• 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!

Status
Not open for further replies.
In a state or overall? If it's overall - look at the states. If it's a state, look at other states. If you fall into the 10% range in too many states then start looking at your assumptions.

Overall, not a state. You keep harping on states but there aren't enough states out there. Even though there are 50 states, the election is really about less than 10 states. The polling is too sparse in places like Montana to matter.

Regardless, it doesn't matter. The solution isn't to look at the states because that says nothing. The 10% chance of a blowout prices in all the swing states, already. If she's blowing out in Fla, Pa, and Wis, then is makes sense she's do the same in NH, Nv, and possibly flip Zona. ETC.

Where you seem to be mistaken is that if she has a 10% chance of an election blowout that she is hitting the 10% blowout in each state. Not true! Not how the math works!

We also have more data to use than just the overall or even state results. What was the voter breakdown? Did our assumptions about how to weight polls based on demographics hold up when compared to who actually voted? Which polls were accurate and what did they use as a voter model? Are there consistent assumptions made across successful or more successful polls of a given place? Did we correctly identify those? There is no lack of data or things to evaluate using that data.

Wrong! We don't know the electorate, definitively. Hence why there's a discrepency between CPS and exit polling. How do we verify? Unless every state starts taking down demographic data for every vote that is cast, it's all just good guesswork.

And besides, even if we could determine it, it does nothing to proving the model right or wrong. If the polls are wrong because of demographic data or whatnot, that means the poll inputs in Nate's model was wrong, not the model. If the polls had the correct weighting, his model might be right. Or wrong, even.

You are talking about looking at the pollsters and their accuracy, not Nate's model. Nate's model, hell every one of these models, depend on the pollsters being somewhat accurate so no fault can lie with them if the polls are wrong.


Nate's model doesn't make demographic assumption or any of that stuff. Nate Cohn's NC model is doing that, but that's different.

Nate's inputs, besides polls, are things like incumbancy, trend, economic data like UE, Jobs, etc, and other things. None of which can be tested in the way you'd like.


Do you understand my criticism here? You're making the same mistake Nate Silver made during the World Cup when he thought his model failed.

When something unlikely happens, it doesn't mean your model is wrong. Your model is saying this unlikely event can happen! In fact, it's saying IT WILL HAPPEN sometimes. If you adjust your model because said unlikely event happened, it means you don't understand the math.

What you're basically arguing is that if I flip a coin 5 times and all 5 times it comes up heads, then we must conclude it's a weighted coin.
 
Why the obsession over oppo?

She was looking to sell her candidacy on its own merits and was going for the upbeat/positive messaging before Comey Sneak Fuck dropped a huge mess on their laps on Friday. We're finally turning the corner on the media narrative, and this is time for the campaign to get back on message instead of waiting for some miracle panacea.

The election is drawning upon us, there is no corner to be turned when it comes to changing the tone.
 

Revolver

Member
I guess Katy Tur better brace for another round of hatred from the deplorables. In her report on Nightly News about Melania's speech she said it was in stark contrast to her husband's campaign which has been built on personal insults accompanied by a reel of his name calling. Then went on to say that the FBI probe into the Clinton Foundation hasn't moved an inch in months. #ImWithTur
 
But that's assuming the conclusion that the 538 model is less accurate in this election. We'll understand much better in like 5 days.

Hell, even if it's more accurate, it'd be nice to get a fix on what is making it more accurate.

But, again, five days!

Actually six, because the polls will still be open at 7 PM on Tuesday
 

Oblivion

Fetishing muscular manly men in skintight hosery
So Hillary has high support among FL Hispanics but Murphy doesn't? What the hell did he do them?
 
Because that's a bad margin when white people heavily vote GOP.

Really now. out of curiosity, where do you think Trump's favorability is with white voters?

Where was Romney's in 2012?

What do you think a spiking latino population does to the margin you need to win white voters by to win the state?
 

DrForester

Kills Photobucket
There is no plan B if Florida is gone.

Pretty much. This is the only "possible" map without Florida for Trump that I can think of

FUFr4s6.png


And Colorado and Pennsylvania both seem like they're well out of reach for him.
 

witness

Member
With the PR vote in Florida I'm expecting a big victory here for Hillary. I'm gonna go with FL being called BEFORE the west coast polls close.
 
This election is over, and has been over for a long, long time.

The Clinton campaign, you know, knows what it's doing. They know where to go. They have endless options of where to send people. They have internals that are run by the same crew that landed Obama a 4 point victory when public polling was either at a tie, slim Obama win, or simply Romney victory.

Do you see the Clinton campaign freaking out because New Hampshire public polling is showing a tight race? Does the Clinton campaign look like a campaign that is freaking out at anything right now? Donald Trump is the nominee they are facing, the understand that if he has a sliver of actual hope of being POTUS it would be GG for the Western world. Does it look like they are reacting that he actually has a chance? They aren't, because he doesn't.

The same thing that happened in 2012 is about to happen now. We're going to wake up on November 9th with a margin of victory many points ahead of current national polling, because national public polling is trash that is being used to drive up click bait or news ratings. What's the point of even looking at this shit if we have to deal with CNN polls that can't even get a proper number of hispanics in a fucking NV poll? Or a proper sampling of under 35 voters? It's insane, and it's so blatant that half of these polls are trash and thrown out into the wild because, shit, we got to talk about something!

Anyways, that's my rant. I'm tired of this oppo-droppo shit. If you guys haven't noticed, Trump is literally a horrible human being and we just spent the last six months going over how fucking horrible he is. It just turns out that as long as you slap an "R" next to someones name, a large portion of the population will vote for them, even if they are racist, sexists shit heads who are literally pawns of a foreign government that is our main global adversary.
 
Really now. out of curiosity, where do you think Trump's favorability is with white voters?

Where was Romney's in 2012?

What do you think a spiking latino population does to the margin you need to win white voters by to win the state?

Alright, so how many points does Murphy need to be up among Latinos in order to offset Rubio's margins with white voters, in your estimation?
 

NeoXChaos

Member
Really now. out of curiosity, where do you think Trump's favorability is with white voters?

Where was Romney's in 2012?

What do you think a spiking latino population does to the margin you need to win white voters by to win the state?

You still need about 35% of the white vote in the South to win statewide and that escpially true in FL and is the only Southern(soon joining NC) state Democrats manage to do that however Latinos in FL are equivalent to AA's in the rest of the South. Meaning you have to keep your base of those groups to win.

Murphy 46-40 performance is like JBE getting the 35-40% white share he did get last year but Vitter got 46-40 of the AA vote. JBE would have lost with such a result

Murphy's not getting a higher percentage than 35%(white share) so he has to compensate that with the Latino share of which Hillary is getting and he is not. So unless he outruns her with white voters by 10% points(guessing) or suddenly miror Hillary in Hispanics he is toast
 
Barring another big event, there really is just one more news cycle left to actually change minds and that's tomorrow. The weekend is limited at best (unless something BREAKING occurs) and Monday just feels too late in the game. We may be in the last meaningful 24 hours of the news soon. Nicole Wallace said more or less the same thing earlier this week.
 

Miles X

Member
This election is over, and has been over for a long, long time.

The Clinton campaign, you know, knows what it's doing. They know where to go. They have endless options of where to send people. They have internals that are run by the same crew that landed Obama a 4 point victory when public polling was either at a tie, slim Obama win, or simply Romney victory.

Do you see the Clinton campaign freaking out because New Hampshire public polling is showing a tight race? Does the Clinton campaign look like a campaign that is freaking out at anything right now? Donald Trump is the nominee they are facing, the understand that if he has a sliver of actual hope of being POTUS it would be GG for the Western world. Does it look like they are reacting that he actually has a chance? They aren't, because he doesn't.

The same thing that happened in 2012 is about to happen now. We're going to wake up on November 9th with a margin of victory many points ahead of current national polling, because national public polling is trash that is being used to drive up click bait or news ratings. What's the point of even looking at this shit if we have to deal with CNN polls that can't even get a proper number of hispanics in a fucking NV poll? Or a proper sampling of under 35 voters? It's insane, and it's so blatant that half of these polls are trash and thrown out into the wild because, shit, we got to talk about something!

Anyways, that's my rant. I'm tired of this oppo-droppo shit. If you guys haven't noticed, Trump is literally a horrible human being and we just spent the last six months going over how fucking horrible he is. It just turns out that as long as you slap an "R" next to someones name, a large portion of the population will vote for them, even if they are racist, sexists shit heads who are literally pawns of a foreign government that is our main global adversary.

Nice post, I'm hoping she still has a shot at Georgia, Iowa, Ohio and within 2 points in Texas just for personal pleasure.

Everything I'm hearing from EV's is good news, it's just polls I'm seeing 'good' Trump news.
 
I really really doubt the networks will call Fla or even NC before the West Coast closes.

Reason being if you call either, especially the former, the election is already over and that could affect voting out West. I think the media will hold off just long enough. Maybe call it 5 mins before Cali closes.

Assuming they even can call it by then.
 

johnsmith

remember me
Why the obsession over oppo?

She was looking to sell her candidacy on its own merits and was going for the upbeat/positive messaging before Comey Sneak Fuck dropped a huge mess on their laps on Friday. We're finally turning the corner on the media narrative, and this is time for the campaign to get back on message instead of waiting for some miracle panacea.

Devastating oppo would help downballot. Also Trump deserves to have his life destroyed.
 

thebloo

Member
Why the obsession over oppo?

She was looking to sell her candidacy on its own merits and was going for the upbeat/positive messaging before Comey Sneak Fuck dropped a huge mess on their laps on Friday. We're finally turning the corner on the media narrative, and this is time for the campaign to get back on message instead of waiting for some miracle panacea.

The campaign doesn't care about the oppo. It's just us, here. The campaign either has it or it doesn't. Either way, they'll play their game till the end, like they've been doing for a year.

If something drops, so be it. If it doesn't, this is a campaign with a kickass infrastructure that will win either way.
 
This election is over, and has been over for a long, long time.

The Clinton campaign, you know, knows what it's doing. They know where to go. They have endless options of where to send people. They have internals that are run by the same crew that landed Obama a 4 point victory when public polling was either at a tie, slim Obama win, or simply Romney victory.

Do you see the Clinton campaign freaking out because New Hampshire public polling is showing a tight race? Does the Clinton campaign look like a campaign that is freaking out at anything right now? Donald Trump is the nominee they are facing, the understand that if he has a sliver of actual hope of being POTUS it would be GG for the Western world. Does it look like they are reacting that he actually has a chance? They aren't, because he doesn't.

The same thing that happened in 2012 is about to happen now. We're going to wake up on November 9th with a margin of victory many points ahead of current national polling, because national public polling is trash that is being used to drive up click bait or news ratings. What's the point of even looking at this shit if we have to deal with CNN polls that can't even get a proper number of hispanics in a fucking NV poll? Or a proper sampling of under 35 voters? It's insane, and it's so blatant that half of these polls are trash and thrown out into the wild because, shit, we got to talk about something!

Anyways, that's my rant. I'm tired of this oppo-droppo shit. If you guys haven't noticed, Trump is literally a horrible human being and we just spent the last six months going over how fucking horrible he is. It just turns out that as long as you slap an "R" next to someones name, a large portion of the population will vote for them, even if they are racist, sexists shit heads who are literally pawns of a foreign government that is our main global adversary.

Yeah. Phew. Yeah.
 
This election is over, and has been over for a long, long time.

The Clinton campaign, you know, knows what it's doing. They know where to go. They have endless options of where to send people. They have internals that are run by the same crew that landed Obama a 4 point victory when public polling was either at a tie, slim Obama win, or simply Romney victory.

Do you see the Clinton campaign freaking out because New Hampshire public polling is showing a tight race? Does the Clinton campaign look like a campaign that is freaking out at anything right now? Donald Trump is the nominee they are facing, the understand that if he has a sliver of actual hope of being POTUS it would be GG for the Western world. Does it look like they are reacting that he actually has a chance? They aren't, because he doesn't.

The same thing that happened in 2012 is about to happen now. We're going to wake up on November 9th with a margin of victory many points ahead of current national polling, because national public polling is trash that is being used to drive up click bait or news ratings. What's the point of even looking at this shit if we have to deal with CNN polls that can't even get a proper number of hispanics in a fucking NV poll? Or a proper sampling of under 35 voters? It's insane, and it's so blatant that half of these polls are trash and thrown out into the wild because, shit, we got to talk about something!

Anyways, that's my rant. I'm tired of this oppo-droppo shit. If you guys haven't noticed, Trump is literally a horrible human being and we just spent the last six months going over how fucking horrible he is. It just turns out that as long as you slap an "R" next to someones name, a large portion of the population will vote for them, even if they are racist, sexists shit heads who are literally pawns of a foreign government that is our main global adversary.
preach.gif
 
Eh. You'll know, even if they aren't called.

Yeah, remember that--even if a state isn't called, it will become increasingly apparent how it's going as the returns come in and the eggheads start digging into the reporting counties. They might not be ready to make a projection but they'll start saying "I don't see where he/she gets the votes" / "there just aren't enough votes left".
 

KHarvey16

Member
Overall, not a state. You keep harping on states but there aren't enough states out there. Even though there are 50 states, the election is really about less than 10 states. The polling is too sparse in places like Montana to matter.

Regardless, it doesn't matter. The solution isn't to look at the states because that says nothing. The 10% chance of a blowout prices in all the swing states, already. If she's blowing out in Fla, Pa, and Wis, then is makes sense she's do the same in NH, Nv, and possibly flip Zona. ETC.

Where you seem to be mistaken is that if she has a 10% chance of an election blowout that she is hitting the 10% blowout in each state. Not true! Not how the math works!



Wrong! We don't know the electorate, definitively. Hence why there's a discrepency between CPS and exit polling. How do we verify? Unless every state starts taking down demographic data for every vote that is cast, it's all just good guesswork.

And besides, even if we could determine it, it does nothing to proving the model right or wrong. If the polls are wrong because of demographic data or whatnot, that means the poll inputs in Nate's model was wrong, not the model. If the polls had the correct weighting, his model might be right. Or wrong, even.

You are talking about looking at the pollsters and their accuracy, not Nate's model. Nate's model, hell every one of these models, depend on the pollsters being somewhat accurate so no fault can lie with them if the polls are wrong.


Nate's model doesn't make demographic assumption or any of that stuff. Nate Cohn's NC model is doing that, but that's different.

Nate's inputs, besides polls, are things like incumbancy, trend, economic data like UE, Jobs, etc, and other things. None of which can be tested in the way you'd like.


Do you understand my criticism here? You're making the same mistake Nate Silver made during the World Cup when he thought his model failed.

When something unlikely happens, it doesn't mean your model is wrong. Your model is saying this unlikely event can happen! In fact, it's saying IT WILL HAPPEN sometimes. If you adjust your model because said unlikely event happened, it means you don't understand the math.

What you're basically arguing is that if I flip a coin 5 times and all 5 times it comes up heads, then we must conclude it's a weighted coin.

The 538 model weights and adjusts polls based on a number of factors. After the election you can see if the weighting was done properly based on comparing which polls were accurate and what voter models they used. If you favored a poll or assumed less bias for a poll that turned out to be very innacurate, you need to evaluate the assumptions you made. There are lots and lots of polls and each is evaluated for weighting and "unskewing", which means many opportunities to check your work.

Only considering the overall election result is neccesssry to your argument but doesn't make sense otherwise.
 

Atlagev

Member
I really really doubt the networks will call Fla or even NC before the West Coast closes.

Reason being if you call either, especially the former, the election is already over and that could affect voting out West. I think the media will hold off just long enough. Maybe call it 5 mins before Cali closes.

Assuming they even can call it by then.

If I recall (maybe I'm remembering wrong), the networks called NC for Obama in '08 before the West Coast closed.
 

Angry Grimace

Two cannibals are eating a clown. One turns to the other and says "does something taste funny to you?"
The real question I'm wanting answered is how long it will take before I can go to the weed store
 
Great story on the rising Hispanic vote in Arizona. From 8% of early voters in '08 to 13 this year. Pretty clearly the reason why the Clinton campaign is putting resources in there.

http://www.azcentral.com/story/news...s-nation-early-voting-surge-latinos/93129112/

Yeah. This is why AZ is so much more winnable than GA despite similar polling numbers.

We know who the voters are in GA, and while it's close, I'm not sure where that last 80,000 votes to put us over the top will come from.

We know there's this massive pool of voters in AZ that hasn't been tapped into yet. Maybe they get us there.
 
Yeah. This is why AZ is so much more winnable than GA despite similar polling numbers.

We know who the voters are in GA, and while it's close, I'm not sure where that last 80,000 votes to put us over the top will come from.

We know there's this massive pool of voters in AZ that hasn't been tapped into yet. Maybe they get us there.

The problem is that there's no Reid turnout machine there like in Nevada.
 
Status
Not open for further replies.
Top Bottom