Presidential Approval Polls 2017

User avatar
Addie
Posts: 21350
Joined: Mon Jun 15, 2009 6:22 am
Location: downstairs

Re: Presidential Approval Polls 2017

Post by Addie » Sat Oct 21, 2017 1:13 pm

Newsweek
Donald Trump's Approval Rating Plunges As Popularity Nears All-Time Low in Latest Polls ...

Notably, Trump's approval rating neared his all-time low in the Gallup poll. The latest figure in Gallup's tracking survey, released Thursday, pegged Trump's approval at just 35 percent, down from 38 percent at this point last week. He's just one percentage point higher than his lowest rating ever of 34 percent. To make matters worse for the commander-in-chief, Trump's disapproval was nearing his all-time high, as well. It stood at 60 percent, just one percentage point off from his all-time high of 61 percent in early September. The Gallup poll surveys 1,500 U.S. adults and has a margin of error of plus or minus 3 percentage points.

Trump fares a bit better in the tracker from data-focused website FiveThirtyEight. The site aggregates public polls and comes up with an average approval rating that accounts for each poll's quality, recency, sample size and any partisan leanings. It had Trump at 37.8 percent approval Saturday, 1.2 percentage points higher than his all-time low of 36.6 percent in the FiveThirtyEight tracker. Trump's disapproval stood at 56 percent Saturday, according to FiveThirtyEight.

Still, the former reality TV star is the most unpopular president in recent history. In the history of modern polling, no other president had as low an approval rating on day 274 of his presidency, according to FiveThirtyEight. The closest was former President Gerald Ford, who stood at 38.4 percent at the same point in his first term. It's worth noting that Ford's popularity plummeted after he pardoned his predecessor Richard Nixon, who resigned in disgrace amid the Watergate scandal. Ford's decision was unpopular at the time but has largely been accepted as a correct and courageous decision years later. And at day 274, Trump's predecessor, former President Barack Obama, had an average approval rating of 53 percent.

Trump has never really come close to even hitting 50 percent in the FiveThirtyEight tracker. (He got as high as 47.8 percent shortly after inauguration but quickly fell a few percentage points.)
¡Qué vergüenza!

User avatar
Addie
Posts: 21350
Joined: Mon Jun 15, 2009 6:22 am
Location: downstairs

Re: Presidential Approval Polls 2017

Post by Addie » Sat Oct 21, 2017 6:23 pm

Gallup
Trump Job Approval Slips to 36.9% in His Third Quarter

WASHINGTON, D.C. -- Throughout Donald Trump's third quarter as president, from July 20 through Oct. 19, an average of 36.9% of U.S. adults approved of the job he was doing. That represented a nearly two-percentage-point decline from his second quarter, after a similar decline from his first to second quarter.



Trump's job approval rating was below 40% in all but one of the Gallup Daily tracking three-day rolling averages reported during the third quarter.

This includes Trump's personal low 34% job approval rating, measured multiple times in mid- and late August. Those ratings came after Trump made provocative statements about North Korea on Twitter and offered late and equivocal condemnation of white supremacists involved in violent protests in Charlottesville, Virginia.

His latest Gallup Daily tracking job approval rating, based on Oct. 17-19 interviewing, is 35%.

Trump's Third-Quarter Average Lower Than Any Other President's

Trump has consistently had much lower approval ratings than any elected president in his first year. His third-quarter ratings continued that trend, as they were more than 10 percentage points lower than the previous low third-quarter rating, 47.7% for Bill Clinton in 1993. All other elected presidents had approval ratings above 50% in their third quarters, with five of these at 60% or higher.
¡Qué vergüenza!

User avatar
Slartibartfast
Posts: 6849
Joined: Mon Aug 10, 2009 2:52 pm

Re: Presidential Approval Polls 2017

Post by Slartibartfast » Mon Oct 23, 2017 3:02 pm

Please note that this is a professional rant, not a personal one. It is, therefore, in no way directed at anyone here, especially not Addie, who has done a wonderful job documenting the crappy state of data punditry these days. I'm trying to educate you to a manner in which you are being, either intentionally or negligently, misled.

--------------------------------------------------------------------------------------------

Whenever you see the phrase "at this point in his presidency" or something similar in an article about polling you should be telling yourself that what is being discussed is not a lie, nor is it a damned lie, it is a statistic - a piece of information that is being used to mislead you in some way or other. I'd be hard pressed to find an article on this thread for which this description isn't apt.

Using statistics to build or support narratives is exactly what Twain was talking about. Any scientific use of data includes three steps: collecting or generating the data and verifying its quality, mining it for information that has been determined to be useful, and, finally, analyzing it to produce intelligence that has value -- an accurate scientific prediction of some sort.

Statistics are a form of information -- data that has been processed into groupings that make it easier to work with -- information that, on it's own, can never be part of a scientific argument. When you use statistics in the manner that is nearly ubiquitous these days, you are taking information that has been stripped of its context and thus any kind of scientific rigor. Which is fine for an informal argument, but becomes problematic when it is implied to be under the aegis of science ("3.1% margin of error" and "95% confidence interval" would be examples of this).

If you are going to consume this kind of information, you should look at it yourselves rather than trusting what ANYONE says about it. Go to 538's approval tracking page, scroll down to "How Trump compares to past presidents", and select "Net approval" and a time span of 8 years. I'm not going to pull out statistics myself, but suffice it to say that there are plenty of other presidents (but not Eisenhower, he was apparently a god) with approval ratings lower than Trump's -- even for sustained periods of time (we know where the buck stops, Harry).

As an aside, there are ways to structure and use polls so that they provide more insight (both rigorously and informally), such as having the answer to approval correlated with the person's level of support of the president (same range: strongly support, support, neutral, oppose, strongly oppose). Then the information will come through 5 by 5*.

* Which is both a Buffy reference and an analogy to the original meaning of rating a radio signal for strength and clarity.
"Strategy without tactics is the slowest route to victory. Tactics without strategy is the noise before defeat."
---Sun Tzu (quoting Thomas Jefferson)
nam-myoho-renge-kyo---Thomas Jefferson (quoting Slartibartfast)

User avatar
Foggy
Posts: 24510
Joined: Tue Jan 20, 2009 12:00 pm
Location: Fogbow HQ (Rawly NC)
Occupation: Dick Tater

Re: Presidential Approval Polls 2017

Post by Foggy » Mon Oct 23, 2017 7:26 pm

I took a course in statistics at the University of Vermont before I transferred to -- and ultimately graduated from -- Georgetown University. In Vermont, we studied the average length of a fish in Lake Champlain, and the average amount of compactification of the soil on farms due to the use of tractors and other machinery, things like that. These were practical statistics, in the sense that they measured and analyzed real things about the real world in which we lived and worked. We studied the theory and the mathematics behind statistics and the reasons you could say things like "3.1% margin of error" and "95% confidence interval".

So I did, to a degree, learn to trust statistics. If the data is gathered validly, based on accurate measurement and lack of bias, the statistics produced by that data can be trusted. A statistic, in my world, is not "a piece of information that is being used to mislead you in some way or other". In my world, people literally bet the farm on the accuracy of statistics, and, in most cases, people won those bets.

Maybe the statistics generated during the 2016 election were not based on data that was measured accurately. It's harder to accurately measure the intent of a voter than it is to measure the length of a fish, I understand that much. Maybe there was some bias involved. I probably had a bias myself, in that I kept telling people that America had not collectively gone batshit insane and we would never elect Donald J. Trump president. I wanted to believe that so strongly, and I did believe that so strongly, that I thought Michael Moore was full of shit.

Nevertheless, the mathematics behind the study of statistics is valid, IMHO. Statistics, for all its weaknesses, is not about misleading the public. There's obviously some room for improvement in the measuring and analysis of the data that goes into statistical predictions. Political statistics are more iffy than practical statistics. But ... even after the failure of people like Sam Wang at the Princeton Election Consortium, who predicted a 98% chance that Ms. Clinton would win the election ... I don't think it's time to throw statistics on the trash heap of history.

In 2012, Nate Silver predicted the electoral outcome in every one of the 50 states. He and Sam Wang had a very unfortunate experience in 2016, and I'm certain they'll make adjustments with a goal of improvement in 2018 and 2020.

I spent 16 hours last week volunteering at the Democratic Party booth at the NC State Fair, four shifts of four hours each. That's something I do every year, and I really enjoy it. Of course, I spent a lot of time commiserating with my fellow Democrats about the 2016 election. But I decided that if Ms. Clinton had won, we'd be in the same place we were in with President Obama, where the RWNJs would obstruct everything she did or said, and they'd be in a state of perpetual outrage for the entirety of her administration. Besides, people have to learn to vote one of these days, or we're going to be in real trouble. Maybe, in the long run, Trump's election was a kick in the ass that we needed to get.

If so, maybe Sam Wang and Nate Silver needed a kick in the ass too also. I think they'll try extra hard for accuracy in the coming years.
If dogs run free, why not we?

User avatar
RVInit
Posts: 4177
Joined: Sat Mar 05, 2016 4:31 pm

Re: Presidential Approval Polls 2017

Post by RVInit » Mon Oct 23, 2017 8:11 pm

I wonder how much of 2016 "bad" predictions had to do with people outright lying. My younger sister lamented, totally unprovoked, that she had no idea who she would vote for, denying without even being asked that she would ever vote for Trump. This was the day before election day. The following day she was bragging and rubbing everyone's face in the fact that she voted for Trump and has been a nonstop Trump fanatic ever since. I have no doubt that she knew damn well she was going to vote for Trump and wasn't willing to admit it until after he won. I highly suspect her behavior of lying about her plans to vote for Trump had to do with being truly embarrassed to admit it before hand, and "see, lots of other people voted for him too" after it became obvious that he won.
"I know that human being and fish can coexist peacefully"
--- George W Bush

User avatar
bob
Posts: 21990
Joined: Sat Dec 05, 2009 12:22 pm

Re: Presidential Approval Polls 2017

Post by bob » Tue Oct 24, 2017 2:18 am

Foggy wrote:Maybe the statistics generated during the 2016 election were not based on data that was measured accurately. It's harder to accurately measure the intent of a voter than it is to measure the length of a fish, I understand that much.
:yeah:

Too also: "At this point into his presidency" is technically correct, as we have data about the president's popularity on his 277th day in office, just as we have data for Obama's 277th day in office, etc.

But it is ultimately a misleading comparison because presidencies are not created equal: the recession drew Obama's numbers down, 9/11 inflated GWB's numbers, etc.
Imagex4 Imagex2 Imagex2 Imagex2

User avatar
Slartibartfast
Posts: 6849
Joined: Mon Aug 10, 2009 2:52 pm

Re: Presidential Approval Polls 2017

Post by Slartibartfast » Tue Oct 24, 2017 1:28 pm

Okay, let me make it clear up front that this is a flat-out appeal to authority. Mine. And I have the expertise to back it up.
Foggy wrote:
Mon Oct 23, 2017 7:26 pm
I took a course in statistics at the University of Vermont before I transferred to -- and ultimately graduated from -- Georgetown University.

I'm probably qualified to teach the class you had. I think I just found a co-author for the book I want to write on predictive analytics -- but until I get it written I don't have a text to refer to. What I'm saying is my professional opinion and is based on my own reasoning. I don't claim anyone else subscribes to it, but I can defend these ideas on their merits.

In Vermont, we studied the average length of a fish in Lake Champlain,

My advisor, Harold Layton, once wrote a paper on counter-current exchange of oxygen in the gills of fish. The moral of the story is that countercurrent exchange only gets you a small advantage, but you would have to swim a lot faster to get the same effect. It happens in penguins' legs and your kidneys right now (at least if they are functioning properly). This has no relevance but I found out my advisor has Alzheimers last winter (he doesn't know who I am anymore), so if you want another gratuitous story about Harold and his research, ask me about the penguins.

One would assume his work applies to fish in Lake Champlain, whether they are of typical or atypical length.


and the average amount of compactification of the soil on farms due to the use of tractors and other machinery, things like that.

Yep. Good solid stats class -- puts hair on your chest. Nothing wrong with that.

These were practical statistics, in the sense that they measured and analyzed real things about the real world in which we lived and worked.

There's a philosophical issue here, but it's nuanced and it's not really on point, so I'll just say I know what you mean.

We studied the theory and the mathematics behind statistics and the reasons you could say things like "3.1% margin of error" and "95% confidence interval".

Which is where things get a little dicey with what you learned. You were taught a frequentist view of statistics. Pretty much everyone was. It is crap (technical term and expert opinion). I subscribe to the Bayesian model, which has a towel and is froody. Mr. Gneiss and I were teaching a master's class in this last fall, but we were scolded and shamed and run off. I would point out that figuring out how to use Bayesian inference to create a metric to drive an evolutionary analysis of biochemical pathway models to help develop better cancer drugs is pretty much my job right now. The Bayesian inference metric part is something I pulled out of my ass one afternoon trying to impress a potential colleague. Then I figured out how it will work. Now I just have to implement it.

The point is, this is exactly the shit I think I'm good at -- like better than anyone else -- and I'm working to demonstrate it. Whether my ideas succeed or fail, they are cutting edge (bleeding edge, really) and mathematically and scientifically sound. This is where you have to decide whether to bet on my expertise or what you learned from your professor at Vermont because I'm all in.
:towel:


So I did, to a degree, learn to trust statistics. If the data is gathered validly, based on accurate measurement and lack of bias, the statistics produced by that data can be trusted.

Which is true as far as it goes, but is part of a much bigger picture. I call this bigger picture "predictive analytics". Essentially, you need a rigorous, scientific framework for the entire process (from the user that order the job to the user that consumes the intelligence -- or from the patient to the patient if you're talking about personalized precision medicine).

There is nothing inherently wrong with the statistics, and certainly nothing that you were taught in skool is violated by these articles that Addy posts. It is data that can be "trusted" (at least in the sense you meant here, other trust issues addressed below). It is the act of taking the statistic, removing it from its context, and using it in data punditry that breaks the scientific chain of custody, as it were.


A statistic, in my world, is not "a piece of information that is being used to mislead you in some way or other".

Nor in mine. As I said, statistics are types of information and I use information to create valuable intelligence. However,
that is exactly how statistics are being used in all of these articles. Which was, in my opinion, exactly what Sam Clemens was nattering on about.


In my world, people literally bet the farm on the accuracy of statistics, and, in most cases, people won those bets.

I know, I have many friends who have gone the actuarial route. Good jobs, good money, low stress. They are the reason why the house always wins. I think the mortgage-based security debacle was based on the management not understanding what their bean counters were telling them (and being drastically overconfident about sketchy financial instruments as a result). I don't do that, I do predictive analytics and I need to be able to show -- literally -- that people should bet their lives on them. That's a much higher bar and not one that statistical models tend to do well with. There are other techniques that can work just fine though.

I'm not saying you're wrong, I'm just looking from a different perspective and measuring by a different standard.


Maybe the statistics generated during the 2016 election were not based on data that was measured accurately.

It's more complicated than that. What should have been a known unknown was treated as an unknown unknown and ignored.
As a result, there was a bias that wasn't being controlled for. The study that everyone was bitching about (LA Times/USC?) is, by far, the most important data from the election cycle, in my opinion. But that's another issue entirely. The upshot is that, given the failure of predictive modeling last election cycle, the idea that anyone should trust polling information until the people proffering it have explained why it is not susceptible to or how they have addressed the issues that caused the problems in 2016 is absurd.

I gave a talk on predictive analytics the day after the election. I had to rewrite the whole damned thing to address what happened. It really sucked, but it was the only honest thing I could do. I haven't seen anyone doing data-driven punditry or analytics show the same level of integrity (or, frankly, insight). The possibility of a systemic bias in the polling was severely underestimated or ignored. In my book, everyone who was guilty of that failure needs to own it. If you look back at the Polling 2016 thread you can see that I did just that with my own pre-election comments. It was great to see Rikker back from trolling over at Fox, but I've got no link for him -- just search my posts in that thread and you can find it no problem.


It's harder to accurately measure the intent of a voter than it is to measure the length of a fish, I understand that much.

Um... er... sort of?

I wouldn't exactly say it was "easy", but the Obama campaign and Cambridge Analytics both did something that could legitimately be called "measuring the intent of a voter very accurately". The Russians may have been playing this game as well. Clinton had it too (also), but she just used it to play video games, more or less.


Maybe there was some bias involved.

It is assumed that all polls have some sort of bias. Which is perfectly reasonable and no problem at all. It was further assumed that all of those biases were essentially distributed randomly with some sort of balanced distribution, which turned out to be...
not so good. Most likely what happened was that Trump voters were being systematically under-sampled (possibly exacerbated by people lying -- call this [appropriately] "the Trump effect") and this problem was amplified by the likely voter models being used. These are all problems that I think can be addressed and mitigated, if not eliminated, but I haven't seen anyone with the guts (and the knowhow) to say "we've got to do this before we can trust the numbers again".


I probably had a bias myself, in that I kept telling people that America had not collectively gone batshit insane and we would never elect Donald J. Trump president.

That's where you were wrong back in November 2016 and your Doomseers are wrong now. America has not gone collectively batshit insane, this election was the culmination of at least a half a century of escalating partisanship. And I cannot imagine a worse candidate for the time than Hillary. Except maybe someone who had literally shot someone in Times Square on nationable TV. Fair or not, there were a great many people who wouldn't vote for Hillary because she represented everything that was wrong with the Democratic Party to them. Not to mention that her political playbook hasn't been current since last millennium.

I wanted to believe that so strongly, and I did believe that so strongly, that I thought Michael Moore was full of shit.

I think Michael Moore is pretty much always 95% full of shit. He went to high school with my ex-wife's sister, by the way. She probably sold him drugs at some point. But the solution to your problem is simple: when you want to believe in something then you need to try with all your might to prove it false. For instance, if you want to believe that Michael Moore is full of shit, then try to prove that he knows what he is talking about and provides useful insight and accurate information. If you do so, you'll find out whether or not he is trustworthy and to what degree.

Nevertheless, the mathematics behind the study of statistics is valid, IMHO.

The difference between you and me is that it's not a matter of opinion for me. I can prove it. But the frequentist math only goes so far and once you violate scientific rigor it's gone.

Statistics, for all its weaknesses, is not about misleading the public.

Yeah, it's not bad, it's just drawn that way. Statistics, in most of it's popular uses, is pretty much entirely about misleading the public. People need to become educated consumers of statistics or ignore them in my opinion.

There's obviously some room for improvement in the measuring and analysis of the data that goes into statistical predictions.

We don't need to worry about redecorating until we've dealt with the tree that fell on the house. Too, also, there are,
as I mentioned, scientific methodologies that don't run afoul of the issues that sunk 2016 election predictions.


Political statistics are more iffy than practical statistics.

They both are what they are, scientifically speaking. It's what you do with them that is at issue.

But ... even after the failure of people like Sam Wang at the Princeton Election Consortium, who predicted a 98% chance that Ms. Clinton would win the election ... I don't think it's time to throw statistics on the trash heap of history.

In my opinion, it's long past time to look at how we use statistics informally and understand the difference between that and when they're used scientifically. Sam Wang's failure was not predicting Hillary had a 98% chance of winning the election -- that's what the data said according to the scientific best practices that he favors. He was tripped up by his (and everyone else's) egregious underestimation of the chances that the basic assumption of random bias in the polls was true. His goal (as stated clearly and repeatedly on his website) was to tell people (on both sides) where to effectively spend their money. And he gave Democrats pretty much the worst advice possible: throw your money into unwinnable Senate races (his targeting on which House races were close was probably off too) instead of into the Presidential race in Michigan, Wisconsin, and Pennsylvania where action might have made a difference in the outcome.

In 2012, Nate Silver predicted the electoral outcome in every one of the 50 states.

Which is to say about a half-dozen races that were actually in play (anyone could have predicted the outcomes in New York and Texas). Which is not to downplay his achievement, but we really need to look at more robust metrics of the performance of Sam and Nate over all of the election cycles they've covered.

He and Sam Wang had a very unfortunate experience in 2016, and I'm certain they'll make adjustments with a goal of improvement in 2018 and 2020.

This case is being heard in the court of science and in that court Sam and Nate had a very fortunate experience: they found out they were wrong. As a result, it should be assumed that their approach is flawed until they can explain why it isn't and demonstrate it to scientific standards. Otherwise they are little more than data pundits who are, in my opinion, using statistics in exactly the manner that Mark Twain warned about.

I spent 16 hours last week volunteering at the Democratic Party booth at the NC State Fair, four shifts of four hours each.

I love the State Fair. What (very tasty) new deep-fried abomination was there this year? I last went with my fambly a couple of years ago -- bigly fun!

That's something I do every year, and I really enjoy it.

And if something deep-fried happens to get et in the process, who's to know?

Of course, I spent a lot of time commiserating with my fellow Democrats about the 2016 election.

Nothing wrong with some gallows humor while you walk off a metaphorical kick in the nuts, but it's been almost a year --
the Democratic party should have long since come up with a plan to fix this problem, that, let's be honest, they have enabled over decades. We should be seeing bold action to bring fundamental change right now. Instead, we're seeing nothing, like we've seen before.

This isn't a failure of yours or of your fellow Democrats, it is a failure of your leadership. Until the rank and file realize that and demand accountability and competence in their leaders and the people back it up with their votes I think the party is doomed to remain moribund.


But I decided that if Ms. Clinton had won, we'd be in the same place we were in with President Obama, where the RWNJs would obstruct everything she did or said, and they'd be in a state of perpetual outrage for the entirety of her administration.

I think it would have been much, much worse. So far as we know, President Obama didn't have many skeletons in his closet and Republican attempts at drumming up outrage for things like fast and furious or Benghazi didn't really make sense. There were real reasons to criticize President Obama (claiming the right to assassinate US citizens overseas, for instance), but no one really cared about them, certainly not the Republicans. Hillary, on the other hand, has skeletons both legitimate and perceived (and the perception of the voters is the reality that one must deal with). One would think, after decades of unprecedented scrutiny of her life, she would understand the need for being above the appearance of impropriety. Many people, right or wrong, are unwilling to give Ms. Clinton the benefit of the doubt. When she gives them reason to doubt, such as by using a private server or deleting emails, the results are predictable. They impeached her husband for lying about a blowjob, how much worse would her impeachment (and you know it would have happened by now)
have been?

The other inescapable conclusion is that, as a result, we've been handed the biggest opportunity for real, transformative change since the Civil War, or maybe even the Revolution. Just like the Cold War finally came to an end, it is time that the age of partisan politics did so as well. It's time for all of the adults in the room to act like it and stop being bullied by the children of both sides who do not understand that we are all in a lifeboat together and are throwing a tantrum.



Besides, people have to learn to vote one of these days, or we're going to be in real trouble.

Actually, I think a lot of the problem is the choices they've been presented with. No one is articulating a vision for the changes they would make, what the results would be, and how they could get that done. In election after election we've been given principals not principles. As we are seeing with President Trump's obsession with destroying President Obama's legacy, the influence of principals often doesn't last longer than their term of office. Policy based on principle, like the New Deal or the Great Society, have fared better. I think the state of Obamacare is a stark reminder that, in the end, any cult of personality must fade. Principles, on the other hand, can still inspire people centuries later.

We hold these truths to be self-evident...


Maybe, in the long run, Trump's election was a kick in the ass that we needed to get.

We've lost 75% of the world's flying insects. It's time to stop thinking of Trump or the Republicans or the Alt-Right or any other group as the enemy and realize that the real enemy is our utter lack of self-control. It's time for the rational adults to take control of the dialogue and discuss our problems honestly and openly with the understanding that, if we can't find a way to work together, even if none of us is doing exactly what we want, then this Great Experiment in Democracy that we call the United States is probably about to fail and will likely take our civilization and possibly our species with it.

Even if it is just raging against the dying of the light, my choice is to go down believing that we can change the world if we all work together and doing everything I can to make that happen. This dialogue is a part of that. I know that you see doom, but I believe that news isn't good or bad, it's just what is -- good or bad comes in our reaction to it. All we can do is be the change we want to see in the world to the best of our ability and embrace whatever happens. It may come to naught in the end, but I think that how we live is more important than how we die. And it starts with a simple question:

What do you want?



If so, maybe Sam Wang and Nate Silver needed a kick in the ass too also. I think they'll try extra hard for accuracy in the coming years.

From my perspective (as a data scientist), Nate is a highly skilled data engineer and Sam is a neuroscientist doing rigorous work outside of his primary area of expertise. I would certainly like to see both of them address the issue of the reliability of polling and poll aggregation in a forthright and insightful manner. I think Sam would be more likely to try it, but either of them would have to win me over on the merits before they regained my trust in the value of their methodologies. That's what I'd do if I were in their places. The scientific credibility of your methodology is pretty much everything, in my opinion.
"Strategy without tactics is the slowest route to victory. Tactics without strategy is the noise before defeat."
---Sun Tzu (quoting Thomas Jefferson)
nam-myoho-renge-kyo---Thomas Jefferson (quoting Slartibartfast)

User avatar
Slartibartfast
Posts: 6849
Joined: Mon Aug 10, 2009 2:52 pm

Re: Presidential Approval Polls 2017

Post by Slartibartfast » Tue Oct 24, 2017 1:41 pm

bob wrote:
Tue Oct 24, 2017 2:18 am
Foggy wrote:Maybe the statistics generated during the 2016 election were not based on data that was measured accurately. It's harder to accurately measure the intent of a voter than it is to measure the length of a fish, I understand that much.
:yeah:

Too also: "At this point into his presidency" is technically correct, as we have data about the president's popularity on his 277th day in office, just as we have data for Obama's 277th day in office, etc.

But it is ultimately a misleading comparison because presidencies are not created equal: the recession drew Obama's numbers down, 9/11 inflated GWB's numbers, etc.
Yes, but it's more than that. There is information in the relationships in the data that is lost when statistics are extracted. The purpose of data visualizations (or one of them, I should say) is to pre-process data so it can be input into our best pattern recognition devices -- human visual cortexes. It's essentially Plato's allegory of the cave.

Go look at the data I pointed out and see what you think it means. Ask yourself how the comparisons used have been misleading and what additional understanding you get from looking at the data in context rather than just considering a few statistics. It's one of those things that I think everyone should do for themselves, instead of just looking at the shadows cast on the wall of the cave.
"Strategy without tactics is the slowest route to victory. Tactics without strategy is the noise before defeat."
---Sun Tzu (quoting Thomas Jefferson)
nam-myoho-renge-kyo---Thomas Jefferson (quoting Slartibartfast)

User avatar
Slartibartfast
Posts: 6849
Joined: Mon Aug 10, 2009 2:52 pm

Re: Presidential Approval Polls 2017

Post by Slartibartfast » Tue Oct 24, 2017 1:46 pm

RVInit wrote:
Mon Oct 23, 2017 8:11 pm
I wonder how much of 2016 "bad" predictions had to do with people outright lying. My younger sister lamented, totally unprovoked, that she had no idea who she would vote for, denying without even being asked that she would ever vote for Trump. This was the day before election day. The following day she was bragging and rubbing everyone's face in the fact that she voted for Trump and has been a nonstop Trump fanatic ever since. I have no doubt that she knew damn well she was going to vote for Trump and wasn't willing to admit it until after he won. I highly suspect her behavior of lying about her plans to vote for Trump had to do with being truly embarrassed to admit it before hand, and "see, lots of other people voted for him too" after it became obvious that he won.
That's a "known unknown". Look up the Bradley effect. There was discussion about this before the election (both ways) and probably more afterwards (I wasn't paying attention then), but I haven't seen any kind of analysis of its significance. In any case, it is something that must be taken into account, if only to be noted in how it increases the chance that the model is way off.
"Strategy without tactics is the slowest route to victory. Tactics without strategy is the noise before defeat."
---Sun Tzu (quoting Thomas Jefferson)
nam-myoho-renge-kyo---Thomas Jefferson (quoting Slartibartfast)

User avatar
NotaPerson
Posts: 2643
Joined: Mon May 18, 2015 9:33 pm

Re: Presidential Approval Polls 2017

Post by NotaPerson » Tue Oct 24, 2017 4:11 pm

It's worth noting that 538/Nate Silver actually did a damn fine job of predicting the 2016 national vote.

His final forecast (Nov. 8th) was that Clinton would get 48.5% of the popular vote, and that Trump would get 44.9%.

https://projects.fivethirtyeight.com/20 ... -forecast/

What happened? Clinton got 48.2% of the popular vote, to Trump's 46.1%.

https://en.wikipedia.org/wiki/United_St ... tion,_2016

It was at the state level (and thus the electoral college voting) where, for some reason, the polling forecasts were way off. I remember watching PA closely during election season, as it's a huge state. Poll after poll showed Clinton was going to win PA, often by a large margin. Yet she lost it (by less than a percent). :madguy:
Am I being detained?

User avatar
Addie
Posts: 21350
Joined: Mon Jun 15, 2009 6:22 am
Location: downstairs

Re: Presidential Approval Polls 2017

Post by Addie » Tue Oct 24, 2017 5:28 pm

Reuters
Fewer than a third of Americans back Trump tax plan: Reuters/Ipsos poll

WASHINGTON (Reuters) - Fewer than a third of Americans support Donald Trump’s tax-cut plan, according to a Reuters/Ipsos poll released on Tuesday, as the U.S. president went to Capitol Hill looking for Republican backing for his proposal to slash tax rates for individuals and companies.

As the 2018 midterm congressional election campaigns grow nearer, the poll found that more than two-thirds of registered voters said reducing the U.S. federal budget deficit is more important than cutting taxes for the wealthy or for corporations. ...

Among Republicans surveyed, 63 percent said deficit reduction should take priority over tax cuts for corporations, while 75 percent said deficit reduction should take priority over tax cuts for the wealthy. ...

More than half of the adults surveyed in the poll agreed that “cutting taxes for the poor is more important than reducing the federal deficit,” with 68 percent of Democrats and 47 percent of Republicans in accord with that statement, the poll showed.

The Oct. 20-23 poll found that only 15 percent of registered voters said Republicans in Congress should prioritize tax reform over other issues. About a quarter of those polled, including 23 percent of Democrats and 30 percent of Republicans, agreed that Congress should continue working on a healthcare bill.
¡Qué vergüenza!

User avatar
Addie
Posts: 21350
Joined: Mon Jun 15, 2009 6:22 am
Location: downstairs

Re: Presidential Approval Polls 2017

Post by Addie » Wed Oct 25, 2017 7:08 pm

The Hill
Trump's approval hits new low in Fox News poll

President Trump’s approval rating has hit a new low in Fox News polling.

Trump now has a 38 percent job approval rating in the Fox News poll released Wednesday, down from 42 percent last month. And his disapproval rating has hit a new high, up four points to 57 percent.

Fifty-six percent of those polled among white men without a college degree, a core constituency among Trump's base, now say they approve of the president, a drop from 68 percent in September and 71 percent last November.

Trump also got low marks for his handling of health care, with 60 percent disapproving of his approach on the topic.

White House chief of staff John Kelly earned a higher approval rating than Trump, with 42 percent approving of his job performance.

Melania Trump’s favorability rating also took a hit, falling from 51 percent in June in the same poll to 45 percent now.
¡Qué vergüenza!

User avatar
Slim Cognito
Posts: 4007
Joined: Fri Aug 29, 2014 8:37 am

Re: Presidential Approval Polls 2017

Post by Slim Cognito » Wed Oct 25, 2017 10:34 pm

Was not expecting to see Melania’s numbers fall.

User avatar
Azastan
Posts: 2331
Joined: Sat Feb 20, 2016 9:42 am

Re: Presidential Approval Polls 2017

Post by Azastan » Wed Oct 25, 2017 10:53 pm

Slim Cognito wrote:
Wed Oct 25, 2017 10:34 pm
Was not expecting to see Melania’s numbers fall.
I see my conservative acquaintances trying to boost Melania Trump as 'classy'. She did herself no favours when she didn't understand the joke Ivanka Trump made about being 'first lady'.

In addition, people were waiting for her to get going on her 'anti-bullying' stuff. She's full of good advice to school children, but can't see the fat forest for the trees.

User avatar
kate520
Posts: 13378
Joined: Mon Oct 26, 2009 3:02 pm

Re: Presidential Approval Polls 2017

Post by kate520 » Wed Oct 25, 2017 11:46 pm

Azastan wrote:
Wed Oct 25, 2017 10:53 pm
Slim Cognito wrote:
Wed Oct 25, 2017 10:34 pm
Was not expecting to see Melania’s numbers fall.
I see my conservative acquaintances trying to boost Melania Trump as 'classy'. She did herself no favours when she didn't understand the joke Ivanka Trump made about being 'first lady'.

In addition, people were waiting for her to get going on her 'anti-bullying' stuff. She's full of good advice to school children, but can't see the fat orange forest for the trees.
Patriotism is the last refuge of a Scoundrel.
Samuel Johnson
April 7, 1775

User avatar
pipistrelle
Posts: 2413
Joined: Wed Oct 19, 2016 11:26 am

Re: Presidential Approval Polls 2017

Post by pipistrelle » Thu Oct 26, 2017 12:02 am

I can't take pleasure in Trump being unpopular with his base because he's not destroying the country fast enough and in the right ways. I doubt there have been epiphanies.

User avatar
RVInit
Posts: 4177
Joined: Sat Mar 05, 2016 4:31 pm

Re: Presidential Approval Polls 2017

Post by RVInit » Thu Oct 26, 2017 9:57 am

pipistrelle wrote:
Thu Oct 26, 2017 12:02 am
I can't take pleasure in Trump being unpopular with his base because he's not destroying the country fast enough and in the right ways. I doubt there have been epiphanies.
I have seen one, and only one, Trump voter that is disappointed in anything he's actually done. That was a schoolteacher who voted for him and was disappointed in the appointment of Betsy DeVos and mumbled other things about Trump's Cabinet in general. He said he wouldn't vote for him again. Other than that, all the disappointment I have seen is because he has not burned the whole country down as he promised. But, hey, maybe there will be someone even more destructive that runs against him in 2020? No, he will get them fired up again, and they will vote for him. Look at all the help in destroying the country he will be getting from the new Congress members we may get in 2018.

We had better hope that Dems are able to take Flake's seat, which may be doable. And hopefully take some others, too. Because we are in for some years of hell if not. Oh, and let's not forget getting decent and sane people out to vote. That is if they can even stand to think about politics again after Trump.
"I know that human being and fish can coexist peacefully"
--- George W Bush

User avatar
Addie
Posts: 21350
Joined: Mon Jun 15, 2009 6:22 am
Location: downstairs

Re: Presidential Approval Polls 2017

Post by Addie » Thu Oct 26, 2017 11:35 am

The Hill
Poll: Majority of military officers view Trump unfavorably

A majority of U.S. military officers have an unfavorable view of President Trump, according to a survey released this week by the Military Times.

According to the poll, 53 percent of respondents said they hold an unfavorable opinion of the commander in chief, while about 31 percent said they view him favorably. Sixteen percent said they were neutral on the matter.

The findings come amid the controversy surrounding Trump's disputed call to the widow of a U.S. Army soldier killed in Niger earlier this month. ...

According to the Military Times poll, Trump's favorability is higher among enlisted military personnel, at about 47 percent. Just under 37 percent of those polled view him unfavorably.

Trump sees his highest enlisted favorability ratings in the Marine Corps — 58.9 percent — and the Army — 46.3 percent. The Air Force and the Navy have the highest rate of unfavorable opinions about the president, at 38.1 percent and 39.5 percent respectively.
¡Qué vergüenza!

User avatar
DejaMoo
Posts: 3210
Joined: Tue Jan 25, 2011 1:19 pm
Occupation: Agent of ZOG

Re: Presidential Approval Polls 2017

Post by DejaMoo » Thu Oct 26, 2017 11:44 am

Azastan wrote:
Wed Oct 25, 2017 10:53 pm
Slim Cognito wrote:
Wed Oct 25, 2017 10:34 pm
Was not expecting to see Melania’s numbers fall.
I see my conservative acquaintances trying to boost Melania Trump as 'classy'.
"Classy" doesn't do soft-porn nudie shoots. Share some of her men's magazine photos with your friends and ask them which one should be her official First Lady portrait. While you're at it, you can remind them that Trump made a cameo in a Playboy soft-porn video.

This is relevant because it proves the party that claims the moral high ground has no morals to speak of, since if they actually did, they couldn't in good conscience have voted for Trump, much less pretended he was the only choice from their christian religious perspective.

It also means we never have to concern ourselves with their manufactured moral outrage ever again, because this proves it's all fakery and only employed by them as a weapon.

User avatar
Sam the Centipede
Posts: 4990
Joined: Thu Aug 30, 2012 3:25 pm

Re: Presidential Approval Polls 2017

Post by Sam the Centipede » Thu Oct 26, 2017 11:50 am

Surely the most disgusting aspect of public opinion shown up by these polls is the relatively high approval rating, over one-third of the electorate. I can understand people with no interest in politics (yes, they exist) neither approving nor disapproving, but how does anybody get to approve of the Orange Shitgibbon's record to date? He's not done the things the deplorables wanted, nor has he done anything constructive that anybody sane wanted. All he has done is vomit a torrent of juvenile, hateful and narcissistic tweets and light a few fires. Regardless of what one wishes for, I can't see why anybody could approve of the Orange Shitgibbon.

That, I think, demonstrates a real problem in the US: that over one-third of the electorate are thoroughtly vile or idiotic.

User avatar
RVInit
Posts: 4177
Joined: Sat Mar 05, 2016 4:31 pm

Re: Presidential Approval Polls 2017

Post by RVInit » Thu Oct 26, 2017 12:03 pm

The one third of Americans that approve of Trump very likely are seeing much different "news" than you and I watch. I was subjected to quite a bit of Faux "news" and sometimes I even deliberately check out what they are saying just to stay in the loop about what they have to say. What I find fascinating is that Faux gets away with telling their viewers, over and over again, about all of Trump's accomplishments by making general statements, they never, ever, ever actually mention a single one. And their viewers don't even seem to notice that they are never given a single example. That is very typical of Faux, they talk in generalities and their viewers are totally conditioned to this and don't even question it.
"I know that human being and fish can coexist peacefully"
--- George W Bush

User avatar
Slartibartfast
Posts: 6849
Joined: Mon Aug 10, 2009 2:52 pm

Re: Presidential Approval Polls 2017

Post by Slartibartfast » Thu Oct 26, 2017 5:59 pm

RVInit wrote:
Thu Oct 26, 2017 12:03 pm
The one third of Americans that approve of Trump very likely are seeing much different "news" than you and I watch. I was subjected to quite a bit of Faux "news" and sometimes I even deliberately check out what they are saying just to stay in the loop about what they have to say. What I find fascinating is that Faux gets away with telling their viewers, over and over again, about all of Trump's accomplishments by making general statements, they never, ever, ever actually mention a single one. And their viewers don't even seem to notice that they are never given a single example. That is very typical of Faux, they talk in generalities and their viewers are totally conditioned to this and don't even question it.
This.

The Democrats have lost a propaganda war. Their would-be allies have been made to see them as unstable, irrational, and even malevolent enemies. They are now the resistance and if they hope to ever succeed, they will need to fight the Fox news framing from the only place that it can be done: inside the circle of trust of Trump voters. Calling these people thoroughly vile and idiotic and other anti-democratic* slurs, aside from being a truly stunning example of blaming the victim, as RVinit points out, is completely counterproductive. All you are doing is adding bricks to Trump's wall of hate instead of tearing it down.

* Anti-democratic because you are essentially saying that someone's fitness to vote depends on their voting for a candidate that is acceptable to you.

Now that I think about it, "Wall of Hate" is a good metaphor for Trump's fear mongering and jingoism. Anything that helps divide us only fuels that hate. It's time that we stopped playing the partisan games that Trump epitomizes and raising our own cults of personality to fight his and realize that we move forward together, or not at all.
"Strategy without tactics is the slowest route to victory. Tactics without strategy is the noise before defeat."
---Sun Tzu (quoting Thomas Jefferson)
nam-myoho-renge-kyo---Thomas Jefferson (quoting Slartibartfast)

User avatar
pipistrelle
Posts: 2413
Joined: Wed Oct 19, 2016 11:26 am

Re: Presidential Approval Polls 2017

Post by pipistrelle » Thu Oct 26, 2017 6:14 pm

But I thought it was stated here (somewhere) Fox doesn't actually have that many viewers. Although I don't know if that counts all the people stuck with it at work lounges, offices, terminals and stations, etc.

Dan1100
Posts: 1410
Joined: Sat Oct 18, 2014 3:41 pm

Re: Presidential Approval Polls 2017

Post by Dan1100 » Thu Oct 26, 2017 6:22 pm

U.S. elections are won and lost on turn out. As President Bartlett said, "Decision are made by the people that show up."

You aren't going to convince very many of the hard core Trump voters of the error of their ways. But, you can demoralize the hard core Trumpers, make them feel defeated even though their guy won, that their cause is futile, and make them stay home while getting the Democrats (who are the majority after all) to actual show up for a change.
"Let's say you're on trial for armed robbery. You say to the judge, 'I forgot armed robbery was illegal.' "

-Steve Martin

User avatar
Addie
Posts: 21350
Joined: Mon Jun 15, 2009 6:22 am
Location: downstairs

Re: Presidential Approval Polls 2017

Post by Addie » Thu Oct 26, 2017 8:57 pm

The majority self-identify as Independents, according to Gallup, as of Oct 2017: Republicans 24%; Independents 42%; Democrats 31%

Combined with leaning Independents, Democrats are in the majority: Republicans get 39% and Dems 46%
¡Qué vergüenza!

Post Reply

Return to “President & Executive Branch”