clock menu more-arrow no yes

Filed under:

Guest Post: Joel Armia, Finnish God of Shots

New, comments

A new metric for spotting tomorrow's stars.

Timothy T. Ludwig-USA TODAY Sports

Aside from working in video & analytics for the McGill University women’s hockey team, I have also been playing around with a new metric for evaluating prospects and minor-leaguers. As is the case in major junior and Canadian university hockey, the completeness of league-level stats-keeping in the AHL sometimes leaves to be desired. Goal-based stats, as we know, are tremendously affected by luck and random variation, which means the basic box scores are generally a poor source of predictive information.  And since we currently don’t have any real-time stats to scour through, it is a lot more difficult for those interested to evaluate players at a higher level.

What can we do about this? Well, by taking a look at the Icecaps’ AHL stats page, we can see that Games Played, Points and Total Shots on Goal (both individual and team) are supplied. By opening the player profile of each individual, we can also get his year of birth. Knowing the importance of shot generation for future success, as well as the large role played by relative age in a player’s development path, we may intuitively get the sense that there is insight to be gleamed just by mixing all numbers together. (Note that it would have been nice to have access to even-strength production, as opposed to total production, but we don’t get that on AHL.com, so there’s that.)

Now, I have not had much time to play around with the weighing of the formula in order to get the maximum correlation to future NHL point per game (which we assume is what most teams are ultimately "buying" when drafting), but the basic concept is the following:

1) It’s good for a player to take shots, whether they go in or not. At least it means they frequently possess the puck and are generating opportunities in the offensive zone.

2) It’s good for a player to take a lot of shots relative to his teammates. Some teams average 35 shots a game, others average 25. A three-shot-per-game player on each team will probably have a different true ability level.

3) It’s good for a player to put up big numbers in their late teens or very early 20s. Being a point-per-game player in the AHL at 18 means you are well on your way to superstardom. Being a point-per-game player in the AHL at 30 means you’re probably a replacement-level NHLer.

If we look at the top 20 AHLers (min. 10 games) ranked by their AASS since 2010, we get a list that looks like this:

1) Jason Zucker

2) Kyle Palmieri

3) Reid Boucher

4) Morgan Rielly

5) Tyler Toffoli

6) Brendan Gallagher

7) Brad Marchand

8) Max Pacioretty

9) Filip Forsberg

10) Sean Couturier

11) Roman Josi

12) Ryan Strome

13) Dmitrij Jaskin

14) Michael Grabner

15) Brayden Schenn

16) Nikita Kucherov

17) Brock Nelson

18) Michael Bournival

19) Jeremy Morin

20) Ryan Sproul

We also have guys like Mika Zibanejad, Taylor Hall, Ryan Nugent-Hopkins and Brandon Saad ranked not too much farther down the list, well inside the 90th percentile of all AHLers observed in the last 5 years.

Things get interesting at this point, because we are starting to see a bit of a pattern. The vast majority of guys I just named are players you’d either like to have on your team, or players for whom you'd would sacrifice a first-born child to have on your team.

So maybe we’ve got something here.

In any case, the true test for the AASS metric is how well it predicts the NHL performance of prospects currently in the AHL. Two examples I had discussed in the past over at Habs Eyes on the Prize are Christian Thomas and Sven Andrighetto, who have experienced some success in their NHL stints with the Habs this season.

So how does this relate to the Winnipeg Jets and Joel Armia? Well, the Finn happened to be one of the highest-ranked AASS players in the Buffalo Sabres system before being traded to Winnipeg in the Evander Kane deal. Despite not being talked about as a huge piece in the trade, the 21 year-old's age-adjusted individual shot production puts him in the neighborhood of aforementioned Zibanejad, Hall and Nugent-Hopkins. Jets fans, you may just have a pretty good one on your hands! (But only time will tell.)

Jack Han is the Video & Analytics Coordinator for the McGill Martlet Hockey team. He also writes occasionally about the NHL for Habs Eyes on the Prize. You can find him on Twitter or on the ice at McConnell Arena.