​I’m something of a traditionalist when it comes to goaltending, probably a byproduct of watching the greatest to ever do it backstop the Devils more times than I can count over two decades. Add in the revolving door we’ve had in the crease the last five years, and yes, I’ve kept my therapist busy. When I wrote about a possible Markström trade a few months back, I didn’t think it was likely—and if it happened, I assumed we wouldn’t get real value back—but here we are.
Behind the scenes, in every sport, the clash between long-held methods and a newer wave of thinking has drummed on for a century. That newer approach often comes with better video, chip tracking technology, and analytics, and evolution tends to move along a gradual curve as old guard resistance erodes. It takes time, though. Baseball shifted gradually from a four-man rotation to five across the 1960s through the 1980s (give or take a decade). Football coaches now go for it on fourth downs more often than not, and basketball has transformed into a three-point shootout. Nowadays, several baseball teams deploy “openers” in a committee approach.
Hockey isn’t immune. With analytics and data taking a louder seat at the table, teams have evolved in how they attack and score, how they defend, how they select players, how they construct rosters, and how they keep players healthy and at their peak. Carolina’s crew has even started to explore 3D spatial relationships to optimize offensive-zone presence. This new thinking is finally breaking through after years of resistance from the Hockey Men who have long held sway. Sunny’s background is well documented, and it seems he’s trying to ride Carolina’s success to rethink goaltending from the ground up. I should caveat this by saying I’m a layperson trying to interpret what Sunny—a sharp mind—might be doing, and I want to credit JP Gambatese’s goalie blog for sparking my line of thinking here.
I’ve used this chart before—the relationship between xGA and GA at 5-on-5. When I first tinkered with the axes, I was struck by how linear the correlation is between those two metrics. In plain terms: the better you are in front of your goalie, the better your team tends to be. Not exactly earth-shattering, but there’s more to it. If you look at the standard deviation (0.21 goals against per 60 at 5v5) of actual goals against versus expected goals against, you’ll find that only a handful of teams sit more than 1.5 standard deviations from the mean, and only a few more hover around 1.0 standard deviations in the positive direction—meaning they gave up less than expected.
So, Tim, what am I getting at here? My take is that you shouldn’t throw money at goaltenders unless you’ve got a true stud in net. And honestly, there aren’t many stud goalies in the league these days who justify the big-ticket price. The analytics suggest a cautious approach: invest in goalies only if you’ve identified a rare, game-changing talent.