-
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
Trust the Polls
Yeah. Their model, like most, is entirely Poll based. Nothing you can really do about that at this point. Maybe something for forecasters to consider going forward.
CoffeeJanitor
Member
I doubt Harry and Nate agree on a broad level based on Harry's bullish analysis
gutter_trash
Banned
day 2 of CNN milking their crappy CNN/ORC polls
i dont believe those Nevada numbers one bit
i dont believe those Nevada numbers one bit
BenjaminBirdie
Banned
Poker Face
Member
The "they're both the same I won't vote for either" people are the worst.
Particle Physicist
between a quark and a baryon
He said some bullshit statement that they did not find evidence that there was a breach but that doesn't mean there wasn't!
Ah, so it's possible they found something, but it would be indirect evidence, and it would be from Weiner's computer. I'd fucking die laughing if Weiner guessed email passwords and downloaded everything to his PC.
Edit: But even that doesn't explain the "hacked by at least five foreign intelligence agencies". So, genuinely is just fucking with people, because they're annoyed they didn't get to go on a FBI-sponsored witch-hunt against Hillary/Clinton Foundation.
cartoon_soldier
Member
Enten can't change Nate's model. But he can drop truthbombs on twitter.
balladofwindfishes
Member
We know they aren't right. We have early voting. We can easily see their sample was terrible with Hispanics and young people.
They're going to look ridiculous next week when she wins the state by several points.
banefirelord
Banned
What does the past correlation between early vote numbers and Election Day results generally look like?
Actually reading that Fox article shows what a sensationalized headline it is. First off as others have said it's Fox and other more reputable outlets like WSJ are telling a different story. But the big thing is that it doesn't say they're heading for an indictment, it says that there may be push for that if there really is pay for play.
Poodlestrike
Banned
Aye. And Polls+ is basically entirely broken.
Accounting for early voting returns seems like a pretty easy step, though. Should definitely be in place for next time.
Honestly, I still have faith in the 538 model, it's just that it's getting fed garbage inputs; bad polls weighted heavily in the interest of hyping the horserace. Aggregates are only as good as the polls you feed into them.
User 463088
Banned
Wait, that was the NH poll that was BOMBSHELL going to get everyone talking!
Abinash117
Member
Sometimes certain polls ask who they voted for the poll from TargetSmart had her leading early votes. They interviewed who they voted for and they verified if they actually voted. The large lead lines up with other pollsters who polled people about the same thing. The totals seen in Dems vs GOP vs NP is the number of people who voted early that is it. It can be seen here http://www.electproject.org/early_2016 .
Now everyone will talk about Wikileaks and huma going to jailWait, that was the NH poll that was BOMBSHELL going to get everyone talking!
I don't know what or who to trust anymore because I hear so many conflicting things. It's contributing to me being an anxious wreck around the clock and I hate it.
Is Hillary still poised to win, or are we just dabbling in the "poll unskewing" that led the Republicans astray last time? I almost wish I could enter a medically induced coma until next Wednesday and just take the final result full force. The uncertainty and the waiting are killing me.
User 463088
Banned
FUCK.
What if Huma IS doing the cyber on that laptop and she IS Wikileaks.
balladofwindfishes
Member
No unskewing. Polls show her with a decisive win.
The electoral college doesn't add up for Trump.
TestOfTide
Banned
Wait, that was the NH poll that was BOMBSHELL going to get everyone talking!
Well what do you expect when the Trump himself sees an AZ poll with him ahead by only 1 point and thinks "Yes!! I'm winning!!! I gotta show this to people on Twitter."
What if Anthony Weiner was actually sexting Julian Assange?
newshinycd
Member
balladofwindfishes
Member
The Trumpers are going to cry rigged when she wins Nevada by 5+ points, and point to the junk CNN poll as proof.
Ugh
Ugh
Abinash117
Member
Still today and tomorrow. As long as the Dems turnout more early voters in many states they will have a huge cushion.
RustyNails
Member
You guys seriously need to follow @yottapoint. Cuts straight through all this haze of bullshit created by the rightwing regarding Clintons. And he doesn't even work for them.
User 463088
Banned
I think Julian may be too old for Anthony.
Weiner's phone was actually Hillary's phone> Cause after she had one of her lesbian seizure parties where she was basically dying, she decided to get on Kik and look at some pictures. But then she recognized the sheets from Weiner's house (because she's secretly having an affair with Huma, duh) and was like OH MY GOD. So she took the bleach to the phone. I mean, not her, obviously. She was already in her medically induced coma, because she's like 93% dead already. (could be more or less when we adjust for TRENDZ) And then she texted her friend Harambe, but he was going to leak to the failing NYT, so she invoked Brexit and then had him murdered by the League of Assassins. And something about PAY TO PLAY
The notion that the FBI is going to swoop in and save the day for Trump is little more than a Republican fantasy. The Clinton Foundation has already been probed ad nauseam and no wrongdoing was found, so maybe it really is a respectable charity! They didn't just happen to do the right thing 99% of the time, oh except for that one indictment-worthy pay-for-play incident that we're somehow just finding out now. Contrast it with Trump's "charity" which had a whole laundry list of shady dealings.
Same thing with Weiner's emails. We've already established that Clinton is not especially computer-savvy and she relied on the judgement of others for the proper handling of classified materials. So the idea that Weiner somehow wound up with the emails that are going to change everything just isn't plausible.
The election will go on, Clinton will win, and Comey will announce that they found nothing significant sometime afterward.
Same thing with Weiner's emails. We've already established that Clinton is not especially computer-savvy and she relied on the judgement of others for the proper handling of classified materials. So the idea that Weiner somehow wound up with the emails that are going to change everything just isn't plausible.
The election will go on, Clinton will win, and Comey will announce that they found nothing significant sometime afterward.
I get the vibe that Harry Enten is growing skeptical of the 538 model. He's been calling out outlier and bad polls on Twitter more, deviating from the group in podcasts and conversations more and then he wrote that article about Nevada early.
Obviously these guys aren't dumb and have to recognize how odd some of the movements in their model are. Like the near 2% drop because of a Missouri poll last night.
Obviously these guys aren't dumb and have to recognize how odd some of the movements in their model are. Like the near 2% drop because of a Missouri poll last night.
How about a quick chuckle? Has this Trey Gowdy video been posted? Ya gotta love the stupid tag at the end...
https://www.facebook.com/gowdyforcongress/videos/1223257727697160/
https://www.facebook.com/gowdyforcongress/videos/1223257727697160/
Actually, doesn't his model allow for systemic polling error both ways? Aren't his odds for a Clinton blowout higher than other models?
D
Deleted member 231381
Unconfirmed Member
They are, yes. His upper bound for Clinton's performance is significantly higher than Wang's.
balladofwindfishes
Member
Now are those real polls or more ripped from their 50-state sample?
So cam someone help me understand why some polls are so different?
I mean the Princeton poll gives Hillary 98% chance to win, but Nate Silver only gives her a 67 percent chance now. In CNN she is only up 4 points over all? So why the difference and which one is more correct?
I mean the Princeton poll gives Hillary 98% chance to win, but Nate Silver only gives her a 67 percent chance now. In CNN she is only up 4 points over all? So why the difference and which one is more correct?
Is that from a 50 state megasample type poll?
Joe Shlabotnik
Banned
Hope those are actual battleground polls and not part of a junky 50-state run.
edit: hahaha
User 463088
Banned
There are two NH polls coming today. The one this morning shows Trump +1. I'm assuming the other is a Hillary lead. The Trump +1 had a 4.5% MoE fwiw.
TestOfTide
Banned
Does Nate own 538 or could all the reasonable people kick him off the team?
Like it's not even subtle anymore, Nate it at peak "HORSE RACE" bullshit. I don't know if it's because he doesn't want to admit to wanting Trump to win or if he just has PTSD from making a fool of himself during the primaries, but someone needs to intervene and tell him that he needs to accept the fact that Trump won the primaries and Hillary will win this Tuesday.
balladofwindfishes
Member
No. Just Fox News being Fox News and this thread being this thread
Stop it
Haha!
D
Deleted member 231381
Unconfirmed Member
Princeton and 538 aren't polls. They're models. They use polls as data points to create a prediction of the election result. Both Princeton and 538 have approximately the same prediction - 538's Nowcast has Clinton leading the popular vote by 3%, Princeton's Snapshot has Clinton leading the meta-margin (which for them is functionally equivalent to the popular vote) by 3.1%. However, 538 incorporates a wider number of variables, which produces a greater degree of uncertainty; Princeton uses a relatively small number, which produces a smaller degree of uncertainty. So 538 thinks that right now Clinton is 3% ahead, but things could still change a reasonable amount; Princeton thinks right now Clinton is 3.1% ahead and that is de facto locked in.
The CNN poll is within the margin of error of both of the current snapshot predictions from 538 and Princeton and is entirely consistent with both.
- Status
Similar threads
- Locked
- Locked
- Locked
- Locked