The Prospect Pyramid: A Response
7 years ago
While my typing fingers were silent last week, I did spend a little bit of time reading the blogs to see what everybody was up to. One of the articles that came across my eyes was on our very own TheLeafsNation.com, written by our pal and Senior Vice-Present of Loud Noises, Steve “Dangle” Glynn. In it, he proposed something called a “Prospect Pyramid“.
First off, it was neat to see Steve turn to written words on this platform; something we haven’t often seen since the days where Dave Nonis was the Leafs GM and our editors had noted girlfriends. But the idea was intriguing; a counter to the prospect rankings so many blogs use, published just before we concluded our own series.
Steve’s argument can be simplified as such:
- Prospect Rankings make for plentiful content, which is why platforms do them
- They also make for excessive arguments over minuscule differences
- Arguments can be eliminated by sorting prospect pools into a tiered pyramid, which separates certain players from the pack while lumping others who are at similar levels together.
It’s a noble effort, and I get where Steve is trying to get at. Micro-analyzing things can be frustrating at the fan level, and this allows for a bit more wiggle room. But I don’t know if it’s enough room to make the difference that it sets out to.
The reality here is that Steve’s pyramid doesn’t eradicate available arguments, it just adds to them.
Chandeliers inside the pyramids
- How exactly are we deciding who should be eligible and who shouldn’t be? Josh Leivo is ineligible Steve’s pyramid because he’s made cameos in a couple NHL seasons and has amassed some experience. Kerby Rychel is in despite playing more games last year than Leivo has played ever. Justin Holl isn’t out because he’s 24 and about to be an NCAA Senior + 2 player. Zach Hyman is in, despite being 24 and an NCAA Senior + 2 player with NHL games played. Nikita Zaitsev isn’t in because he’s a sure-fire NHLer this year, but that doesn’t apply to the Big 3, where Marner is the only one who even has a slight question attached to him.
- Why does the pyramid have exactly five tiers? There’s no explanation as to what each tier was, just that they exist. I presume Steve starts a new group whenever he feels there’s a pocket that is noticeably better and cuts to “the field” after 25 players. My end result involved 12 tiers (I didn’t cut off at “the field”) and by player 25 I was in Tier 9. Where’s the line?
- Most importantly; even if everyone uses the same model, your results are going to be much different.
For the sake of making this argument, I scrapped my own chart and built a pyramid in Steve’s 1-2-4-9-9-Field shape (nobody said drawing triangles was easy), and stacked it next to his. Here’s what it looks like:
Like I said, I personally don’t feel that my pyramid accurately represents the tiers of the organization; I had to break up or combine certain “Jeff tiers” into pieces to properly fill the “Steve Tiers”. But what’s more important here is that an argument-suppressing format still has a lot of differences.
- Steve has Kasperi Kapanen in Tier 3. I have him in Tier 4.
- Steve has Brendan Leipsic in Tier 4. I have him in Tier 3.
- Steve has Tobias Lindberg in Tier 4. I have him in “the field”
- Steve has Zach Hyman in Tier 4. I had him in Tier 5.
- Steve has Yegor Korshkov in Tier 5. I had him in Tier 4.
- Steve has Adam Brooks in Tier 5. I had him in Tier 3.
- Steve has JD Greenway and JJ Piccinich in Tier 5. I have them in “the field”.
- I have Pierre Engvall, Vladimir Bobylev, and Jack Walker in Tier 5. Steve has them in “the field”.
Of 25 players on our pyramids, that’s 11 that we have in distinctly different tiers. Possibly more; I didn’t include Trevor Moore because I wasn’t sure if Steve had him in mind while making the rankings; if he left him out with intent, I’d knock Nielsen down to Tier 5 and Loov into the field, making 14 positions of disagreement, or over half the list. We’d also disagree on how to formulate the list, and who is on the list.
Again, the intentions of the format are good; but rather than solve problems, it just creates new ones.
Where a Pyramid is useful
I don’t think that a tiered system is without merit, to be fair. The most practical use for it, from a writer’s standpoint, is a single-feature, casual look at the organization where you aren’t looking to say a heck of a lot. A situation where you want to acknowledge that you’re aware that these guys exist, and are generally aware of what they do with their lives, but you can’t commit to outlining every fine deal.
So it’s no shock that Steve is the one who is sold on the idea. That’s not at all a knock on him; it’s just something that best fits his own format of easy-to-consume content that doesn’t stretch into longer projects (does a full season of LFRs count as a project?). He has to get a point across in ten minutes or less (the associated video, coincidentally, is 10 seconds shorter than Frank Ocean’s ambitious concept track ‘Pyramids’), spend a bunch of time editing it, publish it, and switch to a new subject for the next video (or work on one of his eighty-bazillion other things). Our prospect rankings, on the other hand, are a month long project spanning eight writers and 50,000 words. This gets the general point across quicker, and while it’s not as thorough, for many it’s worth the opportunity cost of getting the basics covered.
Why rankings reign supreme
As Steve admits very early in the video, there’s a lot to be gained from the time investment of reading through list projects like ours and Pension Plan Puppets’ Top 25 under 25 (which still has four more spots to go; I’m betting on Steven Stamkos or Anatoly Golyshev at #1). When you break down rankings into a couple of dozen posts, you absorb not just an opinion of where a player stands, but information on what they’re about.
Our posts on TLN, for example, included everything from contextualization of box score stats, to projection analytics, to video, to back stories, to a look into what the player’s individual future holds. You can’t go into all of that detail for every prospect in one shot without having the info go out of your brain as fast as it goes in; it’s the same reason none of us remembers much about the books we did all-nighter reports for in high school. You either get bored or look away or get lost in a train of words.
But as long as you consume them at your own pace, these profiles, whether you agree with the order or not, give even the most casual level fan an understanding of players they’ll see in the NHL in a few years. That’s the real end game here; the rankings are mostly to set the order (and a little bit for hindsight bragging rights).
I also believe that once you get a big enough crowd of people, a consensus ranking will sort itself out reasonably well. We had eight writers with drastically different approaches to grading the system and much different opinions, but there were very few “50/50” moments in the final rankings once we put everything together. The same goes for the fan vote; we had nearly 400 responses, and next to no margin-of-error decisions.
The levels of variance are hurdles at an individual level; especially if you’re the one who has to decide the rankings. But that just acts as encouragement to scrub through more data, watch more video, and talk to more people, and learn more about the players so you can pass those thoughts on to others. I had players who took huge leaps and falls from my first draft to my final decision, and that came simply from doing due diligence on them. Do I think I’m going to bat 1.000 on my list? Probably not; great players can be ruined by injuries, underwhelming kids can take huge steps with things as simple as a skating pattern change. But the decision-making process can be as educating as it is predictive, which is why I feel that going one-by-one, while the most difficult approach, is the best option for both writers and readers.
Conclusion
I feel like Steve set out to solve a problem with his Pyramid and managed to solve a different one instead. If the goal was to get rid of the “this guy should be placed here” arguments, which is what the text and the video implies, it completely and totally misses the mark based on the level of variance that there is in people’s player evaluation, and the fact that there are no set rules to the pyramid.
The problem it solves, though, is creating an easy to consume baseline for an organizational depth chart. We used to have HockeysFuture’s upside number, probability letter system and table presentation for that, but with their site no longer being updated for present and future draft classes, a tiered system is something non-proprietary that hockey journalism platforms can easily adopt to give a first glance at an organization. It gives you a general idea of who to be excited for and who to not worry about for now without having to comb through a ton of information.
Think of it like the Pharoahs; the best way to learn about them is to read encyclopedias and go to museums, but for those who just want a reminder of their place in the world, you head to Egypt and look at, well…
Recent articles from Jeff Veillette
