My Model of the GOP Nomination: The Five Northeastern States

 

shutterstock_252327457After plugging the results of the New York primary into my updated model (PDF here), Cruz has lost a lot of ground in terms of predicted votes in the future primaries. Trump’s probability of winning on the first ballot has risen from 22 percent to 41 percent (See p. 21 of the PDF). This increase is not due to delegates in New York (after all, Trump only gained three delegates in New York relative to my model’s median of 87 for him), but due primarily to Cruz’s poor vote totals in New York and the resulting declines in votes expected for him in future states due to the assumed correlations.

In today’s battle for 118 bound delegates in Maryland, Connecticut, Rhode Island, Pennsylvania, and Delaware, the model’s median values are 108 for Trump, four for Cruz, and six for Kasich. I expect Cruz to get more than 30 out of Pennsylvania’s 54 unbound delegates, but I’m focusing now on bound delegates.

The model’s 80 percent confidence intervals for today’s 118 bound delegates are: Trump 91–111, Cruz 1–7, and Kasich 5–19, according to table 2.5 on p. 22.

As described in section 2.2 on p. 20, the model performed well last week. The New York delegate model was almost perfect, and the model’s confidence intervals were hit at just about ideal rates.

Published in General
Like this post? Want to comment? Join Ricochet’s community of conservatives and be part of the conversation. Join Ricochet for Free.

There are 6 comments.

Become a member to join the conversation. Or sign in if you're already a member.
  1. BrentB67 Inactive
    BrentB67
    @BrentB67

    Thanks Chuck.

    • #1
  2. Josh Farnsworth Member
    Josh Farnsworth
    @

    Thank you Chuck.  If you were to game out a Cruz victory in IN by moving 60 percent of Kasich’s vote total to Cruz and 20 percent of Kaisch to Trump and 20 percent of Kasich to stay with Kasich, does your model account for what a Cruz victory in IN would do to his CA totals.  Or would Cruz have to get more than 60 percent of the Kasich support to pass Trump in IN?

    Also, the Pollsters podcast noted how difficult predicting CA delegates would be due to low numbers of GOP voters in certain Congressional Districts.  How did your model account for low-turnout by Congressional district?

    • #2
  3. Chuck Walla Member
    Chuck Walla
    @ChuckWalla

    Josh Farnsworth:Thank you Chuck.

    You’re welcome!

    If you were to game out a Cruz victory in IN by moving 60 percent of Kasich’s vote total to Cruz and 20 percent of Kaisch to Trump and 20 percent of Kasich to stay with Kasich, does your model account for what a Cruz victory in IN would do to his CA totals. Or would Cruz have to get more than 60 percent of the Kasich support to pass Trump in IN?

    Yes, I’ve been thinking about this too.  I am almost sure that Cruz’s outlook in Indiana is better than my model thinks, but I will wait until after today’s results to update my Indiana assumptions.  Today’s PDF does not include any calibration to recent markets or polls, but I’ll do that before Indiana voting.

    Check my thinking on this, but I expect a Cruz surge in Indiana that did not happen in New York due to the huge importance of Indiana and resulting emphasis on it by the Cruz campaign.

    Also, the Pollsters podcast noted how difficult predicting CA delegates would be due to low numbers of GOP voters in certain Congressional Districts. How did your model account for low-turnout by Congressional district?

    Yes, my model is supposed to take that into account.  Had I ignored this consideration, my Trump delegate projections for New York would have been too high.

    • #3
  4. iWe Coolidge
    iWe
    @iWe

    Impressive, indeed!

    • #4
  5. Doctor Listener
    Doctor
    @Doctor

    Fascinating analysis!  Thank you!

    • #5
  6. Chuck Walla Member
    Chuck Walla
    @ChuckWalla

    Thanks for all your kind words.  I plan two updates before Indiana:  The first will reflect on yesterday’s results; the second will look forward with emphasis on Indiana.  I’m thinking of comparing two outlooks for the first ballot depending on whether or not Cruz wins Indiana.

    Any other requests?

    • #6
Become a member to join the conversation. Or sign in if you're already a member.