Since our post earlier this afternoon suggesting that Norm Coleman was the slight favorite to win the recount, Al Franken has gotten three pieces of good news which cloud the picture and may tilt the probabilities in his favor.
The first is that, according to TPM's Eric Kleefeld, the Franken campaign estimated that it was just 50 ballots behind as of this morning, assuming that all challenges will be rejected. This standard is different from the one the Secretary of State uses, as the Secretary of State treats all challenged ballots as nonvotes until they are addressed by the Canvassing Board, effectively allowing either campaign to deduct votes from the opponent's total by challenging legal ballots. However, since the vast majority of such challenges will be rejected, the Franken campaign's standard is probably more reasonable.
I have also had conversations with senior Franken officials and been quoted similar numbers, and believe there is strong reason to regard the Franken campaign's claims as credible -- they are tracking every challenged ballot in every precinct, and have far more information about the nature of ballot challenges than is publicly available. (The Coleman campaign likely has such information too, but have been less forthcoming about it). In addition, being about 50 ballots behind at this stage would also be more consistent with the consensus of our statistical models, which suggested that while Franken probably remained behind, the race was much closer than it appeared to be from the Secretary of State's count.
The second piece of cheery news for Franken is that officials in Ramsey County have "discovered" a stack of 171 ballots that were never counted in the first place; those ballots returned a net of 37 votes for Franken. This is a very big deal; a net gain of 37 votes makes a huge amount of difference in a race that could easily be decided by a small, double-digit margin. If the 37-vote gain in added to the 50-vote margin that the Franken campaign estimated that it trailed by at the start of the day, that would put them just 13 votes behind with 8-9 percent of the state's votes still left to be counted. That is to say, according to the Franken campaign's methodology, the race will likely be effectively tied as we head toward the challenge stage of the recount process.
The final piece of news is that the Canvassing Board has instructed county officials to sort through rejected absentee ballots, and identify those ballots which did not appear to have a valid reason for rejection. This is potentially (although far from certainly) a precursor to those ballots actually being counted, a process which we estimated would result in a net gain of 25 to 100 votes for Franken (the Franken campaign believes a similar number of ballots are at stake).
All in all, it is suddenly much more difficult to make the case that Coleman is the favorite in the recount. If the absentee ballots wind up being counted, the opposite may in fact be true.