POTUS 2016 General Election Discussion Thread

[JonInMiddleGA]

We have enough trouble around these parts with getting tacos that passed a health inspection with a full-on facility. There's a relatively minor food truck thing here -- even has city backing & encouragement -- but no buzz over it to speak of.

(which struck me at least slightly odd given that it's a college town & all)

[JPhillips]

Given Trump's complete lack of campaign infrastructure, it's very likely he underperforms whatever his polling predicts. He has only one field office in all of Florida and none in North Carolina!

[Ben E Lou]

So his only positive on that list is Georgia. Here's a hint, Donald: if Georgia is close enough for you to need to put serious effort into it, you've already lost.

[JAG]

Why would you put 20+ offices in Wisconsin and 1 in Florida? Just a bizarre use of resources.

[JonInMiddleGA]

Originally Posted by Ben E Lou

So his only positive on that list is Georgia. Here's a hint, Donald: if Georgia is close enough for you to need to put serious effort into it, you've already lost.

Not sure whether that data is indicative of that or not, and I'll explain why:
Is the data only showing largely paid field offices?

If so, then it's a fair extrapolation perhaps.

But it could also be merely an indication of a larger than normal amount volunteers/interest if it's including any/all offices. In 2012 there were some outliers that weren't related to campaign situation at all, caused by both campaigns opening offices in fundraising states.

[Ben E Lou]

Originally Posted by JonInMiddleGA

Not sure whether that data is indicative of that or not, and I'll explain why:
Is the data only showing largely paid field offices?

If so, then it's a fair extrapolation perhaps.

But it could also be merely an indication of a larger than normal amount volunteers/interest if it's including any/all offices. In 2012 there were some outliers that weren't related to campaign situation at all, caused by both campaigns opening offices in fundraising states.

Ahhhh...got it.

[JPhillips]

The chart shows physical spaces operated by the candidates. My understanding is staff often comes through the state party, but the space itself is owned/operated by the campaign.

[Ryche]

Originally Posted by JAG

Why would you put 20+ offices in Wisconsin and 1 in Florida? Just a bizarre use of resources.

There doesn't seem to be a plan, at least not one that's being followed.

Inside Trump Tower: Facing grim reality - POLITICO

[JonInMiddleGA]

Originally Posted by JPhillips

The chart shows physical spaces operated by the candidates. My understanding is staff often comes through the state party, but the space itself is owned/operated by the campaign.

Thanks, I genuinely didn't know precisely what it measuring.

Honestly, still not sure there's much that's all that notable about it tbh. This is about turnout, I'm convinced of that, so if you're using space in a supposedly secure state to work that turnout angle then it's not even close to the worst idea ever. For either candidate.

[JPhillips]

The GOPers that commented said the problem is the local volunteers, the ones that know the terrain best, don't have a place to work. Those local volunteers do the pre-vote calling and canvassing, but they need a professional directing them. If there isn't a field office, there isn't a staff, and there isn't a way for volunteers to do their work.

[JonInMiddleGA]

Originally Posted by JPhillips

The GOPers that commented said the problem is the local volunteers, the ones that know the terrain best, don't have a place to work. Those local volunteers do the pre-vote calling and canvassing, but they need a professional directing them. If there isn't a field office, there isn't a staff, and there isn't a way for volunteers to do their work.

I thought the "issue" here was supposed to be that there were too many offices in the wrong states. So now it's that there aren't enough offices?

Here's the funny thing about this stuff, honestly: all the criticism ... about the guy who basically dismantled an entire clown car field.

[JPhillips]

Trump won with 13 million votes in the primary. Romney lost with almost 61 million votes. Or put another way, Trump needs more than twice as many votes as were cast for all the candidates in the GOP primary. He has to greatly increase his vote totals, and running a campaign with little infrastructure makes that much more difficult.

[Atocep]

Originally Posted by JPhillips

Trump won with 13 million votes in the primary. Romney lost with almost 61 million votes. Or put another way, Trump needs more than twice as many votes as were cast for all the candidates in the GOP primary. He has to greatly increase his vote totals, and running a campaign with little infrastructure makes that much more difficult.

It's also highly unlikely Trump wins the primary if there were only 1-2 other serious candidates. He won largely because of the huge, chaotic field the republicans had.

[SackAttack]

Originally Posted by JonInMiddleGA

I thought the "issue" here was supposed to be that there were too many offices in the wrong states. So now it's that there aren't enough offices?

Here's the funny thing about this stuff, honestly: all the criticism ... about the guy who basically dismantled an entire clown car field.

Bolded for emphasis. Nothing about his victory in the GOP primary suggests that his political instincts are any good. He parlayed his ability to build a solid plurality into a game of attrition.

If the field had been Trump, Cruz, Bush and Rubio with the other 15 or however many sitting out, I'm not sure Trump comes out on top. With 19 candidates, he was able to use his celebrity to leverage essentially free press coverage and overshadow everybody else, even the boys with the money.

It's one thing to put away Ted Cruz in a heads-up matchup when the options on the table are "I win" or "contested convention."

It's another to put away Hillary Clinton - or anybody else - in a general election when you aren't bothering to engage with retail politics. Trump's fundamentals are underwater. So are Clinton's, though, so that needn't harm him, necessarily.

But he isn't making the effort to do the things that need doing. He's making stump speeches, feeding "build the wall" and "put her in jail" red meat to the base, but that doesn't organize volunteers to drive turnout.

And having done volunteer canvassing before, I can tell you that those volunteers work for campaigns all the way down the ballot. They aren't given the option to canvass just for Trump or just for (to use a local example) Ron Johnson. You show up, you get your script, and then you go knock on doors for Trump, Johnson, whoever the local House candidate is, and the candidate for governor, if your state's gubernatorial election is in a presidential year.

But if there's no field offices and no organization for volunteers, none of that happens.

That affects the GOP far beyond just 1600 Pennsylvania Avenue. It hurts their ability to limit the down-ticket damage Trump might inflict in a loss, and it puts a ceiling on any coattails a Trump victory might bring.

[JonInMiddleGA]

Originally Posted by JPhillips

Trump won with 13 million votes in the primary. Romney lost with almost 61 million votes.

And Romney got that number in the general despite only 10m votes in the primary. (I won't quibble that Trump actually had 14m primary votes instead of 13m, irrelevant difference)

And the gen election totals aren't apples-to-apples anyway if, as I expect, the turnout is actually lower than four years ago.

And the total is even further irrelevant if, as Ipsos projects most recently, Trump is neck & neck with HRC in popular vote but still faces an electoral landslide loss.

[SackAttack]

Originally Posted by JonInMiddleGA

And the total is even further irrelevant if, as Ipsos projects most recently, Trump is neck & neck with HRC in popular vote but still faces an electoral landslide loss.

I think I need a definition of terms for what you consider an "electoral landslide," because otherwise I'm having a difficult time seeing how Trump loses in a landslide with a roughly split popular vote.

[EagleFan]

Trump will succeed in his goal of getting hillary elected.

[OldGiants]

Richmond Times-Dispatch endorsed Gary Johnson today.

[tarcone]

Originally Posted by EagleFan

Trump will succeed in his goal of getting hillary elected.

This

[molson]

Trump's national poll numbers don't seem to line up with his state poll numbers. He's very close, and even sometimes leading, in national polls, but he is way behind in battleground states, and some strong red states are much closer than you'd think they'd be. I don't know the reason for the disparity, but places like 538 are putting much more stock in the state polls.

[digamma]

Haven't really seen this mentioned but could you imagine if this was the Clinton Foundation?

Trump pays IRS a penalty for his foundation violating rules with gift to aid Florida attorney general - The Washington Post

Short version: The Trump Foundation gave the Florida Attorney General a $25,000 campaign contribution during the time she was deciding whether to pursue charges related to Trump University.

[cuervo72]

Originally Posted by molson

Trump's national poll numbers don't seem to line up with his state poll numbers. He's very close, and even sometimes leading, in national polls, but he is way behind in battleground states, and some strong red states are much closer than you'd think they'd be. I don't know the reason for the disparity, but places like 538 are putting much more stock in the state polls.

Media trying to make it look like it's a horse race?

[nol]

Originally Posted by digamma

Haven't really seen this mentioned but could you imagine if this was the Clinton Foundation?

Trump pays IRS a penalty for his foundation violating rules with gift to aid Florida attorney general - The Washington Post

Short version: The Trump Foundation gave the Florida Attorney General a $25,000 campaign contribution during the time she was deciding whether to pursue charges related to Trump University.

Yeah, stuff like that just underscores how New York-based media just assumes that everyone already knows Trump is somewhat of a joke with his sketchy business dealings without realizing that millions of people in the heartland have little knowledge of him beyond watching The Apprentice.

[JonInMiddleGA]

Originally Posted by SackAttack

I think I need a definition of terms for what you consider an "electoral landslide," because otherwise I'm having a difficult time seeing how Trump loses in a landslide with a roughly split popular vote.

Just to clarify, that wasn't my phrasing, it was their's. ("they" being Reuters or whomever wrote the piece I saw)

Basically, it's along the lines of what Molson said a few posts after mine. Here's a quote from an early version of the Reuters article that made the rounds (I don't have the one that used the word "landslide" but I'd swear it having been used in the version I saw)

As of last Friday, the separate Reuters/Ipsos States of the Nation polling project estimated Clinton was on track to win the Electoral College, by about 332 votes to 206.

Oddly enough, that's the exact finishing electoral total of Obama - Romney. A decisive win but maybe not quite landslide territory (I figure a 2:1 win like '08 Obama qualifies for that)

My guess is that the subsequent one I saw was anticipating better numbers in the separate States of the Nation poll than actually turned up in the latest update. They actually had her down in the 270s but with him in the 180s with quite a few thrown into undecided/too close to call territory.

[JonInMiddleGA]

And the thing to keep in mind about Ipsos is that it differs from a lot of polls in that it uses "likely voter" weighting and a number of projections about turnout within specific demographics.

[SackAttack]

interesting note there, Jon? Since 1972, the only Presidents-elect to win fewer electoral votes than Barack Obama did are George W. Bush (both times) and Jimmy Carter. If either of Obama's victories qualify as landslides, so does nearly every Presidential election since the lower 48 completed their respective marches to statehood.

1948 and 1968 saw 3rd party candidates carry a few states in each election, which differentiates them from 1960 and from the aforementioned '00, '04 and '76.

So by that metric, 20 of the last 26 Presidential elections have been electoral landslides, then.

[JonInMiddleGA]

Originally Posted by SackAttack

interesting note there, Jon? Since 1972, the only Presidents-elect to win fewer electoral votes than Barack Obama did are George W. Bush (both times) and Jimmy Carter. If either of Obama's victories qualify as landslides, so does nearly every Presidential election since the lower 48 completed their respective marches to statehood.

If you beat somebody 2:1, I don't have a huge problem with the term "landslide". Whether that's a presidential race or 5th class hall monitor. It's maybe more of a very solid win but landslide doesn't feel like a huge stretch. My ass got kicked with something in the upper 70s and I'd certainly say that was a landslide.

Wiki cites "Presidential Elections" American Presidency Project. University of California, Santa Barbara in referring to Bush 88, Reagan 84, Reagan 80 and Nixon 72 as landslides.

That seems to set their bar at roughly 80% of the electoral votes (Bush 88 was 79.18%).

I'd not have a problem with referring to any in that time frame that way with the exception Bush2's wins, Obies 2nd, and Carter with the term. YMMV {shrug}

[SackAttack]

Oh, certainly. I'm just commenting that if that's the metric - and I'm not objecting to that, per se - then despite all the talk about swing states and polarization, Presidential elections just really haven't been all that close in living memory. The few exceptions have been extraordinary circumstances - Truman running for his first 'own' term after the 1-2 whammy of the Depression, World War II, and four straight terms for Roosevelt (if ever election 'fatigue' was a thing, that could have been the year); Ford running for his first 'own' term after pardoning Nixon; third-party candidates with decent showings; and Dubya, for whatever reason, was unable to put away either Gore or Kerry as easily as Barack Obama put away McCain and Romney.

[JonInMiddleGA]

Originally Posted by SackAttack

Oh, certainly. I'm just commenting that if that's the metric - and I'm not objecting to that, per se - then despite all the talk about swing states and polarization, Presidential elections just really haven't been all that close in living memory.

That's actually "a thing" or so I've read in the past couple of years in particular. that basically "swing states" really just frequently follow a larger national trend most of the time and they really aren't causing those trends, merely following them.

[Ben E Lou]

CNN/ORC new poll: Trump 45, HRC 43

Presidential poll: Donald Trump pulls ahead of Hillary Clinton - CNN.com

[Marc Vaughan]

Originally Posted by Ben E Lou

CNN/ORC new poll: Trump 45, HRC 43

Presidential poll: Donald Trump pulls ahead of Hillary Clinton - CNN.com

Largely meaningless poll if you look behind the curtain (disclaimer - the vast majority of reported polls are meaningless) .... they interviewed 1,000 voters so a whopping 20 per state roughly.

Then compound that with the fact that 60% of their interviews were done on landlines which biases the demographics hugely towards older more conservative voters and you can see why unsurprisingly its skewed towards trump looking strong.

(if you look behind the curtain on some high Clinton ones then expect to see that their surveys are conducted nearly exclusively on mobile phones etc. ... its really easy for the media to get the results they want in a poll with a little manipulation of method)

[QuikSand]

Originally Posted by Marc Vaughan

Largely meaningless poll if you look behind the curtain (disclaimer - the vast majority of reported polls are meaningless) .... they interviewed 1,000 voters so a whopping 20 per state roughly.

So, how many well-selected people would you suggest it takes to come up with a valid survey? 300 million?

[Marc Vaughan]

Originally Posted by QuikSand

So, how many well-selected people would you suggest it takes to come up with a valid survey? 300 million?

From a statistical perspective a damn sight more than 1,000 - 3m is 1% of 300m and you're expecting to have a reliable survey from 0.0003% (approx) of that population with obvious built in bias from the selection method used to choose the people interviewed.

To be meaningful a survey would require sufficient people to allow for potentially 'lucky choices' for one side or another and a selection methodology which was unbiased ... this is something which would be incredibly difficult to do, not least because the media has no inclination to even try - it'd give more boring results and make for less interesting stories.

Quite frankly I think the entire concept of polls for elections is purely media razamatazz and anti-democratic - they're used to rabble rouse people into voting through fear that the 'other side wins' or lull them into complacency because their side has it in the bag .... they serve no useful purpose.

(if I'm incorrect and they serve a purpose other than providing a false bias to what is meant to be a democratic election based on principles and concrete political standpoints - then please enlighten me)

[QuikSand]

I think there's a major innumeracy issue with the concept of polling and surveys. It doesn't pass the smell test for people not well versed in math that surveying a seemingly small number like 1,000 can be a reasonable representation of a huge number like 300 million. And then you hear things like "well, nobody ever called me for a poll" and then you have what passes in modern discourse as a legitimate attack on the very concept.

The fact is that a well-selected set of 1,000 from a population of any size at all is a perfectly good way to get a good representative number -- essentially with a +/- 3% degree of confidence. Go to about 1,600 to get your number down to 2.5%, which is a fairly common standard. Sorry, math-averse people, that's just how it works, despite your insistence that it sounds wrong.


Now... if your quarrel is with the method (what kind of phone lines, etc), then you'are invalidating part of my question... "well-selected." And that's fine. If your method is biased, your results will be skewed and of less value. Agreed there.

Just lay off the "1,000 voters" as a centerpiece argument. Yes, that's only 20-ish per state. I don't think anyone is suggesting that the 23 people from Maryland in the survey can tell us anything about Maryland -- it's a national poll, and not billed otherwise.

"The vast majority of reported polls are meaningless." No. Some are flawed in method, and sometimes there are undercurrents that don't get revealed until election day, sure. But this isn't witchcraft or voodoo. It's math.

[TCY Junkie]

Originally Posted by QuikSand

The fact is that a well-selected set of 1,000 from a population of any size at all is a perfectly good way to get a good representative number -- essentially with a +/- 3% degree of confidence.

I remember the day in college. .... that makes feel old....... when the teacher showed us how small number needed polled to get consistent numbers.

[PilotMan]

It's all about the sample population. Without a representative sample the numbers aren't worth anything at all. It's companies that pull bad data, then try and generalize it that give it all a bad name.

[albionmoonlight]

538's polls-plus model now has Trump at a 1/3 chance to win.

By any metric, he's had a very good couple of weeks.

[Ben E Lou]

Originally Posted by albionmoonlight

538's polls-plus model now has Trump at a 1/3 chance to win.

Worked out ok for this guy.


One Chance in Three - YouTube

[Mizzou B-ball fan]

Originally Posted by albionmoonlight

538's polls-plus model now has Trump at a 1/3 chance to win.

By any metric, he's had a very good couple of weeks.

He'll win if he can stay out of his own way. That's an AWFULLY big if.

[panerd]

Originally Posted by lighthousekeeper

I can see why people from the midwest might start thinking this way, but this election will be a strong reminder of how little the midwest voting bloc matters.

Michigan, Wisconsin, Iowa, Ohio?

[Marc Vaughan]

Originally Posted by QuikSand

Just lay off the "1,000 voters" as a centerpiece argument. Yes, that's only 20-ish per state. I don't think anyone is suggesting that the 23 people from Maryland in the survey can tell us anything about Maryland -- it's a national poll, and not billed otherwise.

"The vast majority of reported polls are meaningless." No. Some are flawed in method, and sometimes there are undercurrents that don't get revealed until election day, sure. But this isn't witchcraft or voodoo. It's math.

Fair enough but the polls aren't then used for a simple 'yes/no' - they're then sub-divided down into subsets which are then further analysed and presumed to have the same +/- 3% accuracy ... for instance in their 'voters of type x are 60% likely to vote for Trump etc.'.

If you consider the 1,000 people original sample then when they say college educated white males aged above 60 then the actual sample set they're taking that discussion point from might be as few as 25 people which IS statistically insignificant for that purpose.

(presuming 50% chance of college education, 10% of population aged above 60 and a 50% white population - all stats purely guessed from my rectal area)

I agree if you use the polls sensible AND if the people interviewed are carefully chosen to try and remove obvious biases (which doesn't appear to be the case in any of the ones I've looked at) then its possible they can be good predictors ... presuming people answering them tell the truth, which they often won't for fear of feeling stupid or appearing racist, whatever.

I do however stand by my stance that they serve no worthwhile purpose apart really when you think about it ...

[Ajaxab]

I wish jbmagic was here to solve our math problems.

[JonInMiddleGA]

Originally Posted by Marc Vaughan

Then compound that with the fact that 60% of their interviews were done on landlines which biases the demographics hugely towards older more conservative voters and you can see why unsurprisingly its skewed towards trump looking strong.

You can't kick the data on that basis alone, not without looking at the likely voter component on phone service data.

You could argue just fine that it isn't a representative cross sample of the population but that isn't the same as likely voters.

[molson]

Originally Posted by TCY Junkie

I remember the day in college. .... that makes feel old....... when the teacher showed us how small number needed polled to get consistent numbers.

I've still never been able to wrap my mind around that - I've just learned to accept the science of it.

But, there's definitely something off with some of these polls. They vary more than the margin of error suggests they should, even when released the same day and covering the same time period. But, I think 538 has a proven track record of taking into account all of the polls, which creates a much larger sample size and the ability to weed out bad polls. Even with that though, you have situations like Sanders winning Michigan despite Clinton being up huge in the polls.

I also think a lot of people are in denial about Trump's chances. I buy the 1/3rd chance (both 538 and the PredictIt lines have it around there), and that's a pretty decent chance.

[Thomkal]

to add to that Van Sustern is leaving Fox, they have also settled the Gretchen Carlson lawsuit for 20 mil. And they released a public apology:

Fox News Will Pay Gretchen Carlson $20 Million To Settle Sexual Harassment Suit : The Two-Way : NPR

[flere-imsaho]

Originally Posted by lighthousekeeper

Well I have some bad news for you buddy.

I think you just saved us from another 3 pages of math nerds arguing. CHEER!

[JPhillips]

So now Russians, probably connected to Putin, are leaking hacked DCCC emails directly to Trump's son-in-law's newspaper.

[QuikSand]

Originally Posted by molson

I also think a lot of people are in denial about Trump's chances. I buy the 1/3rd chance (both 538 and the PredictIt lines have it around there), and that's a pretty decent chance.

I sold my shares of DEMSLOSE at 31c there today. Bought on 7/14 at 24c. Did okay with that, as well as a move on some of the "close race" bands of electoral outcomes. Not completely sold that it cannot tighten up from here, but I got what I wanted - just some obvious regression to the mean after everyone was writing the race completely off a month or so ago.

[panerd]

Originally Posted by JPhillips

So now Russians, probably connected to Putin, are leaking hacked DCCC emails directly to Trump's son-in-law's newspaper.

I have Trump ranked 4th in preference of the potential presidential candidates so don't count me as a fan of Trump. But maybe when President Hillary Clinton takes office and the CIA comes to her about another "regime change" she will think twice about what she does... Who am I kidding?

[flere-imsaho]

What are you talking about, panerd?