Trump vs. Biden - the Final Showdown

[malchior]

Amazing clip. We've never dealt with this level of lying and worse -- despite Trump turning in performances like he did last night -- Trump still has a real chance. Polls are already starting to tighten. I can't say I'm not sinking into a feeling of hopelessness. Trump may not win but tonight illustrated that we're deep, deep, deep in the abyss and the reality is that many of us have to prepare for a worse future.

https://twitter.com/oliverdarcy/status/ ... 1373303809

[YellowKing]

The difference between Trump and past Presidents (aside from the frequency of his falsehoods), is that most of what we would consider "lies" in the past were actually spin, half-truths, or misrepresentation. Trump does all of the above, but adds in demonstrably false *actual* lies that anyone can just look up and refute. And he doesn't give a shit. That's a level of mental illness that we've never had to deal with before.

[malchior]

The difference between Trump and past Presidents (aside from the frequency of his falsehoods), is that most of what we would consider "lies" in the past were actually spin, half-truths, or misrepresentation. Trump does all of the above, but adds in demonstrably false *actual* lies that anyone can just look up and refute. And he doesn't give a shit. That's a level of mental illness that we've never had to deal with before.

It's gotten so far that some people don't want Biden to debate him unless they have a fact checking referee sitting on the stage with them. It's flat out insane and he has a very real chance of re-election. It is maddening.

[Smoove_B]

I had to watch former NJ Governor Chris Christie on Colbert last night as he's a guest that usually has a little back and forth with him. I was *astounded* to see Christie doubling down on supporting Trump and claiming it's because of his "conservative Republican values". Like...he said all that with a straight face. I guess I shouldn't be surprised - a guy that was absolutely in the room and without question involved in the bridge closures but smart enough to have plausible deniability is going to support someone like Trump on election day.

But anyway, it provided (to me, at least) some real insight to hear someone talk about how the Biden administration will raise taxes and hurt the economy by having government grow larger and imposing more rules and restrictions on people and businesses. Therefore, Trump.

Never mind everything that's been happening the last 4 years and what's currently happening now. He can't vote for Biden or support the (D) party because there's going to be environmental regulations that cripple businesses.

[Octavious]

So far realclear politics has him still getting his ass kicked. I expect a bump for the next couple of weeks, but it a fair election he's toast. This of course won't be a fair election.

[LawBeefaroni]

My wife has always side-eyed me every time I get a first-aid trauma kit or another flashlight or whatever she considers prepper stuff. She is very "ick" on my firearm hobby.

However, recently she said we need to be ready to be for a complete lockdown in chaos in November and we should also have a backup bug-out plan. This is unprecedented.

[Skinypupy]

My wife has always side-eyed me every time I get a first-aid trauma kit or another flashlight or whatever she considers prepper stuff. She is very "ick" on my firearm hobby.

However, recently she said we need to be ready to be for a complete lockdown in chaos in November and we should also have a backup bug-out plan. This is unprecedented.

Earlier this week, Mrs. Skinypupy said - in all seriousness - that she would be perfectly fine with exploring options to live somewhere outside the US if Trump wins in November (not that I really have any, but still...)

For someone who has lived in the same town her entire life (her parents have lived in the same house for 52 years) and has vehemently opposed the couple times we've had opportunity to relocate, that's unprecedented.

[LawBeefaroni]

Never thought I'd have to go looking for vegetarian MREs.

[Holman]

I legit need to go take a shot of liquor after seeing this photo

Why not frame it? You paid for the event, after all.

[$iljanus]

My wife has always side-eyed me every time I get a first-aid trauma kit or another flashlight or whatever she considers prepper stuff. She is very "ick" on my firearm hobby.

However, recently she said we need to be ready to be for a complete lockdown in chaos in November and we should also have a backup bug-out plan. This is unprecedented.

On the bright side, your firearm hobby is looking up!

[LawBeefaroni]

On the bright side, your firearm hobby is looking up!

I've actually limited my range trips because ammo is impossible to get. I keep a reserve but can't replenish every time I go shoot. Pretty much just burning 22lr when I do go.

Plus shit like that kid in Kenosha kind of take the fun out of it.


I will say this, historically the range I go to usually has mostly white, male shooters. Because of the location, a decent amount are Orthodox jewish but that's about the limit of diversity. There are always BIPOC (that's the term, right?), including myself, and women but it's mostly white guys. Since May or so, though, it has been half or less white males. More women (solo, not with BF/husband making them shoot) and a ton of black shooters. Same with gun purchases.

[malchior]

I've been quietly prepping too. I'm not expecting it to be unrest..moreso preparing for another lock down though I do expect it to be real dicey post Nov. 3.

[Kraken]

So far realclear politics has him still getting his ass kicked. I expect a bump for the next couple of weeks, but it a fair election he's toast. This of course won't be a fair election.

Up until the past couple of weeks it looked like trump had given up on winning the vote and was concentrating on invalidating the election instead. He's still laying the groundwork for that -- prevent voting, suppress or challenge results, and generally de-legitimize any swing states that don't swing his way. But this new "lawn order" schtick gives him a chance to actually win the vote if he can successfully provoke enough violence to make suburbanites fearful. Ordinarily you wouldn't expect burning cities to boost the incumbents, but "ordinarily" hasn't pertained for at least the past four years now.

It's obviously not going to be a fair election; they've been openly colluding with Russians ever since the Mueller report fizzled. But it's our last shot.

[stessier]

I wonder if it add or subtracts from the danger that it will likely be at least a week before we have a final result. Probably depends on the result.

[Ralph-Wiggum]

I wonder if it add or subtracts from the danger that it will likely be at least a week before we have a final result. Probably depends on the result.

It 100% adds to the danger. Part of Trump's railing against mail-in voting is that he can then use any delay in the finally tally to sow mistrust of the result.

[$iljanus]

On the bright side, your firearm hobby is looking up!

Plus shit like that kid in Kenosha kind of take the fun out of it.

Yeah, reasonable people suffer the consequences of asshats who have no sense of reason.

[stessier]

I wonder if it add or subtracts from the danger that it will likely be at least a week before we have a final result. Probably depends on the result.

It 100% adds to the danger. Part of Trump's railing against mail-in voting is that he can then use any delay in the finally tally to sow mistrust of the result.

Right, but what if he's leading going into the mail-in counting and ends up winning? I mean, it would be darkly amusing that he would spend all that time railing against the votes only to come out on top, but I don't think it would increase the danger like the opposite result.

[El Guapo]

https://twitter.com/NateSilver538/statu ... 6441418754

FWIW. In general at this point there's more worry about the race moving towards Trump than there is evidence that the race is moving towards Trump.

[malchior]

I wonder if it add or subtracts from the danger that it will likely be at least a week before we have a final result. Probably depends on the result.

It 100% adds to the danger. Part of Trump's railing against mail-in voting is that he can then use any delay in the finally tally to sow mistrust of the result.

Right, but what if he's leading going into the mail-in counting and ends up winning? I mean, it would be darkly amusing that he would spend all that time railing against the votes only to come out on top, but I don't think it would increase the danger like the opposite result.

That doesn't really matter. If he's leading, he'll silently drop his argument upon winning. He'll just move to quickly consolidate more power immediately. Unrestrained he'll move onto to building out a personalistic autocracy. The entire purpose for the "warnings" are to prep for the high chance that he loses a fair election and then makes some move to keep power. He had parts of his family break the law on tv last night and Ivanka implicitly staked a claim as his successor. This is a very dangerous time. More dangerous than again I think people understand.

[coopasonic]

FWIW. In general at this point there's more worry about the race moving towards Trump than there is evidence that the race is moving towards Trump.

Can you blame us for worrying?

[El Guapo]

FWIW. In general at this point there's more worry about the race moving towards Trump than there is evidence that the race is moving towards Trump.

Can you blame us for worrying?

Oh, we should absolutely be terrified, especially given the shenanigans that Trump will engage in and his electoral college advantage, and the consequences of his winning. I'm just saying that the polling situation isn't too worrisome at the moment.

[Skinypupy]

Guess who now owns KeepAmericaGreat.com?

Go on, guess.

[malchior]

FWIW. In general at this point there's more worry about the race moving towards Trump than there is evidence that the race is moving towards Trump.

Can you blame us for worrying?

We need more battleground polling to see how close to doom we sit. This national average stuff doesn't mean much right now. We don't have much but we do have some evidence that Trump might hold in Wisconsin. Polls have Biden at similar levels or underperforming Clinton's numbers in 2016 as compared to the same period in 2020. And that's before Kenosha. That ain't great.

[stessier]

I wonder if it add or subtracts from the danger that it will likely be at least a week before we have a final result. Probably depends on the result.

It 100% adds to the danger. Part of Trump's railing against mail-in voting is that he can then use any delay in the finally tally to sow mistrust of the result.

Right, but what if he's leading going into the mail-in counting and ends up winning? I mean, it would be darkly amusing that he would spend all that time railing against the votes only to come out on top, but I don't think it would increase the danger like the opposite result.

That doesn't really matter. If he's leading, he'll silently drop his argument upon winning. He'll just move to quickly consolidate more power immediately. Unrestrained he'll move onto to building out a personalistic autocracy. The entire purpose for the "warnings" are to prep for the high chance that he loses a fair election and then makes some move to keep power. He had parts of his family break the law on tv last night and Ivanka implicitly staked a claim as his successor. This is a very dangerous time. More dangerous than again I think people understand.

I've been busy, which laws did family members break last night? If it's the Hatch Act, how much of a law is it if there are no penalties or enforcement mechanisms?

[El Guapo]

FWIW. In general at this point there's more worry about the race moving towards Trump than there is evidence that the race is moving towards Trump.

Can you blame us for worrying?

We need more battleground polling to see how close to doom we sit. This national average stuff doesn't mean much right now. We don't have much but we do have some evidence that Trump might hold in Wisconsin. Polls have Biden at similar levels or underperforming Clinton's numbers in 2016 as compared to the same period in 2020. And that's before Kenosha. That ain't great.

That's part of the electoral college advantage. Biden's up ~ 8 - 9 points nationally, and up ~ 5ish points in the tipping point states (which right now are I think PA & FL). Trump's not doomed yet because he only needs to nudge the polls a point or two in his direction to be within a polling error / shenanigan bump of victory (and because there's still time for him to make a comeback).

[malchior]

Right. But the 5% we saw last time isn't a hard level. It fluctuates depending on the turnout and importantly who and where they turnout. It might be 2% or 8% this year for all we know. That is why I just can't be bothered to care about national level polling except as a wide thumb in the air metric. 8% sounds great but until we get into the details it really isn't all that meaningful.

[El Guapo]

Right. But the 5% we saw last time isn't a hard level. It fluctuates depending on the turnout and importantly who and where they turnout. It might be 2% or 8% this year for all we know. That is why I just can't be bothered to care about national level polling except as a wide thumb in the air metric. 8% sounds great but until we get into the details it really isn't all that meaningful.

The 5% is based on current polling, not based on 2016 or anything. I agree that the 8% isn't consequential in and of itself, but it's useful as a supplement to state level polling, especially in states where there's less polling available (since there is at least a general correlation between state and national polling).

[LawBeefaroni]

Remember, this is 2020. An 8% margin could Biden victory or bloody civil war. Or free tacos from Taco Bell.

It's 2020.

[malchior]

Right. But the 5% we saw last time isn't a hard level. It fluctuates depending on the turnout and importantly who and where they turnout. It might be 2% or 8% this year for all we know. That is why I just can't be bothered to care about national level polling except as a wide thumb in the air metric. 8% sounds great but until we get into the details it really isn't all that meaningful.

The 5% is based on current polling, not based on 2016 or anything. I agree that the 8% isn't consequential in and of itself, but it's useful as a supplement to state level polling, especially in states where there's less polling available (since there is at least a general correlation between state and national polling).

Right my reference to 5% wasn't intentionally picking up on the 5% you are mentioning. It just was about the level of imbalance in 2016. I'm personally not bothering with tracking the polls or anyone's models until we start to get good data.

[gameoverman]

I wonder if it add or subtracts from the danger that it will likely be at least a week before we have a final result. Probably depends on the result.

It 100% adds to the danger. Part of Trump's railing against mail-in voting is that he can then use any delay in the finally tally to sow mistrust of the result.

Right, but what if he's leading going into the mail-in counting and ends up winning? I mean, it would be darkly amusing that he would spend all that time railing against the votes only to come out on top, but I don't think it would increase the danger like the opposite result.

The way I think his strategy works is this: He sows mistrust of the vote. The election is held. If he wins, for any reason, then he spins it as "We have overcome the leftist fraud" and congratulates his supporters on their great victory. If he loses then he spins it as an "I told you there would be fraud" result. In that case he does something wacky like issuing an executive emergency order suspending the election results pending a fraud investigation.

This theory of mine is based on seeing how his doublespeak works. He always needs to be right so he always needs some way to say he knew what was going to happen. With the election he's just covering all the angles so no matter what the result he can tell his people "I told you so".

[El Guapo]

Right. But the 5% we saw last time isn't a hard level. It fluctuates depending on the turnout and importantly who and where they turnout. It might be 2% or 8% this year for all we know. That is why I just can't be bothered to care about national level polling except as a wide thumb in the air metric. 8% sounds great but until we get into the details it really isn't all that meaningful.

The 5% is based on current polling, not based on 2016 or anything. I agree that the 8% isn't consequential in and of itself, but it's useful as a supplement to state level polling, especially in states where there's less polling available (since there is at least a general correlation between state and national polling).

Right my reference to 5% wasn't intentionally picking up on the 5% you are mentioning. It just was about the level of imbalance in 2016. I'm personally not bothering with tracking the polls or anyone's models until we start to get good data.

What do you mean by "the level of imbalance in 2016"?

[malchior]

Right. But the 5% we saw last time isn't a hard level. It fluctuates depending on the turnout and importantly who and where they turnout. It might be 2% or 8% this year for all we know. That is why I just can't be bothered to care about national level polling except as a wide thumb in the air metric. 8% sounds great but until we get into the details it really isn't all that meaningful.

The 5% is based on current polling, not based on 2016 or anything. I agree that the 8% isn't consequential in and of itself, but it's useful as a supplement to state level polling, especially in states where there's less polling available (since there is at least a general correlation between state and national polling).

Right my reference to 5% wasn't intentionally picking up on the 5% you are mentioning. It just was about the level of imbalance in 2016. I'm personally not bothering with tracking the polls or anyone's models until we start to get good data.

What do you mean by "the level of imbalance in 2016"?

That is what I remember as being the imbalance some academics calculated between the electoral college and the popular vote in 2016 - as I recall it was based on some aggregate analysis. Trump got ~46% of the vote but got ~56% of the electoral college. So the theory (as I recall it) was that he had another 3% of runway to keep the majority of the EC. In other words, he could have lost by 5% and still gotten the EC majority. That number slides around a bit over time but generally persistently favors the GOP.

Edit: It isn't very predictive (at least for now) and you need to do a look back to calculate it. That's why I don't really care about the national polls. You can't really use them predictively. There is a correlation perhaps between a state and national but it has a major limitation that you can't fit it into the puzzle on its own.

[LawBeefaroni]

I wonder if it add or subtracts from the danger that it will likely be at least a week before we have a final result. Probably depends on the result.

It 100% adds to the danger. Part of Trump's railing against mail-in voting is that he can then use any delay in the finally tally to sow mistrust of the result.

Right, but what if he's leading going into the mail-in counting and ends up winning? I mean, it would be darkly amusing that he would spend all that time railing against the votes only to come out on top, but I don't think it would increase the danger like the opposite result.

The way I think his strategy works is this: He sows mistrust of the vote. The election is held. If he wins, for any reason, then he spins it as "We have overcome the leftist fraud" and congratulates his supporters on their great victory. If he loses then he spins it as an "I told you there would be fraud" result. In that case he does something wacky like issuing an executive emergency order suspending the election results pending a fraud investigation.

This theory of mine is based on seeing how his doublespeak works. He always needs to be right so he always needs some way to say he knew what was going to happen. With the election he's just covering all the angles so no matter what the result he can tell his people "I told you so".

That's exactly what his campaign did in 2016. Calling the upcoming election rigged and planting seeds of doubt. When he was elected that all disappeared, until now when he's doing it again.

Russia had an avalanche of bots and other strategies ready to question the validity 2016 results. They were all pulled back once Trump won.

[El Guapo]

Right. But the 5% we saw last time isn't a hard level. It fluctuates depending on the turnout and importantly who and where they turnout. It might be 2% or 8% this year for all we know. That is why I just can't be bothered to care about national level polling except as a wide thumb in the air metric. 8% sounds great but until we get into the details it really isn't all that meaningful.

The 5% is based on current polling, not based on 2016 or anything. I agree that the 8% isn't consequential in and of itself, but it's useful as a supplement to state level polling, especially in states where there's less polling available (since there is at least a general correlation between state and national polling).

Right my reference to 5% wasn't intentionally picking up on the 5% you are mentioning. It just was about the level of imbalance in 2016. I'm personally not bothering with tracking the polls or anyone's models until we start to get good data.

What do you mean by "the level of imbalance in 2016"?

That is what I remember as being the historical imbalance some academics calculated between the electoral college and the popular vote in 2016 - as I recall it was some aggregate measure. Trump got ~46% of the vote but got ~56% of the electoral college. So the theory (as I recall it) was that he had another 3% of runway to keep the majority of the EC. In other words, he could have lost by 5% and still gotten the EC majority. That number slides around a bit over time but generally persistently favors the GOP.

I'm not familiar with that, but that sounds like an overly complicated and less than ideal way of calculating the effective EC edge. Sounds like it's based off of the national popular vote? But that's a bit misleading because that doesn't matter - like if Clinton had gotten every vote in California that would skew the result this would produce, even though that wouldn't impact the electoral college math at all (since there was no plausible scenario where Clinton wouldn't win CA). Though again, I'm not sure that I understand the analysis.

The key question is what combination of states produce a Trump electoral college win, and what does polling look like in those states. I believe from my recollection of the polls is that Trump could win an electoral college majority by winning all states where the polls no worse for him than Biden +5. He could do that while losing the popular vote massively (e.g., give Biden every vote in states that Biden wins, give Trump a plurality of one vote in every state that Trump wins), or while losing the popular vote narrowly.

Anyway, I think it's fair to look at Biden's 'true' margin as more like Biden +5ish given the electoral college, which is doable for Trump but not easy.

[malchior]

I wonder if it add or subtracts from the danger that it will likely be at least a week before we have a final result. Probably depends on the result.

It 100% adds to the danger. Part of Trump's railing against mail-in voting is that he can then use any delay in the finally tally to sow mistrust of the result.

Right, but what if he's leading going into the mail-in counting and ends up winning? I mean, it would be darkly amusing that he would spend all that time railing against the votes only to come out on top, but I don't think it would increase the danger like the opposite result.

That doesn't really matter. If he's leading, he'll silently drop his argument upon winning. He'll just move to quickly consolidate more power immediately. Unrestrained he'll move onto to building out a personalistic autocracy. The entire purpose for the "warnings" are to prep for the high chance that he loses a fair election and then makes some move to keep power. He had parts of his family break the law on tv last night and Ivanka implicitly staked a claim as his successor. This is a very dangerous time. More dangerous than again I think people understand.

I've been busy, which laws did family members break last night? If it's the Hatch Act, how much of a law is it if there are no penalties or enforcement mechanisms?

Yes Hatch act. Ivanka is a federal employee. And there are penalties if enforced. She could be terminated from employment, she could be fined, or forbidden to work in the government if the OSC and DOJ did their jobs and the President acted on the recommendations. This is the point though. They are openly breaking the law because they can get away with it which is one of key indicators of late-stage fascism.

The penalty structure for violations of the Hatch Act by federal employees includes removal from federal service, reduction in grade, debarment from federal employment for a period not to exceed 5 years, suspension, reprimand, or a civil penalty not to exceed $1,000.

[Defiant]

One reason for Clinton's large popular vote lead was because one of her strategists focused on running up the score in California at the expense of attention in the swing states (in the hope that a large popular vote victory would eliminate accusations of the election being rigged). One hopes that Biden and his advisors don't make the same mistake and focus on the swing states and specifically the tipping point states.

[Smoove_B]

Yes Hatch act. Ivanka is a federal employee. And there are penalties if enforced. She could be terminated from employment, she could be fined, or forbidden to work in the government if the OSC and DOJ did their jobs and the President acted on the recommendations. This is the point though. They are openly breaking the law because they can get away with it which is one of key indicators of late-stage fascism.

Which is also where the media is failing. I'm of the opinion they should be asking every sitting GOP member about this (and anything else where laws have been casually ignored) at every single press conference they attend. In addition, people like Nancy Pelosi should be on camera daily calling this stuff out and demanding that people like Mitch McConnell answer as to why things like this are permitted. Particularly now that we're remote, she can hop online and give a 5 minute press conference and take a few questions. There needs to be constant reminders and constant pushback on what started as low level but is now escalating into common and frequent because they weren't pressed on it initially.

That's why sports figures refusing to play is so amazing. They're making sure people aren't getting distracted and lulled into normalcy by tuning in (and dropping out) for a few hours.

[malchior]

Right. But the 5% we saw last time isn't a hard level. It fluctuates depending on the turnout and importantly who and where they turnout. It might be 2% or 8% this year for all we know. That is why I just can't be bothered to care about national level polling except as a wide thumb in the air metric. 8% sounds great but until we get into the details it really isn't all that meaningful.

The 5% is based on current polling, not based on 2016 or anything. I agree that the 8% isn't consequential in and of itself, but it's useful as a supplement to state level polling, especially in states where there's less polling available (since there is at least a general correlation between state and national polling).

Right my reference to 5% wasn't intentionally picking up on the 5% you are mentioning. It just was about the level of imbalance in 2016. I'm personally not bothering with tracking the polls or anyone's models until we start to get good data.

What do you mean by "the level of imbalance in 2016"?

That is what I remember as being the historical imbalance some academics calculated between the electoral college and the popular vote in 2016 - as I recall it was some aggregate measure. Trump got ~46% of the vote but got ~56% of the electoral college. So the theory (as I recall it) was that he had another 3% of runway to keep the majority of the EC. In other words, he could have lost by 5% and still gotten the EC majority. That number slides around a bit over time but generally persistently favors the GOP.

I'm not familiar with that, but that sounds like an overly complicated and less than ideal way of calculating the effective EC edge. Sounds like it's based off of the national popular vote? But that's a bit misleading because that doesn't matter - like if Clinton had gotten every vote in California that would skew the result this would produce, even though that wouldn't impact the electoral college math at all (since there was no plausible scenario where Clinton wouldn't win CA). Though again, I'm not sure that I understand the analysis.

That's the point. It can change based on the factors you mention. It is intended as a metric for electoral legitimacy but also as a look back number (more on this below). Not who won by the rules but how representative the Democracy is. That is why I personally ignore the national polls. Unless you know where they polled and got sufficient representation...it doesn't mean squat really.

The key question is what combination of states produce a Trump electoral college win, and what does polling look like in those states. I believe from my recollection of the polls is that Trump could win an electoral college majority by winning all states where the polls no worse for him than Biden +5. He could do that while losing the popular vote massively (e.g., give Biden every vote in states that Biden wins, give Trump a plurality of one vote in every state that Trump wins), or while losing the popular vote narrowly.

This is the problem with modeling. It can get convoluted. We don't have portions of that battleground data and using the national polls as some indicator where they are is a flimsy rationale at best. In fact, I think that +5 number is derived from the analysis I mentioned above -- the look back analysis is what is used to set Biden +5 as a bar in battlegrounds. It isn't predictive but it is a metric to measure against what happened last time like a stock signal. You can't know how accurate it is though. It is a rough measure.

Anyway, I think it's fair to look at Biden's 'true' margin as more like Biden +5ish given the electoral college, which is doable for Trump but not easy.

Sure - it's as rough as guess as any other. It'll get clearer when the data comes in.

[Zaxxon]

I'm personally not bothering with tracking the polls or anyone's models until we start to get good data.

Are... Are you saying it's too early for polls?

[El Guapo]

Right. But the 5% we saw last time isn't a hard level. It fluctuates depending on the turnout and importantly who and where they turnout. It might be 2% or 8% this year for all we know. That is why I just can't be bothered to care about national level polling except as a wide thumb in the air metric. 8% sounds great but until we get into the details it really isn't all that meaningful.

The 5% is based on current polling, not based on 2016 or anything. I agree that the 8% isn't consequential in and of itself, but it's useful as a supplement to state level polling, especially in states where there's less polling available (since there is at least a general correlation between state and national polling).

Right my reference to 5% wasn't intentionally picking up on the 5% you are mentioning. It just was about the level of imbalance in 2016. I'm personally not bothering with tracking the polls or anyone's models until we start to get good data.

What do you mean by "the level of imbalance in 2016"?

That is what I remember as being the historical imbalance some academics calculated between the electoral college and the popular vote in 2016 - as I recall it was some aggregate measure. Trump got ~46% of the vote but got ~56% of the electoral college. So the theory (as I recall it) was that he had another 3% of runway to keep the majority of the EC. In other words, he could have lost by 5% and still gotten the EC majority. That number slides around a bit over time but generally persistently favors the GOP.

I'm not familiar with that, but that sounds like an overly complicated and less than ideal way of calculating the effective EC edge. Sounds like it's based off of the national popular vote? But that's a bit misleading because that doesn't matter - like if Clinton had gotten every vote in California that would skew the result this would produce, even though that wouldn't impact the electoral college math at all (since there was no plausible scenario where Clinton wouldn't win CA). Though again, I'm not sure that I understand the analysis.

That's the point. It can change based on the factors you mention. It is intended as a metric for electoral legitimacy but also as a look back number (more on this below). Not who won by the rules but how representative the Democracy is. That is why I personally ignore the national polls. Unless you know where they polled and got sufficient representation...it doesn't mean squat really.

The key question is what combination of states produce a Trump electoral college win, and what does polling look like in those states. I believe from my recollection of the polls is that Trump could win an electoral college majority by winning all states where the polls no worse for him than Biden +5. He could do that while losing the popular vote massively (e.g., give Biden every vote in states that Biden wins, give Trump a plurality of one vote in every state that Trump wins), or while losing the popular vote narrowly.

This is the problem with modeling. It can get convoluted. We don't have portions of that battleground data and using the national polls as some indicator where they are is a flimsy rationale at best. In fact, I think that +5 number is derived from the analysis I mentioned above -- the look back analysis is what they used to set Biden +5 as a bar in battlegrounds. It isn't predictive but it is a metric to measure against what happened last time like a stock signal. You won't know if it is accurate though. It is a rough measure.

Anyway, I think it's fair to look at Biden's 'true' margin as more like Biden +5ish given the electoral college, which is doable for Trump but not easy.

Sure - it's as rough as guess as any other. It'll get clearer when the data comes in.

I guess I don't really care about electoral legitimacy at this point, so much as "how likely it is that Biden will win the electoral college and therefore the election?" And the Biden +5 that I'm talking about is not based on abstract electoral legitimacy models, it's based on Biden being up by about that amount in Florida and Pennsylvania, etc., which are disproportionately likely to be the states that decide who wins the electoral college majority. So it's not rough guesswork, it's based on the current state of the race based on state-level polling.

Of course, the state of the race is subject to change, and this week and last week are high news weeks (from the conventions + Jacob Blake and Kenosha, among other things), so polling over the next couple weeks will be particularly significant.