Alabama Presidential Predictions

I've been doing a lot of spreadsheet research lately in my effort to determine who is going to win this upcoming presidential election.

First of all, I took vote information from the 2000 and 2004 elections, broke it down to into Democrats, Republicans, and Unenrolled voters and then I came up with a formula for 2008.

This year, I changed Republican turnout in 7 different scenarios.

Firstly,
In 2004
Republican Identification was at 48%, breaking 99-1 for Bush
Democratic Turnout was at 34%, breaking 92-7 for Kerry
Indy Turnout was at 18% breaking 66-29 for Bush.

The first thing I did was lower Republican turnout to 43, while raising Dem Turnout to 37 and Indy turnout to 20. For both parties, I had them break 95-5 for their respective candidates and I had the unenrolled voters split 60-40 for Mccain. Under this scenario Mccain beats Obama 55-45.

In Scenario 2, 4-7, I had the Republicans break 85-15 for Mccain, and the Democrats break 91-9 for Obama. I kept the the unerolled split the same however.

The differences in scenarios?

#2: Republican turnout -6, Dem +4, Indy +2
#4: Republican turnout -8, Dem +4, Indy +3
#5: Republican turnout -9, Dem +4, Indy +5
#6: Republican turnout -4, Dem +3, Indy +1
#7: Republican turnout -3, Dem +3, Indy unchanged.

In every single scenario but the third one, Obama loses by between 2 (Scenarios 4/5), and 6 points (Scenario 7)

HOWEVER, In scenario 3 it all changes. Scenario 3 is where Obama wins.

To get an Obama win,
Republican Identification needs to lower to 41% while breaking at 85-15.
Democratic UD needs to hover at 38% breaking 91-9 for Obama,
Unenrolled Turnout needs to have Mccain ahead 55-45.

In this scenario, Obama beats Mccain in Alabama by 8,000 votes. Giving Bob Barr 3% brings the election ridiculously close but still doesn't give Obama a win.

After doing this scenarios, I realized it was based on a lot of What-If scenarios so I sought out to do it differently.

In V2 of My elections prediction kit, I took results from 2000 and 2004, separated out the black/white votes for each party and broke it down by percentages. Following that, I used the formula acquired to figure out the results of 2008 in Alabama.

It turned out to be bad. Obama loses 54-46 which is still remarkably close to the results from my first scenario.

The Good news here though, is this; If Republican Identification continues to freefall as it has been, Alabama will be a blue state by 2016.