Election Predictions

Open invitation: make your prediction what the final results will be for the Presidential election tomorrow, both popular and electoral. For a bonus, add how you think it’ll wind up in House and Senate.

Here’s my prediction: to my surprise (because if you’d asked me a week ago I would have given a very different answer) Barack Obama with 51% of the popular vote and about 310 electoral votes, Mitt Romney with 48% of the popular vote and about 230 electoral votes (yes I’m aware that adds up to 540, it’s because I’m rounding and I said “about”). Democrats hold the Senate, and Republicans hold the house.

I freely admit that I am betting that Nate Silver is correct in his predictions, and this is largely what I base my prediction on.

What do you guys think will happen?

(Anyone who starts a catfight or ranting and raving gets the sledgehammer. Just say what you think will happen and let’s all see how we did when it’s over.)

  • Dishman

    I’m seeing Romney at 52-47, taking 318-355 EV. There’s an outside chance of 362 EV, but I don’t think it’s likely.

  • Elizabeth Reid

    I think Obama is going to win, but I think it’s going to be closer than you think it is. I think there’s even a chance he’s going to lose the popular vote and still win the EC, although I think that’s less likely now than I did last week.

    This is a cool widget by the way, it lets you pick the winner in various swing states and see how the EC count comes out.

    http://www.nytimes.com/interactive/2012/11/02/us/politics/paths-to-the-white-house.html?smid=fb-share

  • Eric Rall

    Based on the assumption that the consensus of national polling is correct and the state-level polls are for some reason biased slightly towards the Dems:

    Obama 277 EV, 49.5% PV
    Romney 261 EV, 49.0% PV

    Dems hold the Senate 50-48-2, with both Sanders and King caucusing with the Dems.

    Little or no change in the House.

  • http://www.jaeddy.com John Eddy

    Romney/Ryan 377 53%
    Obama/Biden 161 46%

    Mostly as a result of BGV’s* and some seriously broken or outdated polling methods.

    *Broken Glass Voters, as in “crawl naked over broken glass to vote.”

    And if I’m wrong and Obama wins, I swear to G*d Wednesday morning I will get up, take a shower and go to work.

  • http://madisonforum.net/ Sandi

    EC: Obama 237 and Romney 301. Not going to speculate on the popular vote, but will be a substantial margin for Romney. The enthusiasm is much higher on the red side, and the republicans have a better ground game.

    The interactive map is here on CNN.

  • Scott

    Sandi had my old Totals, yet I don’t know how I calculated them. I don’t think Romney is going to carry Nevada given the democrat machine there but I’m giving him Pennsylvania for a 50.8% R to 48.3% O win.

    I do think most of the state polls are heavily and unrealistic ally biased towards the Dems. I think a fair reading of the percentages of each group shows that.

    At this point it’s anyone’s guess and no one knows for sure. Except Nate Silver who just thinks he knows for sure.

  • http://www.deanesmay.com Dean Esmay

    Nate Silver gives 85% confidence, which isn’t exactly a claim he knows for sure. My take is he’s an honest statistician giving what he thinks will happen. Let’s see if he’s right or if his methodologies are flawed. (One flawed methodology might be trusting the state sampling, obviously.)

  • Scott

    [irrelevant out-of-context comment removed]

  • Scott

    [irrelevant ad hominem at parties not present removed]

  • ArnoldHarris

    Rather than coming up with potential wild and meaningless guesswork of my own, I’m more impressed with Rasmussen’s and Gallop’s 49-48 in national tracking, plus information from these and other polls which show Rasmussen ahead in Michigan, Pennsylvania, Wisconsin and numerous other states which Obama took in 2008 and he presumably needs in order to win tomorrow.

    In any case, Romney-Ryan are in a well-planned and well-played ground game in the swing states, while Romney-Biden are strictly on defense.

    I hope this doesn’t go into numerous extra innings as in 2000 in Florida. But on the other hand, the issues facing this country as so momentous that all of us would be wise to await final counts from all the mailed-in votes.

    With the president pulling only 48% of the likely and now all but certain voters, just one day before the national election, the situation does not bode well for any incumbent office-holder, and especially one with as much hostility going against him as Obama has generated. I think most of the uncommitted 3% of the voters will go for Romney. Which mean he will capture more than 50% of the popular vote. And if Romney does as well in the swing states as incoming polling data now shows, Romney will take the electoral as well as the popular vote.

    I also have a feeling that if Romney and his running mate win, the American people will not be anywhere near as disappointed four years from now as we have been after four years of Obama and Biden. And I sincerely hope this last remark is not a wild and meaningless generalization of the type I warned about in the first sentence of this comment.

    Arnold Harris
    Mount Horeb WI

  • Buddy

    Arnold,

    Oh I think this has the makings to be worse than FL if Ohio is the ‘decider’ and it’s a very close margin:

    http://www.cnn.com/2012/11/05/politics/ohio-ballots/

    http://www.nationalreview.com/articles/332461/ohio-counts-so-waits-nation-john-fund

    If it’s close, we’re going to be waiting for a long time (the provisional ballot count doesn’t even start for 10 days after the election) and running through another fiasco as bad as the Washington Governors race in 2008, except at a national level.

    If it’s close in OH, its going to be a charlie-foxtrot, no matter who wins.

  • Buddy

    P.S. as for my outlook on the results, it seems there is a bit of Democrat oversampling in the polls, and if that turns out to be true, this might not be as close as it seems. Right now, though, I think it’s basically toss a coin.

    Personally I’d like to see Romney win if just to watch Chris Matthews melt down tomorrow night, but that’s a little juvenile on my part.

  • http://blog.beliefnet.com/cityofbrass Aziz Poonawalla

    my prediction is laid out in detail here:

    http://www.patheos.com/blogs/azizpoonawalla/2012/11/2012-election-prediction-obama-303-to-romney-235/

    in a nuitshell, Obama 303 EV to Romney 235.

    Sandi – you’re the only other person in this thread to actually show their state predictions, for which I salute you. I do find it odd you left MI in Obama’s column given how you’ve given Romney a clean sweep of every other swing state. Why would Michigan buck teh regional trend? Just curious.

    In my post I do go through some alternatives but overall the picture is pretty grim for Romney unless – as Nate Silver pointed out – all the state polls are systematically biased.

  • http://blog.beliefnet.com/cityofbrass Aziz Poonawalla

    Eric – over the weekend, the national polls converged and are now in agreement with state polling.

  • http://madisonforum.net/ Sandi

    Aziz,

    I didn’t include MI in Romney’s column for the same reason I didn’t include MN. The Romney ground game is not as strong in either state.

    BTW I agree with Buddy that the dems are oversampled, probably based on 2008. Also I made my prediction fairly conservative.

  • ArnoldHarris

    I see that the the latest Gallop tracking poll results — possibly the last of this campaign — show Romney beating Obama 50-49%. The spread between them has been and remains 1%. If this is accurate, Obama cannot win the popular vote. Will the electoral vote paint a different picture? We may not know for weeks, depending on when a final vote comes out of Ohio. Or it could be all over by Wednesday if either candidate team has enough electoral votes from the rest of the states.

    I still think Romney/Ryan will be the winning ticket, unless and until I see evidence to the contrary.

    After all these years of comments, all Dean’s World commenters know full well that I am anything but a liberal. I do vote for a number of Democrats for local and states offices. But for the presidency and Congress, my choice is Republican leadership.

    Arnold Harris
    Mount Horeb WI

  • Scott

    Aziz,

    Ive talked to aomeone who works at Rasmussen and their polls (the most accurate in the biz two presidential cycles running) showed a convergence last week due to Obama’s post Sandy actions but now are goinng back to the +2-3 lead for Romney. But they are a three day rolling poll with romney at Friday-Saturday-Sunday 49-49-50 and Obama at 49-48-47. I expect to see a complete return to a 50-47 romney lead when he turns in his last poll and that, with a lot of the other stats which one can easily infer a Romney lead nationally and in the important swing states, leads me to believe a Romney win is the most likely outcome at this point.

  • Eric Rall

    Aziz — My understanding is that the gap has narrowed a bit, to the point that they no longer suggest different election outcomes, but there’s still a notable gap. Nate Silver’s post on Sunday noted an Obama +1.3 polling average; he didn’t state the national popular vote implied by the state polls, but his now cast (about 2/3 state polls, IIRC) shows an Obama +2.1 win in the popular vote, implying the state polls alone show an NPV of about Obama +2.5.

    Compared to today’s RCP average of Obama +0.7, that’s a spread of 1.8. Not enough to give Romney the lead if the national polls are right and the state polls are wrong, but enough to swing several states and narrow the margin of victory somewhat.

  • Scott

    Pretty good and honest article from an Obama supporter.

    http://blogs.ajc.com/kyle-wingfield/2012/11/05/presidential-prediction-the-candidate-who-defies-history-and-wins-will-be/

    Best comment (by the author)

    As I wrote in the OP, likely-voter polling is an art, not a science. And the other numbers I based my gut feeling on — numbers about enthusiasm, favorability, etc. — are still, you know, numbers. It’s just a matter of which ones you tend to believe. My head tells me these other numbers ought to be more reflected in the topline poll results. My gut tells me they’ll be reflected in the actual voting totals. We shall see.

    There’s an obvious disconnect between all the polling that’s been done with most of the swing state polls on one side and the most reliable state, national, independent preference on the other. Romney leads in independents, Obama’s lead in early voting is way down from 2008 while still being a lead. Crunch the numbers of Ohio Independent’s preference, early voting differences from 2008 till now and the closing turnout margin and you’ll have to come to the conclusion that Romney is going to take Ohio.

    That is, of course, if you’re not adverse to listening to what the numbers have to tell you. I do that every day, just look at the conclusions built on the cold, hard numbers; and I can’t see Obama taking Ohio. it’s just not in the numbers.

  • http://blog.beliefnet.com/cityofbrass Aziz Poonawalla

    I’d like to see a electoral map based on the RCP averages. What does that predict?

    UPDATE ok, i did it myself. Took the RCP averages state-by-state and plugged them into RCP’s own electoral map widget. I get the same results: Obama 303 and Romney 235.

    http://www.realclearpolitics.com/epolls/2012/president/obama_vs_romney_create_your_own_electoral_college_map.html

    RCP averages have Obama taking MI, WI, OH, PA IA, CO, and VA

  • http://blog.beliefnet.com/cityofbrass Aziz Poonawalla

    Sandi, Bud – I don’t really find Dean Chambers’ “unskewed” argument compelling. He seems to basically misunderstand the concept of voter self-identification. Here’s a pretty persuasive argument by Max Blumenthal that lays out the error in that critique.

    http://www.huffingtonpost.com/mark-blumenthal/unskewed-polls_b_1924293.html

  • http://blog.beliefnet.com/cityofbrass Aziz Poonawalla

    Dean,

    Silver’s figure of 80% is not a confidence interval, it’s the output of a monte carlo simulation of tens of thousands of elections, with random drift in the various variables that go into his (well-documented) model.

    He runs the simulation daily with updated numbers from all the various metrics and thats where the “now-cast” comes from. But of course the model weights averaged polls as well into account, to a degree extrapolated from past elections.

    Silver is the first to admit that doing a straight-up average of the polls (like RCP) will get you pretty close to the right answer as well. Its hardly a black box. This isn’t rocket science :P

    monte carlo simulations are described here: http://en.wikipedia.org/wiki/Monte_Carlo_method

    and you’ll have to agree after reading it that they are well-named :)

  • http://blog.beliefnet.com/cityofbrass Aziz Poonawalla

    Eric, I don’t see any reason why the national polls cant have different results than the state polls. After all, thats what happened in 2000. I don’t see why the national polls have any bearing on the state results, and I don’t really accept that they even have any relevance at all to the outcome of the election. The state polls are all that matter.

  • jaymaster

    Close call, but I’m giving the edge to Romney.

    I’m seeing things I never thought I’d see. Catholics and Evangelicals coming together, and both talking to Jews. Tea Partiers and libertarians shaking hands with social conservatives. Big money Wall Street types who funded Obama big time last election have done a 180. Big “L” libertarians admitting they won’t vote “L” this year. Hell, even some Ron Paul Radicals are holding their noses while sticking Romney signs in their yards. And even my redneck, trailer trash, welfare queen family members gave up on Obama when their unemployment ran out and they lost their homes.

    While on the other side, all I see left pulling for Obama are the die hard, bleeding heart liberals and progressives, the true socialists (and communists), union members, and a few youngsters. And they all appear demoralized.

    IMO, Obama lost the middle, and that will cost him the election.

  • fruitylips

    I had been rooting for 269-269 for pure comedy value.

    I’ve changed my mind, tho.

    I want someone to win 270-268 with the winning vote coming from a split EC vote in Maine or Nebraska.

    I predict that neither of these will happen.

  • Scott

    Pure Comedy Value would be if PA, OH, WI, MI, MN all go for Romney indicating a near-Reaganesque landslide and then watching the entire cast of MSNBC (except for the moderate Joe Scarborough) opening up veins and committing Seppuku on live TV in abject Grief over their Fallen God.

  • Eric Rall

    The state polls, taken collectively (i.e. you add up the projected popular vote from each state) imply a certain national popular vote, which is different from the national popular vote implied by the average of the national polls. This matters because this implies that there’s something wrong with one or both sets of polls. We’ll find out which tomorrow.

  • http://www.deanesmay.com Dean Esmay

    To my surprise, here in Michigan the Republicans have knocked on our door at least three times. Now, this is a Republican district thanks to the gerrymandering that’s done so much to corrupt our political system (and which both parties collude on by the way–it’s an obscenity), so it may be that their “ground game” consists of going into Republican districts and urging people to vote. Nevertheless, for the supposedly nonexistent ground game for Team Romney here, it looks surprisingly strong in this household. We may just be an anomaly, maybe there’s just a group of particularly feisty Republicans in this neighborhood that isn’t reflective of what’s going on statewide.

    Just a point of information. I will actually be shocked if Romney takes Michigan and I’m pretty sure if he does it will mean he’s won a crushing victory, because of Michigan goes that way so will Ohio and maybe even Wisconsin and more. (Not that I’m saying Michigan is bellwether for those other two states, it’s the reverse if anything, just saying, a Romney victory seems so unlikely here that it would be indicative of a much larger trend.)

  • http://madisonforum.net/ Sandi

    When all is said and done—and we have been inundated with much of that—the voter goes into the booth alone. What is probably on most minds isn’t ads, polls, or speeches, but the number one issue on almost everyone’s mind.

    The economy.

    That should be a loser for Obama and a winner for Romney. We will know in about 24 hours.

  • Scott

    At the end of the day you have to believe one of two things: (1) Either all the older, more established polls with a good track records are wrong about the race being statistically tied with maybe Romney ahead by a point or three. That independents going nearly 2-to-1 for Romney doesn’t matter and that Obama’s early voting margin, which he still leads in, is significantly down from 2008 doesn’t matter. Or (2) that all the openly-biased polls, many of which are from new firms with no track record and use shoddy techniques to identify likely voters (Rasmussen and Gallup ask detailed questions like “do you know where your polling place is?” or “Have you voted in the last elections?” to establish likelihood while the pro-Obama polls ask “How likely are you to vote?”) are wrong.

    It’s really Occam’s razor: the simplest answer wins.

    And so does Romney.

  • ArnoldHarris

    Sandi, you are right about the voter going alone into that booth today, and that the overwhelming issue on his or her mind will be the economy.

    Public perceptions of how bad economic conditions have become are now ubiquitous. I think they will select Romney and Ryan as the logical choice to start us all back on the road to rebuilding our industrial base upon which a healthy economy depends.

    Here in Wisconsin, I will vote for Tommy Thompson for the US Senate. He did a fine job as governor, including introducing not a few programs for this state that depended upon bi-partisan support which he helped build.

    Arnold Harris
    Mount Horeb WI

  • http://blog.beliefnet.com/cityofbrass Aziz Poonawalla

    an important message from Nate Silver:

  • ArnoldHarris

    Aziz, all things considered, I think Rasmussen and Gallup tracking polls have a better long-term likelihood of measuring public opinion and immediate voting trends than does a statistician.

    They say that the recent past is a good predictor of the immediate future. But I don’t think that rule applies too well to politics in a country with a 16 trillion dollar national debt, official unemployment levels at about 8-9 percent, 27 million Americans out of work, 47 million Americans feeding themselves and their families on food stamps, American diplomatic personnel being slaughtered by Arab al-Qaeda gangs while the president sits on his ass in the White House trying to cover up the truth about the incident, and much, much else.

    Of course, you could be right and I could be wrong. We will know a lot more about this in about 14 hours. Because this is the day of days.

    Arnold Harris
    Mount Horeb WI

  • http://blog.beliefnet.com/cityofbrass Aziz Poonawalla

    Arnold, the polls you cite are inputs to a model. Statisticians designed them, and the results are scientific so we can use their results in models that have genuine predictive power (as has been the case in *every* election).

    “statistician” is not an arcane art, but a fairly mundane and thoroughly understood field. Its not partisan to say, of you have two polls of 500 responses each, combinining their results will reduce margin of error, which scales as sqrt(N) (in this case, 1000 instead of 500).

    all the issues you cite, debt and war and whatnot – these are represented in the polls. Because those are the variables that determine how each statistically sampled voter responds. So its all in there.

  • http://blog.beliefnet.com/cityofbrass Aziz Poonawalla

    more to the point, Arnold:

    It’s important to be clear about this: If Silver’s model is hugely wrong — if all the models are hugely wrong, and the betting markets are hugely wrong — it’s because the polls are wrong. Silver’s model is, at this point, little more than a sophisticated form of poll aggregation.

    from http://www.washingtonpost.com/blogs/ezra-klein/wp/2012/10/30/the-nate-silver-backlash/ which is worth reading in full

  • maggie – labrat

    I got a bad feeling. I think Obama’s gonna pull this off.
    Tell me it’s just because I’m in blue country…..

  • http://madisonforum.net/ Sandi

    Tell me it’s just because I’m in blue country…..

    It’s just because your in blue country maggie. *grin*

    I think some believe in models and polls because the want/need to believe them. As I said earlier when voters get into the booth alone with their thoughts, the economy looms big no matter who they want to vote for.

    Personally I have never been called for any polls until this year, and it now happened 3 times. Twice I lied through my teeth, and once I was busy and hung up. That only accounts for one vote in the poll, but am I the only one who thinks it is none of their business?

  • Ruth H

    Some very interesting thinking going on here. I wish I could say I had an informed opinion, but I am afraid all I have is an opinion. I read all the blogs, I read all the polls and it tells me we have a very divided nation. I can make no prediction. I can tell you I FEAR Obama will will, I can tell you I PRAY Romney wins, but I cannot make a prediction. Maybe if we roll the dice…….

  • ArnoldHarris

    More or less as I wrote earlier this morning, Aziz. We all will know the trendline tonight, when Drudge Report begins issuing exit polls.

    But at 11:05 am CST, only Allah, Karl Rove and Nate Silver seem to know for certain. And Allah, at least, is keeping his mouth shut about American elections.

    Arnold Harris
    Mount Horeb WI

  • Scott

    Its not partisan to say, of you have two polls of 500 responses each, combinining their results will reduce margin of error, which scales as sqrt(N) (in this case, 1000 instead of 500)

    Wow, you really have zero understanding of the field of statistics, don’t you, Aziz. I hate to bring this up again, but I’m a trained and experienced statistician who teaches students (albeit “lower” level stats to business students). You simply cannot mash together two different polls even if they measure the same thing. Before you begin you have to make absolutely sure the polls were conducted in the same way and then even it’s not so simple as to “reduce the margin of error by Sqrt(N)”.

    Wow. I’ve seen a lot of statistical myths many people seem to hold in my time but never from someone who says he has a good understanding of the field.

    Wow, just wow. Let me ask you, do you believe that the Central Limit Theorem means that you have to have a sample size of 30?

    Wow.

  • Scott

    maggie-labrat,

    From the beginning of this election cycle the media has been playing the theme of “Obama’s re-election is done deal except for the formalities, why are you people fighting it?”.

    Seriously, look at Nate Silver. He called his own paper “stupid” for putting forth the idea in one of their articles last week that the race was “too close to call”. See, someone didn’t toe the party line.

    Heck, even today we have reports that the Obama campaign is going to declare victory early.

    It’s all about depressing the opposition vote and to set up a belief early on that it’s inevitable so just don’t fight it.

  • http://blog.beliefnet.com/cityofbrass Aziz Poonawalla

    You simply cannot mash together two different polls even if they measure the same thing.

    on the toplines, yes you can. As long as teh basic questions are the same, it will work.

    even it’s not so simple as to “reduce the margin of error by Sqrt(N)”.

    actually, it is. The standard error is just the sample stdev divided by the sqrt of teh sample size. Larger sample size means smaller std error. If you’re a stats teacher then this should not be new to you.

    the CLT doesnt really inform minimum sample size in this context., It just justifies the assumption of the normal distribution.

  • Scott

    on the toplines, yes you can. As long as teh basic questions are the same, it will work.

    Uhm no, you can’t. I don’t have the time (nor do I have the responsibility) to educate you on your glaring errors in understanding statistics, but suffice it to say you’re one of the many statistics users I encounter every day who have a deep misunderstanding of the field.

    Please, educate yourself. Go take a class or two on basic Mathematical Statistics to understand why you can’t simply mash together the results of two random samples with different results (heck even with the same results). Oh, and please look into Sampling Theory to understand what happens to the standard deviation when you do mash together two different polls.

    But at least you got the question on the Central Limit Theorem essentially correct.

  • Scott

    I’ll give you this since it came up in both the Graduate-level Bayesian class I am taking and the Business stats class I am teaching. If you take two separate polls with two different results for Romney (say 50% +- 3% and 49% +-2.5%) what you end up with is a union of the confidence intervals. Meaning, in my example, you end up with a confidence interval from (46.5%, 53%). You don’t gain precision when you combine polls you lose it.

  • http://blog.beliefnet.com/cityofbrass Aziz Poonawalla

    ok, Scott, I’ll do another round (though I am also puzzled over where you pulled that number 30 from).

    you’re calling the spread a “confidence interval” which is nonsense. Its just a statement of the mean plus/minus error. And you cannot “merge” two polls in that way.

    By increasing sampling size, you are decreasing standard error in the mean. Two polls may differ on method in some ways, but if both have n = 500 and both ask the question, “will you vote for Obama yes/no? Will you vote for Romney yes/no?” then we can combine teh answers from those questions. (there are a set of questions common to almost every poll).

    since you have 2 sets of polls of n = 500 asking essentially the same question, you can pool that data into one bigger poll of n = 1000. the error calculation (that gives you the spread) will then be different – not just the crude addition you gave as example, but something else based off of dividing by sqtrt(n).

    actuaklly election polling is slightly different from say measuring a physical process. The formula for std error isnt going to be just sigma/sqrt(n) because there actually isn’t a sigma per se. Instead you have a percentage p (ie, 51% support Obama) and the sampling size N (say, 500). The std error is then basically sqrt[(p(1-p)/N].

    lets say we have two polls, each n = 500 and one finding 51% support Obama and the other finding 49% support Obama. That means
    255 people voted for Obama in poll 1 and 245 people voted for him in poll 2.

    The std error for each poll is then 2.2% for the first poll and 2.2% for the second.

    (yes, identical. Do you see why?)

    now, combine them. We have 245+255 = 500 votes for obama out of 1000 votes total. That’s 50%. Again, std error is easy to calculate: 1.6%.

    This is, as you put it, “basic statistics” but if you didn’t understand any of this before then I am a little skeptical of your claim to be teaching stats or taking a “graduate” level course.

    I see you threw in the buzzword “bayesian” which is very fashionable but all it means in context of poling is you can take a-priori performance into account. Thats where teh models differ (and includes LV models, economic indicators, etc). Make sure you grok the std error above before you get into the weeds on Bayesian statistics, though. Likewise before I try to explain Monte Carlo simulations to you.

  • http://blog.beliefnet.com/cityofbrass Aziz Poonawalla

    btw – just for teh real stats aficionados who are wisely staying out of this – i assumed that we approx the bionomial distribution with a niormal distribution. Its possible I am wrong about that and the formula isnt the one I used above. But I think the CLT covers my ass on this. Am I right?

  • Scott

    Ugh, please Aziz stop while you’re behind.

    And yes, I am a professional statistician who has taken both Math Stats and Advanced Math stats and aced most of it. Sampling Theory and Regression etc. etc. It’s really quite rude of you to dismiss all my years of effort and blood, sweat and tears (in mostly the literal sense) simply because I’m showing you up for what you don’t know squat about.

  • Scott

    Any anyhow you wouldn’t use binomial in this case since there’s more than two possibilities. You’d use a Multinomial (you have to account for third party candidates) but even in that your explanation is horribly wrong.

    Just stop, Aziz.

  • ArnoldHarris

    Aziz, Scott is correct. Even as far back as 1961-1962, when I was completing the first of two degrees that I earned at the University of Illinois-Champaign/Urbana, I studied enough about survey research and survey design to understand that different polls with different samples cannot be matched in terms of results without negatively changing the confidence levels. At which point the survey results become little more than propaganda.

    Why don’t you sit still and cool it for another 4.5 hours? After 8pm, you and all the rest of us will be able to see the carefully guarded exit poll results. That should more or less tell you whether your hero of the hour stays in office or is replaced.

    Arnold Harris
    Mount Horeb WI

  • Scott

    And I only bring up Bayesian because they calculate CI differently than the “normal” Frequentist method does (and note, Bayesian has a lot more influence in statistics than just building a posterior distribution given a prior distribution and a current one) but there are some assumptions both schools agree on.

  • http://blog.beliefnet.com/cityofbrass Aziz Poonawalla

    Arnold, I hate to contradict you in public, so I’ll just say that if you google Gallup’s methodology, or any variation of “poll sample size error” keywords, or lookup the definition of sampling error on Wolfram Mathworld, or if you look up “margin of error” on Wikipedia, or read the “Statistics and Political Polls” article on About.com, or basically consult any other reference, you will see that my simple explanation holds quite true. If theres another authority you prefer about polling or about statistics or about math, then consult that instead.

  • Scott

    (edited out because Aziz edited out his rude comment)

  • Scott

    Aziz,

    I’ve studied some of the best literature in the field of statistics on this, both theoretical and applied. Anyone who has expertise in a field knows that open-source internet resources like Wiki and About.com are riddled with errors (ESPECIALLY in the field of statistics. The misconceptions that people hold on the field are truly legion and the number of people who get most of the basics but still have horrendous misconceptions are practically uncountable).

    So, in short, I don’t trust sites like Wiki and About.com to teach me something I already know better than they do.

  • http://www.deanesmay.com Dean Esmay

    Aziz and Scott, may I invite the two of you to actually go ahead and debate statistics and polling methodologies after all this is over? I’d really like people to just make their predictions. Although it’s probably a lost cause at this point, and I am actually finding the discussion interesting and would like to have it, especially in the aftermath of the election, where we may actually be able to look at real data and see how it held up against our predictions and the poll predictions and that of various either predictors like Intrade, the counterintuitive stats guy at the NYT, and all that good stuff.

    Still I’d like to ask you STOP discussing that here and now and invite anyone here who has any NEW predictions they’d like to add, or modifications to their previous predictions, to feel free to share them.

    (I’m not wussing out, I am sticking with my predictions, let’s see how I do, and how the rest of you did.)

  • ArnoldHarris

    Aziz, I’m too old and well-seasoned to get upset about being contradicted in public, which, in any case, is something Stefi does with me whenever she thinks I deserve it. She votes mostly Democrat, by the way, and I vote mostly Republican. Sort of like Carville the Democrat and his Republican wife, but in reverse.

    Nevertheless, I still know you are wrong about what you think is appropriate survey research methodologies. But I admit I probably don’t know more than 30-40 percent of what I think Scott knows about this topic.

    As I wrote earlier, why don’t you just cool off and await the exit polls or even the real voting results, which, obviously, may not be known for a couple of weeks.

    One thing we can all agree on. There will be almost greater ill will after this election than there was leading up to it. I think the social order in this country is fracturing irretrievably.

    Arnold Harris
    Mount Horeb WI

  • Scott

    Sorry, Dean. And no problem. If you want to open a stats discussion I’ll join. Though I won’t know how responsive I’ll be seeing that it’s the middle of the quarter and all.

  • Scott

    Back to prediction. I’m going to predict that, along the lines of that report I wrote about earlier, that the Obama campaign is going to declare victory early in order to suppress the vote of the Romney voters in this very close election. That’s just the kind of scum-bag tactics we’ve grown to expect from this guy.

    Which is why I go back to my statement that we really need the media and the election officials to act like grown ups and not report a single state outcome until every poll in every state is closed. That way everyone who wants to cast a vote can.

  • queenofallevil

    Go big or go home they always say…

    Romney 331 Obama 207

    I think Romney is gonna take some big prizes and I believe Michigan is one of them.

  • http://www.deanesmay.com Dean Esmay

    A Michigan win for Romney would mean a Romney win for sure, I’ll stick with that as a prediction. I will say that turnout was high here in this gerrymandered-Republican district, but there was a very large turnout of black voters to my eye, and while we don’t want to generalize too much we know that’s a solidly Democratic constituency on average, just like Jewish voters and to a lesser extent Hispanic voters (which so far as I know we don’t have a lot of the latter two in this area).

    I once again wish to emphasize that I am not saying Michigan will be the decider state, I am merely predicting that if Obama loses here, he loses big, and further, it will be a very bad night for Democrats in general.

    I am not predicting that to happen, I will be surprised if it does. But if it turns out that Obama lost Michigan, you can probably go to bed without waiting to hear the rest because Romney will be President and Democrats will have suffered a bloodbath. That’s my prediction.

  • Scott

    Heh,

    First Drudge had the results of the exit polls with the caveat of “exit polls tight”. Now he has as his headline “ENOUGH! LET’S COUNT!”. I wholeheartedly agree.

    This is a close one folks. Ain’t gonna be an easy night for anybody. Including the people who don’t care.

  • ArnoldHarris

    Hello, Rosemary. I’m with you in spirit for this election. My wife Stefi voted for Obama. I voted for Romney and for Tommy Thompson for the US Senate. Even so, I don’t bite the hand that feeds me, and we stay in good humor over politics.

    Romney has a lot going for him in this campaign, with simultaneous backing from Evangelical Christians, conservative Catholics, Zionist-oriented Jews, most members of the military, gun owners, small town, rural and suburban populations, and above all, independent voters. If that isn’t enough to pull down a sitting US president who has presided over what has become a major sinkhole of a national economy, than I cannot imagine what else it would take.

    As for Michigan, other than for the Detroit area black population, what other population segment of your state would support Obama against Romney? As guess we will know in a couple of hours.

    Arnold Harris
    Mount Horeb WI

  • http://madisonforum.net/ Sandi

    Keeping tabs on FOX website as well as watching on TV.

    I’m also updating results every half hour with an updated electoral map. Too early yet to get a feel of any trends.

  • ArnoldHarris

    Dean, stay awake so we can argue with you.

    That’s what God created blogsite hosts for.

    Arnold Harris
    Mount Horeb WI

  • ArnoldHarris

    Scott, what do the numbers under “Enough, let’s count!” on Drudge represent? Is that the relative R and O percentages of the cumulative popular vote?

    Arnold Harris
    Mount Horeb WI

  • Scott

    ArnoldHarris,

    That’s my guess, yes.

  • http://madisonforum.net/ Sandi

    Romney leading in Mich Dean 53%-46%

  • ArnoldHarris

    Sandi, where did you pick up that report on MI?

    Arnold Harris
    Mount Horeb WI

  • Scott

    Please don’t do that, Sandi.

    I, for one, don’t want to get my hopes too high only to have them dashed. And, I know you’re not doing it but it comes off as gloating.

  • Scott

    I will say this: the polls which showed a tight or tied race seem to be proven correct so far. The polls that showed an easy Obama victory seem to be false, so far.

    let’s hope Romney wins it! (Well, for those rooting for Romney, that is)

  • http://madisonforum.net/ Sandi

    Arnold, at FOX

    http://www.foxnews.com/politics/elections/2012-election-results/

    That is what I am using for my live update on madisonforum.net Just hover over any state whether called or not it will give you the stats.

  • http://www.deanesmay.com Dean Esmay

    Scott, nah, no worries, I’m not too moved by it, I would merely warn Sandi that in Michigan, the returns tend to come in faster in places like Oakland County (heavily Republican) than they do in places like Pontiac, Detroit, Flint, etc. (which lean or are heavily Democratic). So early returns… not so useful. Don’t get yourself overexcited. Ann Arbor comes in solidly Democratic and will probably be in fairly soon if it isn’t already, but the poorer and the rural areas are going to take longer.

  • http://madisonforum.net/ Sandi

    Correct Dean, only 7% in and already switched.

    53% Obama 47% Romney

  • http://www.deanesmay.com Dean Esmay

    I wish to state now that if Mitt Romney wins, even though I did not vote for him I will do my best never to call him a liar when he may just be wrong, I will do my best not to impute negative motives to him when they may just be things I disagree with, and that I hope the country prospers under his Presidency.

    My gut at the moment though (9:09pm Eastern) is that he’s not going to win; a week ago I was sure Romney had this but Florida’s not looking great for him at the moment and parts of Ohio aren’t going the way Republicans hoped. But let’s see, it’s still early.

  • http://www.deanesmay.com Dean Esmay

    Arguing with myself: still razor thin in Florida. I think Republicans were hoping to be a little further ahead in certain areas and they were but Romney could still take Florida.

    However, Wisconsin and Michigan have already been called for Obama, and probably most important, Pennsylvania. Assuming there are no shocks that reverses those projections, Romney’s in deep trouble.

    (No, Romney fans, I’m not “rubbing it in,” I’m just calling it as I see it, the game is not fucking over.)

  • http://madisonforum.net/ Sandi

    It is down to Ohio now with Wisconsin gone. Whoever wins Ohio is President.

  • Scott

    Yes, and the counties that haven’t reported in in Ohio are all red.

  • Scott

    I’ve been reading Bob Krumm and it really comes down to Cuyahoga County. If Obama can’t get a 230,000 lead from that one county it’s over.

  • http://www.deanesmay.com Dean Esmay

    Slight change of subject, to my (somewhat) surprise but absolute delight, it looks like pro-marijuana referendums are doing very well tonight.

    I’m EXTREMELY pleased by this. That will be something I can walk away from this election feeling VERY happy about.

    I will also be pleased if Bridget Mary McCormack makes it to the Michigan Supreme Court.

    At this point I’m still seeing no reason to move from my prediction on the Presidency or the Congress.

    I’m having a little trouble here, if Romney loses Ohio but wins Florida, he still won’t win? It really is down to Ohio you guys think? (I’m getting too tired to try the math.)

    I also notice that at this writing, Romney/Ryan has a solid lead in the popular national vote. My hopes for Obama to win the Electoral College but lose the popular vote look like a slim chance of happening. It would please me on multiple levels, most of them I admit frankly sadistic but also principled. ;-)

  • Scott

    I’m optimistic about Ohio and let me tell you why. Obama had a “lead” in the early voting since more Democrats voted early than Repubs. Well, they’re through counting the early votes (except for the provisional I think) and the two counties still outstanding are both red. Also, Obama”s lead is shrinking statewide.

    I think it really could be a very narrow victory for Romney.

    [snip]

  • Scott

    Oooooh, one red county left to report and O’s lead is shrinking! Will it be enough for R to win?

    Wait and see, Kiddies!

  • Scott

    Last county in, 55% in and his lead has shrunk drastically.

  • Scott

    58% and still shrinking. At 65K votes ahead.

  • http://madisonforum.net/ Sandi

    Slim possibility that Wisconsin could get reversed. Obama projected winner but still behind by 3% with about a quarter counted. It looks the opposite of Ohio. Also most of the unreporting counties are fairly red.

    Can’t believe Obama is taking Iowa.

  • ArnoldHarris

    Unless there is some political miracle in Ohio, Obama will win re-election. Those of you who have been cheering him on will get a president who will:

    — appoint two more leftist members of the United States Supreme Court for purposes of subverting the original United States Constitution;

    — raise the national debt to 20 trillion dollars in the next four years, creating a situation in which much of the resources of this country must be diverted just to pay the interest on such a stupendous debt; in other words, a large North American version of Greece, Spain, France, Portugal or other once-significant societies;

    — oversee an economy in which the public sector will attempt unsuccessfully to create make-work employment;

    — see tax revenues shrink with a steady stream of private businesses that will collapse, shut down and move their production elsewhere, including out of the USA;

    — attract millions more illegal immigrants to this country for purposes of establishing a permanent class of quasi-Americans for cheap labor and votes for the new Democrat professional ruling class;

    — witness a nuclear war in the Middle East as Iran raises the enrichment level of its uranium stockpiles to weapons grade, meanwhile threatening the State of Israel with extinction — a threat that Israelis will be compelled to answer with an armed attack, possibly requiring use of their own nuclear weapons;

    — the gradual continued social degradation of a population surviving largely on food stamps and other government hand-outs, with burgeoning minority populations forming more than half the country’s prison populations and most of the rest of them jobless and giving birth to children of dubious parentage and no future whatsoever;

    But cheer yourselves with your victory in the name of democracy. Everyone gets what’s coming to them in life. But some of you lack the imagination to think ahead what this country is coming to.

    I see ahead growing and endless social disorders, city streets only barely policed and largely in the control of violent criminal gangs, growing use of dangerous drugs — crack cocaine by blacks, methamphetamine by their white counterparts, shrinking employment in all industries, large cities falling into the decay one sees in photographs of what used to be Detroit, Michigan; growing number of homeless everywhere; never-to-end dependence on foreign petro-energy that one day this country will not be able to afford to import; continued crumbling of the overbuilt physical infrastructure used by private transportation, but with no public transportation to take its place.

    Could Mitt Romney and Paul Ryan have helped stem this terrible tide? At least they had a plan to do so. But Obama will continue offering excuses in that an increasingly hostile Congress will not raise taxes to bring down the wealthy investor class into the common misery that awaits all of you. Nor will there be much private investment under Obama. The smart money will flee this country as fast as possible.

    But cheer yourselves on.

    Arnold Harris
    Mount Horeb WI

  • http://www.deanesmay.com Dean Esmay

    Looks like Obama’s won. Let’s see but it’s looking like it.

    Fox News has just called it for Obama.

  • http://madisonforum.net/ Sandi

    It appears it will be another 4 years of Obama, trillion dollar deficits, high unemployment.

    A stagnant economy, higher taxes for everyone that pays taxes with the healthcare tax.

    No growth because all business including small is burdened with stifling regulations and more taxes.

    A friendly administration to dissident Muslims.

    Because I have no children I can live with it, however I feel a bit of shame in what we ( as a country ) have become. My heart goes out to future generations that will have a Greece like economy.

  • Scott

    I call on every Obama supporter who just FUCKED the long-term unemployed to give up their jobs to qualified non-Obama supporters.

  • http://www.deanesmay.com Dean Esmay

    How ’bout you just bring back the jobs those guys you call “our most productive citizens” (you know, the ones who never worked a real job a day in their lives, and who claim that corporations are people) who shipped those jobs overseas to slaves and indentured servants back, and we’ll call it even?

    Also, how ’bout you learn to lose with grace and humility instead of crying like the little bitch Democrats did over Florida in 2000?

    By the way, looks like Obama didn’t declare early victory as predicted by some, and is still waiting politely for the concession speech. Oh, and Fox News called it for Obama before anyone else.

    Now let’s just see how the electoral college works out, and how our predictions worked out.

  • Scott

    1) They shipped those jobs overseas because their businesses couldn’t exist in the environment with heavy-handed union rules and regulations. It was move or die, period.

    2) Dean, I’ve known friends who fucking killed themselves for the insurance money because they worked on an oil rig in the Gulf and Obama illegally and immorally killed their fucking job. so he and every Obama voter has their blood on their hands.

    No I won’t “lose with grace and humility”. We’re fucked if his first four years of open hostility to opposition, endless debt to give his cronies money and illegal assassinations and imprisonment is any clue.

  • Elizabeth Reid

    Hey, Donald Trump is calling for revolution. It could be worse.

    If Romney had won I would be the one filled with anger and sorrow right now. But I truly believe that 98% of Americans, including the ones I disagree with, are not imbeciles or cold-hearted or greedy or lazy or… whatever epithet leaps to mind (and the remaining 2% probably consists of as many Democrats as Republicans). I think it’s really silly to attribute *any* single defining characteristic to fifty million people for one thing, and for another I know enough people who voted for Romney to know that they are good people, smart people, caring people, so whatever motivated them to do this thing I strongly disagree with, it wasn’t because they just fucking suck. Try to give us the same benefit of the doubt, OK?

  • http://www.deanesmay.com Dean Esmay

    Obviously my mistake for opening my mouth. But you aren’t the only one who knows people who died due to poor government decisions. And obviously we wildly disagree on why those jobs went to slaves and indentured servants.

    By the way, your friends at Rasmussen obviously had bad models, didn’t they? They might want to go back and re-tool. I would also suggest Republicans start listening harder to the likes of Haley Barbour and Lindsey Graham. I voted mostly-Republican in every election from 1994 to 2008, and I don’t think it’s because I lost touch with reality or all that much of a change of heart on most of the issues that I changed my pattern of voting. I didn’t leave the Republican Party, the Republican Party left me. And the new generations are coming along, and they’re facing different issues and challenges. It’s time to recognize that.

  • http://madisonforum.net/ Sandi

    I for one and very disappointed, and yes disgusted. However I am mostly disgusted with much of the electorate who don’t take the time to know what their candidates are really about. It is them that put people in office that do not agree with them, and do not vote them out when they say one thing and do another.

    However life goes on, and the world goes on. So I go on. As I am retired and 72 the effect of the election is not as bad for me as most. However as I said above, I have little faith in this countries people. They got what they asked for. They will get what they deserve. Those who will suffer the most is the next generation.

    It should be a rude awakening for 2016.

  • Scott

    OK, I apologize for my outburst and language. I’m still mad at Obama voters, but at least I can forgive their short-sightedness. If not understand it.

    I just hope Obama has the humility to understand one thing: the mess we’re in now? He owns it. He can’t blame Bush or the GOP anymore for 25 million unemployed, for 45 million on food stamps, for the skyrocketing price of food and gas, for the $6 Trillion in debt, for the soon-to collapse SS/Medicare/Medicaid systems. It’s all on him.

    He needs to internalize this find compromise, truly compromise from now on, else he’ll leave a country far worse than he found and his name will become a curse for generations to com like “Nero”, “Nixon”, etc. are.

    Oh, and Dean, all those young people? They’re about to get a rude awakening if Obama doesn’t adopt a humble and compromising approach to governing. And if the next four years are like the last four I foresee a massive youth move to the GOP.

    And I see the GOP becoming more libertarian.

  • jrogge

    If the GOP drops the attack on homosexuals, contraception, and latinos they will win. The main reason for the loss today was the loss of much of the latino vote. There are many people who are fiscal conservatives, but socially liberal. The GOP needs to stop thinking that it needs to become “more conservative”. Whether I live in a socialist utopia or a corporate gulag, I would care less. However if I can’t come home from my crappy job and have group sex with my husband and two wives after smoking a giant blunt, then I just don’t feel free. (BTW I do none of these things, but I feel like I should be able to).

  • http://madisonforum.net/ Sandi

    There are many people who are fiscal conservatives, but socially liberal.

    That is a concept I have always had a hard time understanding. Being socially liberal means liberally spending big money on social programs that are in direct opposition to the term “conservative.” Not to mention social programs are one of the biggest budget expenses.

    It is like saying I am conservative in how much I drive to save on gas, but like to do it in a Lamborghini.

  • Elizabeth Reid

    Those of you who are predicting doom and destruction over the next four years? With all due respect, on Monday you couldn’t accurately predict Wednesday, on the most studied subject in the history of ever, with every news channel over the last few months devoting substantial time to that one subject and what was likely to happen. I swear I am not trying to be a jerk, but doesn’t it give you even the slightest doubt about your ability to predict what’s going to be going on in 2014?

  • http://blog.beliefnet.com/cityofbrass Aziz Poonawalla

    Elizabeth – I hope you’ll agree that my recent post makes this same point with perhaps a bit more finesse :)

  • http://www.deanesmay.com Dean Esmay

    I agree with Jrogge and I agree that the Republicans will probably become more libertarian, and this will help them. But they’re going to have to do something to address the behavior of publicly traded corporations beyond saying “smaller government will fix it.” They aren’t going to believe that “smaller government” will fix that, and that’s because there’s no reason to believe it’s true. In fact, there’s every reason to believe the opposite is true, and that what you’re essentially doing under the guise of “letting the market decide” and promoting “competition” with corporate behemoth looks like this: “Listen if we have a lot fewer cops, there will be a lot fewer arrests and prosecutions, therefore, obviously, less crime!”

    The Republican Party’s roots were in small business and entrepreneurs. They no longer are. This is painfully obvious to many of us who’ve run our own small businesses and is painfully obvious to more than one small business owner I know. The big corporations go in and do pretty much what government designed them to do (oh did I mention government created and sustained corporations? That’s because they did, which is another thing you guys need to acknowledge when you’re talking about “the free market.”)

    Time and again individuals who would like to be self-employed or actually start their own small businesses (whether they incorporate or not) find the big corporations come in and crush them, not through legitimate fair-playing-field “competition” but rather through the brute force power of monopolistic, rent-seeking, and anti-competitive behavior that gobbles these small business up and/or just crushes them, using the sledgehammer brute force of massive wealth accumulation and their ability to operate 24/7.

    Watch the movie “Office Space” and realize that this is the reality countless young people actually live in, and how fucking miserable and dehumanizing it really is. That’s a movie from the ’90s, but today it’s worse than ever, and it’s NOT because we “regulated too much,” in fact, we’re in a period where “deregulation” has been the war cry of every politician between Reagan all the way to George W. The increasing trend for young people is to see that these corporations view them as DISPOSABLE. You sit there and say “well these corporations are just acting rationally to get a better deal elsewhere,” but what they hear is, “we have no loyalty to you whatsoever, and there is no reason for you to be loyal to us, we don’t even appreciate your loyalty, and the minute some middle management fuck decides she doesn’t like you you’re gone motherfucker.”

    My advice is: become the party of small business and entrepreneurs again, and start turning a much more jaundiced eyes to the corporate monoliths. Most of them are terrible places to work, most of them rent-seek, monopolize, and seek market capture any way they can: if they can’t do it through regulatory capture, they’ll just do it through the old fashioned means that caused the creation of anti-trust laws in the first place.

    Guys, I know Romney did reasonably well, but Tea Party candidates took a drubbing across the country, and that’s because “smaller government” is *not* something that impresses, because it’s too simple. What are you going to do in a *positive fashion* to make life better for people who want to be self-employed, who want to start small businesses or partnerships or family businesses, and what are you going to do to limit the behavior of corporate behemoths to set the agenda for everyone else?

    If you can’t do that, don’t keep thinking you’re going to have the young people on your side. In fact, for the first time in history, for not just one but two election cycles in a row, the youth vote was a major factor, and it went overwhelmingly for Obama. You need as Republicans to start listening to these young people and their concerns, and I’m telling you right now “smaller government” increasingly sounds like a shallow, empty slogan. Do you have a positive agenda that will help people gain independence that goes beyond “you’re on your own motherfucker?” If you don’t, you can count on continuing to have problems.

  • Scott

    Guys, I know Romney did reasonably well, but Tea Party candidates took a drubbing across the country, and that’s because “smaller government” is *not* something that impresses, because it’s too simple. What are you going to do in a *positive fashion* to make life better for people who want to be self-employed, who want to start small businesses or partnerships or family businesses, and what are you going to do to limit the behavior of corporate behemoths to set the agenda for everyone else?

    Well, the autopsy of the Romney campaign is just starting. And a lot of people are pointing out things that many libertarians and conservatives thought but kept silent about. They’re things that might have helped Romney but we didn’t (we libertarians for Romney) want to “rock the boat”.

    Things like:

    1) Yes, Romney should have distanced himself from the Senate candidates who made stupid (if horribly misinterpreted by the Pro-Obama media) remarks about abortion and rape.

    2) Romney should have gone after Obama about Benghazi. As the election closed we learned more and more about the culpability and lies of the Obama administration on the matter. Romney let it drop and that was a mistake.

    3) While doing the above he should have at the same time been more on point about “working with the other side”. It would have been a strong contrast with Obama’s and Reid’s out-and-out refusal to work with the other side (heck Reid said that in the case of a Romney win he wouldn’t work with him).

    4) Romney was never a “small government” guy. He was a moderate-sized government guy. He wanted to shrink the rate of increase of government programs. While Obama wanted to build on his (and Bush’s) massive government increase in spending and intervention in the economy. Romney should have emphasized that it was a “reasonably-sized government” vs. a “Massive-sized governmnet” choice by promising to go back to Clinton-era government spending levels as a percentage of GDP.

    The small government argument is not a loser since no one, aside from Gary Johnson, was really for it this election cycle. The Big Government as a Magical Solver of All Problems will be a loser in the years to come as our collapse increases.