The question I have is, what happens when people take their ball and go home? Are they just ignored completely (that is, the remaining votes for viable candidates are normalized to add up to 100%), or does the precinct as a whole simply get less weight if more people walk out? If it’s the former, then does that mean that the candidates get phantom votes to represent the people who left? If it’s the latter, then doesn’t that mean that you should walk out rather than support a rival, even if you prefer the rival to one of the alternatives? In other words, wouldn’t Warren’s supporters prefer to walk away in precincts where she can’t get above 14% than to throw support to Bernie and let him pull away state-wide?
I believe caucusers who choose to just leave after the initial count are counted as uncommitted delegates.
Since it’s unlikely for “uncommitted” to win anything, it’s more valuable for them to switch to another candidate in many cases, if they have a strong enough dislike for another specific candidate.
One of the interesting things about Nate Silver’s influence on political punditry is that the focus on probabilistic framing is both more accurate and more confusing. You could easily imagine a scenario where Iowa breaks down like this (after realignment):
Sanders 30%
Biden 28%
Buttigieg 21%
Warren 16%
Klobuchar 5%
Then, if you looked at the 538 forecast you might say, “Wow, Mayor Pete really overperformed, and Sanders and Biden did poorly.” Except that the two things are measuring very different variables.
I think they would be better off translating their probabilities into expected-value in terms of the measures the networks will actually report (and I think the NYT Upshot folks should do the same with the damn needles). They could report the probabilities elsewhere, but the topline should be a prediction, with error bars, of where they think the reported outcome will land. It would have saved a lot of headache in 2016 if instead of “90% chance of victory” the needle had been saying “265-325 expected EV”
Yep, it’ll be a huge test of Biden’s South Carolina bulwark. And it’s possible that could completely collapse because Biden’s campaign has been so poorly managed thus far.
A recent poll (usual caveat) had Steyer (!) and Sanders significantly cutting into Biden’s lead in SC.
Reading rumors that some reporters have seen the cancelled poll and it looked similar to the one released in early January (which is not good news for Biden.)
Wishcasting FTW!
I’m still hoping for a miracle win by Warren. She has a 10% chance of winning! That’s not zero!
I have it on good authority that the cancelled poll was actually buried because it had Sanders getting 80% of the vote. This Buttigieg being left off thing is just a smoke screen.
Deep state strikes again!
NYT returns page is up!
Final prediction by me: Sanders by 7 over…someone. :)
Nah, I don’t care if Biden wins. He probably has the best chance at winning FL, and if Dem’s win FL it’s game over. But Biden’s team are out there lowering expectations (which I realize is what they’re supposed to do.) But thanks for assuming!
https://twitter.com/ElectProject/status/1224451865632366597?s=20
What caused them to change the realignment rules this year?
So if you are with someone, and they are getting more than 15%, then you can’t leave them? That’s not how it used to work.
It seems like they are making the classic mistake of changing the things that people complained about last time to try and eliminate complaints because that they don’t realize that some people just want to weaponize complaints.
Caucusing is some bullshit. Every state needs to just use mail-in ballots as a standard option and we can leave this ridiculousness in the past!
And here you go, DMR poll fans.
YIKES BIDEN!