O-H-I-O. Not Just a Song By the Pretenders. Or CSNY.

Long-time readers of this blog – all two of you – might remember that in 2008 I was posting a lot in terms election prognosticating.  I haven’t done nearly so much this year.  Part of the difference is that I’ve been flat-out busy the last month or so with paying work and the other part is that I’ve concluded there’s almost zero value I can add to a discussion on this topic.  Everybody I know checks Nate Silver every morning (and afternoon and evening) along with Real Clear Politics, Votamatics and the other sites that aggregate, evaluate and weigh polling data.  What was once the purview of high-priced consultants and their client campaigns is now available to all of us for the price of a mouse click.

As of this morning, Silver is making the following predictions:


A 50-vote margin in the electoral college seems like a pretty good margin, but what Mr. Silver knows – but what few of his readers seem to appreciate – is that the margin hinges almost entirely on Obama’s very thin lead in Ohio (like Silver, I believe Romney is likely to win Virginia, North Carolina and Florida).  To put it simply, if Obama wins Ohio, he wins re-election as the only way Governor Romney can then win is to win every remaining toss-up state in which the President currently leads in the polls.  Ain’t gonna happen.

By contrast, however, if the President doesn’t win Ohio, we’re talking about a very narrow Obama victory – 272-266 – at best.  A Romney win in Ohio, by contrast, flips the script that’s been playing out all fall: suddenly it’s the President who has the much more challenging task of winning all of the remaining toss-up states.

Ohio’s role as the fulcrum in this election is nothing new, but the stakes there are increasing with each passing hour.  And, with Sandy taking east-coast campaigning off the table for at least another couple of days, the campaigns may as well just move there for the duration.

Mr. Silver’s analysis of Ohio polling gives President Obama a 2.6 point lead over Governor Romney in the state, a lead that, according to his methodology, indicates that the President should win the state 73 percent of the time or about 3 times out of every 4 tries (for those who would find comfort in this number, please remember that the odds in Russian roulette are 5 times out of 6 you live so don’t get too complacent).

The Romney campaign’s latest effort to move the needle – an ad seeking to scuff up Mr. Obama’s assistance to the auto industry that is so false (see here, here, here, here, here, here) that it may well be the most inaccurate ad of the cycle by either candidate or party (there are some whoppers from the independents that are worse) – is an indication of how important the state has become and how desperate Team Romney is to move the numbers.

With a week left and early voting well under way in Ohio, it’s hard to predict what – if anything – can change the dynamic of the race there, but let’s for a moment give Governor Romney a win in Ohio.  Assuming that win comes with more likely pick-ups in Florida, Virginia and North Carolina, the only path left for the President to get to 270 is to sweep the remaining battleground states of New Hampshire, Iowa, Colorado and Nevada (just as I gave the Red Team three states, I’m giving the Blue Team a win in Wisconsin).  That combination would lead to a 272-266 Obama win that would look like this:

But…that fallback position is by no means secure.  Mr. Silver assesses the chances of an Obama win in Nevada to be 80 percent, 71 percent in Iowa and 70 percent in New Hampshire, but in Colorado the split is 55 percent to 45 percent, meaning that if you held the election 20 times, the President would win 11 times and the Governor 9 times.

Much better to win Ohio and be done with it. Or to surprise me with a win in Florida (which would be instant game over but is also where Mr. Silver gives Governor Romney a 65 percent chance of winning).

Whatever happens, I’m looking forward to election night as usual.  I’ll be pulling out the TVs, rigging the laptops and making chili.  It beats watching the World Series.

– Austin

29 thoughts on “O-H-I-O. Not Just a Song By the Pretenders. Or CSNY.

  1. Dennis Lang says:

    I’ve been waiting for the always anticipated Austin pre-game analysis. As I recall you had the celebrated Mr. Silver in a deadheat (or better) in 2008. But I believe from photos provided of the Austin prognostication laboratory, your technology had Silver slightly outgunned.

  2. I got close to matching Mr. Silver last time in part by plagiarizing him so I’m not going to pretend that I play in his league. His analysis this time, though, is so good (even if there are legitimate quibbles about his methodology) that there’s no use in even pretending to compete. Far better for all of us that I just say, “Whatever Nate Silver says…” and be done with it.

    I did just buy his book. It’s on my post-election reading list.

    – Austin

    1. Dennis Lang says:

      Self-deprecation admirable but astute observers will remember Austin has no trouble at all “playing in his league”.

      Personally, I never thought this election could be as close as its become. The Obama ground force better be very good at getting people out to vote.

  3. Brian Lambert says:

    Did I make my official prediction yet? And who is holding the SRC betting pot? Benidt, in a crab trap somewhere off Tarpon Springs?

    Obama by 1.5% and … well like Silver says, around 290 electoral votes. Iowa and Colorado seem iffy to me.

  4. PM says:

    I will be taking a cooking class on election night……something new and different! (actually, something to do until the returns from Ohio start to be close to complete)

    1. PM says:

      Yes, he is being trashed by people who do not like what he has to say. They really do not have any good arguments against what he has to say, but that doesn’t mean that they like him saying it.

      Basically, Silver has pointed out that there is not a “horserace” narrative going on, but that the race has been both close and unchanged for quite some time–no Mittmentum, etc.

      Bottom line is that the critics really don’t seem to understand statistics–and, often, even algebra.

      1. Erik says:

        The critics are skeptical of Silver’s faith in state samples that show Democrat turnout exceeding 2008 levels.

        Is that a good argument? Do they misunderstand statistics? Do they misunderstand algebra?

      2. PM says:

        Here is a good explanation of the problems people have with election statistics:


        People who are criticizing Silver clearly do not understand what he is saying–that if the election were held today, given what we know now, and there were 100 iterations of the election, Obama would win 72 times. But that means that Romney would win 28 times–a .280 batting average, which is respectable.

        If a .280 batter comes to the plate and a hit will be the difference between a win and a loss–there is still a good chance of winning. in that sense, calling the race “too close to call” and saying that there is a 72% chance of Obama winning are not inconsistent.

        I am hoping for an Obama win. I think that Obama will win. But I also think that the race is extremely close, and i certainly can’t say that Obama will win–but the odds are in his favor (not Romney’s). I’d much prefer to be in Obama’s position as opposed to Romney’s.

      3. Erik says:

        Baloney. As usual, you’re ignoring the substantive critique to say the conservative critics “misunderstand”. Which is a strawman and a veiled insult.

        The critics are saying the underlying polls are bad, thus he’s got no basis from which to call an 80% probability.

      4. PM says:

        The critics are saying that the polls are wrong? How are the polls wrong? Are they falsifying the answers that people give? Is there some kind of conspiracy at work here (the lame stream media, again?) Are they also complaining that the sun is rising in the east and setting in the west?

        Just how are the polls supposed to be wrong? And how could so many polls be so consistently wrong?

      5. PM says:

        Interesting read. I think he had a good point where he noted that the results of the election will verify (or not) his analysis, given his confidence in a massive Romney win.

        Frnakly, given that the national tracking polls are also moving in Obama’s direction (thus coming in to line with the state level polls) i think that he is massively wrong…..but I certainly could be wrong.

      6. PM says:

        Now both Rove and Morris are trying to cover their ass, and come up with reasons why their predictions of a massive Romney win will not come to pass….


        (given the massive size of their asses, this is quite an undertaking)

        WSJ poll has Obama up by 6 points in Ohio and up by 2 points in FL.


        Still, Romney could win…..no slam dunks here, no sure things.

  5. PB says:

    If Ohio does end up being the “decider” state, we are in for a long wait (and a likely battle) for the answer to who wins the race. Due to a very large number of absentee ballot requests and the likelihood that many will not be sent back in, there could be a very large number of provisional ballots. By law the provisional ballots cannot be counted until November 17th. Eric Black has a good run down: http://www.bit.ly/Q7baRB

    That would be a real treat after an already long campaign.

    1. PM says:

      Yes, i saw that. Given that Rove has a stake in a specific outcome as opposed to accuracy of his prediction, I tend to doubt his analysis.

      All of the betting sites (intrade, etc., where people stand to make or lose money based on their predictions) echo Silver, etc. They see an Obama victory as more likely (in the range of 60 to 75%). And Silver has no stake in the outcome–just his reputation and business on the line.

      Do you know if Joe Scarborough has taken up Silver’s bet yet?

    1. PM says:

      Sam Wang’s site is great. Very close to Nate Silver–but a bit less restrained in predicting a Romney loss.

Comments are closed.