-
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 |OT11| Well this is exciting
- Thread starter Crayons
- Start date
- Status
Aaron Strife
Banned
Marquette releasing their Wisconsin poll in 7 minutes.
Their last poll had Clinton up 15 but that was right after the convention, so I wouldn't expect something so gaudy.
Their last poll had Clinton up 15 but that was right after the convention, so I wouldn't expect something so gaudy.
their last poll actually had her at +2
Actually their last one was Clinton +3 but that was right at the start of the tightening
Which is it?!?!?
Fragamemnon
Member
Great data point out of NH. Still points to that +3-ish national lead.
nature boy
Member
Neo is right.
D
Deleted member 231381
Unconfirmed Member
Monmouth don't publish their RV -> LV methodology and it is really annoying me.
EDIT: blargh and an eighth of the sample was treated differently why aren't contamination effects isolated boo
EDIT2: y'all guys are really weird about Nate. As the number of factors you include goes up, so does the uncertainty. In exchange, you get more accuracy (assuming those factors do have causal effect and you have determined the causation properly). Nate's explicitly said he's built his model to include as many factors as possible - so it naturally has higher uncertainty. Wang's model is... incredibly simple. He even treats states as independent events, which in my personal opinion is too simple and produces an unrealistically high level of certainty. There are advantages and disadvantages to either choice. Nate didn't kill your dog.
EDIT: blargh and an eighth of the sample was treated differently why aren't contamination effects isolated boo
EDIT2: y'all guys are really weird about Nate. As the number of factors you include goes up, so does the uncertainty. In exchange, you get more accuracy (assuming those factors do have causal effect and you have determined the causation properly). Nate's explicitly said he's built his model to include as many factors as possible - so it naturally has higher uncertainty. Wang's model is... incredibly simple. He even treats states as independent events, which in my personal opinion is too simple and produces an unrealistically high level of certainty. There are advantages and disadvantages to either choice. Nate didn't kill your dog.
Aaron Strife
Banned
Oh, nevermind then. That was just the first one that came up in a google search so I assumed it was the most recent.
ItWasMeantToBe19
Banned
nature boy
Member
PLOUFFE!
ItWasMeantToBe19
Banned
What about Maine though.
Aaron Strife
Banned
Clinton at most would lose 1 EV.
I visited Ellis Island for the first time yesterday, and what struck me the most was the museum's section on the political environment during the early 1900s. Reading about the immigration opponents and their propaganda sounded EERILY like what we're witnessing now from Trump and the R party...100 years later.
eBay Huckster
Member
Alternately, if New Hampshire is the tipping point state at that margin then she's winning by 8%.
GutsOfThor
Member
Is it just me or did we go from saying last month that she was going to get as close to 400 ev's as possible to now just hoping she gets 270? Has it really gotten that bad? I haven't been paying attention much the past few weeks.
RustyNails
Member
WI 9/15 - 9/18
HC 44
DT 42
No change
HC 44
DT 42
No change
Plouffe has been MVP lately. Dude is so confident it's funny.
We won't, really. All of these models are going to be pretty close together at the start of November. And of course the election is only one sample - they're offering probabilities. Trump winning wouldn't tell you that much about whether 538 was doing a better job all along than the others.
The "did they call it right?" question is just not that interesting. That's basically just averaging the latest polls, though there's a bit more you can do like trying to figure out if a pollster has a house effect. The reason these sites are interesting is that they try to forecast the election well in advance, without the benefit of polls released November 1st. But you never get a great check on this because even if one was moving by 20 points every week and the other barely moved, they're all going to end up in basically the same place when the election actually happens, and the election result doesn't tell you anything about how the voters got there.
RustyNails
Member
4 way race WI
HC 41
DT 38
GJ 11
JS 2
HC 41
DT 38
GJ 11
JS 2
ItWasMeantToBe19
Banned
Super lame result.
nature boy
Member
Wisconsin stable, in both H2H and 4way.
Small advantage though.
Small advantage though.
Right? Colorado was the tipping point state in '08 and '12, and both times it was about 1.7% more Democratic than the national average.
StopMakingSense
Member
New Hampshire is too educated to not be a blowout.
RustyNails
Member
Wouldnt call it stable. H2H is within moe.
worse than bad , man. losing florida, ohio and barely above water in viriginia, nc, and wisconsin. also rumors that kaine's absence on trail is related to prepping by clinton team of eventuality that hillary will succumb to the common cold after election.
eBay Huckster
Member
we went from a high point to a low point and now we're back on an upswing
otherwise, listen to incognito. place bets according to that post right there.
nature boy
Member
Didn't move from last poll.
Spiritreaver
Member
Are there any CO polls happening this week?
Other than probably one or two of those 50-state polls you should take with a grain of salt, not that we know of.
Aaron Strife
Banned
Meh WI poll isn't great but at least it's a lead.
Wanna know how Feingold is doing too.
Wanna know how Feingold is doing too.
eBay Huckster
Member
it works by poisoning all anime
GutsOfThor
Member
whyamihere
Banned
There's literally no movement from there last poll, so yes, it is exactly what you'd call stable.
balladofwindfishes
Member
this took me too long to figure out what the heck you were talking about
ItWasMeantToBe19
Banned
Arabs have to be the whitest group hatred by white people for not being white since pre-WW2 Jews.
Feingold is +6 (up from +3 last month).
Feingold is +6 (up from +3 last month).
whyamihere
Banned
Jeeeesus, I missed this.
Eh, I kind of agree, but if we devalue the importance of the Nov 8th results, then what are we even arguing about with 538 currently? What even is the point of these forecasts other than fun ways to look at poll trendlines? For that we could just look at the RCP poll aggregate and call it a day. Silver and Wang became famous for a reason and that wasn't just calling Obama or McCain the winner.
Their models are meant to lead to Nov 8th and give us more than just who wins, but how close they end up being to the predictor, in addition to the individual state margins and state congressional races. I'd argue that if those margins are closer than the other models, then that does add more veracity to the methodology used by the "winner".
Ignatz Mouse
Banned
Go Andy!
Aaron Strife
Banned
There's my African-American!
- Status
Similar threads
- Locked
- Locked