% | $
Quotes you view appear here for quick access.

Xinyuan Real Estate Co., Ltd. Message Board

  • hmmm26 hmmm26 Nov 6, 2012 5:34 PM Flag

    Midwestern Firewall; Or, Why Obama's a Lock

    I've reviewed every poll in existence on this election, and here's the bottom line: unless the polls have systematically mis-estimated their voter turnout models -- and that hasn't happened in 25 years -- Governor Romney's toast.

    I. Obama's Midwestern Firewall.

    Commentators on TV today are focused on the wrong states. Let's say Romney wins both true tossup states: Virginia and Colorado. Let's also assume he wins all four leaning tossup states: Iowa, NH, NC, Florida.

    Even if Romney goes 6-0 in those battleground states, he STILL LOSES this election unless he also pulls off an upset in one of Obama's Midwestern Firewall states: Penn, Ohio, Mich, Wisconsin, Minn.

    II. Why There's a 91% Chance the Firewall will Hold.

    Obama's "poll of the polls" lead in the Firewall states is between 3% and 8%. Leads that size a month out from an election might fairly be considered only slight leans or tossups.

    But, that's NOT still true when a candidate has the same size lead only a day before the election. When a candidate's down 3% to 8% heading into election day, his only real hope is that the polls are wrong. That does sometimes happen, in races with very few polls or races with very few polling firms, but it's a practical impossibility in a US Presidential election.

    Worse, from the Romney point of view, breaking through in the Firewall states isn't 5 different shots at an unlikely outcome, as pundits would have you believe. Instead, to win ANY of them, the polling data's got to be systematically wrongly sampled everywhere (best hope, if you're a Romney supporter, is that every polling firm has over-estimated African American or Hispanic turnout).

    [Hi, Walt. I know you're the only person still reading this far down the post. You agree this thing's a lock?]

    So, instead of having 5 independent shots to break through the Firewall at the following likelihoods (using Nate Silver's numbers): 9% chance in Ohio, 2% chance in Penn, 2% Wisconsin, 1% Mich, 1% Minn, because all of those upset chances rely on the exact same phenomena -- systematically skewed polls -- they either all exist, or none of them do.

    If the 9% chance doesn't hit in Ohio, it means that the polling data's not systematically skewed, and all the other states' chances for upsets were always 0%.

    III. Even if Romney Does Breach the Firewall, He Might Still Lose.

    George Will and some other Republicans I respect have picked Minnesota as the potential weak link in the Firewall, owing to the fact there's a gay marriage amendment on the ballot. My own opinion is that Wisconsin might be the best of the long shots for Romney.

    So, let's say Romney did breach the Firewall by pulling off an upset in Minn or Wisconsin. He still doesn't win the election unless the President is unable to replace the upset state's electoral votes -- for Wisc or Minn, that'd be 10 -- from those 6 tossup states we just assumed Romney would sweep at the beginning (Va, Col, Iowa, NH, NC, Flor).

    Therefore, even if Minn did provide an upset due to the gay marriage thing, all the President has to do to right the ship is any of the following:

    1. Win Virginia, OR
    2. Win NH + Iowa, OR
    3. Win Col + NH (Col alone would produce a 269-269 tie -- chaos!)
    4. Win Col + Iowa

    IV. Conclusion: Romney's Toast.

    Governor Romney's only remaining hope is that the polling data's systematically skewed. The bad news is that hasn't happened since at least 1980 (as far back as I researched), and, in my opinion, really couldn't happen today, due to the massive expansion of polling firms and polling data over the last 30 years.

    If there's a silver lining for the Romney campaign, it's that if he does hit the lottery tonight on systematically skewed polls, he won't just win the election, he'll win it in a landslide: 350 EC votes, something like that.

    SortNewest  |  Oldest  |  Most Replied Expand all replies
6.44+0.04(+0.62%)4:02 PMEDT