Ogmius815 posted:The entire method of the unskewing guy was he assumed, without good evidence, that all the polls were over sampling democrats and under sampling republicans. He then systematically reweighted all the polls to accord with what he assumed the electorate actually looked like. He was an idiot. Don’t be like him. You don’t have secret knowledge of how polls should be weighted. The polling models may be wrong (indeed, most of them have to be wrong since they don’t agree with one another), but speculation about how and in what direction they’ll be wrong is almost always motivated and uninformed. Unskewing is fine for what-if analysis, especially in the context of the primary where turn-out is less certain. It allows people that believe the sample population is wrong to understand the sensitivity to various scenarios. For example, what would the under 30 turnout have to be for Sanders to win given his advantage there and disadvantage with 65+. It helps discourage the worst, uninformed posting that all polls are invalid because they can't possibly predict eventual turn-out by demographics. If a candidate leads or nearly leads all subgroups in the crosstabs then sample bias doesn't really matter other than higher MOE within those subgroups.
|
|
# ? Sep 23, 2019 03:26 |
|
|
# ? Jun 7, 2024 02:48 |
|
KingNastidon posted:Unskewing is fine for what-if analysis, especially in the context of the primary where turn-out is less certain. It allows people that believe the sample population is wrong to understand the sensitivity to various scenarios. For example, what would the under 30 turnout have to be for Sanders to win given his advantage there and disadvantage with 65+.
|
# ? Sep 23, 2019 03:30 |
|
Groovelord Neato posted:the polls are disheartening i'm not gonna lie. i really did think he'd be polling first and it'd be a breeze to the nomination. I never thought he'd have it easy, because Bernie represents an existential threat to the Democratic party to an only somewhat lesser than that which he poses to the Republicans. twodot posted:I think Sanders' 28 years of experience in Congress is going to make it difficult to come across as an outsider. Not so much, when you consider how he actually spent those years.
|
# ? Sep 23, 2019 03:35 |
|
i thought it'd be easier because i thought name recognition for a multiple primary loser infamous for being birdbrained wouldn't be enough to rocket him ahead of the guy that actually took on the establishment.
|
# ? Sep 23, 2019 03:36 |
|
Captain_Maclaine posted:Not so much, when you consider how he actually spent those years. edit: Just objectively Sanders is not an outsider. He has spent 3 decades in some of the most powerful political positions, and that isn't a bad thing. It's just not true that he is running with a perspective untainted by the day to day politics of politicians. twodot fucked around with this message at 03:44 on Sep 23, 2019 |
# ? Sep 23, 2019 03:39 |
|
https://twitter.com/SinaDutaWinyan/status/1175942961730859008
|
# ? Sep 23, 2019 03:49 |
twodot posted:You can only do this if you also think that the under 30 group was correctly polled. Which you can't because the premise behind unskewing is that there was a problem polling the people you are unskewing. You can't start from a premise of "These polls are wrong" and wind up with "If I arbitrarily multiply the wrong polls by a good number I will end up with good numbers". twodot you keep going back to this argument but it's nonsense. First, pollsters are splitting their sample based on what they think is accurate. They could be wrong, and of course will be wrong because it's possible to 100% predict turn-out. The under 30 group could be correctly polled in isolation from anyone above 30. All that matters is that the other demographics within the 30 group, eg race, education, income are reflective of the broader universe within that subgroup. So long as you collect a meaningful n-size within that group it has statistical validity. The only problem arises where the polling undersamples this group so much that the MOE is large and projecting/weighting off this small sample leads to increased error. Getting the sample split wrong does not automatically mean the subgroup results are wrong. If you could take a magic poll where sample captures 90% of the universe but some subgroups are undersampled such that we only get 85% of the universe while others are 95%, the data from that 85% is still valid and highly, highly statistical significant on its own!!!
|
|
# ? Sep 23, 2019 03:50 |
|
I just had a weird thought that it's possible bernie has more Cherokee blood than warren LOL
|
# ? Sep 23, 2019 03:52 |
|
KingNastidon posted:Getting the sample split wrong does not automatically mean the subgroup results are wrong.
|
# ? Sep 23, 2019 03:52 |
|
Mellow Seas posted: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. Who cares
|
# ? Sep 23, 2019 03:53 |
|
Typo posted:I just had a weird thought that it's possible bernie has more Cherokee blood than warren LOL this is less fun than redneck nazgul's gimmick
|
# ? Sep 23, 2019 03:55 |
twodot posted:Getting the sample split wrong automatically implies that everything else is at best questionable and you need better data. Concluding "That's expensive/hard so I will just massage my bad data" is wrong and dumb. twodot, you can't be right or wrong about sample split in a prospective poll. Voter turn-out in primaries is so low that it's impossible to predict with 100% certainty. The best people have is either anchoring to historical polls or their best guess based on trends or their gut. If I told you to tell me the specific age, race, gender, income split of the Iowa/NH exit polls I can guarantee with 110% certainty you will be wrong.
|
|
# ? Sep 23, 2019 03:57 |
|
KingNastidon posted:twodot, you can't be right or wrong about sample split in a prospective poll. Voter turn-out in primaries is so low that it's possible to predict with 100% certainty. The best people have is either anchoring to historical polls or their best guess based on trends or their gut. edit: If you think a poll has under sampled a demographic it is just garbage data. If you think it is impossible to know whether a poll is under sampled, then you still can't multiply polls by nonsense to create good data. I suppose you could use that process to create fan fiction if you were inclined. twodot fucked around with this message at 04:02 on Sep 23, 2019 |
# ? Sep 23, 2019 03:58 |
twodot posted:Of course, which is yet another reason you can't take a poll that has under sampled a demographic and then multiply it by bullshit to produce better numbers. twodot do you have any higher education in mathematics or specifically statistics, probability, or experimental design? Work experience? I'm loathe to pull this card so soon after getting banned telling you not to commit suicide by cop if Warren wins, but you're just spreading nonsense and people may believe it.
|
|
# ? Sep 23, 2019 04:04 |
|
twodot posted:I think Sanders' 28 years of experience in Congress is going to make it difficult to come across as an outsider. Trump's 28 years of hobnobbing with Hillary and the like didn't make him less of an outsider.
|
# ? Sep 23, 2019 04:13 |
|
HootTheOwl posted:Trump's 28 years of hobnobbing with Hillary and the like didn't make him less of an outsider.
|
# ? Sep 23, 2019 04:14 |
|
'Not even a democrat' has probably backfired.
|
# ? Sep 23, 2019 04:16 |
|
Ogmius815 posted:I observe that there seems to be some consternation about polls ITT. I ask posters to remember the commandments of reading polls: Jesus Christ. I take it all back, gently caress polls forever and always.
|
# ? Sep 23, 2019 04:16 |
|
twodot posted:Is the number of years that Trump spent as an elected official more or less than 28? When Bernie talks about how special interests and big money have a hold of our government, he's referring to people like Trump.
|
# ? Sep 23, 2019 04:21 |
|
Pollsters are trying their best imho
|
# ? Sep 23, 2019 04:23 |
|
I give no fucks about people voting for Yang as much as his ideas and messages of unity and to get facts about him and his character straight. I know that Cory Booker is a Stanford grad, Rhodes Scholar, etc. and is probably a good person - Yang appears to be friends with him prior to the election cycle. I try my best to learn actual facts about candidates and it's loving exhausting.Majorian posted: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. 1. Yang's background is minimal in technology (some health software company for like 2 years that flopped that he described in his book showing empathy yet disdain for doctors) and his money has come from two primary sources - he sold Manhattan Prep and also Venture 4 America, a non-profit literally built on the super bleeding heart liberal Teach 4 America (my sister was a teacher in it and she is a much braver soul than I am). When he was doing V4A in 2009 he recruited founders to work improving the opportunities for people in communities hit really hard by the recession and was able to lure people away from Wall Street and Big Tech jobs. His net worth is currently estimated at about $2.1MM, which makes him among the least wealthy candidates in the primaries so far. Although he has not released his tax returns, he has expressed a desire to do so. See his funding stats at https://www.opensecrets.org/2020-presidential-race/candidate?id=N00042308 Until maybe a month or so ago he took literally $0 in PAC money, so I dunno what the $2546 is from. Bernie's campaign funds look very different to me at https://www.opensecrets.org/members-of-congress/summary?cid=N00000528&cycle=career but it doesn't make him an rear end in a top hat or anything but I do worry about money in politics even for him. Bernie used to be rather poor for a senator and what I would expect from someone who has had public service for as long without taking money from corporations - it is demoralizing when everyone around you is making tons of money just slightly compromising principles here and there while you are clipping coupons and driving a 20+ year old clunker just to make it to retirement without being in a loving double wide, and Bernie had moral superiority over someone like myself. Yang to me is somewhere in that territory but was around a lot more private sector multi-millionaires and billionaires. If you go back and find a video when he announced his candidacy, he explains why he had no choice but to run and that ultimately he's doing this for his kids. That's both selfish and selfless. Yang's references from those he mentored, his former classmates, and business colleagues are insanely immaculate for anyone successful in business. As someone that's struggled to start businesses myself, he's got my respect for doing well and having a long line of non-rear end in a top hat people that would gladly work with him again. He has managed to do all of this despite the really, really corrupt governments and cultures of many impoverished regions of the US that plague the skeletons of most politicians and businessmen that stick to one metro alone or go multi-national globalist FYGM capitalist. I am expecting some hit pieces, but Yang has a small army of those he has helped from different walks of life that will help him. If there is any businessman that deserves to be in any public office position of power I will put Yang forward instead of any of these career assholes in DC/MD/VA that pretend they put their country before themselves while relentlessly screwing over their employees and the government over an extra $500k or $1MM for themselves. I take serious issue with people that write someone off for their work or occupational background as some vague statement of their character or fitness for an office that has been evolving since its creation or that doesn't necessarily directly need technical knowledge to be effective, for example. In the same way people are prejudiced about income, race, gender, etc. I think what we have done for a living is a part of us but not definitive - Maya Angelou was briefly a sex worker and produced inspirational work, but I heard enough BS from the right wing about how awful she is just for that one stint in poverty. Yang has always been the weird guy that wanted to stick up for people but never went into public service - he is no saint, but he tries hard to be honest and to speak his mind with increasing clarity. He has straddled the lines of different opposing forces of American culture and struggled much as I have exactly like him as the son of immigrants. In this respect, Yang is fighting a lot of the same image and perception battle that I have also fought my whole life but he is now in the open. Julian Castro is another person I would vouch for in character with more political experience, but gosh is he having trouble getting anywhere polling-wise because I think he doesn't have any resonating policies (possibly about as weak as, dare I say, Biden). I will gladly listen to your concerns and try my best not to be reactive and defensive, but most of us in YangGang are fighting a lot of prejudice from the left and the right that don't hold up under any reasonable scrutiny or are at least open for productive discussion. We're not trying to be martyrs or anything IMO. We're just dumbfounded at some of the reactions moreso than acting smugly like we're better people. I was volunteering for Yang the other day and I wanted to chase some people and ask simply why they feel they do not to convince them about Yang but to understand where Yang is not getting through. In all honesty, most of us are so goddamn happy we found someone that speaks with our direct to a fault, identity politics allergic approach when nobody ever got really close we are super excited (maybe Jimmy Carter did for myself and perhaps Ross Perot). We're loving annoying, we get it. 2. Yang's intent is not to destroy safety nets but to ensure that they are so universal that no rear end in a top hat could try to prevent people from accessing it through legislation, intimidation, etc. that are para-legal means of oppression and to me against everything that non-rear end in a top hat Americans stand for. I know people that have been on welfare, disability, medicaid, unemployment, etc. and the programs are extremely difficult to get into (such as TANF) or so delayed such that people do not receive assistance in time, give up during the paperwork, and it is extremely hard to get people a specific cocktail of aid without treating it as a fulltime job, especially if you're stressed out about bills, health, etc. Less than half the people that qualify for welfare or food stamps receive them. We need to de-stigmatize and change the culture of aid, and the FYGM people also getting it is one approach (they are in some respects the minority). What made medicare and social security so politically impossible to repeal by assholes could be done with some bottom. Again, socialism / humanism is loving dead culturally and electorally in my worldview so these specific programs are probably given tombstones and if UBI is written where these assholes also benefit in some measure and start squabbling over poo poo like "well I didn't take any of that socialism dividend" in their form of virtue signalling - but their bullshit hypocrisy is clear when people like Michelle Bachman will take hand-outs for farmers and apply it to their own "farms" among others. If some redneck gets $1k / mo for guns and his truck hobby then making sure that a poor urban family doesn't get shut out because they hosed up a form or their phone service got cut off is almost worth it to me. And honestly, most rednecks are pretty decent people and identity politics' damaging effects is a nuclear arms race that liberals are losing the forest for the trees for angers me to no end. His massive policies pages are (eye rollingly, I'm quite aware) essentially every loving Ted Talk put together along with some policies that don't necessarily poll best but are aligned around improving democracy and physical / mental well-being for everyone, even if they are loving assholes. I have worked with a lot of assholes and have made a career of calming them down and finding their strengths to keep us all working together while protecting who I can. One of my career highlights now is successfully managing a team split of Trump supporters and Democrats from 2015 - 2016 in Ohio, and I saw he was going to win for the same reasons that Yang has given. I am of the firm opinion a shitload of people are assholes and drawn to the GOP because they have been conditioned like Pitbulls in a harsh, adversarial environment and that maybe some warmer connection to their community and some hope of finding a place where they are valued not to make $$$ and FYGM would make them less defensive and able to empathize better at least temporarily. I believe in Yang's notions of what the government could realistically tell people what to do with the current attitudes of people toward government as well as the political / cultural situation - banning things doesn't necessarily work (it can cause a counter gray-market effect culturally), and taxation is tricky because corporations are almost designed through accelerated evolution to avoid taxes now. When I listen to Bernie and his rhetoric through the years, I get angry and incensed. When I listen to Yang, he gives me both acknowledgement of powerlessness and positivity - I believe that kind of energy is something transformative and therapeutic when we're all so stressed out, and that feeling is similar to what I had when I was younger listening to Obama. Yang was speaking in Iowa yesterday and said if he gave the same speech in California about how your vote matters that he'd be lying - 1 Iowan is worth 1000 Californians in the end. That kind of brutal reality is what Yang is about and he usually has a solution of some sort that reaches people that take the time to parse the words out. I implore you to understand Yang as a person first and a policymaker-hopeful far second (yes, that is kind of the #1 thing as president... not disagreeing its importance and weakness / reality) and ask yourself if this man would be able to find the most effective people at least to fill in for policies where he is weak and if this man is able to bring over Trump voters at the rate he does in our broken system that democrats will be able to work across the aisle effectively when we the people need something right the gently caress now when the GOP holds basically every card except for a president. He is really likable unless you're a really conservative person that is attracted to the stoic "Presidential" archetype that I think has really hosed up the country over time because we cared more about an image than their personal character, associates, policies, and values behind the face of the MSM. Even if you don't like his policies you may be curious to understand how the same person can manage to excite figures like Ben Shapiro, Tucker Carlson, Elon Musk, Van Jones, Whoopi Goldberg, and fuckin' Al Sharpton. Trump hasn't insulted Yang on Twitter once when he's attacked all the other democratic candidates up and down the popularity charts - this is a mystery to me. Many of us interpret this as a universal appeal and charisma that the Democratic party desperately needs before the political boot is put on its face forever. Everything I'm hearing indirectly about Yang as a person is that the guy is a massive sponge for information that can summarize it out using direct language that is a big, sloppy liberal at heart like myself that just wants the country to stop stabbing itself in the face to stop the gushing wound of political division. VitalSigns posted:Have you ever worked in the private sector. I can only speak with conjecture as someone that is not studied in policy, but I have some anecdotal observations to summarize. The government is not structurally built for that kind of speed in changes IMO and would cause some unnecessary disruption because so much policy is done at the speed of human relationships, not machines. Florida's whole law having to read out everything is kind of weird but is closer to what we'd need to re-process and re-vote on old laws that we need to keep going and congressmen would need to have some auto-voting kind of setup going and one that we'd need to make transparent to the public to track their voting ahead of time to hold them to task on why they changed their mind. We absolutely do have a lot of cruft, but the layers of bureaucracy generally favor a culture of cronyism and moneyed organizations over innovation and energy. The energy in trying to fight fraud in contracting / sourcing is crippling government's ability to... govern IMO. Meanwhile, private sector is great at crunching through data and overwhelming the government with speed and people just give up and rubber stamp stuff a lot. The book The Chickenshit Club describes the reasons we fail to prosecute white collar crime anymore and I believe Yang may be one of the few to be thinking about things like this openly outside the halls of DC. Yang's proposal to create a wall of separation between corporation and state is a start based upon the reality that wealthy, famous people kind of want comfort AKA money eventually for their work. Raising the wages of public officials dramatically but lifetime bans on speaking, lobbying, etc. is a variation on clergy minus the whole vow of poverty part but may work to keep massive rear end in a top hat grifters out and paying people to not be corrupt as gently caress in our money obsessed American culture. He is not married to this like most of his 2nd tier policies because Yang will listen to someone that can show with data and drivers of behavioral economics how we could stop high-influence politicians from working for moneyed institutions instead of oh... people. So it's a lot of well defined goal posts that needs legislation designed to execute it right. Regarding shaking up government, I will directly criticize Yang's proposal to move a lot of departments and functions outside DC because the area is expensive. To parody-quote him for something much more serious "DC is expensive to run a government, but you know what's even more expensive to the public? NOT being in DC." Trump has tried to do this by moving an agency to literally Kansas and it hasn't gone well at all. This kind of a move is not like a corporate HQ move at all to me. However, Yang I believe has a proposal to modernize communications in the federal government to better support remote work and there are some deep risks and rewards here (our IT security is complete dogshit... but I work for a security software company so I may be biased). I suspect this one is a "meh, feed something to the Trump voters that sounds reasonable if you're not familiar with government." To me, a lot of how DC works is similar to how manufacturing in China is so competitive and not possible to replicate elsewhere - a lot of really connected people working in close proximity to each other because their work is too intertwined with physical presence to offshore or decouple much. Yang is fully on record saying that trying to run the government like it's a business is ridiculous because they are not the same and their goals are nothing alike - this is a complete 180 from I've come to expect even from a lot of former private industry execs getting brought into government positions for the past couple decades (an executive whose name escapes me from Disney was hired into the intelligence community way back, for example - that is what he was expected to do per Bush / Cheney / Obama it seems). necrobobsledder fucked around with this message at 04:36 on Sep 23, 2019 |
# ? Sep 23, 2019 04:28 |
|
Groovelord Neato posted:the polls are disheartening i'm not gonna lie. i really did think he'd be polling first and it'd be a breeze to the nomination. Warren is trending up! It's not all bad.
|
# ? Sep 23, 2019 04:35 |
|
Bernie has absolutely no chance of winning. First, many Dems are still mad at him for targeting Hillary last time. Second, he is a cranky blowhard who has openly declared himself a socialist, and he is too far left for the majority of the party (much less, the country). Elizabeth Warren is a much better choice in every possible way. She was absolutely right on the reasons for the last economic debacle (Bush’s), and everyone knows it. We are headed for the same kind of collapse with Trump at the helm. We need her expert knowledge as an economist and her positive energy to get the ship of state back on course
|
# ? Sep 23, 2019 04:44 |
|
I only briefly skimmed that because there is no way I'm reading more than one paragraph about the bad gimmick candidate but can you address his, to be incredibly polite, stupid as gently caress fascist plan of "move to higher ground" in response to climate change?
|
# ? Sep 23, 2019 04:47 |
|
necrobobsledder posted:Venture 4 America, a non-profit literally built on the super bleeding heart liberal Teach 4 America (my sister was a teacher in it and she is a much braver soul than I am). Hahaha wow stopped reading here. Teach For America's primary goal is to bust teacher's unions by bringing in cheap desperate labor. Scabs for America would be a more accurate title.
|
# ? Sep 23, 2019 04:50 |
|
Wicked Them Beats posted:Hahaha wow stopped reading here. I don't think Elizabeth Warren would nominate a senior educational advisor from a teachers' union busting organization. Please provide a source.
|
# ? Sep 23, 2019 04:52 |
|
necrobobsledder posted:
So you think it's good if Medicare goes away after five years unless Mitch McConnell assents to it, or decides to hold it hostage for something else or demand cuts or privatization in exchange for keeping it? Do you have parents that depend on Medicare. What about Social Security. What about Medicaid. Obamacare regulations on preexisting conditions and lifetime limits. What about the Civil Rights Acts. The Americans with Disabilities Act. The Clean Air Act. You think it would be good if Republicans could repeal any of that simply by doing nothing?
|
# ? Sep 23, 2019 04:56 |
|
It makes sense that Warren would start to attract a number of people who had supported Bernie. I view those voters as coming home to the liberal candidate. Most democrats are liberals, you see. And well they should be, because liberalism is cool and good.
|
# ? Sep 23, 2019 05:01 |
|
redneck nazgul posted:I don't think Elizabeth Warren would nominate a senior educational advisor from a teachers' union busting organization. Please provide a source. Let's take a look~ quote:When I joined Teach For America in the spring of 2011 I had no idea that my belief in social and economic justice was about to be cynically exploited by the corporate class. As a former development manager for a nonprofit that serves low-income Chicago public school students, TFA’s claims that its corps members and alumni are helping lead an educational revolution in low-income communities across the country spoke to me. Naively seduced by TFA’s do-gooder marketing pitch, I charged ahead on a mission to close the academic “ achievement gap” that TFA blames on incompetent (read unionized) teachers.
|
# ? Sep 23, 2019 05:05 |
|
Warren's paper on housing costs for Americans way back is still one of my favorite papers to cite in Internet arguments because it's pretty straightforward to read for doctoral level work while still having some good legs academically. I consider her a decently meaning person but so far her being nominated will lead to a Hillary II kind of situation politically and I'm scared shitless at the thought of a Trump without giving a poo poo about re-election anymore... unless we have a terrible recession late 2019 or in 2020. The problem I have with academic economists is similar to the reasons why Yang cites them... and shows how wrong they were about crucial things (the textbooks say that workers will retrain and move with automation and industries going away... but nobody expected at this rate of job loss and decline of industries where there's increasingly less unskilled labor work available). Yang is also correct to point out that only 17% of Americans ever go to college, so Warren and Sanders talking about student loans sounds to many Americans like we'd be the GOP talking about tax cuts and incentives on second and vacation homes (I loving poo poo you not that I overheard that having dinner in Tyson's Corner years ago as something they were trying to look into). I don't think Bernie is as crazy as Lyndon LaRouche was, but his ability to get the nomination is about as tough in my eyes as Yang because they're both so... unorthodox. I will slightly add some more context to the last economic crisis - I think that it started as far back as Clinton under Greenspan where he thinks that he may have hosed up real bad with lasting repercussions for decades. The real estate bubble to me is a derivative of money that sought new returns as a specter of the dot com bust (which I kinda don't see as a Bush thing) and with bi-partisan legislation de-regulated all these systems just to get people into more housing without understanding how asset bubbles could form with the "innovations" of Wall Street casino mathematics.
|
# ? Sep 23, 2019 05:09 |
|
necrobobsledder posted:I don't think Bernie is as crazy as Lyndon LaRouche was, but his ability to get the nomination is about as tough in my eyes as Yang because they're both so... unorthodox. You do realise this is not a weakness. Being 'orthodox' is a weakness. Trump was unorthodox.
|
# ? Sep 23, 2019 05:18 |
|
https://twitter.com/briebriejoy/status/1175987235600195590
|
# ? Sep 23, 2019 05:25 |
|
https://twitter.com/byyourlogic/status/1175990348281274368?s=21 This, but all of the candidates except Warren and Bernie
|
# ? Sep 23, 2019 05:43 |
|
Wow! That's kind of hard to reconcile with her unionizing her campaign. I assume she's come out against TFA and the dude she picked up was just there for the experience?
|
# ? Sep 23, 2019 06:04 |
|
redneck nazgul posted:Wow! If only she had actually unionized her campaign my friend.
|
# ? Sep 23, 2019 06:28 |
|
Judakel posted:If only she had actually unionized her campaign my friend. They did unionize, though?
|
# ? Sep 23, 2019 06:35 |
|
redneck nazgul posted:They did unionize, though? Not quite. Nothing has been signed and she probably still has unpaid interns doing work.
|
# ? Sep 23, 2019 06:41 |
|
Ogmius815 posted:It makes sense that Warren would start to attract a number of people who had supported Bernie. I view those voters as coming home to the liberal candidate. Most democrats are liberals, you see. And well they should be, because liberalism is cool and good. With the obvious exception of the last sentence, I think this is true for a subset of Warren supporters, though according to polling it's apparently limited to a relatively small subset of Sanders supporters. There's obviously a set of voters who would likely consider themselves "very liberal" who rejected Hillary but would prefer someone like Warren if the option is available. That being said, anecdotally this doesn't seem to have to do with an actual rejection of Sanders for being "too left-wing" and is instead based off their politics being kinda superficial and image-based, and Warren meeting the criteria of "liberal enough." My guess is that these voters are mostly older. Groovelord Neato posted:this is less fun than redneck nazgul's gimmick It's annoying because it's a malicious sort of trolling, though it's kind of funny that he ends up stumbling into correct points in a situation akin to those conservative cartoons where they're like "would you believe that these leftists want healthcare to be a human right?!" Ogmius815 posted:The entire method of the unskewing guy was he assumed, without good evidence, that all the polls were over sampling democrats and under sampling republicans. He then systematically reweighted all the polls to accord with what he assumed the electorate actually looked like. He was an idiot. Don’t be like him. You don’t have secret knowledge of how polls should be weighted. The polling models may be wrong (indeed, most of them have to be wrong since they don’t agree with one another), but speculation about how and in what direction they’ll be wrong is almost always motivated and uninformed. There's nothing wrong with projecting what results would be like if the electorate ends up taking a certain form, as long as you don't try to claim that said projection is some objective truth. vvv KingNastidon's post here actually does a better job of explaining the point I'm trying to make. Basically "unskewing" can make sense as a projection of what might happen if a certain change occurs in the electorate. KingNastidon posted:Unskewing is fine for what-if analysis, especially in the context of the primary where turn-out is less certain. It allows people that believe the sample population is wrong to understand the sensitivity to various scenarios. For example, what would the under 30 turnout have to be for Sanders to win given his advantage there and disadvantage with 65+. twodot posted:Of course, which is yet another reason you can't take a poll that has under sampled a demographic and then multiply it by bullshit to produce better numbers. You don't multiply the undersampled demographic in this situation, though; you just take a random subsample of the oversampled demographic (and end up with a higher MoE as a result, but still potentially valid statistics) Correct me if I'm wrong, but I get the impression part of this confusion is that you're making the (fairly reasonable) assumption that if age demographics are mis-sampled that you can't trust each of those individual sub-groups to be appropriately sampled with regards to other parameters. But that doesn't necessarily follow from a single demographic being over/under-sampled, even if there's a decent chance of it being the case.
|
# ? Sep 23, 2019 06:52 |
Judakel posted:People already view him as such. Probably the most common sentiment that comes up when I talk with people irl about Bernie is that he's "the only honest politician".
|
|
# ? Sep 23, 2019 06:53 |
|
|
# ? Jun 7, 2024 02:48 |
|
Ytlaya posted:You don't multiply the undersampled demographic in this situation, though; you just take a random subsample of the oversampled demographic (and end up with a higher MoE as a result, but still potentially valid statistics) edit: Like it may be the case that despite having objective evidence the poll is garbage, the problems don't extend past the demographics we already know that exist, but you still have to assume the problems affect the whole study twodot fucked around with this message at 07:03 on Sep 23, 2019 |
# ? Sep 23, 2019 07:00 |