538 Democratic primary model is up

[Starry Eyed Jagaloon]

Also, 538 has Biden winning TX, NC, VA, TN, AL, OK, and AR on super tuesday. Klob is up in MN and Sanders takes CA, MA, CO, UT, ME, VT, and American Samoa. So much for a Sanders landslide.

Those results (aside from TX and maybe NC) are pretty consistent with the CW prior to South Carolina - most people were not expecting a Sanders landslide across the board.

A lot (most?) people were predicting a comfortable Sanders win in TX and narrow ones in VA and NC.

[Fmr. Gov. NickG]

The 538 model has been updated with the SC results, and Biden’s chances of winning a majority went down.

Because nobody is now at 60%. If Tuesday goes as expected, we should probably be modeling types of pluralities rather than majorities.

It is certainly my impression that the SC result made a contested convention less likely (relative to what the model was predicting, which was a much smaller Biden win in SC).  Why is the model interpreting this so differently?  And how could Biden winning a much bigger victory than predicted reduce his probability of winning a majority, regardless of the contested convention probability?

[The Ex-Factor]

Current 538 projections:
Sanders - 1672 (42%)
Biden - 1287 (32%)
Bloomberg - 590 (15%)
Warren - 203 (5%)
Buttigieg - 162 (4%)
Klobuchar - 63 (1.6%)

[Interlocutor is just not there yet]

Also, 538 has Biden winning TX, NC, VA, TN, AL, OK, and AR on super tuesday. Klob is up in MN and Sanders takes CA, MA, CO, UT, ME, VT, and American Samoa. So much for a Sanders landslide.

Outside of TX and NC, these predictions seem to have been largely agreed upon here. TX & VA could go either way though. It's about a 1-2 difference in points & delegates in both states.

But don't let me stop your gloating.

[Vaccinated Russian Bear]

Majority:
No one 41%
Sanders 36%
Biden 12%
Bloomberg 9%
Buttigieg 1%
Warren 1%

Plurality:
Sanders 54%
Biden 22%
Bloomberg 18%
Buttigieg 3%
Warren 3%

Majority:
Sanders 46% (+10 since FEB 19)
No one 40% (-1)
Biden 9% (-3)
Bloomberg 4% (-5)
Warren 0.7% (-0.3)
Buttigieg 0.3% (-0.7)

Plurality:
Sanders 69% (+15)
Biden 17% (-5)
Bloomberg 11% (-7)
Warren 2% (-1)
Buttigieg 1% (-2)


Obviously NV (and polls from SC?) was a really good for Bernie, 538 model thinks.

Majority:
No one 44% (+4 since FEB 23)
Sanders 43% (-3)
Biden 11% (+2)
Bloomberg 2% (-2)
Warren 0.3% (-0.4)
Buttigieg 0.1% (-0.2)

Plurality:
Sanders 69% (-)
Biden 23% (+5)
Bloomberg 7% (-4)
Warren 1% (-1)
Buttigieg 0.3% (-0.7)


So it is basically 70% Bernie vs Biden 30% rn according to the model.

Majority:
No one 60% (+14 since FEB 27)
Sanders 28% (-15)
Biden 11% (-)
Bloomberg 0.6% (-1.4)
Buttigieg 0.1% (-)
Warren 0% (-0.3)

Plurality:
Sanders 64% (-5)
Biden 32% (+9)
Bloomberg 4% (-3)
Warren 0.1% (-0.9)
Buttigieg 0.1% (-0.2)


Majority - "No One" is up taking 99% from Bernie's chances, while Biden is stable.
Plurality - Biden up, taking 50/50% from Bernie and the field.


If not Bloomberg drops out now or/and Biden gets a huuge bump, "No one" likely becomes unstoppable.

[Vaccinated Russian Bear]

Majority:
No one 66% (+14 since FEB 27)
Sanders 28% (-15)
Biden 11% (-)
Bloomberg 0.6% (-1.4)
Buttigieg 0.1% (-)
Warren 0% (-0.3)

Plurality:
Sanders 64% (-5)
Biden 32% (+9)
Bloomberg 4% (-3)
Warren 0.1% (-0.9)
Buttigieg 0.1% (-0.2)

Majority:
No one 60% (+14 since Pete dropped out March 1)
Sanders 22% (-6)
Biden 112% (+1)
Bloomberg 0.2% (-0.4)
Warren 0% (-)
Buttigieg 0.1% (-)

Plurality:
Sanders 60% (-4)
Biden 37% (+5)
Bloomberg 3% (-1)
Warren 0.1% (-)
Buttigieg 0.1% (-0.2)


Delegates:
Sanders 1626
Biden 1400
Bloomberg 624
Warren 236
Klobuchar 66


We'll likely get a much clearer picture of the race after ST...

[StateBoiler]

Current 538 projections:
Sanders - 1672 (42%)
Biden - 1287 (32%)
Bloomberg - 590 (15%)
Warren - 203 (5%)
Buttigieg - 162 (4%)
Klobuchar - 63 (1.6%)

Do they have an "Other" category that's not being considered? I have to think Gabbard would at least get delegates in Hawaii.

[bilaps]

New model has no majority at 60%, but Biden narrowly trailing Sanders being pretty close in delegates. When Bloomberg drops, this probably means Biden and Sanders finish with a very similar number of delegates. As such, I'd give Sanders a 1/3 shot of getting the nomination outright, Biden a 1/3 shot of getting the nomination outright, and Biden a 1/3 shot of getting the nomination on the second ballot. As such, Biden is the overwhelming favorite to be the nominee, has been the frontrunner from the start, and has honestly run a solid campaign. It's time to unite behind our likely nominee.

Delusional

[Sorenroy]

The model has been updated post Klobuchar dropout.

Majority:
No one — 69%
Sanders — 17%
Biden — 14%
Bloomberg — 0.1%

Plurality:
Sanders — 54%
Biden — 45%
Bloomberg — 2%
Warren — 0.1%

Klobuchar is still the second most likely candidate to win Minnesota. She has a 23% chance to win outright and is projected to get 16% of the vote and 12.2 pledged delegates.

[Mr. Morden]

The model now gives Biden a slightly better chance at getting a majority of delegates than Sanders has, while Sanders is given a slightly better chance at a plurality.

Mean number of delegates won across all runs of the model for...
Sanders: 1542
Biden: 1520

[Bidenworth2020]

Biden is now more likely to get a plurality of delegates, according to the model--

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

[Mr. Morden]

The model also predicts that Gabbard will get all of 1 delegate over the course of the primaries. 

[ElectionsGuy]

I want to remind people that this model is way too reliant on polls. It has a close race in Arkansas (because of one junk poll from a while ago) but Virginia as a 16 point blow out because of the poll that just got released tonight. It has Sanders behind Bloomberg in Oklahoma only because of that crappy Sooner poll, and Biden way down at 14% in 4th place in Utah because of that poll that had him in single digits there. At some point, you have to ignore the polls and go based on the fundamentals of each state. A lot of these projections are going to be terribly wrong tomorrow.

[ElectionsGuy]

I want to remind people that this model is way too reliant on polls. It has a close race in Arkansas (because of one junk poll from a while ago) but Virginia as a 16 point blow out because of the poll that just got released tonight. It has Sanders behind Bloomberg in Oklahoma only because of that crappy Sooner poll, and Biden way down at 14% in 4th place in Utah because of that poll that had him in single digits there. At some point, you have to ignore the polls and go based on the fundamentals of each state. A lot of these projections are going to be terribly wrong tomorrow.

Now Utah radically changes and goes to 29%-25% Sanders over Biden.

[Vaccinated Russian Bear]

Majority:
No one 60% (+14 since Pete dropped out March 1)
Sanders 22% (-6)
Biden 12% (+1)
Bloomberg 0.2% (-0.4)
Warren 0% (-)
Buttigieg 0.1% (-)

Plurality:
Sanders 60% (-4)
Biden 37% (+5)
Bloomberg 3% (-1)
Warren 0.1% (-)
Buttigieg 0.1% (-0.2)


Delegates:
Sanders 1626
Biden 1400
Bloomberg 624
Warren 236
Klobuchar 66

Last before ST was:

Majority:
No one 61% (+1 March 2)
Biden 31% (+19)
Sanders 8% (-14)
Bloomberg 0.1% (-0.1)
Warren 0% (-)

Plurality:
Biden 65% (+28)
Sanders 34% (-26)
Bloomberg 0.9% (-2.1)
Warren 0.1% (-)


Delegates:
Biden 1738 (+338)
Sanders 1363 (-263)
Bloomberg 555 (-69)
Warren 283 (+47)

[°Leprechaun]

The forecast is frozen. I eagerly await the day that it thaws.

[Gass3268]

So, they unfroze the model:

https://twitter.com/NateSilver538/status/1235954698596872192?s=20

[GeorgiaModerate]

After a slew of polls that were good for Biden, he's up to 96% chance of a majority.  "No one" is at 4% and Sanders at 0.6%.  I suspect this thread is nearing the end of its useful life.

[Sorenroy]

It's interesting, the fact that 538's model freezes on the morning of each election day means that we get a map that's not dissimilar to predictions here on the Atlas (in the sense that they have previous states baked in). Anyway, here's an update of what the model shows as of their pre-March 10th results freeze. Shown are the odds of the leading candidate winning, not the percentage they are expected to get.