|
Ciaphas posted:Do we know who leaked the taxes to NYT?
|
#
?
Oct 3, 2016 05:11
|
|
|
|
| # ? May 20, 2024 00:03 |
|
|
lozzle posted:Not a clue. Suspects so far include: Edit: joking aside, Marla makes a ton of sense based on what the person above me just said. 
|
|
#
?
Oct 3, 2016 05:13
|
|
|
It was Putin
|
|
#
?
Oct 3, 2016 05:14
|
|
|
GalacticAcid posted:so who's excited for all the trump costumes this halloween Even if it's topical, going around as a klansman is still pretty gauche.
|
|
#
?
Oct 3, 2016 05:14
|
|
|
Honestly, I get why people are still Arzying despite all the good news this past week. I mean, I personally think Clinton's going to win unless something absolutely game-changing happens, but I'm not going to be confident until the results are in and I see TRUMP LOSES BIGLY in big flashing letters on the screen. There's so much at stake this year - the presidency obviously, but a Trump win gives the GOP control of all three branches and 35 states, never mind how much worse President Trump could be than a generic R - which is going to make a lot of people nervous. Democratic votes have this year have produced some awful results. And most of all, Trump is an open bully, a grifter, a narcissist with fascistic tendencies who doesn't even try to hide it and produces an ever-expanding list of scandals, controversies, and unbalanced statements, any one of which would be enough to sink a conventional campaign, and still he's come this far with none of his abhorrent poo poo sticking to him or shaking his base. And then one week of bad decisions from Hillary's campaign nearly put him level. This past week has been really encouraging that finally it's sticking, finally it's catching up with him, and I know that realistically even if he stops self-imploding the damage is likely done. But there's still six weeks to go and it's still really hard to shake the fear of what if. What if Assange really has something concrete behind his bullshit this time? What if Trump stumbles his way into some killer attack lines at the debates? What if, at the end of all, enough people see Trump for what he is and decide that that's what they want for the country?
|
|
#
?
Oct 3, 2016 05:16
|
|
|
Shimrra Jamaane posted:Find a poll that isn't the LA Times tracker that has Hillary down. One that isn't at least 2 weeks old. Go on. http://elections.huffingtonpost.com/pollster/polls/upi-cvoter-25825
|
|
#
?
Oct 3, 2016 05:18
|
|
|
GalacticAcid posted:It was Putin You have failed me for the last time, Donald 
|
|
#
?
Oct 3, 2016 05:18
|
|
|
Crain posted:Even if it's topical, going around as a klansman is still pretty gauche.  Apraxin posted:The Stakes Yep there's a lot at stake. If you're nervous, volunteer.
|
|
#
?
Oct 3, 2016 05:19
|
|
|
Because the poll is conducted online and individuals self-select to participate, a margin of error cannot be calculated.
|
|
#
?
Oct 3, 2016 05:19
|
|
|
Corncob is a dumb term and using it unironically is 4chan-level unfunny.
|
|
#
?
Oct 3, 2016 05:19
|
|
|
NaanViolence posted:Corncob is a dumb term and using it unironically is 4chan-level unfunny. Hmmm interesting point Have you considered perhaps, as a counter-argument, that you should suck on my rear end in a top hat?
|
|
#
?
Oct 3, 2016 05:21
|
|
|
Night10194 posted:The biggest sign she's winning right now is that loving Rasumussen has her up. Seriously, Rasmussen had R-money up like 3 points or something on ELECTION DAY last year. If even they think Hillary is winning she's doing fine.
|
|
#
?
Oct 3, 2016 05:21
|
|
|
"Because the poll is conducted online and individuals self-select to participate, a margin of error cannot be calculated." Sounds legit. e: goddammit
|
|
#
?
Oct 3, 2016 05:21
|
|
|
I suppose part of it is also just the uncertainty of waiting to see if the latest major hit actually did anything. I felt the same way right after the DNC until a bunch of polling came out and showed he got destroyed.
|
|
#
?
Oct 3, 2016 05:22
|
|
|
I love when people who never post in this thread try to tell people what to post.
|
|
#
?
Oct 3, 2016 05:23
|
|
|
computer parts posted:Because the poll is conducted online and individuals self-select to participate, a margin of error cannot be calculated. Oh, huh. Haven't taken stats class in years, what's the difference between a margin of error and credibility interval? "Because the poll is conducted online and individuals self-select to participate, a margin of error cannot be calculated. The poll has a credibility interval of 3 percentage points." The huff post page (http://elections.huffingtonpost.com/pollster/polls/upi-cvoter-258250) says the margin of error is 3%, so they must be conflating them somehow.
|
|
#
?
Oct 3, 2016 05:23
|
|
|
NaanViolence posted:Corncob is a dumb term and using it unironically is 4chan-level unfunny. Nope
|
|
#
?
Oct 3, 2016 05:25
|
|
|
A Fancy 400 lbs posted:Seriously, Rasmussen had R-money up like 3 points or something on ELECTION DAY last year. If even they think Hillary is winning she's doing fine. I think it's kind of ridiculous to discredit the Rasmussen poll on that point when it was 4 years ago and they certainly could have updated their methodology to fix the problems they had back then. They may still have a bias, but they were wrong last time and it would make sense if they tried to fix things and be better this time around.
|
|
#
?
Oct 3, 2016 05:25
|
|
|
vyelkin posted:Trump will take that property he just bought down the street from the White House and renovate into the Trump White House and pretend he's the president and hold court there with all his cronies making up a fake cabinet, while he pursues aggressive litigation trying to prove the election was stolen. I envision at least once a week he'll get all the press together and point towards the WH and bitch about Hillary and her election rigging and blaming her for everyone's and everything's problems
|
|
#
?
Oct 3, 2016 05:27
|
|
|
I wonder how much jealously rear end has for Snowden.
|
|
#
?
Oct 3, 2016 05:27
|
|
|
legit corncob owns and I will force it into national memetic status through sheer force of will
|
|
#
?
Oct 3, 2016 05:27
|
|
|
Built 4 Cuban Linux posted:I think it's kind of ridiculous to discredit the Rasmussen poll on that point when it was 4 years ago and they certainly could have updated their methodology to fix the problems they had back then. They may still have a bias, but they were wrong last time and it would make sense if they tried to fix things and be better this time around.
|
|
#
?
Oct 3, 2016 05:28
|
|
|
NaanViolence posted:Corncob is a dumb term and using it unironically is 4chan-level unfunny. Your red text is deserved. Built 4 Cuban Linux posted:Oh, huh. Haven't taken stats class in years, what's the difference between a margin of error and credibility interval? "Because the poll is conducted online and individuals self-select to participate, a margin of error cannot be calculated. The poll has a credibility interval of 3 percentage points." The huff post page (http://elections.huffingtonpost.com/pollster/polls/upi-cvoter-258250) says the margin of error is 3%, so they must be conflating them somehow. It's some kind of Bayesian thing. I haven't taken stats or probability in forever either but I'll look into it. Built 4 Cuban Linux posted:I think it's kind of ridiculous to discredit the Rasmussen poll on that point when it was 4 years ago and they certainly could have updated their methodology to fix the problems they had back then. They may still have a bias, but they were wrong last time and it would make sense if they tried to fix things and be better this time around. Rasmussen leans substantially R like every election year.
|
|
#
?
Oct 3, 2016 05:29
|
|
|
Built 4 Cuban Linux posted:I think it's kind of ridiculous to discredit the Rasmussen poll on that point when it was 4 years ago and they certainly could have updated their methodology to fix the problems they had back then. They may still have a bias, but they were wrong last time and it would make sense if they tried to fix things and be better this time around. They lean R so consistantly that if you add +x(I wanna say 3 or 4?) D to any result they have they actually become one of the most accurate pollsters out there.
|
|
#
?
Oct 3, 2016 05:33
|
|
|
And the news cycle said HA! https://twitter.com/cnnbrk/status/782796191419465728
|
|
#
?
Oct 3, 2016 05:38
|
|
|
Built 4 Cuban Linux posted:Oh, huh. Haven't taken stats class in years, what's the difference between a margin of error and credibility interval? "Because the poll is conducted online and individuals self-select to participate, a margin of error cannot be calculated. The poll has a credibility interval of 3 percentage points." The huff post page (http://elections.huffingtonpost.com/pollster/polls/upi-cvoter-258250) says the margin of error is 3%, so they must be conflating them somehow. According to Wikipedia: quote:For example, in an experiment that determines the uncertainty distribution of parameter t, if the probability that t lies between 35 and 45 is 0.95, then Which doesn't really help things.
|
|
#
?
Oct 3, 2016 05:38
|
|
|
computer parts posted:According to Wikipedia: I found a pretty in-depth explanation from Ipsos: https://ipsos-na.com/dl/pdf/research/public-affairs/IpsosPA_CredibilityIntervals.pdf Bayesian probability is very much not my thing though so I can't say I fully understood it.
|
|
#
?
Oct 3, 2016 05:42
|
|
|
Night10194 posted:I suppose part of it is also just the uncertainty of waiting to see if the latest major hit actually did anything. I felt the same way right after the DNC until a bunch of polling came out and showed he got destroyed. The debate did serious damage too.
|
|
#
?
Oct 3, 2016 05:42
|
|
|
Oxxidation posted:So is Robert Costa, like, capable of astral projection or something, because this guy's privy to a downright eerie amount of insider info. Costa worked at the National Review for a while and he's basically spent his entire early career cozying up to conservatives and learning how to speak their language. He's simply really good at what the does without having drunk the kool-aid himself. Shimrra Jamaane posted:I believe that a bunch were thrown out because they want to redo them to take into account this weekends bombshell. I really doubt it because polls aren't exactly cheap.
|
|
#
?
Oct 3, 2016 05:46
|
|
|
Night10194 posted:I suppose part of it is also just the uncertainty of waiting to see if the latest major hit actually did anything. I felt the same way right after the DNC until a bunch of polling came out and showed he got destroyed. I follow Predictwise because I think betting markets are a good way to estimate the impact of events before polling can catch up, and also account for the possibility of big surprises that would shake up the race. Oddly, it hasn't moved at all since this tax story came out yesterday, which says the consensus is: this won't really hurt Trump any more than everything else that's happened over the last week.
|
|
#
?
Oct 3, 2016 05:48
|
|
|
Polls as a whole shouldn't really be taken seriously more than a week out from the election. Humans are simply too fickle for them to be reliable this far out. It's a lot more predictive (and reassuring) to see how Romney did in the last election and then try to realistically figure out how Trump wins any states that Romney didn't win. Hilldawg has this in the bag.
|
|
#
?
Oct 3, 2016 05:49
|
|
|
Islam is the Lite Rock FM posted:Vince Foster? "Vince, I need you to fake your death. Only from beyond the grave can you become the most dangerous campaign operative in the world." "Anything for you, Future Madam President."
|
|
#
?
Oct 3, 2016 05:52
|
|
|
NaanViolence posted:Polls as a whole shouldn't really be taken seriously more than a week out from the election. Humans are simply too fickle for them to be reliable this far out. Yes I too trust the predictive power of nebulous speculation over scientific polling and statistical models. 
|
|
#
?
Oct 3, 2016 05:52
|
|
|
lozzle posted:Not a clue. Suspects so far include: The postmark and return address suggested that it had been mailed from Trump Tower.
|
|
#
?
Oct 3, 2016 05:56
|
|
|
lozzle posted:I found a pretty in-depth explanation from Ipsos: It's basically the Bayesian thing of updating your beliefs about the world based on your initial beliefs about the world + the data you observed. For example, say you thought you probably had a fair coin. And then you flip it N times and get H heads, and T tails. Then you can calculate the probability that the actual probability of getting a heads is p, using Bayes rule and your prior of probably having a fair coin. And then you can add up the probability that p = p1, p = p2, p = p3, etc., until you've got the probability that p is in some interval [p1, pn]. And then what you do is find the interval where there's a 95% probability that p is in that interval (and most likely set it up so that there's a 2.5% chance you're above the interval, and a 2.5% chance that you're below the interval). That's all fairly straightforward Bayesian calculations. The big thing is: Ipsos posted:In order for this approach to hold true, one must assume that the sample design is “conditionally ignorable”, meaning that, once we control for the biases mentioned above, there is not a relationship between one’s likelihood to participate in an online survey and the variables we want to measure. Ipsos is taking steps to ensure that its online samples are conditionally ignorable, such as combining multiple opt‐in online panel and non‐panel sources. In essence, by combining multiple sample sources, the “holes” in any one sample source can be filled by the other source. It looks like you really need to make sure you've designed your sampling technique correctly. And I think you'll also need to have a fairly good idea about the current state of the world (in terms of how likely a given poll result is, and how likely an individual based on whatever variables you're controlling for will answer the poll and answer it in a given way).
|
|
#
?
Oct 3, 2016 05:58
|
|
|
Instant Sunrise posted:The postmark and return address suggested that it had been mailed from Trump Tower. The postmark merely confirms that it was mailed from New York and you can write literally whatever you want as the return address. lozzle fucked around with this message at 06:03 on Oct 3, 2016 |
|
#
?
Oct 3, 2016 05:58
|
|
|
Dems got their work cut out for them to convince the younger crowd to vote for Hillary.
|
|
#
?
Oct 3, 2016 06:00
|
|
|
Instant Sunrise posted:The postmark and return address suggested that it had been mailed from Trump Tower. you can write whatever you want on an envelope for a return address, and then just mail it from a collection box in that area an old trick to scam the mail system is to write your address as the sender, and you recipient address as the return address, and then put insufficient postage on the envelope so it is 'returned' to the person you wanted to send it to anyway 
|
|
#
?
Oct 3, 2016 06:01
|
|
|
punk rebel ecks posted:Dems got their work cut out for them to convince the younger crowd to vote for Hillary. That was published September 19th. PPP's polling on the debate said it made like 44% of the millennials watching more likely to vote for her. I don't think this is one to worry about too much.
|
|
#
?
Oct 3, 2016 06:02
|
|
|
|
| # ? May 20, 2024 00:03 |
|
|
punk rebel ecks posted:Dems got their work cut out for them to convince the younger crowd to vote for Hillary. Obama won bigly in 2012 despite white Millennials going for Romney by like +10. They're an important bloc, but not a kingmaker because a huge chunk of them are still in the 18-24 bloc that never loving shows up. Is there data showing that she's down significantly among minority Millennials?
|
|
#
?
Oct 3, 2016 06:02
|
|
