Polls (Page 2)

Polls are still crazy.

LA Times has the race tied
Ipsos has Clinton +4
McClatchy has Clinton +15  (lol)

It's really interesting how they can be so far apart.

One fun fact:  Obama was 10 points down to McCain in September 2008.
I do enjoy all my FB friends who have completely written off the Trump.  They need to be careful how much the push the "he's over" narrative because there would be no need for anyone to go out and vote for Hillary.  I mean, she's already won and all. 

You can split hairs about how big a difference there is between the two candidates, but there's no arguing the trend over the last month.

Might as well call the election.   

It's more likely that Fox News was suffering from selective poll syndrome

Agree.  It is hard to motivate you base when you are up by double digits.  One thing for certain, Trump's campaign is looking at the same polls and know something has to change or he is done.  So they will do something.  I heard this weekend one of the talking heads say the best news for Trump is that this is only August.

My guess it there will be more emails to come out.  I've said all along that ALL the emails are out there somewhere.  It is naïve to think that foreign intel agencies did not even try to hack the US Secretary of State's email....and from what the FBI said, her server lacked even the most basic of anti-hacking software such as what you'd find in gmail. 

So I do think there will be another shoe to drop.....but sadly, I don't think people care.

lol, what I was referring to was the time when McCain took the lead in the first part of September.  He had a great convention and everything was rolling.  Everyone was bashing on Obama being the least qualified candidate in history and that he was so dangerous he couldn't even pass a security background check (which was and is true ironically enough).  McCain was up in almost all the polls like Hillary is now with some outliers in the double digits.
What I'm getting at is I got over confident and felt as the election was in the bag.  Obviously I was very wrong and events such as the economic collapse had a massive impact on the election.

I guess my main point is that there's a lot of time, debates, and an October surprise or two yet to impact the election.  The only reason Hillary is up on Trump is because she's doing a better job of attacking Trump personally.

On a side note I read an article that was a little interesting over the weekend.  It mentioned how there are several different major polling methodologies that people are using.  There are Clinton vs. Trump with registered voters, Clinton vs. Trump with likely voters, Clinton vs. Trump with Johnson (both registered & likely voters), and Clinton vs. Trump with Johnson and Stein (both registered & likely voters).

They all have a trend of Hillary on top, but the most accurate scenario with Trump/Hillary/Johnson/Stein with likely voters are mostly low single digit leads with Hillary on top.  Basically within the margin of error.

RCP does a good job of showing the polls and lists RV vs. LV as the poll and has a tab with the 2, 3, and 4 person race.  You'll notice they're very heavy on RV polls and the LV polls are a lot closer.

With these two going at it I'm almost certain there will be about 5 more shoes to drop on both sides.  hah

This is what I don't get about polls.  41% to 38% in a poll that likely has an admitted margin of error of +/- 3%.  Also 41 + 38 = 79 so still 21% undecided, other candidates or refuse to answer.  I will vote for Trump but I refuse to take a poll.  People need to make up their minds and not be swayed by how winds are blowing in the polls.

the takeaway here isn't the number, its the fact that a democratic candidate is leading, or tied if you want to take the margin of error, in a very red state.

Not sure about the full poll you were looking at but the Libertarian and Green parties were sitting at something around 9% and 5%, which gets you into the mid-90s for "decided".

Yeah, you are still missing the point about polls.  You cannot say with certainly that one democrat candidate is leading or tied if your data are only accurate to +/- 4% (on a good day).  All you can say is the candidates are close enough that we cannot tell you who is ahead.....however are data suggest a trend towards one candidate.  In engineering you assume the WORST case not the best case....which is what they are doing. 

For a 4 point MOE, they should be reporting: "In the best case Clinton is leading 3 points and in the in the worst case, Trump is leading my 1 %."

I'm not just talking this particular poll but all polls.  They really should report best and worst case and not ignore the margin of error. 

This isn't a poll, but I've seen it posted several times on conservative sites and thought it was interesting.
Some of these are skewed because of Trumps celebrity status prior to politics such as twitter and facebook, but I found the Reddit one most fascinating.  My understanding is that Reddit is a heavy young male demographic which Trump certainly draws some support, but not to that level of skew.

I know these numbers mean nothing, but just a curious datapoint for us political watchers.  btw, these could also be a symptom of Hillary's team not being as engaged on social media which was a big problem the Republicans had in 2008 and 2012. 

Polls on the popular vote, of course, are a little silly to begin with.

It's more useful to look at state-specific polls in the battlegrounds (Florida, Ohio, etc).  California's boost to Clinton and Texas to Trump nationwide mean squat because those states are already decided anyhow.

The press likes quick tidy summaries though.

Even Breitbart agrees that Clinton is in the lead.

http://www.breitbart.com/2016-presidential-race/2016/08/14/breitbartgravis-poll-hillary-clinton-leads-donald-trump-42-to-37/

There's no question she's in the lead in the polls overall.  My only concern is how accurate are the polls.

If Clinton is truly up as the polls indicate (and she may very well be) then I'd expect to see some sort of confirmation outside of polls such as general enthusiasm and social media buzz etc, but it's like Crickets anywhere outside of the polls.

Even NBC News (who has a more liberal following) did an online poll which is obviously not scientific in any way last week and Trump was at something like 60% with Clinton only getting 10% of the vote.  Even Stein had more votes that Hillary in that poll. (ironically, this poll is nowhere to be found anymore and my old link is dead)

Obviously Trump is in no way shape or form up that big, but if Hillary is more popular than Trump nationally and Trump is on such a slide why isn't it showing at rally's, in social media, or with any populous mechanism such as online polls.    Obama did really well in non-scientific online polls if I recall in 2008/2012.

I mentioned before that I've been burnt in the past trying to dissect polls and not trusting them.  Overall, they ended up being fairly accurate in the end so I came to the point of just believing them blindly.  However, yet again I find myself in the camp struggling to understand them.  haha

Sorry, it's not that I don't understand them because they are fairly transparent on their methodologies.  It's more of a question of the same methodologies that have been working historically continuing to work in this election or not.  I know I'm playing devils advocate a little, but I also am genuinely wondering if they're legit.

If you recall during the GOP primaries they were way off on several occasions where they failed to capture the actual feel of the people. 
(These are from RCP averages)
Wisconsin had Trump up by 10% (he lost by 13%)
New York had Trump at 50% (he got 60%)
Indiana had Trump at +10% (he won by 17%)

I guess what I'm getting at is that the pollsters failed to capture Trump accurately in many of the races in the Primaries so how can we trust they're now completely accurate.  They all use similar methodologies.
I haven't looked at Hillary and Bernie to see how they compared to the actuals.