Hockey Analytics Competition?

Lately, I’ve been thinking about creating a public hockey analytics competition – So, this is just me thinking out loud.

I really think it’s a shame that there currently isn’t a public analytics competition or a public analytics forum (like Hockey-Graphs used to be). I know that the Big Data Cup exists, but unfortunately none of those submissions are published, so there really isn’t any peer-review process. I’m not sure I quite understand why the Big Data Cup submissions aren’t published (especially considering the data is anonymized).

Anyway, this is not an attack on the Big Data Cup or Stathletes by any means – It’s a great competition. I just believe there is a need for a public version and hopefully also a public forum for open discussion and knowledge sharing.

An open Hockey Analytics Competition

The idea is to make a hockey analytics competition soon after the NHL season finishes. I would obviously put my own database to free use. It would be NHL, PWHL and AHL data for sure, but there might be more data by then.

I don’t really want there to be any limitations or constraints on the submissions. The goals are to teach, engage and develop hockey analytics. There’s no gatekeeping – Everyone with an interest in hockey can participate.

Here are some of the areas I imagine could be covered:

• Storytelling with data: Could be player- or team specific analysis. Or perhaps a comparison between the goal scoring races of Gretzky and Ovechkin. It could be anything as long as there’s a focus on data.
• Statistical modelling: Building an xG model, Game Projection model, RAPM model or something else.
• Data Visualization: Building a viz in Tableau, Power BI, Excel, Shiny App or any other relevant tool.
• Data Tool: Could be building a Shot plotter, a Data Scraper or something entirely different.
• Data Research: Asking a specific question and finding the answer – e.g. What is the difference between playing a defender on his strong side vs. weak side?
• Educational Piece: Could be an explanation of the different concepts in hockey analytics or a tutorial how to build a Shot Map or something else.

The only rule is that all entries must be publicly available for free – That’s the whole point.

I have no idea if anyone would even participate. This article is in part to find out if there’s any interest at all. The hope is to have a broad panel of judges to pick the winner(s) and give feedback to all participants.

Creating a Hockey Analytics Forum

The dream is that the competition could lead to a more permanent hockey analytics forum like Hockey-Graphs used to be.

I’ve stated before that anyone is more than welcome to post data related articles on my website. Not that I have a large platform or anything.

However, a new Hockey Analytics platform doesn’t have to be on my website by any means. If a group of people are willing to go together and create a new site or even revive hockey-graphs, then I’m all for that… And I would gladly help out.

A Virtual Analytics Conference

It would be amazing if the competition could end with a virtual hockey analytics conference, where finalists could present their projects. However, this is well outside of my area of expertise, so I would definitely need some serious help if we were to put something like that together.

The Gap between Public and Private data

Finally, I want to talk about the growing gap between public and private hockey data. I do have access to some SportLogiq data, because I help out a team with access. That data is absolutely amazing in terms of micro statistics.

To me it’s very concerning that the gap between public and private data is widening, because it will eventually leave public work irrelevant. I’m a firm believer that innovation comes from data- and knowledge sharing – Transparency beats secrecy.

I wish the data companies would be more willing to share actual state of the art data. I’m not saying they should share any new data to the public – That’s what they sell for a living. However, sharing a full season worth of data from the 18/19 QMJHL season wouldn’t devalue their current product in any way. It would allow the public community to work with actual relevant data and at the same time it would show the world how amazing their data is. I really don’t see a downside to this.  

Anyway, that was just my little rant.

Help will be needed

If any of this is to come true, I will clearly need some help. First of all, I need help spreading the word. Please help share this article, so we can reach as many as possible.

Let me know if you would be interested in participating in the analytics competition. This whole project is nothing without submissions/content creators.

Please reach out if you want to help out in any capacity. Could be as a judge, as an organizer or perhaps you want to grant access to some of your data.

You can contact me on: hockeystatistics.com@gmail.com

I know most of you are much smarter and more organized than me, so I will need all the help I can get.