Announcement

Collapse
No announcement yet.

2016 Presidential Election Poll Update Thread

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Originally posted by Surrey View Post
    Thanks, so California has 42 times the population but only 18 times the votes.
    Welcome to the Electoral College.

    Comment


    • Originally posted by Daemon of Decay View Post
      Welcome to the Electoral College.
      What do you want, Mob Rule?
      The people in the rural areas and in the rust belt of our country finally got a say, they are tired of this shiete, HRC would have screwed up this Country for decades to come with her SCOTUS appointments, that's why I am glad Trump won, I don't like Trump but I voted for him just to keep that Power Hungry Bi-ch away from the WH, she lied even in her concession speech "this election was not about one person but our country" I call BS on that, she lied again!
      Last edited by Trung Si; 09 Nov 16, 21:16.
      Trying hard to be the Man, that my Dog believes I am!

      Comment


      • Leaving aside drawbacks and merits of different electoral systems, but the same people that screamed and shouted about Trump not accepting the election results last week are now protesting. You have a better case when you aren't a hypocrite.

        I don't like Trump, but all this hysteria is a bit childish.
        It is always more difficult to fight against faith than against knowledge.

        Косово је Србија!
        Never go to war with a country whose national holiday celebrates a defeat in 1389.

        Armored Brigade

        Armored Brigade Facebook page

        Comment


        • These were the final polls posted before the election. A 4 to 6 point lead would equate to a 5 - 7.5 million vote lead and an EC landslide.

          IBD/TIPP - Trump +2
          McGlatchy/Marist - Clinton +1
          Rasmussen - Clinton +2
          Bloomberg - Clinton +3
          Reuters/Ipsos - Clinton +3
          Econ/YouGov - Clinton +4
          FOX - Clinton +4
          CBS - Clinton +4
          ABC/WaPo - Clinton +4
          NBC/SM - Clinton +6
          Monmouth - Clinton +6
          {}

          "Any story sounds true until someone tells the other side and sets the record straight." -Proverbs 18:17

          Comment


          • Originally posted by BorderRuffian View Post
            These were the final polls posted before the election. A 4 to 6 point lead would equate to a 5 - 7.5 million vote lead and an EC landslide.

            IBD/TIPP - Trump +2
            McGlatchy/Marist - Clinton +1
            Rasmussen - Clinton +2
            Bloomberg - Clinton +3
            Reuters/Ipsos - Clinton +3
            Econ/YouGov - Clinton +4
            FOX - Clinton +4
            CBS - Clinton +4
            ABC/WaPo - Clinton +4
            NBC/SM - Clinton +6
            Monmouth - Clinton +6
            You'll notice that most of these will have a 4-5% margin of error, which would make the final results (roughly Clinton +1) within the 95% confidence interval.

            The polls say nothing about electoral votes at all, much less electoral landslides. These polls only reflect pre-election data concerning national voting (aka the popular vote) which has a correlation to the Electoral College, but it isn't a direct causal one. As we saw, Clinton can win the popular vote (according to the polls) but lose the Electoral College.

            When it comes to electoral college, one has to look at individual states if one wants to draw conclusions, since each state's vote is essentially an independent vote.

            Comment


            • Originally posted by nikolas93TS View Post
              Leaving aside drawbacks and merits of different electoral systems, but the same people that screamed and shouted about Trump not accepting the election results last week are now protesting. You have a better case when you aren't a hypocrite.

              I don't like Trump, but all this hysteria is a bit childish.
              Oh it is. My girlfriend shared was worried about losing a friend when her die-hard democrat friend began lashing out at anyone and everyone for Trump's victory. The petulant (and thankfully limited) number of people screaming "not my president" are pathetic.

              Anyone who has that attitude against a duly elected official, whether it was Trump, Obama or Bush (and I've seen it with all three now) spits on the nation's principles.

              Comment


              • Originally posted by Daemon of Decay View Post
                The polls say nothing about electoral votes at all, much less electoral landslides.
                I didn't say they did. I said ...a 4-6 point margin equates to 5 to 7.5 million votes.

                125,000,000 (votes cast in election) X .04 = 5 million

                Such a lead would create an electoral landslide. Inevitable.

                As we saw, Clinton can win the popular vote (according to the polls) but lose the Electoral College.
                Her lead in that category is a very slim 0.2%, not a full 4 or 6 point lead.

                You'll notice that most of these will have a 4-5% margin of error
                Check again. Margin of error in parentheses-

                IBD/TIPP - Trump +2 (3.1)
                McGlatchy/Marist - Clinton +1 (3.2)
                Rasmussen - Clinton +2 (2.5)
                Bloomberg - Clinton +3 (3.5)
                Reuters/Ipsos - Clinton +3 (2.3)
                Econ/YouGov - Clinton +4 (n/a)
                FOX - Clinton +4 (2.5)
                CBS - Clinton +4 (3.0)
                ABC/WaPo - Clinton +4 (2.5)
                NBC/SM - Clinton +6 (1.0)
                Monmouth - Clinton +6 (3.6)
                {}

                "Any story sounds true until someone tells the other side and sets the record straight." -Proverbs 18:17

                Comment


                • Originally posted by BorderRuffian View Post
                  I didn't say they did. I said ...a 4-6 point margin equates to 5 to 7.5 million votes.

                  125,000,000 (votes cast in election) X .04 = 5 million

                  Such a lead would create an electoral landslide. Inevitable.

                  Her lead in that category is a very slim 0.2%, not a full 4 or 6 point lead.


                  Check again. Margin of error in parentheses-

                  IBD/TIPP - Trump +2 (3.1)
                  McGlatchy/Marist - Clinton +1 (3.2)
                  Rasmussen - Clinton +2 (2.5)
                  Bloomberg - Clinton +3 (3.5)
                  Reuters/Ipsos - Clinton +3 (2.3)
                  Econ/YouGov - Clinton +4 (n/a)
                  FOX - Clinton +4 (2.5)
                  CBS - Clinton +4 (3.0)
                  ABC/WaPo - Clinton +4 (2.5)
                  NBC/SM - Clinton +6 (1.0)
                  Monmouth - Clinton +6 (3.6)
                  Let's use the Monmouth University poll to discuss as an example of a poll in error since they're both off the mark and not a poll held by a television news media outlet (NBC, MSNBC, Fox, etc. love their own numbers, but considering how they've abused polls in this election and others, they're not the most reputable sources).

                  You'll note that their figures are much further off the mark than other polls (the top 5 there) which were on the mark, while still claiming roughly the same margin of error.

                  Why is that?

                  Well, one has to start with their methodology to see the warning signs.
                  The Monmouth University Poll was conducted by telephone from November 3 to 6, 2016 with 802 registered voters in the United States. The results in this release based on all registered voters have a margin of error of 3.5 percent. Results based on likely voters only have a margin of error of 3.6 percent. The poll was conducted by the Monmouth University Polling Institute in West Long Branch, NJ.

                  Sounds good... on the surface.

                  But their numbers are small, and more importantly their criteria for "likely voter" is not expanded upon. Since we don't have their data on hand to analyze ourselves, the only way to measure how they failed is to examine their results to judge how effective their equations have been.

                  Thankfully, 538 has done the leg-work there. There are other resources like 538, but I'm using them for convenience.

                  When we look up Monmouth, there are two figures that stand out. The first is their "85% races called correctly" figure. That in isolation says a bit, because it illustrates that even their flawed methodology is still far superior to just random guessing.

                  But it's also a red flag. If you look at the rest of the pollsters studied on the 538 page you'll see many (roughly 2/5ths or so) of the polls there have a much higher rating. Now, that too can be misleading: 85% accuracy over 71 separate races is actually a more useful resource than one that has 100% accuracy in 1 race, but it's also something to note.

                  The second figure that we have to look at is "simple average error", which is 5.5. Now, what makes that interesting is when you take that simple average error and apply it to Monmouth's provided figures for the general election.

                  That means that when you adjust Monmouth's Clinton +6% figure according to their "real" error rate, and you get Clinton +.5% - a number that reflects the actual election results.

                  So why is it that Monmouth University is claiming 3.6%, when analysis shows their actual margin is much higher?

                  Well, that's because there is a flaw in their methodology they've not addressed. Like The Literary Digest, the best reasoning for their failure this election has to do with a mixture of different mistakes in their data created by a flawed methodology.

                  For example, read this from Monmouth's website:
                  The Monmouth University Poll was sponsored and conducted by the Monmouth University Polling Institute from November 3 to 6, 2016 with a national random sample of 802 registered voters. Interviews were conducted by a live caller in English, including 401 drawn from a list of registered voters (201 landline / 200 cell phone) and 401 using random digit dial (201 landline / 200 cell phone). Monmouth is responsible for all aspects of the survey design, data weighting and analysis. The final sample is weighted for age, gender, race and partisanship based on voter list and U.S. Census information. Data collection support provided by Braun Research (field), Aristotle (voter list sample), and SSI (RDD sample). For results based on this sample, one can say with 95% confidence that the error attributable to sampling has a maximum margin of plus or minus 3.5 percentage points (unadjusted for sample design). Sampling error can be larger for sub-groups (see table below). In addition to sampling error, one should bear in mind that question wording and practical difficulties in conducting surveys can introduce error or bias into the findings of opinion polls.

                  Now, reading this, there is one MAJOR flaw I see staring me in the face. Can you see it?

                  Yes, that's right: landlines. The Monmouth poll artificially ensures that half of their respondents come from those households possessing a landline telephone.

                  Why is that important? Because landline telephones are not a ubiquitous item anymore. The possession and use of a landline in the age of smart phones is a major red flag to me.

                  Consider this article from last year stating that only 40% of American households still had landlines.

                  Or how about this slightly older article pointing out that "Older and poorer people are more likely to have landlines. For people age 25 to 29, 66 percent have no landline. For people over age 65, its 14 percent."

                  If we assume those trends were continuing, it means these numbers would have been even more striking now. It also means that Monmouth's sample was being tainted by over sampling from the poor and the elderly, which as you can imagine would skew the data. They were getting 50% of their sample from less than 40% of the population. Again, I recommend looking up the story of the Literary Digest to see why this is a bad thing for pollsters to do.

                  Now, compare Monmouth and their erroneous figures with IBD and their accurate one. Well, once again, to see the difference we have to look at their methodology.
                  Results based on survey of 1107 likely voters conducted from 11/4 - 11/7. Margin of error: +/- 3.1 percentage points. Party identification breakdown: (Unweighted) 361 Democrats/367 Republicans/362 Independents; (Weighted) 361/333/381. Results in table above are based on weighted sample. IBD's polling partner TechnoMetrica uses "traditional" telephone methodology using live interviewers for data collection for its public opinion surveys. Roughly 65% of interviews come from a cell phone sample and 35% from a Random Digit Dial (RDD) land line sample.

                  You know what stands out here? That their results from the "cell phone sample" match up with the previous data concerning cell phones and landline use. They were getting 65% of their interviews from cellphones, and less than 40% of Americans have landlines, meaning their sample already more closely reflects the American population than an arbitrary decision to demand half of their sample must be from that shrinking landline minority (a la Monmouth).

                  Again, without seeing the hard data first hand I can't say for sure, but in this superficial once-over my impression of why Monmouth was so off while IBD was absolutely correct comes down to flaws in their methodology like that. My money is on Monmouth having over sampled from the poor, skewing their figures towards Clinton.

                  But the common talking head on your favorite cable news network wouldn't consider that when bringing up "the latest polls say..." in their evening broadcast.

                  All polls are not equal. What this election has done is illustrate that divide, shine light on previously hidden flaws in methodology, and helped challenge some erroneous assumptions based on previous track records.

                  Which is why this election is thrilling to any political junkie out there!

                  Comment


                  • Originally posted by Daemon of Decay View Post
                    Oh it is. My girlfriend shared was worried about losing a friend when her die-hard democrat friend began lashing out at anyone and everyone for Trump's victory. The petulant (and thankfully limited) number of people screaming "not my president" are pathetic.

                    Anyone who has that attitude against a duly elected official, whether it was Trump, Obama or Bush (and I've seen it with all three now) spits on the nation's principles.
                    I didn't like Obama....ironically the only speech of his I thought was well delivered and not pandering to his base or otherwise being a jackwagon was his speech acknowledging Trump. And I appreciated that, since the Hillary Camp was playing the petulant child card. That being said, I tried to give him a fair shake until he did something ridiculous that poisoned him in my eyes forever.....it didn't take long since Obamacare happened so fast, but I did give him a chance.

                    Everyone should give Trump a chance. I'm not asking for 100 days, that's too long to wait for the Twitter generation. But one month, so around the end of February, at least. If he hasn't imposed the Gestapo, started rounding up Mexicans to go to the showers, or nuked someone by then, it should be pretty obvious that he's not Voldemort.
                    Tacitos, Satrap of Kyrene

                    Comment


                    • Originally posted by TacCovert4 View Post
                      I didn't like Obama....ironically the only speech of his I thought was well delivered and not pandering to his base or otherwise being a jackwagon was his speech acknowledging Trump. And I appreciated that, since the Hillary Camp was playing the petulant child card. That being said, I tried to give him a fair shake until he did something ridiculous that poisoned him in my eyes forever.....it didn't take long since Obamacare happened so fast, but I did give him a chance.

                      Everyone should give Trump a chance. I'm not asking for 100 days, that's too long to wait for the Twitter generation. But one month, so around the end of February, at least. If he hasn't imposed the Gestapo, started rounding up Mexicans to go to the showers, or nuked someone by then, it should be pretty obvious that he's not Voldemort.
                      You deserve some rep because your intent is glorious and I support it wholeheartedly, but the reality is that if you're a political partisan you'll never give the other side a real "chance". Think about with Obama: how could he have followed the Democratic platform and succeeded in bringing you aboard? How can Trump fulfill his campaign promises to Republicans and convince Democrats he's worth supporting?

                      I concede that if the "giving them a chance" concept is tied into panic and fear - giving him a chance to prove he's not a super monster - then that could work, but I don't think it would, sadly.

                      Comment

                      Latest Topics

                      Collapse

                      Working...
                      X