Disclaimer:
Disclaimer: While I am a paid undergraduate researcher for Texas Tech University, the opinions expressed on this blog are my own. The statements on this blog are not endorsements of views of Texas Tech University, The Honors College of Texas Tech University, or the math department of Texas Tech University, or any other department or campus organization of Texas Tech University.
Saturday, January 30, 2016
Guide to Election Calls
Just a clarification right before I begin calling Iowa. First, let's talk about my methodology. For Iowa, New Hampshire, South Carolina, and Nevada I will use a t-interval to generate a range. The other states will use a Bayesian approach using the results of the prior states and poll data to rank candidates based on probability of winning. Then I will also generate a range of percentage points. For both sets of races I will follow similar calling procedures. Based on that range I will pick a number based on recent events and polls that are regarded as more reliable that is in my opinion the most likely percentage on election day (My Center Percentage). I will then assign delegates based on the rules of that state for delegate allocation allowing for possible combinations of candidates if applicable. My primary goal is predicting delegates vs. percentages because delegates ultimately decide the race. Closer to July I will reorganize delegates if a possibly of a brokered convention still exists.
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment