↓
 
Home » On those polls

On those polls

The New Neo Posted on by

Here’s Nate Silver on where it stands at the moment:

Donald Trump has a 33 percent chance of winning the election according to our polls-only forecast and a 34 percent chance according to polls-plus. These roughly 1 in 3 odds are close to Trump’s highs since the party conventions.

Still not what you’d call great. He’s been there once before and fallen again, so it’s hard to say what will happen this time.

However, this time just feels different. I’m not sure why. I’ve never been the least bit sanguine about his chances of winning, but I’ve always conceded they are far from zero and will depend on a set of unpredictable circumstances. Hillary Clinton’s “deplorables” remark and her health problems last weekend are two things that I think have great potential to hurt her as long as Trump doesn’t do anything too awful before November—as long as he keeps being at least semi-presidential; and in particular, as long as he can acquit himself decently during the debates.

It’s hard to imagine these debates, except to say they are likely to be very atypical, and that the candidates are likely to showcase very different approaches from each other. We can expect Trump to be free-wheeling, non-detail-focused, and insulting. We can expect Hillary to be wonky and more subdued but strident nonetheless. But will they stay true to form, or will they mix it up? Will Trump suddenly talk about the minutiae of his plans? Will Hillary come out swinging? No one knows; all bets are off.

Here’s more from Silver:

My best guess on the effect of the weekend’s news, based on what the model shows so far, is that the race is continuing to trend moderately toward Trump, when the momentum toward him might have stalled out if not for the events of the weekend. But we can’t rule out a more acute shift toward Trump or that the “Hillary’s bad weekend” meme is a false alarm ”” there isn’t quite enough data yet.

Whether or not the race will continue to tighten is a guessing game, in other words. But my impression is that the commentariat has been slow to recognize how much the race has tightened already. It’s never a good idea to freak out over any one poll. But the trend toward Trump has been clear for a few weeks now, and it’s been just as clear in state polls as national polls. Yes, the data is noisy. Polls are all over the place in Ohio, for instance. But over the course of all of this, Trump has whittled down an 8-point lead for Clinton into about a 3-point lead instead ”” about a 5-point swing. With there having been several shifts of that magnitude since the primaries ended, with there being a large number of undecided voters, and with the debates still ahead, neither Clinton nor Trump should feel all that secure.

I have generally respected Silver’s poll analyses, with the obvious caveat that polls are polls and always imperfect. But he—unlike many other poll analysts—looks at the trends over time with a relatively clear eye. I also happen to agree with him here; what he says is that same as my gut impression.

I can’t imagine that Hillary Clinton is feeling good right now. I also can’t imagine that Donald Trump has ever felt all that bad about any of the results in the entire race, and he certainly has reason to feel especially good right now with some momentum going in his direction. Trump’s boundless self-confidence and faith in himself (also his narcissism) seem to have immunized him against that sort of doubt and anxiety, at least so far as we can tell.

Comments

On those polls — 36 Comments

  1. I believe that Trump is demonstrating a bit more discipline. I hope that his “handlers” will get him up to speed on issues–by force feeding if necessary.

    Hillary will certainly know the issues; but, he can trap her into obvious lies if he is prepared. Just do it without obvious bullying. Please.

  2. The subterranean anti-Hillary factor (which becomes pro-Trump by default) must also be kept in mind. All of my friends, indeed everyone I know, is anti-Trump, and because I have MS and need some of these people to pick up prescriptions and the like, I’m not going to contradict them about politics.

    My doctor went off on a minor rant about how the Clinton Foundation is actually pretty good and how Trump has stiffed this person and that, and I just nodded, said “Oh really?” and we parted fine friends.

    Meanwhile I associate Hillary not only with a continuation of Obama-ism but with every noxious manifestation of neo-fascistic political correctness out there, and dearly want to stop this in its tracks.

    But I’m not going to say it out loud. Not to anyone. Well, here, I guess, but using an internet pseudonym. Otherwise it’s unsafe.

  3. So much plaintive yearning is still be felt from Silver even after:

    1. June 16, 2015: Why Donald Trump Isn’t A Real Candidate, In One Chart

    2. July 16, 2015: Two Good Reasons Not To Take The Donald Trump ‘Surge’ Seriously

    3. July 20, 2015: Donald Trump Is The World’s Greatest Troll

    4. Aug. 6, 2015: Donald Trump’s Six Stages of Doom

    5. Aug. 11, 2015: Donald Trump Is Winning The Polls, And Losing The Nomination

    6. Nov. 23, 2015: Dear Media, Stop Freaking Out About Donald Trump’s Polls

    http://dailycaller.com/2016/05/04/7-times-nate-silver-was-hilariously-wrong-about-donald-trump/

  4. For those who haven’t seen it Salena Zito’s most recent article (in The Atlantic) it is an interesting take.

    http://www.theatlantic.com/politics/archive/2016/09/why-democrats-in-western-pennsylvania-are-voting-trump/499577/

    Granted it is anecdotal, and a single anecdote at that, but IMO it highlights the potential unannounced anti-Hillary feeling (and thus pro-Trump support) that seems to be undocumented in the polls.

    If she is correct (remember this is “Pennsyltucky”) between the crossover Dem vote in Southwestern PA and in PA’s Lackawanna and Luzerne counties (also coal country) I find it plausible that the Philadelphia Dem vote will be cancelled out. Pittsburgh, I don’t know about because although, like many major cities, it is a long-term Democrat redoubt, it also has a long-term relationship to coal country and the working class that doesn’t exist in Philadelphia or its I-95 corridor.

    Pittsburgh aside, a Trump victory in PA becomes more likely as/if this trend continues.

  5. Let’s just keep in mind how she dealt with Bernie. When the votes are close and she gets nervous, she cheats. (Remember the six coin tosses IN A ROW that landed in her favor?)

  6. Two weeks ago I fit a linear plot to the 538 odds data and it projected Trump reaching near 75% by Nov 8. I just did a recheck today, and the fit is still the same. I assume the “deplorables” and HRC’s health hasn’t really seeped into the polling sufficiently yet, so the slope of the fit may actually increase.

    Again, a simple linear fit to data for a complex system such as an election is simplistic, but can be illuminating. Back then I said the Clinton campaign should be a bit concerned. Now I would say they should VERY concerned.

  7. We’ve seen this before:

    trump down in the polls – polls are rigged

    trump up in the polls – competitor / clinton is a loser

  9. Physicsguy,

    I have been saying for at least several weeks, if not a month or two, that I think Trump will win. There’s nothing I want less but I think we get the government we deserve. Our country pretty much is asking to have a historically horrible person as President, so either of these will do, and I’m figuring it will be Trump just to make me miserable.

    With that being said, you wrote: “Again, a simple linear fit to data for a complex system such as an election is simplistic, but can be illuminating.”

    No it can’t. Try doing a linear fit on a stock you like that’s currently moving up, or any other object in motion. “Hey Marge, Apple went up 5% today! That means it will be up 100% in 19 more days! Buy! Buy! Buy!” 🙂

    Unless we’re talking about an object in a vacuum with nothing else acting on it, things just don’t move that way. I’ll eat crow, of course, if it’s 75% on Nov 8th. But I don’t think it will be. I think the election will be tight, Trump will probably win, and then everyone who voted for him and (of course) those of us who didn’t will get what they asked for, good and hard, provided the Alt-R thugs aren’t already rounding us up before then. #neverTrump!.

    He is exceptionally fortunate that HRC is such an awful, awful politician. People have said that a key selling point of Trump is that he’s “not a politician”. He’s actually quite a good politician, and way> better than she is. Maybe the Democrats are actually the Stupid Party after all.

    [fetal position]

  10. I know what miklos000rosza is saying. I think that when Trump becomes the leader in the polls, despite the obvious bias in the media, those who have educated themselves on the issues and what they really mean to our future will be emboldened and will begin to speak up. If Trump was going to lose (and worse, do so without dignity) few people wanted to be remembered as that idiot who was for Trump. However, as Trump becomes respectable, that is changing.

    I have manned a voter registration booth at a local market on Saturdays. An amazing number of people deduce that we are Trump supporters (no signage is allowed, nor any overt mentions of political issues) and give us a “thumbs up” sign or say something positive about Trump or negative about Hillary.

    Reflecting the enthusiasm edge noted in polls, most people changing address or newly registering are Republican (which, of course, is why we are out there in the hot sun).

    I was a Cruz supporter in the primaries. As the race developed and the issues became clear I have become an enthusiastic Trump supporter. I even have a yard sign “Veterans for Trump”, the only one in my condo group so far.

    As to the commenters here who have had such a difficult time with the Trump campaign – I just don’t understand. I was a registered Independent who changed so that I could vote in the primary, so maybe I lack the emotional attachment to Conservative tradition that causes them such anguish.

    I believe that the “Never Trumpers” will be the idiots of the future along with those Republicans who do not vote for their party’s candidate. But that is just me.

  11. Miklos is correct. While my household and family are strongly conservative, I am one of those people who in ordinary life is surrounded by angry, closed-minded lefty ideologues – the kind of people who would not only believe that 2+2=5 but berate you for believing otherwise, if that was the lefty-left party line. I was not and am not a Trump fan, but I’ll vote for a dead house plant before I vote for Hillary, who is not just a Dem but the worst kind of Dem. I just don’t need these jackboot lefties making my everyday life miserable.

    The number of people like us out there is hard to guess.

  12. notherbob2: “I believe that the “Never Trumpers” will be the idiots of the future along with those Republicans who do not vote for their party’s candidate. But that is just me.I believe that the “Never Trumpers” will be the idiots of the future along with those Republicans who do not vote for their party’s candidate. But that is just me.”

    Well, if Trump is elected and ends up being a great president those of us who are #neverTrump will have reason to eat crow, certainly. If he’s not elected, and HRC is awful (as I expect her to be) a lot of people will have pitchforks out for guys like me as well.

    Can it be OK with you that I am voting my conscience? This isn’t easy, but I cannot vote for Donald Trump. I’m tired of being psychoanalyzed by Trump supporters (“neverTrumpers are just self-righteous pansies/put your big boy pants on/you must be making money off this/etc.”)

    I am all for going hammer and tongs on the web, respectfully, about who we should vote for. But at the end of the day I have to respect the votes of those who vote differently than me. They are voting for what they believe is the best choice.

    I’m willing to be thought of as an idiot or a traitor – it’s the price of voting your conscience. It gets old, though. I say this as an Ex-GOPer who voted basically straight ticket in every election since I proudly voted for Ronald Reagan in 1984.

  15. I sympathize with Never-Trumpers but I part with them here:
    If scoundrel 1 wins, nobody of her party in Congress will challenge her, the dominant media will continue to cover for her, and the largely liberal judiciary picked by her predecessors will twist rulings to serve her. If scoundrel 2 wins, he will be confronted by many party members in Congress, he will be given a daily colonoscopy by an energized media ready to do journalism again, and the judiciary will be looking for him to serve them, not vice versa.
    .
    Let’s begin to get our arms around the prospect of preferring scoundrel 2 and work with the checks and balances that will tend to be operational in that event.

  16. There’s the possibility that Trump may attract some new voters who would not otherwise be there, due to his personality and the issues he has put forth. (Which should show up in the polls, but I’m not sure of the numbers. And I’ve only voted since 2000, but there are many others who never vote, even for President.)

    Like they say about war, the other side also has a vote. And in this case, there may be external actors, and external events that could influence this election.

    Right now, some people are ready to blame the Russians for anything, but regardless of what happens, we can be sure that Russia will look out for its own interests, and that Putin knows that the Clintons can be bought.

    Besides that, there is the possibility of more terrorist attacks, or more revelations from Wikileaks. The stock market could go down–but it seems less likely, after last year’s decline. Something more might come out about Bill and Hillary’s many scandals. Or Hillary could collapse again, or have another coughing fit. (She’s already been absent from the campaign so much that it just looks bad, regardless.)

  17. I think it’s the endless revelation of Clinton lies and corruption that are wearing down her supporters. Even the hard-core loyalists are getting disgusted. I don’t know that any of them will ever vote for Trump, but I expect the turnout among usually reliable Democrats will be lower than usual.

    Of course, the Dems do have the zombie vote to count on.

  18. If, in the debate, Trump happens to not say anything untrue, but Hillary does – if Hillary looks sickly or makes any unusual movements that people can interpret (even if wrongly) as evidence she is not well – or if Trump can hit one of those electric statements that he does every once in a while, any of those can move the dial. There are likely other possibilities of “things that could happen.”

    The point to notice is that only one of them has to happen.

  19. While there were varying predictions about Trump’s ability to win the nomination, there was widespread agreement that he gave Hillary the best chance possible.

    This shouldn’t have been close.

  20. Teh Donald is certainly trending upward.

    This weekend was terrible for Sec. Clinton.

    First, her deplorable remark.

    Second, her staff (benefit of the doubt here) saying it was the heat. THEN saying it was/is pneumonia. Teh Donald can wish her well AND still score points about her ‘truthiness’.

    Why couldn’t she just have ‘I have pneumonia’ to begin with?

  21. Tuvea, 11:52 pm — “Why couldn’t she just have ‘I have pneumonia’ to begin with?”

    The Clintons will reflexively default to the lie, whatever lie, even when it is more advantageous to them to go with the truth.

    Incredible, isn’t it?

  22. Tuvea Says:
    September 14th, 2016 at 11:52 pm

    HRC was wearing prescription BLUE sunglasses — specifically known to suppress Parkinson’s symptoms, while at the 9-11 memorial.

    The average Joe is not aware that said lenses are a ‘tell’ for the malady.

    You just don’t find them on sale at your usual retail outlet.

    https://www.youtube.com/watch?v=q5mYx5oCxEg#t=117

    Jump to 3:26

    She’s so well medicated that many of the symptoms that folks think of are repressed in her.

    Nonetheless she needs to have two physicians with her at all public events — she doesn’t settle for nurses. Her tab has to run $20,000 per day.

    A vote for Clinton is a vote for Kaine — as she can’t last much past the oath of office.

    That’s her health reality.

    BTW, check out her co-ordinating blue outfit to hide her need for those striking BLUE lenses… as if they were a fashion statement.

  23. I think Trump’s paid maternity leave proposal is a direct appeal to normally Democratic voters in a time of weakness for HRC.

    HRC is truly one of the worst politicians to ever make it this far. Regardless of what happens, both parties could have done so much better this time around. But Democrats were fixated on trying to relive the glorious Clinton years and stuck in “it’s her turn” thinking (and really, really lacking a bench of any kind) and too many people on the right (I won’t say Republicans) were in a “burn it all down” mood, so here’s what we got.

    It’s funny, now, as a conservative who is ex-GOP, to see all the “lions of conservatism” like Hannity and Rush spin positively for the non-conservative proposals Trump is putting forth. How’s that institute for advanced conservative studies working for you now, el-Rushbo?

    I quit listening to those blowhards years ago.

    The good thing about 2016: it is illuminating. People who have extolled their own conservative purity for years are being unmasked. It was never about ideology. It was never about principles. It’s about one thing: will to power. As long as it’s their guy as the authoritarian strong-man everything is A-OK.

  24. “No it can’t. Try doing a linear fit on a stock you like that’s currently moving up, or any other object in motion. “Hey Marge, Apple went up 5% today! That means it will be up 100% in 19 more days! Buy! Buy! Buy!” :-)”

    Quite true. However the DATA from Aug 16 to the present is best fit with a linear function. Do I believe it will stay that way? No. But until the DATA shows signs of such non-linear oscillations, then linear is the best fit.

    One huge Trump foot-in-mouth, or another Hillary fainting session can throw the whole thing into another scenario. Why impose non-linear solutions on data that doesn’t show such tendencies at this point in time?

  25. “Why impose non-linear solutions on data that doesn’t show such tendencies at this point in time?”

    Because the conclusion you came to (75% for Trump) doesn’t appear reasonable, and because the data has, when a wide view of it is taken, been non-linear.

    I’m not disagreeing with you that Trump will end up winning.

  26. Not the last word, but insiteful none the less:

    “It is sometimes the case, though few analysts care to admit it, when it is simply impossible to objectively predict the prior probability of an event. We peer into the crystal ball but it remains silent. In the choice between the candidates some supporters of Hillary though well aware of her flaws have persuaded themselves she is the lesser risk than Donald, the same reasoning is probably being invoked by Trump’s backers. ”

    Richard Fernandez

    https://pjmedia.com/richardfernandez/2016/09/14/raising-the-hood/2/

  27. blert:

    I used to wear a pair of blue sunglasses a lot. They were a fashion statement, and fortunately I wasn’t wearing them for any medical reason. I’ve seen plenty of people who do the same.

    Of course, it’s a possibility that Hillary wears them for medical reasons. But certainly not the most likely one.

    Oh, and it’s not Parikinson’s the Zeiss glasses help. It’s a very specific type of epilepsy, photosensitive epilepsy (where epileptic fits are triggered by flashing lights). It’s a rare form of epilepsy accountable for only 3% of epileptic sufferers.

    I think you are grasping at straws.

  28. “In sum, it appears that Democrats have an even bigger advantage than they did in 2012 because their margin over Republicans has increased. That’s the good news for Democrats. The bad news is Clinton has not yet sealed the deal with weak and leaning Democrats–and these are voters who are extremely familiar with her and the Clinton brand. Unless and until she (clinton) reaches Obama’s 90 percent-plus of Democrats in 2012, the race will remain fluid.

    Meanwhile, Trump has the same problem with weak and leaning Republicans and is well short of Romney’s 93 percent-plus 2012 numbers. However, because there are fewer Republicans now, his (trump’s) situation is more dire. Among Independents indicating a preference for president, Trump leads Clinton but by nowhere near enough to make up for the partisan gap.

    However, the 31 percent of respondents who intend either to vote for a third-party candidate, are undecided or are saying they won’t vote keeps us from concluding that the election is on the way to being decided. This is significant. Obama had a similar lead over Romney four years ago, and won by that margin. But so much of the electorate had its mind made up by this time, there wasn’t much wiggle room.

    While it is true that when pushed to vote, the 31 percent (undecideds)
    prefer Clinton to Trump by 39 percent to 24 percent, this preference is not strong    , and more than one-third of these voters still won’t state an inclination one way or the other.

    These data show that in early September, Clinton holds a small lead, but it is not conclusive and she will have to firm up her support among Democrats and continue to stay close to Trump among Independents and undecideds in order to win. Trump also has to shore up his Republican Party support and begin to increase his percentage of Independent voters by converting the undecided to him.

    Of the two candidates, he (trump) has the longer row to hoe.“ RCP article

    http://www.realclearpolitics.com/articles/2016/09/15/party_identification_shift__may_hobble_trump_131778.html
    .

    Yes, trump is moving up in the polls, but there remains an unusually large undecided vote for this point in the race.

    And, it is hard to say if those not yet in clinton’s camp , but were in obama’s in 2012 (sander’s supporters?) would really swing trump’s way anyway.

    GOTV will be critical this year. On that, Dems seem to have a superior, and better funded operation. But, who knows for sure, as not much has been public about trump’s GOTV operation.

    This still seems to be a “clinton’s to lose” election.

    That said, neither are getting my support.

  31. I think some of the criticisms of physicsguy’s liner progression (above @ 4:21 pm) are missing something:

    Two weeks ago I fit a linear plot to the 538 odds data and it projected Trump reaching near 75% by Nov 8.

    Whether such criticisms are fact baased or not, remember that physicsguy is saying that his current progression indicates that Trump will have a 75% chance of winning, not that he will win by 75%. That’s very different. Physicsguy, correct me if I’m wrong.

  32. @T – think we understood that part, but, in any case, it is still a linear progression in a world that doesn’t operate linearly.

    The predictive power of that progression and $4.15 will get one a Venti Mocha at Starbucks drive thru, and that’s about it.

  33. If 6 weeks’ maternity leave (Hillary wants 12 weeks!) will let the famous suburban Republican women we’ve heard so much about vote for Trump, or at least stay home and not vote for the Evil Empress, it’s a price I’m willing to pay.

  34. “Whether such criticisms are fact baased or not, remember that physicsguy is saying that his current progression indicates that Trump will have a 75% chance of winning, not that he will win by 75%. That’s very different. Physicsguy, correct me if I’m wrong.”

    That’s basically correct. ALL I’m saying is that the data CURRENTLY is best fit with a linear function, AND if extended out to Nov 8, intercepts the 75% probablility for Trump. NOTHING MORE. I put in all the caveats about using linear functions for complex systems, but no one seems to read those.

    I fully expect toward mid October that the fit will be more logarithmic, than linear. But until the data shows such a trend, nothing more to do.

    What is a hard lesson to learn in doing science is that what we EXPECT to happen can’t color what the data is showing now.

  35. Big Maq,

    “The predictive power of that progression . . . .” I agree, as I noted in an earlier discussion with you, this is all speculative. Only on November 9th will we really know who/which polls were accurate and to what extent.

    I disagree with you about how the world operates though. To say that the world doesn’t operate in a linear fashion is not correct. The theory of large numbers presumes precisely that, and with that data, insurance companies can predict, to within about one quarter of one percent, the number of people who will die in any given age group in a given year as well as the delta, the change, in that rate over longer terms. You are right that the world certainly does not always or consistently operate in a linear fashion, especially on a short term basis, but sometimes it does do exactly that.

  36. “To say that the world doesn’t operate in a linear fashion is not correct. The theory of large numbers presumes precisely that” – T

    We are talking about the predictive value for an election. In that situation the world does not work linearly.

    We are not talking about rolling the dice 1000 times, where the chance of each outcome is known ahead of time. We are not talking about people driving a car every day for their entire lifetime to assess the probability of an accident. LLN does not apply as the underlying factors / behaviors are not nearly so predictable.

    We are talking about a single election, and whether or not a linear progression of polls has much predictive value.

    It does not.

    It is a misuse of mathematics / statistics to make a point with it, as if it were meaningful, even when adding caveats.

Leave a Reply

Your email address will not be published. Required fields are marked *

HTML tags allowed in your comment: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>

↑