Elections Clinton vs Trump Polls thread (Clinton's Bounce Larger than Trump's)

[JudoThrowFiasco]

There was a gap between "announced his candidacy" and actually appeared to be seriously in the race, as there is for most candidates who end up among the top few. Again (and again, and again--sigh), the point is not that the race can never change or that someone leading at X point in the race is 100% certain to win.

OH, so when you said that clinton was up by double digits in polling in states she lost in, it was due to Bernie not announcing his candidacy - you really meant - not actively being an candidate

Jack Backtrack seems to be a common thing here.

and you haven't acknowledged that several states had clinton leading polls months before and still ended up with her losing - so, why should we take these polls as indicative of her future win?

Let's try another angle: People who study this say that the r squared between polling this far out and the final result is about 0.7. Your assertion is that it is what? 0? And you base your claim on what? The fact that you recall some high-profile cases of someone coming from behind?

where did you get .7 from?

Not really. If the election were held today, polling would show Clinton with a very large lead (and her odds would be over 99%).

A +5 for the month of June is a very large lead? 99% chance of winning? Mind showing us your algorithm

Where is this evidence? Don't give me badly recalled anecdotes. Give me some data.

i did not give a single shred of anecdotal evdience - i produce that wired article - feel free to read it

heres one RE: the British federal election in 2015:

http://www.theguardian.com/politics/2015/nov/13/new-research-general-election-polls-inaccurate

"Opinion pollsters are very good at making their samples reflect the general population. But the general population and the electorate are very different things, because around 40% of adults don’t vote. There are far too many young voters in polling samples, and not enough young non-voters. Because young people tend to be more Labour leaning, this means we end up with too many Labour voters in the polls”.

Ugh. When you're as slow as you are, you need to be a lot more humble.

I said that if her odds were 3-1 in October, that would mean that the race would have gotten tighter. That's because the current odds take into account how tight the race is and also uncertainty, given how far out the election is. If Trump doesn't close a lot of ground by October, his odds will change greatly for the worse. Do you get that? As the time remaining shrinks, the polling leader's odds get better if the polls look the same. If the odds don't change, that means that the race is getting closer.

Backtrack Jack, strikes again. UGhh, i used to think Rama and Ultra and 99% of the members of this thread were wrong about you...but alas, they kind of have a point.

First you said 3-1 would show a tightening in the closeness of the race - in which i said it it was already 3-1 clinton. So, then, because you thought the current odds would be more in Clinton's favor - you started some rhetoric of the odds being skewed at this point due to uncertainties. So i ask, again, if we should take the odds right now as a byproduct of uncertainties, why shouldnt we take polls with uncertainties as well?

I'm sorry, but professionals aren't as stupid as you think. Hence my point about monetizing your insight.

please provide a poll that has a breakdown of its demographics? If their methodology is true, why is there no breakdown to demonstrate valid quantitatives

“polls are snapshots, meant to capture voter sentiment at a particular moment.” People are not very good at predicting what they will do in the future" - Nora Kelly

 

[Jack V Savage]

and you haven't acknowledged that several states had clinton leading polls months before and still ended up with her losing - so, why should we take these polls as indicative of her future win?

Because "we" (not you) know what "indicative" means.

I would say her CoD is ??? i would guess a .3 - where did you get .7 from?

.7 is based on historical data. What are you basing .3 on? Just bias and wishful thinking, right?

A +5 for the month of June is a very large lead? 99% chance of winning? Mind showing us your algorithm

Latest RCP average is 6.7. That is a very large lead. That's just based on watching elections. A big polling lead shortly before the election means a virtual lock. Look through the primaries if you doubt it.

i did not give a single shred of anecdotal evdience - i produce that wired article - feel free to read it

Maybe you don't know what "anecdotal evidence" means? You've been saying that since you are aware of a few high-profile surprises, polling is not predictive at all. In fact, if you want to make that case, you need to look more systematically. You have backtracked (funny that you accuse me of that when I didn't do it and you are) to now saying that you acknowledge a correlation--just a weaker one than the data show.

"Opinion pollsters are very good at making their samples reflect the general population. But the general population and the electorate are very different things, because around 40% of adults don’t vote. There are far too many young voters in polling samples, and not enough young non-voters. Because young people tend to be more Labour leaning, this means we end up with too many Labour voters in the polls”.

I guess hints aren't good enough here. Pollsters are well aware of this stuff. It's, like, a job and something that people get educated for. A whole field of study.

Backtrack Jack, strikes again. UGhh, i used to think Rama and Ultra and 99% of the members of this thread were wrong about you...but alas, they kind of have a point.

Don't get mad because you can't follow a simple point. Just try harder to understand.

It's funny because Anung was even worse at understanding probability. I made the point that if you think the betting odds are wrong, you should take advantage. He responded by saying that if you think they're right, you should bet, which made no sense at all and nothing could convince him that he was missing that simple point.

First you said 3-1 would show a tightening in the closeness of the race - in which i said it it was already 3-1 clinton.

Um, I'm still saying that. If the race doesn't get tighter, the odds will diverge more. This is important so think hard about why that is. There is uncertainty built into the current betting odds. As time passed, uncertainty fades, and with no change in the state of the race, the favorite's odds get better (the correlation between polling and the final result strengthens). So 3-1 in late October means that the polls are very close. 3-1 today doesn't. Get it?

 

[TheComebackKid]

So your "analysis" is just "Republicans say that Clinton is evil; therefore, she can't win." In other words, it's just wishful thinking. The thing is, even if people who don't vote for Clinton don't like her, that doesn't mean that she's going to lose. In our system, the candidate who gets more votes wins. I thought that the GOP taking over the Senate in 2014 was terrible for the country, but I was still able to acknowledge the fact that polling that year showed that that was a likely outcome. Don't know why you can't be similarly honest and willing to face reality head on.

Hillary can definitely win the election and the polling is in her favor.

Is that what you were looking for? I don't think I've ever said anything that would contradict that.

 

[Jack V Savage]

Hillary can definitely win the election and the polling is in her favor.

Is that what you were looking for?

At this point, she's the likely winner.

 

[VulcanNervPinch]

Fair, but brexit was remain up until the vote started and projected up to a few hours after.

Brexit is an interesting and rare case.
I don't know anything about the nature of Britain's major polling institutions.
However, I think it's is probably an example of the principle that, when you have a close race and strong emotional/illogical/nationalistic hot-button issues involved, that side tends to be undercounted by a little.

Another similar example (presuming there wasn't any outright fraud involved) would be W's win in 2004.
9/11 had really inflamed nationalistic fever and the subsequent invasion of Iraq spawned even more militaristic chest-thumping. Even though it was becoming apparent that the pretexts for the Iraq invasion were utter lies and no Mission was "Accomplished", lots of people weren't willing to accept it yet.
Polls in that election undercounted W's support by a bit, most likely because some people voted for him, but were too ashamed (with good reason) to admit it to pollsters.

Still, that's only 1 instance in the last how many decades?

So - either polling months in advance is at best a rough gauge not to carry any weight or merit and, polling can flat out be wrong - i.e Clinton in Ohio / Sanders in Michigan

State primary polls are definitely done less frequently and intensively than the general election. And there's also the much more uncertain nature of turnout, local races, primary vs caucus, which ballot people will choose to participate in, and usually > 2 active candidates. So certainly they are subject to being wrong a little bit more often than a general election Presidential poll.

So, putting any weight in the federal election at this point seems nebulous

Things certainly can change and the leader at any given time isn't guaranteed to win.
Nonetheless, the leader in June is more likely to win than the trailer.
The leader in July is more likely to win than the trailer.
The leader in August is more likely to win than the trailer.
The leader in September is more likely to win than the trailer.
The leader in October is more likely to win than the trailer.
So they certainly do matter. Dismissing polls in advance, just because they're not always reflective of who wins in the end, is typically just something a fan of guy who's trailing does to keep their morale up.

 

[JudoThrowFiasco]

Because "we" (not you) know what "indicative" means.

Considering you have totally ignored the several states that followed Clinton's early polling lead - still resulting in her loss, i would say you may be a little fuzzy of "indicative" - - Clinton was leading Colorado in November - 4 months later, she lost. You dont see how this scenario can be applied to this case?

.7 is based on historical data. What are you basing .3 on? Just bias and wishful thinking, right?

i would give her shaky line at best -- based on how many bernie voters support her (only about 50% in race that has independent nominee). Currently, Trump has about 8% of bernie supporters and no reason to think he will not sure up a higher percentage of the right aligned voters than Hillary will with the left aligned.

Please provide a single source for the historical data you are referring to.

Latest RCP average is 6.7. That is a very large lead. That's just based on watching elections. A big polling lead shortly before the election means a virtual lock. Look through the primaries if you doubt it.

Like Clinton had in Michigan? Indiana?

Maybe you don't know what "anecdotal evidence" means? You've been saying that since you are aware of a few high-profile surprises, polling is not predictive at all. In fact, if you want to make that case, you need to look more systematically. You have backtracked (funny that you accuse me of that when I didn't do it and you are) to now saying that you acknowledge a correlation--just a weaker one than the data show.

I have said early polling provides little insight besides the mood of the day - and i have posted several articles relaying that fact - no anecdotal at all - also, pointing out times in which it's been wrong = fact

Not my problem if you choose to ignore them - great thing about facts is that they remain the same if jack reads them or not.

I guess hints aren't good enough here. Pollsters are well aware of this stuff. It's, like, a job and something that people get educated for. A whole field of study.

Again, provide a single poll that gives its demographics - especially in a GE poll. The "just take their word approach does not work

read that wired article.

Don't get mad because you can't follow a simple point. Just try harder to understand.

It's funny because Anung was even worse at understanding probability. I made the point that if you think the betting odds are wrong, you should take advantage. He responded by saying that if you think they're right, you should bet, which made no sense at all and nothing could convince him that he was missing that simple point.

just admit you had no idea the current lines were so close and that you had to do some gymnastics to revision your point to fit your narrative.

Um, I'm still saying that. If the race doesn't get tighter, the odds will diverge more. This is important so think hard about why that is. There is uncertainty built into the current betting odds. As time passed, uncertainty fades, and with no change in the state of the race, the favorite's odds get better (the correlation between polling and the final result strengthens). So 3-1 in late October means that the polls are very close. 3-1 today doesn't. Get it?

so polls and odds are pretty useless in June for a november election...i think so too.

 

[JudoThrowFiasco]

Brexit is an interesting and rare case.
I don't know anything about the nature of Britain's major polling institutions.
However, I think it's is probably an example of the principle that, when you have a close race and strong emotional/illogical/nationalistic hot-button issues involved, that side tends to be undercounted by a little.

Another similar example (presuming there wasn't any outright fraud involved) would be W's win in 2004.
9/11 had really inflamed nationalistic fever and the subsequent invasion of Iraq spawned even more militaristic chest-thumping. Even though it was becoming apparent that the pretexts for the Iraq invasion were utter lies and no Mission was "Accomplished", lots of people weren't willing to accept it yet.
Polls in that election undercounted W's support by a bit, most likely because some people voted for him, but were too ashamed (with good reason) to admit it to pollsters.

Still, that's only 1 instance in the last how many decades?

I agree with your premise that each pass month will give a slight better insight - but plenty of evidence not to put too much merit in polls this far from the election date.




State primary polls are definitely done less frequently and intensively than the general election. And there's also the much more uncertain nature of turnout, local races, primary vs caucus, which ballot people will choose to participate in, and usually > 2 active candidates. So certainly they are subject to being wrong a little bit more often than a general election Presidential poll.



Things certainly can change and the leader at any given time isn't guaranteed to win.
Nonetheless, the leader in June is more likely to win than the trailer.
The leader in July is more likely to win than the trailer.
The leader in August is more likely to win than the trailer.
The leader in September is more likely to win than the trailer.
The leader in October is more likely to win than the trailer.
So they certainly do matter. Dismissing polls in advance, just because they're not always reflective of who wins in the end, is typically just something a fan of guy who's trailing does to keep their morale up.

check out how the polls also got the 2015 British fed election wrong

http://www.theguardian.com/politics/2015/nov/13/new-research-general-election-polls-inaccurate

and dont forget 2012: Gallups missed polled RE: A Romney win

http://theweek.com/articles/617109/problem-polls

 

[TheComebackKid]

At this point, she's the likely winner.

I think this is Trump's election to lose given the different variables and factors at play. But there is a lot of ways for Trump to lose the election especially given his particular weaknesses.

Why did that cause you to have a meltdown?

 

[Jack V Savage]

Considering you have totally ignored the several states that followed Clinton's early polling lead - still resulting in her loss, i would say you may be a little fuzzy of "indicative" - - Clinton was leading Colorado in November - 4 months later, she lost. You dont see how this scenario can be applied to this case?

No, you really don't seem to know what it means. If X is "indicative" of Y, it doesn't mean that X always, inevitably leads to Y. Note that despite a big polling lead, Clinton's betting odds indicate that she has only a 75% chance. That means that there's a one in four chance that Trump can win *even if the odds are exactly correct*. Do you understand that last part?

i would give her shaky line at best -- based on how many bernie voters support her (only about 50% in race that has independent nominee). Currently, Trump has about 8% of bernie supporters and no reason to think he will not sure up a higher percentage of the right aligned voters than Hillary will with the left aligned.

Please provide a single source for the historical data you are referring to.

Did that earlier in this thread. As for your earlier analysis, what is the normal percentage of crossover voters? Do you know? If not, it would seem to be pretty irrelevant. Further, that's a reason to think that Clinton's lead will widen, as it is common for supporters of the loser in a primary to take some time to warm to the winner, but it would be very uncommon for that not to happen.

Like Clinton had in Michigan? Indiana?

Er, I don't know how to explain this to you so you'll understand. It's not good analysis to pick out extreme outliers and then assume that they invalidate all evidence. If you are claiming that polls are unreliable, you need to look at all the polls. What percentage of the time did the polling leader win in all the primaries?

http://www.dailykos.com/story/2016/4/21/1516967/-How-accurate-are-those-538-primary-forecasts-anyway

The polls-only forecast for 23 states predicted a delegate-weighted average lead of17.6%19.18% for Clinton. The actual delegate-weighted lead was 18.63%, a difference of only -1.020.55 points. There’s some variance in individual forecasts, as can be seen in the scatter-plot above, but 90% of the polling errors fall between -14.3 points (favoring Clinton) and 10.1 points (favoring Sanders). The correlation coefficient is roughly 0.96.

Granted, that's using later polls (closer to the election results), but the reason Michigan was such a surprise was that it went against the predicted results based on late polls (where even a small margin means a lot).

I have said early polling provides little insight besides the mood of the day - and i have posted several articles relaying that fact - no anecdotal at all - also, pointing out times in which it's been wrong = fact

I don't think you know what "anecdotal evidence" means. I'm not claiming that the anecdotes are false; just that you don't build a data-based case by citing them. And, again, if I say that you have a one-in-six chance of rolling a six, and you roll a six, that doesn't prove me wrong.

Again, provide a single poll that gives its demographics - especially in a GE poll.

What difference would it make? If you're asserting that it never occurred to pollsters to adjust that stuff, you're wrong.

just admit you had no idea the current lines were so close and that you had to do some gymnastics to revision your point to fit your narrative.

LOL! You're really dense, dude. If the betting odds stay the same as the election gets closer, that means that the race is getting closer because uncertainty based on time is falling.

so polls and odds are pretty useless in June for a november election...i think so too.

I have trouble believing that anyone can really be this dense. Please tell me you're just trolling.

I think this is Trump's election to lose given the different variables and factors at play. But there is a lot of ways for Trump to lose the election especially given his particular weaknesses.

Why did that cause you to have a meltdown?

??? You should see someone about your extreme sensitivity. Just because I'm disagreeing with you doesn't mean I'm mad at you, bro. Calm down. You're cool. Ehug.

 

[TheComebackKid]

??? You should see someone about your extreme sensitivity. Just because I'm disagreeing with you doesn't mean I'm mad at you, bro. Calm down. You're cool. Ehug.

No I know you're not mad at me and it's not about being sensitive. I'm just saying...you kind of had a meltdown back there. I simply said I think the tide is turning and you responded by calling me a dishonest partisan who doesn't want to acknowledge polling data.

I was just curious about the reaction to a rather benign comment. But it's cool bro. Ehug accepted!

 

[Jack V Savage]

No I know you're not mad at me and it's not about being sensitive. I'm just saying...you kind of had a meltdown back there.

Hmm. Is English not your first language?

I simply said I think the tide is turning and you responded by calling me a dishonest partisan who doesn't want to acknowledge polling data.

Oh, I see the problem. That's not what happened. Read through the thread again.

 

[TheComebackKid]

I simply said I think the tide is turning and you responded by calling me a dishonest partisan who doesn't want to acknowledge polling data.

Oh, I see the problem. That's not what happened. Read through the thread again.

Ok, reread...

The tide has turned. Trump is on the popular side on most of the issues and Clintons corrupt meter is off the charts. If he can not go off the rails expect those poll numbers to start to switch.

The tide has indeed turned, though not in the direction that you're saying it has. Look at the polling numbers. That's not to say it can't turn again or something (and candidates usually get a temporary boost after their party's convention), but at this point, your post is just wishful thinking from one of the most committed Republican partisans here. And the people who believe the dishonest smears about Clinton's character are already in that 43% or so that is planning to vote for Trump.

So your "analysis" is just "Republicans say that Clinton is evil; therefore, she can't win." In other words, it's just wishful thinking. The thing is, even if people who don't vote for Clinton don't like her, that doesn't mean that she's going to lose. In our system, the candidate who gets more votes wins. I thought that the GOP taking over the Senate in 2014 was terrible for the country, but I was still able to acknowledge the fact that polling that year showed that that was a likely outcome. Don't know why you can't be similarly honest and willing to face reality head on.

Spoiler

 

[Jack V Savage]

Ok, reread...

Doesn't seem like you did, if you're standing by your story. But maybe you're not so big on telling the truth.

 

[carlos cadona]

We've got a lot of new polls released today, all of them show Clinton leading, the GE polls have her up by a significant margin and the swing state polls are fairly close.
http://www.nbcnews.com/meet-the-pre...w&hootPostID=f1bc2ae86e0f1f07a0afb55a7cdbdf67
http://www.langerresearch.com/wp-content/uploads/1178a12016Election.pdf

 

[Social Distance Warrior]

No I know you're not mad at me and it's not about being sensitive. I'm just saying...you kind of had a meltdown back there. I simply said I think the tide is turning and you responded by calling me a dishonest partisan who doesn't want to acknowledge polling data.

I was just curious about the reaction to a rather benign comment. But it's cool bro. Ehug accepted!

Anything Hillary related will set this dishonest POS off..

You committed the Cardinal error of suggesting that Hillary might lose...

Jack loves Hillary Clinton....kind of obsessed..

But watch...he'll now deny that he supports her. Lol.

Such a little bitch.

 

[Jack V Savage]

Such a little bitch.

Funny you use that, when you're the guy who can't even engage me directly, and yet you talk about me constantly, and you lie as bad as TCK.

 

[spin]

Anything Hillary related will set this dishonest POS off..

You committed the Cardinal error of suggesting that Hillary might lose...

Jack loves Hillary Clinton....kind of obsessed..

But watch...he'll now deny that he supports her. Lol.

Such a little bitch.

I'll deny it as well.

Literally none of this happens.

From what I see, the very thought of jack sets you off, and you seem kind of obsessed.

You dont like him, fine, but you're doing the same shit

 

[wcisdoomed]

They didn't even predict it that wrong. The noteable one they were wrong on was Iowa where they called him winning but it was very close in the polls and there was shifting with Rubio in the last week so it was a decent toss up at that point. After that, they called, NH, NV, SC all correct in a row for Trump. They called Florida for Trump. I don't see where they did extremely bad with calling those primaries.

They might be talking about their commentary where they did seem to see Trump as a short term thing but that never was reflected in their predictions state by state.

To the best of my recollection, the statistician said he would have forecasted Trump to win the nomination as early as summer 2015 had he focused solely on the polls.

 

[wcisdoomed]

I feel the same way about you guys and Clinton. Or I guess the boogeyman would be a better comparison.

Hillary is a crooked politician, but no more so than most others. I'll concede that some people hate her way beyond her merits. Definitely a boogeyman for some.

 

[Madmick]

In new poll, support for Trump has plunged, giving Clinton a double-digit lead (June 25th, 2016)