WTF! 538's Delegate Projection still shows Sanders in the lead?

[bluewater]

Ok, I don't know what to say about 538 and their wildly gyrating Delegate Projections...

When Sanders won Nevada, getting 66% of the delegates, 538 projected him ahead by 700 delegates!

Now Biden had a superb SC primary getting 77% of the delegates and yet 538 STILL has Sanders projected to get more delegates.

Over 400 hundred more in fact!

Sanders 1701
Biden 1292
Bloomberg 568
Warren 196
Buttigieg 159
Klobuchar 62

https://projects.fivethirtyeight.com/2020-primary-forecast/?ex_cid=rrpromo

SO, what can we expect after Super Tuesday, another wild gyration in 538's model?

[jmg257]

1. I think we can we expect what so many have been considering...that going into the convention,

it will be a closer Sanders lead than originally expected (after Nevada), close enough that super-delegates can 'justifiably' give the nomination to Biden.

[caraher]

2. SC is not incredibly delegate-rich

plus I'm sure their previous projections already had Biden with most of those delegates.

What happened yesterday wasn't really anything surprising. Biden did somewhat better than expected but not shockingly so

And yes, I do think you'll see some more wild swings. The model is very "twitchy" because of how fine-grained it tries to be, and the input data are both sparse and volatile. I put very little faith in it at this stage.

[underpants]

6. Yeah the first 4 only have 4% of the total delegates

There's a long way to go.

[Sloumeau]

3. Based on the way 538 does its model,

which includes a number of recent polls, Sanders still is projected to get more delegates. This is based on things like Sanders's current polling in California. However, th polls have changed after each state win in favor of whomever won each state.

So, as new polls come in, if they reflect movement in Biden's direction, 538's projections will move accordingly. Rome wasn't built in a day, and 538's projections will not completely change overnight.

[Codeine]

4. Bernie's poll numbers still indicate

he’s going to have a very good day on Tuesday, so it’s still reasonable to assume he’s the frontrunner at the moment.

[underpants]

5. The numbers have changed since your post

Bernie 1635
Joe 1352

Not sure how it works but I'd think that's Pete's projected 158 will go to Bernie and Warren and Amy's combined 260 will go to Joe once those three drop out.

[squirecam]

7. No one getting a majority

Is now the clear favorite in the forecast.

[drray23]

8. I don't think Pete's would go to Bernie.

He is closer to being a moderate than espousing Sander's views.

[underpants]

9. Okay.

I was basing that on the age of his supporters or at least my take on them as a group.

[PTWB]

10. What were you basing their age on?

[underpants]

13. It seems to me that Pete has a lot of support with the younger crowd

That's just my perception of things.

[PTWB]

14. Okay, so based on perception. I didn't know if there was data available.

[highplainsdem]

11. Just before you posted your OP, 538 had changed that to Sanders 1645 Biden 1352.

Cutting Sanders's delegate lead by more than 100.

Yet for some reason you used the old numbers, and still haven't changed them...

[thesquanderer]

12. What I don't get is why you expected a bigger change.

Even before the South Carolina primary, Biden was projected to win it. The forecasting was already assuming he would almost certainly win SC. That's what happened. He won it by a large margin, so that skewed the odds a bit more in his favor, but fundamentally, there wasn't a huge change in the variables.

SO, what can we expect after Super Tuesday, another wild gyration in 538's model?

It seems to me that you're complaining that they're not gyrating wildly enough. You expected one win--which was expected to happen anyway--to move things a lot more than they did. THAT would have been a wild gyration.