• Hey, guest user. Hope you're enjoying NeoGAF! Have you considered registering for an account? Come join us and add your take to the daily discourse.

Analysts were blind to one of the biggest upsets in primary history. What went wrong?

Status
Not open for further replies.

injurai

Banned
How is that your takeaway when superdelegates haven't come into play yet?

This logic sounds like something the banks said back in 2007, when defaults were looming but hadn't happened yet.

Do you really think I would be unprincipled enough to say having superdelegates is democratic having waited to see if they went my way?
 

womfalcs3

Banned
They were wrong about this, but Clinton still came out on top when you add the delegates and super delegates she's won. Sanders doesn't have a chance at victory.
 
My major takeaway in this election, is how undemocratic superdelegates are. I'm disgusted by them.

Superdelegates have had no effect on the election.

Clinton is winning in pledged delegates which are awarded on the proportion of the actual vote
 

injurai

Banned
Superdelegates have had no effect on the election.

Clinton is winning in pledged delegates which are awarded on the proportion of the actual vote

Well that is patently false. They can swing the popular primary vote to in the primaries upon the party's convention.
 

royalan

Member
Well that is patently false. They can swing the popular primary vote to in the primaries upon the party's convention.

But they haven't come into play yet. Hillary is winning among pledged delegates. You can't just hand wave that.
 

Xe4

Banned
Yeah its pretty shocking. I wasn't expecting it at all myself. I think it had to so with the fact that it was an open primary and that indipendants skew heavily in favor of Sanders.

Also Michigan is in the shit heaps and that probably skews towards populist candidates (Trump, Sanders) a lot.

Whether being open will also effect Ohio and others, remains to be seen, but the night was definetly intresting.

Congrats to Sanders and co. Lets see if it goes anywhere.
 
The thing to remember is that the polling has been pretty dead-on in most of the contests this year (and in fact called the GOP side perfectly tonight) so beware of any analysts or loud Facebook friends insisting that polling is now meaningless ("People have cell phones now! You can't poll them!"). That's just not true.

Now, if a week from now the polls in Illinois and the rest are just as wrong, they'll definitely need to adjust their models. But weird outlier data points are just part of life. Could have been due to a lot of cross-voting due to it being an open primary, could have been a ton of late deciders, could have been a broken likely voter model, could be something unique to the demographics there that made it hard to gauge support.

But as far as I can tell, the Bernie fans on Reddit (admittedly some of the worst) seem to be following this logic:

Polls said Clinton had a 99% chance of winning Michigan, but she narrowly lost

Therefore

Polls saying Clinton has a 99% chance of winning Illinois/Ohio/Florida mean Sanders has a 100% chance to win.

And of course that's silly. And yes, even if he repeats this improbable scenario in those states, he still loses - he can't make up a 214 delegate deficit without some huge wins in some big states. And by "huge" I mean "huge margins" not "narrow wins that are huge in comparison to expectations."
 
Thing about it is, after next Tuesday, Sanders could win 12 states in a row. And if the trend of him winning big in the smaller stakes by 30 points continue, he will rack up delegates fast.

Ooooooh, I was looking at the numbers and was surprised there was still so much excitement even though today still increased Hillary's lead. How good are his chances really?

I'll be honest I thought Bernie was gonna fade. Fair play to his supporters, they've been hustling like hell.
 
Ooooooh, I was looking at the numbers and was surprised there was still so much excitement even though today still increased Hillary's lead. How good are his chances really?

I'll be honest I thought Bernie was gonna fade. Fair play to his supporters, they've been hustling like hell.

Well, this is what 538 put out about Sanders' best chances to win.

sliver-sandersbenchmarks-1.png

3/22 - 4/19 could bring him a really big streak. And the thing about it is, I don't really know if she can turn around the Idahos, Utahs, Dakotas, of the world. I know they are really small, but when he nets 10 delegates from them, they starts to add up. Then you take into consideration Washington state, Oregon, Wisconsin, all three have modest amount of delegates to them and are heavily white. Though, I don't think they are caucus states. It will be interesting to see how far the anti-hillary vote goes in these smaller states.

One thing diffrent between this and the 2008 primary is, Obama one the western smaller states.
 
Re: Super delegates

Didn't Hillary lock up a ton of them early on in the 2008 race, but then they changed their loyalty to Obama when the polls started going in his direction?

Honestly, I've been kind of like "what?" when it comes to superdelegates, but this makes a bit more sense. I wonder what kind of shift, if any, may happen if Bernie really gets going.
 
I would be very interested in seeing some of sort statistical research or scientific polling indicating, nationally, how many non-Trump-supporting Republicans would vote for Sanders in a Trump/Sanders general election.

Everything I have seen is purely anecdotal, but supportive of the theory that, at the very least, some sort of measurable amount of non-Trump-supporting Republicans would vote against their own party if Sanders were the Democratic nominee, versus those same voters abstaining from voting altogether if the match-up was Trump/Clinton. I have seen online polling that supports this theory, but nothing concrete.
 

Freshmaker

I am Korean.
I'd like to see someone analyze how accurate his analysis is relative to the other analyses we've seen like 538. Getting Michigan right is impressive. Honestly I really am interested in finding out just why every single poll in Michigan was so wrong.

Landline phone surveys were probably the bulk of their polling data. That tends to trend much older.
 

Joeytj

Banned
So with the news that Bernie has won the Michigan primary, it comes as a shock to all of us... except for one person. His name is Tyler Pedigo.

I've been following Tyler's predictions for a couple of weeks now, and I've noticed that his statistical model has been getting astonishingly more accurate with every primary/caucus day. He predicted every Super Tuesday state with the exception of Minnesota, and he's corrected predicted every state since then, with fairly similar margins to the actual results.

As I understand it, Tyler uses a combination of social media and Google trends data points, and refines his predictions by honing in on what variables are key to showing how likely a state is going to vote for one candidate over the other.

Well, yesterday, he looked at his data, as was surprised to find that it was predicting an upset, despite the poll numbers.

Here's what he had to say about it:




Link


What's incredible is that this guy doesn't rely on polls; just the internet and some fancy math.

Personally, I think that he should be on national television explaining his model, after seeing tonight's results.

Whoa! Awesome. Watch out Nate.

On the issue that Bernie might have made "inroads on African Americans", Silver pointed out earlier too that there was never much data to show that northern African Americans were voting for Hillary by the same huge margins than those in the South, and her huge win in Mississippi proves that there's some geography involved too.

Big win for Bernie, but not the game-changer that he needs. Not yet.
 

Xe4

Banned
Ooooooh, I was looking at the numbers and was surprised there was still so much excitement even though today still increased Hillary's lead. How good are his chances really?

I'll be honest I thought Bernie was gonna fade. Fair play to his supporters, they've been hustling like hell.

Still pretty bad. He had a good night tonight, but him winning the primary would make tonight look like nothing in comparison to the fluke that that would be. He needs HUGE wins in a lot of states.

There will be a clearer answer on what Bernie's true chances (or lack of chances) are. It depends on whethe he can tie/barely win/barely loose Ohio and Florida.

His chances are a lot better tonight than they were last night, but it's now a 5% instead of a 1% chance.
 

lenovox1

Member
Well good if they don't. The fact that they exist systemically undemocratic. All I stated was such. They simply should not be a thing.

Don't be fooled into thinking that a private non-profit operation's selection for whom will be their Presidential candidate is in any way inherently democratic. It isn't necessarily intended to be.
 

Xe4

Banned
Don't be fooled into thinking that a private non-profit operation's selection for whom will be their Presidential candidate is in any way inherently democratic. It isn't necessarily intended to be.

Yeah. The primaries have never been democratic. Nor do they have any responsibility to be so. The US has one of the most open primary systems as it is. The Super Delagates and brokered conventions exist to self preserve the party.

Landline phone surveys were probably the bulk of their polling data. That tends to trend much older.
There's probably a lot of reasons pollsters were wrong but this is not one of them. Yes they largely do landlines, but they account for them, and this does not affect the numbers much.
 
Whoa! Awesome. Watch out Nate.

On the issue that Bernie might have made "inroads on African Americans", Silver pointed out earlier too that there was never much data to show that northern African Americans were voting for Hillary by the same huge margins than those in the South, and her huge win in Mississippi proves that there's some geography to involved too.

Big win for Bernie, but not the game-changer that he needs. Not yet.


Also I'm not convinced losing the vote by 70% should be called making inroads.
 
Also, this is, what I consider to be, a very good read, albeit a bit of a long one: http://www.salon.com/2016/03/08/hil...re_rushing_to_nominate_the_weakest_candidate/

A few points to consider, from the article:

Clinton owes some of her early success to the frontloading of Southern states. Super Tuesday is a scheme hatched in the ’80s by a bunch of white, male, mostly Southern Democrats who thought a regional primary would help “centrists” like themselves get a leg up on liberals. But they forgot, not for the first or last time, about African-Americans, lots of whom live in the South and vote Democratic. In ’88, Jessie Jackson and Al Gore split the region, thus allowing Northern social liberal Mike Dukakis to slip through the net.
This year Super Tuesday finally worked as planned; hindering a progressive, aiding an insider. There was a twist: African-Americans who now dominate the party in the South made it work. I doubt they prefer Clinton’s neoliberalism to Sanders’ democratic socialism. The win owed more to loyalty to Obama and other trusted leaders, and to Hillary’s skills and connections. By Saturday, eight of the 11 states of the old Confederacy had voted. In them she won 68 percent of the vote. Ten of 39 states outside the South had voted. In those states Sanders took 57 percent of the vote. On March 15, the Confederacy will be all done voting. The race begins then
 

Hazmat

Member
To be fair, analysts weren't blind to an upset, they created the upset by being so wrong.

It was a staggering failure on their part though. I'd hate to be one of them tomorrow morning, but they all got it wrong so I doubt anyone gets taken to task over it.
 

royalan

Member
Honestly, I've been kind of like "what?" when it comes to superdelegates, but this makes a bit more sense. I wonder what kind of shift, if any, may happen if Bernie really gets going.

It's not impossible.

What Bernie Sanders would need to do to possibly get super-delegates to switch is this:

1) Win from here on out. Bernie needs to gather a string of major victories against Hillary from here on out. And not tiny wins like tonight. He needs major blowouts that net him enough delegates to close in the on Hillary's lead.

2) He needs to grow his coalition. Winning alone won't be enough at this point, especially if the wins only happen because we're moving into whiter parts of the country. Like it or not, the minority vote is important for Democrats, especially in the general. A political revolution just will not happen on the backs of young white liberals. He needs to show he can mobilize minority voters in the same way Obama and Hillary have.

3) And this might be the MOST important thing: Bernie needs to show that he is a True Blue Democrat, and not just using our brand to legitimize his run for office. It may make Bernie's fans feel butthurt, but there is a REASON Hillary is getting all of this establishment democrat support (read: superdelegates). Hillary Clinton is not just the wife of a VERY popular Democratic president, but she's a champion for the party in her own right. She works with fellow dems, she supports fellow dems, she raises money for the party, she supports the party in the off seasons. Hillary goes to bat in a way that Bernie Sanders just hasn't. It may seem unfair, but Hillary Clinton has the lion's share of supperdelegates because she's earned them. And if Bernie is going to make a compelling case for them to switch, he has to not only prove that his campaign has the momentum to take the general, but that he will do for the party what Hillary has done.

It'll be a tough sell, but if he can do the above, it wouldn't be impossible to get huge chunk of supers to switch.
 
It's not impossible.

What Bernie Sanders would need to do to possibly get super-delegates to switch is this:

1) Win from here on out. Bernie needs to gather a string of major victories against Hillary from here on out. And not tiny wins like tonight. He needs major blowouts that net him enough delegates to close in the on Hillary's lead.

2) He needs to grow his coalition. Winning alone won't be enough at this point, especially if the wins only happen because we're moving into whiter parts of the country. Like it or not, the minority vote is important for Democrats, especially in the general. A political revolution just will not happen on the backs of young white liberals. He needs to show he can mobilize minority voters in the same way Obama and Hillary have.

3) And this might be the MOST important thing: Bernie needs to show that he is a True Blue Democrat, and not just using our brand to legitimize his run for office. It may make Bernie's fans feel butthurt, but there is a REASON Hillary is getting all of this establishment democrat support (read: superdelegates). Hillary Clinton is not just the wife of a VERY popular Democratic president, but she's a champion for the party in her own right. She works with fellow dems, she supports fellow dems, she raises money for the party, she supports the party in the off seasons. Hillary goes to bat in a way that Bernie Sanders just hasn't. It may seem unfair, but Hillary Clinton has the lion's share of supperdelegates because she's earned them. And if Bernie is going to make a compelling case for them to switch, he has to not only prove that his campaign has the momentum to take the general, but that he will do for the party what Hillary has done.

It'll be a tough sell, but if he can do the above, it wouldn't be impossible to get huge chunk of supers to switch.

Considering the utter incompetence of Sander's campaign staff, I doubt he can do any of that.
 

Tingle

Member
It's not impossible.

What Bernie Sanders would need to do to possibly get super-delegates to switch is this:

1) Win from here on out. Bernie needs to gather a string of major victories against Hillary from here on out. And not tiny wins like tonight. He needs major blowouts that net him enough delegates to close in the on Hillary's lead.

2) He needs to grow his coalition. Winning alone won't be enough at this point, especially if the wins only happen because we're moving into whiter parts of the country. Like it or not, the minority vote is important for Democrats, especially in the general. A political revolution just will not happen on the backs of young white liberals. He needs to show he can mobilize minority voters in the same way Obama and Hillary have.

3) And this might be the MOST important thing: Bernie needs to show that he is a True Blue Democrat, and not just using our brand to legitimize his run for office. It may make Bernie's fans feel butthurt, but there is a REASON Hillary is getting all of this establishment democrat support (read: superdelegates). Hillary Clinton is not just the wife of a VERY popular Democratic president, but she's a champion for the party in her own right. She works with fellow dems, she supports fellow dems, she raises money for the party, she supports the party in the off seasons. Hillary goes to bat in a way that Bernie Sanders just hasn't. It may seem unfair, but Hillary Clinton has the lion's share of supperdelegates because she's earned them. And if Bernie is going to make a compelling case for them to switch, he has to not only prove that his campaign has the momentum to take the general, but that he will do for the party what Hillary has done.

It'll be a tough sell, but if he can do the above, it wouldn't be impossible to get huge chunk of supers to switch.

History says he just has to do the first. They always switch over if they believe the voters are for it. Unless we find out Bernie is a felon or some really strange occurrence happens, he really just needs to win over voters as far as the super delegates are concerned.

Whether that is likely or not is obviously up for debate, though.
 

Joeytj

Banned
Also, this is, what I consider to be, a very good read, albeit a bit of a long one: http://www.salon.com/2016/03/08/hil...re_rushing_to_nominate_the_weakest_candidate/

A few points to consider, from the article:

"the race begins then"

Not really. There are still a lot of very good states for HIllary left, especially the big ones like NY, CA and Florida. Sanders has more chance in those than others, of course, but Hillary's win in Texas with Hispanics puts more chips on Hillary's side of the bets going forward in states like Arizona and California, or Illinois.

But, again, there are going to be few blow-out states for both candidates. The only other Southern states left are NC and the Florida panhandle, where Hillary can expect huge margins, but it gets murkier for her, although the same can be said for Sanders.

Sanders should continue to do well in norther, whiter states like Alaska, Montana, Washington (it has some African Americans, but not a lot) and Wisconsin, but nothing big that will flip the delegate count.
 
Also, this is, what I consider to be, a very good read, albeit a bit of a long one: http://www.salon.com/2016/03/08/hil...re_rushing_to_nominate_the_weakest_candidate/

A few points to consider, from the article:

57% isn't enough to make up for the annihilation in the South.

Also what a condescending article. Another implication that somehow southern black voters are ignorant and that southern democrats don't matter.

Had Sanders been even a bit competitive in those states that don't matter he could be in a much more stable path to a potential victiory instead of being severely behind the eight ball.
 
It's not impossible.

What Bernie Sanders would need to do to possibly get super-delegates to switch is this:

1) Win from here on out. Bernie needs to gather a string of major victories against Hillary from here on out. And not tiny wins like tonight. He needs major blowouts that net him enough delegates to close in the on Hillary's lead.

2) He needs to grow his coalition. Winning alone won't be enough at this point, especially if the wins only happen because we're moving into whiter parts of the country. Like it or not, the minority vote is important for Democrats, especially in the general. A political revolution just will not happen on the backs of young white liberals. He needs to show he can mobilize minority voters in the same way Obama and Hillary have.

3) And this might be the MOST important thing: Bernie needs to show that he is a True Blue Democrat, and not just using our brand to legitimize his run for office. It may make Bernie's fans feel butthurt, but there is a REASON Hillary is getting all of this establishment democrat support (read: superdelegates). Hillary Clinton is not just the wife of a VERY popular Democratic president, but she's a champion for the party in her own right. She works with fellow dems, she supports fellow dems, she raises money for the party, she supports the party in the off seasons. Hillary goes to bat in a way that Bernie Sanders just hasn't. It may seem unfair, but Hillary Clinton has the lion's share of supperdelegates because she's earned them. And if Bernie is going to make a compelling case for them to switch, he has to not only prove that his campaign has the momentum to take the general, but that he will do for the party what Hillary has done.

It'll be a tough sell, but if he can do the above, it wouldn't be impossible to get huge chunk of supers to switch.

'A tough sell' is an understatement. I'm not all that confident that the superdelegates will switch, but we'll see.
 

Suikoguy

I whinny my fervor lowly, for his length is not as great as those of the Hylian war stallions
'A tough sell' is an understatement. I'm not all that confident that the superdelegates will switch, but we'll see.

Seriously, some of the overt attacks on the entire Democratic party last debate did him no favors.
 

harSon

Banned
I just don't see how the math adds up after March 15th. Florida, Ohio, Illinois, North Carolina, Missouri. That's like 600 delegates at stake, and which of those is he ultimately going to run up big numbers on - let alone win?
 

royalan

Member
History says he just has to do the first. They always switch over if they believe the voters are for it. Unless we find out Bernie is a felon or some really strange occurrence happens, he really just needs to win over voters as far as the super delegates are concerned.

Whether that is likely or not is obviously up for debate, though.

Oh, I definitely agree. I just added 2 and 3 because I don't think Bernie will actually be able to pass Hillary in pledged delegates (she's too far ahead and campaigning too well for that to happen), but he can get close. And if he gets close, and survives to the convention, I can see him having a case if he can accomplish 2 and 3.

But, if I'm being honest, I don't see it happening. Bernie's proved to be too stubborn.

Seriously, some of the overt attacks on the entire Democratic party last debate did him no favors.

Then there's this. After throwing democrats under the bus like he did last debate, there will be very little love for him at the convention. lol
 

Miletius

Member
Lets get ahead of ourselves here. He even says so himself, in his first post. Maybe after primary season progresses and more results come in (in his favor) then people ought to start asking about his magic sauce.
 

Tingle

Member
Oh, I definitely agree. I just added 2 and 3 because I don't think Bernie will actually be able to pass Hillary in pledged delegates (she's too far ahead and campaigning too well for that to happen), but he can get close. And if he gets close, and survives to the convention, I can see him having a case if he can accomplish 2 and 3.

But, if I'm being honest, I don't see it happening. Bernie's proved to be too stubborn.

Ah, okay. I don't see the delegates switching- in favor or against Bernie.

In my eyes 2 and 3 won't matter to him- the delegates probably do favor Hillary anyway, I don't see a situation they would break historical patterns and convert to Bernie unless its just by the vote from the public.
 

Tesseract

Banned
Bern or burn

The superdelegates will switch once Bernie bests Hillary in the upcoming states. Bernie will then defeat Trump, get elected, and rip shit up with real talk against neoliberlism and neoconservatives.
 

Brakke

Banned
The language around this is so weird. Hard to call it a "victory" for Bernie when the soberest headline is something like "Bernie falls less far behind than we expected". I guess the hope is that the polls vs reality gap gives us reason to think that other Great Lakes states could be stronger for Bernie than polls indicate? That's a pretty slim hope to hang a hat on.
 
The language around this is so weird. Hard to call it a "victory" for Bernie when the soberest headline is something like "Bernie falls less far behind than we expected". I guess the hope is that the polls vs reality gap gives us reason to think that other Great Lakes states could be stronger for Bernie than polls indicate? That's a pretty slim hope to hang a hat on.

This is a major historical event for primary elections just because of the sheer disparity between the statistics and the end result, and the fact that it's unprecedented. Even if Bernie lost, it would still be significant.
 
Gotta give this one to Bernie. Except for one person no one saw it coming. I mean many may have claimed it but didn't have statistical analysis backing it up.

I can see people trying to poach this guy already.
 

benjipwns

Banned
As I understand it, Tyler uses a combination of social media and Google trends data points, and refines his predictions by honing in on what variables are key to showing how likely a state is going to vote for one candidate over the other.
I began to utilize data from Google Trends to develop new estimates for Super Tuesday. Looking in retrospect, the correlation between the Google data and results in the first four states is striking once one has massaged the data set to produce usable measures (I use relative search frequency averaged over the couple days prior to the election).
This model is experimental in the sense that I am using data that has been aggregated from social media to make my estimates. To my knowledge this has never been done before, but after this election season we will be able to determine the viability of social media political sentiment as a proxy for broader public political sentiment.

There are many elements and controls to any good statistical model, and at this point there is not enough variance in some of the variables that I would like to include; nor is there enough observations to truly call anything statistically significant.
Well, I'm certainly convinced.

This is a major historical event for primary elections just because of the sheer disparity between the statistics and the end result, and the fact that it's unprecedented.
is it?
 
There's probably a lot of reasons pollsters were wrong but this is not one of them. Yes they largely do landlines, but they account for them, and this does not affect the numbers much.

Can you or anyone explain to me how exactly they account for the landlines skewing older? I've actually been wondering this for a while. Are a great number of polls still using landlines or are some able to get cell phone numbers and call them? It seems to me the landline thing isn't even going to be an option much longer
 
It's great that Bernie won a state, for his campaign anyway (before anybody assumes no I do not support Clinton I think every candidate this cycle is crap), but he needs to show that he can do more than win 1-2 to Hillary's 3-4 every Tuesday if he wants to stay in this thing.

He needs to chain wins, not just scrape and claw his way to a couple where they basically split the delegates, anyway.
 

evanmisha

Member
If Trump doesn't manage to secure enough delegates and the party chooses a different candidate, and he runs as an independent like he's talked about doing, and Clinton secures enough delegates to become the Democrat nominee, might Sanders also run as an independent?

Granted, I don't think he would consider doing that and splitting the Democrat vote if Trump doesn't split the Republican vote himself first
 
If Trump doesn't manage to secure enough delegates and the party chooses a different candidate, and he runs as an independent like he's talked about doing, and Clinton secures enough delegates to become the Democrat nominee, might Sanders also run as an independent?

Granted, I don't think he would consider doing that and splitting the Democrat vote if Trump doesn't split the Republican vote himself first

Then they'd split the pop vote, nobody would hit 270, and the republican congress would pick whoever the GOP fielded which at this point if a brokered convention happens (it's looking less likely by the day), would be Ted fucking Cruz.
 

evanmisha

Member
Then they'd split the pop vote, nobody would hit 270, and the republican congress would pick whoever the GOP fielded which at this point if a brokered convention happens (it's looking less likely by the day), would be Ted fucking Cruz.

Derp misread

Oh god please don't let this happen
 
Well, this is what 538 put out about Sanders' best chances to win.



3/22 - 4/19 could bring him a really big streak. And the thing about it is, I don't really know if she can turn around the Idahos, Utahs, Dakotas, of the world. I know they are really small, but when he nets 10 delegates from them, they starts to add up. Then you take into consideration Washington state, Oregon, Wisconsin, all three have modest amount of delegates to them and are heavily white. Though, I don't think they are caucus states. It will be interesting to see how far the anti-hillary vote goes in these smaller states.

One thing diffrent between this and the 2008 primary is, Obama one the western smaller states.

Another thing that is different is that Bernie is much further back in delegates than Hillary was in 2008
 
Status
Not open for further replies.
Top Bottom