clock menu more-arrow no yes

Filed under:

Could Advanced Stats Be The Key to Success in the NHL

New, comments
"Shoot it THAT way!"
"Shoot it THAT way!"
Brace Hemmelgarn-USA TODAY Sports

If you follow "Hockey Twitter," which is just a community of hockey fans who talk about hockey... On Twitter.. You'll have heard a lot about "fancy stats.."
The hockey stats community has vastly grown over the past two years, constantly innovating ways to evaluate players, and how they are playing, and how they will fare down the road. At the center of the universe, sits the wonderful website ExtraSkater. The mecca for all interested in advanced analysis of their favorite team and players. Among the Corsi, Fenwick, PDO, and Zone-Adjusted stats, there are endless options to analyzing the game. Advanced stats have proven one thing: Possession is everything in this game. When you have the puck, good things are happening more often than not. We've even seen some teams openly speak of finally investing in stats analysts watching games, tracking zone entries, gathering data, and meticulously adapting it into the teams' philosophies and everyday systems. Some teams that have hinted towards it are Tampa Bay, Los Angeles, and Chicago. Three teams who, wouldn't you have guessed, are pretty good as of late. Sure, having good players helps, but having stats to show you where a player is excelling is the key to taking it to the next level, I think.

Not long ago, I think some of us can remember Ken Holland stating that his philosophy was going with "gut analytics." I'll admit, when I heard him say that, I wanted to throw-up all over myself and lay face down on my floor. There is nothing I crave more than a team who thinks progressively, applies new ideas and tries new things. What does Corsi provide you? It provides raw data for a player's shot attempts. It takes out zone-relative data, goalie save percentage, shooting percentage, and shows you if he is getting outshot or is the one doing the majority of the outshooting. Players like Henrik Zetterberg, are possession monsters who anchor the puck, and create shot attempts, where players like Luke Glendening are usually doing the exact opposite.

Earlier this month, Ken Campbell of The Hockey News wrote a piece about NHL teams and how they are beginning to warm up to the untapped world of statistics. The article, which can be found here, mainly cites two NHL general managers. New Jersey's Lou Lamoriello, and our very own Ken Holland. What Kenny said, wasn't TOO much of a surprise. He seems a bit reluctant, but he's obviously starting to entertain the option of investing in it:

"We’ve been talking about it, but I wouldn’t say it’s a big factor in any of our decision making," Holland said. "Let’s say you’re Pavel Datsyuk’s linemate. You move to another team and not playing with Pavel Datsyuk is going to have an effect on your lack of success. With baseball, it’s more black and white because the pitcher is on the mound and he’s going against the batter. But in hockey, you’ve got four teammates and five opponents who are going to have some impact on what’s going to happen."

So, they've been talking about it.. That's enough to make me excited over it. But speaking to somebody inside the organization, I learned Ken Holland directly told players at this year's Development Camp that they would begin implementing "shot attempt" analysis into their evaluation. Shot attempts = Corsi/Fenwick. Period. And yes, I trust this source.

Detroit has always been a "possession" team. Get the puck, enter the zone, setup, and shoot the puck. This is bare-bones advanced stats stuff, without the statistical jargon attached to it. It's there... The more I think about it, Ryan Martin (Assistant GM) has always been a numbers guy. He deals a lot with the salary cap formulaic effects. He also helps with player analysis at both NHL and AHL levels.

So am I jumping on the wagon too quickly thinking that Detroit is about to start thinking progressively? Maybe.. But then NHL writer Dan Rosen released a few quotes from Mike Babcock about statistics:




Is it happening? Are the gut analyses on their way out? We aren't too sure, but as a fan of analyzing players using advanced statistical data, I sure hope so. It shows signs of forward-looking enlightenment for a team who is in need of a new approach to player analysis. With advanced statistics, you can see that players like Luke Glendening are badly overplayed (even though it should be obvious by the common eye-test), and players like Riley Sheahan are turning to prodigious possession powerhouses. Is this it? Is it time for the advanced stat revolution in Detroit? Don't know.. But boy oh boy, do I sure hope so.