I want to qualify this post with the notion that I am not, in any way, a hockey analytics expert. I don’t create content, and my statistical background is limited. That’s not to say I don’t understand or appreciate the significance/meaning of analytics within ice hockey. I merely want to suggest that, as an observer, I’m not a genesis of content.
The Stats vs. Non-Stats Crowd: A War of Some Against All
I’ve followed the hockey analytics community for about a year and a half, and Garret Hohl recently posted an article on Hockey Graphs about some things I’ve noticed; hostility and arrogance among several content creators within the community. The article suggests that some of this is justifiable, if misguided. As an outsider to analytics, I have to disagree. I’ve often struggled with the same feelings of arrogance and superiority. When critiquing the work of those who actively resist analytics, my mind often jumps to, “How is this not apparent to that person? The evidence is right there in front of you!”
The fault, however, lies with us who pretend that analytics is the be-all, end-all of interpreting and enjoying hockey. When I first studied hockey analytics, I felt empowered. Here was a relatively revolutionary way of understanding a sport I loved, one that spurned some conventional wisdom in favor of evidence-based methodologies. Most of my educational background was vastly in favor of the analytical approach, and I appreciated the great work the hockey analytics community put out.
As Garret has pointed out several times in the past, analytics shouldn’t be taken at face-value. It’s meant to be one tool in a decision-maker’s arsenal. The eye test has its uses, whether it’s determining a proper application of technique in-net or positioning on the ice. Analytics and observation are inherently tied together; one does not exist without the other. That distinction is important, especially when considering Garret’s bar-room discussion example.
Synthesizing an assessment of a hockey game and debating the merits of said assessment requires a holistic view. Arguing in favor of “just stats” or “just the eye test” does a disservice to the value of both methods. Using the bar-room debate scenario (albeit via Twitter), it feels like we’ve taken analytics for granted. I’m guilty of this too, especially when I bear down on a player with some of the great charts available. Sure, analytics provides some very useful insights into on-ice performance. The danger, however, is in assuming that analytics provides the only truth necessary to understand the game. It’s not fair of me or others to assume the superiority of stats to other forms of hockey analysis.
Unfortunately, a feeling of superiority continues to creep in to some of my conversations. I’ll be honest; a foundational understanding of hockey analytics can make one feel like an expert. In a number of cases, this is absolutely true. Hockey analytics is, after all, based on evidence collected in a relatively-scientific manner. The crux of the conflict is when those who support stats adopt a combative nature with those who do not.
The basic assumption, at least from my own experiences and observations, is that non-stats supporters are ignorant. At times, it feels like the stats community can view them as willfully stupid. That assessment might be unfair, but to outsiders examining the analytics community, that sort of pretentiousness has to be a turn-off. I don’t think it’s a consistently conscious feeling either. It’s a by-product of the age-old “educated vs. uneducated” conflict.
As ambassadors, content creators, or supporters of hockey analytics, we need to do be better than that. This is more than just seeking support and acceptance within the hockey community. It’s about treating each other with respect. It’s hypocritical to suggest everyone listen to us if we aren’t willing to listen ourselves. Listening to someone we don’t agree with can be the most challenging aspect, but failing to do so only breeds more conflict. Again, we have to do better and prove that stats is for the masses, not the few.
The Stats Crowd vs. Itself: A War of Some Against Some
The other troubling thing I’ve seen is how combative the stats crowd can be with its own members. Pioneering work in a new field is always a prickly endeavor, but it still surprises me to see the degree to which people dress up insults as “constructive feedback”. Plenty of criticism and debate is valid (and a good thing, I might add). There have been a number of times, however, where those who feel their work is the best continue to assert themselves upon others.
I understand that the young and exciting nature of hockey analytics is a prime breeding grounds for conflict. A number of content creators are competing to earn positions with hockey teams or gain notoriety for their research in the public space. That’s totally fine, and we should embrace that. Doing so by actively trying to tear down the valuable work of others, however, is unacceptable. Analytics relies on a diversity of interpretation and feedback, not baseless hostility.
I don’t want to make it sound like the analytics community is tearing itself apart. It isn’t. There are countless members who work collaboratively on some fantastic projects. It is, however, important to acknowledge that hockey analytics is an inherently-political endeavor. We should strive to support each other and actively improve the quality of research, not spend our time providing meaningless critiques that end in impasses.
Unjustifiable Arrogance: A War of All Against All
My final point is to bring all of this back to something Garret said in his post. He suggested that the arrogance and hostility of the analytics community is justifiable. From the article, it sounded like he meant justifiable when actively confronted with those who attack content creators. I’ll agree that it’s unfair to expect analytics community leaders to be perfect. We’re all human, and we all understand that every now and then, one of us will say something harsh. I’m in the same boat as everyone else.
That does not excuse or justify the attitudes that are often prevalent in this community. Analytics should be accessible and enjoyable by everyone. Instead, it often comes across as the domain of the elite. Maybe I’m baseless in my own critiques, but this is what outsiders see. I think we can do a heck of a lot better than that.