The first significant breakthrough in hockey analytics occurred in the mid-2000’s when analysts discovered the importance of Corsi in describing and predicting future success. Since that time, we’ve seen the creation of expected goals, WAR models, and more. Many have cited that the next big breakthrough in hockey analytics will come once the NHL is able to provide tracking data. We’ve already seen some of the incredible applications of the MLB’s Statcast data and the NBA’s SportVu data. Unfortunately, the NHL has no immediate plans to publicly provide this data and as such, many analysts have decided to manually obtain the data.

We’ve seen some of the potential applications of manually obtained tracking data. Using zone entry data obtained by Corey Sznajder, Eric Tulsky and colleagues were able to generate zone performance scores. Sznajder’s All Three Zones Project shed light on the importance of possession entries and exits on shot generation. Ryan Stimson’s passing project demonstrated the impact of shot assists on shot quality and individual point production. Finally, Arik Parnass’ NHL Special Teams’ Project led to important findings about the habits of powerplays and penalty kills.

However, obtaining this type of tracking data is cumbersome and can be subject to tracker bias. Thus, on behalf of Rushil Ram, MetaHockey is happy to announce the creation of Tape to Tape, a tracking system designed to help standardize the way that we track passes, zone entries, and zone exits. With this system, you will be able to track zone entries, zone exits, and passes by pointing and clicking on a rink template.

Once you have selected a game, you can load the events recorded by the NHL play-by-play and the system will autogenerate the period, time remaining, and the players on the ice. It will also provide the X,Y coordinates for the recorded shots. Once you have selected a shot from the NHL play-by-play, you can add setup passes to record the shot assists.

With entries and exits, once you enter the period and time remaining, it will autogenerate the players on the ice at that time. You can then point-and-click to show the location of the entry or exit. From there, you can select if the entry was controlled or uncontrolled. If the entry was controlled, you can indicate if there was a setup pass for the entry.

Additionally, you will be able to link events from the NHL play-by-play to events that you’ve tracked. For example, if there’s a zone entry that resulted in multiple shots on goal, you will be able to link the zone entry to the recorded shots from the NHL play-by-play. You will also be able to tag the event with a phrase such as “one-timer” or “2-on-1” rush to help further stratify plays.

After the game has been tracked, users will be able to export the plays and roster as a .CSV file for analysis. Tracked games will be stored under your login for future use. The goal of this project was to make tracking games more accessible in the hopes that we can crowdsource enough data to discover meaningful conclusions. Long-term, my hope is that we can publicly quantify the importance of pre-shot movement and include this in the next iteration of expected goal models.

The idea for this project was conceived by Mike Gallimore (Twitter: @mikegallimore) back in July. The tracking system was developed over the last few months by Rushil Ram (Twitter: @__rushil__ ) and his team. Ram, a graduate of University of California at Berkeley, has worked as a full stack engineer in Silicon Valley in addition to working part time for Stats LLC to provide live play-by-play feeds for San Jose Sharks games.

If you have any questions, comments, or suggestions please direct them to either Rushil or the Meta Hockey Twitter account (@meta_hockey). The system is far from perfect and there are a number of improvements already in the pipeline. We hope that this system will be able to provide the data that we so badly need to advance hockey analytics.

Access the tracking system here at tapetotapetracker.com

Access the full list of How-To videos here