Is this the earliest a Presidential election outcome has been determined?

[john "teach me how to" dougie]

Shouldn't it tell you something either about yourself or the person you are voting for if you're too embarrassed to admit to voting for that person?

More a lib problem than a candidate problem.

[star seed 7]

lol, ok trumpax

[star seed 7]

Shouldn't it tell you something either about yourself or the person you are voting for if you're too embarrassed to admit to voting for that person?

More a lib problem than a candidate problem.

 

[chum1]

What's the correlation between late season undecided voters and greater chance of a disproportionate break?

dunno.  not sure if there has ever even been a comparable number of undecideds in the era of modern polling.

So, why bother with it? There are lots of circumstances that obtain only this year yet have nothing to do with the election.

I could be wrong, but given the polling in this particular case, I don't think any realistic turnout scenarios could change the outcome.

i think you're wrong.  clinton seems to be up by about 4-7 pts nationally, but she is doing better in non-competitive states than competitive states, so the mean difference is less in swing states. 

If we grant Trump OH, FL, and NC, he'd still need one of PA, VA, or CO to get to 270 where Clinton leads by and 4.4, 7.2, and 6.8.

[chum1]

like right now 538's best model has it about 85:15.  those are good odds for clinton, but it's not at all like a trump win would be some crazy shock.  and not to 2nd guess the model, but if i had to bet on whether it is overly certain or underly certain, i'd bet the over.

Out of all major models, 538's has consistently had the lowest odds for Clinton. Personally, I don't see how you can legitimately get to 85% when straight poll numbers say >97%.

[MakeItRain]

I wouldn't tell anyone if I was voting for Trump. Too many smug liberals everywhere just waiting to laugh in your stupid face. But, they won't be around in the voting booth.

And when they pass the exit pollers it's all "HILLARY! YEAH"

I don't have casual conversation about voting for Jill Stein but I don't know what that has to do with polling or November 8th.

I think the trump people are afraid to be honest with players talking point is the dumbest thing I've ever heard.

[gatoveintisiete]

You guys seem pretty engaged for folks whose candidate is up 9 points with 13 days left, interesting....

[sys]

Out of all major models, 538's has consistently had the lowest odds for Clinton. Personally, I don't see how you can legitimately get to 85% when straight poll numbers say >97.

there was a post a few days back on 538 exploring why their model is more conservative than others.  i think, if anything, there is more uncertainty than is represented in their model, not less.  it's important to remember that all these models are based on small samples and it is not certain that the population from which those samples were drawn is the same (similar enough) to the population currently being predicted.

[sys]

dunno.  not sure if there has ever even been a comparable number of undecideds in the era of modern polling.

So, why bother with it? There are lots of circumstances that obtain only this year yet have nothing to do with the election.

i mean, you can't just discard a variable because you can't model it (not saying it can't be modeled, i don't know the data well enough to say how well it can or can't be modeled), you still have to account for it.

If we grant Trump OH, FL, and NC, he'd still need one of PA, VA, or CO to get to 270 where Clinton leads by and 4.4, 7.2, and 6.8.

these things aren't independent events.  if trump wins oh, fl and nc, it is very likely that he will outperform his poll #s in other states as well.  but keep in mind, i'm not arguing that trump is likely to win.  i'm arguing that there is more uncertainty remaining than this thread title proclaims.  it is also more likely that clinton blows the crap out of trump than it would be if the election were more typical (given the same difference in polls).

[Fake Sugar Dick (WARNING, NOT THE REAL SUGAR DICK!)]

Clinton supporters like lib making fun of trump supporters makes me think of a guy with downs calling a guy with severe autism a Fake Sugar Dick (WARNING, NOT THE REAL SUGAR DICK!).

[chum1]

dunno.  not sure if there has ever even been a comparable number of undecideds in the era of modern polling.

So, why bother with it? There are lots of circumstances that obtain only this year yet have nothing to do with the election.

i mean, you can't just discard a variable because you can't model it (not saying it can't be modeled, i don't know the data well enough to say how well it can or can't be modeled), you still have to account for it.

If we grant Trump OH, FL, and NC, he'd still need one of PA, VA, or CO to get to 270 where Clinton leads by and 4.4, 7.2, and 6.8.

these things aren't independent events.  if trump wins oh, fl and nc, it is very likely that he will outperform his poll #s in other states as well.  but keep in mind, i'm not arguing that trump is likely to win.  i'm arguing that there is more uncertainty remaining than this thread title proclaims.  it is also more likely that clinton blows the crap out of trump than it would be if the election were more typical (given the same difference in polls).

I'm not discarding a variable that isn't modeled. I'm ignoring things not evidenced to be relevant. Doing otherwise requires factoring in anything and everything. It's absurd.

For the second part, I was directly responding to your claim in the conext of the possibility of turnout affecting outcome. I agree that if all suburban housewives in the Philadelphia suburbs don't vote, Trump may win PA. I disagree that this is something that could actually happen in a significantly quantifiable way.

[sys]

well it's ridiculous to suggest that the % of undecided voters isn't relevant, so i don't understand what you're driving at.  pretty much the same thing with turnout (obviously, the two are correlated).

[chum1]

well it's ridiculous to suggest that the % of undecided voters isn't relevant, so i don't understand what you're driving at.  pretty much the same thing with turnout (obviously, the two are correlated).

There's nothing supporting the idea that they'll vote differently than the decided voters. No reason to think this is uncertain.

[chum1]

Out of all major models, 538's has consistently had the lowest odds for Clinton. Personally, I don't see how you can legitimately get to 85% when straight poll numbers say >97.

there was a post a few days back on 538 exploring why their model is more conservative than others.  i think, if anything, there is more uncertainty than is represented in their model, not less.  it's important to remember that all these models are based on small samples and it is not certain that the population from which those samples were drawn is the same (similar enough) to the population currently being predicted.

I think you either take the polls as representative of the population or ignore them completely. Anything in between is a pretty big guess.

[sys]

well it's ridiculous to suggest that the % of undecided voters isn't relevant, so i don't understand what you're driving at.  pretty much the same thing with turnout (obviously, the two are correlated).

There's nothing supporting the idea that they'll vote differently than the decided voters. No reason to think this is uncertain.

there is actually, but there doesn't need to be in order for their presence to contribute to increased uncertainty.  it should be self-evident that 46-40 with 14% undecided is less certain than is 51-45 with 4% undecided in the same way that a 2 run lead in 5th is less certain than is a 2 run lead in the 9th.

[sys]

I think you either take the polls as representative of the population or ignore them completely. Anything in between is a pretty big guess.

i was talking about the population of elections, not the population of the electorate.  but, regardless, your assertion doesn't make a lot of sense.  believing something is not an all or nothing proposition, there are levels of certainty.  some of that certainty can be estimated.

[chum1]

Oh, yeah, I'd agree that the populations of elections aren't really very informative. One reason I don't care that much for 538.

Probability is supposed to be an all or nothing propsition. (Or close enough.) The more guesswork you do, the more you stray from it. I don't see the point of that. If you want your feelings to be a factor, if you want to see 85%, just skip all of the polls and everything and make a chart of your feelings.

[Trim]

Half-baked election idea:

Election still decided by electoral votes, but to get the presidency the winner has to have gotten some minimum percentage of all votes cast nationally (40%?).  If the electoral winner doesn't hit that mark, new election between new candidates. 

Voters are motivated to vote none of the above or write in something ridiculous if they don't like either.

I need help thinking how the calendar for this would work to be able to get someone picked in time.  Would parties need to have someone on deck if their 1st round candidates don't have what it takes?

[SkinnyBenny]

You guys seem pretty engaged for folks whose candidate is up 9 points with 13 days left, interesting....

Yeah, and I stayed until the end of the 67-0 game too, but it wasn't because I was worried about the outcome -- it was because the losing team was pure comedy, dumbass.

[Mrs. Gooch]

Half-baked election idea:

Election still decided by electoral votes, but to get the presidency the winner has to have gotten some minimum percentage of all votes cast nationally (40%?).  If the electoral winner doesn't hit that mark, new election between new candidates. 

Voters are motivated to vote none of the above or write in something ridiculous if they don't like either.

I need help thinking how the calendar for this would work to be able to get someone picked in time.  Would parties need to have someone on deck if their 1st round candidates don't have what it takes?

I like this idea but there are going to be complications on how you determine which ballots count in the "all votes cast". If the field is left blank, does that count? If they write in a fictional character, does that count? If they write in a name that was not an official write in candidate, does that count?

[chum1]

well it's ridiculous to suggest that the % of undecided voters isn't relevant, so i don't understand what you're driving at.  pretty much the same thing with turnout (obviously, the two are correlated).

There's nothing supporting the idea that they'll vote differently than the decided voters. No reason to think this is uncertain.

there is actually, but there doesn't need to be in order for their presence to contribute to increased uncertainty.  it should be self-evident that 46-40 with 14% undecided is less certain than is 51-45 with 4% undecided in the same way that a 2 run lead in 5th is less certain than is a 2 run lead in the 9th.

There may be more uncertainty in a 100 run lead in the 5th than a 100 run lead in the 9th, but not enough to change expectations for the outcome of the game.

[Trim]

Half-baked election idea:

Election still decided by electoral votes, but to get the presidency the winner has to have gotten some minimum percentage of all votes cast nationally (40%?).  If the electoral winner doesn't hit that mark, new election between new candidates. 

Voters are motivated to vote none of the above or write in something ridiculous if they don't like either.

I need help thinking how the calendar for this would work to be able to get someone picked in time.  Would parties need to have someone on deck if their 1st round candidates don't have what it takes?

I like this idea but there are going to be complications on how you determine which ballots count in the "all votes cast". If the field is left blank, does that count? If they write in a fictional character, does that count? If they write in a name that was not an official write in candidate, does that count?

Any ballot cast, no matter what it has on it - blank, absurd, whatever. Might have to lower the minimum % of those ballots won for the electoral winner to earn the presidency.

[Trim]

Half-baked election idea:

Election still decided by electoral votes, but to get the presidency the winner has to have gotten some minimum percentage of all votes cast nationally (40%?).  If the electoral winner doesn't hit that mark, new election between new candidates. 

Voters are motivated to vote none of the above or write in something ridiculous if they don't like either.

I need help thinking how the calendar for this would work to be able to get someone picked in time.  Would parties need to have someone on deck if their 1st round candidates don't have what it takes?

I like this idea but there are going to be complications on how you determine which ballots count in the "all votes cast". If the field is left blank, does that count? If they write in a fictional character, does that count? If they write in a name that was not an official write in candidate, does that count?

Any ballot cast, no matter what it has on it - blank, absurd, whatever. Might have to lower the minimum % of those ballots won for the electoral winner to earn the presidency.

An alternative method would be having a quorum on the number of ballots that have to be cast for the election to stand, but that would discourage voting on the non-presidential matters on the ballot.  Maybe have an official "none of the above" or "abstain" option on the ballot that would be scored to determine if the electoral winner has enough national support to take office.

The timing is the tricky part.

[Mrs. Gooch]

Half-baked election idea:

Election still decided by electoral votes, but to get the presidency the winner has to have gotten some minimum percentage of all votes cast nationally (40%?).  If the electoral winner doesn't hit that mark, new election between new candidates. 

Voters are motivated to vote none of the above or write in something ridiculous if they don't like either.

I need help thinking how the calendar for this would work to be able to get someone picked in time.  Would parties need to have someone on deck if their 1st round candidates don't have what it takes?

I like this idea but there are going to be complications on how you determine which ballots count in the "all votes cast". If the field is left blank, does that count? If they write in a fictional character, does that count? If they write in a name that was not an official write in candidate, does that count?

Any ballot cast, no matter what it has on it - blank, absurd, whatever. Might have to lower the minimum % of those ballots won for the electoral winner to earn the presidency.

An alternative method would be having a quorum on the number of ballots that have to be cast for the election to stand, but that would discourage voting on the non-presidential matters on the ballot.  Maybe have an official "none of the above" or "abstain" option on the ballot that would be scored to determine if the electoral winner has enough national support to take office.

The timing is the tricky part.

You'd need a None of The Above option to differentiate between people who specifically want to vote for no one and people who don't want to vote for President at all but submitted their ballot so they could vote for other offices.

[XocolateThundarr]