Manmademan
Member
Polls don't matter
-
Hey, guest user. Hope you're enjoying NeoGAF! Have you considered registering for an account? Come join us and add your take to the daily discourse.
PoliGAF 2016 |OT14| Attention NV shoppers, democracy is on sale in aisle 4!
- Thread starter eBay Huckster
- Start date
- Status
Polls don't matter
Godzilla's Left Toe
Member
Seems like it
banefirelord
Banned
I say yes. I'm happy to eat crow, but it makes very little tactical sense to leave it this late since early voting has been going on for weeks.
excelsiorlef
Member
Fragamemnon
Member
I think Hillary knows about how explosive the Latino break is for her ( this is reinforced by what we're seeing out of early voting in TX, FL, and VA) and AZ is the most likely win-more state in the most common scenarios. Makes perfect sense.
GA is part of the long term mid-atlantic project where Democrats fuse African-Americans, other minorities with college white urban and core suburban voters who don't share GOP values. See VA 08/12 and NC 08/12/16 for the model of what's in store for GA in the future.
GA is part of the long term mid-atlantic project where Democrats fuse African-Americans, other minorities with college white urban and core suburban voters who don't share GOP values. See VA 08/12 and NC 08/12/16 for the model of what's in store for GA in the future.
The thing about GA, is that it depends on historic AA turnout and white-college transplants to Atlanta. With AA at best being at 2012 levels, the Clinton campaign is guessing that there's more hope in turning out Latinos in AZ than AA in GA.
If GA goes there way, it would be icing on the cake, but not as important or plausible as AZ.
I think that's accurate. I was born in 81 and don't identify too well with the mid 90s kids. I think the cutoff should be mid to later 80s, not early 80s.
1. Just because a methodology is made public does not mean it is valid or correct. See LA-Time tracking poll.
2. Do we hate models now? Because it's basically impossible to get an accurate and representative simple random sample right now, so every poll taken has some level of bias in its results. Even an aggregator that just reports those results is going to show you some level of bias. We just don't know where that tilts yet.
3. Past success does not always match future success. Better to have more information and consider the wide range of viewpoints out there.
The dynamics of the race point to something more dropping about Trump, but we should just forget about it and work with the assumption that there's not going to be anything else to drop.
I've seen people over 35 who are very Millennial like, but yeah, maybe early 80s is too much of a stretch to be called Millennial. Then again, maybe not. Not all Baby Boomers or Gen Xers are alike also.
Everything I've read from Nate about his model's behavior makes perfect sense to me. He's trying to account for the inherent uncertainty that exists in all elections, and his model is more concerned with the various unpredictable elements than most of the others.There are factors that make this election more unpredictable than most, and his model is reacting to them, as it is meant to.
I can absolutely understand why the big increase in undecideds and 3rd party voters this time makes this election more of a tossup than usual. And if there are national trends, they are likely to be reflected in all states, whether polling exists or not.
Election day results being off by 3 or 4 percentage points from the polling, and going in Trump's favor doesn't seem impossible to me.
I honestly think they just would like to try and help the AZ Senate race where the GA one has even less of a chance of happening. Neither will, but there's more on the table in Arizona.
Manmademan
Member
AZ has a potentially competitive Senate race. Georgia doesn't.
BotoxAgent
Member
I agree with this. Clinton's win is guaranteed at this point....gotta ensure the Senate is safe too.
newshinycd
Member
Aww yiss. Georgia goin' blue this year y'all. No way Aleppo + Wifi goes over 5% here.
ItWasMeantToBe19
Banned
I mean, we have an insider in this thread who just posted that it's still coming.
So we'll see!
eBay Huckster
Member
at least we'll know why PEC's methodology was or wasn't correct, because it isn't hidden from public view.
do we not hate unskewing polls? because that's literally what 538 does above and beyond the huffpost aggregate (between the weighting and national trend adjustments), and that arguably introduces more bias than just reporting the polls as is.
then why the hell does everyone piss themselves over one viewpoint's wild swings?
don't answer that, i already know what the answer is
McCain's seat? Yeah that's not happening.
TheCochese
Banned
Ahh, you say that, but Isakson doesn't have very favorable numbers. 57% or so the last I saw?
We're doing work here fam.
Carpe Libertatem
Member
Was gonna say, probably Kirksville. I'd imagine the students at Truman would be more likely to lean liberal, too. At least compared to Missouri State, SEMO, etc
Hooligan69
Member
1975 checking in. Tweenial i guess. We're the generation that, like a middle child, kinda gets ignored and may just get away with more shit because everybody's worried about the oldest and youngest. I say, instead of beating up the younger generations, let's go after baby boomers. They had the chance to usher in real, progressive change, but choked hard in the 70's, fucking us all and creating what mess we're in now. (not all but most).
I'm sorry is there a reason we're supposed to be panicking
I thought EV was looking pretty good lately
The methodology is explained on their site and they comment on it all the time.
Properly correcting for objective biases and methodological differences in polls is good. "Unskewing" by the republican campaigns in the past wasn't evidence based or rational. That's bad.
whyamihere
Banned
I mean, I know that I'm going a little nuts, but the early vote in the states that has it is generally great. That's why I said I was having a bit of an irrational panic. You agreed!
MadLaughter
Member
Imagine the scene in Stranger Things Where the guy goes through the opening with the cable around his waist.
I feel like this guy. I need people to pull the rope. Get me out of this place where I can't convince myself that Trump is not an agent of Putin (knowingly or not) who is undermining the stability of our democracy.
I feel like this guy. I need people to pull the rope. Get me out of this place where I can't convince myself that Trump is not an agent of Putin (knowingly or not) who is undermining the stability of our democracy.
A Link to the Past
Banned
I mean, not to suggest that McCain is enfeebled, but his seat is not going to be his for very long; there's a very real opportunity for Democrats to push a competitor for his seat against whomever the GOP pushes as McCain's replacement.
newshinycd
Member
Yea I'd like Barksdale to win here in GA too. Everyone likes Isakson for whatever fucking reason.
Dubbedinenglish
Banned
Runoff rules in GA. If the stupid hat doesn't get 50%, it triggers a runoff after the election.
metalslimer
Member
Is there a reason we shouldn't be?
TheCochese
Banned
1980 is an accepted cutoff. 77 is on the short end.
Bam Bam Baklava
Member
Other than Nevada and maybe CO, I have no idea wtf is going on. One person says black turnout is up, the other says it's down and on and on.
Fragamemnon
Member
Speaking as a 77'er, I can say that I've always felt more gen-x. I feel the defining breakpoint for the two generations is that one graduated college without a lot of debt and was able to get jobs at the end of the Clinton era pretty easily.
Millennials graduated with much, much more student debt and had to really fight their way to get good jobs. It's those things that define the generations, in my opinion.
Millennials graduated with much, much more student debt and had to really fight their way to get good jobs. It's those things that define the generations, in my opinion.
TheCochese
Banned
And the runoff would likely go sharp R.
Isakson looks like he's going to drop dead any day.
I'm not comfortable with anything, and I would rather be told right now that Trump is going to win outright than to continue reading EV tea leaves and polls, but even AA turnout, the lone red alert in the early vote, is looking good again.
True.
A Link to the Past
Banned
Pretty much, yeah. How able would she be about getting immigration reform through?
whyamihere
Banned
Think of it as black turnout being up 3-1.
I think AZ senate could end up closer than expected but Kirkpatrick win is still a long shot.
She needs to push for it... depending on Senate/House results though it might not be easy to pull it off. GOP is already prepared for 100% obstruction and they will hold House.
She needs to push for it... depending on Senate/House results though it might not be easy to pull it off. GOP is already prepared for 100% obstruction and they will hold House.
Laevateinn
Member
How the fuck is she going to get that done without the House?
I think it is weird that 538 adjusts for the trendline in their Now-cast. But considering that their other models are trying to make a prediction about election day and not just reflect the immediate mood of the electorate, then it makes a lot of sense that they would create a methodology that makes adjustments to polls based on historical data of previous elections and pollster bias. Republicans arbitrarily giving extra points to Trump is problematic. But the problem is the method used.
I'm not. We see a trend of undecideds towards Trump. There could be and probably is a cap on how far that trend goes. But the trend exists and we can't ignore it. 538 is reacting more quickly than other methodologies on this trend which may or may not be a flaw. I just wouldn't dismiss it out of hand nor would I believe things are happening just because 538 is saying it is.
TheCochese
Banned
You going to kill some people in here.
eBay Huckster
Member
it's explained, sure - it's still not transparent. especially not in comparison to PEC, which flat-out has its MATLAB script (and statistical priors, such as pre-election drift and the meta-margin's standard deviation) publicly available.
it's good, particularly in comparison to dean chambers! but it still, ironically, introduces bias into the equation in a way that no one else in the aggregation game does.
e: to be clear to both of you - i'm not saying we should discount 538. ironically (given the contest of yr last post, Zealous) i'm saying we should pay equal attention to everything else in case the trends don't actually show up elsewhere. because they do tend to show up in PEC/DKE, just significantly more muted.
Clinton is much more likely to accept incremental change than Obama did with the ACA.
If she doesn't see a path for comprehensive immigration reform, she'll salvage what she can, like getting DAPA and DACA into law, at the very least, and maybe a path to citizenship in a latter bill.
minor effort
Banned
1980 here. I feel closer to Generation X to be honest, even if some definitions of Millennials put 1980 in the definition.
Growing up with internet ubiquity seems to be the most immediate defining factor of Millennials. My first experience with the internet wasn't until 1995 or so, by which point I was already in high school, so I still have clear memories of not even knowing the existence of the internet. By the time I graduated in 1999 and was legally an adult, the internet was still relatively limited and isolated compared to today.
Those experiences seem largely different than someone born in 1995, never having known a world without the internet, and having a smartphone on hand from the age of 12 onward.
Growing up with internet ubiquity seems to be the most immediate defining factor of Millennials. My first experience with the internet wasn't until 1995 or so, by which point I was already in high school, so I still have clear memories of not even knowing the existence of the internet. By the time I graduated in 1999 and was legally an adult, the internet was still relatively limited and isolated compared to today.
Those experiences seem largely different than someone born in 1995, never having known a world without the internet, and having a smartphone on hand from the age of 12 onward.
Manmademan
Member
There was a bullshit article politico put out that had black voter turnout in FL at an abysmal 55% of their 2012 pace. The organization that put those numbers out wasn't even a pollster, but a conservative business group out of florida.
As it turns out, that number was completely fabricated. Black voter turnout in FL is up 33% over 2012.
That article came out around the same time other outlets were looking at the black voter turnout in NC, which was down due to the state shutting down early voting locations in NC and other voter suppression tactics.
Lazy journalists took those two things and ran with "black voters aren't going to polls!" as headlines for a couple of days, which lines up well with the diet racist conservative mantra that Clinton won't have good black support because she's not black herself.
from what we've seen of actual early voter data, that one isn't true at all.
- Status
Similar threads
- Locked
- Locked
- Locked
- Locked