I spent six years as an elections lawyer. I handled a lot of very wacked election canvasses and recounts. One thing they all had in common: when new votes were found, they generally went both ways -- a few for one candidate and a few for the other candidate. It didn't quite follow the polling, but then close races only happen at the 50-50 margin.
Here's something that never happened: the votes did not all go 100% for one candidate except in the two instances where there was only one misplaced ballot discovered.
In Minnesota, though, the votes are all going for Al Franken — pretty much all 600 of them. That defies statistical probability.
Here's something else canvasses of ballots have in common: a lot of ballots drop off. It's an undervote, e.g. 20 races on the ballot, but only 15 get votes. Sometimes people don't want to vote for the Republican, but they can't stomach their own Democratic candidate.
Al Franken's campaign strategy is now premised on the undervotes in Minnesota. His legal team would have us believe that a vote for Barack Obama is also a vote for Al Franken — never mind the Democrats who could not stand voting for an ass like Franken.
Of course, this is fatuous nonsense. Just look at the polling. Franken trailed Obama by 12.6% in the final poll conducted before Election Day. In fact, for the last months of the race, Franken chronically trailed Obama.
But Franken would have us believe that if a ballot shows up with no vote for Franken, despite having votes for other Democrats, the voter intended to vote for Franken and just, I guess, forgot to.1
Al Franken must be the only man on earth who doesn't recognize how many in his own party hate him.
Undermining this whole theory is the Democratic effort to steal the votes. During the canvassing, Franken's vote total increased at a greater percentage than Obama's. Should we also assume that the people intended to vote for Obama and just forgot or marked the wrong box? In fact, suspiciously, Franken's vote went up well outside the margin of machine error and well outside the margin of error overall for the canvass. In other words, Franken's now operative theory on undervotes does not match (A) the polling showing Obama out performing him; (B) the results from election day; nor (C) the pattern of vote shifts to Franken in excess of other candidates on the ballot during the canvassing.↩