The Minnesota Senate Recount Episode III: Revenge of the Lizard People
by CA Pol Junkie
After two days of the recount, one thing has become clear: we have a nail-biter on our hands. It is looking very likely that the winning margin will be in the double digits (if that). Although goofy things can come up at any time during a recount, the projections are zeroing in on a near tie after the recount. Then it will be up to the State Canvassing Board to resolve the challenged ballots. No matter how this turns out, let's all hope it isn't decided by Lizard People.
In the middle of the day yesterday, Coleman had built up a sizable lead in challenges, which was good news to Franken since he would likely get votes back when those challenges were rejected. By the end of the day, Franken had caught up and slightly surpassed Coleman in the challenges. There are two types: Type A and Type B. Type A are challenges by candidate X of a ballot that would be recorded for candidate Y. Candidate Y loses that vote until the State Canvassing Board rules on the challenge. If the challenge is rejected, candidate Y get the vote back. Type B is a ballot that would be recorded for a third candidate or nobody. Candidate X can challenge that the ballot should be for candidate X. This does not affect the recount totals, but could eventually add a vote to candidate X's total if the State Canvassing Board agrees with the challenge.
Type A challenges are bad for a candidate, since they are likely to be denied by the State Canvassing Board, thus adding to the other candidate's total. Type B challenges are good, since they can only add to a candidate's total. Almost all challenges will be denied, as we can see by comparing the Twin Cities. Ramsey County (St. Paul) works through the challenges with both campaigns before submitting them to the State Canvassing Board to get them down to just the ones with a serious chance of being agreed to. Hennepin County (Minneapolis) does not reduce the challenges that way. Ramsey County has only 13% of the challenges per vote as Hennepin. That means that statewide up to 87% of challenges are truly hopeless. Some fraction of the remainder will be accepted. To find out how the challenges will turn out, we need to estimate how many Type A (bad) and Type B (good) challenges each candidate has. Based on the expected number of ballots gained without challenges (around 0.02% by regression analysis), I can estimate how many challenges in each county are Type A and Type B for each candidate. Overall, about 84% of both candidate's challenges are Type A. As of the end of the day yesterday, Franken had a small lead in challenges, so Coleman would stand to gain a few votes from the State Canvassing Board. As a vague rule of thumb, 2/3 of the difference in the number of challenges could be added to the vote total of the candidate with fewer challenges.
Below are running updates of the current margin between the candidates and a spreadsheet projection of the final tally before the challenges are addressed by the State Canvassing Board.
November 20, 9:24 PM CST (46% recounted): Current: Coleman +136 votes Projected: Franken +12 votes
November 21, 3:40 PM CST (49% recounted): Current: Coleman +145 votes Projected: Coleman +19 votes
November 21, 9:43 PM CST (64% recounted): Current: Coleman +120 votes Projected: Coleman +40 votes