|
CapnAndy posted:I appeal to you, wise goons, to unfuck some dissonance for me. This is my absolute nightmare map, and quite frankly it felt dishonest to turn NC red: Trump's chances in Nate Silver's model are almost entirely "wait, the polls might all be wrong, and when I mean wrong, I mean WAY Wrong!!! Like beyond the MoE!" If you change that assumption to "nah, the polls are basically right, maybe off by a point or two, and maybe 1 or 2 states will be off by 4 or 5", then Trump's chances dive near zero.
|
#
¿
Nov 1, 2016 19:52
|
|
|
|
| # ¿ May 10, 2024 02:53 |
|
|
ReV VAdAUL posted:Silver's clickbait stuff is probably him just hedging his bets for after the election. Polling does seem to be getting less reliable with the British general election last year and the Spanish general election this year having right wing parties doing considerably better than the polls suggested they would. Consequently the reputation of pollsters have taken a significant hit. One of the more anti-arzy thoughts from Nate, is the couple times he's talked about his worry that the polls might be off, he's said he's more concerned that the polls might be greatly underestimating Clinton, rather than underestimating Trump.
|
|
#
¿
Nov 1, 2016 19:56
|
|
|
Supercar Gautier posted:As far as I've read, the reason 538's model is the most pessimistic for Clinton is because it's the model that most heavily accounts for a Total Polling Failure in either direction. This uncertainty inherently favours the underdog unless the leader is WAY ahead, like Clinton was a couple weeks back. 538's simulations are more likely than others to produce a Clinton landslide (which doesn't actually improve her victory odds because it's just running up the score), but also more likely to produce a Trump win. and given that n for presidential elections without an incumbent is always going to be low, it is almost impossible to prove whether its better to have lower variance or higher variance in election simulations.
|
|
#
¿
Nov 1, 2016 20:13
|
|
|
computer parts posted:Incidentally if anyone's interested in how political calculus will change in the future, here's a map on how apportionment will happen in 2020: Thats the price of more moderate, progressive people leaving the frozen wastelands. In the long run hopefully your vote will be more efficiently spread out, making it impossible for the GOP to win a national election, but there will be a short window of time where they benefit from more electoral votes in red states they haven't lost yet. It won't be a problem if we can flip AZ and Texas obviously, or at least make it so that Florida is no longer a swing state, but it needs to happen by 2024 when we have the new map.
|
|
#
¿
Nov 1, 2016 20:17
|
|
|
Mind_Taker posted:Also the thing about Nate's model is that he is basically giving a 30% chance that the polls will be wrong enough in Trump's direction that the EC will swing in Trump's favor (like a 4%+ swing in the popular vote). Whether Hillary wins bigly or just wins only matters downticket.
|
|
#
¿
Nov 1, 2016 20:21
|
|
|
spacetoaster posted:I don't think anyone is super excited to get out and vote for Hillary. At least not anything like we saw for Obama in either of his presidential elections. They are excited to block Trump. Not many people are excited about Hillary. As for crowd size, Romney and Mondale drew big crowds, too. I'm not really interested in the size of Trump's hate rallies.
|
|
#
¿
Nov 1, 2016 20:27
|
|
|
Crowsbeak posted:Its dumb, but its less dumb then not expanding the house. Seriously are their any good resources for why we haven't, because we have over three times what the population was in 1020 and we have same number of reps now as we did then? Also Wisconsin and my state seem to have been having the same population growth. They won't even consider expanding the house unless the Dems achieve total, overwhelming control over government and can trample any and all GOP and moderate objections. The reason why is a larger house helps the Dems in the electoral college. The GOP benefits a lot more from the 2 extra free EV that every state gets because of all the safe red flyover states. A bigger house means more total EV, which then dilutes the artificially high influence that small states have.
|
|
#
¿
Nov 1, 2016 20:48
|
|
|
FuzzySlippers posted:Why are people with anime avatars suddenly so afraid of war with Russia? We got through the cold war. Syria isn't causing WW3. One of the most shocking things about this election is how the GOP has suddenly become a bunch of weeping, trembling cowards who want to appease Putin so the big red bear doesn't kill us.
|
|
#
¿
Nov 1, 2016 21:06
|
|
|
BarbarianElephant posted:I think they mostly want to team up with Putin because they think he's the bee's knees and a model for how a leader of a country should act. I'm talking about the idiots who attempt to make the argument, with a straight face, that if we elect Hillary Clinton she'll start WW3 against Russia.
|
|
#
¿
Nov 1, 2016 21:11
|
|
|
Shimrra Jamaane posted:Considering all we know how the gently caress is that even possible? 5% of polls should be expected to be outside the margin of error. If we see several similar results from reputable pollsters though, then thats not good.
|
|
#
¿
Nov 1, 2016 21:12
|
|
|
ImpAtom posted:+7 from a reputable pollster is a bit more than a weird outlier though. Like even if their MoE is entirely towards Trump that's still huge. Your right, I looked up the math on this. If Clinton is actually up in NC, say by +1 or +2, then the chances that a random representative sample from a well-designed poll with a MoE of like 3 or 4% (which is supposed to be 2 standard deviations), would come up with Trump +7 (about 4 sd) is extremely low, like a tenth of a percent. Unlikely events happen though when you get hundreds of polls every week. So, this is either an extremely unlikely outlier, or Trump is winning in North Carolina, or there are serious structural problems with how that poll (which was A-rated in 2012) attempts to get a representative sample.
|
|
#
¿
Nov 1, 2016 21:23
|
|
|
Dexo posted:Why are we arzying again? If NC is tied, then we're fine. If Trump opens up a solid lead in NC, then that probably means that states which Clinton DOES need get uncomfortably close.
|
|
#
¿
Nov 1, 2016 21:26
|
|
|
theblackw0lf posted:Don't weekend only polls, especially ones right near a holiday, tend to skew Republican, since younger voters will more likely be out or busy? Thats a good point. Now obviously they adjust for that demographically, but I would argue that young people who are actually at home and answer the phone on Halloween weekend are probably more Republican-leaning than young people who are out getting trashed. Northjayhawk fucked around with this message at 21:31 on Nov 1, 2016 |
|
#
¿
Nov 1, 2016 21:28
|
|
|
Crowsbeak posted:Why was it restricted after the 20s though? It wasn't "restricted", congress can choose to expand the house any time they want. They have just chosen not to.
|
|
#
¿
Nov 1, 2016 21:35
|
|
|
Also, when a pollster is brave enough not to bury what looks like an obvious outlier, thats a good thing, and they should be commended for it (unless they are partisan with intentionally lovely methodology). I don't want to be lied to, if Trump is pulling ahead, I want to know as soon as possible. We don't want pollsters to herd to the conventional wisdom, because thats where shocking election night results and madness lies.
|
|
#
¿
Nov 1, 2016 21:37
|
|
|
New poll in CA from SurveyUSA Clinton is +21, down from +25 before email poo poo. Pot legalization is +15 with Yes at 54% and only 7% undecided
|
|
#
¿
Nov 1, 2016 21:44
|
|
|
lol, the Log Cabin Republicans endorsed Trump.
|
|
#
¿
Nov 1, 2016 21:48
|
|
|
Crowsbeak posted:I mean why did they decide to suddenly not support expanding it. Politics. They did the math and the ruling party found out that expanding would be bad for them. Same in 1940. By 1950, everyone was used to 435 and the new normal was not expanding. Expansion today would seem like a radical new idea to a lot of people.
|
|
#
¿
Nov 1, 2016 21:49
|
|
|
Hypothetically, if on election night the polls are where they are now, we expect Clinton to get a close win maybe a 50/50 senate, and..... inexplicably Clinton just cruises to a 12-point wipeout and the Dems take all of congress. What would that mean for polling this year and how could they adjust to a huge miss? Maybe a correlation between people who enjoy answering polls and GOP voters? Missing latinos showing up at the polls? Or just a one-time anomaly driven by the fact that Trump is Trump?
|
|
#
¿
Nov 1, 2016 22:17
|
|
|
BROCK LESBIAN posted:I read a post earlier that suggested there was low turnout though. Lower AA turnout. Latinos seem to be compensating for it.
|
|
#
¿
Nov 1, 2016 22:19
|
|
|
Tir McDohl posted:This election has really turned me into a hardline partisan. At this point it is clear that I cannot get along with the kind of psychopath who would vote for Trump. Kinda scary that around 40% of the electorate might as well be aliens to me now. I don't even want to engage in discussion with them. Scorched earth. I'd be pruning my Facebook heavily if I had not quit it years ago because of the conservative bully pulpit there. Yep. I've never voted straight Democrat. Federally, Governor, and the state legislature sure, but at the state and local level I do my research and usually find at least 1 or 2 Republicans somewhere that look more competent (or their opponent is a crazy-rear end) running for what is basically a non-partisan non-legislative office. This year, I'm doing no research except for races where party is not listed and not known, then I guess I'll figure those out. This year, if you have a D next to your name, you got my vote. If a Republican is running unopposed, I'll leave that race blank.
|
|
#
¿
Nov 1, 2016 22:26
|
|
|
canepazzo posted:FWIW, that SurveyUsa poll of NC had Trump at 14% afro-american vote and 43% 18-35 vote which seems a bit off. Thats 43% of 18-35 who had no Halloween weekend plans, stayed home sober and watched TV. So, good young Republicans.
|
|
#
¿
Nov 1, 2016 22:28
|
|
|
Radish posted:Seriously any Democrat that appoints a Republican to any office unless there's no alternative whatsoever after Obama's term is an absolute fool. Well, its probably fine it you want to appoint a Republican to Commerce or Energy. FBI director and Justice department are now clearly politically partisan offices, and it should become routine for the sitting FBI director to offer a resignation if the other party wins.
|
|
#
¿
Nov 1, 2016 22:35
|
|
|
Spacebump posted:Hi, I think Donald Trump is unacceptable so I'm going to light my vote on fire. He's running in 2020 and wants to establish his #NeverTrump bona fides in the event that Trump gets wiped out.
|
|
#
¿
Nov 1, 2016 22:43
|
|
|
Radish posted:If one possibly/possibly not an outlier poll from a state which Obama lost last election is all it takes for you to freak out and proclaim Trump the next president, you probably shouldn't watch the news for the next seven days or you will have a nervous breakdown. The consequences of a loss, however unlikely, is a hell of a lot worse than normal for an election. If I was offered a choice between a 1 term Trump followed by a 2 term Democrat (if we're still alive) or a 1 term Clinton followed by a 2 term Republican and those were the only options, I would probably choose the latter.
|
|
#
¿
Nov 1, 2016 22:45
|
|
|
Crowsbeak posted:Then why not McMullin? Egg McMuffin is not the highly-respected governor of a swing state.
|
|
#
¿
Nov 1, 2016 22:48
|
|
|
here, have an outlier in the other direction https://twitter.com/PpollingNumbers/status/793569865789542404 Problem with this poll is that half of it was before email poo poo.
|
|
#
¿
Nov 1, 2016 22:50
|
|
|
Guy Goodbody posted:They all get banned. If you toxxed for Trump and he loses, you get banned and if you toxxed for Clinton and she loses you get banned In addition to this, if Hillary loses, then anyone with the gang tag shown on my posts gets perma-banned.
|
|
#
¿
Nov 1, 2016 23:12
|
|
|
turn it up TURN ME ON posted:Oh nooooo lol, bullshit
|
|
#
¿
Nov 1, 2016 23:12
|
|
|
Charlz Guybon posted:Some bad polls came out? There were also some OK polls and some good polls. The bad poll was recent and we're goons, so we're still in the process of getting off the ledge.
|
|
#
¿
Nov 1, 2016 23:22
|
|
|
There aren't really many good things about the FBI's email poo poo, but one of the greatest is that its given Trump hope, and since they literally spend more on red MAGA hats than on polling, Trump will go into election night thinking he's going to be president. If he gets wiped out, he'll be hilarious, and the tears online from his supporters will be glorious.
|
|
#
¿
Nov 1, 2016 23:26
|
|
|
canepazzo posted:Trump's GOTV is going to win Nevada: lol, that is just hilariously incompetent. I assume this guy has never done anything which would have ever gotten him on a GOP phone list. If they are just calling everyone in certain areas (not latinos or casino workers) where they think Republicans live, then that "GOTV" effort is stupidly inefficient.
|
|
#
¿
Nov 1, 2016 23:30
|
|
|
VikingofRock posted:Yeah, I mean they are my friends and family, too. It's just in previous elections I never had this same level of disgust for Republican voters. I could see how you could be a decent person and still vote Romney or McCain or even Bush. Now, with Trump's and the alt-right's hate being so incredibly blatant, that feeling is gone. We're learning the psychological limits of how far people are willing to go, what bullshit they'll eaccept and truths to ignore, to convince themselves that they are not actually voting for a monster when the supreme court is on the line. The people who are unable to bridge that gap are the #NeverTrumpers, and they are depressingly few, but good on Utah for having some morality. I believe that if one side or the other had an unassailable large majority on the court and it was obvious that it wasn't going to change in the next decade, then the stakes would be lower and people would be less willing to rationalize voting for Trump.
|
|
#
¿
Nov 1, 2016 23:37
|
|
|
Mel Mudkiper posted:I mean one of the clearest signs that Nate's model might be over-estimating uncertainty is that his model is an outlier of all the other aggregates. Even Sam Wang's very high 99% is still matched by two other pollsters. Its also difficult to prove that Nate is the real outlier. That may be part of his decision, it'll take another 100 years of elections to prove that he's got too much uncertainty, but it only takes one shocking night where Trump somehow wins by +4% or something to discredit all of the 99% guys. If this was my livelihood rather than a hobby, I'd be careful too.
|
|
#
¿
Nov 1, 2016 23:41
|
|
|
This is the first time we'll be able to judge the absolute value of GOTV. We could judge the relative value when Obama outperforms the polls because he was better at it then Romney, but at least Romney tried. This is probably the only election we'll ever see where one side has chosen to do as close to zero GOTV as we can ever reasonably expect. We're assuming Hillary might outperform her polling a point or two, but hell, maybe the absolute value of GOTV is more like 6-7 points when the other side doesn't even show up to match your efforts. If election night polls are H+4 and she wins by double digits, that'd be nice.
|
|
#
¿
Nov 1, 2016 23:54
|
|
|
Anomalous Blowout posted:I think everyone who's posted their reasons (financial crisis, tox bans, etc.) has been right to some extent. It's a combination of factors. To add in my own, I'd also say that part of it might simply be an age thing. It seems to be a rite of passage as an American youth to go through a libertarian phase in your late teens/early 20s, especially if you're middle-class or higher on the income bracket. This is a good post. I went through this as well, but the swing was more dramatic for me. I bought Limbaugh's first two books and listened to his show a couple times a week. The first crack was when I really liked McCain and the conservative bubble poo poo all over him, which made me angry. Then as you described, the daily unfairness of life and seeing good people working hard getting screwed over, unpopular groups (gays, minorities) ostracized and kicked around eventually made me realize my side was pretty terrible. I only spent maybe 1 or 2 years passing through as a libertarian when I left the conservative movement, when I started to move left it was pretty quick.
|
|
#
¿
Nov 2, 2016 00:00
|
|
|
AlouetteNR posted:For people who know more about polling, how does it handle early voting? Are people who already voted still included, and is there a lesser likelihood of people who have already voted responding to the poll? Because I feel like a decent partial explanation for Clinton's drop in very recent polling would be more Clinton voters taking part in early voting and being excluded from future polls, making it seem like Trump is surging. If you tell a pollster that you already voted, then you obviously make it through the likely voter screen.
|
|
#
¿
Nov 2, 2016 00:19
|
|
|
Muggsy posted:Are there any handy guides out there that explain American election data and statistics? Like, things that explain the implications of states having a candidate at +4 or something? I have some friends that are interested in learning about the electoral system, and I am admittedly curious too. Sure. In all cases states can change categories slowly over time as people move around and ethnic minorities grow. For example, California and Missouri used to be competitive, now they are really not, at all. Dark Red Hellholes In these states, a Generic Republican should win by 20% or more. No one even think about these states as competitive, and if a Dem is within 10, it is shocking. UT (weird exception this year because Mormons are aghast at Trump and Even McMullin is running a strong 3rd party) WY Nebraska's 3rd CD and Nebraska Statewide ID OK WV AR KY KS AL ND TN Very Red States In a gigantic landslide, these might be close, but the Republican will probably still win. SD LA AK Nebraska's 1st CD Texas (keep an eye on it, demographics changing fast with Latinos growing) MT MS Reliable Red States These should usually be around +10% GOP. Dems don't target these unless they are running up the score. A Dem sometimes wins one, Obama picked off Indiana once. IN SC MO AZ (omg Hillary might win this. Like Texas, demographics are changing) GA Competitive Some lean more to the left than the right. I'll start with the most GOP leaning state and go down. Nebraska's 2nd CD (OK, this isn't a state, but should usually be +5% GOP) OH VA IA NC FL (This is about where the zero point is for a generic GOP vs a generic Dem) NH CO NV PA (The GOP's white whale, should normally be about +5% Dem) WI Maine's 2nd CD MN Reliable Blue States GOP only targets these to run up the score, should be about +10% Dem. Its a short list. NM (Good old New Mexico. For a long time it was the inexplicable big blue square in the sea of red. Now they are joined by CO, and soon AZ and hopefully TX in future elections) MI (lol Trump, why the gently caress are you wasting time in Michigan?) Very Blue States OR WA Maine Statewide Blue Strongholds There's a lot of them. These should normally be near +20% Dem on up. The GOP isn't winning here anytime soon, unless something weird happens or the GOP shifts a lot farther left. IL CT NJ DE Maine's 1st CD CA MA MD RI NY VT HI (lol, GWB's campaign once thought this state was in play for a couple days and sent Cheney there) DC (Obama won the district by 84% in 2012) Northjayhawk fucked around with this message at 00:45 on Nov 2, 2016 |
|
#
¿
Nov 2, 2016 00:43
|
|
|
Samuel Clemens posted:That's a good write-up, although calling PA "competitive" is a bit of a stretch. It pretty much never turns red outside of landslide elections. Its competitive in the same way that Charlie Brown is always close to kicking the football. The GOP usually doesn't get wiped out in PA and they are obviously able to elect the random GOP senator here and there, but they always seem to fall several points short.
|
|
#
¿
Nov 2, 2016 00:55
|
|
|
|
| # ¿ May 10, 2024 02:53 |
|
|
Lemming posted:Even if the Dems take the Senate, there's basically 0 chance they keep it in 2018 Yeah, if the Dems take the Senate and the old liberal lions on the court are smart, they have a 2 year window to retire. Hillary also has 2 years to maybe try to get any legislation or reforms passed. The odds that the Dems lose the Senate in 2018 are not 100%, but they are really drat high.
|
|
#
¿
Nov 2, 2016 03:27
|
|