Big Data Cup

As a sumbission for Stathlete's Big Data Cup 2021, my partner and I looked at a large set of play-by-play data from the NWHL and 2018 Women's Olympic tournament to determine the most effecitve and most efficient types of passes to generate quality scoring chances.

View the Github repository and writeup for this project above

Whether you’re a casual viewer or a lifelong hockey analyst, you know that efficient passing is key to a great offense. Cross crease, outlet, stretch or just a good ol’ fashioned tape to tape, good passing leads to goals. While a gorgeous goal might make the highlight reel, a scoring play or a good look always starts with crisp, effective passing. For our question of interest, we wanted to focus on what types of passes resulted in the high efficiency and effectiveness. Specifically, which types of passes and how many of those passes resulted in the best scoring opportunities. And to measure the effectiveness of passing, we used expected goals per each shot as well as an adjusted effectiveness expected goals.

When you’re in the offensive zone, the goal is to create the best chance possible with the smallest margin of error. We looked at what passes best suited a team in achieving that goal We identified two types of passes — low-to-high and lateral — that we thought would be common during the power play and would create high danger scoring opportunities. Taking data from the 2018 Women’s Olympic Hockey Tournament and the 2019-20 NWHL season, we looked at offensive zone possessions on the power play and at even strength to see what the most common passing patterns during the power play and to determine which of those passes produces the best chances and which of those passes produces the best chances efficiently.