As many Coloradoan's watched their beloved Broncos crush the caught and identified unethical Saints, (isn't that ironic but then again just read Chris Hayes recent book about how absolute competition leads to all forms of cheating), anyway I was crunching numbers.
First some good news, Mail in Ballot and Early vote casting is closing according to party ID lines. Both parties have now reached 41% of all Permanent Mail In Voter (PMV) with D's in CO now within .5%. Early vote as of Saturday is now moving in D direction where in both cases there are upticks in rate of change progressions. (I have been asked not to put up precise numbers as they are not official from the Secretary of State Office.) Voter turnout by PMV is running in two day intervals 10-23 (20%), 10-25 (29%) 10-27 (41%) and Early Vote on those who are active registrants not PMV, 10-23 (7%), 10-25 (10%), 10-27 (16%).
As importantly are the Unaffiliated voting block which is running about 32% on PMV voters and 10% of active non-PMV voters.
Interestingly I pulled up my old spreads from the 2010 US Senate Bennet-Buck race and these voter turnout rates are eerily very similar, the later surge of D PMV voting in the second week of PMV voting as well as the similar spread between partisan registrants and unaffiliated. To review Bennet won 48.1% to Buck 46.4% (while 3rd Party candidates polled 5.5%* remember this, Colorado electorate is splintering on the edges.)
Now my number crunching prediction.
Looking at the aggregate polling where Obama has a 1.2% advantage in his post: State Poll Averages Usually Call Election Right so I went through my historical data and projected voter turnout, Gallup predicts voter turnout to be the similar to 2008, and then factoring in increased voter registrations from 2008 Colorado 2008 Active Voter, 2,587,835 (adding in inactive 3,204,000) to 2012 Active voter 2,700,182, (Plus inactive 3,644,344) increase of of 112,347 active voters.
Cross referencing 2008 and 2010 partisan ID turnout with exit polls on voting preference and turnouts with projected turnouts in partisan ID and polling preference averages I came up with a projected vote and result. Furthermore understand when percentages are this close 3rd Party candidates have a profound effect.PPP poll gave the clearest projected effect;
When Gary Johnson's included he gets just 2%. He's polled at higher levels over the course of the year but this is what tends to happen with third party candidates as the election gets closer- their support evaporates. That's why we have tended not to include Johnson in most of our polling over the course of the year. Johnson draws slightly more from Obama than Romney, leaving the President ahead 49-46 when he's in the mix.
I then looked at the
2010 Bennet Buck race where 5.5% of the votes bleed to 3rd Party and Independent candidates. I decided to look at 2008 and 1.6% bleed from Obama-McCain race. Therefore I gave the 3rd Party number of 2.7%, and taking the accumulative effect of 52% to 48% from Obama as PPP said Johnson will have the largest impact.
I then went to Lynn Vavreck of UCLA who said that on Up With Chris Hayes that women undecided voters were splitting 7/3 for the President. I lowered the number to 65% and in the PPP poll they stated that 6% remained undecided in the Unaffiliated voter universe.
Crunching the numbers where again I deflated and accounted for D's voting for Romney (11%) and R's voting for Obama (12%) and figuring traditional turnout for Unaffiliated and split via PPP's 54-43% for Obama among independents, (started with 51% on expected turnout minus undecided 6% and then added in the 6.5/3.5 split) and then factoring in the late split towards Obama from unaffiliated independent women and I came up with the following:
1,306,916 votes for Obama 49.5%
1,272,855 votes for Romney 47.8%
Johnson 2%
Others 0.7%
This is almost what Bennet-Buck was (1.4% in 2010), 1.7% now