Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Electoral College Decision Tree

observablehq.com

Post reply on HN
Print view View original

201–210 of 373 posts

Re: Electoral College Decision Tree

#201

Post by remarkEon »

rayiner wrote: post #155
alistairSH wrote: post #85

Earlier quoted context omitted.

Not quite. The EC (along with the 3/5 compromise) was designed to protect the interests of slave-holding states, while also avoiding direct election of the president (ie, the people were too stupid to elect directly). And now, instead of tyranny of the majority, we have tyranny of the minority. Due to capping the number of members of the House, those small-population states have disparate impact on federal issues.

This is false for several different reasons. There were two initial proposals for Congress: the New Jersey Plan and the Virginia Plan. (Virginia, of course, being the leading slave-owning state.) Under both plans, the President would have been elected by Congress. There were concerns that having Congress elect the President would jeopardize separation of powers. So the Electoral College was created, with one elector…

I remember reading about all of this in AP US History back in high school, and it's going to forever confound me that we seem to have uninstalled this history from our collective memory.

Re: Electoral College Decision Tree

#202

Post by dragonwriter »

therealjumbo wrote: post #153
AnthonyMouse wrote: post #111

Earlier quoted context omitted.

The biggest problem with the National Popular Vote Interstate Compact is that in order to get a majority of the electoral college delegates into it, you would need states to sign on who benefit from the existing arrangement. You can't get the swing states because the existing system makes them the focus of national politics and you can't get always-red states because the party that controls them benefits from the ext…

The biggest problem with it is that it's unconstitutional. States are not allowed to make agreements with each other parallel to the system. EDIT: if you can get that many states to agree to it, then you could have just done a constitutional amendment. since all you need for that is for the states to agree, and the senate to agree. Senators come from the states. So... if you can't get it passed that way, it doesn't p…

> The biggest problem with it is that it's unconstitutional.

There is considerable debate about that.

> States are not allowed to make agreements with each other parallel to the system.

Yes, they are. In certain circumstances, such agreements require affirmative Congressional consent, though its questionable if the National Popular Vote "Interstate Compact" would meet the standards for requiring such consent, whether it would require only negative consent, or whether it would be outside of Congress' authority because of the assignment of the power of choosing electors.

> if you can get that many states to agree to it, then you could have just done a constitutional amendment

No, that's not at all necessarily even approximately true. For it to be effective, you need jurisdictions with a bare majority of electoral votes to implement it. For a constitutional amendment, you need 3/4 of states.

The minimum number of states to hit a majority of EVs is 16, as opposed to 38 for a constitutional amendment.

> since all you need for that is for the states to agree, and the senate to agree.

You don't need the Senate to agree to a Constitutional Amendment. You need 3/4 of state legislatures to agree. You can propose a constitutional amendment with 2/3 of both houses of Congress, or call a Constitutional Convention on the application of 2/3 of states.

Both proposal and ratification (presuming senators vote in line with their states, regardless of the House) have a higher minimum threshold than would be otherwise required for states representing a majority of EVs to assign electors in line with the national popular vote and determine Presidential elections that way.

Re: Electoral College Decision Tree

#203

Post by russellbeattie »

Here's a fact: Regardless of who wins the electoral college, Biden is going to receive at least 5 to 6 million more votes than Trump. This isn't a biased prediction, it's based on national polling which has proven itself to be accurate to within a couple points. This makes sense, given the law of large numbers.

Biden is currently up +10, though this will drop before election day. Hillary ended up winning with 48.2% vs 46.1% - +2.1 points equaling 2.8 million votes more than Trump. Biden will most likely double that number, and yet still easily lose the election.

The system we have is amazingly broken.

Re: Electoral College Decision Tree

#204

Post by hyperion2010 »

bryanlarsen wrote: post #87
dragontamer wrote: post #81

Earlier quoted context omitted.

> election night, as state results come in. Not this year. With mail-in ballots, the ballots won't necessarily be finalized until a week or more after the election. Its necessary due to COVID19. But it will be until multiple days after the election before we know the true winner. A large number of states will certainly be called on election night (because many states, such as California, are far from competitive). Bu…

Not necessarily. There are several paths where a Biden win could be called early on election night. For example, Texas has a large fraction of early voting and a small fraction of mail ballots. So the chance that Texas can be called early is quite high. And if Texas gets called for Biden, then the networks will call the entire country for Biden.

For context, if Texas goes for Biden, then the 538 model predicts some amusing things. 1. The next most likely state that Biden could win is South Carolina at 25%. 2. The next most likely state that Trump could win is Ohio at 28%. 3. Biden would have the same chance of winning Mississippi as Trump would have of winning Iowa, 21%.

Calibrate your expectations accordingly.

Re: Electoral College Decision Tree

#205

Post by erichocean »

standardUser wrote: post #4
atonse wrote: post #2

Nerd Me: This is a really nice use of this visualization. Human Me: I saw all this stuff in 2016 and Trump still beat the overwhelming odds. (Yes 16% chance of winning is still a chance, I get it... there's just so much deja vu with all these visualizations and articles about his campaign imploding).

He "beat the odds" because most models were incorrect. They have been corrected. Are the models wrong again? Maybe, but it would need to be in a totally different way, and by a much larger margin than in 2016.

> They have been corrected.

Hispanic vote numbers are wrong this cycle (too high of support for Biden in the models). Might not matter, I'm still expecting a Biden victory.

Re: Electoral College Decision Tree

#206

Post by SllX »

jedberg wrote: post #110
SllX wrote: post #80

Earlier quoted context omitted.

Only way I’m on board with DC representation is if we give it to Maryland. Maryland would in effect gain a House seat during the next redistricting after. Easier solution? Kill the cap on the House seats. Doesn’t require a constitutional amendment, doesn’t make a State out of DC (although I’m still onboard with returning it to Maryland, there is precedent for this as DC used to have part of Virginia), and it evens ou…

If you combine DC and Maryland, it makes Maryland go from the 19th largest state to the 10th. DC is bigger than 19 states, why not let it have its own representation? Edit: I mixed up DC and Puerto Rico. Forget what I said.

No worries about the mixup, but I’ll at least answer the core of your question anyway: DC is bigger than a handful of States by population, but it is still just a carve out from Maryland. You could make a case for any of the top 20 cities by population to be their own State, but you need a limiting principle.

DC does not have proper representation in Congress, and I think that’s a shame and worth addressing, but there is also a case to be made for Congress having some territory that it directly governs. Only problem is DC became cool and a place where a lot of people live. My compromise is to leave the city itself to be administered by Congress and the government it setup there, but to have Maryland administer elections and put DC back under Maryland law for civil, criminal and tax purposes.

Re: Electoral College Decision Tree

#207

Post by Balgair »

speeder wrote: post #197
CyanLite2 wrote: post #157

Earlier quoted context omitted.

Florida is one of the few states that are reporting mail-in ballots the night of the election. They've already started counting them and tabulating the results. After polls close, they press a button and it spits out the final results. With early voting getting nearly 60% of all registered voters in FL, I suspect we'll get 99% of precincts reporting FL numbers within 30-45 minutes after the polls close. Thus, I perso…

The signature verification part was dropped by court. And to be honest I don't understand why.

Unanimously dropped by the State Supreme Court. Not one justice voted to require signatures.

Re: Electoral College Decision Tree

#208

Post by Balgair »

67868018 wrote: post #152
elicash wrote: post #150

Earlier quoted context omitted.

All but 8 states can start processing ballots before election day. Unfortunately, 3 of those 8 are MI, WI, PA. So it may take longer for those. Florida will likely be an early sign, since those mail ballots get processed and counted early. IF (and this is a big if) Biden carries Florida on election night, then it's actually possible we know who wins that night.

Madison Wisconsin pre-sorts mail in ballots by voting ward and has them taken to the polling places on Election Day morning and feeds them through the tabulator after someone cosplays as the voter to ensure the voter can't cast two ballots. I assume the same process is done for all the major metro areas. They'll be counted on time. (They verify a few absentee ballot envelopes for required signatures etc and then open…

> after someone cosplays as the voter to ensure the voter can't cast two ballots.

I'm sorry, what? This needs more explanation.

Re: Electoral College Decision Tree

#209

Post by gpm »

dragontamer wrote: post #90
bryanlarsen wrote: post #87

Earlier quoted context omitted.

Not necessarily. There are several paths where a Biden win could be called early on election night. For example, Texas has a large fraction of early voting and a small fraction of mail ballots. So the chance that Texas can be called early is quite high. And if Texas gets called for Biden, then the networks will call the entire country for Biden.

You're right that if Texas goes blue, I think its safe to say that Biden won the election. But I also think that Texas going for Biden is very small: Trump is still leading by 2% or so by official polls. Its important to push Texas a little bit since there's a chance for victory. In the most likely event: Texas goes red (as it has gone for the past several decades), and we're left wondering about Florida, Ohio, and P…

I'm not sure what you mean by "official" polls, but in 538's polling average Trump is only leading Texas by 0.5%.

I also tend to like 538's model. 538's model gives Trump an average vote share of 2.2% more than biden (more than the polling average because the model thinks the demographics and fundamentals favor trump) but still gives Biden a 35% chance of winning. Biden winning is not very likely, but it's also not a not very small chance.

https://projects.fivethirtyeight.com/2020-election-forecast/...

Re: Electoral College Decision Tree

#210

Post by Balgair »

bryanlarsen wrote: post #24

Another similar tool, with very different visualization. https://projects.fivethirtyeight.com/trump-biden-election-ma...

That tool has been a total productivity drain on me since it was launched.

When I play around with it, the real 'turning point' states tend to be PA and FL. You can fill in the likely EC votes and the chances don't really change all that much. I mean, CA going Blue is not a surprise.

But when you fill in PA or FL for Donny, then the whole race changes. If FL or PA go Red, then Donny has ~1/3 chance, if one goes Blue, then it's 1/100 [0].

PA is unlikely to be called on election night. But FL is very likely to be called. So that is the one to watch. Just like every other election, really.

It's a great tool and I'd totally suggest spending some time with it. Kudos to the team that built it!

[0] Note that these models are updated all the time by 538 as new polls roll in.

Post reply on HN