Milestone 1- NHL Expected Goal Model
Motivation
In the sports analysis domain, there already exists models that attempt to predict expected statistics using various amounts of data. For this project, we wanted to create our own expected goals model for the NHL where we can control the inputs and the weight they have that affect the model’s prediction. Finding the likelihood of scoring a goal for for every shot is nothing new, however the question we wanted to answer is are those models accurate. By framing our project in the mindset of building a credible model that can be used in a practical way, we were able to identify requirements and limitations of the model to best optimize its performance.