One week away:

[pbrower2a]

First I can tell us all at this stage that Trump has done nothing to bring any state or district that he lost by 10% or more in 2016 into play. New Mexico was on the fringe of contention in 2016, and it spirals past the event horizon of the black hole for Trump chances:

Not even light can pass from beyond the event horizon of a black hole. That is 188 electoral votes completely lost to Trump a long time ago.

It gets worse. Colorado, Minnesota, Maine at-large, New Hampshire, and Virginia probably went behind the event horizon a couple months ago.

 

Virginia is about the least populist of states, and it has usually been sympathetic to Republicans on issues of national security. Obama could win on that in 2008 and 2012 (the latter especially after whacking Osama bin Laden). Donald Trump is about as attractive on national security as was George McGovern in 1972. I have seen some double-digit leads for Biden in Colorado, Maine at-large, Minnesota, and New Hampshire. Nevada polls are super-solid and likely undercount Hispanic voters for whom Trump has little to offer. Whether an object went past the event horizon two months ago or two years ago makes no obvious distinction. I have the "new" states in the "black hole of Trump chances" in medium red for now.  Add those 44 votes and Trump has lost 232 electoral votes. Late news: anyone who thinks that President Trump has done anything to help himself in Michigan after the plot to kidnap the Governor must be using some very potent, dangerous, and illegal drugs. 248 electoral votes are now in the grasp of the black hole of legitimate despair for anyone who wants to see a Trump win. All that keeps Trump from being wolverine food in the wilds of Michigan is that Michigan doesn't really have wolverines.

Trump is 22 electoral votes away from losing this election because of states and districts that he has no realistic chance of winning.

 

[pbrower2a]

........

Trump is 22 electoral votes away from losing this election because of states and districts that he has no realistic chance of winning.

Time to election  |1 point|5 points||10 points|20 points|
one day............. |...64%|....95%|.....99.7%|.99.999%|
one week........... |...60%|....89%|.......98%|...99.97%|
one month......... |...57%|....81%|.......95%|.....99.7%|
three months..... |...55%|....72%|.......87%|........98%|
six months..........|...53%|....66%|.......79%|.......93%|
one year.............|....52%|...59%|.......67%|.......81%|

Trump did not do what he needed to do to get re-elected. His election campaign looks more like an ego trip than a competent effort even if winning were simply unlikely.  As in any timed sporting contest the unit of measurement (minutes on a clock or days on a calendar) ultimately decides whether the team or contestant behind has any chance at all.

Here I am using a map without electoral votes. with states and districts  that I deem Trump to have lost irrevocably (note the "black hole" reference) in the darkest shade of red:

 

OK. Everything in deep red is basically out of reach for Trump.

More data:


Michigan
52% Biden
42% Trump

Pennsylvania
52% Biden
44% Trump

Wisconsin
53% Biden
44% Trump


 https://mobile.twitter.com/kabir_here/status/1320372045943799809

Other great data for Biden in FL the EV polling has Biden winning those who already voted early 61-37 while ED is 59/40 in Trump’s favor and NC has similar data
https://mobile.twitter.com/kabir_here/status/1320372061873885190

Leads with a week to go, and the chance of the leader winning based on Senate races (Nate Silver with my interpolation)

[/quote]

L  %W
0    50
1    60
2    67  
3    76  
4    85  
5    89
6    92  
7    94  
8    96  
9    97
10  98

Directly from Nate Silver's probabilities of wins in Senate races with one week left in the race; others interpolated

Numbers on states not in the very darkest (saturation 9) shade of red indicate chances of a Biden win; subtract from 100 for the chance of a Trump win.


 

Except that I do not have an extremely-recent number for Arizona,  Iowa, Ohio, ME-02, or NE-02, this is all that you need to know at this point. Trashy, falsehood-laden, garbage polls (you know from whom) do not count.

[pbrower2a]

This is what a 4% lead looks like for Trump. It is Montana.

lead and likelihood of winning:

L  %W
0    50
1    60
2    67  
3    76  
4    85  
5    89
6    92  
7    94  
8    96  
9    97
10  98

Directly from Nate Silver's probabilities of wins in Senate races with one week left in the race; others interpolated

Numbers on states indicate chances of a Biden win; subtract from 100 for the chance of a Trump win.


 

Except that I do not have an extremely-recent number for Arizona,  Iowa, Ohio, ME-02, or NE-02, this is all that you need to know at this point. Trashy, falsehood-laden, garbage polls (you know from whom) do not count.

[pbrower2a]

This is what a 4% lead looks like for Trump. It is Montana.

lead and likelihood of winning:

L  %W
0    50
1    60
2    67   
3    76   
4    85   
5    89
6    92   
7    94   
8    96 
9    97
10  98

Directly from Nate Silver's probabilities of wins in Senate races with one week left in the race; others interpolated

Numbers on states indicate chances of a Biden win; subtract from 100 for the chance of a Trump win.


 

Except that I do not have an extremely-recent number for Arizona,  Iowa, Ohio, ME-02, or NE-02, this is all that you need to know at this point. Trashy, falsehood-laden, garbage polls (you know from whom) do not count.

I missed a day on the Internet yesterday to see the Museum of the Air Force in Dayton, Ohio. If you are within 500 miles it is well worth the trip. It's four hours from where I live and four hours back, so I was too tired to get on the Internet when I got back home.

If there is one political conclusion from what I saw on the drive, it is that rural Ohio is strongly pro-Trump... as if people are compelled to put up "Trump 2020" signs under social pressure. Dayton and Lima are far from sympathetic. People vote; milk cattle don't. 

Now back to polls:

https://twitter.com/gelliottmorris/status/1321437444466724871?s=19
[/quote]

With this number, Donald Trump cannot win either the popular vote or the Electoral College. Cheating will be unlikely because it will stick out and practically beg investigations by the next Attorney-General. I expect President Biden to show all the sympathy to corruption of any kind that Barack Obama did: absolutely none.

I'm not predicting who will get the opportunity to cheat, but Democrats have no need for electoral fraud, and... a side seeing itself losing nationally might as well let honest results prevail, whatever those honest results are.

This may be the biggest matter six days from now.

   Oct. 20-25

MI (789 LV, MoE: 4%)
Biden 51%
Trump 44%

SEN: 52-46 Peters

Trump approval: 46/52 (-6)

WI (809 LV, MoE: 4%)
Biden 57%
Trump 40%

Trump approval: 41/58 (-17)

https://www.langerresearch.com/wp-content/uploads/1216a52020StateBattlegrounds-MIWI.pdf

If anything I would expect these to be inverted. Over ten for Biden in both states is a reasonable assumption.

Virginia, Christopher Newport University

https://twitter.com/13DanKennedy/status/1321388406748925953

Changes with September 9-21

Biden 53% (+5)
Trump 41% (-2)

 I'm not sure that Virginia yet has alligators, but after global warming that Trump denies, the Potomac (as well as the James, and possibly even the Shenandoah) will have them. So will the Schuykill and perhaps even the lower reaches of the Hudson....maybe even the Connecticut. But in the meantime, Chesapeake Bay already has sharks.... Virginia has been the equivalent of shark-bite for the Trump bid for re-election. We can all remember when we thought Virginia was reliably R... Carter lost it in 1976 although winning every former Confederate state, and Bill Clinton never won it. Virginia is now a solid "four-peat" for the Democrats.

Biden 51%
Trump 38%

SEN: 47-43 Gideon (+3 pre-rounding)
CD-02: 46-42 Biden
CD-02 House: 56-31 Golden

https://www.pressherald.com/2020/10/28/gideon-maintains-narrow-3-point-lead-over-collins-in-latest-colby-college-poll/

Maine has been polled rather little... the Senate race and ME-02 might not be called early. About what I expect.

Two polls in Arizona average Biden +2. Arizona polling has been remarkably stable. I think that Democrats located every vote that they could get. It is within the margin of error, but with so many polls outstanding, I can easily see Biden holding this one.


Iowa:

 

https://twitter.com/jacobfhsmith/status/1321196396985360384

Previous: Biden 48 Trump 46 in September.

Over 50, Biden lead at the margin of error. 

Nevada:

https://twitter.com/Nate_Cohn/status/1321135387092389888
[/quote]

My system compels me to reduce Biuden's chance of winning Nevada to 92%. On the other hand, pollsters generally underestimate Democratic support here, and I really expect the same to happen this year. Either way, Trump loses Nevada.

South Carolina:

 https://twitter.com/DataProgress/status/1321239265762828288?s=20
[/quote]

Joe Biden will not be the first President to win all 13 original states since Ronald Reagan in 1984.

[Protect Trans Hoosiers]

I look at the map like this. These states are safe Biden:

226 EVs

I'll now add the states where Biden is averaging 50% or more on 538:

273 EVs

Biden is under 50% in Nevada, but leads by 6.1 points:

279 EVs

Now let's add all the states where Biden has any kind of lead in the polls:

357 EVs

So first, Trump would have to win every state where he's in the lead in the polls, which is a highly unlikely. Trump didn't even accomplish this in 2016 (he lost Nevada). Ohio and Texas are very tight.

Next, he's going to have to win in at least four places where Biden currently leads (NC, FL, and GA are must-haves, and he's going to need at least one more). In 2016, he only won three states led by Clinton.

If he accomplishes this amazing feat, he'll still be short without winning a state where Biden is polling above 50%

Also, 75 million people have already voted.

I struggle to see how Trump wins this.

[pbrower2a]

Monmouth, Georgia. Biden up 5.

lead and likelihood of winning:

L  %W
0    50
1    60
2    67   
3    76   
4    85   
5    89
6    92   
7    94   
8    96 
9    97
10  98

Directly from Nate Silver's probabilities of wins in Senate races with one week left in the race; others interpolated

Numbers on states indicate chances of a Biden win; subtract from 100 for the chance of a Trump win.


 

Except that I do not have an extremely-recent number for Arizona,  Iowa, Ohio, ME-02, or NE-02, this is all that you need to know at this point. Trashy, falsehood-laden, garbage polls (you know from whom) do not count.

I didn't except one for Georgia. I would love to see what Selzer has for Iowa this late.

[pbrower2a]

One update because it is obvious: Trump is up 12% in Kentucky, which is rather weak. But Trump still has an above-98% chance of winning Kentucky with that number. No further updates here as we are fewer than 100 hours away from the opening of the first polling stations on Election Day.

Quinnipiac, Ohio: Biden up 5

No further updates here despite a flood of polls.  Note well: what Trump did not do in the last six months he cannot do in fewer than four days.

L  %W
0    50
1    60
2    67  
3    76  
4    85  
5    89
6    92  
7    94  
8    96  
9    97
10  98

Directly from Nate Silver's probabilities of wins in Senate races with one week left in the race; others interpolated

Numbers on states indicate chances of a Biden win; subtract from 100 for the chance of a Trump win.


 

Except that I do not have an extremely-recent number for Arizona,  Iowa, Ohio, ME-02, or NE-02, this is all that you need to know at this point. Trashy, falsehood-laden, garbage polls (you know from whom) do not count.

I didn't except one for Georgia. I would love to see what Selzer has for Iowa this late.

I am closing this thread.

[pbrower2a]

For some strange reason I rediscovered this post and found it potentially embarrassing. How could I be wrong, whether in accepting the model (I treat Presidential campaigns in states as if they were Senatorial or Gubernatorial races because such is how most candidates run for President -- almost all Presidential nominees since 1944 have been Senators or Governors while or before running for the Presidency) and applying it to a race in which the incumbent President operates a campaign as no Governor or Senator ever would? Or is the statistical inference simply wrong in itself? Or was the late polling wrong?


Here I try to explain myself out of a prediction of a Biden landslide that turned out to be a bare Biden win.


 The numbers, where shown, suggest what the system showed as a chance of a Biden victory. I saw Biden having a 67% chance of winning both Arizona and Florida, separately. The states are very different in their politics, and they split. The chance that Trump would win one of those state4s but not the other was 4/9, which is just over half. Probabilities of independent events are multiplicative; the chance that Biden would win both was only 4/9, which is slightly less than half. I saw both Michigan and Pennsylvania nearly out of reach, which doesn't look like a good prediction. It still ended up right. I would expect that a state with a 99% chance of going against the President to not go against him by a razor-thin margin than showed in historical reality. All three states did go for Biden. 

Late polls suggested that Trump was losing both Iowa and Ohio... based on the polls in question I would have given Trump a 2.25% chance of winning both... and somehow he did. I was surprised to see both Georgia and North Carolina as likely wins for Biden, but Biden won Georgia but not North Carolina. Go figure.  The subtle difference between an 89% chance and an 85% difference was the difference between the two states. Finally, Texas was a genuine toss-up in the system and it could not have been predicted either way. 

....................................................

So what went wrong with my prediction of a Biden landslide? My assumption, along with that of the statistician Nate Silver, that a candidate can do practically nothing to win in the last week that he did not do in the previous six months. Biden was hiding from COVID-19, and Trump was holding rallies in which people got away with not wearing masks. Damn the virus! Full speed ahead! Trump showed physical courage that Biden thought more likely to do harm than good. Democrats did not canvas for votes late in the campaign because such seemed pointless... and too dangerous. 

Trump had some late ads. One in Florida tried to equate Joe Biden and Kamala Harris politically with Fidel Castro and Hugo Chavez, anathemas among the large Cuban-American population of southern Florida. Trump got big gains because Democrats lacked a swift and effective response to such messaging.   

I remember this one in Michigan. With a female voiceover (it is not an exact quote):

Let's look at the Democrats. They'll raise taxes. I can't afford that. Biden will take away the Trump tax cuts. My husband is out of work, and it will be harder for him to get a job. 

Well, that's simple. I'm voting Republican"

In essence, no matter how much one despises Donald Trump and the interests for which he stands, if you know what is best for you you will still vote for him. Life is all about money, and never forget that -- peon! Remember to suffer in This World on behalf of cruel and rapacious plutocrats and executives so that you can deserve pie-in-the-sky-when-you-die.. and give suffer in great cheer in This World to your Divinely-appointed masters because hardship in This World is preferable to ending up with muggers, rapists, Commies, and Nazis in the Next World.  That may not be your religious heritage, but it is what many that I know were brought up with. Many of those still believe it and more have come to believe it. That's the equivalent of the Stockholm syndrome for people who see themselves just one slip-up from hunger and homelessness. 

That ad may have worked. Trump had won over perhaps a couple million votes and had sealed victories for his right-wing Senate allies other than McSalley (R-AZ, and wholly inadequate) and Gardner (R-CO, and in a state spiraling away from the GOP), and the two Senators from Georgia (much too early to decide the run-off elections for their Senate seats). 

Either the system is wrong, the polling is wrong a week before the election, or the 2020 election is not normal. 

This little fellow



makes the 2020 election unique. It is a genocidal killer. 

Oh, yes -- small late changes in the margin of a lead make all the difference in the world in that model. Those are rare in normal elections.