. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Nap Time!!!

Wednesday, February 25, 2009
Election news

News from the Elections Council:

Turnout: 3786. If folks didn't care before, they probably still didn't care. Under normal circumstances, I would say this looks horrible for Moghtader, but given his poor showing when presenting his case last Friday, I actually think a high turnout number would have been even more ominous for him. The recall needs 2/3 of the vote, but I'm far more interested in finding out how many people went through the effort to vote, only to abstain on the only question.

The cost of the election is being estimated at $20,000, with the general election coming up estimated at $35,000. (The filing period starts on Monday!) The Publicity Coordinator for the election seems confident that he can plaster the campus with publicity and boost turnout from last year (he should have said "last election" and made beating the recall turnout his goal). My only recommendation is that he find a way to avoid including announcements of the programs that the ASUC has done during the year and/or is putting on at the time of elections in the campus-wide e-mail he's planning, because I think that's enormously shady. In fact, aside from giving permission to use the system, I don't think ASUC elected officials should have any say in the content of that e-mail.

The Elections Council is planning on waiting until next Tuesday to count results after all. In fact, the chair says they're expecting a delay until Thursday on account of "the AirBears issue." The AirBears issue is that for about 9 hours on Monday, folks could vote using AirBears, which is prohibited in the By-Laws for some reason (I've heard two very different ones). She wants to look at the data and find out how many did, and whether it's enough to possibly affect the election. If so, she wants to find a way to link those who did vote using AirBears to their votes (though not their identities).

First of all, if this is possible, it's a flaw in the system. Being able to attach votes to IP addresses compromises the closed ballot. But more importantly, the Elections Council will never be able to get away with not counting the AirBears votes, and I don't even understand why she would want to. I don't think the voting rules can even be read to place an obligation on voters. It's the job of the Elections Council to make it impossible, not the voters to avoid doing it. As I understand the voting program, a person who voted from AirBears would not be able to vote from another computer, and would have no notification that such a vote might be invalid. Someone could probably ask the Judicial Council to give the Elections Council the okay to announce the results regardless of the AirBears issue, and then results should come out Tuesday.

posted by Beetle Aurora Drake 2/25/2009 10:26:00 PM #
Comments (2)
. . .
Comments:
You say "She wants to look at the data and find out how many did [vote on airbears], and whether it's enough to possibly affect the election. If so, she wants to find a way to link those who did vote using AirBears to their votes"

What I heard reported from Calvin (her tech assistant) was that the votes would be tallied twice: once including the airbears votes, and once without. Then there would only be an issue if the legit votes + airbears result was different than the result just from the legit votes.

This implies that there is a method of distinguishing which votes come from Airbears, which is in direct contradiction to what Emily went on to say about not being able to match the two.
 
The impression I got was that if the margin of victory was bigger than the number of AirBears votes, it won't be a big deal. And perhaps they'd use some statistical techniques for for determining how much we should care even if the margin is smaller. If the margin is too small, then they'd try to match the two.

I'm not entirely convinced they have access to the information they need even to count the AirBears votes, and I really doubt they can match those IP addresses to votes. That information may be recoverable by crosschecking some log files, but again, I'm unconvinced they have access to the necessary logs, even if they are kept.

I think you're putting far more confidence in her suggestion of "data mining" than she did, and her use of the term suggests that she really didn't know.
 
Post a Comment


. . .