The HALO Project

Exploring AHL Player Tracking Data

This project digs into Teamworks and Sportlogiq AHL player tracking data to surface insights that go beyond traditional box-score statistics. From shift-level visualizations to expected goal models built from raw event data, the goal is to find meaningful signals for player evaluation and team success.

What You Can Do Here

This site lets you explore AHL player and team performance at a level of granularity you won't find on traditional stats sites:

  • Shift-by-shift breakdowns — See how a player performed across every shift in a game, not just the final stat line.
  • Live rink visualizations — Watch tracking and event data unfold play-by-play on an interactive hockey rink.
  • Expected goal analytics — Go beyond goals and assists with xG models derived from raw event data, broken down by situation (5-on-5, power play, etc.).
  • Deep-dive blog posts — Analytical write-ups that use this data to answer specific questions about player evaluation.

Start with Team Standings to pick a team, drill into a player, and follow the data all the way down to individual shifts. You can also jump straight to any player from the Player Stats page.

The Data

The dataset contains a 30-game sample for 1,172 AHL players across parquet files for players, games, stints, events, and tracking data. That adds up to:

  • 480 games
  • 1.8 million+ event rows
  • 2.2 million+ stints
  • 13.5 million+ tracking data rows

You can browse the raw data directly in the Sample Raw Data section.

Get in Touch

Have questions, feedback, or ideas for new analyses? I'd love to hear from you.