This spring, right after our season ended, I was staring at a recruiting list full of players from Europe and Canada, stuck on a question that sounds simple and isn't: does a 2008 from Norway playing meaningful games in a low-level Latvian pro league actually matter? Does any of that translate to my league, my level, my team?

I had tape. I had stat lines. And look — I trust my eyes. I know what I see in a player. But I was hungry for more. The variables overseas are brutal. Different ice, different pace, leagues that don't map onto anything over here. And these aren't fantasy hockey picks. When I recruit a kid from overseas, I'm asking a teenager and his family to take my word for it and cross an ocean. If I'm wrong, I didn't just make a roster mistake — I set a kid up to fail in a foreign country. I wasn't doing that on a hunch.

So I started asking myself questions.

The Questions Built the System

I never sat down to build anything. I sat down with one question: what's a point in one league actually worth in mine?

That question led to another one. If I can't compare leagues directly, can I find players who moved between them and see what actually happened? Then another. Does a 17-year-old doing this mean more than a 19-year-old doing the same thing? And another. Is this kid producing because he's a driver, or is he a passenger on a stacked team? Or flip it — is he better than his numbers because he's buried on a bad one?

Then one more that changed how I recruit entirely: what does a kid's first junior season tell me about his third? If I like a 2009 right now, knowing what he is today isn't enough. I need a realistic picture of what he becomes. So the system doesn't just evaluate players — it forecasts them. Development curves. Projections two, three years out. Because I'm not trying to be right about every player today. I'm trying to hit on more than I miss over time. That's the whole game.

Every answer opened two more questions. I chased them for about three weeks — ten hours a day, most days. What came out the other side was an evaluation and forecasting system built on real player data. How production actually translates when players change leagues, adjusted for age and for all the context a raw stat line hides.

I'm not getting into the mechanics. The principle is what matters: every number in it comes from what real players actually did. Not what I assumed they'd do.

The Moment It Earned My Trust

Early on, the system spit out something that looked broken. It rated a kid as elite — a kid who had 15 points all season. My gut said the model's wrong. Fifteen points is fifteen points.

But instead of overriding it, I dug into why it saw what it saw. And the deeper I went, the more I realized it was catching something real that the stat line buried. Chasing that answer exposed a blind spot in how I was reading players — and fixing it made the whole system sharper.

That's when it stopped being a spreadsheet and became a tool. Not because it agreed with me. Because it disagreed with me and had receipts.

The Showcase Problem

Here's where this actually shows up in my life.

Walk into any showcase. Sixty players on the ice. Nobody gets a real read on sixty players in a weekend — it's not humanly possible. So what do most of us do? We watch whoever pops. And here's the problem with that: the guy taking over a showcase might just be the best player in a wildly mixed talent pool. Does that mean anything when the competition turns up? Sometimes. A lot of times it doesn't. That's noise. Showcases are full of it.

There's a saying in the analytics world — the numbers give you the base, your eyes verify. That's exactly how I use this thing.

Now I walk in with a targeted list. Instead of skimming sixty guys, I go deep on twenty the data says can actually play at my level. If a kid is the best player at camp and a legit projection analytically? Great — my eyes confirm it. And now they're free to watch everything the numbers can't touch. His habits. His body language after a bad shift. How he treats his linemates. What he does when he's matched up against the one other real player out there.

And I don't burn a weekend getting seduced by a noise guy who was never translating anyway.

The data doesn't replace my evaluation. It aims it.

What I Actually Believe

I'm not an analytics guy who found hockey. I'm a hockey guy who got tired of guessing at questions that had findable answers.

Am I still going to miss on players? Of course. Everyone does. But there's a difference between missing because hockey is unpredictable and missing because I walked into a rink unprepared.

The best thing I built this offseason wasn't the system. It was the habit of asking, on every single evaluation: what do I actually know here — and what am I just assuming?

Every coach at every level can ask that question. Most of us don't ask it enough.