I just finished an 8-part series on roles and statistics with a specific intent of showing that certain types of players or playing styles have impacts on save percentage. It was received well by some, not so well by others. Clearly there are detractors however this is a case of been there, done that for me. I went through the same debates 5-6 years ago with on-ice shooting percentage. I am confident down the road the hockey analytics community will begin to accept that players can influence save percentage just like most have now accepted players can influence on-ice shooting percentage. It’ll just take some outside the box thinking to get there. Hockey Analytics needs more outside the box thinkers.
I have been a part of hockey analytics for a long time now. According to Rob Vollman this blog is the longest continually active hockey analytics blogs (just past the 11-year mark) and of course my stats sites are among the longest running and most popular as well. I was also the first to generate and post salary tables for every NHL team (then CapGeek came around and did it 10 times better so I refocused on stats). I was also one of the first to seriously question whether zone starts really have much of an impact on a players statistics and the importance of quality of competition. I am also confident that we will eventually look back at current expected goal models and realize they don’t do that good of a job at accounting for shot quality. I also came up with the idea of how to identify “rush shots” from play by play data which is now used in many of those expected goal models. I have at times been a harsh critic of some of the work done in hockey analytics however I only do so because I want to see hockey analytics constantly push itself to be better and not stagnate. I hope that I have helped push hockey analytics forward.
Looking forward there is a lot more I want to do in hockey analytics. I want to explore more areas where others feel we have reached dead ends. For example, I want to do some investigation into giveaway/takeaway stats which I wonder if we have prematurely cast aside. I am also starting to think we need to do more investigation into how we calculate quality of competition as I now wonder if we may be understating it. I want to do more work at the game by game level and shorter term trends though that will require a significant reworking of my current code base. I also want to do more with some of the data collected by the various tracking projects. Longer-term, when the player and puck tracking data is available I really want to dig into it as there are some really interesting ways that we could look at that data.
With all that said, I need a break. I need a break from blogging as it can be time consuming and take time away from some of the other things I want to do. Mostly though I need a break from hockey (analytics) twitter. There is just too much negativity and closed mindedness on hockey twitter these days and I think twitter brings out the worst of everyone (myself included at times). Hockey twitter can be quite closed minded (if not intolerant) of other peoples opinions, particularly those that have’t bought into their view of hockey analytics. Even as I write this I see twitter updating in the background and a half-dozen people have pointed out (in a ‘haha, look what the idiot said’ kind of way) a comment an announcer made about analytics in one of the World Cup hockey games. For a community that can be so forward thinking when it comes to social issues, depression and mental issues it is sad to see the (at times seemingly relentless) attacks towards those who they don’t agree with. There is a lot of good stuff in my twitter feed but there is a lot of nonsense too. I am at the point that hockey twitter often makes me cringe more than smile and that isn’t what it should be. For these reasons I am dramatically scaling back my public presence on twitter and blogging about hockey analytics.
My stats sites should remain up so long as the NHL doesn’t change their stats format so no need to worry about that. Depending on what I do I may even add features. As I said, I am not done with hockey analytics. Far from it. My interesting and work in hockey analytics will continue and I may have opportunities to participate in some very interesting projects in the near future which I am excited about. Also, if you have a project you want to collaborate on, definitely feel free to reach out. I’d be happy to collaborate with some smart, creative and open minded hockey fans, agents or teams.
Long story short, I am not leaving hockey analytics, I am stepping back from public hockey analytics because I need a break from the public nature of analytics. I don’t know whether this will last for 2 months or 2 years but starting today and for the foreseeable future I will be on a public hockey analytics hiatus.
I am enjoying it already.