Bullish on Bernie: The Bernie Sanders 2016 Thread II

[Tsukasa Buddha]

I assure you, I'd be ignoring polls no matter what the score was. It has nothing to do with my viewpoint other than that I think that polling is a completely useless activity. That applies to psych and sociology polling methods as much as it does to politics. It has been a long held viewpoint -- it is hardly dependent upon my support for Sanders.

I also think that demographic polling by news organizations (and others) to increase their popularity and pander to a population segment is the number one thing wrong with our press nowadays.

Statistics in general I have no problem with, but opinion polling tends to be useless due to the limitation of options and the fact that people don't really have a tendency to know what they want until they see it. Furthermore, there are always some major problems in methodology when you are trying to evaluate people in general. If it were otherwise, personality surveys/categories would actually make sense (which they don't). It usually boils down to a matter of measuring the wrong thing.

In the area of political polling, the problems that crop up are:

1) They throw out results based on flawed assumptions.
2) Every single moment in time is a point at which someone can change their mind.
3) The polls themselves can be used to affect the political process via self-fulfilling prophecy.

So why is it that polling has been so accurate?

Are Democratic primary polls accurate? In individual states they are good, with a few exceptions. At an aggregated level, they are remarkably accurate. The delegate-weighted polling margin has a total error of 3.1%. This is better accuracy than one would expect from the reputation of polling these days. The overall error is also not nearly enough to change my calculation from yesterday.

Above are Election Day results so far this season, plotted against poll medians taken shortly before the election. The correlation coefficient is +0.85, pretty good. Red symbols indicate caucus states.

Linky.

 

[TheL8Elvis]

So why is it that polling has been so accurate?



Linky.

Nice link. I enjoyed the prior post as well:
http://election.princeton.edu/2016/03/26/no-path-forward-for-sanders/

 

[angrysoba]

There has been a bit of a backlash against the Hillary-supporting superdelegates; and one of them in my state is facing an already-tough battle against a progressive candidate, who is now seeing an upwelling of support after the incumbent declared that he was going to vote for Clinton regardless of the popular vote's overwhelming support for Sanders. If Sanders supporters can put pressure like this on the superdelegates, I think we might see more of them switch than expected at this point.

If Sanders does win the popular vote, and the superdelegates push the nomination over to Clinton, that is only going to hurt her campaign in the general election, since it will merely reinforce the feelings of disenfranchisement among the younger voters who are the backbone of Sanders' campaign, that persistent meme that "voting doesn't accomplish anything". Too many have already insisted that they will not vote for Clinton under those circumstances, and are more likely to abstain from voting, than they are to make the effort to vote for her, or vote for a protest candidate. That would also have the larger effect of the Democratic party alienating younger voters, and either driving them to third parties, or out of the process entirely (which, given the current attitudes of the old-guard party establishment, is probably something they would consider desirable, given the anti-establishment, anti-corporate, and hard-left leanings of younger progressives).

The Fivethirtyeight people were talking about this on a recent podcast following Sanders's win. They acknowledge that Sanders is doing far better than expected, BUT, that this is still a far cry from saying that he really has much of a chance of winning the nomination. They have already said that Sanders was likely to win a string of caucuses following the March 15th wins for Clinton but that people shouldn't be fooled into thinking that it truly gives Sanders a meaningful momentum. They point out that Sanders also won Michigan in a big surprise but that Clinton wiped out that victory with the March 15th wins. Sanders's popularity with certain voters also galvanizes Clinton voters and in the end there are more of them, so it is extremely unlikely that Sanders will overhaul Clinton's lead which makes the superdelegates irrelevant anyway.

And I would prefer Sanders to win the nomination.

 

[Manopolus]

So why is it that polling has been so accurate?

Already answered in the quoted post.

The problem is, as already stated, that you can't even measure the accuracy of a poll that has been widely publicly released prior to the actual vote without defying the scientific method. It's not properly blinded -- the poll itself can influence the result of the election.

Thus, any result you come up with is scientifically invalid with regards to accurately predicting anything.

Anyone want to answer that one? I'm pretty sure I'm right. I've also given numerous examples of when they weren't right, explained why specific methodology was wrong, and have done just about everything possible to explain myself. If you don't understand my position at this point, you just plain aren't listening.

Heck, I've even pointed at a specific polling group that has always erred on the side of Clinton in every poll they've done and explained why that happened. It just happens to be the group that did New York's and Wisconsin's most recent polls.

 

[Tsukasa Buddha]

Already answered in the quoted post.

The problem is, as already stated, that you can't even measure the accuracy of a poll that has been widely publicly released prior to the actual vote without defying the scientific method. It's not properly blinded -- the poll itself can influence the result of the election.

Thus, any result you come up with is scientifically invalid with regards to accurately predicting anything.

Anyone want to answer that one? I'm pretty sure I'm right. I've also given numerous examples of when they weren't right, explained why specific methodology was wrong, and have done just about everything possible to explain myself. If you don't understand my position at this point, you just plain aren't listening.

I understand your position, I just think it is silly.

 

[Manopolus]

I understand your position, I just think it is silly.

Yeah... you don't get to do that out loud without at least attempting to refute my arguments.


...unless you want to look silly yourself, that is.

 

[angrysoba]

I assure you, I'd be ignoring polls no matter what the score was. It has nothing to do with my viewpoint other than that I think that polling is a completely useless activity. That applies to psych and sociology polling methods as much as it does to politics. It has been a long held viewpoint -- it is hardly dependent upon my support for Sanders.

I also think that demographic polling by news organizations (and others) to increase their popularity and pander to a population segment is the number one thing wrong with our press nowadays.

Statistics in general I have no problem with, but opinion polling tends to be useless due to the limitation of options and the fact that people don't really have a tendency to know what they want until they see it. Furthermore, there are always some major problems in methodology when you are trying to evaluate people in general. If it were otherwise, personality surveys/categories would actually make sense (which they don't). It usually boils down to a matter of measuring the wrong thing.

In the area of political polling, the problems that crop up are:

1) They throw out results based on flawed assumptions.
2) Every single moment in time is a point at which someone can change their mind.
3) The polls themselves can be used to affect the political process via self-fulfilling prophecy.

1) no polls are perfect, but that is why they have margins of error. However, polls are generally roughly accurate.
2) yes, but polls tend to have fairly large numbers in their samples to account to diminish the effect of this possibility.
3) many forecasts involving people do have an effect, of course, but that takes away from the strong claim that polls are useless. They clearly would have some use to some people. Biased pollsters clearly do try to influence voters with their results by skewing the data, but the noticeable dog-that-didn't-bark point here is that pollsters can't just predict landslides for their own candidates otherwise they would quickly lose all credibility. Almost all pollsters therefore have some incentive to provide largely accurate results which they generally succeed at.

 

[Trakar]

At what point will people admit that Bernie basically no chance?...

When the votes are counted at the 2016 Democratic convention and do not favor Sanders, then Sanders will not be capable of being the 2016 Democratic nominee.

At what point would you accept that Hillary is a bad choice for the Democratic party to nominate for the presidency?

 

[Trakar]

I heard the opposite explanation for why Bernie does well in caucuses. That is, that Bernie's supporters are motivated enough to spend considerable time and effort to vote at a caucus, while Clinton's wins in primaries are mostly due to people just having to show up to vote.

Most primary voting doesn't even need you to show up, much of it was mailed in months ago.

 

[angrysoba]

Yeah... you don't get to do that out loud without at least attempting to refute my arguments.


...unless you want to look silly yourself, that is.

If you agree with the idea of red states, blue states and swing states then effectively you agree with the premises that polling is based on. These states are only designated that way because it is forecast how they will vote ahead of time. If the candidates took your view then they would be campaigning in states with the hope of having a fifty-fifty chance of winning any of them. But in the general election no such thing will happen.

 

[Skeptic Ginger]

...
At what point would you accept that Hillary is a bad choice for the Democratic party to nominate for the presidency?

I can't speak for anyone else but I'm sure I'm not the only one who would ask, why would I do that when I think she's the most qualified candidate in the bunch on both sides of the aisle?

 

[Trakar]

Berenson's comments are Machiavellian. Instead of letting Bernie's team soap-box the topic in the belief that the electorate will rally behind it, he's tossed it back to them. How they respond will be interesting, but however they do so, they now look like supplicants.

The only one with a "tone" issue that needs adjusting is Hillary.

Hillary Laughs as Woman Says Bernie Doesn’t Care About Dead Kids
http://www.independentsentinel.com/hillary-laughs-as-woman-says-bernie-doesnt-care-about-dead-kids/

 

[Trakar]

It's not garbage that Bernie has had a problem with getting non-whites to vote for him. That doesn't mean no non-whites do.

Actually, Bernie has carried a majority of non-white voting blocks, he just hasn't done too well with southern blacks.

 

[Tony Stark]

When the votes are counted at the 2016 Democratic convention and do not favor Sanders, then Sanders will not be capable of being the 2016 Democratic nominee.

At what point would you accept that Hillary is a bad choice for the Democratic party to nominate for the presidency?

So you plan on being in denial until the bitter end?

And the answer to your question is if she loses to Trump (with the assistance of "progressives" like you) or signs right wing bills/appoints right wing judges. Neither are likely to happen.

 

[Trakar]

Yes, this certainly aligns with history. Candidates with zealous supporters tend to do well in caucuses, e.g. Ron Paul and Pat Robertson.

Barry (Obama).

 

[The Big Dog]

So you plan on being in denial until the bitter end?

And the answer to your question is if she loses to Trump (with the assistance of "progressives" like you) or signs right wing bills/appoints right wing judges. Neither are likely to happen.

lolz, Rule of So.

Better watch your "tone."

 

[Trakar]

Swing states this year: Ohio, Colorado, Iowa, Nevada, New Hampshire, Virginia, and Florida. Sanders won New Hampshire and Colorado. Five out of seven were won by Clinton.

Swing state is meaningless when discussing Bernie and Hillary because Bernie and Hillary won't be competing for them when, they swing or they don't. Just Like Hillary won't be winning the deep south states in November even if she does win the Democratic nod this summer.

So, which are the states that Hillary has won so far, that you think she will win in November that Sanders wouldn't win in November?
IOW - Red states that Hillary has won in the Primaries that she can turn Blue in November and Sanders couldn't

And I guess turnabout is fair play, so which states has Sanders won that I think Sanders will win in November that Hillary wouldn't win in November:

Alaska and possibly Oklahoma (There are several others that I feel good about Sanders' chances in, that I don't think Clinton could win now or in November,...but those primaries haven't happened yet)

 

[wareyin]

Swing state is meaningless when discussing Bernie and Hillary because Bernie and Hillary won't be competing for them when, they swing or they don't. Just Like Hillary won't be winning the deep south states in November even if she does win the Democratic nod this summer.

So, which are the states that Hillary has won so far, that you think she will win in November that Sanders wouldn't win in November?
IOW - Red states that Hillary has won in the Primaries that she can turn Blue in November and Sanders couldn't

And I guess turnabout is fair play, so which states has Sanders won that I think Sanders will win in November that Hillary wouldn't win in November:

Alaska and possibly Oklahoma (There are several others that I feel good about Sanders' chances in, that I don't think Clinton could win now or in November,...but those primaries haven't happened yet)

If swing state is meaningless, then so is any consideration of red state, blue state, or any state.
We can go by total votes, where Clinton has more than Sanders, but also more than Trump, or Cruz, or Kasich. But this still doesn't make Sanders the winner you want him to be, so maybe polling? Nope, Clinton is the clear winner there, as well.
Waving away Clinton's wins in swing States ignores the plain fact that those are States the Democratic nominee will need to win, and Sanders couldn't win the majority, and especially the large delegate ones needed.

 

[Trakar]

I can't speak for anyone else but I'm sure I'm not the only one who would ask, why would I do that when I think she's the most qualified candidate in the bunch on both sides of the aisle?

I'm sure the many if not most of the supporters of all candidates feel that their candidate is the most qualified candidate in the bunch on both sides of the aisle. But saying that that nothing can change your mind, speaks for itself.

 

[Trakar]

So you plan on being in denial until the bitter end?

And the answer to your question is if she loses to Trump (with the assistance of "progressives" like you) or signs right wing bills/appoints right wing judges. Neither are likely to happen.

I'm not denying anything, until the votes are cast and counted there is no reason to presume the outcome. Right or wrong, guesses are guesses.