly not too current, but it would be like a Republican quoting Lee Atwater media-manipulation techniques.  In other words, both people know what they're doing, but that doesn't mean they're God or something, and that tends to annoy people.

[Sam Spade]

Too bad the Star-Tribune is not smart enough to release that this isn't a Senate race challenge (another recount actually)

http://senaterecount.startribune.com/media/ballotPDFs/Morrison_LittleFallsW2_challengedballot1.pdf

Either the Star-Tribune are idiots (again) or the SOS is.  Btw, this Doty challenge in the Doty-Lemieur race sucks as well.

http://senaterecount.startribune.com/media/ballotPDFs/Morrison_Pierz_challengedballot1.pdf

[Sam Spade]

Check page two of this ballot:

http://senaterecount.startribune.com/media/ballotPDFs/Olmsted_W3P6_challenged%20ballot%2041.pdf

[Sam Spade]

No, Obama is not Jesus...

http://senaterecount.startribune.com/media/ballotPDFs/LacQuiParle_Madison_challengedballot4.pdf

[Sam Spade]

Check page two of this ballot:

http://senaterecount.startribune.com/media/ballotPDFs/Olmsted_W3P6_challenged%20ballot%2041.pdf

How does a judge loving sexual abuse constitute an identifying mark i wonder?  Unless it is proven that only one person knows about the Judge's ... uh, dark side, but even then, the judge himself could have written that.

What is a pedifile anyway?  Is this someone who cuts his toenails too short?

[Sam Spade]

Probably someone pissed about a ruling the judge made.

This brought up an interesting question: Does it count as identifying the ballot if someone writes themselves in for all the minor offices? (As many have done. Almost all Coleman supporters interestingly.)

The Franken campaign has argued this but I remember reading that the judges generally feel that it's not sufficient proof (you and your friend might agree to write your friend in for everything or you might be writing in your sixth grade teacher)

I tend to agree with the judges (ignoring statute for a moment).

Your ballot should not be thrown out simply because you have the gall to write in your own name or someone else you know for a position where you don't like the alternatives.  Maybe it would be different if the ovals weren't colored in, but I still think it's not a reason for disqualification.  Keep in mind the "Lizard People". 

One of the things that most annoys me with the lever system in New York is how difficult it is to write-in a name for a spot.  I so wanted to vote for "Daniel Patrick Moynihan" in the NY Senate race in 2006, but it requires filling out a separate card and some other procedure, which is useless.

[Sam Spade]

Probably someone pissed about a ruling the judge made.

This brought up an interesting question: Does it count as identifying the ballot if someone writes themselves in for all the minor offices? (As many have done. Almost all Coleman supporters interestingly.)

The Franken campaign has argued this but I remember reading that the judges generally feel that it's not sufficient proof (you and your friend might agree to write your friend in for everything or you might be writing in your sixth grade teacher)

I tend to agree with the judges (ignoring statute for a moment).

Your ballot should not be thrown out simply because you have the gall to write in your own name or someone else you know for a position where you don't like the alternatives.  Maybe it would be different if the ovals weren't colored in, but I still think it's not a reason for disqualification.  Keep in mind the "Lizard People". 

One of the things that most annoys me with the lever system in New York is how difficult it is to write-in a name for a spot.  I so wanted to vote for "Daniel Patrick Moynihan" in the NY Senate race in 2006, but it requires filling out a separate card and some other procedure, which is useless.

At least you have the option of writing in candidates in New York. Stupid Texas.

I thought they had that silly system where you move a dial to enter letters (remember Sekula-Gibbs?).  Or maybe that was just Harris/Fort Bend County

[Sam Spade]

Apparently, Franken's campaign says they will withdraw "some" challenges.  I personally have no doubt both campaigns will, eventually...

http://www.startribune.com/politics/national/senate/35329829.html?elr=KArksLckD8EQDUoaEyqyP4O:DW3ckUiD3aPc:_Yyc:aUnciaec8O7EyUsr

[Sam Spade]

Which meant I had to vote for some random Republican for some random judge because there was no Democrat.

Skip the office.

Yep.

[Sam Spade]

As of today's update...

Counties to Be Completed
Beltrami (1 precinct) - Star Tribune has reported no change in this precinct already.
Brown (100%) - Star Tribune has already reported this.  SOS hasn't.
Dodge (40% or 6 precincts)
Hennepin (35 precincts) - Pretty much all precincts are located in urban Minneapolis.
Le Sueur (61% or 14 precincts)
Ramsey (15 precincts) - Pretty much outer ring St. Paul suburbs.  All of urban St. Paul has reported.
Rock (100%)
Scott (100%)
Winona (100%)
Wright (100%)

[Sam Spade]

As of today's update...

Counties to Be Completed
Beltrami (1 precinct) - Star Tribune has reported no change in this precinct already.
Brown (100%) - Star Tribune has already reported this.  SOS hasn't.
Dodge (40% or 6 precincts)
Hennepin (35 precincts) - Pretty much all precincts are located in urban Minneapolis.
Le Sueur (61% or 14 precincts)
Ramsey (15 precincts) - Pretty much outer ring St. Paul suburbs.  All of urban St. Paul has reported.
Rock (100%)
Scott (100%)
Winona (100%)
Wright (100%)

Update (Using My Method)
SOS
Coleman: 1,100,922 (1,103,291)
Franken: 1,105,030 (1,107,528)

Coleman: +344 (+215)

Actual Changes (With Assumptions in Parentheses)
Coleman: +112 (+21)
Franken: +157 (+46)
Total Change: Franken +45 (Franken +25)

Challenge Gap (With Assumptions in Parentheses)
Coleman (CH by Franken): -2481 (-2390)
Franken (CH by Coleman): -2655 (-2544)
Total Gap: Franken +174 (Franken +154)

Non-Subtracted Challenges
Coleman: 485
Franken: 426

Actual Total (With Challenge Gap Included, Assumptions in Parentheses)
Coleman +170 (Coleman +190)

[Sam Spade]

No one with any sense wants Franken on their "team".  That, and the fact that after today, 60 seats will no longer be a option and the fact that such a maneuver would be bound to infuriate the other side (probably the moderate Repubs too) leans strongly towards not using such a manuever.

[Sam Spade]

But come 2010, you want to be able to blame at least some of your problems on obstructionist Republicans.  Otherwise, every little thing that gets worse (crime, education, economy) will be because of the largely meaningless "blank check" (the public won't understand cloture votes).

It'll be near impossible to do anyway even without 60.  Public perception will be that Democrats control everything, whether you like it or not, for good or for bad.  You can only pull that gambit in divided government (see 1948/1996, for example), at least historically.

Public perception now is that Republicans control everything - though this is, of course, not true - but the Bush administration has zero talent in relating to people that it is otherwise, so it is, in fact, reality.