Hockey Analytics

Apply data science to hockey and broader sports analytics

---

Project maintained by justinjjlee Hosted on GitHub Pages — Theme by mattgraham

Data Science and Hockey Analytics

This repository contains data science toolkits I applied to analyze and visualize National Hockey League (NHL) statistics. This includes descriptive, diagnostic, predictive, and prescriptive component in the context of sports analytics.

This is a selection of many ways to collect, compile, clean, analyze, model, and predict player (skaters and goalies), team, and market performances statistics. All original codes (including generic and model algorithms) can be used freely with a proper citation to this repository.

1. Data Collection and Processing
2. Analyses and insights
3. NHL game prediction model - TBD
4. Other relevant sports analytics

This repository does not claim any ownership of data nor represents points of views of organizations or representations mentioned.

Data Processing

Following link shows example workflow process of collecting and processing scraped and called (via NHL API) data. All original source code is located in src of this repository.

• NHL data scraping, collection, and cleaning
• NHL Stats pull and consolidations
• NHL in-game data collection via API and processing - Github Action

link to the top

Analyses

• Better than coin flips? Money line and market expectation in NHL
• Positional success by Stanley Cup Winners: A case for Oilers and Maple Leafs
• Decision Science - What happends after blocking shots
  • Relates to My blog post Human Shields on Ice: The Valor of Blocking Shots

link to the top

Other Sports Analyses

Following analyses are not hockey-related but incorporated to provide conceptual views applied in other analyses. Concepts and analyses methods used in the analyses are referenced with the analyses used for the game of hockey.

• How Michian Wolverine’s undervalued prospect became NBA’s top rookie

link to the top