Sister Shirley, my first and second grade teacher, introduced me to the Magic Math Machine.

I remember it vividly, up to the point of near-practicality.  It was something she drew on the overhead.  It was square, with ears.

Think of Mickey Mouse with a square head, and you have the image.  Instead of a nose, the Magic Math Machine/square Mickey Mouse had a rectangle in the middle of its face, where the magic answer would appear.

Because I was a geek even at six, I re-drew the Magic Math Machine for my older brother on my chalkboard in the basement.  I invited all my friends to have a go at the Magic Math Machine.

Each ear contained numbers that were to be added.  This is the point, though, that my memory gets fuzzy.

I can't actually remember where the magic came in.

In retrospect, I don't think there was any magic at all to how this thing worked.  I believe Sister Shirley wrote one number in one ear, another in the other ear, and then told us how we added them together to write the answer in the rectangle.

Then, we could erase the numbers in the ears and the rectangle, and start again!

No wonder my friends weren't impressed.

(I should point out that Sister Shirley was my favorite teacher of all time, but when I was in the third grade, she left her vocation to get married.  I went to her wedding and yet I still only remember her as Sister Shirley.  Thus, I'm also beginning to think there is further evidence of gaps in my memory from the 1970s).

Anyway, I've been on the hunt for the equivalent of an Elections Magic Math Machine.

What I'd like to know, in some easy way, is the relationship between advance voting and polling place turnout.  My imaginary Magic Math Machine has sliders that let me increase the number of advance voting sites or machines at a site, and see instantly how that impacts the number of voters at each location on election day.

I took a crack at all of this following the 2008 presidential election and plan to follow this post with some numbers and the beginning of how such a Magic Math Machine might work.

One of the things I determined is that advance voting clearly lets us "sweat the assets," getting utilization of our voting machines at a rate of about 85 percent (meaning, of the time we are open, a machine would be idle only 15 percent of the time).  At the polls, the utilization rate is about 25 percent.

A robust Magic Math Machine is one of the things I'd like to see as an answer to the now infamous, "We've got to fix that," comment regarding lines at polling places by President Obama.  If the Election Assistance Commission is amped up as a response to this statement, this is a tool I wish the EAC would develop.

There are some statistical modeling packages out there.  Crystal Ball, by Oracle, has some capabilities and I've tried to look at R (that's what it's called, R) as a hobby, but you can imagine that any statistical package that goes by a letter already insists upon itself.

(Oh, a side note on the "fix that," comment--there has been no shortage of ideas put out by the media, scientists, and election administrators on how "that" might be fixed.  There have so many ideas, in fact, that I'm growing concerned that fixing "that" likely won't happen, primarily because no one seems to be stopping and agreeing upon what "that" is.  The entire community is in solutions-mode without a problem statement).

"That" in Johnson County terms is "this":  What's the right mix of advance voting and voting at the polls?  Does that mix change by type of election (should we offer satellite advance voting for the lower turnout spring elections, for instance)?   What role, if any, should we have in our office to steer someone one way or another, voting in advance or at the polls?  If we think we should have a role, why?

I'm hung up on the fact that 50 percent of our 2008's vote came in advance and in 2012, 43 percent was in advance.  That 43 percent is still big, but in real numbers, it put more than 15,000 people at the polls that we thought would have voted in advance.

If we knew they were coming for dinner, we'd have put a place-setting at the table.

Actually, I've simplified that statement for effect.

In our case, we were prepared for that greater turnout because we always build the election with more machines that we need, but it was an exhausting day for our workers.  We encroached our buffer.

More importantly, 2012 highlighted the lack of control we have in preparing for turnout and no one--in this election--is talking about the other side, planning for a higher turnout than occurs.

There's a cost there.

One of our county commissioners, a former mayor, once wisely compared planning for elections as buying a summer's supply of snacks for the pool.  Terrific weather, and we'd be back needing more money for more snacks.  If the summer was cool and rainy, expect a lot of chips to be wasted.

The uncertainty, and the costs that come with uncertainty, are a big deal.  I don't know the specific issues nationwide, but lack of funding, or at least reasonably cautious spending, was probably a root cause behind "that."

So, nationally, whatever "that" is, I don't think it can be solved as simply as some pundits are making it.

The answers to "that," for me will come from the elusive Magic Math Machine.  More on that soon.

Harry Mullins said...

Thanks for your diligence Brian! I don't have any magic answers, but I certainly hope you can come up with the correct ratio. If you do, this information will be very valuable to have for the campaign teams also.

Thanks again.

Anonymous said...

Love your Magic Math Machine story! Your personality really comes out in your writing and it feels like I'm in your office again, like it was 2002 all over.

JJJJJJulllliiiieee

Election Diary said...

Harry, thank you for all of your support to our office.

Julie, you make me misty. 2002? That's when you told me about your views on RIM, and then you had to nudge me to let me know RIM made Blackberry. And now, RIM really means RIP. You probably called that in 2002, also.

;