You guys are down... Says who?... Polls, All of them.

[MongGrel]

Three. Not that your choice of media would allow you to know it.

Zogby: Clinton and Trump in Statistical Tie

Still technically shows Trump trailing.

WTF is a Zogby

 

[BurnItDwn]

The problem with your post is that while it is true that conservatives supporting Trump have important things to wake up to, liberals should understand that they have a similar irrational flaw that drives them to react as irrationally in the face of conservative denial as conservatives are irrational in their denial. Liberals are quite terrified of how irrational conservatives are and react in a correspondingly irrational way to that fact.

The flaw, then, is that while you are right that conservatives need to cease their irrational denial for their own good and the good of everybody else, they will not do so just because liberals are terrified they won't and they will not respond with anything but denial to being informed they are in denial.

Two things have to happen. One is that they need to see why they are in denial and to accept that they have to be in that state. They were powerfully conditioned as children to conform, both to authority and to their group. The torture to insure they did so was brutal. You are asking them to drop the defense they built, their state of denial, that protects them from suffering the memory of the pain they were subjected to. You are also asking them to do that without facing your own similar condition, your fear of their mental state.

I would suggest that you first need to deal with your own fear of them, so that you can sympathize with their state rather than fear and hate them. Then and only then, when you are emotionally free from any need to change them, you can offer your sympathy for their state and suggest that they are not to blame for their own condition and that they might want to consider the possibility that life can be lived without all the defenses they have built. The only person we have any real chance of changing is ourselves. Every person who does so adds a bit more love to the world.

So, in simple terms
Conservatives are afraid of the insane world
Liberals are afraid of the insane conservatives

It would suck if fear was that big of a motivator.

 

[Jaskalas]

Still technically shows Trump trailing.

If that's your reply to a statistical tie, by all means. Tell everyone that it's in the bag, that they should stay home.

 

[zinfamous]

Three. Not that your choice of media would allow you to know it.

Zogby: Clinton and Trump in Statistical Tie

Why do you keep showing polls where Clinton is leading to try and show that she is not leading? They are only "tied" in these polls because Clinton's lead/Trump's deficit is within the MoE.

You are also using largely discredited polls that will always massage the polling to favor the R candidate....which makes your use of these polls even more hilarious: The fact that Hillary actually leads in polls that favor R candidates shows just how large Trump's gap is. I want to say that Rasmussen/Zogby tend to favor R by about 4 or 5 percentage points.

Since it is fundamentally known (national polling aggregates) that Hillary is leading by about 7-8 points right now (and has held for the last 3 weeks), then her 2 point lead in these polls is about where she should be (4-5 point R handicap +2 point D lead here = 6-7 point advantage for Hillary). Your polls, in fact, follow the real trend that she is trouncing Cantaloupe Mussolini.

 

[dullard]

The USC/LA Times poll has a really unique way of looking at the election which has probably inserted a fairly large pro-Republican bias. It's a panel survey so they keep asking the same people from week to week. The problem is that they weight their survey based on who the people in the panel say they voted for in the last election, and it's very common for people to say they voted for the winner when they actually didn't. Therefore they are almost certainly weighting their poll to in a way in which Republican voters are being counted as former Democratic voters, skewing the sample.

Regardless of the results on election day I would bet money that the USC/LA Times poll will predict a result several points more favorable to the Republicans than what actually occurs.

You are correct that the USC/LA Times survey shouldn't be considered as a poll, as they don't randomly poll voters. As, you said, it is a survey of a set panel instead of a poll. So, accuracy of representing the final election isn't the intent, nor is it necessarily important in the USC/LA Times results.

They could have picked a panel that is heavily biased Democratic or heavily biased Republican. That really doesn't matter. What they are trying to do is track CHANGES over time of the same group. If the group goes from +25% Clinton to +30% Clinton (with a Democrat bias) it shows that people are switching to Clinton. Similarly, if that group goes from +25% Trump to +20% Trump (with a Republican bias), it still shows that people are switching to Clinton.

So, accuracy isn't important with that website, precision is.

So, what has the trend shown so far? Clinton's supporters have gone from 80.4% likely to vote to 82.7% likely to vote (a swing in her favor). Trump's supporters have gone from 83.0% likely to vote to 81.3% likely to vote (again a swing in Clinton's favor). If these 3000 people are swinging in Clinton's favor, then the assumption is that America is swinging in Clinton's favor. That is all that website is telling you. It doesn't tell you if one candidate is truly in the lead or not, since that isn't the way they poll people.

 

[dullard]

You are also using largely discredited polls that will always massage the polling to favor the R candidate....which makes your use of these polls even more hilarious: The fact that Hillary actually leads in polls that favor R candidates shows just how large Trump's gap is. I want to say that Rasmussen/Zogby tend to favor R by about 4 or 5 percentage points.

Your intent is correct, but I think you are exaggerating which doesn't help your cause. 538 rates Zogby as historically +0.8R biased and Rasmussen as historically +2.0R biased (using their mean-reverted bias rating). Yes, biased in the Republican's favor, but not by 4 or 5 percentage points.

 

[Strk]

Still technically shows Trump trailing.

WTF is a Zogby

Except Clinton still has the electoral votes necessary, Trump rarely breaks 200.

 

[MongGrel]

If that's your reply to a statistical tie, by all means. Tell everyone that it's in the bag, that they should stay home.

The wife and I always vote, you don't have to worry about me at any rate. We encourage people to do so.

You're still digging for anything that shows Trump ahead nationally at this point just to call the thread a lie and have not provided anything substantial.

Of course that could change again, polls always do.

 

[fskimospy]

Your intent is correct, but I think you are exaggerating which doesn't help your cause. 538 rates Zogby as historically +0.8R biased and Rasmussen as historically +2.0R biased (using their mean-reverted bias rating). Yes, biased in the Republican's favor, but not by 4 or 5 percentage points.

More importantly though Zogby is rated as a C-, meaning it is a poll of low predictive value.

 

[fskimospy]

You are correct that the USC/LA Times survey shouldn't be considered as a poll, as they don't randomly poll voters. As, you said, it is a survey of a set panel instead of a poll. So, accuracy of representing the final election isn't the intent, nor is it necessarily important in the USC/LA Times results.

They could have picked a panel that is heavily biased Democratic or heavily biased Republican. That really doesn't matter. What they are trying to do is track CHANGES over time of the same group. If the group goes from +25% Clinton to +30% Clinton (with a Democrat bias) it shows that people are switching to Clinton. Similarly, if that group goes from +25% Trump to +20% Trump (with a Republican bias), it still shows that people are switching to Clinton.

So, accuracy isn't important with that website, precision is.

So, what has the trend shown so far? Clinton's supporters have gone from 80.4% likely to vote to 82.7% likely to vote (a swing in her favor). Trump's supporters have gone from 83.0% likely to vote to 81.3% likely to vote (again a swing in Clinton's favor). If these 3000 people are swinging in Clinton's favor, then the assumption is that America is swinging in Clinton's favor. That is all that website is telling you. It doesn't tell you if one candidate is truly in the lead or not, since that isn't the way they poll people.

These are all good points. While I agree that panel surveys are best used this way I don't think the results of this survey have generally been reported on in that way. Likely the fault of the media more than the researchers, but still.

 

[dullard]

Except Clinton still has the electoral votes necessary, Trump rarely breaks 200.

Yes, the electoral votes are what really matters:

Clinton 272 (270 needed to win), Trump 154, with 112 toss-ups
http://www.realclearpolitics.com/epolls/2016/president/2016_elections_electoral_college_map.html

Clinton 356, Trump 180 (no toss-up map)
http://projects.fivethirtyeight.com/2016-election-forecast/?ex_cid=rrpromo

Clinton 316, Trump 221 (no toss-up, polls plus map)
http://projects.fivethirtyeight.com/2016-election-forecast/?ex_cid=rrpromo#plus

Clinton 358, Trump 180 (no toss-up map)
http://www.electoral-vote.com/

Clinton 279, Trump 191, 68 toss-ups
http://www.270towin.com/news/#.V7crwK1wXGg

Clinton 347, Trump 191, (scroll to the How the Analysts See It map since this is one you can adjust yourself)
http://graphics.wsj.com/elections/2016/2016-electoral-college-map-predictions/

Clinton 272, Trump 190, 76 toss-ups
http://cookpolitical.com/presidential/charts/scorecard

I could go on and on.

 

[First]

Three. Not that your choice of media would allow you to know it.

Zogby: Clinton and Trump in Statistical Tie

Huh? Your attempt to debunk that "all the polls" show Clinton ahead is to list a poll....showing Clinton ahead? And your two previous examples also had her ahead. Are you unaware of this or what?

 

[Jaskalas]

What I'm aware of is you guys have no appreciation for statistical ties.

 

[fskimospy]

What I'm aware of is you guys have no appreciation for statistical ties.

You do realize don't you that even in 'statistical ties' of that sort it is not actually a tie, right? From a probabilistic perspective the person ahead is most likely to be in the lead. More importantly though, you have yet to provide even a single poll in which Clinton doesn't lead to back up your statement that she wasn't leading in all the polls.

And all that is totally aside from the fact that cherry picking polls is a fools errand that does nothing other than misinform you. It's pointless nonsense.

 

[dank69]

You do realize don't you that even in 'statistical ties' of that sort it is not actually a tie, right? From a probabilistic perspective the person ahead is most likely to be in the lead. More importantly though, you have yet to provide even a single poll in which Clinton doesn't lead to back up your statement that she wasn't leading in all the polls.

And all that is totally aside from the fact that cherry picking polls is a fools errand that does nothing other than misinform you. It's pointless nonsense.

On top of all that, she clearly said "Most polls. All polls?" which means she wasn't even claiming it was definitely all polls. It just happens to be that it is all polls, haha.

 

[Mxylplyx]

What I'm aware of is you guys have no appreciation for statistical ties.

So just to be clear, you methodology for fooling yourself involves taking the stated margin of error in the poll results, subtracting it from the Hillary number, adding it to the Trump number, and voila!, it's actually a close race.

Yeah, good luck with that.

 

[First]

What I'm aware of is you guys have no appreciation for statistical ties.

You're wrong, and maybe someday you'll come to terms with that.