I had the itch yesterday and wanted to dust off the ole’ blog.
I don’t really consider myself a journalist, although I did minor in the practice in undergrad, and have written a couple of longform pieces that fall outside of my normal practice of statistical hockey analysis. That said, I feel pretty strongly that what’s out there on average, particularly in the field of sports journalism — especially in terms of hockey journalism — could be a lot more convincing than it is..
I tend to come back to this clip from Moneyball, because it’s what I use as my litmus test of quality whenever I hear or read a person in the public media.
Obviously the clip involves front offices, but there is overlap with the media. “You guys are just talking, lalalalala.” If I’m doing that crocodile hand gesture while listening to you talk, it’s probably not a good sign.
Now I want to be clear that there a lot of good media people doing good work out there, and it can’t be said enough how refreshing that is, people from hugely varying backgrounds who are constantly seeking to improve themselves and deliver more than the minimum.
But I don’t see it as the industry standard; it’s certainly not a requirement.
Let’s start with former players. If you’re a former NHL player — on the radio, on TV, in print — let’s be honest, you’re probably in that position because you’re a former player. I don’t want to paint a broad stroke, but it’s unlikely you worked your way up through unpaid internships and fruitless freelancing and part-time jobs at Burger King or the local pub to get you through journalism school. If you’re a former player in the media, you need to be able to provide something that justifies you getting that position over somebody with the experience, with the advanced degree, with the likely heightened vocabulary and articulation. You need to be able to provide anecdotes and high-level insight from the perspective of a player. If there are 20 games left in the season and a team is out of the playoffs, tell me what that’s like for a player; tell me what it was like for you. And I don’t mean conventional platitudes like “well it’s been a tough year for them but they’ve gotta strap on their workboots and show some character.” I mean tell me about that one time Joe Smith and Billy McKenzie got into a pre-game argument over commitment. Or about maintaining a grinding day-to-day routine (what is that routine? Did you ever cheat on it?) when you knew you weren’t playing for anything. My general rule is, if I’m reading or listening or watching a former player be paid to talk about hockey, and I can’t tell they were a former player by the insight they’re providing, then they’re not doing a good enough job. That may sound harsh, and it’s doesn’t mean every second sentence has to be an anecdote, but former players are in a unique situation to be able to provide new and original viewpoints, and if they’re not doing it, then how are they enhancing their medium?
Now let’s talk about analytics. As I’ve written and spoken about before — probably to an obnoxious extent — “analytics” isn’t a human being, and it’s not a viewpoint. Analytics is a way of doing things, a method of analysis, a data-based and informationally-inclusive (a made-up term but a fitting one) practice. So the “analytics” don’t say something about a player. I cringe anytime I hear that. I know it’s just conjecture, but I think it can be damaging in the sense that it paints a picture of the practice as being linked to one number — usually Corsi — when analytics involves weighting numbers and scouting and reputation in a manner that allows one to get as close as possible to an accurate evaluation. Knowing a metric’s strengths and weaknesses is vital to proper analytics, and knowing how to use certain metrics and figuring out which metrics should be generally avoided is critical as well. Data analysts disagree all the time, so “analytics” can’t possibly say any one thing about a player or team, in the same way you would be scolded for saying that “scouting” leads to a specific conclusion about a player. If that were so, there would be no point in teams having their own scouts. The idea is to use analytics more effectively than another team, which is why simply reading public work, while important, is far from exhaustive.
But I digress. Analytics have largely been ignored in the media, for the most part because journalists aren’t schooled in math and just don’t know what metrics are out there and what they mean. But for the media, implicit analytics are perhaps more important than explicit ones. What does that mean? You don’t need to know every number, you just need to know on a league-wide level what the numbers have been finding. For example, implicit analytics mean understanding the concept that the best defense is a good offense, to the extent that you don’t think a purely defensive defenseman whose shot differentials are among the lowest in the league is an excellent pickup. Or, maybe more importantly, you know the two sides of the argument so that you can outline them for viewers in print or on-air. It’s very true that the average hockey fan doesn’t want complex metrics shoved down their throats, but they do want good analysis, and just because they haven’t read the piece that mathematically proves the relevance of a metric, doesn’t mean they aren’t interested in the analysis that comes from it if presented at a basic level. Case in point…
“Hockey is different from a sport like baseball or even basketball because offense and defense are so intertwined. Players often will play far more of one than the other in a game because there isn’t an imposed even amount of both. So you can be a less than great defender in your own zone but still be an incredible defenseman because you’re so good at driving play forward that you spend very little time in your defensive zone and are deadly when on offense.”
“Shot attempts are an important measure because they accumulate quickly so we get bigger samples. That means that the complaint folks always have about plus/minus — a player could get a plus for just jumping over the boards — isn’t as relevant because that will all even out more quickly. Of course all shots aren’t equal, but over time it’s been found shot quality somewhat evens out, and in fact shot attempts because of all of these factors do a good (though not perfect) job of predicting future goal differentials and wins and losses, better than basically anything else we have.”
Any time I’ve explained these concepts to a friend or old school colleague they’ve understood. There’s no reason the average hockey fan wouldn’t. There are no numbers involved, but if the average fan understands these implicit analytic-based ideas, then next time when somebody mentions “shot differentials” they won’t change the channel or turn the page — which, after all, is generally the idea.
So far, the use of analytics in the media has been confined to mostly “Analytics Websites” or “Analytics Shows,” but as I mentioned earlier, “analytics” isn’t a viewpoint. Regular writers should be able to incorporate implicit — and maybe even basic explicit — analytics into what they’re doing because it is now more and more how front offices are looking at the game, and rightly so. Hockey fans don’t want complex numbers, but they don’t want to be lied to or deceived, and they want to feel smart. If they can understand it, I believe they will tune into the network that delivers the highest quality analysis (they already do as evidenced by the ratings over the years for one certain network relative to another). It’s the responsibility of the journalist to seek out explanations and to educate him or herself about these concepts, so that they can pass those lessons on to the public. It’s also incumbent upon the network or publication to educate its staff, ideally with workshops but at least with suggested readings or videos. It wouldn’t be difficult to make either of these things happen.
The essence of all this boils down to a few main ideas.
- As journalists, hockey media people and networks have a responsibility to deliver accurate and high quality analysis
- That can come in different ways, as former players for example need to be able to draw on their experiences to deliver unique and insightful content
- “Analytics” aren’t a single viewpoint, they’re a practice. Implicit and eventually basic explicit analytics should be contained within every media broadcast, just as they’re becoming essential for accurate evaluation in NHL front offices.
If you read about baseball or watch a baseball studio show (well, some baseball studio shows) you will see that what I’m talking about has happened in that sport. It’s only a matter of time before it takes place in hockey. If you’re a media network that wants to boost ratings without going full TMZ, or a media person simply wanting to get better at what you do, be a part of this new generation, no matter your age. You don’t need a calculator, only an open mind.