Proofiness - Charles Seife [124]
53
Of course, the campaigns tended to go to the press crying foul whenever their opponent picked up votes, and they tended to be silent when errors were resolved in their favor: classic cherry-picking of data.
54
Their high-tech randomization device: drawing little slips from brown paper bags. Minnesota secretary of state Mark Ritchie was a bit embarrassed about the procedure, but said that the state had given up the tried-and-true fishbowl method when one official got his hand stuck in the bowl.
55
If you restrict the numbers to the case where none of the ballots were challenged by either party, the error rate drops by roughly a factor of two, to 0.02 percent, but that underlying error rate is as low as one can go. Counting errors are unavoidable.
56
This motivation became blindingly obvious as the two campaigns grudgingly withdrew their sillier challenges. Coleman’s campaign wouldn’t budge until Franken’s campaign withdrew some challenges first. Then, when Franken’s campaign withdrew a few challenges, the Coleman campaign would withdraw almost exactly the same number. This was the only way the Coleman campaign could ensure that they wouldn’t lose their pseudo-lead in the vote tallies.
57
The “no identifying marks” rule was particularly troublesome. Quite a number of people filled out the wrong circle on their ballot in pen. When they discovered the mistake, they put an X through the wrong candidate’s oval and carefully filled in the oval of their true vote. These hapless voters then put their initials on the side of the ballot to designate that they approved the change—as they would on a contract—thereby putting their vote in jeopardy because they had inscribed an identifying mark.
58
For the record, I think that the lizard people decision was correct; it was an overvote. However, I believe that overstrict interpretation of the statute forced the pro-Coleman forces into a ridiculous lie—to get their way, they had to pretend that they honestly believed “Lizard People” was a living, breathing individual who had somehow convinced a Minnesotan to put him up for the Senate.
59
In the United States, presidential elections have a curious structure. Citizens vote to get “electors” to represent them in a body—the electoral college—that formally determines who will be the next president. The rules of how to appoint those electors are left to the states. Most states have a winner-take-all rule; whoever gets the most votes in the election (usually) gets all of the state electors’ votes. As a result, in a close race, the votes of a relatively large number of electors—even enough to determine the winner of a contest—can hinge on a few hundred votes. This is what happened in Florida.
60
The number of Buchanan votes, roughly 3,400 of them, was more than ten times the number of votes Buchanan was expected to receive given the history of the county and the voting patterns of Florida residents, even under extraordinary conditions. When Bush spokesman and later press secretary Ari Fleischer explained away the anomaly by insisting that “Palm Beach County is a Pat Buchanan stronghold,” even the Buchanan staff called the claim “nonsense.”
61
Another example of this role reversal came from a set of lawsuits in Seminole and Martin counties. These lawsuits hinged on allegations that Republican operatives were allowed to alter improperly filled-out ballot applications—a shady procedure that tainted thousands upon thousands of absentee ballots in these counties. The pro-Gore forces (not Gore himself, even though some of his legal team tried to convince him to join in the suit) tried to get those ballots thrown out, as Florida law would seem to dictate. (Rules are rules!) If they succeeded, the suits would likely have eliminated a net of 2,000 to 2,500 votes for Bush, handing Gore the victory. Predictably, the pro-Bush forces (including Bush’s team) fought hard to keep those ballots in the contest. (Count every vote!) In this particular battle,