Break Glass - 2020 USA Elections

1505153555682

Comments

  • Gramps49 wrote: »
    Indeed.

    This is probably a silly question, but if Mr Biden draws ahead quite quickly (as, please God, he will), could he declare himself the winner at that point?

    Or IOW, if Trump thinks he can do it, could Mr. Biden do the same?

    Mind you, I dread to think what the immediate consequences might be, so it's a hypothetical question...

    Traditionally, the person that is elected will wait until he gets a concession call from his opponent. But this is not a traditional election.

    Trump may just jump ahead and declare he has won as soon as the Eastern Time Zone polls close in order to try to suppress the vote in states in the latter Time Zones.

    Trump will not go down without a fight (in the courts, at least) but my bet is that he will sorely lose the court challenges.

    I think the greatest fear is that the militias pledged to him will try to interfere. They already harassed a bus dedicated to the Biden campaign in Texas and blocked traffic on a freeway between New York and New Jersey. But I think if they do act up, they will be facing a military that is dedicated to defending the constitution. Their little pea shooters will not have much effect on tanks and armored personnel carriers.

    In the end, we will know by 6 January when the electoral votes are counted.

    O dear. Yes, there's perhaps rough weather ahead, but I guess Mr Biden is not one to act in such a way as to make the violence even more inevitable (IYSWIM).

    Armed militias versus the regular military is not a happy thought.

  • edited November 2020
    I've voted in 2 elections thus far this fall, both by mail given pandemic. Provincial and city/school board. The goals of the agencies managing the elections are to ensure everyone gets to vote. It's actively encouraged. There are additional rules for signs not allowed with some distance of polling places when people go in person. initimation or discussing votes with people queued up to vote isn't allowed. They publicize continually how long the wait is for advance polls. Mostly people are in and out in about 5-10 mins on election days, with polling places within walking distance of where people live.

    Which leads me to ask if "managing" polling places to have long wait times, signs and intimidation of voters while lined up to vote, and putting polling places in difficult places for people to get to - are these considered legitimate tactics in elections, and if they're not, why do they occur? --unless the news is reporting wrongly about this election, and suggesting that various tactics to suppress or cancel voting rights isn't true.
  • Also there could be fraud of a ballot form 'going astray' in the household and being fraudulently filled in and sent in the voter's name. 'cos he's young and doesn't understand what's at stake; let's MAGA!

    There has long been suspicion of this kind of low-grade vote harvesting in mail-in elections. I suspect it will be about as rare in this election as it has been in previous elections, and is unlikely to matter.

    I remain of the firm belief that any small effect due to voter fraud is vastly dominated by the larger effect due to various forms of voter suppression.

    A fair election is one in which each legal voter who wishes to express a preference has their vote counted. Having your vote be the subject of fraud is not worse for the election than just not having your vote counted because of problems with the electoral roll, or not being able to vote because of long lines at your polling place - it's the same.
  • I'm sure the majority of Americans would agree that these *tactics* are indeed unfair, if not downright illegal - but since when have Trump and/or his minions been bothered as to whether or not anything they do is fair or legal?
  • Riots coming? 15 sec video of Washington DC boarded up. https://twitter.com/RexChapman/status/1323294727316361222?s=20
  • Riots coming? 15 sec video of Washington DC boarded up. https://twitter.com/RexChapman/status/1323294727316361222?s=20

    Possibly. But it's just a few seconds of boarded up streetfronts, with the credits stating bereft of any other evidence that this is the state of every building in DC right now, and that the obvious purpose is to protect against riots. I think we need a little more information to draw conclusions.

    FWIW, even up in peaceful Canada, when revellers and hockey fans riot, they've been known to smash in bank windows and torch police cars. I think an enraged mob in DC would make short work of a few wooden boards.
  • IMHO, it isn't for a candidate to declare themself a winner, or shouldn't be. The media shouldn't decide it, either. People get really upset when the media "makes the call" that So-and-so is the clear winner, when the votes are still being counted. Even more upset when the media gets it wrong.
  • Golden Key wrote: »
    IMHO, it isn't for a candidate to declare themself a winner, or shouldn't be. The media shouldn't decide it, either. People get really upset when the media "makes the call" that So-and-so is the clear winner, when the votes are still being counted. Even more upset when the media gets it wrong.

    Pronouncements by the media, or by the candidates themselves, are more or less meaningless. Al Gore called George W Bush to concede the 2000 election, then un-conceded when he realized how close Florida was.

    It is courteous and civilized for a candidate to concede defeat once it's clear that their opponent has won. But it doesn't mean anything, and it doesn't stop the votes from being tallied.
  • Barnabas62Barnabas62 Purgatory Host, 8th Day Host, Epiphanies Host
    edited November 2020
    Back to 538 at close of day. Current probabilities are as follows.

    Biden 90% to win overall, 85% to win in Pennsylvsnia, 69% to win in Florida, 68% to win in Arizona, 65% to win in North Carolina, 58% to win in Georgia.

    Bookmakers have Biden at 65%, Trump at 35%. Someone in the UK has bet £1 million on Biden, said to be the biggest political bet ever, and stands to win over £0.5 million if Biden wins.

    I'd say that person got good odds.

    But Trump can still win.


  • Wasn't 538 giving similarly astronomical chances to Hillary right before the 2016 election?

    I remember thinking it was odd that there was so much focus on "X% chances of winning" as opposed to the more usual "X% of people planning to support this candidate", and after Trump won, thinking that that might be the death knell for the new style of forecasting.
  • DafydDafyd Shipmate
    stetson wrote: »
    Wasn't 538 giving similarly astronomical chances to Hillary right before the 2016 election?
    IIRC 538 gave Clinton roughly a 2/3 chance of winning. That's far from a certainty. 90% is more comfortable but not a certainty either.

  • stetson wrote: »
    Wasn't 538 giving similarly astronomical chances to Hillary right before the 2016 election?
    No. In 2016 just before the election, 538 had Clinton with a 71.4% chance and Trump with a 28.6% chance, which was higher odds for Trump than most other polls or models had.

    And 538 banged the drum that a 28.6% chance was a real chance, just as they’ve been clear this cycle that while a 10% chance means the odds aren’t in Trump’s favor, it also means he can still win.

  • Thanks for correcting my misrememberings.
  • The New York Times podcast The Daily discussed the election this morning and noted that although many Democrats are leery of polls given the pain of 2016, that the models are showing Biden so clearly over Trump that if polls are wrong again then it means that there's something fundamentally wrong with how polling is done. It seems more likely to me that Trump steals the election outright through the courts, as he's always been most easy running to the courts to solve his problems, but we'll have to wait and see.
  • Amanda B ReckondwythAmanda B Reckondwyth Mystery Worship Editor
    ECraigR wrote: »
    there's something fundamentally wrong with how polling is done

    Well, for starts, no poll has ever contacted me.
  • ECraigR wrote: »
    there's something fundamentally wrong with how polling is done

    Well, for starts, no poll has ever contacted me.

    I know you mean this whimsically, but it gets me thinking...

    As someone who has actually worked the phones at a polling firm, it's funny to hear people complain "No one ever polled me". Because I'd be willing to bet that there's a pretty big venn overlap between people who complain about not getting polled, and people who would immediately hang up on any pollster who did call them.
  • I don't hang up on pollsters. My answering machine has been off for several months -- if I don't recognize the person calling (on caller i.d.) it isn't answered.
  • Pigwidgeon wrote: »
    I don't hang up on pollsters. My answering machine has been off for several months -- if I don't recognize the person calling (on caller i.d.) it isn't answered.

    Which is cool. As long as you don't complain about not getting polled. (And I'm not saying you do.)
  • stetson wrote: »
    Pigwidgeon wrote: »
    I don't hang up on pollsters. My answering machine has been off for several months -- if I don't recognize the person calling (on caller i.d.) it isn't answered.

    Which is cool. As long as you don't complain about not getting polled. (And I'm not saying you do.)

    I don't.
    :smile:
  • ECraigR wrote: »
    there's something fundamentally wrong with how polling is done

    Well, for starts, no poll has ever contacted me.

    Well that a pollster has never contacted you personally doesn’t indicate much. There’s over 300 million people in the country, one can’t reasonably expect that every one one of them will talk with a pollster at some point. I’ve been called for jury duty three times, while I have friends decades older than me who’ve never been called.
  • Barnabas62Barnabas62 Purgatory Host, 8th Day Host, Epiphanies Host
    It’s by no means unknown for horses to win at odds of 10 to 1.

    It’s a bit more problematic when the owner wishes to bring the winning post forward to a point where the horse may be in the lead.

    As Huffington Post puts it succinctly, Trump’s closing message is “I will cheat”.

    If necessary he will. If his loss is obvious on election night he won’t.

    Re 538 accuracy. In 2012 they were giving Romney an 8% chance of winning in the EC despite the National polls showing only a narrow Obama lead. They got the EC just about on the nose.

    The 538 model is pretty sophisticated in its assessment of probabilities but that’s basically all it does. And it has no way of assessing the impact of these attempts to disenfranchise voters. In the end that might be a matter of how only 9 people vote.
  • Barnabas62 wrote: »
    It’s by no means unknown for horses to win at odds of 10 to 1.

    True, and Trump could still win, but a horse at 10 to 1 could be the favourite in a crowded field (or there could be a number of 10 to 1 shots in a given race, so one of them winning would be less surprising) whereas in a 2 horse race like this the 10 to 1 shot is pretty unlikely to be first over the line.
  • Barnabas62Barnabas62 Purgatory Host, 8th Day Host, Epiphanies Host
    Trouble is the two horses are running in 51 races!

    Minor adjustments to 538 forecasts.

    Biden win Pennsylvania 84%, North Carolina 64%. And there will be no more forecasts from 538.

    It’s been a pretty stable race according to the pollsters. Trump needs more things to go his way than in 2016.

    I wonder what tomorrow will bring?
  • Two old nags ...
  • Boogie wrote: »
    Two old nags ...

    Which one's got the bob-tail?

  • Furtive GanderFurtive Gander Shipmate
    edited November 2020
    Is this equine frivolity to break the tension? As we all (not just Americans) wait to see what happens in the next few days - we want to know how the voting goes, to watch people queueing, vox pops, related news and news punditry, we lap up news of trends going the way WE want, watching anxiously for signs of things going the other way, or voter suppression ... the first results, maps, swings ... and so on, probably for days.

    Most people here on the Ship want Biden to win (as we've seen the disaster the current man has been) but I'm wondering whether if there was a more even split of desired outcomes we'd have a better discussion, or is there such a big division that in any forum that's near impossible? Is is like comparing watching a sports game with fellow supporters of your team or with a mix of supporters from both sides?
  • Far too much at stake to compare it to sports, sorry.
  • Gramps49 wrote: »
    Indeed.

    This is probably a silly question, but if Mr Biden draws ahead quite quickly (as, please God, he will), could he declare himself the winner at that point?

    Or IOW, if Trump thinks he can do it, could Mr. Biden do the same?

    Mind you, I dread to think what the immediate consequences might be, so it's a hypothetical question...

    Traditionally, the person that is elected will wait until he gets a concession call from his opponent. But this is not a traditional election.

    Trump may just jump ahead and declare he has won as soon as the Eastern Time Zone polls close in order to try to suppress the vote in states in the latter Time Zones.

    Trump will not go down without a fight (in the courts, at least) but my bet is that he will sorely lose the court challenges.

    I think the greatest fear is that the militias pledged to him will try to interfere. They already harassed a bus dedicated to the Biden campaign in Texas and blocked traffic on a freeway between New York and New Jersey. But I think if they do act up, they will be facing a military that is dedicated to defending the constitution. Their little pea shooters will not have much effect on tanks and armored personnel carriers.

    In the end, we will know by 6 January when the electoral votes are counted.

    O dear. Yes, there's perhaps rough weather ahead, but I guess Mr Biden is not one to act in such a way as to make the violence even more inevitable (IYSWIM).

    Armed militias versus the regular military is not a happy thought.
    there is no guarantee that it won’t be the armed militias with the regular military. I read an article by a military participant of the Trump Bible photo op bullshit. He said his unit was briefed that the situation was violent and the protestors were aggressive. That set the tone for the encounter that he didn’t realise was bogas until AFTER the encounter was over.
    Not until everything was done and dusted did his brain register that he and his unit had behaved as if they were in a combat situation with an enemy instead of fellow citizens of his own country.
  • Barnabas62Barnabas62 Purgatory Host, 8th Day Host, Epiphanies Host
    Inclined to agree about the nags tangent!

    Anyway I’m steering clear of the tension tonight, for the sake of my somewhat dodgy almost 78 year old heart. What’s the phrase? Wake me up when it’s over.

    Thinking about nags I’ll probably watch some reruns of ‘Heartland’. Most characters in that are a lot more wholesome, and encourage the view that many differences can be sorted out by talking about them. And the Alberta scenery is beautiful. Yes, a few hours escapism will be better for my health than being glued to CNN.
  • Barnabas62 wrote: »
    But Trump can still win.

    Any major party nominee has a non-trivial chance of winning the presidency. A 10% chance is not a 0% chance.
  • @Barnabas62 - yes, I think many of us on this side of the Pond will be praying for America, but still keeping clear of the tension (we have enough shite of our own to contend with).

    BTW, Mr Biden is also 78...
    :wink:
  • Amanda B ReckondwythAmanda B Reckondwyth Mystery Worship Editor
    ECraigR wrote: »
    There’s over 300 million people in the country

    And I'll never understand how polling 300 of them (or however many) gives an accurate prediction of anything.
  • Nick Tamen wrote: »
    stetson wrote: »
    Wasn't 538 giving similarly astronomical chances to Hillary right before the 2016 election?
    No. In 2016 just before the election, 538 had Clinton with a 71.4% chance and Trump with a 28.6% chance, which was higher odds for Trump than most other polls or models had.

    And 538 banged the drum that a 28.6% chance was a real chance, just as they’ve been clear this cycle that while a 10% chance means the odds aren’t in Trump’s favor, it also means he can still win.

    For those who are interested, 538's final 2016 predictions are still available online.
    ECraigR wrote: »
    There’s over 300 million people in the country

    And I'll never understand how polling 300 of them (or however many) gives an accurate prediction of anything.

    There might be ~330 million American residents at this point, but there are only ~230 million voting age citizens, according to best estimates.
  • ECraigR wrote: »
    There’s over 300 million people in the country

    And I'll never understand how polling 300 of them (or however many) gives an accurate prediction of anything.
    This article is hardly an in-depth look into polling methodology, but ti does give an idea on how polls should be more accurate than in 2016 on one hand and on the other why all bets are off super conditional in 2020.
  • lilbuddha wrote: »
    This article is hardly an in-depth look into polling methodology, but ti does give an idea on how polls should be more accurate than in 2016 on one hand and on the other why all bets are off super conditional in 2020.

    Which article?
  • Barnabas62 wrote: »
    It’s by no means unknown for horses to win at odds of 10 to 1.

    Or as I like to say, it would be surprising if nothing surprising ever happened.
  • For those who want to watch events unfold on Election Night (and live in a time zone conducive to such activity), the states to watch are Florida, North Carolina, and Ohio. Why these states? Three reasons:
    • They're swing states, so the outcome of the presidential race there is uncertain.
      -
    • The polls there will close (relatively) early. That's 7:30 pm EST in North Carolina and Ohio. Florida has asynchronous polling times due to the fact that part of the Florida panhandle is in the Central Time Zone. Polls close at 7:00 pm local time in Florida, which means some won't close until 8:00 pm EST (which is 7:00 pm CST), at which point media outlets will feel free to report on results there.
      -
    • And the most important reason for keeping an eye on these states is that they all start processing and counting mail-in ballots before Election Day. So unlike a lot of states which don't start counting mail-in votes until the day of the election (or in some cases after the polls close), the results in these states have a decent chance of being known in full (or close to full) on Election Night. The one caveat here is that Ohio and North Carolina will accept mail-in ballots that arrive after Election Day provided they're postmarked on or before Election Day, so if either of those states is really close the winner might not be known for a few days.

    Donald Trump needs to win all three of these states to be re-elected*. He can win all three and still lose, but if he loses even one of these three states his chances of re-election* grow very remote. If he loses two or all three he has no realistic path to victory.
  • Looks like we won't get the results from North Carolina until at least 8:15 pm EST.
    Plainview fire station in Sampson County had printer issues and precinct officials did not go to backup labels. Polls opened 45 minutes late. At NE Clinton site in Sampson Co there was also printer issues. Polls opened 24 minutes late

    NCSBE is extending voting at the Plainview fire station for 45 minutes and NE Clinton site for 24 minutes. Vote was 3-2
  • Crœsos wrote: »
    lilbuddha wrote: »
    This article is hardly an in-depth look into polling methodology, but ti does give an idea on how polls should be more accurate than in 2016 on one hand and on the other why all bets are off super conditional in 2020.

    Which article?

    Give me a break, it is 2020. This article.
  • lilbuddha wrote: »
    Crœsos wrote: »
    lilbuddha wrote: »
    This article is hardly an in-depth look into polling methodology, but ti does give an idea on how polls should be more accurate than in 2016 on one hand and on the other why all bets are off super conditional in 2020.

    Which article?

    Give me a break, it is 2020. This article.

    lilbuddha:

    Are you implying Croesos should have known how to find the article without a link? Honestly, I don't think I would have been able to, since the potential search words seem fairly generic to me.

  • ECraigR wrote: »
    There’s over 300 million people in the country

    And I'll never understand how polling 300 of them (or however many) gives an accurate prediction of anything.

    Statistics and stuff. I’m all for challenging experts, but there’s a bit of math and science behind what they do so unless one is competent in such areas (which I am not) then challenging their methods is difficult.

    That being said, most polls routinely use a larger number than 300.
  • ECraigR wrote: »
    Statistics and stuff. I’m all for challenging experts, but there’s a bit of math and science behind what they do so unless one is competent in such areas (which I am not) then challenging their methods is difficult.

    Suppose, for example, that you have a bag that contains 100,000 balls. Some are red, and some are blue, and you have no a priori knowledge about how the colours are distributed.

    Now suppose you take ten balls out of the bag, at random, and they're all blue. It's theoretically possible that the bag contained 99,990 red balls, and you just happened to select the ten blue ones, but you can see that that's unlikely. Without doing any math at all, you can see that even just picking ten balls at random and having them all come out blue, it's probable that the bag is mostly blue, and you'd be very surprised if the balls in the bag were actually evenly split between red and blue.

    We could do math and quantify that statement - it's not very hard.

    What pollsters do is a little more complicated than that. To start with, they tend to perform stratified random sampling (not all balls are the same, so rather than just grabbing balls at random, they select some big ones, some small ones, and some knobbly ones in proportion to how those balls appear in the big bag population.) Then they apply correction factors to account for things like how likely particular kinds of people are to actually vote, how likely they are to lie to the pollsters (the "shy Tory" phenomenon) and end up with a number.

    And then aggregators like 538 take each individual poll, and combine it with historical data for how well that poll has predicted previous elections, or whether it tends to bias one way or the other, and produce a combined estimate.

    But at root, it all rests on the question "if you have a big bag of mixed red and blue balls, how many do you have to measure to know which one you have most of".
  • Then they apply correction factors to account for things like how likely particular kinds of people are to actually vote, how likely they are to lie to the pollsters (the "shy Tory" phenomenon) and end up with a number.

    That's going to be particularly difficult to adjust for this year, since easier early voting seems to have skewed the American electorate to a higher level of voter participation. It's not going to be greater numbers of the same distribution. By definition higher voter turnout means more people who would have been modeled as non-voters in previous elections showing up at the polls.
  • Crœsos wrote: »
    That's going to be particularly difficult to adjust for this year, since easier early voting seems to have skewed the American electorate to a higher level of voter participation.

    Yes indeed. At root, everyone's model is pinned to previous data, and the more unusual the year is, the more uncertainty you have to add. In normal conditions, higher turnout tends to favour Democrats. Whether that's true this year to the same extent (given that the effect of Covid-19 and widespread mail-in ballots on voters is particularly hard to model), we'll see.
  • stetson wrote: »
    lilbuddha wrote: »
    Crœsos wrote: »
    lilbuddha wrote: »
    This article is hardly an in-depth look into polling methodology, but ti does give an idea on how polls should be more accurate than in 2016 on one hand and on the other why all bets are off super conditional in 2020.

    Which article?

    Give me a break, it is 2020. This article.

    lilbuddha:

    Are you implying Croesos should have known how to find the article without a link? Honestly, I don't think I would have been able to, since the potential search words seem fairly generic to me.
    No. I could have worded it more clearly, perhaps, but it was self-deprecating.
    Meaning 'I screwed that up, give me a break it is 2020.

  • Crœsos wrote: »
    Barnabas62 wrote: »
    But Trump can still win.

    Any major party nominee has a non-trivial chance of winning the presidency. A 10% chance is not a 0% chance.

    /tangent - the number of times I've seen people say "either could be true so it's 50/50"....
  • lilbuddha wrote: »
    stetson wrote: »
    lilbuddha wrote: »
    Crœsos wrote: »
    lilbuddha wrote: »
    This article is hardly an in-depth look into polling methodology, but ti does give an idea on how polls should be more accurate than in 2016 on one hand and on the other why all bets are off super conditional in 2020.

    Which article?

    Give me a break, it is 2020. This article.

    lilbuddha:

    Are you implying Croesos should have known how to find the article without a link? Honestly, I don't think I would have been able to, since the potential search words seem fairly generic to me.
    No. I could have worded it more clearly, perhaps, but it was self-deprecating.
    Meaning 'I screwed that up, give me a break it is 2020.

    Thanks. That interpretation also occured to me. Sorry for jumping the gun.
  • KarlLB wrote: »
    Crœsos wrote: »
    Barnabas62 wrote: »
    But Trump can still win.

    Any major party nominee has a non-trivial chance of winning the presidency. A 10% chance is not a 0% chance.

    /tangent - the number of times I've seen people say "either could be true so it's 50/50"....

    Despite my efforts to beat such thinking out of them in a maths classroom.
  • KarlLB wrote: »
    Crœsos wrote: »
    Barnabas62 wrote: »
    But Trump can still win.

    Any major party nominee has a non-trivial chance of winning the presidency. A 10% chance is not a 0% chance.

    /tangent - the number of times I've seen people say "either could be true so it's 50/50"....

    Despite my efforts to beat such thinking out of them in a maths classroom.

    It's not even just a maths fail though. It's a complete thinking fail.
  • In any case, whether it's an "either" of a 50/50, I'd say the Donald's chances of winning at this point are pretty good. I'm calling this for Trump.
This discussion has been closed.