Discover how the NHL's groundbreaking player and puck tracking technology is revolutionizing the way we understand and analyze the game. From improved accuracy to better predictive models, this data-rich initiative promises a brighter future for hockey analysis and fan engagement.

On Monday morning, the NHL launched a brand new arm of its site that debuts the fruits of player and puck tracking. This is exciting for a number of reasons and for everyone from the average fan to the nerds. There are so many different ways to use various statistics and this launch seems to mirror the NBA advanced stats site accessible to fans. 

The NHL announced player and puck tracking nearly half a decade ago, however it was unknown whether or not the stats would be made public. The past few seasons have seen the tracking stats available on television broadcasts in the form of shot velocity, skating speed, skating distance and shot location. However, none of the statistics were available for public use. 

Public modelling is most difficult in hockey because of the lack of available metrics to use. Quantifying defence has been extraordinarily difficult compared to the NFL or NBA metrics. Currently, zone entry details and controlled zone entries were the most we had to quantify a defenceman's ability to defend in transition. While there is no skater proximity available yet, it is being tracked and will give the public access to understand how good or bad the gap control truly is. 

A second reason that modelling hockey is difficult is the accuracy. There is always human error in tracking and using technology will help reduce the errors with the goal of eliminating them altogether. Accurate shot locations and shot types will improve the accuracy of public expected goals models because the data is better. Accurate shooting statistics such as on net, blocked or missed should remove bias that have been rampant across the league for years. Generally speaking, the most important part of this release is the improved accuracy of the data. Models are only as good as the data they draw from. Inaccurate data means inaccurate models. The improvement in data accuracy should have a tremendous impact on models going forward. 

Given that most journalists refer to models in their analysis, it means the analysis should improve as well. While the dangers of making this Pandora's box of data available exist, they are heavily outweighed by the positives. Sure, journalists and columnists can cheery pick data to fit their narrative. But they were already doing that. This plethora of data won't change that. What it will change is the ability for more people to understand their cherrypicked narratives and form their own conclusions. More access to data will generate interest from fans who want to be educated on the sport, specifically for gambling reasons. It will also allow the casual fan to understand how fast a given player is, how much skating is being done and how goalies are performing relative to their beliefs. 

From a modelling perspective, the NHL has made it clear that it invites gambling on the sport through various partnerships. Making more data available allows those who build predictive models (like me), to build models using better features with increased accuracy. This will automatically improve the quality of the models and should make them more accurate. This is only the beginning of what could be released, as behind the scenes, there is far more data being tracked and used. Stats are here to stay, nerds aren't going anywhere, and the NHL Edge site is a great way to target a different segment of people who are curious about different stats in hockey. 

This site is in its infancy and will undoubtedly improve in the coming years. Off the hop, it's a win for casual fans, a win for analysts, a win for model building nerds and that means, it is a win for the NHL.