Bill Meltzer: Meltzer’s Musings: Tradition and Adaptation


On November 16, 1997, I covered an American Hockey League game at the Philadelphia Spectrum. It wasn’t anything that happened in the Philadelphia Phantoms’ 6-3 win over the Kentucky Thoroughblades that made it memorable and special to me.

Rather, it was the first professional hockey game I was ever credentialed to cover in the pressbox. Of all the assorted press passes I have kept since then — including the ones from my first credentialed Flyers games during the 1999-2000 regular season and playoffs, my first full-season Flyers press pass, first road-game pass as visiting media, IIHF World Junior championships in Sweden in 2006-07, my Stanley Cup Finals credentials from 2010, ones from the NHL Draft, etc.

— that first one is still the most special one to me in some ways.

At my first game in the Phantoms’ pressbox, Bruce “Scoop” Cooper gave me a piece of advice (which he has repeated frequently over the years): If you really want to know what’s going on around this game, talk to the scouts.

Back then, I was too reticent to bother anyone. Nowadays, one of the things I enjoy the most before games, between periods or via electronic communications is asking scouts their take on a team, a player or a game. Sometimes you can learn a lot simply being a fly on the wall as they talk shop amongst themselves.

It was not until rather recently that I ever heard NHL pro scouts even make a passing mention of the existence of advanced stats/analytics. Most NHL scouts tend to be traditionalists, and the ones I speak with the most frequently were varying degrees of dismissive. Several flat out said there was nothing useful to be gained from early-generation of analytics that still get cited the most frequently (Corsi, Fenwick, etc) that they couldn’t already observe with their own two eyes.

“As I understand it, the Corsi thing is just a different type of plus-minus,” said one scout, picking at his food in the Wells Fargo Center commissary before a game. “Well, plus-minus is a pretty flawed stat, anyway. It’s a team thing, not an individual thing. The difference is Corsi is related to shots and not just goals. To me, that’s pretty flawed, too.”

Another scout said to me, “Here’s the problem. Hockey isn’t math. In math, if A is greater than B and B is greater than C, then A is always greater than C. Matchups in hockey don’t always work that way. Entire playoff series can turn on emotion, hot goaltending and leadership. Those are things that can’t be measured on any chart.”

Over the last year or so, some of the resistance to analytics has lessened. Hockey people have learned there are ways to add context to data. In the last year, I have heard more and more scouts say there is useful information about how a player is used and how well he performs that role. However, the majority of NHL scouts with whom I communicate still maintain they do not and will never primarily assess a player’s value to his team based on advanced stats.

For example, this is why the “big and physical defensive defensemen” still has a role in the NHL game. NHL clubs still want a big-bodied defender or two in their lineups who excel at non-puck related duties such as applying muscle in front of his own net, pinning opponents to the boards, and blocking shots. On the flip side, we are also seeing the demand for that role starting to lessen just a bit — or least the puck-handling competency demands increase in addition to performing the traditional parts of the job.

What I have found is that isn’t really so much the analytics per se that traditionalists within the game object to. As a matter of fact, more and more are beginning to see that analytics and traditional observation often support the same conclusions.

If you have not already done so, I strongly recommend reading the outstanding article penned by longtime NHL player Jason Strudwick on the changing views on analytics among those within the game. It has gone from analytics never even being discussed around teams to discussions of finding the proper context and fit for their use.

One point that Strudwick raised is that the analytics advocates weren’t the only ones who thought the Toronto Maple Leafs were in danger of a collapse down the stretch last season. Many hockey traditionalists fell the same way but they came to the same predictions by somewhat different angels.

I can vouch for the truth of what Strudwick wrote. Shortly after the Olympic break last year, I asked two NHL pro scouts which team that was currently in a playoff position would be the most likely to fall out by the end of the year. Both of them immediately said Toronto.

“Who is a real leader on that team?” the scout said. “Who is the tone setter when the checking gets tight? Who is going to lift them when they have to chase a game? They don’t have that leadership in the room to win when it counts. They don’t play good team D, so it’s all on the goalie. Combine those things and they’re going to go down because of it.”

Never once did either scout refer to the Leafs’ negative puck-possession differential making them vulnerable. However, in the bigger picture, they were really talking about many of the same things that the advanced stats metrics pointed to a mathematical way. Strudwick’s article cogently explains how the two “warring” sides were really on the same page all along, and it was easy to realize that once you stripped away the noise of people talking at (instead of with) one another.

Strudwick’s views are shared by an ever-widening community within the game, including Flyers general manager Ron Hextall. They still use observation and years of first-hand experience as their main baseline for assessing the game and drawing conclusions, but analytics are now a piece of the puzzle for testing and adjusting their observations.

More and more, people in hockey are realizing that it is not analytics per se that are turn off. Rather it’s the way some of the proponents deliver the message. You can call it the Dubas/Dellow Dichotomy.

Recent Toronto Maple hire Kyle Dubas has always framed his stance on analytics in a positive and respectful manner: basically a message of “this is how we can all do our jobs a little better.” He also worked within the game successfully at the major junior level before graduating to the NHL. I have had private communications with established NHL scouts who have said verbatim that “the kid deserves a chance to see if he can help” in Toronto.

Conversely, recent Edmonton Oilers hire Tyler Dellow — while undeniably ground-breaking in some of the ways he applied metrics — took a much harsher, caustic and negative approach. His framework was one of saying “you are incompetent at your jobs and I am smarter than you are.”

Using that approach gained him notoriety and worked well as an outsider but also made a lot of enemies who want to see him fail now that he’s got an opportunity to actually work for an NHL team. He is now faced with a challenge of not only being good at what he does but also convincing others to get on the same page.

Wrote one longtime pro scout by email, “You have to give respect to get respect. [Dellow] is going to learn that real quick. He’s got an uphill battle coming in to prove himself.”

Hockey people often talk about the sport being one that is every bit as much about adjustments as strategy, perhaps even more so. What we are seeing now within hockey is a process of adjustments being made. Those — on both sides — who adjust will stay within the game and thrive. Those who are too rigid and refuse to adapt will fail.

This post originally appeared on and we thank them for permission to rebroadcast it here.

Your sports. Your teams. The ISN Daily Digest.

Sign up to the ISN Daily Digest and sit back while we pick the previous day’s best headlines and speed them straight to your inbox every morning.
Email address
First Name*
We abide by all applicable emailing laws including 100% CAN-SPAM/CASL/US CAN-SPAM Act compliance. No spam!