Yesterday I wrote Mish Projection: Trump Wins Nomination with 1,246 Delegates (9 More than Needed) .
I almost held off for a day posting that. Typically, when I do that, I regret it. Someone else beats you to the idea (or appears to), even though you had the idea first.
Today, Nate Cohn at the New York Times writes Donald Trump's Path to 1,237 Is Not Mission Impossible.
I predicted a winner, Cohn didn't. However, our reasoning, and even the states we mentioned were nearly identical.
Not Mission Impossible - Cohn vs. Mish
Until Wisconsin, Donald Trump was considered likely to win a majority of pledged delegates. In truth, his plan to reach 1,237 was already very vulnerable; Ted Cruz had built enough support by March 15 that even adding a modest share of Marco Rubio's voters was likely to start him on the road to deny Mr. Trump a majority.
But after Mr. Trump's loss Tuesday, the conventional wisdom has gone too far in the other direction. His path to 1,237 is still clear. It is certainly narrow, but it may require him to do only two challenging things: win two tossup states, Indiana and California. There's an argument he's currently favored in both.
The polls in the Northeast suggest that Mr. Cruz would be lucky to end with the split of anti-Trump voters (38 percent for him and 29 percent for Mr. Kasich) that he got in Madison. Mr. Kasich has generally run ahead of Mr. Cruz in the Northeast.
His path is narrow enough that every delegate counts. He might need as much as 90 percent of the delegates from the Northeastern states and West Virginia to keep his delegate target in California manageable. But it's not very difficult to imagine Mr. Trump doing so well.
My big pickups [over Silver's expert panel] are New York (+15), Indiana (+8), and California (+25).
Trump is polling extremely well in New York, over 50%. I gave Trump 14 at the state level.
At the district level, if at least 2 presidential contenders receive 20% or more of the vote, the candidate with the most votes receives 2 delegates and the candidate with the second most of votes receives 1 delegate.
Curiously, this math helps Trump. Kasich will take votes from Cruz, not Trump. Is Kasich about to drop out? If not, he helps Trump at the district level, but potentially hurts Trump at at the state level.
[Note: I awarded Trump 86 of 95 NY delegates, just over 90%]
The most important state that no one is talking about is Indiana. The contest there comes one week after the Northeastern primaries, and it's arguably the most balanced state left in the race. It has a mix of both Mr. Cruz and Mr. Trump's strengths, somewhat like the two states where they've fought two of their closest races so far, North Carolina and Missouri.
Indiana awards its delegates on a winner-take-all basis by congressional district, which could easily allow Mr. Trump or Mr. Cruz to claim most of the state's 57 delegates with a modest victory.
There isn't much polling data in Indiana, but Mr. Cruz would be a favorite if he could get the level of support among non-Trump voters that he did in Wisconsin. It would be a closer race than Wisconsin; it's the type of state where Mr. Trump ought to approach or exceed 40 percent of the vote, not the 35 percent he won in Wisconsin.
Provided he dominates in the Northeast as expected, Mr. Trump will have a good chance to win the nomination if he can carry Indiana. Without it, it's very difficult for him to reach 1,237.
[I guess I am no one. Regardless, this is what I had to say about Indiana]
The experts gave Indiana to Trump. My rationale for Indiana was to simply go with the experts and work backwards.
Indiana awards 30 delegates to the overall winner. I put those in Trump's column.
Indiana has another 27 delegates (9 districts of 3 delegates each). I gave Trump a modest 5 of those 9 to reach my total of 45 [out of 57].
If the experts are correct overall, I will beat them. If the experts are wrong, I will do much worse.
If Mr. Trump wins big in the Northeast, carries Indiana and picks up a few proportional delegates in New Mexico, Oregon and Washington (as he is all but assured to do), the race will come down to California on June 7.
It's too far away to be very confident about whether Mr. Trump would have a realistic chance to win the 70 percent or so of California delegates that he would need to win an outright majority.
But, at least right now, it looks realistic.
California is a curious bird.
Silver's Polls-Only California Forecast says Trump has a 65% chance of winning. However, based on his secret sauce, Silver says Cruz has a 61% chance of winning.
This is how the secret sauce projects things.
I find Silver's secret sauce all the more incredible because Silver projects Trump to have a 49% chance of winning the nomination.
Adding to the confusion, Silver's expert panel has Trump winning the state.
Here's the deal: If Cruz were to win California by those margins, Trump would not now realistically have a 49% chance of coming up with the delegates, unless I am way off on my assignment of a mere 15 of 125 uncommitted delegates in Trump's camp.
Alternatively, Trump can do better in states like Colorado, Connecticut, Oregon, Washington, and New Mexico, coupled with 25 of 125 of the uncommitteds.
[I gave Trump 118 out of 172 California delegates. That's 68.6%. So Cohn and I are thinking similarly about percentages needed. I am not at all confident about that number but I underplayed Trump's chances in Colorado, Connecticut, Oregon, Washington, and New Mexico]
Provided he dominates in the Northeast as expected, Mr. Trump will have a good chance to win the nomination if he can carry Indiana. Without it, it's very difficult for him to reach 1,237. He would either have to win nearly all of California's delegates or win a state where he's an underdog -- most likely Montana -- and post a clear win in California. It's possible, but it's hard to see how he would be poised to do either of those things if he's losing in Indiana.
Obviously a lot could change between now and June, but this is not a crazy scenario. It's consistent with the current state polling and it's consistent with how demographically similar states have voted so far this cycle. If Mr. Cruz can't unify Republican voters, it might just happen.
Mish - Bottom Line - Close to Tossup:
Some states I know very little about and there are no recent polls. Five states, shaded in blue, (Colorado, Connecticut, Oregon, Washington, and New Mexico) are places where I went with the expert opinion. However, I suspect Trump can pick up a handful somewhere.
Montana appears to be rigged as delegates are selected by the County Central Committee, but perhaps Trump can unexpectedly pick up a few of those.
My projection is not fearless.
All I did was extrapolate expert opinions in the logical direction based on state-by-state analysis of delegate assignment rules from the Green Papers.
We are very close to a tossup here. I placed Trump over the top, but barely, explaining my sauce. Other sauce remains secret.
Things can dramatically change one way or another between now and California.
If Silver gets to update his forecasts, so do I.
Cohn and I wrote similar posts. Our logic was nearly identical. Indiana, a state allegedly "no one is talking about" was was a key state in my analysis.
I mentioned Montana, as did Cohn.
I was first.
This is not an accusation. I strongly suspect I Cohn is one of those few people who actually dive into the details to look at what's cookin'.
The key difference in what we wrote is I made a prediction, he didn't.
To see my state-by-state projections and detailed reasons fore each state, click on the above link.
I reserve the right to be wrong, but my overall analysis was sound.
Thanks for the confirmation of my analysis Nate Cohn!