Every election year it is exciting to make predictions about who will win each state by how much. However, you may wonder how accurate the predictions are. With this simple election prediction tool, you can quickly assess the accuracy of your prediction!

What it does

It uses regression models to predict the accuracy of your election predictions based on the actual results of the 2020 US Presidential Election. Predict a few, or all 50, and see how well you do! Challenge your friend to do so too! See who gets the best percentage! Anyone can login with google to make a prediction.

How I built it

I used a JavaScript Framework: React, and I integrated it with Firebase on the backend. It featured authentication and data storage.

Challenges I ran into

Coming up with the right regression model was difficult, but I settled for a simpler and more objective approach of calculating residuals for the results based on the mean vs. user predictions, and comparing them for a percent accuracy number.

Accomplishments that I'm proud of

I built this in a short amount of time, the week of the election, in a timely manner. It is also extremely simple and easy to use by anyone, and you don't have to predict all the states to get accuracy percentage calculated for you.

What's next for Election Predictions 2020

I will enhance the UI/UX, add add more info about each election prediction

Share this project: