There is a non-zero chance that Hillary Clinton will have a bad day tomorrow.
My model is estimating two Sanders wins on Tuesday, in Missouri and Illinois. However, Illinois and Ohio are both effectively coin flips with such thin margins between victory and defeat (if you recall, I put Bernie at 53.48% in Michigan and he won by less than 1%, though my model should be more accurate now). It is also estimating two wide victories for Hillary in North Carolina and Florida, which is and has been expected. Here are tomorrow’s projections:
Only one Bernie win in Missouri will not likely lead to any permanent change in the perception of Hillary being the candidate that is destined to win the nomination. Two upsets will likely change the narrative of the presidential race, and bolster Bernie’s image as a threat to the prospect of Hillary being the Democratic nominee. Three upsets tomorrow will likely transform Bernie from “challenger” status to “probable nominee”status, and I say this because early numbers indicate to me that Bernie will win (at least) the next eight states in a row, all the way until April 19th. If Sanders wins three states tomorrow, this means that in mid-April he will be able to say that he has won eleven of the last thirteen state primaries. That’s some serious momentum.
I’ve also been putting together a GOP model over the past week. Though the model seems to fit previous elections extremely well, the GOP elections are just far too volatile for me to have much confidence in the numbers. Regardless, it is estimating at least two upsets tomorrow, in Florida and North Carolina. If it turns out to be acceptably accurate, I will begin posting projections for the GOP as well.
-Tyler
Tyler- I am not trying to be a conspiracy theorist, but combining your predictions, normal poll predictions, and exit polls, do you think there is any evidence of foul play in, say, Ohio?
It was something like 10 points off of the exit polls.
Just asking for an informed opinion rather than a strictly BernieorDie person.
Link to exit polls:
LikeLike
Tyler, I noticed that Bernie’s results from two states (FL and MO) fell outside the model’s 99% confidence interval, and another (OH) outside its 95%. https://pbs.twimg.com/media/Cdj8Qu0W8AATe9d.jpg
What’s more, they all fell outside those intervals in the same direction. Does this cause you to rethink the model’s inputs at all?
LikeLike
Yes it does. Still trying to make sense of why everything went the other way.
LikeLike
Clinton gets ~80% of early voting/absentee ballots and outperforms in precincts with electronic voting machines.
LikeLike
Dude, your models need some serious reworking.
LikeLike
I agree. Ready to throw all of it away and start over.
LikeLike
No offense Tyler but I think you should get a job that doenst involve projecting primaries
LikeLike
Why?
He was right so far, Bernie won Missouri and lost OH, FL and NC
IL is close and could still go to Bernie
LikeLike
Bernie won Missouri? I don’t know what results you are looking at, but I wish I was looking at them, too. The NYT show him down by 1 point in MO, and losing IL.
LikeLike
Correction, looks like Hillary pulled ahead of Bernie late last night in Missouri.
So looks like Tyler was about 2-6% too much Bernie leaning in his projections, except NC, where he was nearly spot on.
LikeLike
@Tony
When I posted about Missouri it was around 11PM, and at the time, Bernie had a sizable lead in Missouri
LikeLike
So far his only big errors were margins with Bernie’s actual results being a bit lower than Tyler’s predictions
LikeLike
And not winning any states. That’s kind of a big one.
LikeLike
2-6% off it looks like, too much of a Bernie lean
LikeLike
@Tony
actually when I posted, he was leading in Missouri, he fell behind later in the night, and Illinois had not yet been called
LikeLike
Wow, you are a massive troll.
LikeLike
Turnout seems pretty high everywhere.
Have you got enough data to predict the final turnout and project the results according to your model?
LikeLike
What does non-zero chance on the first sentence mean? Lol I feel so dumb but I googled it with no help
LikeLike
He just means that there isn’t 0% chance Hillary has a bad day, or in other words, there IS a chance (a % chance greater than 0) that she will have a bad day.
LikeLike
Got it! Here’s to Hillary having a bad day cheers!
LikeLike
Tyler was wrong
LikeLike
Not only was I wrong, it’s looking like I was very wrong. We will see what the rest of tonight looks like when all the results come in.
LikeLike
Non zero means a greater than 0% chance. In other words, if Hillary had a 99.9999% chance of winning every state, Bernie would still have a non zero chance.
LikeLike
Three upsets tomorrow will likely transform Bernie from “challenger” status to “probable nominee”status
Is winning three states going to somehow change the delegate math that heavily favours Clinton?
LikeLike
“Three upsets tomorrow will likely transform Bernie from “challenger” status to “probable nominee”status, and I say this because early numbers indicate to me that Bernie will win (at least) the next eight states in a row, all the way until April 19th. If Sanders wins three states tomorrow, this means that in mid-April he will be able to say that he has won eleven of the last thirteen state primaries. That’s some serious momentum.”
The explanation is in the article. Although these wins would not make up much, if any, of the delegate gap, the resulting momentum would be huge, and that momentum could lead to the gap being closed in later states.
LikeLiked by 1 person
Hi Tyler,
May I ask to what degree does your model take into account the GOTV movement in each state? Given that that movement seemed to play such a significant role in Michigan, I can only assume that it plays a significant role in your calculations. Do you anticipate that various of the 5 states tomorrow will have drastically higher or lower GOTV impact (particularly by Sanders supporters)?
Lastly, can you pretty please share your estimation of whether Kasich or Trump will win Ohio? I realize this means going out on a limb a bit, as you mentioned your models are still being tested for the GOP side, but it would mean a great deal to the supporters of reasonable, decent political campaigning to feel a glimmer of hope!
Kindest regards, and kudos on all your great work.
Colby
LikeLike
I have Kasich in Ohio. Who knows though.
As for GOTV, it is only taken into account indirectly in the candidate search interest data.
LikeLike
Any chance of you doing a post explaining your math in detail? 🙂
LikeLike
I doubt he’d reveal his secret sauce in detail, if this model pans out going forward, it could be a pathway to fame and wealth
LikeLike
So essentially in one day, it’s moved:
Missouri, +5.0 to Sanders
Ohio, +0.7
Illinois, +2.0
North Carolina, +1.1
Florida, +0.6
So big movement in Missouri and to an extent Illinois, some movement in North Carolina and little change in Ohio and Florida?
LikeLike
He adjusted some variables as well that he said improved his numbers in Missouri and Illinois, but hurt his numbers in Ohio. So his Ohio improvement is actually bigger than it looks and the two big improvements aren’t as big as they look.
LikeLike
Spot on Nick. Thanks
LikeLike
So what are you saying, is the margin in Ohio closer than it appears, and the leads in MO and IL smaller than they appear?
LikeLike
On the GOP side, who are the upsets in favor of?
Does Cruz benefit in both? Rubio or Kasich?
Or is it split between two of that group?
Would be chaos on the GOP side if all three win a state tomorrow, would ensure a brokered convention
LikeLike
The good news in Ohio is that the trend seems to be moving towards Bernie
(Last prediction had him at 48.09, now its 48.80)
That indicates there may be room for him to over-perform the Prediction tomorrow if there is a strong GOTV drive
LikeLike
Is there a post somewhere explaining how your model works? Even if only in general terms? Link please?
LikeLike
What data are using to come to your Ohio conclusion?
LikeLike
Don’t think he’s going to give away the recipe – and he shouldn’t. This could be a career making moment.
LikeLike
Yeah. Nate Silver blew up from his predictions in 2008. This could be very similar if he outpredicts fivethirtyeight twice in a row.
LikeLike
Except that FiveThirtyEight does explain, in detail, their methodology.
http://fivethirtyeight.com/features/how-we-are-forecasting-the-2016-presidential-primary-election/
LikeLike