Math Error in the New York Times “President Prediction Model”

There is a mathematical error in the algorithm used in the New York Times “President Prediction Model” (“2016 Election Forecast: Who Will Be President?“). 

The model is based on voting history for each state and hundreds of state and national polls. The model then extrapolates the probability that a candidate will acquire sufficient Electoral College votes to win the election and become President of the United States.

This algorithm is mathematically incorrect. It does not account for the time period between the Election and the Presidential Inauguration. The algorithm predicts the odds of a candidate becoming President-elect and then assigns 100% certainty that the President-elect will be Inaugurated President forty days later. While it is highly probable the President-elect will become the President, it is not mathematical certainty.

  1. The President-elect could become incapacitated or die.
  2. The President-elect could voluntarily decline the presidency for personal reasons.
  3. The President-elect could decline the presidency due to political pressure, scandal or high crimes and misdemeanors.
  4. The Electoral College could elect someone other than the President-elect.

These factors have statistical significance. I am not being political – for me it is all about the math 🙂

Isabella Blanchard
Cherry Hill, New Jersey


bellachance™  Copyright © 2016