[RD] Super Tuesday

[onejayhawk]

This is the schedule for the next eight days. 1 March is called Super Tuesday because so many of the delegates are awarded on one day.

Tue, Feb 23 Nevada Caucus (R)
Sat, Feb 27 South Carolina (D)
Tue, Mar 1 Alabama
Tue, Mar 1 Alaska Caucus (R)
Tue, Mar 1 American Samoa Caucus (D)
Tue, Mar 1 Arkansas
Tue, Mar 1 Colorado caucus
Tue, Mar 1 Georgia
Tue, Mar 1 Massachusetts
Tue, Mar 1 Minnesota Caucus
Tue, Mar 1 North Dakota Caucus (R)
Tue, Mar 1 Oklahoma
Tue, Mar 1 Tennessee
Tue, Mar 1 Texas
Tue, Mar 1 Vermont
Tue, Mar 1 Virginia

All of the results are proportional or unbound. All the primaries but Oklahoma are open, meaning a voter can choose either ballot. Caucuses are mostly closed. Many of these states have almost no polling. 270toWin has some very useful maps and interactives. I am including a couple.

Democrats:
http://www.270towin.com/2016-democratic-nomination/
The national polls are very close. Sanders won the most recent one. He is coming off a resounding win in New Hampshire and a close loss in Nevada. Clinton is coming off a win in Nevada that some thought she would not get. Her organization has a better reputation. Ground forces may be critical in some of the less contested states, say Arkansas. She is expected to win South Carolina Saturday which would be a big talking point over the weekend.

Republicans:
http://www.270towin.com/2016-republican-nomination/
Donald Trump has a big lead in the national polls. Some think he is working away from his logical base in the big cities. However, his demographics are blue collar. These states are heavily blue collar. He has also done surprisingly well with evangelicals. This is the Bible belt.

Trump's opposition has thinned substantially, down to Cruz, Rubio, Kasich and Carson. Cruz is by far the best organized and his home state of Texas is the jewel of the day (up double digits). Georgia and Virginia are dicier. Trump is up in both, but by much less than the national averages. The other test is the smaller states. Alabama, Alaska, Arkansas, Colorado, Oklahoma, Tennessee, and Wyoming are poorly polled, but add up to more delegates than Georgia and Virginia combined. Cruz has a shot at all of them. Rubio will be hard pressed to win any of them.

We will get a peek at the Republican side with the Nevada caucuses tomorrow. Trump is polling 20 points ahead. He polls surprisingly well with Hispanics, which are a big part of the Nevada electorate. We will see how that holds out.

J

 

[civvver]

huh I've never looked at how the delegates for primaries are done. Why are the dem and gop maps so different numbers wise? And why is NC for example more delegates on the republican map than MI but less on the dem one? Is it super delegates?

 

[Antilogic]

Delegates are allocated by the national parties to each state, usually as a function of the population of the state as well as the number of elected office-holders for that party in each state. So if the state has a Democratic governor, senators, or House reps, it will have more Democratic delegates and vice-versa for the GOP. How those are allocated are different for each party: the GOP allocates all of theirs through the primary and caucus contests, the Democrats have these unelected, unpledged superdelegates hanging around as an intentional artifact from the pre-reform era.

And the parties are under no obligation to have an equal number of delegates, the Democrats have roughly twice the number as the Republicans.



EDIT: There is a particularly excellent blog I follow, Putnam's Frontloading HQ, that will tell you everything you ever wanted to know and then some about delegate assignment, primary and caucus structure, all the backroom dealing before the race started, etc. I highly recommend it.

 

[Joecoolyo]

I read a pretty good dissection of Super Tuesday on the Sanders subreddit (sue me).

Basically, Clinton has a solid shot at taking South Carolina and a little bit more than half of the Super Tuesday states. Her support lies mostly in conservatives and old people, which are plentiful in the South. Bernie will win a little bit less than half, probably not a split (thought it could happen!). His strength is in Colorado, Massachusetts, Vermont, and Minnesota. Oklahoma and Tennessee are seen as tossups, though I'd wager a close race with Clinton eeking out a small win, like Nevada or Iowa. The tossup status is based on the assumption of tied electorate nationally between Clinton and Sanders. If we use the Nevada results as the barometer, then it is slightly less than a tie with Clinton up by 2 points.

However, after Super Tuesday we enter into very favorable Sanders ground. The West, North, and Pacific are all heavily swinging towards Sanders. It'll be very interesting to see where things go! Super Tuesday is not the end all be all, as much as Clintonties would like the conservative south to define the race.

As for the Republicans, everything I've seen points towards a Trumpist sweep. Not much to comment on there.

 

[Lohrenswald]

Alright so what warrants yet another thread on the presidential nominations

 

[Nexlev]

All the primaries but Oklahoma are open, meaning a voter can choose either ballot.

I'm pretty sure Arkansas is a closed primary.

I posted this in the "Republican nomination" thread before seeing this one.
Alabama - open
Alaska - closed caucus
Arkansas - closed
Colorado - closed caucus
Georgia - open
Massachusetts - mixed open/closed
Minnesota - open
*North Dakota - closed caucus
Oklahoma - closed
Tennessee - open
Texas - open
Vermont - open
Virginia - open
*Wyoming - closed caucus
* delegates from Wyoming and North Dakota are officially "unbound"

Bolded are the states that I could see Cruz winning. I think his popularity in Texas will help him some in Oklahoma and Arkansas, though I'm less sure about Arkansas. Alaska and North Dakota are oil states, with a lot of transplanted Texans. Cruz is winning in the only Alaska poll I know of, for what that's worth. I'm not sure about Wyoming, not knowing much about the people there, but it is rural and conservative which helps Cruz. Colorado and Nevada could possibly go to Cruz but I think they'll go to Trump with support from population centers.

 

[CavLancer]

Thanks for the heads up Jay, think I'll avoid the news for a bit until this settles out. Go take a look maybe on 3/3, see if the bullstuff has stopped flying.

 

[Timsup2nothin]

Alright so what warrants yet another thread on the presidential nominations

J needed a new venue for posting "Hillary can't win."

 

[Camikaze]

It's probably worth looking at the 538 forecasts:

For the GOP:
Nevada - Trump 64% chance, Rubio 25%, Cruz 10%
Georgia - Rubio 41%, Trump 39%, Cruz 15% (this is an interesting one, because all polls show Trump up by about 10 points)
Massachusetts - no forecast, but polls have Trump up by 14-34%
Oklahoma - no forecast, polls have Trump up by about 5%
Texas - no forecast, polls have Cruz 5-15% up
Virginia - Rubio 50%, Trump 34%, Cruz 12%, Kasich 3% (again, polls have Trump in the lead, but not the forecast - this seems to be based on one poll, which saw Carson's 5% lead disintegrate and Rubio jump from 11 to 22%)

So the chances of Trump winning all of Nevada, Georgia & Virginia are apparently only 8.5%. The chances of the same for Rubio are 5%. The chances that Trump fail to win a state are 14.5%, whilst they're 22.1% for Rubio. So it looks like that Trump will not run the field on Super Tuesday, with Rubio having a 75%+ chance of actually winning on of those states (in the forecasted states, Cruz has a 32.7% chance of winning something, but he is up in Texas, which is the big prize of the day). Btw, it's possible I suck at maths.

Rubio seems to be heavy beneficiary of 538's 'polls-plus' forecast model, so it'll be interesting to see how that turns out if Trump is defying such models. If you look at the 'polls-only' forecast, Trump-Rubio is 78-9 in Nevada, 49-21 in Georgia, and 48-26 in Virginia.

For the Democrats:
South Carolina - Clinton >99%, Sanders <1%
Arkansas - Clinton 99%, Sanders 1%
Georgia - Clinton >99%, Sanders <1%
Massachusetts - Clinton 58%, Sanders 42% (two recent polls, one has them tied, one has Bernie up by 7%)
Oklahoma - Clinton 78%, Sanders 22% (two recent polls, showing Clinton up by 16% and 2%)
Tennessee - Clinton 99%, Sanders 1%
Texas - Clinton 98%, Sanders 2%
Virginia - Clinton 97%, Sanders 3%

9% chance for Sanders to take both Massachusetts and Oklahoma, 45% chance that he doesn't get either. 92% chance that Clinton wins everything other than Oklahoma and Massachusetts, and 72% chance that she wins everything other than Massachusetts.

 

[Timsup2nothin]

Cam, on the Republican side there should be some pretty wide margins for error. Any polling done before the SC primary and subsequent change in the field is heavily suspect.

 

[Camikaze]

I imagine that's where the uncertainty comes from - the wider field doesn't help the front-runner in that regard either (see Trump's New Hampshire forecast peak of 71%). I wonder how they reach the 'polls-plus' forecast, though. I would think it largely takes into account the expected poll movements after the SC primary.

 

[Joecoolyo]

Alright so what warrants yet another thread on the presidential nominations

Our country is literally the greatest bestest country there is and that warrants constant attention from the whole world.

We elect, you watch.

 

[dutchfire]

@Cam: I believe 538 mainly uses endorsements in its polls+ forecast (also: facebook likes...)

Also note that the probabilities are not independent!

 

[Camikaze]

@Cam: I believe 538 mainly uses endorsements in its polls+ forecast (also: facebook likes...)

Also note that the probabilities are not independent!

Yeah, just noticed they have a 'how this works' page. I'm sceptical of the merit of using endorsements, but that's probably what has given Rubio a huge recent boost that hasn't yet been reflected in their 'polls-only' forecast.

___________________

Re Democrats, it's probably worth looking at this table again:

Spoiler :

An even race has Sanders winning both Massachusetts and Oklahoma (as well as some others which are not forecasted). Sanders has been getting away with underperforming a 50/50 split so far, but he'll need to start making up ground at some stage, and failing to do so until after Super Tuesday leaves him with much more ground to make up.

Our country is literally the greatest bestest country there is and that warrants constant attention from the whole world.

We elect, you watch.

It's genuinely an interesting sport. It helps that the US is important enough that viewers can have some investment in the outcome.

 

[Timsup2nothin]

Yeah, just noticed they have a 'how this works' page. I'm sceptical of the merit of using endorsements, but that's probably what has given Rubio a huge recent boost that hasn't yet been reflected in their 'polls-only' forecast.

Endorsements correlates to turnout. Having Nikki Haley say "I like Rubio" doesn't really mean much. Having the elected governor provide access to lists of people who have been turned out on past election days is useful. Having the elected governor provide access to lists of people who organized "rides to the polls as long as we like your vote" efforts in their neighborhoods is even more useful.

 

[Ajidica]

My completely unscientific gut opinion is that Sanders and Trump will win in Minnesota.

Why Sanders? Minnesota is a good demographic for him. Generally well educated middle class white people with a strong bent towards progressivism. Plus we like outsiders (Jesse Ventura, Al Franken).

Why Trump? Despite nutjobs like Michelle Bachman, Minnesota isn't that big on the teahadist/fundy crowd. The state GOP is in the midst of a hilarious kerfuffle that is seeing them kicked out of their congressional offices which I'm betting limits their ability to coordinate with the GOP establishment. Kasich might get it because of his very good job appearing moderate (we do like out moderate GOP governors who are so white bread they made ketchup look spicy and ethnic) but then again, we are the state that elected Jesse Ventura as governor.

 

[onejayhawk]

I'm pretty sure Arkansas is a closed primary.

Good catch.

It's probably worth looking at the 538 forecasts:

For the GOP:
Nevada - Trump 64% chance, Rubio 25%, Cruz 10%
Georgia - Rubio 41%, Trump 39%, Cruz 15% (this is an interesting one, because all polls show Trump up by about 10 points)
Massachusetts - no forecast, but polls have Trump up by 14-34%
Oklahoma - no forecast, polls have Trump up by about 5%
Texas - no forecast, polls have Cruz 5-15% up
Virginia - Rubio 50%, Trump 34%, Cruz 12%, Kasich 3% (again, polls have Trump in the lead, but not the forecast - this seems to be based on one poll, which saw Carson's 5% lead disintegrate and Rubio jump from 11 to 22%)

So the chances of Trump winning all of Nevada, Georgia & Virginia are apparently only 8.5%. The chances of the same for Rubio are 5%. The chances that Trump fail to win a state are 14.5%, whilst they're 22.1% for Rubio. So it looks like that Trump will not run the field on Super Tuesday, with Rubio having a 75%+ chance of actually winning on of those states (in the forecasted states, Cruz has a 32.7% chance of winning something, but he is up in Texas, which is the big prize of the day). Btw, it's possible I suck at maths.

Rubio seems to be heavy beneficiary of 538's 'polls-plus' forecast model, so it'll be interesting to see how that turns out if Trump is defying such models. If you look at the 'polls-only' forecast, Trump-Rubio is 78-9 in Nevada, 49-21 in Georgia, and 48-26 in Virginia.

For the Democrats:
South Carolina - Clinton >99%, Sanders <1%
Arkansas - Clinton 99%, Sanders 1%
Georgia - Clinton >99%, Sanders <1%
Massachusetts - Clinton 58%, Sanders 42% (two recent polls, one has them tied, one has Bernie up by 7%)
Oklahoma - Clinton 78%, Sanders 22% (two recent polls, showing Clinton up by 16% and 2%)
Tennessee - Clinton 99%, Sanders 1%
Texas - Clinton 98%, Sanders 2%
Virginia - Clinton 97%, Sanders 3%

9% chance for Sanders to take both Massachusetts and Oklahoma, 45% chance that he doesn't get either. 92% chance that Clinton wins everything other than Oklahoma and Massachusetts, and 72% chance that she wins everything other than Massachusetts.

Good stuff. 538 is going out on some limbs, due to scant polling. For example, Texas is 50-1, but a Bernie upset would not shock me. Massachusetts is close. I think Bernie cruises. Regardless, Clinton will have a big day. The question is whether she gets a knockout.

Virginia, Georgia and Texas are the big states. For the GOP, they are also states where Trump can lose. Rubio has to start winning or his run is finished. Getting either Georgia or Virginia is necessary just to stay alive. Cruz could clean up in the smaller states. Or Trump could land the knockout punch.

J

 

[hobbsyoyo]

Our country is literally the greatest bestest country there is and that warrants constant attention from the whole world.

We elect, you watch.

Well, you're not wrong...

 

[innonimatu]

Alright so what warrants yet another thread on the presidential nominations

I must admit that I'm finding that drama highly interesting! Mind you, the US is now socially and politically more similar to Europe than it ever was before, it seems to me. And as it goes now so will Europe soon, I'm willing to bet. The "spirit of the time" is actually changing and it shows on this election.

 

[warpus]

I find the American way of selecting presidential candidates for the 2 big parties so weird. How did this originate? Is this purely an American creation or does it have roots in anything British or European?

And why do both parties seemingly select their candidate in such a similar fashion? Where does the need to do it this way come from?