I've been "planning" to write this article for nearly a year now yet due to my laziness, never got around to actually doing it. Then I listened to the Steve Dangle Podcast episode with Jeff Marek, and a few quotes by Jeff finally gave me the motivation to start typing:

"...In a lot of ways, I think if we look 10 years from now and look back at how we're looking at hockey analytically, we're gonna laugh at it because it'll be a whole different way to look at the game from an analytics point of view that makes this look rudimentary.."

Adam Wylde jumps in with, "...when you're the only team using them, there's an advantage. But when you're one of thirty teams using them, where's the advantage?"

"...It's like talking about zone entries. If zone entries lead to more scoring chances, eventually everyone is going to tune into that and then wouldn't it be smarter to dump it in?" (paraphrasing)

What Marek mused on here is what I think will eventually become the analytics driven hockey metagame, though it won't be as clear cut as Marek's example.

To understand how this metagame will come about, it's important to know where the first true metagame was developed, and no it's not Moneyball and the Oakland A's; it's the world of competitive Magic: The Gathering.

Strap in kids, it's about to get real nerdy in here.

_______________________________________________

Back in the early days of competitive MTG, it was like the Wild West. People would show up with random decks and strategies to play against one another without rhyme or reason as to why they chose those decks. Eventually everyone caught on to which decks were strongest and blindly starting to play them because, hey, they were the strong decks.

Then as MTG Pro Mike Flores tells in his seminal article "I Never Metagame I Didn't Like: The History of the Magic Metagame", at the 1995 US National's Mark Justice won by playing neither deck, but instead a deck designed to defeat the two major decks. He had played the metagame and won.

After Justice's victory, people starting to play his innovative deck and began to beatdown the once mighty kings that were the mono-black Discard and Black Vise decks, and so those decks saw less play, thus in effect reducing the power of Justice's deck as people started to play decks designed to beat his. And so on and so forth. The true metagame had evolved.

So what exactly is the metagame you ask? Quite frankly, it's the name given to the current condition of Magic; who's playing what decks and in what numbers.

A defined metagame is one where it is generally clear what decks will be played (at a tournament), and the proportions in which they will be present. This can be by word of mouth, tournament trends (if a deck did well in a largely publicized tournament last week, you can expect to see it this week), or by the sheer obviousness and strength of a potential strategy.

So basically metagame is just the definition of what everyone is currently playing, but it doesn't tell us why they're playing what they are. For that, there's skew:

Beyond having a sense of specific decks that will be played, it is also possible to have a sense of general trends. This is skew.....There is also skew in an overall sense. This is what tournament players often mean when they refer to "metagaming": the consideration that others know what you might do to adapt to a given metagame, and the subsequent attempt to go one step further.

If some decks are going to be especially popular then many players will not play them, and instead play decks designed to beat those popular decks. Going further some players might even play decks designed to beat the decks designed to beat the popular decks because those decks might actually end up being the most popular ones. (metagaming)

By now you're thinking, what the hell does this have to do with hockey? Well let's go back to Marek's quotes on zone entries:

If zone entries lead to more scoring chances, eventually everyone is going to tune into that and then wouldn't it be smarter to dump it in?

What Marek is talking about is a potential hockey metagame where carry-ins are the predominant strategy for gaining the zone. Coaches will eventually catch on that most teams are carrying in the puck, and so switch to a defensive system designed to stop carry-ins; skew. So then the smart move for coaches becomes the dump and chase instead if they think the other team is setup to defend carry-ins; metagaming.

And this is where analytics come in.

In the early days, the MTG Metagame was established via word of mouth and tournament news coverage. But in the modern era, it, like most things now, is driven by analytics. There are a multitude of websites devoted to MTG analytics that use many of the same styles of analysis that we currently do for hockey, only they're far more tied into actually game outcomes and influence the MTG metagame far more than their hockey counterparts.

Head-to-head win probability, card advantage numbers, average deck construction...just tons of great great data for any MTG player looking for information to help them decide which deck to play at the next tournament.

Now back to hockey. While we're not yet at the same level of sophistication as the MTG analytics when it comes to defining and analyzing the metagame, I think we're on the cusp of being able to do so thanks to things like SportsVu, SportLogIQ, the great work on zone entries being done about the blogosphere, and the works of numerous others.

Soon we'll have league-wide data on things like which zone entry scheme is the best, what PP formation bests a certain type of PK formation, which side does a team prefer to attack; left or right, are certain defensemen unable to handle speed to the outside, which faceoff formation and play yields the best possession, etc...

Once all this data is collected and put into a useful format, coaches (just like MTG players) will be able to see what the trends around the league are when it comes to things like PK formations for example and then deploy a PP formation designed to counter them; perhaps even switching on a game-by-game basis depending on the tendencies of the teams they're playing.

And perhaps coaches of those opposing teams will also being using data to change their PK formations in anticipation of the PP coach changing his to counter their usual system. The coaches will be metagaming one another on a level beyond what they're currently doing and in every facet of the game as well.

This to me is the ultimate goal of hockey analytics; establishing an analytics-driven hockey metagame and giving coaches the tools they need to know what's being played, what beats that, and even what beats that. Gone would be the days of every team blindly playing the trap and it taking the NHL rewriting the rule book after a lockout to change the state of play; coaches will be redefining what NHL style hockey is on a near nightly basis and it'll be awesome.

Analysis