Princeton Election Consortium

A first draft of electoral history. Since 2004

Presidential prediction 2012 – final

November 6th, 2012, 2:00pm by Sam Wang


I apologize to all for the late update. We wanted to make sure all the polls were in. And it’s a hectic day.

The following are final estimates, based on taking polling data over longer intervals than our usual 1-week rule. Instead, I found the period over which a state’s polling variance was minimized, as a means of identifying stability.

ELECTORAL PREDICTION (mode): Barack Obama 303 EV, Mitt Romney 235 EV. The mode is the single most frequent value on the EV histogram. It corresponds to the map below, and has a 22% chance of being exactly correct. The next-most-likely outcome is Obama 332, Romney 206 EV.

ELECTORAL PREDICTION (median): Obama 305 EV, Romney 233 EV, Popular Vote Meta-Margin Obama +2.76%. This median is almost guaranteed to be off, since 305 EV is not a common combination. It is the midpoint of all possibilities, and reflects the overall shape of the distribution. The nominal 1-sigma band is Obama [293, 332] EV.

TWO-CANDIDATE POPULAR VOTE SHARE: Obama 51.1%, Romney 48.9%.

ALL-STATE PREDICTION (binary outcomes):

As I wrote late last night, Florida is a hard case.  Several new polls came out this morning, making the median basically zero. As a tie-breaker I resorted to mean-based statistics. I will be unsurprised for it to go either way. Nate Silver and Drew Linzer went the other way. We are all tossing coins. I am prepared to lose the coin toss.

In Florida, a recount is triggered by a margin of 0.5% or less (recount rules, Brennan Center for Justice). I estimate that there is a 50-50 chance of a recount. We might not know the exact outcome for some time.

The next-closest state is North Carolina, with margin of Romney +1.0%. All other margins are 2% or greater.

POPULAR VOTE. Using the above Meta-Margin for the last three days was Obama +2.76%. The median of national polls is Obama +1.0 +/- 0.4% (n=21 polls). The approach I described before for combining these measures gives

  • Final predicted popular-vote margin: Obama +2.2 +/- 1.0%.
  • Two-candidate vote share: Obama 51.1%, Romney 48.9%.
  • Allowing 1% for minor-party candidates: Obama 50.6%, Romney 48.4%.

Finally, here is the EV histogram based on the optimized probabilities. The two highest peaks correspond to FL-Romney and FL-Obama.


By 5:00pm I’ll have some graphs for you to print. They will let you compare the returns with polls.

Tags: 2012 Election · President

158 Comments so far ↓

  • Dean

    This was also my prediction, upon looking at the polls. Let us hope fervently that it comes true, and that it may be an early, happy night tonight.

  • cevangelista413

    man i hope this is right. i have faith, but i also make the mistake of reading other things, like some guy at the wall street journal predicting a romney win with him getting pretty much ALL the swing states, somehow. And also this video about voting machines in PA automatically going to Romney even when you press Obama: http://blogs.suntimes.com/ebert/video-of-voting-machine-fraud-.html

    Time to take my brain medicine…

  • Skillethead

    I’m a regular follower, ex-pat, Princeton alum living in New Zealand. My sincere thanks for all your work and your colleagues’ work on this site. Also, as an educational psychologist, thanks for your books on the brain. They will feature prominently this year in my “How People Learn” course.

    Thanks, Sam!

    • ZSdust

      hey skillethead–also following from NZ (Auckland).

      I’m going with 303EV, too. Though it would be nice to see O pick up FL–maybe he’ll win the coin-toss.

      isn’t the internet great?

  • BrianTH

    Looks like my opportunity for relatively unique glory is to predict 319 (which I believe is summarizable by Obama getting FL but not VA).

    So 319 it is.

    • BrianTH

      Oh, and I guess also FL and VA, but not CO and NH. Anyway, go Team 319!

    • CH

      Taking a break from GOTV here in N VA, and I really think (hope) you’ll be wrong. Lines to vote are very very long today, and the GOTV surplus of enthusiasm and labor is really reminding me of 2008. Feeling good.

      I’ll be Team 323 over here in the corner – think NH falls our way too, but I share your doubt re: CO.

    • BrianTH

      I’m probably wrong. I’m selecting my prediction not because I think I can actually improve on the implied ordering of states found here (or any of the other aggregators using the same state polling data), but because on the small chance I am right, I will be one of the very few to have selected this particular number.

  • Mike Jones

    would you have an updated stateprobs.csv? I’d love to plug it into my spreadsheet.

    thanks so much!

  • RJ

    Great work. Many thanks. Hang tight. 5 hours we start the ball.

  • Michael S

    Seems right. Florida is tipping over for the other guys based more on wishful thinking than anything else.

  • Matt McIrvin

    The ghosts and mirages are starting to multiply. Just heard a dismayed Cambridge liberal on 538 claiming that Obama is “getting beat in Philadelphia”, which I can’t see making any sense at all.

  • RandyH

    Would like to see Florida Blue and that could still happen even with the suppression attempts.

    But VA and OH look good so 303 I would be happy with.

  • Montykoolaid

    AHHHHHH!!!!

    The voter suppression reports are flooding in…

    We are dealing with millionaires and billionaires people…

    Is Romney going to win?

    • Howie Weiner

      No Romney is not going to win. Good advice is to stop listening to anecdotal reports they are almost always wildly inaccurate.

    • CH

      Save your sanity! Look away from the news. Go to Obama’s website and get yourself set up to make calls to GOTV in swing states. You’ll panic less, the time will pass, and you’ll help us get there.

  • pidru da pogi

    I am using your numbers to bet with. Most of my friends only know about Nate. I’m going to look like a genius!!!

  • wheelers cat

    im standing my ground at 332 and 4.6%.
    if there is significant systematic error in the legacy polling on cell phone demographies and asymmetry then I will be right.
    If Dr. Wang, the CLT, and the world according to Gauss rules, then Obama wins anyways.

    im not standing by myself.
    Nassim Nicholas Taleb has my back….and…and…evolution!
    ;)

    • Anbruch

      You might be right about the cell phones, but I think we’re going to also see the absent GOP voter, who will find better things to do today than make it to the polls and vote for Romney.

    • wheelers cat

      everyone should watch Olav’s link to Michael McDonald.
      Because of Sandy….im going to reduce my popular vote margin to +4.2 % Obama.

    • 538 Refugee

      Gravis has Flordia tied. That means Florida is Obama’s. ;)

    • Amitabh Lath

      Wheeler, thank you for your kind words on the other thread. I have enjoyed your idiosyncratic take a lot. I confess to not knowing anything about fractals, but never to old to learn.

      I don’t know what part of the country you are writing from, but if you are in the NJ area, drop a line.

    • wheelers cat

      Amitabh, your comments have been delightful epiphanies for me.
      Someday we will meet on the border of the quantum and classical worlds, and I will buy you une tasse of monkey-picked oolong in the tea house at the end of the universe.
      Dr. Wang too of course.
      I will be wearing my Unification Theorem t-shirt so you can recognize me.

      /makes deeply respectful obeiance

    • Some Body

      One second, but if you say Obama takes the PV by 4.2%, or 4.6%, then why only 332, and not 347, in the EC?

    • wheelers cat

      I believe there is systematic error in the nat’l polls because of legacy polling methodology and the inability to accurately capture asymmetry and cell demographies.
      This error doesn’t doesnt covary with the state polls.
      Dr. Wang estimated it at 2.6%

  • Eric

    The 303 or 332 final EV total has been my prediction as well, but I lean more toward the latter because the newest polling data and exit polling data favor Obama so when all is said and done I fully expect history will record the Nov 6th 2012 national election to be Obama 332 to Romney 206. As for the national popular vote, my expectation is that it will be closer to 3%, but my analysis is nowhere near as complete or detailed as yours. What concerns me most is if it is in fact 2.2% or even smaller it will show the handful of billionaires that funded Romney’s campaign that it is indeed possible to affect any race by throwing enough money at it because human beings are so easy to control. What a shame. But at least justice will be served by Obama’s re-election victory tonight. Tomorrow morning we shall all awaken to a more hopeful future under the President’s leadership.

    • Randy

      I don’t think the handful of billionaires will be satisfied with a pyrrhic victory that their millions *almost* won an election. Instead, I hope they conclude that they threw that money away. I think the truth is that money might have made a difference if they had a better candidate. Time will tell.

    • Some Body

      Eric, raw exit polls are very far off the mark (and on average, though not in any case, seem to have a strong Dem bias). I wouldn’t use them as an input. Plus, the ones you’re citing are almost certainly bogus (too early in the day).
      (My source: http://www.huffingtonpost.com/2012/11/06/2012-exit-polls_n_2038617.html?utm_hp_ref=@pollster)

  • RandyH

    With Florida the state officials are saying a record turnout of over 9 million, they say high turnouts mostly help Democrats..We shall see.:)

  • Joel

    I think a unique but plausible prediction could be Obama winning Omaha again. The polling isn’t awesome for him there, but it’s pretty sparse. So I’m going to go Obama 304, Romney 234 for my chance at glory.

    • BrianTH

      Ooh, smart.

      If too many other people jump on Team 319, I might have to switch to Team 320.

    • Craig

      Obama won that district by 1 point four years ago. It’s not happening again. Romney has a better chance of winning the election than Obama has for a NE EV.

  • InmanRoshi

    Larry Kudlow says Romney is going to win because he’s an optimist, and optimists always win elections over pessimists.

    Sounds pretty scientific …

    http://www.realclearpolitics.com/articles/2012/11/06/romneys_optimism_will_win_116077.html

    • RandyH

      Republicans do not believe in Science remember.

    • RDT

      I think he might have something with the optimism thing — but its hard to see Romney as the optimist in this race.

    • Some Body

      An optimist and a pessimist meet on the street. The optimist asks: “How’s life?”, to which the pessimist replies: “Oh, it’s horrible, couldn’t be worse!” to which the optimist cheerfully replies: “Oh, no, of course it could!”

    • listenToMe

      I considered joining the Optimist Club but I figured they wouldn’t have me!

  • Fred

    I gotta admit I hope Sam is wrong and Nate is right. It would sure be sweet to win FL.

  • ajnycajnyc

    Sam, thanks for all the thoughtful and thought-provoking analysis over the last month. By taking a statistics-based approach to the election, you have done the nation a great service by helping voters (like me) gain a greater understanding of the actual state of facts during the election rather than forcing us to rely on the “gut instincts” of professional bloviators with axes to grind, news “stories” to flog, and no accountability.

  • E L

    Forsan et haec olim meminisse iuvabit.

  • Khan

    The voter suppression is making my heart ache.

    • Neil S

      This is the only election site with sufficient information and clarity in a nutshell. Thank you!

  • Dylan Wasserman

    I live in FL ( Pasco County , about +7 Republican, and from what I have heard, it seems as if the Republicans canibalized their base during early voting. Lots of Obama supporters voting today. My Girlfriends best friend told us that while waiting in line a white couple were heckling an Indian couple and telling them that they are the reason our county has no morals and is dying. Its funny how apparently “good” morals involve heckling people, that for all you know are voting for your candidate. Cant wait for the results tonight.

    • Fred

      So your girlfriends best friend told you about something they saw happen to someone else. Sounds like an iron clad fact!

      Best thing that could be done when that happens is to pull out the video phone and get it on youtube. Seriously, this stuff should be out in the open for all to see. That is when things change.

    • Martin

      How do you know their O supporters?

  • Froggy

    New evidence out today that Sunday’s Redskins-Panthers game (the Redskins loss forecasting a Romney victory) may have been wrongly decided:
    http://www.washingtonpost.com/blogs/football-insider/wp/2012/11/05/nfl-acknowledges-officiating-mistake-on-panthers-touchdown/

    • Strangeite

      Now that is a scary black cat superstition.

      Imagine that the universe is stranger than we can possibly imagine and the Redskin meme is the TRUE catalyst for taking the White House.

      If the the Redskins only won because of an officiating mistake, then maybe Romney will win, but only because of fraud.

      Wooooooooooooooooooooo…..

    • Olav Grinde

      Froggy, an “officiating mistake” is the only thing I actually fear today.

    • securecare

      Been quite a few of those “miscalls” lately. Maybe the idea of review needs to be extended.

    • Mason

      It shouldn’t have even needed a review. The LJ blew his whistle and began to kill the clock. The play should have ended right there whether he was right to blow it dead or not.

      I’m not sure if inadvertent whistles are reviewable, but they did review the play because it was a scoring play.

    • Suja P

      Is this considered an “officiating mistake”?

      http://www.youtube.com/watch?v=oOj2IPulNPA&feature=share

    • Chicky

      my wife says it’s all because of Mercury Retrograde. Astrology is an exact science, right?

  • Steven S

    As we prepare to see the final result, it makes me think about how Sam has long since pulled back the curtain for the 10s of thousands of people who have had the privilege of tuning in.

    The October exchange with John Dickerson was an interesting example of the clash between those who have seen behind the curtain and those who haven’t (or refuse to look). During this election cycle, I’ve often thought there are parallel worlds out there. The Big Media and blogger conversation (except for 538) rattles on, and then there are those few people who have been shown the power of data to tell the real story.

    So this makes me wonder about two things:
    1) What will happen, assuming we see a result around 300 EV, if the Dickersons and other pundits of the world begin to accept the existence of our secret reality? We might begin to see sentences like this, “The sense of excitement and momentum at Romney rallies has increased dramatically despite no apparent gains in the major aggregators of state polls such as the New York Times’ 538 and the Princeton Election Consortium.” I think this site and 538 have received enough attention that a tipping point could come.

    Wow. But… then would our glory be short-lived? Would a new wave of manipulation be unleashed, where the name of the game is to confound pollsters and add noise that’s harder to strip out? So that’s one of my big questions going forward: What is the PEC vision of success? How can we permanently elevate the dialogue?

    2) What other areas of punditry could enjoy the same numbers-driven success? I saw some commentary on some economic figures the other day and I thought, “Whoa, am I listening to this opinion just as those who don’t know about PEC listen to poll results? Is there a demonstrable, objective reality out there that I am not privy to or that no one has exposed in the way Sam has done for elections?”

    Are presidential elections a unique piece of low-hanging fruit because of the great disparity between the amount of data and the amount of bloviating? Or are there other data-rich, under-analyzed areas of public life?

    Thank you, Sam, and all PEC commenters for a fascinating ride this season.

    • wheelers cat

      Steven S
      There ARE parallel worlds.
      We just cant get to them.
      ;)

    • Olav Grinde

      Wheeler, I thought that was where you were posting from. ;)

      Seriously, though, I greatly appreciate your lively posts!

    • wheelers cat

      okfine.
      YOU can’t get to them.
      ;)

      /heart Olav

    • securecare

      “…are there other data-rich, under-analyzed areas of public life?”

      The field of economics needs a very heavy dose of the type of non-traditional behavior we see here by interested people. The “priesthood” in that field needs to be revealed for their voodoo behavior.

    • Steve16748

      Regarding economics, I think it would be very helpful if we could all get our friends and associates to read Paul Krugman’s blogs. He is a great teacher. It is helpfully that his pieces are very seldom kept behind the NY Time’s fence.

    • TAW

      Steven S:

      Re: Are presidential elections unique with respect to media bloviating?

      Yea ….. financial and economic news has a lot of parallels, although they don’t converge on a single, widely publicized outcome.

      The biggest take away is the bayesian thought process. In most areas that the daily news covers, people have a reasonable prior belief.

      Use the same general reasoning used on PEC.

      First, eliminate outliers is an efficient manner.

      Then try to figure out to what extent the new information would change your prior belief. You don’t need to do the arithmetic — just try to look at it as a single point and keep in mind the cognative biases that tend to overweight it.

      That’s why Warren Buffett reads 10-Q’s and ignores analyst reports as well as most financial news.

      As far as low hanging fruit, I would say that unreproducable results in psychological research is obvious to me, at least.

      And more importantly, medicine is trying – lots of meta studies – but there is a huge amount to be done.

    • Some Body

      Steven — I find this messianic approach to data and numbers a bit, well, self-defeating. While election forecasting is definitely an excellent example of the power of quantitative data analysis, there are also many examples of over-zealous quantification which simply leads to bad results.

      We’ve seen that happen many times in the social sciences — folks keep using numbers and quantitative (or more broadly — formal) models where there’s no clear evidence that they actually apply, get all the wrong results, and then justify their failure by saying that their model must be superior *because* it’s quantitative /formal (or, worse still, they effectively rig the empirical results through misuse of definition; case in point: defining intelligence as whatever is measured by intelligence tests). I can go on for hours about this sort of fiasco in my own field (the study of language).

      Quantification and statistical analysis are highly powerful tools of scientific research and of understanding reality more generally, but you should also be aware of where, how, and to what extent can these tools be applied, because where applied incorrectly, they can derail science instead of promoting it.

      Sorry for the rant, pet peeve and all.

    • Steven S

      @Some Body: indeed, the key would be that the data lead to a demonstrably correct (low-noise) view of something. Elections may be unique because, barring hellish outcomes like Florida in 2000, there is in fact a single answer at the end. So back to my question 1: Will even that last if aggregation goes mainstream?

  • Dylan Wasserman

    By the way guys a factor NO ONE has considered is that Halo 4 came out today. Millions of young people like myself will be out today picking up the greatest game to come out in years and will help Obamas voting turnout. I have 5 friends that voted early last time, but figured they could make one trip election day to get the game and vote all in the same day. Sales are expected to be the largest release of any media product EVER, and that includes the Avengers movie.
    http://www.sacbee.com/2012/11/06/4964343/the-biggest-entertainment-event.html

    • Matt McIrvin

      Ha! I’ve also heard that Halo 4 will doom Obama because young voters will all play Halo instead of standing in line to vote. I guess that can be spun either way.

    • Stuff

      By the same token, you could argue young people will be too busy buying and playing Halo to vote. Sounds just as plausible, and equally unsubstantiated.

    • Dylan Wasserman

      I dont see that spin at all, What seems more realistic to you, that while people are out buying a game they will vote, or that they will buy the game, rush home and play it all day and not vote. It can be spun but I will take the logical point of view

    • Mason

      Then there are gamers like me who never buy any game until it has gone through at least three price cuts. Also, when things are on sale on Steam.

    • Matt McIrvin

      I’m just upset that the last several updates to Farsight’s Pinball Arcade may not make it to XBox Live for a long time (if ever), because Microsoft makes them use a software publisher as a middleman, and the middleman went bankrupt.

      Oh, well, there’s always Android.

  • Rick in Miami

    I just heard from someone on the Gulf side of FL that the scan machine at his precinct broke. After trying to get it to work for a while, they gave up and had everyone put their ballots in a box to scan later.

    Florida is now rid of butterfly ballots and chads, but we’ve figured out how to mess up optical scan voting!

  • Worried Abroad

    I just watched “Hacking Democracy” on youtube.

    It won an emmy for investigative reporting and it’s test bed for demonstrating how easy it is to do electronic voter fraud was Cuyahoga county in Ohio.

    I’m terrified.

  • Peter TS

    Maybe I’m out of the loop, but I haven’t heard of any significant problems regarding voter suppression anywhere. I heard about some complaints in Philadelphia today, which actually came from Republicans, but little beyond that.

    That situation was resolved, as was the early voting problem in Florida over the weekend (which Democrats considered to be a problem). So what am I missing? ;)

  • Mike B.

    Dr. Wang,

    Watched you on CNNI.

    You are a true national treasure now!!!

    Thanks,
    for all you do.

  • Jack, a former P-townie

    Dr. Wang & Crew: I believe in “the Math.” Thanks for the blood sweat and tears in putting it all together.

  • Pat

    While we wait, I aggregated the margins given by 4 aggregators: below are the average margins from the closest states, ordered from more Dem to more Rep, averaging the margins offered by:
    - Fivethirtyeight
    - PEC
    - Votamatic
    - Pollster
    (in brackets: the Obama cumulated EV total)

    NM – O+8.1 (EV total including NM: 184)
    MN – O+8.0 (cumulated EV: 194)
    OR – O+7.1 (201)
    MI – O+6.1 (217)
    PA – O+5.1 (237)
    WI – O+4.9 (247)
    NV – O+4.4 (253)
    OH – O+3.3 (271)
    NH – O+2.8 (275)
    IA – O+2.8 (281)
    CO – O+2.2 (290)
    VA – O+2.0 (303)
    FL – O+0.2 (332)
    NC – R+1.8 (347)
    AZ – R+7.5 (358)

    • Neil S.

      Not sure aggregating the aggregates makes total sense. There must be a significant overlap between aggregates, e.g. PEC vs. 538..

  • better bettor

    Obama 292; Romney 246
    Popuar vote: 54%O to 46%R

    Obama wins all swing states and NE district 2 in a huge outpouring of enthusiastic support, but OH, FL, and PA are stolen for R by vote suppression and vote tabulation fraud. The surprise wins in NC and AZ give Obama his winning margin.

  • Jon V

    AP- Live Vote tracking map: here is the link. I hope Sam does not mind if I post the link here. If so, I apologize, just wanted to provide another tool for people to follow results
    http://www.wyff4.com/news/politics/election-map/-/12016576/-/vifd4y/-/index.html

  • Montykoolaid

    Seriously guys….is ROmney going to steal this? How is the exit polling looking? (Not that it matters) Is turnout looking good?

    ???

    ?????????!!!!!

    • Khan

      Turnout is enormous. They’re saying turnout in some places is crushing 2008.

      Also sadly there is rampant, verified voter suppression and cheating happening already.

    • Craig

      Exit polls won’t be released to the media until 5 pm. And early exit data is usually pretty bad.

    • BrianTH

      Not that this is what data-starved people want to hear right now, but I don’t think it is worth looking at turnout figures or exit polls until after the polls are closed, at which point you will have more interesting information to look at anyway.

    • Mason

      Don’t worry about exit polls. They said Kerry was winning.

      Cook yourself up a nice dinner, or order out if your hands are trembling too much to hold a knife. Eat it. Have a drink. Flip the TV on around 7:30 ET to watch the fun begin.

  • Jon V

    MSNBC reporting large turnout in key areas for Dems, and record turnout overall in many places. Usually good for Dems. No exit polls can be released until polls for that state are closed

    • otis29

      That’s what I was thinking. Doesn’t a large turnout mean we’re getting closer to a “registered voter” turnout than a “likely voter” turnout?

    • Ralph

      The main difference between the ‘likely voter’ models given higher numbers to Mitt and those giving higher numbers to Lightening (Baraq is Arabic and Hebrew for lightening.) is the latter has a higher turnout.

      I know that Barack’s name is Swahili. I don’t know Swahili but know some Hebrew and a little more Arabic. Since the right is calling him Muslim, then his name is Lightning.

    • Some Body

      @Ralph — Actually, the name means “blessed” (soft k, not hard k). The same root as former Pres. Mubarak of Egypt.

    • Matt

      Ralph,

      Besides your reply sounding a little off the deep end, you are wrong. Barack is a cognate in Swahili from Barakah, which means “blessing.” You are confusing the letters Kaaf and Qaaf.

    • Some Body

      Matt — Jinx!

    • Fred

      exit polling is notoriously unreliable. Don’t understand why anyone cares about exit polling.

    • Matt McIrvin

      I think the Hebrew cognate is Baruch. Barak, like Ehud Barak, is a different name (which does mean “lightning”).

    • P G Vaidya

      But, the TV anchors have now seen the exit polls and they start dropping, not so subtle, hints.

  • SouthernFried

    Is it wrong that I just don’t have faith in the Math? I know Dr Wang has a fantastic track record (better than Nate’s!), but when it comes down to it, I have to trust my inner pessimist.

    I’m increasingly of the belief that the polls in CO, VA, IA, and NH are just flat-out wrong in having Obama ahead; and OH, FL, and PA — which would have sided with the president absent shenanigans — will go to Romney due to a combination of vote suppression and Democratic apathy (I’ve heard that voting has been very light in heavily D precincts in PA & OH).

    Naturally, I will be very happy if I’m once again proven wrong and President Obama can pull off the win he’s predicted to have, but I don’t think I’m wrong this time.

  • Jon V

    Gotta give credit to Rev. Al Sharpton, really made a big issue of voter suppression with minorities, I think this is going to back fire on R’s

  • Annelid Gustator

    Sam, what’d the variance minimizing period end up being?

  • Steven J. Wangsness

    Thanks…but Obama takes FL by +1!

  • xian

    math is my anti-anxiety medicine: yfrog.com/oeheylp

  • badni

    Sam! Did you break the faith of polling-medians-only forecasting?! Have we not learned not to anger the Median gods? Florida wil find a way to betray you!!

    This also puts all your eggs in one 303-ish basket. Having two legit projections was a great hedge. To tell the truth, I admire the boldness in doing that.

    • Matt McIrvin

      He did something similar in 2008. His off-by-one-EV prediction wasn’t the median value from his model, it was some kind of variance-minimized derivation. The median was something unlikely, just like this time. If you want to throw out an exact number it should probably be something more like a mode of the model output.

  • Khan

    I voted.

    I am channeling my inner bayesian until midnight.

  • Kyle

    Hello Sam,

    how does the minimal variance interval compare to the usual one week interval you’ve been using so far? It is much longer?

  • AdamT

    Just wanted to thank Dr. Wang for his excellent analysis. Long live rationality.

  • Olav Grinde

    Question: Does anyone know how Stephen Colbert’s SuperPAC is spending its money?

    • Mason

      He’s going to take all the money and run his own election. With blackjack and hookers! In fact, forget the election.

  • nutjob

    Are you not having two bites of the apple with the mean and mode predictions. Will you be going with one final prediction? Yes you have to call Florida, to not do so would be cheating!

  • Mike M

    British betting odds on Obama have moved from 1/5 to 1/4 on some betting sites.

  • Peter TS

    This site has been fantastic. Thank you Dr. Wang and team. I don’t think it’s an exaggeration to say that well-executed efforts like yours are what keep our democracy strong and vibrant!

  • Jeannius

    #team332 :) ALL THE WAY!!!

  • LTM

    Anyone got any Valium. This is way too much stress for a Tuesday.

    Oh please be right Sam, oh please be right!

Leave a Comment