US Presidential Election 2020 Thread II - Judgement Day(s)

serfboard

Christy42 wrote: »

Biden will have the edge in that he is able to talk to people and to negotiate so he will be able to occasionally work with Republicans.

Doubt it. As was said here before, if the Republicans win the Senate, Dr. No (Mitch McConnell) will be back.

road_high

I’d seen Florida go strongly for Trump and assumed that was that but seems the story is elsewhere this time

Cantstandsya

MJohnston wrote: »

The craziest thing, to me, is how close the race has actually ended up hewing to forecast models. For all the early insanity thanks to Florida, most states have gone the way they were predicted to, even if the margins were less than assumed.

That’s why it’s always worth digging into the data that the likes of 538 provide — the topline numbers are one thing, but always arm yourself with an understanding of margins of error, state by state uncertainties, and potential polling problems (the Latino vote problem was flagged well in advance), and therefore what are the range of plausible outcomes.

I'm the first to put my hand up and admit to being clueless about statistics.

It's something I think schools should put a lot more emphasis on.

MJohnston

road_high wrote: »

The postal vote factor seems to be a big thing this time- went to bed this morning assuming a Trump win tbh but it’s only in the past 3 hours or so things seemed to be shaping up differently

This was another thing that was flagged well in advance, yet the TV networks largely avoided making it clear.

Consult 538s when to expect results guide (https://projects.fivethirtyeight.com/election-results-timing/) and it was always clear that the Midwest states were going to see big blue shifts late in the race.

On CNN though, it was frantic bull****.

jasonb

Trump's lead in Michigan increases to 24,000, Biden hasn't flipped it Blue yet!

Retr0gamer

Michigan swinging back red but still a massive amount of Detroit votes to come in.

Inquitus

Deleted User wrote: »

Does Biden need the Senate to increase the size of the Supreme Court?

Aye if he gets rid of the filibuster for court packing he would need 50 Senators + The VP's deciding vote to proceed.

Timberrrrrrrr

Doctor Jimbob wrote: »

Personally I hope everyone who put up with accused of being a "snowflake" or had people delight in their alleged tears give as good as they got and then some over the next few years.

Oh there are tears still, tears of joy knowing that Trump is most likely gone

rameire

Info from CNN and RTE

MJohnston

Cantstandsya wrote: »

I'm the first to put my hand up and admit to being clueless about statistics.

It's something I think schools should put a lot more emphasis on.

I’m not all that strong on statistics myself, but it’s like when you see a national Biden forecast of +8 then that doesn’t tell you how close things might be state by state. And that’s what’s really important in the presidential race.

FL was forecast as Biden +2 and looks like it went Trump +3. At a state level, that’s not a particularly unexpected flip, because the sample sizes for polls are smaller, and therefore the margins of error are larger.

Anytime you see a Democrat forecast for a red leaning state by less than 4 or 5, you have to understand that this probably makes it a bit of a tossup.

RoryMac

Hard to see Trump hold MI or turn WI around on those numbers

OhHiMark

MJohnston wrote: »

This was another thing that was flagged well in advance, yet the TV networks largely avoided making it clear.

Consult 538s when to expect results guide (https://projects.fivethirtyeight.com/election-results-timing/) and it was always clear that the Midwest states were going to see big blue shifts late in the race.

On CNN though, it was frantic bull****.

CNN were constantly, and I mean constantly to the point that it was getting ridiculous, saying how we can't read much into the early numbers and that the postal votes counted later would be important. They were even giving percentages of how many current votes were postal votes, and the projected total postal votes.

is_that_so

Inquitus wrote: »

Aye if he gets rid of the filibuster for court packing he would need 50 Senators + The VP's deciding vote to proceed.

TBH it's very much a bi-partisan project and they'd all get credit for it.

Christy42

serfboard wrote: »

Doubt it. As was said here before, if the Republicans win the Senate, Dr. No (Mitch McConnell) will be back.

He will. And it will hamstring the Dems but he wasn't able to bring them all along on everything and it will require little to get to a tie that Harris can break

AdamD

MJohnston wrote: »

I’m not all that strong on statistics myself, but it’s like when you see a national Biden forecast of +8 then that doesn’t tell you how close things might be state by state. And that’s what’s really important in the presidential race.

FL was forecast as Biden +2 and looks like it went Trump +3. At a state level, that’s not a particularly unexpected flip, because the sample sizes for polls are smaller, and therefore the margins of error are larger.

Anytime you see a Democrat forecast for a red leaning state by less than 4 or 5, you have to understand that this probably makes it a bit of a tossup.

I would have thought a 5 point flip is quite big? Aren't margins of error usually 3 points?

If your poll is 5 points off its worse than useless

Inquitus

is_that_so wrote: »

TBH it's very much a bi-partisan project and they'd all get credit for it.

The GOP are never going to jeopardise their majority on the Supreme Court working with the Dems. They jammed Coney-Barrett through for precisely that reason. Only way Dems can balance the court is if they control the Senate.

Jimmy Bottlehead

Doctor Jimbob wrote: »

Personally I hope everyone who put up with accused of being a "snowflake" or had people delight in their alleged tears give as good as they got and then some over the next few years.

It'll be very amusing. The cognitive dissonance at play is fascinating; their man has had four years to 'drain the swamp', in the most powerful position in the world, and he STILL couldn't rid the state of these shadowy conspiratorial powers who've sabotaged his election win?

Doesn't sound like a tough talking success story to me...

is_that_so

Christy42 wrote: »

He will. And it will hamstring the Dems but he wasn't able to bring them all along on everything and it will require little to get to a tie that Harris can break

Biden's inclination and ability to deal with people may serve some good purpose even if numbers favour the GOP in the Senate.

prawnsambo

AdamD wrote: »

I would have thought a 5 point flip is quite big? Aren't margins of error usually 3 points?

If your poll is 5 points off its worse than useless

They're always +/- MoE. So going +3 to -2 is inside a 3 point MoE.

MJohnston

AdamD wrote: »

I would have thought a 5 point flip is quite big? Aren't margins of error usually 3 points?

If your poll is 5 points off its worse than useless

In a red swing state? No that’s what I’d call a normal polling error. The MoE on state polls is usually around 4% because of the smaller sample sizes.

MJohnston

prawnsambo wrote: »

They're always +/- MoE. So going +3 to -2 is inside a 3 point MoE.

Yep - margin of error applies to both candidates in each poll. This is something that I don’t think many people are aware of. Didn’t realise it myself until this year

mirrorwall14

Cantstandsya wrote: »

I'm the first to put my hand up and admit to being clueless about statistics.

It's something I think schools should put a lot more emphasis on.

Depending on when you came out of school there was a huge shift to statistics for all with Project maths. I’m very pro it being on the course though I’ll admit it can be difficult for students to get their head around. But it’s always fun to teach misleading statistics

MJohnston

To put it another way, if a poll says Biden 52 and Trump 48, and the MoE is 3%, then that’s saying Biden could get anywhere from 49 to 55 and Trump anywhere from 51 to 45.

A lot of people think the MoE applies to the margin, so they’d see 52-48 and say that Biden was outside the MoE because there’s a 4 point gap.

murpho999

OhHiMark wrote: »

CNN were constantly, and I mean constantly to the point that it was getting ridiculous, saying how we can't read much into the early numbers and that the postal votes counted later would be important. They were even giving percentages of how many current votes were postal votes, and the projected total postal votes.

I really think CNN did a great job on this and really said that you have to wait for votes to be counted. Going in at country level and showing what votes are still to come on and how they compared to 2016 really showed to me that Biden always had a chance.

People calling this as a done deal for Trump and relying on betting exchanges for results were wide of the mark.

terryduff12

terryduff12 wrote: »

4 years later and people couldn't accept the fact trump won, tried everything to get him out. So if trump goes down the same route it will go to show both sides are as bad as each other.

Who threatened to go to the Supreme court when things looked like they might turn against them?

BorneTobyWilde

Fox still pushing the line, '' how far democrats are out of touch with the american people, with their polling, that Trump has won all the states he was suppose to be losing in.''

Mellor

MJohnston wrote: »

To put it another way, if a poll says Biden 52 and Trump 48, and the MoE is 3%, then that’s saying Biden could get anywhere from 49 to 55 and Trump anywhere from 51 to 45.

A lot of people think the MoE applies to the margin, so they’d see 52-48 and say that Biden was outside the MoE because there’s a 4 point gap.

That's just people being bad at maths tbh.
It should be obvious that if 52% is 2points away from being equal, not the 4pts from 52 to 48.

UpBack1234

BorneTobyWilde wrote: »

Fox still pushing the line, '' how far democrats are out of touch with the american people, with their polling, that Trump has won all the states he was suppose to be losing in.''

There is a degree of truth in this tbh- not on individual state polls necessarily (did anyone REALLY think he'd lose TX/FL?) - but to the extent that Trump's IRL popularity among middle Americans has been underplayed by many on "our" side.

MJohnston

Mellor wrote: »

That's just people being bad at maths tbh.
It should be obvious that if 52% is 2points away from being equal, not the 4pts from 52 to 48.

Yeah but these races are usually reported in terms of vote margins (Biden +5, Trump +3 or whatever) which can obscure for most people what the error is actually referring to.

Mr.Nice Guy

I'm worried Trump will take Michigan. It looks very tight.