Weird 538 maps

[Roronoa D. Law]

Biden 379 - Trump 159

Craziest parts are Louisiana, South Carolina, and Mississippi going to Biden while Arizona stays in Trump's column.

That map is plausible if there is a super high black turnout.

[Amenhotep Bakari-Sellers]

Map going the way of Senate race plus 278 of course, without OH, FL or TX
 Twitter had Biden leading in 278 plus AZ and NC

[Ferguson97]

Map going the way of Senate race plus 278 of course

Honestly that's not even the craziest map I've seen. Florida is gonna Florida, and with very few split-ticket voters in Kansas with a Boiler surge... it's unlikely, but not *ridiculous*.

[Amenhotep Bakari-Sellers]

Map going the way of Senate race plus 278 of course

Honestly that's not even the craziest map I've seen. Florida is gonna Florida, and with very few split-ticket voters in Kansas with a Boiler surge... it's unlikely, but not *ridiculous*.

Not really this is a Biden 334 map, this is a Biden plus 6 map

[YL]

In which Biden somehow loses Oregon while winning in a huge landslide:

[𝕭𝖆𝖕𝖙𝖎𝖘𝖙𝖆 𝕸𝖎𝖓𝖔𝖑𝖆]

Today I found this nugget:

The really funny thing is that this is a 278-260 Biden win.

THIS IS THE REAL 278 FREIWAL GUYS

[Amenhotep Bakari-Sellers]

The one thing that Trump changed was the EC map, dont expect a typical EC map anymore, that's why we can get another not so typical map again, but this time, with a Biden win

Dems not necessarily gonna win with OH, FL or TX

[DabbingSanta]

proof that JHK beats 538 any day

[Progressive Pessimist]

In which Biden somehow loses Oregon while winning in a huge landslide:

I guess Oregon appreciated their federal occupation more than we thought.

[The Mikado]

I love how somehow this is a Biden map.

I know 538 checks state correlations etc and I just cannot imagine a universe in which Trump loses Pennsylvania and wins New Jersey even in the most wild of hypotheticals.

[JRP1994]

I love how somehow this is a Biden map.

I know 538 checks state correlations etc and I just cannot imagine a universe in which Trump loses Pennsylvania and wins New Jersey even in the most wild of hypotheticals.

#AmishForJoe

[💥💥 brandon bro (he/him/his)]

The issue is that they're allowing a lot of subgroup variance. If you think about it you can tell how many of the weird maps happen.

It can be, say, "extremely good Biden black voter scenario" (black turnout 80% or something) combined with "extremely good Trump Hispanic voter scenario" (Trump wins Hispanics!) and then suddenly Biden is winning places in the Deep South with all his black voters while Trump wins New Mexico, etc.

Right, but my point is, in an era as polarized as this one it's pretty unlikely to simultaneously see contrasting trends like that. These trends are generally going to go in the same direction. The subgroup variance needs much more covariance. You'd really only build a model which does this frequently if 2016 realignment takes melted your brain.

One of the specifically weird things is that they allow Corona to be a group thing for states. So when you combine a bunch of different tail end scenarios for different groupings, including some that group states along non-political lines (like Corona) you get these really wild outcomes.

This is interesting - didn't know about this.

Is that a Jorgensen win in WY or did Wyoming literally just disappear?

It's a weird aesthetically non-intuitive way of highlighting only the states of the winner - Trump wins WY here.

[Amanda Huggenkiss]

I've received this one today:

It's a 344 EV Biden win.

[Gustaf]

The issue is that they're allowing a lot of subgroup variance. If you think about it you can tell how many of the weird maps happen.

It can be, say, "extremely good Biden black voter scenario" (black turnout 80% or something) combined with "extremely good Trump Hispanic voter scenario" (Trump wins Hispanics!) and then suddenly Biden is winning places in the Deep South with all his black voters while Trump wins New Mexico, etc.

Right, but my point is, in an era as polarized as this one it's pretty unlikely to simultaneously see contrasting trends like that. These trends are generally going to go in the same direction. The subgroup variance needs much more covariance. You'd really only build a model which does this frequently if 2016 realignment takes melted your brain.

One of the specifically weird things is that they allow Corona to be a group thing for states. So when you combine a bunch of different tail end scenarios for different groupings, including some that group states along non-political lines (like Corona) you get these really wild outcomes.

This is interesting - didn't know about this.

Is that a Jorgensen win in WY or did Wyoming literally just disappear?

It's a weird aesthetically non-intuitive way of highlighting only the states of the winner - Trump wins WY here.

I'm not really disagreeing with the bolded part, just noting that this is the underlying thing in the model which causes these weird maps.

[JRP1994]

[emailking]

The 2008 map definitely falls in this category, and that happened, so...

[The Mikado]

Why does their model think there's any chance at all of Hawaii going Republican?

[ultraviolet]

The 2008 map definitely falls in this category, and that happened, so...

The only thing weird about that map is Indiana so I don’t think it would fit

[emailking]

The 2008 map definitely falls in this category, and that happened, so...

The only thing weird about that map is Indiana so I don’t think it would fit

NE-2 while losing Missouri

[VAR]

GREAT PLAINS JOE

[Penn_Quaker_Girl]

I've received this one today:

It's a 344 EV Biden win.

That Four Corners Pinwheel is beautiful!

[The Mikado]

PREFER THE SUNBELT STACK

[Granite City]

https://twitter.com/NateSilver538/status/1300825856072454145?s=20

Not sure if this can justify Biden winning everything but NJ but it's relevant.

[VAR]

and



Titanium Tilt R Florida

[Ferguson97]