Political Betting Thread

[Omni1]

I do think the democrats shouldn't nominate Biden lol.

Favorability ratings weren't a great predictor in 2016 due to Hillary's comparably low favorability. She was a deeply disliked candidate, one of the most hated people in the democratic party. What's becoming clear here is that Trump doesn't have an incumbency advantage and his performance as president is pretty poorly received by most of the population. All this despite relative normalcy in our economic situation and no major conflicts. That's a very bad sign. Trump has no comeback factor within sight to win people over here.

I can give you the four most recent people who showed comparable situations to Trump at this point in time: Gerald Ford, Jimmy Carter, George HW Bush, and Barack Obama. Obama's numbers improved for the reasons I mentioned. Anyone with confidence in Trump is fooling themselves.

Chuckle. Just days ago President Trump spoke in front of a packed Dallas stadium (20,000+). I find it absolutely hilarious that people like you still exist in 2019. Mainstream media = Fake news. I understand that you want to cling to your illusions & that no amount of facts will motivate you to accept reality. Speaking of reality. Here is footage of Presidents Trumps rally in Dallas, Texas.

 

[rev0lver]

Chuckle. Just days ago President Trump spoke in front of a packed Dallas stadium (20,000+). I find it absolutely hilarious that people like you still exist in 2019. Mainstream media = Fake news. I understand that you want to cling to your illusions & that no amount of facts will motivate you to accept reality. Speaking of reality. Here is footage of Presidents Trumps rally in Dallas, Texas.

This is a sarcasm post, right

 

[longway]

For a good chunk of them, it really is variance. I'll reiterate that Trump won some of these states by less than 1%, a few thousand votes. I think the factor of enthusiasm and apathy toward Hillary absolutely came into play on election day and it's why it pushed Trump to a win in these states, though generally within reasonable margins. I've mentioned before that I predicted a Trump win in 2016 (though I didn't have the bankroll to win that much at the time :/ ), and this is the exact data I was looking at to inform that decision.

I'm talking about all this because I see BOTH the opposite effect happening in 2020, with more enthusiastic support against Trump and possibly for a populist democrat, and a data trend toward democrats gaining a bigger lead in the swing states where an underestimated pro-Trump turnout wouldn't even be a concern.

You call it variance, then contradict yourself talking about enthusiasm.

 

[rev0lver]

You call it variance, then contradict yourself talking about enthusiasm.

Let me clarify a little bit. The polls were already accounting for these known and unknown factors in their error margins. The 538 article you cited showed Trump's variance against the polls in these swing states averaging about +3%, about the same as I said and within a standard margin of error in election polling. I expected Trump to reach more favorable ends of the margin on election day with his higher enthusiasm+lower Hillary enthusiasm, and edge out enough states to win. But that he'd generally still be within these margins because they're well-representing the real data. And that's what happened. Some of that you can attribute to enthusiasm, some of it normal variance, some of it a mix of both.

538's model in 2016, which showed a much better chance of Trump winning than other media outlets, was a pretty well-developed model and there's nothing really wrong with the approach. Hillary had the edge by standard data-driven logic, but Trump had a chance to edge out tight races in some key states if you account for some variance with these factors, and he did.

 

[Roads55]

biden is accepting money from super pacs now. desperation from lack of donors i guess. surely he will get called out on this. latest poll has warren leading as well. not sure which if not both is responsible for biden drifting back to 5.0 area.

 

[longway]

The polls were already accounting for these known and unknown factors in their error margins.

Error margins are set by a formula determined by sample size, and the error can go either way over different polls which when aggregated tend to cancel each other out.

The problem was the polled sample populations were not representative of the voting public, so multiple polls were skewed.

Factors like enthusiasm is not part of natural variance, it meant the likely voter models were wrong, so they did not poll representative samples.

Trump support was underestimated in multiple states in multiple polls outside the 95% margin of error.

You also seemingly cant understand that variance should go both ways equally.

Its safe to say that variance is a far less likely explanation of final results than skewed sampling in polls.

 

[longway]

biden is accepting money from super pacs now. desperation from lack of donors i guess. surely he will get called out on this. latest poll has warren leading as well. not sure which if not both is responsible for biden drifting back to 5.0 area.

Its incredible isnt it. Biden has no money to fight a general election, and warren and sanders wont have a broad enough appeal.

Will an outsider swoop in?

 

[waiguoren]

I'm not sold on America being ready for an openly gay president. Homophobia is still a thing, particularly among swing states that still have heavy religious influence

"America" does not choose the president. Florida, Ohio, Wisconsin, Pennsylvania, North Carolina, Virginia, Michigan, Iowa and New Hampshire choose the president. I think it's crazy to think that independents in those states are doing to fail to turn out or will vote Trump because Buttigieg identifies as a homosexual or because he is married to a man. I think if you look over some opinion polls of independent voters on the gay marriage issue, you'll see it my way.

Error margins are set by a formula determined by sample size, and the error can go either way over different polls which when aggregated tend to cancel each other out.

The problem was the polled sample populations were not representative of the voting public, so multiple polls were skewed.

Factors like enthusiasm is not part of natural variance, it meant the likely voter models were wrong, so they did not poll representative samples.

Trump support was underestimated in multiple states in multiple polls outside the 95% margin of error.

You also seemingly cant understand that variance should go both ways equally.

Its safe to say that variance is a far less likely explanation of final results than skewed sampling in polls.

Good post.

 

[Omni1]

You call it variance, then contradict yourself talking about enthusiasm.

Facts, reason & logic do not have a place in dev0lved's wild eyed fantasy world. In regards to him. This old saying comes to mind. "Oh, foolish man, what can you not be made to believe?"

 

[waiguoren]

Facts, reason & logic do not have a place in dev0lved's world. In regards to him. This old saying comes to mind. "Oh, foolish man, what can you not be made to believe?" LOL!

Come on, at least try to offer some analysis instead of dropping in to flame people. You're the guy who posted five months ago that Biden was going to drop out of the race imminently. A little humility is in order.


==================

I'm curious what effect, if any, Biden's decision to work with PACs will have on this race. It's hypocritical, but that doesn't necessarily matter.


My Current bets:


April 15 Bernard Sanders +410 $2439.02
May 19 Peter Buttigieg +550 $1818.18
August 31 Peter Buttigieg +2000 $500
September 20 Peter Buttigieg +2026 $493.58
September 25 Elizabeth Warren +123 $4268

 

[Roads55]

Come on, at least try to offer some analysis instead of dropping in to flame people. You're the guy who posted five months ago that Biden was going to drop out of the race imminently. A little humility is in order.


==================

I'm curious what effect, if any, Biden's decision to work with PACs will have on this race. It's hypocritical, but that doesn't necessarily matter.


My Current bets:


April 15 Bernard Sanders +410 $2439.02
May 19 Peter Buttigieg +550 $1818.18
August 31 Peter Buttigieg +2000 $500
September 20 Peter Buttigieg +2026 $493.58
September 25 Elizabeth Warren +123 $4268

let's see how much he actually gets in the short term, but it is a sizable event for an opponent to potentially attack him over.

 

[waiguoren]

let's see how much he actually gets in the short term, but it is a sizable event for an opponent to potentially attack him over.

I don't think most of the people telling pollsters they prefer Biden really care whether Biden works with PACs. Overall I can see this helping Biden. The November debate will still have at least nine people. I'm hoping the threshold for the December debates is higher so that we see no more than six candidates on stage.



My Current bets:


April 15 Bernard Sanders +410 $2439.02
May 19 Peter Buttigieg +550 $1818.18
August 31 Peter Buttigieg +2000 $500
September 20 Peter Buttigieg +2026 $493.58
September 25 Elizabeth Warren +123 $4268

 

[rev0lver]

Error margins are set by a formula determined by sample size, and the error can go either way over different polls which when aggregated tend to cancel each other out.

The problem was the polled sample populations were not representative of the voting public, so multiple polls were skewed.

Factors like enthusiasm is not part of natural variance, it meant the likely voter models were wrong, so they did not poll representative samples.

Trump support was underestimated in multiple states in multiple polls outside the 95% margin of error.

You also seemingly cant understand that variance should go both ways equally.

Its safe to say that variance is a far less likely explanation of final results than skewed sampling in polls.

These are the types of things error margins can account for, you just wrote a longer explanation for it. Aggregated polling does not necessarily cancel it out with different sampling methods, methodology, and framing of questions, and can have consistent sampling issues that stack error rather than reduce it. The margin only represents a hypothetical, in these cases it doesn't reduce by combining samples of other polls (especially when given at different times).

There was some variance on both sides. It skewed more toward Trump's side for various reasons. Voter enthusiasm cannot necessarily be accounted for in polling and doesn't mean the methodology was wrong. The polls likely were well-representative of the populations they were polling. I guess it's inappropriate to say something like enthusiasm is a part of "variance" in the scientific sense, because that's not what it refers to in political polling, I only mean that in the context of the results with reference to the predictive nature of polls vs the base intention of showing voter attitudes, and that we were within these margins of what these attitudes were in the final results. There's a good reason Nate Silver gave Trump a more favorable outlook in 2016 than all the other orgs, and applying a natural variance to the data along with these factors showed he had the right idea.

 

[rev0lver]

Facts, reason & logic do not have a place in dev0lved's wild eyed fantasy world. In regards to him. This old saying comes to mind. "Oh, foolish man, what can you not be made to believe?"

Wait so you're serious? Lol. Did you not just post a rally in Texas on this page implying 20k Trump supporters converging together means that all the negative data for Trump (aka most of the data out there) is fake? I guess your kind isn't really known for looking at information though.

 

[longway]

Voter enthusiasm cannot necessarily be accounted for in polling and doesn't mean the methodology was wrong.

lol

But I give up, you win.

 

[rev0lver]

"America" does not choose the president. Florida, Ohio, Wisconsin, Pennsylvania, North Carolina, Virginia, Michigan, Iowa and New Hampshire choose the president. I think it's crazy to think that independents in those states are doing to fail to turn out or will vote Trump because Buttigieg identifies as a homosexual or because he is married to a man. I think if you look over some opinion polls of independent voters on the gay marriage issue, you'll see it my way.

Well for the most part, I agree that independents and dems (and maybe most republicans at this point) don't care. Gay marriage clearly has majority support in the country right now and debate on that issue is pretty much done for. My issue is that there still are holdouts and people uncomfortable with the idea, and we don't even have to worry about that issue with any other candidate. The potential voting base for Pete has to get over the sexuality issue before they even consider his policy platform. I guess you could say the same for minority+female candidates, but there's no real religious objection to gender or skin color, and you can't underestimate the political power of religious groups even in this day and age.

 

[longway]

^ buttboy is the only candidate that might be able to beat trump in this cycle.

 

[rev0lver]

lol

But I give up, you win.

I know this is getting super drawn out now lol but I just want to clarify this point - political polling is meant to represent attitudes of voters/the population. It doesn't tell you about behavior on election day, but will usually give you a general idea of what will happen within a certain range that can skew to one side or another with other factors. So a predictive model can be a bit different than showing who people like more.

 

[longway]

^ a very long winded admission that the polling was off. Finally.

 

[rev0lver]

^ a very long winded admission that the polling was off. Finally.

I'm saying the polling was probably correct that x-percent of people in whatever states supported one candidate more than the other. But that's not what an election does, and the data showed applying other factors could easily let Trump win where he was some error spots away from leading. Which is what I did, and why I laid my money on Trump.