Who do you wish to win the Democratic primaries? This poll is closed. |
|||
---|---|---|---|
Joe Biden, the Klansman | 8 | 0.91% | |
Bernie Sanders, the Hand Flailer | 578 | 65.76% | |
Elizabeth Warren, the Plan Maker | 185 | 21.05% | |
Kamala Harris, the Cop Lord | 4 | 0.46% | |
Cory Booker, the Super Hero Wannabe | 0 | 0% | |
Julian Castro, the Twin | 3 | 0.34% | |
Kirsten Gillibrand, the Franken Killer | 3 | 0.34% | |
Pete Buttigieg, the Troop Sociopath | 9 | 1.02% | |
Robert Francis O'Rourke, the Fake Latino | 2 | 0.23% | |
Jay Inslee, the Climate Alarmist | 4 | 0.46% | |
Marianne Williamson, the Crystal Queen | 19 | 2.16% | |
Andrew Yang, the $1000 Fool | 19 | 2.16% | |
Tulsi Gabbard, the Muslim Hater | 8 | 0.91% | |
Amy Klobuchar, the Comb Enthusiast | 1 | 0.11% | |
Just like in real life, nobody voted for Hickenlooper | 2 | 0.23% | |
Jeffrey Epstein, the MCC Most Hated | 9 | 1.02% | |
KKKillary KKKlinton | 16 | 1.82% | |
Some other idiot not in this list | 9 | 1.02% | |
Total: | 879 votes |
|
Tir McDohl posted:Hell yeah now that the polls show Bernie in the lead I agree with them. yes that's the point of polls
|
# ? Aug 26, 2019 19:22 |
|
|
# ? Jun 12, 2024 05:18 |
|
https://twitter.com/thehill/status/1166030050371784704 meanwhile
|
# ? Aug 26, 2019 19:22 |
|
hobotrashcanfires posted:Wow Warren took over the #1 spot and Biden #2? This is huge! You joke but... https://twitter.com/Redistrict/status/1166050274345140225
|
# ? Aug 26, 2019 19:24 |
|
Also one important thing about the poll, this is actually about how they weighted it before with Biden winning https://twitter.com/archibaldcrane/status/1166051114254340096
|
# ? Aug 26, 2019 19:25 |
|
Main Paineframe posted:Before anyone gets too excited, please note that only 298 people were asked this question in the poll, resulting in a ~6% margin of error. it's the same dem subsample size as previous monmouth national polls
|
# ? Aug 26, 2019 19:26 |
|
I mean, I wouldn't be shocked if one of the 2 did that if it looked like Biden would win otherwise. Hopefully Warren endorsing Bernie, but it would probably be whoever was in 3rd doing it
|
# ? Aug 26, 2019 19:26 |
|
Sadly it's one of the easiest predictive jokes to make after any poll. List?!
|
# ? Aug 26, 2019 19:27 |
|
Mahoning posted:I guess the question then is, why is Monmouth rated so highly by Nate? If sample size and MOE aren't being taken into account, then what is even the point of saying "These are the good polls, these are the bad ones." Based on the chart, because they do cell phone polling, they release raw data to transparency groups, and have a high long term accuracy: https://projects.fivethirtyeight.com/pollster-ratings/
|
# ? Aug 26, 2019 19:31 |
Joe 2020: his brain is normal
|
|
# ? Aug 26, 2019 19:32 |
|
goethe.cx posted:Joe 2020: his brain is normal I promise you if you vote for me we will ALL be having a normal one!!!
|
# ? Aug 26, 2019 19:35 |
I know I shouldn't read comments but those comments sure are a thing.
|
|
# ? Aug 26, 2019 19:43 |
|
Trabisnikof posted:Based on the chart, because they do cell phone polling, they release raw data to transparency groups, and have a high long term accuracy: https://projects.fivethirtyeight.com/pollster-ratings/ Right, and while those are all valuable things to look at, if the first thing the person doing the grading scale does when a poll comes out is say "don't trust it, look at the sample size and factor in the margin of error" then maybe it's NOT one of the best polls, no? Or maybe their chronic under-sampling should be factored into their grade?
|
# ? Aug 26, 2019 19:44 |
|
They are a really good pollster on average, but the chance of any given poll being a statistical outlier still exists. Taking smaller more frequent polls with really good fundamentals seems to be their MO and it's not a bad approach, but it does make outliers more possible (though even with that MOE this looks bad for Bideb) It's usually best to wait for a second poll to confirm before getting too excited no matter how good the source.
|
# ? Aug 26, 2019 19:47 |
|
Mahoning posted:I guess the question then is, why is Monmouth rated so highly by Nate? If sample size and MOE aren't being taken into account, then what is even the point of saying "These are the good polls, these are the bad ones." 538 does periodic reassessments of pollsters following elections based on their methods and how their methods turned out vs actual vote totals mind you the reassessments are roughly every year or two, since it's based off of federal or gubernatorial assessments so if a pollster suddenly poo poo the bed it's going to lag in the ratings (though you wouldn't be able to tell its making GBS threads the bed until the actual elections) Mahoning posted:Right, and while those are all valuable things to look at, if the first thing the person doing the grading scale does when a poll comes out is say "don't trust it, look at the sample size and factor in the margin of error" then maybe it's NOT one of the best polls, no? Or maybe their chronic under-sampling should be factored into their grade? see above, the grades were last computed in 2018 OJ MIST 2 THE DICK fucked around with this message at 19:52 on Aug 26, 2019 |
# ? Aug 26, 2019 19:49 |
|
VitalSigns posted:Incorrect, delegate allocation in the primary is proportional. As long as your preferred candidate gets above the 15% threshold there is no need to unite behind the lesser evil. Your analysis ignores key features of the race. For one thing, if voters unite around a candidate early, that candidate receives more delegates over time due to their larger coalition netting them a higher percentage of the vote in each state. So when they go to a floor flight at a brokered convention, they go in with more delegates to start with and are thus better positioned to get to 50% over successive rounds of voting while they argue that other delegates, including superdelegates, should support them. Momentum also matters because momentum influences voter behavior. When candidates win early contests, that makes them prospectively more likely to get a higher share of the delegates in later contests. For instance, in order to beat Biden across the preponderance of delegate-rich Super Tuesday states, Sanders or Warren will probably need to have won IA and NH. If instead Biden goes into Super Tuesday in the lead and with the momentum of all the early state victories, he'll come out of it with an almost insurmountable advantage. Splitting the vote now does help Biden, even accounting for proportional delegation. If Biden wins a plurality of states, that will very likely also mean he has a plurality of the delegates. And if it's a close overall delegate count between him and next highest candidate, he can use the "won the most states" argument to boost his electability narrative and the party hacks at the DNC will eat it right up, granting him the nom. Beating Biden requires us to be united. I'd prefer Sanders. But if Sanders really has hit his ceiling, then no, I'm not gonna say "gently caress it" and acquiesce to a Biden dystopia. I would vote Warren, who would be lacking but not apocalyptically bad. Then I'd use whatever bought time she'd give us to work to move the Overton Window left. But if we're split then Biden gets it so he can pass Republican bills and cut entitlements while the planet burns beyond the tipping point. So what then for your socialism?
|
# ? Aug 26, 2019 19:51 |
|
Mahoning posted:Right, and while those are all valuable things to look at, if the first thing the person doing the grading scale does when a poll comes out is say "don't trust it, look at the sample size and factor in the margin of error" then maybe it's NOT one of the best polls, no? Or maybe their chronic under-sampling should be factored into their grade? Something he (and most of the rest of the media) rarely if ever do on polls with similar or worse MOE or other issues so long as it's good enough for who they like and unremarkable or bad for Sanders. Strange how much noise so many of these types are making about this one, huh? Even one of the guy's running it threw a jab at "bernie stans" prior to it's release. It's almost like whats going on here is blatantly obvious and not some mystery. Almost.
|
# ? Aug 26, 2019 19:51 |
|
A 300 person sample size is not abnormal and is just fine assuming that your methodology is actually selecting a representative sample of the population. You can get a smaller MoE with a larger sample but not much smaller, a 1000 person sample only takes you to ~3% iirc And if your methodology isn't polling a representative sample then adding more people doesn't correct that bias.
|
# ? Aug 26, 2019 19:55 |
|
Mahoning posted:Right, and while those are all valuable things to look at, if the first thing the person doing the grading scale does when a poll comes out is say "don't trust it, look at the sample size and factor in the margin of error" then maybe it's NOT one of the best polls, no? Or maybe their chronic under-sampling should be factored into their grade? The significance of the moe changes based on the results. The fact that it has the three of them so close together is a big reason the moe makes this one kind of a head scratcher. You could argue that being that position is a very bad trend in it's own right. Also I didn't see what they chose for a significance level but moe shouldn't be read as it having an equal probability of being anything within the moe. Assuming they went with .05 for example means you should read those results as "there's a 95% chance that the correct value is in a standard distribution with a range +/- the moe and centered on the reported value".
|
# ? Aug 26, 2019 19:55 |
|
Any further Biden cratering/sundowning is good news for Bernie, most Biden voters are non-ideological and will switch over to Bernie if they think he has the best chance to win. (Bernie is the second choice of Biden voters according to most polling.) Keep spazzing out Biden, my man. This is the most hopeful I've been about the primary.
|
# ? Aug 26, 2019 19:57 |
|
HootTheOwl posted:There should really be a name how the beliefs of the center and the right kind of loop around like a fish fin. It’s called reactionary.
|
# ? Aug 26, 2019 20:04 |
|
To make my earlier point in a more succinct manner: there's a much higher probability for the real results to be closer to reported results than there is for them to be at the edge of the moe. A poll with that high of a moe would be be much much more statistically useful if the leaders were even a little more separated.
|
# ? Aug 26, 2019 20:04 |
|
VitalSigns posted:A 300 person sample size is not abnormal and is just fine assuming that your methodology is actually selecting a representative sample of the population. You can get a smaller MoE with a larger sample but not much smaller, a 1000 person sample only takes you to ~3% iirc The square root of the sample size should normalize MOE and thus bound it by the square foot of the sample size, so yeah that's about right (square root of 1000 is about 31, square root of 300 is about 17), you get slightly less than twice the accuracy if you increase the sample size by three times same caveats on representative samples
|
# ? Aug 26, 2019 20:07 |
|
Jarmak posted:The significance of the moe changes based on the results. The fact that it has the three of them so close together is a big reason the moe makes this one kind of a head scratcher. Also worth pointing out that Bernie and Warren polling 5 points ahead of Biden with a ~6% MoE means that it's 95% likely that Biden is anywhere from leading Bernie/Warren by 1 point to trailing them by 11. While the poll doesn't tell you who is 'really' ahead (although it is saying 'hey guys chances are it's not Biden'), it's still a remarkable change from Biden having double digit leads (usual caveats apply that this is only one poll, MoE is not the only source of polling errors, etc)
|
# ? Aug 26, 2019 20:13 |
|
Tacier posted:Practically every liberal I know says they’re not voting for Bernie because he’s said or done some things they find “problematic”, or because they just don’t like him for reasons they can’t articulate. I can never pry out more details. It’s infuriating and worrying. Against class interests is my guess.
|
# ? Aug 26, 2019 20:18 |
|
Even if that poll is an outlier, Biden still has to be hemorrhaging support for his number to drop that much. If we could get down to Bernie vs. Liz, that might actually resemble a sane-world dem primary. Also encouraging is that Kamala and Mayor Pete have been far closer to irrelevant than relevant over several polls now.
|
# ? Aug 26, 2019 20:18 |
|
gohmak posted:Against class interests is my guess. It's antisemitism.
|
# ? Aug 26, 2019 20:20 |
|
hobotrashcanfires posted:Uh 6%? Even a rough calculation makes it 6/100ths moe. Stop posting irresponsibly. 1. https://youtu.be/qf1Nbg1g7dQ 2.
|
# ? Aug 26, 2019 20:26 |
|
VitalSigns posted:Also worth pointing out that Bernie and Warren polling 5 points ahead of Biden with a ~6% MoE means that it's 95% likely that Biden is anywhere from leading Bernie/Warren by 1 point to trailing them by 11. I might be having a broke brain moment but aren't we talking about the poll that has Biden as 1% behind? But otherwise yes, the major takeaway here is Joe Biden doesn't have a clear lead anymore. Also I've been noticing for awhile now that the early primary state polls haven't been looking like the national polls... many I've seen look like this. The optimist in me would like to think that early primary states were more engaged and this is a sign of the rest of the country starting to pay attention.
|
# ? Aug 26, 2019 20:27 |
|
kinda weird everyone freaking out about this poll when last week's economist/yougov had a similar top 3 (22/19/18 biden/sanders/warren)
|
# ? Aug 26, 2019 20:30 |
|
Mahoning posted:I guess the question then is, why is Monmouth rated so highly by Nate? If sample size and MOE aren't being taken into account, then what is even the point of saying "These are the good polls, these are the bad ones." FiveThirtyEight's ratings are based primarily on historical accuracy, not the properties of their polling. Basically, they take polls of previous elections and compare them to the actual results of those elections to establish a base score, and then they add a point to that score if the pollster calls cellphones as well as landlines, and then they add a point to that score if the pollster is a member of certain pollster associations (such as the National Council on Public Polls). They don't really look at methodology at all. Then it adds a bunch of minor penalties and skews depending on how many polling firms polled that particular race. That's all. It really doesn't look at good polls vs bad polls, it looks at which pollsters have a history of ending up at the right result by election day and meet Nate's particular filtering criteria. Note that I wasn't originally saying that this poll was bad so much as I was pointing out that we're unlikely to see other polls showing similar numbers. The Monmouth poll covers a sample that's much less friendly to Biden, since their sample was nearly a third of their sample was between ages 18-35, while most other polls are closer to 10% millennials. So the Monmouth polls had Biden lower than most others from the beginning, and we're unlikely to see this being the start of a polling trend. And with the MOE, Sanders and Warren (or Sanders and Biden) could easily be a few points apart. Mainly, I just wanted to drop a reminder that "polling is bullshit, all polls are fake" doesn't go out the window the instant you see a single poll you like, so hopefully people will be a little less likely to melt down when another polling firm posts radically different numbers in a couple days.
|
# ? Aug 26, 2019 20:30 |
|
3. I'm statistically incapable of feeling shame
|
# ? Aug 26, 2019 20:32 |
|
InnercityGriot posted:Any further Biden cratering/sundowning is good news for Bernie, most Biden voters are non-ideological and will switch over to Bernie if they think he has the best chance to win. (Bernie is the second choice of Biden voters according to most polling.) Keep spazzing out Biden, my man. This is the most hopeful I've been about the primary. My guess is that, should this poll be accurate, the issue for Biden is his run of various people totally vouching for his mental state. Biden being Biden is a long, slow kill but Biden assuring everyone that his brain is totally not mush is a different animal. Blind Pineapple posted:Even if that poll is an outlier, Biden still has to be hemorrhaging support for his number to drop that much. If we could get down to Bernie vs. Liz, that might actually resemble a sane-world dem primary. Assuming that Biden implodes by Iowa, it's been looking like the race will be between Bernie and Liz for a while. Harris was in the mix for a little but she quickly dropped back down.
|
# ? Aug 26, 2019 20:34 |
|
New Jersey Primary Poll (16-20 August, Change Research): Biden 26%, Warren 23%, Sanders 21%, Buttigieg 12%, Harris 8%, Booker 5%, O'Rourke 2%
|
# ? Aug 26, 2019 20:36 |
|
This latest poll is only significant to the extent that it is another poll where Bernie is trending up (+6 in last week's economist, +2 on Morning Consult, +1 on CNN, +1 on Harris X) even as at least two separate media companies have asked him about his decline in the polls, and countless other places have speculated that he might drop out because of that. And as a general point, Bernie isn't dropping out early regardless of results. He didn't in 2016. His entire logic is of building a movement, and so he's not going to let go of campaign staff and volunteers early. As a reminder, the NY primary in 2016 happened long after it wasn't mathematically possible for him to win. But that late, losing effort set up the infrastructure for AOC to run and win.
|
# ? Aug 26, 2019 20:37 |
|
Armack posted:Your analysis ignores key features of the race. For one thing, if voters unite around a candidate early, that candidate receives more delegates over time due to their larger coalition netting them a higher percentage of the vote in each state. So when they go to a floor flight at a brokered convention, they go in with more delegates to start with and are thus better positioned to get to 50% over successive rounds of voting while they argue that other delegates, including superdelegates, should support them. I think we can agree that Bernie Sanders would endorse Warren at the convention over Biden, and that his delegates would follow him and are very unlikely to be low-info voters who have Biden as their second choice. So this only makes sense if you fear Warren might endorse Biden or her delegates might go against her endorsement, and in that case it's even more crucial to support Bernie to keep that from happening. I don't think superdelegates change the analysis. If the voting continues past the first ballot, and Warren or Bernie get the majority of elected delegates, and the DNC steals it for Biden with superdelegates it will tear the party apart and hand the election to Trump. Nobody is going to see the DNC blatantly outright steal an election and then be won over with a bunch of "well " hairsplitting about how the will of the voters doesn't count because Warren only had 30% on the first ballot instead of Biden's 40% or whatever the gently caress. If the DNC really would rather lose to Trump than win with Warren then the primary is pointless they'll just stab her in the back like they did when McGovern won. E: Nobody cares who won more states with a plurality. Nobody will accept that. If we wanted that to matter then the primary would be a first-past-the-post system with winner-take-all delegates in every state. It's not that because that's undemocratic and a proportional system better represents the will of the voters. If the DNC tries to argue that Biden should win despite a majority of the electorate preferring someone else because he would have won under the hands-down worst most undemocratic voting system ever then they just want to straight up steal the election and no amount of tactical voting is going to shame them out of it. Armack posted:Momentum also matters because momentum influences voter behavior. When candidates win early contests, that makes them prospectively more likely to get a higher share of the delegates in later contests. For instance, in order to beat Biden across the preponderance of delegate-rich Super Tuesday states, Sanders or Warren will probably need to have won IA and NH. If instead Biden goes into Super Tuesday in the lead and with the momentum of all the early state victories, he'll come out of it with an almost insurmountable advantage. This analysis favors uniting behind Sanders who is stronger than Warren in every pre Super Tuesday state except maybe Iowa which is close. He consistently leads in NH, then Nevada and South Carolina have large Hispanic and African American populations where Warren underperforms, then its Super Tuesday. Most SC polls have Warren underneath the 15% threshold. Also I don't live in any of those states so it doesn't matter who I'm supporting until Texas votes on Super Tuesday and even from a purely tactical standpoint most polls don't even have Warren breaking 15% in Texas meaning that if I switched to her my vote will probably be wasted, so if they want to defeat Biden Warren supporters should unite behind the only non-Biden candidate consistently polling above 15%. VitalSigns fucked around with this message at 21:15 on Aug 26, 2019 |
# ? Aug 26, 2019 20:40 |
|
Gyges posted:Assuming that Biden implodes by Iowa, it's been looking like the race will be between Bernie and Liz for a while. Harris was in the mix for a little but she quickly dropped back down. i've been curious if harris actually can retain a 4th (or 3rd if biden collapses) position up to iowa or if some other b-tier ends up filling the slot six months down the road
|
# ? Aug 26, 2019 20:42 |
|
https://twitter.com/GarbageApe/status/1166070927307788290
|
# ? Aug 26, 2019 20:52 |
|
Main Paineframe posted:
yougov had a poll in field at the same time with similar results, more than twice the sample size, and an unweighted sample pretty close to the 2016 primary electorate
|
# ? Aug 26, 2019 20:56 |
|
Jarmak posted:I might be having a broke brain moment but aren't we talking about the poll that has Biden as 1% behind? No you're right I misread the tweet Jarmak posted:But otherwise yes, the major takeaway here is Joe Biden doesn't have a clear lead anymore. Yeah this is still true, I was wrong to say that he is almost definitely behind, but the poll is still a notable change. Also with them all within one point, it's unlikely that a smaller MoE would actually help much at distinguishing them because sampling error is a normal distribution. Assuming it's a 95% confidence interval, the +/- 6 points MoE is 2 standard deviations. Which means for a poll with a +/- 3 point MoE to distinguish them the 'real' numbers would need to be outside of 1 standard deviation in Monmouth's poll. But there's a 68% chance that these results are within 1 standard deviation. So it's still more likely than not that a 1000 person poll would put them within MoE of each other anyway. It'd be a little more helpful but probably not a slam dunk unless this poll really is an outlier. (E: well since they are separated by 1 point the chances are less than 68% by some amount but I don't think that would be enough to overturn my point) But I'm not really a stats nerd so like ya know, don't believe me without checking my work.
|
# ? Aug 26, 2019 21:00 |
|
|
# ? Jun 12, 2024 05:18 |
|
Eh, I get the feeling most PSA guys support Warren at this point By the way, I've been really impressed with Bernie's policy output recently.
|
# ? Aug 26, 2019 21:17 |