History has really turned against that belief. Clinton won by a lot, and Perot's voters were like Johnson voters in that it was a pretty even split between sides. Perot may have damaged Bush in his run, but Clinton did not win because of Bush votes going to Perot. It would have to be like an 75-25% split, or something, which no polls indicated.
How Clinton lost Michigan and blew the election. Interesting read.
- Thread starter bangbang i'm a timpani
- Start date
- Status
History has really turned against that belief. Clinton won by a lot, and Perot's voters were like Johnson voters in that it was a pretty even split between sides. Perot may have damaged Bush in his run, but Clinton did not win because of Bush votes going to Perot. It would have to be like an 75-25% split, or something, which no polls indicated.
AbortedWalrusFetus
Member
Yes, they needed to maximize turn out, but SHE didn't need to be there to do that, nor did they need to focus so much effort on the area to do it. She supremely over-invested in Philly and ignored the rest of the state. She could have reduced resources allocated to Philadelphia by a quarter and it wouldn't have made a lick of a difference. The democratic machine here is well oiled and gets its work done. Even then, the biggest mistake made here was not giving street money to the local dems. Do that alone and you solve your turnout problem.
whyamihere
Banned
Eh, I don't think the Democratic machine is good enough in any city in any swing state that no one needs to ignore it. She should've gone to Philly a lot. She also should've gone to Erie and Scranton a lot.
lol I have watched that.... it's literally "how Perot stole the election from Bush" until the end where they quote the almighty exit poll and say that's what did it.
Exit polls mean jack shit
Also you can't discount that Perot explicitly targeted Bush far more heavily than Clinton.
whyamihere
Banned
This has been consistently disproven by political scientists since 1992. It is a wrong analysis.
tearsintherain
Member
Did Hillary use the same set of advisors in 2016 as she did in 2008? While ultimately her failures rest on her shoulders, some of the stuff you read in that article makes you wonder at how utterly incompetent her campaign staff and/or the DNC staff must be.
In many ways it mirrors her failures in 2008, where Obama was behind in the primary popular vote the entire way, but strategically chose battleground states to focus on and maximize delegate counts, ensuring his victory. To completely miss so many key things, in hindsight, a second time is pretty crazy.
In many ways it mirrors her failures in 2008, where Obama was behind in the primary popular vote the entire way, but strategically chose battleground states to focus on and maximize delegate counts, ensuring his victory. To completely miss so many key things, in hindsight, a second time is pretty crazy.
If they are so accurate, why did we never have a President Kerry? Why did Trump win? Why didn't Gore win Florida? Exit polls said all these things, but were dead wrong.
They called Florida in 2000 for Gore pretty early due to a big exit poll lead, and later had to rescind it. And don't get me started on 2004.
AbortedWalrusFetus
Member
Funnily enough, before this election democrats were convinced Pennsylvania was not actually a swing state.
I was here for both of Obama's elections and his presence here was nowhere near the level of what Clinton had. The GOTV effort was at levels I've never seen in Philadelphia in my lifespan. She thought the city alone would carry the state and overinvested in it, and underinvested in the rest of the state.
whyamihere
Banned
I didn't say anything about exit polls in the post you quoted, but it's nice for you to edit my own posts with your own thoughts.
She should've done both!
cooljeanius
Member
At one canvass kickoff I went to the organizer emphasized how important it was to stick to the script because they were "scientifically tested" to maximize turnout. That's like the only canvass kickoff I've been to where they've said something like that, usually in previous campaigns organizers trusted canvassers to go off script to say things relevant to the specific person you're talking to. Oh and they also wouldn't let me make my own sign at a rally because people higher up in the campaign had spent "a lot of time and money" perfecting the official sign design. Both instances fit with the narrative of Hillary being too manufactured and "fake" for a lot of people.If GAF ran a campaign.txt
Then what the hell were you talking about? The only reason given for it not being Perot's fault is exit polls.
whyamihere
Banned
The actual thing that I said in the post that you quoted, the consensus from political scientists since 1992.
But what are they basing it on? Exit polls.
Exit polls are bad for demographics, or bad for things that correlate highly with demographics. Random-Sample Polling, Vote Tallies and Voter File analysis are far more useful tools.
For actually determining reasoning and rationales from voters, there really isn't a better source than exit polls. It gets dicey when there's, like, second-order relationships and correlations, but its really the best we have for that kind of thing. Think when the reason for young voters to go to candidate A differs wildly than the reasons older voters go to Candidate A. Then it gets messy.
That's really a stretch of an argument, because turnout in primaries/caucuses differs wildly between states in a way that GE voting doesn't. Caucuses are terrible and totally skew things. And then there's the Michigan shitshow. I think you're really stretching the circumstances to fit your argument.
MI, WI and PA are the focus because Trump underperformed Obama 12 in all three and still won. PA can be argued a bit more because Trump came damn close to Obama's total votes, but WI and MI aren't even close.
FL isn't part of the narrative because Hillary exceeded. I think, anyone's expectations in the state. There's simply nobody, including the GOP, that saw 500k+ voters in the Panhandle coming out of nowhere to vote Trump.
I came across a story about a week before the election that went into detail about what it was exactly that motivated Hillary to use a private email server. The conclusion after pouring through mountains of deposition testimonies is that Hillary's main motivator was an aversion to change and the adoption of new systems. The extent to which Hillary is a luddite when it comes to technology goes well beyond the stereotypical grandma not wanting to figure out computers. She apparently had no interest in upgrading tech from what she was accustomed to and actively fought against having to learn how to use new systems, hence the reliance on a private email server she and Bill have used for years.
This is admittedly a weak link to her campaign choices but it does give you an insight into what kind of person she is. And that's not to say that being tech-averse inherently makes you a bad person, but there's was a stubbornness that played a huge part in her motivations for the private server and I believe it can also be seen in how she constructed a campaign that was incredibly insular and unwilling to challenge their pre-set plans in any way.
foxuzamaki
Doesn't read OPs, especially not his own
Gonna keep it real with you, this is a case of common sense losing out, common sense said that Clinton should have won by every account, even people who wouldn't have voted trump were saying that they most likely wasent going to
People act like it is incompetence but to me this just seems like the inherent flaw in drawing conclusions from statistics and models. What happened in this campaign is the same thing which undermines so many conclusions based on economic, social, and even scientific models based on statistics.
Due to scarce resources in polling and modeling, you distill infinitely complex systems like how people are going to vote on election day to a few categories you can manageably address. You make assumptions about other factors in your model based upon historical precedent and trends. Basically you focus on things that have a wide range of possible outcomes and ignore the things that have a very predictable outcome. You say for example, we know with 95% confidence that black people will turn out at a rate of 60%+/- 4% and vote 95% for the democrat +/-1 5% so we are not going to do much to influence that group. We know that rural evangelicals turn out at 40% +/- 15% and vote 40% democratic +/- 10%. That is a high degree of variability so we are going to focus on that group.
Now you have tens if not hundreds of assumptions you make due to the fact you can only poll so many people a day, and you have even more assumptions you make that you don't even realize you are making. (A huge snowstorm doesn't shut down Philly turnout.) Even if results only have a 1% chance of being wrong, you have 100s of these so some of them are going to be wrong. What typically happens is they are not all wrong in the same direction, so you never even notice. But every now in then you get results outside of the model. And by the way, the field of statistics is all about quantifying the probability of these type of results. It is just people inherently interpret something that has a 99% certainty to meaning it has 100% certainty ignoring the fact that when you have 1000 things of 99% uncertainty that a handful of them are going to be out of range all the time.
The solution I see advocated is that you throw out what statistics and modeling tells you in favor of anecdotes. A lot but not all the critiques of the campaign seem based on this. They should have recognized these anecdotal examples as being indicative of flawed assumptions. Yet to me that seems worse, you would have a very reactionary campaign based on nothing. What basis do you have on trusting some anecdotes over others. Volunteer X says a 20 year democrat is going to vote for Trump, the sky is falling. Volunteer Y says a 20 year Republican is voting for Hillary, we are doing it guys!
TL/DR statistics and models based on historical trends cannot predict outcomes with 100% certainty because of limited historical data and limited budgets for polling, but it is better than running a campaign reacting to anecdotes.
The ignoring of primary results criticism I can see. But it is still a different situation. There is a huge difference between voting for Bernie and voting for Trump. But probably the real choice was voting for Hillary or not voting at all. What they probably should have done is put turnout numbers at the lower bound of possible outcomes, if not even lower to be conservative and run that as a worse case model. They probably thought Trump was so bad though that turnout couldn't possibly be as low as it ended up being.
Due to scarce resources in polling and modeling, you distill infinitely complex systems like how people are going to vote on election day to a few categories you can manageably address. You make assumptions about other factors in your model based upon historical precedent and trends. Basically you focus on things that have a wide range of possible outcomes and ignore the things that have a very predictable outcome. You say for example, we know with 95% confidence that black people will turn out at a rate of 60%+/- 4% and vote 95% for the democrat +/-1 5% so we are not going to do much to influence that group. We know that rural evangelicals turn out at 40% +/- 15% and vote 40% democratic +/- 10%. That is a high degree of variability so we are going to focus on that group.
Now you have tens if not hundreds of assumptions you make due to the fact you can only poll so many people a day, and you have even more assumptions you make that you don't even realize you are making. (A huge snowstorm doesn't shut down Philly turnout.) Even if results only have a 1% chance of being wrong, you have 100s of these so some of them are going to be wrong. What typically happens is they are not all wrong in the same direction, so you never even notice. But every now in then you get results outside of the model. And by the way, the field of statistics is all about quantifying the probability of these type of results. It is just people inherently interpret something that has a 99% certainty to meaning it has 100% certainty ignoring the fact that when you have 1000 things of 99% uncertainty that a handful of them are going to be out of range all the time.
The solution I see advocated is that you throw out what statistics and modeling tells you in favor of anecdotes. A lot but not all the critiques of the campaign seem based on this. They should have recognized these anecdotal examples as being indicative of flawed assumptions. Yet to me that seems worse, you would have a very reactionary campaign based on nothing. What basis do you have on trusting some anecdotes over others. Volunteer X says a 20 year democrat is going to vote for Trump, the sky is falling. Volunteer Y says a 20 year Republican is voting for Hillary, we are doing it guys!
TL/DR statistics and models based on historical trends cannot predict outcomes with 100% certainty because of limited historical data and limited budgets for polling, but it is better than running a campaign reacting to anecdotes.
The ignoring of primary results criticism I can see. But it is still a different situation. There is a huge difference between voting for Bernie and voting for Trump. But probably the real choice was voting for Hillary or not voting at all. What they probably should have done is put turnout numbers at the lower bound of possible outcomes, if not even lower to be conservative and run that as a worse case model. They probably thought Trump was so bad though that turnout couldn't possibly be as low as it ended up being.
Rentahamster
Rodent Whores
I warned ya!
They got cocky.
They got cocky.
hydragonwarrior
Member
I hate referring to the "campagn". Blame should be at actual people within that campaign - like, the managers who refused to listen or the advisors who never thought twice about their strategy.
TBH I was thinking it was simply the Obama campaign team that migrated over to help Hillary - unless it IS and they simply got arrogant from the 2012 win...
TBH I was thinking it was simply the Obama campaign team that migrated over to help Hillary - unless it IS and they simply got arrogant from the 2012 win...
That, and the fact that for a period of time, he stopped running.
This is fundamentally not true.
https://twitter.com/NateSilver538/status/809105926993178624
"A coherent case is that the "fundamentals" (no incumbent, middling economy, high partisanship) predicted a close race & that's what we got."
"Clinton actually outran most of the econometric models—which predict the popular vote & not the Electoral College—by a couple points."
The assumption that Trump is a weak candidate discounts partisanship, or that the reasons liberals and democrats hate him are reasons that his base would also hate him. Clearly, the latter wasn't true; Republicans have no qualms about voting for an admitted serial sexual-assailant, who spews constant & unending lies, or is blatantly racist, or blatantly sexist.
Both of which had justification. Caucuses are an un-democratic trainwreck, so drawing conclusions from them is just dumb and misleading. Michigan has had 1 primary in the previous 16 years, so the demographic weightings were wildly inaccurate.
hydragonwarrior
Member
And with absolutely nothing to challenge it. The internet can't do that for people in the boonies - that requires mandatory feet on the ground to reach out to them.
jchap
Member
Clinton - Trump: 0 - 1
1. e4 ... e5
2. Qh5 (Send the bus back to Iowa it's time for trickery) ... Nc6 (Shit)
3. Bc4 (We still got this) ... g6 (NOOOO)
4. Qf3 (PLZ WORK) ... Nf6 (resigns)
Netherscourge
Banned
I said it before. I'll say it again.
She lost because the last bit of information to come out about the candidates was the FBI reopening it's investigation against her.
Had the media outlets sat on those Trump tapes until the last week prior to the election, Trump would have lost. It would have been fresher in everyone's minds than the stupid email investigation.
Undecided voters are undecided for a reason - they wait until the very last moments to make a decision. And because of that, the most recent bit of information that was circling in the media just days/hours before Election Day was the stuff about Hillary and Emails.
All those nasty tapes with Trump's misogynistic comments fizzled out right after the debates and everyone stopped caring in the following weeks. Then, the FBI reopened their investigation and all attention was back on Hillary. And that was the freshest bit of information on the undecided voter's minds going into the voting booths.
She lost because the last bit of information to come out about the candidates was the FBI reopening it's investigation against her.
Had the media outlets sat on those Trump tapes until the last week prior to the election, Trump would have lost. It would have been fresher in everyone's minds than the stupid email investigation.
Undecided voters are undecided for a reason - they wait until the very last moments to make a decision. And because of that, the most recent bit of information that was circling in the media just days/hours before Election Day was the stuff about Hillary and Emails.
All those nasty tapes with Trump's misogynistic comments fizzled out right after the debates and everyone stopped caring in the following weeks. Then, the FBI reopened their investigation and all attention was back on Hillary. And that was the freshest bit of information on the undecided voter's minds going into the voting booths.
Rentahamster
Rodent Whores
lol, like I'm going to take branding advice from the genius team who came up with gems like:
"I'm with her" (way too self-centered, not inclusive)
"Love Trumps Hate" ("Love Trump" is literally in this slogan, it doesn't parse well for many people and sounds like "Love Trump's Hate")
...and who made the awesome official logo that looks like it's pointing which way to go to get to the hospital.
I agree with this, but there is some bit of truth to the argument that there could have been a better plan to respond to October Surprises (even completely BS ones), or that it "should" have been a safer lead, or some oppo could have been stored until the end.
But, yeah, Comey letter + horrible media response to it unambiguously cost ~0.7% of the vote. I don't even see how it is debatable. And were it not for that, we'd be talking how brilliant her resource assignments were, having won ~6 or 7 states by 0-4%.
GoldenEye 007
Member
IIRC, they had data that showed them greatly over performing their projections in the rust belt. To the point they didn't truly believe them until election night.
Here's one article on Trump data: http://mashable.com/2016/11/10/donald-trump-polling-data/#DLt.cqtGASqk
And another: https://www.bloomberg.com/news/arti...m-saw-a-different-america-and-they-were-right
KO Traveling Hobo
Member
A red arrow cutting straight through Hillary to get to the White. Nailed it.lol, like I'm going to take branding advice from the genius team who came up with gems like:
"I'm with her" (way too self-centered, not inclusive)
"Love Trumps Hate" ("Love Trump" is literally in this slogan, it doesn't parse well for many people and sounds like "Love Trump's Hate")
...and who made the awesome official logo that looks like it's pointing which way to go to get to the hospital.
edit: I could even see it being a Truther fuel tank with its depiction of a pointed object crashing through a pair of tall rectangles.
A Fish Aficionado
I am going to make it through this year if it kills me
Brexit told that, but we still believed.
Suikoguy
I whinny my fervor lowly, for his length is not as great as those of the Hylian war stallions
When are people who keep bringing up Brexit going to actually look at the polls taken before Brexit?
It's sad.
Enosh
Member
it makes sense if your polls show you leading in all the other places that you want to win and your goal isn't to just win but to completely discredit your opponent and his views
Hillary was obviously going for a big landslide win, hubris or trying to show that Trumps views have no place in America? idk probably a bit of both
but Trumps team also saw something happening in MI, PA, NH, MI and WI, it's the reason they did the crazy last 2 weeks of the campaign there full of events while everyone was mocking them for wasting time in blue states
Lot to unpack here, so I'll be brief.
The main data based sign that was missed was that a red flag should have gone up about being down in Ohio and Iowa so much, but yet the polls showing that she was so far ahead in MN MI WI and PA. You had to believe that the states were not correlating this year nearly as much as they had in the past, or that the polls in one of those two areas were wrong. Hindsight being 20/20, you could have assumed that Ohio and Iowa polling was far more likely to be accurate (due to their history as swing states) then Wisconsin and Michigan.
The issue with the "demographic victory" that the Democratic party seems to be aiming for is that as voting blocs get bigger, they tend to vote less like a bloc. So as the hispanic / asian voting bloc gets bigger (and older, notably), the idea that they will vote at the same percentages for the Democratic party is one that is based on some assumptions that may not hold true.
The main data based sign that was missed was that a red flag should have gone up about being down in Ohio and Iowa so much, but yet the polls showing that she was so far ahead in MN MI WI and PA. You had to believe that the states were not correlating this year nearly as much as they had in the past, or that the polls in one of those two areas were wrong. Hindsight being 20/20, you could have assumed that Ohio and Iowa polling was far more likely to be accurate (due to their history as swing states) then Wisconsin and Michigan.
The issue with the "demographic victory" that the Democratic party seems to be aiming for is that as voting blocs get bigger, they tend to vote less like a bloc. So as the hispanic / asian voting bloc gets bigger (and older, notably), the idea that they will vote at the same percentages for the Democratic party is one that is based on some assumptions that may not hold true.
Suikoguy
I whinny my fervor lowly, for his length is not as great as those of the Hylian war stallions
Also, the best way to make gains in the Senate.
If the polling was right, this would have be lauded as a fantastic strategy and victory as Dems would likely have a majority in the Senate.
Yup. Senate & President were perfectly correlated for the first time ever. If she had won the Presidency, she'd also have the Senate.
Poli-GAF is about the feels, not the data.
Dash_Riprock
Member
Hilary truly was the Romney of this election, right down to having a fancy fireworks show planned which was canceled at the last minute once polls numbers started coming in.
That was because his own polling showed him winning in a landslide.
That was because his own polling showed him winning in a landslide.
Suikoguy
I whinny my fervor lowly, for his length is not as great as those of the Hylian war stallions
No, Romney believed the polling was wrong that showed Obama winning.
There probably would not have been a perfect correlation had Clinton won - IIRC Johnson outperformed Trump in Wisconsin against Feingold.
She lost to the "grab em by the pussy" guy, bro
"Deplorables" was identical to the "47%" gaffe and's crazy how so many people failed to see the similarity. Many of the same people who berated Romney for dismissing so much of the electorate defended the deplorables comment when both statements were functionally true and both objectionable for the same reasons. Hillary's was probably worse because of the immediate negative connotation with "deplorable."
No one ever comes to Washington state (except Trump -- who got crushed like a bug*) so it's always interesting reading what it's like for the states where the candidates are actually expected to campaign. Kind of a window into an alternate dimension.
*I think Hillary only came for a fundraiser
*I think Hillary only came for a fundraiser
SodiumBenzoate
Member
Jezebel just posted a recap of this article that ends with a pretty good recap at the end of how I sort of feel:
The real gem is the comment section; here's a comment that echoes all the other anecdotes I have seen elsewhere, in the Politico article, this Jezebel article, and in many other places: (Hey, it wasn't just Michigan!)
and I think this zinger says it all.
The real gem is the comment section; here's a comment that echoes all the other anecdotes I have seen elsewhere, in the Politico article, this Jezebel article, and in many other places: (Hey, it wasn't just Michigan!)
and I think this zinger says it all.
It is true that the Hispanic vote will become more split as it gets larger, but that will be in part because they have a larger say in politics. In the two party system, there will always be a back and forth battle between two parties, while a certain party can dominate for a decade or 2 nationally (Republicans in the post-reconstruction era and Democrats following the Great Depression until the end WWII), there will always be State and local competition between parties and the other party will find a way to become nationally relevant again. Parties are in it to win so they will attempt to appeal to enough voters to win elections.
The Democrats have been around since the 1700's, and they even managed to "elimate" the opposing party, but that lasted maybe 6 years before a new party took the Federalist place. We will always return to 2 strong parties.
So while it is almost assuredly that Hispanic voters will break toward less monolithic than they are now, it will be because they vote against the GOP and actually force the GOP's hand in addressing their issues. Parties have also flipped from being conservative and liberal, it is entirely possible that in 2050 the Democrats are the Conservative party and Republicans are the liberal party. Remember, the parties only care about one thing: maximizing their vote total, every thing they say and do is toward that end.
hodayathink
Member
Johnson barely outperformed Trump, IIRC. If she wins, Dems definitely win Pennsylvania, more than likely win Wisconsin, and probably win Missouri (where she would have lost but Kander ran so much ahead of her that he probably would have pulled it out). Which gives Dems a 51-49 Senate where they don't even have to rely on Kaine as a tie-breaker.
Felt like she got the Obama b-team mixed w/ her own people.
She lost because being in a state does not mean being in the area you are most likely to win people over to your view.
Dems like to talk how she campaigned on the economy etc. But truth is her entire campaign was a get out the vote campaign - she never really campaigned against Trump - just played to her crowd.
Dems like to talk how she campaigned on the economy etc. But truth is her entire campaign was a get out the vote campaign - she never really campaigned against Trump - just played to her crowd.
MrCunningham
Member
And I'm sure that Trump doesn't understand how to play Chess at all.
boiled goose
good with gravy
When you lose the the most unpopular candidate in History who is also grossly incompetent and has less money than You...
YOU BLEW THE ELECTION.
- Status