Friday, September 19, 2008

IMW Electoral Vote Prediction #1

I will do these electoral projections every Friday between now and November. Using my own proprietary formula (I'll tell you after the election), IMW projects the following Electoral Vote:

McCain: 315
Obama: 223

We shall see how this changes over time.


g. stelzer said...


Seems like a highly processed product you're feeding us. You know what else is highly processed?...baloney.

I'd like to see at least some break down of the electoral votes analysis. What's your opposition to showing the breakdown, at least in the battleground states? (PA, OH, FL, MI, MN, WI, VA, NV, CO, NM)

Or, does your magic formula not even take individual states into account?

You may be running a very interesting model, perhaps including global cooling measurements. I don't think you need to expose your method, just a little more data.

The Iconic Midwesterner said...

Ugh...that would take a lot of time to break down state by state...(to type out I mean..yes the formula uses individual states).

IN case you are wondering about the states you list:





These things are fluid and I expect some states to swap back and forth...but thats the way the formula spits it out right now.

(And yes, Political Science and sausage making have lots in common.)

g. stelzer said...


Better'n a magic 8-ball. :)

The Iconic Midwesterner said...

Well...when you put it like that every one of these states (plus Indiana and New Hampshire) are in the "Ask Again Later" category.

g. stelzer said...

Perfect. Then, each week, all you need to post, in addition to the total EVs, is the movement (if any) on these 12 states between candidates.

If your model doesn't introduce excessive variability, it should only be a max of 4 states swaying per week; and less than that most weeks.

I'm piqued.