|
Schwarzwald posted:Wouldn't people betting on who they want to win rather than who they expect to win make them more in line with polls, rather than otherwise? My guess is that Trump supporters are disproportionally emotionally invested in their candidate winning. Just look at how many of them flat-out reject the positive sign for Clinton and clamour to the 'monster vote'. But I admit I have no statistical data to back up my assertion.
|
#
?
Nov 6, 2016 00:36
|
|
|
|
| # ? Jun 8, 2024 21:01 |
|
|
lol at how nate thinks 538 exists to spread some message about trusting data and empiricism and math and poo poo we all know the only reason why it's here today is just so liberals can watch their numbers go up
|
|
#
?
Nov 6, 2016 00:43
|
|
|
e: wrong thread
|
|
#
?
Nov 6, 2016 00:43
|
|
|
tankadillo posted:we all know the only reason why it's here today is just so liberals can watch their numbers go up It's pretty terrible at that then, given how it gives Trump decent odds.
|
|
#
?
Nov 6, 2016 00:44
|
|
|
Schwarzwald posted:It's pretty terrible at that then, given how it gives Trump decent odds. yeah funny how that's exactly when everyone suddenly realized he's not so hot anymore
|
|
#
?
Nov 6, 2016 00:59
|
|
|
tankadillo posted:yeah funny how that's exactly when everyone suddenly realized he's not so hot anymore I don't think I get where you're going with this. Are you saying Nate or 538 or whatever is loving up by not just giving liberals what they want, or that liberals were only ever in favor of 538 when it was showing good numbers for the Democrats (and have turned now that it's model is less positive), or something else?
|
|
#
?
Nov 6, 2016 01:16
|
|
|
Nate is more shook than a SIDS baby in a paint mixer.
|
|
#
?
Nov 6, 2016 01:16
|
|
|
FMguru posted:Nate is more shook than a SIDS baby in a paint mixer. His coworkers on 538 are talking about it on their latest podcast. It's hilarious.
|
|
#
?
Nov 6, 2016 01:21
|
|
|
Schwarzwald posted:I don't think I get where you're going with this. Are you saying Nate or 538 or whatever is loving up by not just giving liberals what they want, or that liberals were only ever in favor of 538 when it was showing good numbers for the Democrats (and have turned now that it's model is less positive), or something else? The latter. I think Nate is pretty well-intentioned in his passion for numbers and data, but he rose to his current level of popularity because he gave his audience what they wanted. The fact that he's not doing that anymore, combined with the mediocre journalism of this year's 538, means that there's not much to cheer for.
|
|
#
?
Nov 6, 2016 01:28
|
|
|
https://twitter.com/NateSilver538/status/795041766877646848 this meltdown is so delicious edit: also https://twitter.com/NateSilver538/status/795015340539473924 huge pile of hamburger has issued a correction as of 01:56 on Nov 6, 2016 |
|
#
?
Nov 6, 2016 01:29
|
|
|
https://twitter.com/SamWangPhD/status/788544053059153924?ref_src=twsrc%5Etfw will nate silver eat a bug if he's wrong? doubtful. sam wang once again proves superior
|
|
#
?
Nov 6, 2016 03:21
|
|
|
huge pile of hamburger posted:https://twitter.com/NateSilver538/status/795041766877646848 Think the polls are skewed on that one
|
|
#
?
Nov 6, 2016 03:41
|
|
|
Concerned Citizen posted:https://twitter.com/SamWangPhD/status/788544053059153924?ref_src=twsrc%5Etfw Nate basically can't be wrong with his model since it predicts that we don't really have that much information at all from polling. If that's the case then he'll be vindicated and basically be out of a job.
|
|
#
?
Nov 6, 2016 03:43
|
|
|
I got the Huffpo dude to retweet the Medium article while Nate was melting down at him, so there's a non-zero chance he read it.
|
|
#
?
Nov 6, 2016 05:08
|
|
|
Mukaikubo posted:poo poo, sorry, misinterpreted what you wrote then! I don't think you understand statistics. That article is imprecise in some of its language, but first of all, if you are going to call Wang's 99% chance flat out bullshit, show your math, because you clearly don't get margin of errors. You may disagree with some of their assumptions, but the claim that they are saying "that systematic polling errors like have happened a number of times in recent history are impossible" is laughable. Second, the article isn't calling for 99%. Their model and the upshot are in the 80s. Finally, they correctly call out Nate on two things: trend line adjustment and poll adjustment. Those two things are bullshit that is being done in a proprietary model that is not open to scrutiny. Nate claiming that he does those things based on empirical evidence is something that you have to take at his word, because people can't see it. In any case, regardless of the outcome of the election, this cycle has shown that Nate's model is underconfident by default and therefore useless. Hillary has ranged from underdog to almost 90% favorite at 538 despite poll aggregates never having her trailing (RCP is an outlier aggregator because they cherry pick right leaning polls and even then they have Hillary leading virtually the entire race). joepinetree has issued a correction as of 06:14 on Nov 6, 2016 |
|
#
?
Nov 6, 2016 05:29
|
|
He isn't the be all and end all of forecasting, but he never was. He's just a dude who knows some about stats who is good at communicating them to the public. That's not a skill set to be thrown out.
|
lock him up
|
|
#
?
Nov 6, 2016 06:02
|
|
|
Concerned Citizen posted:https://twitter.com/SamWangPhD/status/788544053059153924?ref_src=twsrc%5Etfw eat a bug? no. catch a bug? maybe.
|
|
#
?
Nov 6, 2016 06:09
|
|
|
Did we ever find out what happened in the Michigan primary cuz that was p weird
|
|
#
?
Nov 6, 2016 09:59
|
|
|
androo posted:I got the Huffpo dude to retweet the Medium article while Nate was melting down at him, so there's a non-zero chance he read it. Beautiful.
|
|
#
?
Nov 6, 2016 10:19
|
|
|
Tayter Swift posted:Did we ever find out what happened in the Michigan primary cuz that was p weird the podesta emails actually answered this - public polls predicted an electorate that was older and blacker than it really was. hfa actually had more accurate polling #s showing it was real close
|
|
#
?
Nov 6, 2016 14:45
|
|
|
Credit where it's due, this is actually pretty funny.
|
|
#
?
Nov 6, 2016 15:10
|
|
|
huge pile of hamburger posted:https://twitter.com/NateSilver538/status/795041766877646848 
|
|
#
?
Nov 6, 2016 15:41
|
|
|
CottonWolf posted:This seems like the place to ask people's opinions on this: There might be some hedging going on inflating Trump's odds, i.e. "If Trump wins my life is ruined but I'll have made a couple of thousand bucks on the bet to move out of the country with."
|
|
#
?
Nov 6, 2016 15:50
|
|
|
I just listened to their podcast and one of their speakers called their sandwich "a Jairo" deleted and unsubscribed.
|
|
#
?
Nov 6, 2016 15:53
|
|
|
|
|
#
?
Nov 6, 2016 16:24
|
|
|
Antti posted:There might be some hedging going on inflating Trump's odds, i.e. "If Trump wins my life is ruined but I'll have made a couple of thousand bucks on the bet to move out of the country with." And the fact that Trump is basically a walking antithesis to any polling that was against him in the primary
|
|
#
?
Nov 6, 2016 16:39
|
|
|
Dead Cosmonaut posted:And the fact that Trump is basically a walking antithesis to any polling that was against him in the primary trump was ahead in polling like the entire gop primary
|
|
#
?
Nov 6, 2016 16:46
|
|
|
The problems with Nate's model have been well documented by now, but the garbage input part is underserved. Here's a Nevada poll that came out yesterday: https://drive.google.com/file/d/0B4lhKxf9pMitYTV0T1hDMnpDVkU/view This thing is Trump+5 because... quote:Using automated voice response technology allows each participant to hear the questions exactly the same way, from the same voice, spoken with the same tone and nuance. Calls are placed to randomly-selected phone numbers through a stratified process that insures approrpiate {sic} geographic representation. These calls were placed from 6:00pm to 8:30pm local time from Tuesday, November 1st though Friday, November 4th. Yeah, they called landlines between 6 and 8:30 PM and then asked how the neighbors voted and threw the answers into the topline. Trump magically gets 24% of blacks, 37% of Hispanics and 53% of Asians. Nate gives this poll a weight of 2.21 because we never exclude a poll and the bad ones will get averaged out by the good ones (they're not being averaged out by the good ones).
|
|
#
?
Nov 6, 2016 16:49
|
|
|
Eifert Posting posted:I just listened to their podcast and one of their speakers called their sandwich "a Jairo" Yo when even fuckin Arby's gets this right this is shameful
|
|
#
?
Nov 6, 2016 17:03
|
|
|
theflyingexecutive posted:Yo when even fuckin Arby's gets this right this is shameful I worked at a Greek gyro place for 2 years and the only thing that got our owner more riled up then people mispronouncing gyro was a worker spending more than one minute making one... I got a free one for dinner everyday I work today though and they were the best I ever had. A+ would work again.
|
|
#
?
Nov 6, 2016 17:08
|
|
|
I worked for Greek people at both their annual huge church festival and for catering. Oh man it really made the Greek financial crisis understandable
|
|
#
?
Nov 6, 2016 17:11
|
|
|
Oh man, I still dream of my hometown's local Greek gyro place in Munster, IN just under the protecting eaves of Chicago. We also have a great Lebanese gyro place in town here, in central IL of all places. I'm still not used to how they grill the meat instead of just slicing it straight onto the pita, but it's still pretty delicious and the owner is a great dude. Really all I want is a giant mountain of gyro meat with enough feta to flavor the whole thing. I don't need the sauce or anything else. Well, black olives are okay I guess. And if Nate Silver or one of his cronies pronounces it like the first part of "gyroscope" they are dead to me.
|
|
#
?
Nov 6, 2016 17:43
|
|
|
dwarf74 posted:Oh man, I still dream of my hometown's local Greek gyro place in Munster, IN just under the protecting eaves of Chicago. That's schwarma dumb rear end.
|
|
#
?
Nov 6, 2016 17:47
|
|
|
Bip Roberts posted:That's schwarma dumb rear end. Nah, it's very definitely gyros with some shawarma-like preparation thrown in. http://therockrestaurant.net/our-menu/
|
|
#
?
Nov 6, 2016 18:20
|
|
|
dwarf74 posted:Nah, it's very definitely gyros with some shawarma-like preparation thrown in. Oh, what's up fellow Bloomington-Normal goon? The Rock is definitely the best around here, but Uncle Nick's is the only reason to ever go to Rockford. It's worth it to check that place out if you're ever anywhere near that shithole of a town. http://www.seeyounextgyros.com/welcome.html
|
|
#
?
Nov 6, 2016 20:43
|
|
|
Bip Roberts posted:That's schwarma dumb rear end. also lol
|
|
#
?
Nov 6, 2016 21:09
|
|
|
Adar posted:The problems with Nate's model have been well documented by now, but the garbage input part is underserved. Here's a Nevada poll that came out yesterday: Nate includes google consumer surveys in his model as well. He gives Trump a 10% chance in loving Delaware because 2 google consumer surveys with samples of 74 and 107 had Hillary ahead by just single digits (and one was somehow adjusted to Trump +1). Besides the sample size, google consumer surveys are those surveys you are forced to do to see certain webpages, and they then estimate demographics from Google information.
|
|
#
?
Nov 6, 2016 21:27
|
|
|
Lager posted:Oh, what's up fellow Bloomington-Normal goon? Can't say I spent much time in Rockford. But if I'm up there I'll do it!
|
|
#
?
Nov 6, 2016 21:45
|
|
|
Patter Song posted:Polling is only going to get shittier every cycle. By 2024, polls will all be "poo poo, I don't know." Nah, by 2024, Facebook will release some kind of polling mechanism and it'll be Google Consumer Surveys but actually effective and polling will return to being fairly accurate.
|
|
#
?
Nov 7, 2016 00:36
|
|
|
|
| # ? Jun 8, 2024 21:01 |
|
|
lol @ the idea of elections still existing or mattering in 2024 
|
|
#
?
Nov 7, 2016 00:51
|
|