Canadiens continue to show hockey's analytics problem (Trending Topics)

On Wednesday night, it came out the the Montreal Canadiens had parted ways with the analytics expert they’d hired at the start of last season.

Matt Pfeffer, who has worked with NHL teams in the past, is well-regarded in the hockey numbers community, and the decision not to renew his contract was therefore thought of as another bad move by a team that has recently exhibited a propensity toward them.

Then Sportsnet’s Eric Engels reported as part of his ongoing coverage of the baffling P.K. Subban trade that there might be a connection behind Pfeffer’s “vehemence” on not trading Subban (he reportedly prepared an “elaborate presentation” on why the swap was a bad idea from Montreal’s perspective) and the team’s decision not to bring him back for another season.

However, the Habs are also quick to note they’re still in the analytics game, and currently evaluating their options for next season.

Of course, one has to wonder what kind of investment they’re going to make in this regard, because if a whip-smart guy like Pfeffer basically gets told not to bother coming back because he disagreed with management’s decision on a franchise-altering transaction. So basically what they’re looking for is a few yes-men who will tell them what they’re doing has them on the right track.

It’s bizarre, frankly. There is a tendency in the analytics community to look down on the Canadiens and Michel Therrien in particular, but the Pfeffer hire and the on-ice performance last season is kind of indicative that this is, or perhaps was, a franchise making good decisions. For all the stick Therrien got — and his handling of Subban is indeed worthy of derision — there’s no doubt that what Montreal did last season probably would have had them in the conversation for one of the better teams in the league if not for that whole “injured All-World goalie” thing.

Therrien was hired in 2012-13, and truly earned his reputation as a behind-the-times coach who made a lot of the wrong decisions when it came to helping his team succeed. He favored the wrong players, didn’t have good systems in place, etc. Honestly, just ballpark their adjusted possession numbers from before and after the start of 2015-16: in the three 82-game seasons since he took over, Montreal went from approximately 47.7 percent to 48.9 percent to 51 percent. (I’m ignoring the 48-game lockout season, because Montreal played a short schedule against nothing but the Eastern Conference, but maybe it’s worth nothing Montreal was nearly 54 percent in adjusted possession over that campaign.)

Now, 51 percent isn’t great or anything — Montreal finished the regular season in 13th last season — but it’s better than what they’d been doing and probably indicates an improving process. Subban was, of course, a big part of that, but the entire lineup got better on a season-long basis.

While Subban was playing by far the best competition on the team and therefore suffered in some regards, it’s worth noting that his adjusted relative possession tumbled from plus-5.2 in 2014-15 to minus-0.4 last season, and it was mostly because the rest of the team caught up to him; his on-ice CF% actually dipped just half a percentage point.

One might be able to safely conclude they identified a serious problem that was masked only by elite-level goaltending and, with Pfeffer and a more top-down view of implementing an analytical approach, improved more than marginally, in some respects.

But one might also be able to conclude that they blew that up because of one season in which they materially improved their process but missed the playoffs because Carey Price went out for the year in November. It’s also worth noting that the Canadiens started well but declined a lot as the end of the season approached.

This is some mystical thinking that seems in many ways to be divorced from reality. The team made actual on-ice strides, but because their goaltending cratered (to .919 at 5-on-5 from the previous year’s .935) they lost a whole lot more games. Anyone could reasonably tell you that the Canadiens had some decent talent on the roster, but the reason they’d had so much success in recent years should be directly attributed mostly to Price being incredible.

It is unfortunate that while they were taking judicious steps to ensure they weren’t totally reliant on Price for most of their wins, a freak serious injury set them back in terms of perceived success. “We did all this and finished 28 points lower in the standings!” seems to be a fair evaluation of their thinking here.

What often gets ignored about the point of analytics, as far as professional sports franchises themselves ought to be concerned, is that they help you identify value. That is to say that they can or at least should be used in a way that maximizes the value of every dollar spent while still maintaining or improving your on-ice/field/court product.

That’s especially important when it comes to leagues like the NHL, where there’s a hard and fast salary cap to which teams must adhere, and therefore should also plan how they’re spending both now and in the future.

For this reason, it’s important to note that what the Montreal Canadiens have done in the past few weeks probably doesn’t hurt them too badly in the near-term, if we’re being totally honest. Say what you want about the Andrew Shaw contract (too much money, too many years), but he’s a player that does in fact help you on the ice. Adding Al Montoya as a backup to Price is a huge upgrade. And even Shea Weber — though no Subban himself — will at least provide somewhat comparable service for a few more years before the contract descends into chaos around his 34th birthday.

This is the Canadiens, though, kind of lucking into getting good players in their pursuit of “culture” (as discussed at some length in the PDPR earlier this week) and probably being harder to play against. Shaw is, like several ex-Chicago players before him, an eminently useful player who just happened to be a contributor to multiple Stanley Cups. He’s not Andrew Ladd and he’s not Dustin Byfuglien, but he can play. Add in the “culture” aspect and you can see very easily why Montreal overpaid for his services. That at least costs Montreal very little in terms of assets and commitment.

Acquiring Weber — the Olympic-gold-medal-haver — obviously comes with the commitment until he’s in his 40s, and at a price of one of the most dynamic defensemen alive. It saves the Habs a marginal amount of money against the cap, but from where they sit, they get the better of the deal both because Weber is still viewed as the player he was in 2012, and because he Hates To Lose.

As was pointed out time and again when that Engels story about Pfeffer going in hard for keeping Subban, the Canadiens are not alone here. Hell, some teams are proud enough of these decisions to televise them. We’ve all seen the horror — fraught with dramatic irony — of Bruins executives justifying a Tyler Seguin trade to themselves on Behind the B.

While the whole video is filled with weirdo ideas, the key exchange here is this:

Director of Player Personnel Scott Bradley (since promoted to assistant general manager): “He’s not a physical player, he relies on his skill.”

GM Peter Chiarelli (currently GM of another team): “Does that sound familiar?” [They’re taking a shot at Phil Kessel here, obviously.]

Bradley: “Yeah it does. He’s a star player. There’s no doubt. But does he fit with our culture?”

That clip should end with Werner Herzog handing the tape to Chiarelli’s family and saying, “You must never listen to this.” It’s that bad.

Then there’s this from the Flyers after the 2013 draft, in which they drafted Samuel Morin based on little more than his size (he’s 6-foot-6) and nastiness (117 penalty minutes in 46 games for Rimouski in his draft year). Certainly, they did not select him for his offensive capabilities, because at that time he had 4-20-24 in 108 regular-season games in the QMJHL.

The fine folks at Broad Street Hockey looked at all the evidence in that video and put together a list of who probably fell where on the Flyers’ final prospect rankings. They had Morin sixth, ahead of guys who have been in the NHL for at least a year, like Darnell Nurse, Sean Monahan, Max Domi, and so on. Morin was projected to go somewhere in the mid-20s by most estimates.

“I still like the hardness on him,” one Flyers official said in the pre-draft meeting. “His natural ability to be hard and aggressive. I like that, and that’s what I like most about Morin.”

“I disagree,” answered another. “I think Morin knows his game better than Nurse does. I think Morin knows what he’s gonna be, I think Nurse is gonna get it. He might. … You’re reaching but he’s a 6’8” guy. Where do you find these guys? … I think Morin has more meanness. He’ll whack you with his [expletive]ing stick and give you that stuff.

Again, woof. Especially because they acknowledge in the video that this is viewed on a max of maybe 13-15 viewings of a player over the course of a season. Will it surprise you at all to learn that Morin is one of only two players taken in the first round of that draft to never appear in an NHL game? He also put up just 18 points in 76 games in his first season in the AHL. And granted, he just completed his draft-year-plus-3, and he’s only 21, but most of his peers seem to have passed him by. It’s safe to say that the big ol’ reach the Flyers took because of intangibles didn’t pan out for tangible reasons.

The Canadiens obviously had a lot more time to evaluate Subban and Weber, plus stats and video and all that kind of stuff as well. But you can rest assured, if there were cameras in the room in Montreal for the Subban trade decision, you’d hear a lot of talk along similar lines.

But we’re over and over and over again that toughness is a relatively unimportant attribute. Taking penalties makes you far more likely to concede goals, and throwing hits just means you didn’t have the puck. In that regard, hits are actually less valuable than blocked shots. Blocks at least have some amount of tangible benefit; they prevent the puck from getting on net. What does the average hit accomplish besides tacking an extra $2,500 onto Cal Clutterbuck’s next contract demand?

Both hits and shot blocks aren’t actually an indicator of anything other than not-having the puck. Nick Lidstrom is a freak of nature, obviously, but the fact that from 2002-12 he only blocked 373 shots at 5-on-5 in 697 games is telling. He was also credited with just 212 hits in 388 games from 2007-12. There are few players who can get away with those kinds of numbers, but shouldn’t we be modeling our ideal players after guys like that?

Because guess what: there’s little overlap between things like hitting a lot or blocking a lot of shots and winning meaningful hockey games. In discussing what made the Pittsburgh Penguins so good on their quest to win a Cup this year, Pensburgh found that, like almost every other Cup winner since 2007, they didn’t throw a lot of checks. The only teams that finished above-average in terms of hits-for — among all clubs with at least 14 playoff games played per season in that time — were the two Kings Cup winners. And we can be certain that the hitting wasn’t the reason why. And if you look at the chart, you’ll see that even that famously nails-tough Bruins club from 2011 actually didn’t throw many checks at all. Obviously there are other ways to be “tough” besides hitting, but it’s still strange, then, that they pursued toughness to the ultimate downfall as an elite team.

And look, everyone inherently understands that this kind of thinking is never going away. Even the Smart Florida Panthers buy into the ideas of leadership as an issue for them to consider when making personnel decisions:

One supposes that the idea behind effective team management in the modern NHL, then, is to acknowledge intangibles as existing, but also acknowledge their likely small impact on outcomes. Qualities such as being well-liked by teammates, leadership, fitting into a culture, and “hating to lose” are good to have.

But if they’re the things that drive your decision-making process, you’re going to end up making a lot of bad choices, and in the long run that makes your team worse. Good luck to the Canadiens finding analytics guys who will tell them otherwise.

Ryan Lambert is a Puck Daddy columnist. His email is here and his Twitter is here.

All stats via Corsica unless otherwise stated.