Mining NHL Draft Data and A New Value Pick Chart

It appears your Web browser is not configured to display PDF files. Download adobe Acrobat or click here to download the PDF file.

Click here to download the PDF file.


Wilson, David




For every hockey player, getting drafted to the NHL is a dream come true, but the real goal is to reach 160 games played (GP). A simple model is fit to NHL draft years 1998 to 2009 with an aim at predicting the proportion of players who will play 1 or 160 games. For individual players drafted between 1998 and 2011, predictive models (GLM, Neural Network, SVM and LOESS) are created to predict a player's career GP, then combined to create a voting model to decide whether or not a player will reach 160 GP. Players drafted from 1998 to 2008 were analysed using non-linear multi-variate model, with a modified weighted least squares, to predict a player's Time-On-Ice (TOI) for their first seven seasons. The results were analysed for positional advantage, then converted to value pick charts (used by teams when choosing a draft pick or considering a trade).






Carleton University

Thesis Degree Name: 

Master of Science: 

Thesis Degree Level: 


Thesis Degree Discipline: 

Applied Mathematics

Parent Collection: 

Theses and Dissertations

Items in CURVE are protected by copyright, with all rights reserved, unless otherwise indicated. They are made available with permission from the author(s).