|
OgreNoah posted:In Michigan at least the only "early" voting is for absentee ballots, which is for people not here, and old people. Old people are more conservative naturally. And Clinton still seems to be winning. This is also true in PA.
|
#
¿
Nov 2, 2016 14:25
|
|
|
|
| # ¿ May 12, 2024 18:50 |
|
|
Mind_Taker posted:There are three buckets: He does too, FWIW. In the histogram for number of electoral votes, you'll notice that HRC has two peaks. One is right over the 270 mark, and the other is at 370. quote:That means that the polls being more or less correct only has a 40% chance of happening under Nate's model, which I call bullshit on. To me, 40% seems a little low, but there are at least two big reasons polls should be less informative this election than they have been in the previous few: - Numerous other people have pointed out the problems polls have reaching non-traditional voters, and both parties have made those voters key parts of their strategies. Polls can't make judgments about populations they systematically can't/don't reach. - The GOP GOTV operation will be unlike any in recent memory, both in quality and quantity. (Doing GOTV by calling in on talk radio?) The asymmetry is unprecedented. What does that mean about the accuracy of the polls' likely voter screens? Hard to say. I don't know whether 538 is predicting a lot of uncertainty for the "right" or "wrong" reasons, but those two things alone are huge question marks. It's reasonable to assume that the polls are much more likely to miss (for or against HRC) than they have in the past few elections.
|
|
#
¿
Nov 2, 2016 20:59
|
|
|
Northjayhawk posted:Here though race is an issue and you have the unfortunate angry black man stereotype so its probably not a good idea to go there here, but people used to it on other forums and probably not even thinking about race don't realize that. It's also quite understandable to be angry about the issues discussed in D&D. If someone is angrily making good points, they're making good points. (Didn't catch the start of this, so no idea if Crain was actually making good points or what.)
|
|
#
¿
Nov 2, 2016 21:22
|
|
|
mcmagic posted:I feel that "Trump engages in misconduct with That might have legs. I mean, not this close to the election. Although "Trump! Child rape!" might crowd out "Hillary! Emails!" for a few days, which might be nice. And then, you know, maybe a woman gets a small measure of justice for being sexually abused. I mean, it's a long shot, but it could happen.
|
|
#
¿
Nov 2, 2016 21:34
|
|
|
joepinetree posted:For the record, and once again. PEC and Drew Linzer (now at KOS) have better track records than Nate over the last few presidential elections. It's difficult to evaluate investment funds until they've gone through both good and bad times, preferably with at least one crisis in the mix. It's easy to overoptimize for one environment, then do poorly when that environment changes. It's not hard to point to multiple sea changes that Trump's candidacy has brought to this election. I mentioned the changes in the parties' voter outreach and GOTV strategies upthread; another is the potential decoupling of the top of the GOP ticket from the bottom. Even if the models you cite were the most accurate in 2008 and 2012, it remains to be seen if they will be in 2016. Also: SNYPE 4 HRC
|
|
#
¿
Nov 3, 2016 02:31
|
|
|
I signed up to phone bank this morning. Can't sit around stressed and not do anything about it. I haven't done it yet, but it looks dead easy; log in, get a number and a script, talk to someone. It had me sign up for shifts, but AFAICT you just log in and do it when you have time. (Was the shift-picking just for motivation? idk.) It lets you pick the state you want to call if you like, and even has gamification. I already feel better, and Inhavent even done it yet. If you're Arzying, I highly recommend it. I think giving people nudges in these last days is really going to make at least a small difference, and Senate control (hell, maybe even POTUS) is likely to hinge on small differences.
|
|
#
¿
Nov 3, 2016 17:29
|
|
|
Fojar38 posted:Remember when a presidential candidate bragged about being a rapist on tape and everyone forgot about it a month later Republicans: Literally that Dude from Memento
|
|
#
¿
Nov 4, 2016 11:20
|
|
|
Blurred posted:They've spent 8 years doing nothing but saying no to Obama, they're being led by the nose by a demographic that is increasingly indifferent to the laissez-faire idealism of the party's establishment, they've forgone all talk of boiler-plate social conservative positions (abortion, gay marriage and "family values" have hardly been raised at all this election), they have a leader who has openly insulted most of them, and all that they are capable of saying is, "well, at least he is not Hillary Clinton". This is not new. IIRC, it was Karl Rove who said that GWB would have lost re-election if he had been running against nobody. The whole premise of their re-election campaign was "Look how awful Kerry is, we're not that." It's a lot more obvious this election, but it's been a defining characteristic of the GOP since at least the 1960s.
|
|
#
¿
Nov 4, 2016 16:53
|
|
|
WampaLord posted:Because his model is so far out of whack with the other various polling models. Either they are all wrong, or he is. You're neglecting the possibility that they're all wrong!
|
|
#
¿
Nov 4, 2016 20:45
|
|
|
Bloops Crusts posted:It's not the case that Silver's model is "out of whack" with the rest of the models. It simply expresses a lot more uncertainty about the outcome, which manifests as a higher probability of winning for the candidate behind in the polls and a lower probability for the candidate out front. Also very asymmetric (and for the GOP, unorthodox) GOTV operations and both camps targeting non-traditional voters. You can argue whether 538 has modeled uncertainty properly. (The t-distribution seems reasonable; the state-correlation may or may not be; the relative lack of poll curation seems unwise.) But I find it hard to argue there isn't good reason to be uncertain.
|
|
#
¿
Nov 4, 2016 21:02
|
|
|
cant cook creole bream posted:That's the sort of poll which is really relevant, but doesn't flow into most polling aggregates at all. Unless for some reason the people lied about who they voted, those are guaranteed numbers rather than some vague notion of "likely voters". And this is where Nate's correlated states thing can really bite him. "Polls are tightening in AZ! I bet it's tightening in CO, too!" when, even if that were true, it would affect a tiny part of the electorate that hasn't voted yet. He really needs some kind of scaling parameter in there for early voting if he's gonna do that.
|
|
#
¿
Nov 4, 2016 21:34
|
|
|
|
| # ¿ May 12, 2024 18:50 |
|
|
Eugene V. Dabs posted:Are CO and AZ correlated in his model? Because the two states have very little in common in terms of culture or educational attainment. Literally the only thing similar is racial/ethnic makeup and their geographic location. Full disclosure: I pulled AZ (and CO, to a lesser extent) out of my rear end. The general problem of accounting for locked-in votes remains a problem in 538's model, though, AFAICT.
|
|
#
¿
Nov 5, 2016 01:56
|
|