Register a SA Forums Account here!
JOINING THE SA FORUMS WILL REMOVE THIS BIG AD, THE ANNOYING UNDERLINED ADS, AND STUPID INTERSTITIAL ADS!!!

You can: log in, read the tech support FAQ, or request your lost password. This dumb message (and those ads) will appear on every screen until you register! Get rid of this crap by registering your own SA Forums Account and joining roughly 150,000 Goons, for the one-time price of $9.95! We charge money because it costs us money per month for bills, and since we don't believe in showing ads to our users, we try to make the money back through forum registrations.
 
  • Post
  • Reply
KingNastidon
Jun 25, 2004

twodot posted:

Yes acknowledging you had a sampling problem and then assuming it had no other issue is the problem.

You can look at the crosstab and see whether the other demographics are accurate within each age group. It only becomes a problem when they aren't. This is possible when you're attempting to control for multiple variables. For example, let's say a poll wants to hit 10% African American across the entire sample. Ideally you can do that where each age subgroup has 10% (if it doesn't vary by age in reality), but you could also throw them all in one age bucket. If race has strong predictive value on voter preference then you run risk of under/oversampling that demographic post re-weighting of age and skewing the output.

But again, crosstabs allow you to verify all this poll by poll. It seems to be a larger problem to write off polling altogether because you believe one sample attribute is being under/oversampled.

Adbot
ADBOT LOVES YOU

Civilized Fishbot
Apr 3, 2011

My Twitter Account posted:

Super-short answer: in a primary, people go to a polling location and cast ballots for their candidate; the ballots are counted and a winner is determined by a plurality of votes. In a caucus, they go to caucus locations where they are physically counted as supporters; "viable" candidates are determined by plurality of caucus supporters, and the remaining supporters regroup themselves to support those "viable" candidates until a winner is determined.

This is all true. I'd just add that the difference means that caucuses take way longer and don't allow early voting, so they're decided by a smaller group of more enthusiastic voters who have the free time and energy to go through the whole caucus process. In 2016 Bernie did a little better in caucuses than in normal primaries because Hillary voters weren't excited enough about her to do a caucus all night.

VitalSigns
Sep 3, 2011

Chilichimp posted:

Voter suppression and EC gerrymandering is a bitch, ain't it

like... polls weren't super turbo wrong, but pollsters were with their prognostications. The "there's not way Trump can win" brain rot was loving strong. Besides, polls are just one data point, and folks shouldn't hang on them, but we're all guilty of it.

Primary polls in 2016 were horrible. 20-40 point misses weren't uncommon right before the vote, and the polls in Sep 2015 were complete trash as far as predicting specific results

twodot
Aug 7, 2005

You are objectively correct that this person is dumb and has said dumb things

KingNastidon posted:

But again, crosstabs allow you to verify all this poll by poll. It seems to be a larger problem to write off polling altogether because you believe one sample attribute is being under/oversampled.
No one is writing off polling altogether, you just can't fixing a sampling problem by saying "Whelp the results I got for the demographic I under polled are probably right anyways, I will just make those numbers bigger and no more sampling problem". Sampling problems happen because of reasons, these reasons have unpredictable impacts, you can't turn bad data into good data.

necrobobsledder
Mar 21, 2005
Lay down your soul to the gods rock 'n roll
Nap Ghost
Give no fucks anymore here but this is the Bernie echo chamber and people here act like it's a complete joke, I'm YangGang. We exist, and a lot of people are at least uh... Yang-Curious even with single digits, but it's mostly a catch-22 situation (he converts pretty well once you watch him speak outside a debate for a while). All of his supporters understand he sounds loving insane or a capitalism apologist / cuck or whatever - we were all there. I've evolved on my Bernie stance as he's laid out more policies. As a former federal employee I would hardly wish employment in it upon my worst enemies ever so can't agree on Bernie's FJG because we may be the first civilization to destroy itself through bureaucracy in both public and private sector. I consider Yang in the same human dignity centered socialist spirit of Bernie but acknowledging that anyone labeling themselves socialist in any way is going to not get anything passed even with a house that's democratic when the GOP cultist capitalists are running our election commissions, districting, Senate, and now even Supreme Court.

My general political position as a bleeding heart is that the left lost politically decades ago so peeling off the now-ruling caste right-wing to help working Americans via the same cultural channels perpetuated by the winners would fix 70%+ of our issues at home because being poor or stressed from overwork (if you're not a poor) is literally making us stupider, more racist, more fearful, and angrier perpetuating the cycle of democratic fail (we are already at Iran or Saudia Arabia levels of "lol, democracy"). I might be crazy enough to give up on gun control permanently if we got M4A on condition that's a Constitutional Amendment somehow, for example. Lots of the rear end in a top hat Party are only there for guns, which is weird as gently caress to me because it's not like us gun owners go out shooting every day on principle when our health, infrastructure, and environment is measurably more important statistically for most people than defending ourselves from a shooter, but I'm convinced after living in GA, NC, and VA as a minority that if liberals gave up on gun control 70%+ of elections would go better for them. Are guns that much more important than life with dignity, infrastructure that helps all Americans, healthcare under control, and corruption of predatory corporate influences? My gut instinct tells me no.

Cool with Bernie if he gets the general nomination in all honesty though I am concerned about the GOP blockade of facts in place. I have zero love for the "centrist" (read: win 3 battles and lose the war, rinse, repeat) establishment so much that if Biden gets the nomination I might literally go vote Trump in 2020 to show how much I hate the DNC. Honestly, I'll probably do the same thing I did in 2012 because I hated Righter Than Reagan Obama then - vote Green Party like the whole 2k other people in my entire state (that was crushed by the Constitution party nearly 10:1 because it's the home state of the NRA, ugh).


With love, respect, and camaraderie - the people crazy enough to believe late stage capitalism and militant libertarianism can be reformed to not destroy us and to use its strangehold on the American psyche to advance us beyond petty bullshit and join the ranks of countries that care about human beings. If you want to hate on us, Bernie Bros, feel pity that late stage capitalism has driven us to this point

#YangGang

Nonsense
Jan 26, 2007

YANG!

Typo
Aug 19, 2009

Chernigov Military Aviation Lyceum
The Fighting Slowpokes

VitalSigns posted:

Primary polls in 2016 were horrible. 20-40 point misses weren't uncommon right before the vote, and the polls in Sep 2015 were complete trash as far as predicting specific results

"this poll is the end of the bernie campaign" -increasingly nervous man for the 8th time this year

Majorian
Jul 1, 2009

I'd rather not vote for a tech millionaire whose version of the UBI lays waste to the social safety net and has very little else to offer.

reignonyourparade
Nov 15, 2012

Civilized Fishbot posted:

This is all true. I'd just add that the difference means that caucuses take way longer and don't allow early voting, so they're decided by a smaller group of more enthusiastic voters who have the free time and energy to go through the whole caucus process. In 2016 Bernie did a little better in caucuses than in normal primaries because Hillary voters weren't excited enough about her to do a caucus all night.

Multiple caucuses do allow early voting.

Wicked Them Beats
Apr 1, 2007

Moralists don't really *have* beliefs. Sometimes they stumble on one, like on a child's toy left on the carpet. The toy must be put away immediately. And the child reprimanded.

twodot posted:

No one is writing off polling altogether, you just can't fixing a sampling problem by saying "Whelp the results I got for the demographic I under polled are probably right anyways, I will just make those numbers bigger and no more sampling problem". Sampling problems happen because of reasons, these reasons have unpredictable impacts, you can't turn bad data into good data.

I sort of am.

You write off polling because it is impossible to ask a person today what they will do six months from now and get a reliable answer. People lie, sometimes to the pollster and sometimes to themselves. If the person conducting the poll reads a name in a slightly negative tone of voice people might not pick that candidate because they're worried the person asking the question won't like them. If your poll is too long people stop answering honestly just to get it over with, and while you can try to rotate question order to overcome this even the order you ask questions in influences the outcome. Phrasing guides responses (see: push polls), leading to fights over whether questions are worded neutrally or not, or even what wording something neutrally looks like. And since you can't force people to take your poll your sample is ultimately self-selecting. Which raises the serious question of whether the people who select into polls are the same as the people who actually vote or get involved, and we can safely say from previous polling failures that there isn't a 1:1 relationship between those two groups.

I like numbers and going through crosstabs, but you have to be honest and admit that the only poll that actually tells you anything at the end of the day is the actual vote itself. And the only similarity between the polls and the election day vote is that they're conducted in equally arbitrary fashions.

ded redd
Aug 1, 2010

by Fluffdaddy

necrobobsledder posted:

Give no fucks anymore here but this is the Bernie echo chamber and people here act like it's a complete joke, I'm YangGang. We exist, and a lot of people are at least uh... Yang-Curious even with single digits, but it's mostly a catch-22 situation (he converts pretty well once you watch him speak outside a debate for a while). All of his supporters understand he sounds loving insane or a capitalism apologist / cuck or whatever - we were all there. I've evolved on my Bernie stance as he's laid out more policies. As a former federal employee I would hardly wish employment in it upon my worst enemies ever so can't agree on Bernie's FJG because we may be the first civilization to destroy itself through bureaucracy in both public and private sector. I consider Yang in the same human dignity centered socialist spirit of Bernie but acknowledging that anyone labeling themselves socialist in any way is going to not get anything passed even with a house that's democratic when the GOP cultist capitalists are running our election commissions, districting, Senate, and now even Supreme Court.

My general political position as a bleeding heart is that the left lost politically decades ago so peeling off the now-ruling caste right-wing to help working Americans via the same cultural channels perpetuated by the winners would fix 70%+ of our issues at home because being poor or stressed from overwork (if you're not a poor) is literally making us stupider, more racist, more fearful, and angrier perpetuating the cycle of democratic fail (we are already at Iran or Saudia Arabia levels of "lol, democracy"). I might be crazy enough to give up on gun control permanently if we got M4A on condition that's a Constitutional Amendment somehow, for example. Lots of the rear end in a top hat Party are only there for guns, which is weird as gently caress to me because it's not like us gun owners go out shooting every day on principle when our health, infrastructure, and environment is measurably more important statistically for most people than defending ourselves from a shooter, but I'm convinced after living in GA, NC, and VA as a minority that if liberals gave up on gun control 70%+ of elections would go better for them. Are guns that much more important than life with dignity, infrastructure that helps all Americans, healthcare under control, and corruption of predatory corporate influences? My gut instinct tells me no.

Cool with Bernie if he gets the general nomination in all honesty though I am concerned about the GOP blockade of facts in place. I have zero love for the "centrist" (read: win 3 battles and lose the war, rinse, repeat) establishment so much that if Biden gets the nomination I might literally go vote Trump in 2020 to show how much I hate the DNC. Honestly, I'll probably do the same thing I did in 2012 because I hated Righter Than Reagan Obama then - vote Green Party like the whole 2k other people in my entire state (that was crushed by the Constitution party nearly 10:1 because it's the home state of the NRA, ugh).


With love, respect, and camaraderie - the people crazy enough to believe late stage capitalism and militant libertarianism can be reformed to not destroy us and to use its strangehold on the American psyche to advance us beyond petty bullshit and join the ranks of countries that care about human beings. If you want to hate on us, Bernie Bros, feel pity that late stage capitalism has driven us to this point

#YangGang

You got the nofap candidate lol

(USER WAS PUT ON PROBATION FOR THIS POST)

twodot
Aug 7, 2005

You are objectively correct that this person is dumb and has said dumb things

Wicked Them Beats posted:

I sort of am.

You write off polling because it is impossible to ask a person today what they will do six months from now and get a reliable answer.
If you are a sane person who is reading a poll, which is to say you are high level staff in a campaign who has to make strategic decisions, you are not expecting polling today to tell you what individuals will do six months from now. You're using polling today to make resourcing decisions today, and it won't be perfect, but it's going to be better than guessing (though people will also just guess because people do that).

KingNastidon
Jun 25, 2004

Wicked Them Beats posted:

I like numbers and going through crosstabs, but you have to be honest and admit that the only poll that actually tells you anything at the end of the day is the actual vote itself. And the only similarity between the polls and the election day vote is that they're conducted in equally arbitrary fashions.

All of your limitations and potential biases of polling are true, but despite them, market research is a multibillion dollar industry. Some data on anticipated behaviors is better than absolute conjecture. It'd be nice of a larger portion of the country was fluent in statistics and polling methodology such that they could better interpret and mentally caveat any data they see, but I don't know of a better alternative than asking people who they plan to vote for.

VitalSigns
Sep 3, 2011

necrobobsledder posted:

As a former federal employee I would hardly wish employment in it upon my worst enemies ever so can't agree on Bernie's FJG because we may be the first civilization to destroy itself through bureaucracy in both public and private sector.

Have you ever worked in the private sector.

What do you think of Yang's proposal that every law expire automatically after 5 years. I'm sure he could easily get that passed through even a GOP Congress, what do you think the effect will be.

Groovelord Neato
Dec 6, 2014



don't steal redneck nazgul's bit.

Ytlaya
Nov 13, 2005

Yang poster may have a bad opinion, but at least they're explicit about what they think and want (which is sadly not something you can say about a bunch of other people).

KingNastidon posted:

All of your limitations and potential biases of polling are true, but despite them, market research is a multibillion dollar industry. Some data on anticipated behaviors is better than absolute conjecture. It'd be nice of a larger portion of the country was fluent in statistics and polling methodology such that they could better interpret and mentally caveat any data they see, but I don't know of a better alternative than asking people who they plan to vote for.

The better alternative is for the media to not report polling, or to at least not report it very frequently or give much emphasis to it. Polling serves no purpose other than guiding the decisions of campaigns themselves, but it should not factor into the way people vote (or otherwise form political opinions). As is, it just sends the message to people that they shouldn't support unpopular (or "unpopular") ideas out of a misguided sense of "pragmatism."

edit: The only possible purpose I can think of for normal people is determining "who is remotely viable," but that doesn't require frequent polls and could be conveyed through some other measure (like who qualifies for debates).

Wicked Them Beats
Apr 1, 2007

Moralists don't really *have* beliefs. Sometimes they stumble on one, like on a child's toy left on the carpet. The toy must be put away immediately. And the child reprimanded.

twodot posted:

If you are a sane person who is reading a poll, which is to say you are high level staff in a campaign who has to make strategic decisions, you are not expecting polling today to tell you what individuals will do six months from now. You're using polling today to make resourcing decisions today, and it won't be perfect, but it's going to be better than guessing (though people will also just guess because people do that).

True, but that's not how polls are explained to the masses. Polls are explicitly sold as "if the election were held today" predictive devices.

I think having polling would be better than not in an ideal world, but in the world we live in they're just another tool of the elite to disenfranchise and demoralize the proles who might dare demand change. "You want to impeach Trump? Well take a look at this poll that says no one agrees with you! Guess you're just an idiot, huh?"

KingNastidon posted:

All of your limitations and potential biases of polling are true, but despite them, market research is a multibillion dollar industry. Some data on anticipated behaviors is better than absolute conjecture. It'd be nice of a larger portion of the country was fluent in statistics and polling methodology such that they could better interpret and mentally caveat any data they see, but I don't know of a better alternative than asking people who they plan to vote for.

Markets aren't rational and a lot of money spent on something doesn't mean it actually does anything.

And your comparison being market research is fitting. Market research is more often about how to convince someone that they need or want something, not to gauge actual needs or desires. The same with political polling: how can we convince someone that Biden/Harris/Warren/etc is the winner and if they want to be a winner they should vote for that person? It's not about finding out what the people want, it's about the most effective way to tell them what they want and make them believe that the idea originated internally instead of externally.

Majorian
Jul 1, 2009

KingNastidon posted:

All of your limitations and potential biases of polling are true, but despite them, market research is a multibillion dollar industry.

There, uh...there might be a reason for that.

HootTheOwl
May 13, 2012

Hootin and shootin

KingNastidon posted:

All of your limitations and potential biases of polling are true, but despite them, market research is a multibillion dollar industry. Some data on anticipated behaviors is better than absolute conjecture. It'd be nice of a larger portion of the country was fluent in statistics and polling methodology such that they could better interpret and mentally caveat any data they see, but I don't know of a better alternative than asking people who they plan to vote for.

You can't artificially increase a sample size by over-counting it.
Think of it this way: If your survey only catches only one person from a demographic that represents 30% of the population, you can't multiply them by 100 million.

Typo
Aug 19, 2009

Chernigov Military Aviation Lyceum
The Fighting Slowpokes

Majorian posted:

There, uh...there might be a reason for that.

:thunk:

VitalSigns
Sep 3, 2011

Ytlaya posted:

Yang poster may have a bad opinion, but at least they're explicit about what they think and want (which is sadly not something you can say about a bunch of other people).


Yeah it's definitely a nice change to have someone advocate in good faith for a non-Bernie candidate they sincerely believe in and give actual reasons to back it up. Even if I think those reasons are wrong or ill-considered, I'll take it over the people who hate Bernie but can't articulate why so they resort to fakeposting dumb poo poo that they think makes Bernie Bros look bad

KingNastidon
Jun 25, 2004

Wicked Them Beats posted:

And your comparison being market research is fitting. Market research is more often about how to convince someone that they need or want something, not to gauge actual needs or desires. The same with political polling: how can we convince someone that Biden/Harris/Warren/etc is the winner and if they want to be a winner they should vote for that person? It's not about finding out what the people want, it's about the most effective way to tell them what they want and make them believe that the idea originated internally instead of externally.

The only difference between market research and political polling is that political polls are communicated externally while market research is often not. If you do market research to understand preference of various potential Doritos flavors and intent to purchase you aren't influencing demand in any way. First, you wouldn't design market research to bias any single option if your goal was to arrive at the option that would maximize sales. You can manipulate research methodology to arrive at the wrong answer, but there's no motivation to do so for a for-profit company. Second, the sample size necessary for the data to be statistically significant is inconsequential relative to total potential consumers.

The alternative of not sharing public polling is worse than showing it. If 50% voters are telling pollsters that they intend to support Bernie Sanders then the media must react accordingly to that data. The alternative is that 50% of people may support Bernie, but far fewer of those people exist at MSNBC or NYT. Therefore the attention they spend on any given candidate is completely subjective and unable to be criticized because there's no data to show their coverage is not representative of broader public opinion. This especially helps someone like Sanders because conventional wisdom would indicate a socialist independent would not be viable given historical precedent. The data is the only thing to show otherwise.

It's not the pollsters or media's problem that voters hide their true preferences in an anonymous survey because they're concerned (why?) their choice may not align with the eventual plurality.

HootTheOwl posted:

You can't artificially increase a sample size by over-counting it.
Think of it this way: If your survey only catches only one person from a demographic that represents 30% of the population, you can't multiply them by 100 million.

Issues with sample are never this extreme. You're correct I wouldn't advocate re-weighting a subgroup within a sample by 10,000,000X or even 2X. The vast majority of sample biases that people freak out about are on the magnitude of changing 30% to 35% or whatever. If the sample size that made up the 30% was reasonably large then there are generally few issues with slightly overweighting that subgroup data in the final aggregated result.

KingNastidon fucked around with this message at 23:07 on Sep 22, 2019

Judakel
Jul 29, 2004
Probation
Can't post for 9 years!

Nonsense posted:

Remember when Lizzo posted positively about Bernie and succdems shamed her into deleting it?

We gonna obliterate these ghouls.

https://twitter.com/leslieleeiii/status/1175422761205993472

We need to keep bringing it up. The only people eyerolling about it are white chuds and trust fund brats also hoteps.

We will either obliterate them, or show them centrists are no longer viable and humiliate them.

Unoriginal Name
Aug 1, 2006

by sebmojo

KingNastidon posted:

The only difference between market research and political polling is that political polls are communicated externally while market research is often not. If you do market research to understand preference of various potential Doritos flavors and intent to purchase you aren't influencing demand in any way. First, you wouldn't design market research to bias any single option if your goal was to arrive at the option that would maximize sales. You can manipulate research methodology to arrive at the wrong answer, but there's no motivation to do so for a for-profit company. Second, the sample size necessary for the data to be statistically significant is inconsequential relative to total potential consumers.

The alternative of not sharing public polling is worse than showing it. If 50% voters are telling pollsters that they intend to support Bernie Sanders then the media must react accordingly to that data. The alternative is that 50% of people may support Bernie, but far fewer of those people exist at MSNBC or NYT. Therefore the attention they spend on any given candidate is completely subjective and unable to be criticized because there's no data to show their coverage is not representative of broader public opinion. This especially helps someone like Sanders because conventional wisdom would indicate a socialist independent would not be viable given historical precedent. The data is the only thing to show otherwise.

It's not the pollsters or media's problem that voters hide their true preferences in an anonymous survey because they're concerned (why?) their choice may not align with the eventual plurality.


Issues with sample are never this extreme. You're correct I wouldn't advocate re-weighting a subgroup within a sample by 10,000,000X or even 2X. The vast majority of sample biases that people freak out about are on the magnitude of changing 30% to 35% or whatever. If the sample size that made up the 30% was reasonably large then there are generally few issues with slightly overweighting using that subgroup data in the final aggregated result.

Which poll was it that had 9 people under the age of 30 again

MSDOS KAPITAL
Jun 25, 2018





KingNastidon posted:

You can manipulate research methodology to arrive at the wrong answer, but there's no motivation to do so for a for-profit company.
Motherfucker have you never heard of the Pepsi Challenge? Companies conduct motivated sampling all the time and then blast the results out for everyone to see. And there is probably even more incentive to do so in politics: just like more people will try a new Doritos flavor if they think a lot a people like it, "electability" is hugely important to primary voters.

KingNastidon
Jun 25, 2004

MSDOS KAPITAL posted:

Motherfucker have you never heard of the Pepsi Challenge? Companies conduct motivated sampling all the time and then blast the results out for everyone to see. And there is probably even more incentive to do so in politics: just like more people will try a new Doritos flavor if they think a lot a people like it, "electability" is hugely important to primary voters.

Right, a corporation sharing their own internal market research outputs without survey design, sample information, data collection methodology, etc should be taken with a grain of salt. I'd hope most consumers are savvy enough to know this. I'm not going to keep drinking Pepsi just because Pepsi polling tells me it's great.

Thankfully political polling, what we're talking about, is nothing like this. They provide transparency into all those factors and independent third parties like Monmouth or Selzar have no obvious personal stake in arriving at a certain outcome. Unless, of course, there's a vast conspiracy among numbers shitbergsteins to secretly manipulate the raw survey data against certain candidates. If you think herd mindset hinders individual expression of preferences then what do you think would happen when there's zero data driven evidence that others share your views?

Mahoning
Feb 3, 2007

KingNastidon posted:

Right, a corporation sharing their own internal market research outputs without survey design, sample information, data collection methodology, etc should be taken with a grain of salt. I'd hope most consumers are savvy enough to know this. I'm not going to keep drinking Pepsi just because Pepsi polling tells me it's great.

Thankfully political polling, what we're talking about, is nothing like this. They provide transparency into all those factors and independent third parties like Monmouth or Selzar have no obvious personal stake in arriving at a certain outcome. Unless, of course, there's a vast conspiracy among numbers shitbergsteins to secretly manipulate the raw survey data against certain candidates. If you think herd mindset hinders individual expression of preferences then what do you think would happen when there's zero data driven evidence that others share your views?

Except we gained a lot of insight into polling with the recent “outlier” poll where several prominent pollsters admitted that they toss out polls that don’t fit their predefined narrative and the pollster that conducted the outlier (was it Monmouth?) was basically bullied and ridiculed into retracting the poll and declaring it an outlier.

If EVERYONE is doing this, the whole thing is essentially rigged since we’re not actually seeing all the polls. Only the ones that “fit”.

It makes me wonder how many polls were conducted in 2016 that showed Trump leading in certain states and the pollsters just thought “hmmmm that doesn’t seem right, keep going until we get one that shows Clinton leading.”

KingNastidon
Jun 25, 2004

Mahoning posted:

Except we gained a lot of insight into polling with the recent “outlier” poll where several prominent pollsters admitted that they toss out polls that don’t fit their predefined narrative and the pollster that conducted the outlier (was it Monmouth?) was basically bullied and ridiculed into retracting the poll and declaring it an outlier.

They didn't retract the poll. The said it was a likely outlier, but explained why they still publish outliers. There are good pollsters and there are bad, unethical pollsters. Which is why it's useful that badman Nate attempts to grade them such that people can better interpret results poll by poll.

https://twitter.com/MonmouthPoll/status/1166729267457265665?s=19

Mellow Seas
Oct 9, 2012
Probation
Can't post for 10 years!

VitalSigns posted:

Yeah it's definitely a nice change to have someone advocate in good faith for a non-Bernie candidate they sincerely believe in and give actual reasons to back it up. Even if I think those reasons are wrong or ill-considered, I'll take it over the people who hate Bernie but can't articulate why so they resort to fakeposting dumb poo poo that they think makes Bernie Bros look bad
There may be people doing this, I don't know, but it's definitely way less than the amount of people you think are doing it.

And thank you to YangGuy for his post, it was refreshing and honest. If we put any stock in the polls, it's enough to know that a majority of people aren't supporting Bernie at the moment, so it's a nice change of pace to have someone from (a minority of) that majority explain their reasoning in this thread. It's theoretically what the thread is for!

Ytlaya
Nov 13, 2005

KingNastidon posted:

Right, a corporation sharing their own internal market research outputs without survey design, sample information, data collection methodology, etc should be taken with a grain of salt. I'd hope most consumers are savvy enough to know this. I'm not going to keep drinking Pepsi just because Pepsi polling tells me it's great.

Thankfully political polling, what we're talking about, is nothing like this. They provide transparency into all those factors and independent third parties like Monmouth or Selzar have no obvious personal stake in arriving at a certain outcome. Unless, of course, there's a vast conspiracy among numbers shitbergsteins to secretly manipulate the raw survey data against certain candidates. If you think herd mindset hinders individual expression of preferences then what do you think would happen when there's zero data driven evidence that others share your views?

Uh, you realize that there's a problem in Actual Science with manipulating data/statistical criteria to achieve the desired outcome. The idea that political pollsters, who are anything but ideologically neutral, might tailor their assumptions in a way that de-emphasizes the Democrats who support Sanders is not nearly as Wild And Crazy as you seem to be implying.

I don't necessarily think that this is happening (since they can effectively make his chances look low by just assuming turn-out will be similar to how it was in the past), but it's hardly outlandish, especially in light of the tweet by the pollster that explicitly advocated against releasing polling results that go against what they've arbitrarily decided to be "the correct results" (that tweet in response to the Monmouth claim of their poll being an "outlier" is what people are referencing).

And none of this means that polling (or at least the reporting of polling) is a good idea. It simply isn't useful information, and only has the impact of making the status quo more resistant to change.

Wicked Them Beats
Apr 1, 2007

Moralists don't really *have* beliefs. Sometimes they stumble on one, like on a child's toy left on the carpet. The toy must be put away immediately. And the child reprimanded.

KingNastidon posted:

independent third parties like Monmouth or Selzar have no obvious personal stake

KingNastidon posted:

They didn't retract the poll. The said it was a likely outlier, but explained why they still publish outliers. There are good pollsters and there are bad, unethical pollsters. Which is why it's useful that badman Nate attempts to grade them such that people can better interpret results poll by poll.

https://twitter.com/MonmouthPoll/status/1166729267457265665?s=19

Their ability to get funding for polls is tied directly to how their results are viewed, so they have an explicit stake in adhering to the dominant narrative. These primary polls are advertising for the contract polling they provide to politicians and policy groups. That they stepped outside of the orthodoxy established by the Council of Nate and were forced to make amends to get back into the good graces of the polling community does not strengthen your claim that these are aloof scientists, handing down unbiased fact. In fact, the very existence of a polling community, that all these people know each other and work with each other and tweet at each other and can convince each other to write letters downplaying their own work, should be very concerning.

tylersayten
Mar 20, 2019

by LITERALLY AN ADMIN

Typo posted:

lol identity politics voters are literally going to destroy the planet so a few minorities can be billionaires

This is what American liberalism has always been about.

KingNastidon
Jun 25, 2004

Wicked Them Beats posted:

Their ability to get funding for polls is tied directly to how their results are viewed, so they have an explicit stake in adhering to the dominant narrative. These primary polls are advertising for the contract polling they provide to politicians and policy groups.

So all of these pollsters are colluding to have similar results that favor specific candidate(s) such that they all can sell their contract polling to them when that candidate wins? Why would that chosen candidate need so many pollsters? Under this conspiracy wouldn't it make sense for each polling outfit to align with a given candidate to ensure their services are used if they win? Why hasn't this cabal been exposed given campaign managers routinely bounce around from candidate to candidate every 2-4 years?

Wicked Them Beats posted:

That they stepped outside of the orthodoxy established by the Council of Nate and were forced to make amends to get back into the good graces of the polling community does not strengthen your claim that these are aloof scientists, handing down unbiased fact.

Still don't understand where this is coming from. People questioned the results from a poll. It was generally agreed upon to be an outlier given it fell outside the 95% CI of similar polls. Monmouth, a good and credible polling org, reiterated their rationale for publishing it. They could have just never published the poll and no one would have ever known.

Wicked Them Beats posted:

In fact, the very existence of a polling community, that all these people know each other and work with each other and tweet at each other and can convince each other to write letters downplaying their own work, should be very concerning.

I talk shop with people in my industry because I don't know people outside of my industry nor have the subject matter expertise or credibility to contribute to their discussion. People communicating with others in their field isn't an indication of a vast conspiracy. Again, the survey and samples are publicly available. The only thing you should be concerned about is manipulation of raw data or tossing out entire polls. I don't doubt the latter exists among less credible outlets, but there's no evidence of the former. I don't think it's appropriate to assume data manipulation is prevalent just because polling outcomes don't align with my preconceived notions.

TheOneAndOnlyT
Dec 18, 2005

Well well, mister fancy-pants, I hope you're wearing your matching sweater today, or you'll be cut down like the ugly tree you are.

Wicked Them Beats posted:

Their ability to get funding for polls is tied directly to how their results are viewed, so they have an explicit stake in adhering to the dominant narrative. These primary polls are advertising for the contract polling they provide to politicians and policy groups.
This makes no sense. Who would want to contract a pollster who only "adheres to the dominant narrative"? The whole point of the contract polling is to give those politicians and policy groups accurate information, so they can figure out whom to target and where to spend their money. A pollster who only tells their clients what they want to hear would be the most worthless pollster imaginable.

Majorian
Jul 1, 2009

TheOneAndOnlyT posted:

This makes no sense. Who would want to contract a pollster who only "adheres to the dominant narrative"? The whole point of the contract polling is to give those politicians and policy groups accurate information, so they can figure out whom to target and where to spend their money. A pollster who only tells their clients what they want to hear would be the most worthless pollster imaginable.

One of the purposes behind polling firms, and a big part of why they get the funding they do, is to shape the media narrative.

Typo
Aug 19, 2009

Chernigov Military Aviation Lyceum
The Fighting Slowpokes

TheOneAndOnlyT posted:

This makes no sense. Who would want to contract a pollster who only "adheres to the dominant narrative"? The whole point of the contract polling is to give those politicians and policy groups accurate information, so they can figure out whom to target and where to spend their money. A pollster who only tells their clients what they want to hear would be the most worthless pollster imaginable.

the whole point of polls is to drive the "electibility" narrative so the proles stay down and don't dare want anything better than slightly better prison cells under capitaism

tylersayten
Mar 20, 2019

by LITERALLY AN ADMIN

necrobobsledder posted:

Give no fucks anymore here but this is the Bernie echo chamber and people here act like it's a complete joke, I'm YangGang. We exist, and a lot of people are at least uh... Yang-Curious even with single digits, but it's mostly a catch-22 situation (he converts pretty well once you watch him speak outside a debate for a while). All of his supporters understand he sounds loving insane or a capitalism apologist / cuck or whatever - we were all there. I've evolved on my Bernie stance as he's laid out more policies. As a former federal employee I would hardly wish employment in it upon my worst enemies ever so can't agree on Bernie's FJG because we may be the first civilization to destroy itself through bureaucracy in both public and private sector. I consider Yang in the same human dignity centered socialist spirit of Bernie but acknowledging that anyone labeling themselves socialist in any way is going to not get anything passed even with a house that's democratic when the GOP cultist capitalists are running our election commissions, districting, Senate, and now even Supreme Court.

My general political position as a bleeding heart is that the left lost politically decades ago so peeling off the now-ruling caste right-wing to help working Americans via the same cultural channels perpetuated by the winners would fix 70%+ of our issues at home because being poor or stressed from overwork (if you're not a poor) is literally making us stupider, more racist, more fearful, and angrier perpetuating the cycle of democratic fail (we are already at Iran or Saudia Arabia levels of "lol, democracy"). I might be crazy enough to give up on gun control permanently if we got M4A on condition that's a Constitutional Amendment somehow, for example. Lots of the rear end in a top hat Party are only there for guns, which is weird as gently caress to me because it's not like us gun owners go out shooting every day on principle when our health, infrastructure, and environment is measurably more important statistically for most people than defending ourselves from a shooter, but I'm convinced after living in GA, NC, and VA as a minority that if liberals gave up on gun control 70%+ of elections would go better for them. Are guns that much more important than life with dignity, infrastructure that helps all Americans, healthcare under control, and corruption of predatory corporate influences? My gut instinct tells me no.

Cool with Bernie if he gets the general nomination in all honesty though I am concerned about the GOP blockade of facts in place. I have zero love for the "centrist" (read: win 3 battles and lose the war, rinse, repeat) establishment so much that if Biden gets the nomination I might literally go vote Trump in 2020 to show how much I hate the DNC. Honestly, I'll probably do the same thing I did in 2012 because I hated Righter Than Reagan Obama then - vote Green Party like the whole 2k other people in my entire state (that was crushed by the Constitution party nearly 10:1 because it's the home state of the NRA, ugh).


With love, respect, and camaraderie - the people crazy enough to believe late stage capitalism and militant libertarianism can be reformed to not destroy us and to use its strangehold on the American psyche to advance us beyond petty bullshit and join the ranks of countries that care about human beings. If you want to hate on us, Bernie Bros, feel pity that late stage capitalism has driven us to this point

#YangGang

I have more respect for #YangGang people because at least they’re honest about why they’re voting for him, which is infinitely more than I can say for Warren voters.

There are a lot of Yang Gangers in the city I live in, and they get along with us Bernie people quite well. Most Yang people will vote for Bernie in the general - at least the ones I talk to - because Yang’s base is mostly people who’s lives have been completely destroyed by late capitalism, and the popularity of that $1000 a month just shows you how desperate people have become under our oligarchical, fascist system.

Calibanibal
Aug 25, 2015

"polls aren't real," i assure myself as i close my eyes and donate $5 to a campaign that obviously isn't going that well.

Majorian
Jul 1, 2009

Calibanibal posted:

"polls aren't real," i assure myself as i close my eyes and donate $5 to a campaign that obviously isn't going that well.

But enough about Kamala, Pete, Delaney, et all...

Adbot
ADBOT LOVES YOU

Typo
Aug 19, 2009

Chernigov Military Aviation Lyceum
The Fighting Slowpokes

Majorian posted:

But enough about Kamala, Pete, Delaney, et all...

you are engaging with a gimmick don't waste your time

  • 1
  • 2
  • 3
  • 4
  • 5
  • Post
  • Reply