Latest posts by Phil Kocher (see all)
- Tony DeAngelo Will Miss 3-to-4 Weeks with Sprained Ankle - 03/14/2018
- Quick Hits - Rome Wasn't Built in a Day - 03/10/2018
- Quick Hits: Rangers Suddenly Deep at C, but D May Kill Hank - 03/03/2018
While making his rounds on a number of hockey podcasts this week in the wake of the surprising firing of now ex-head coach Gerard Gallant by the Florida Panthers, Sportsnet’s Jeff Marek—one-half of the wonderful Marek vs. Wyshynski podcast that you should absolutely be listening to if you aren’t already—whimsically suggested that those arguing in favor of hockey analytics start substituting the word “analytics” for “facts”.
For example, when responding to the popular “I don’t care for/believe in analytics” rebuke from traditional/conservative-thinking hockey people, simply swap “analytics” for “facts”. Their rebuke can be directly countered by the question “so, you don’t believe in facts?”.
It’s facetious (and mildly insulting) by design, but surprisingly enough all you’d need to do is spend a fair amount of time listening to Canadian media coverage or cruise around hockey Twitter on any given game night to cross paths with a proverbial army of people who think and speak in this manner.
Not me and not anyone who understands both. It's the analytics crew that want to live and die with numbers only. https://t.co/Jmqp2ZCyQC
— Patrick O'Sullivan (@realPOSULLIVAN) July 18, 2016
In many ways, the firing of Gallant has reignited the false dichotomy that there are only two schools of thought to analyzing hockey—analytics or eyeballs. There aren’t. The two are not diametrically opposed. They’re actually two sides to the same coin, but to this point, as former Canadiens analytics consultant Matt Pfeffer said after Montréal walked away from his services back in June of 2016:
“Analytics hasn’t really reached maturity in the NHL yet,” Pfeffer said. “Teams get a lot of different solutions and offers from people and companies and they just don’t know what’s what and they tend to lean conservative in those instances.”
That lack of maturity extends well beyond individual NHL teams through to individual fans, which is why the Panthers’ decision is now annoyingly acting as a referendum on the value of analytics in the NHL. Here we go again.
To circle back, Marek isn’t wrong. His suggestion was somewhat tongue-in-cheek and is more sharp humor than a realistic proposal, but hockey analytics are facts. That’s all they are. When you read that Player X had a Corsi-for percentage of 51.1, or Player Y owns a 46.9 Fenwick-for percentage, or that Team Z is playing with a 103.9 PDO, all you are actually reading are facts. It’s all statistical data that is recorded and reported on in the same way traditional metrics are. To what degree you (or an NHL team) allow those numbers to influence and shape your opinions on the game or on specific players is a secondary, often contentious matter, but the facts themselves never stop being facts regardless of personal beliefs. They’re objective, verifiable scientific observations that exist independent of whether you believe in them or not. As famed astrophysicist and director of New York City’s Hayden Planetarium Neil deGrasse Tyson often quips, “the good thing about science is that it’s true whether or not you believe in it”. So is/are analytics. They, too, are science, which is why this tiring black-or-white thinking is so frustrating.
As a quick aside, I’m well aware of how intimidating analytics can be, especially for first-timers or anyone who is genuinely not comfortable with math. That’s coming from someone who is genuinely not comfortable with math and who was wholly intimidated by them, too. It just takes a little time and effort to understand, however, in the same ways you’d learn any new technology. I don’t want to take up a huge chunk of this article explaining how these statistics are determined, but I also don’t want to alienate anyone who doesn’t actually understand them, either. My suggestion, if you’re one of these readers, is to take a few minutes to check out Sam Page’s primer piece he wrote for SI back in 2014, as it does a wonderful job of explaining the basics. Then come back here and read the rest of this. It’ll be happily waiting for you.
I believe in analytics. I think much value can be gleaned from understanding and accounting for things like Corsi, Fenwick, and PDO when judging player and team on-ice performance. I think that when these things are accounted for properly, in conjunction with traditional metrics and the eyeball test, they can be wonderfully predictive. I also believe that the Rangers don’t value them nearly enough. That’s not to say they don’t value them at all, but a greater emphasis should be placed on them if the front office has any interest in acting in accordance with the undeniable trends shared by recent Stanley Cup winners. Trends like not often being on the negative side of shot attempts margins, prioritizing mobile, puck-moving defensemen over poor-skating brutes, and embracing offensively-oriented depth forwards over far less productive grinders whose best attributes (like hitting or fighting) come while playing without the puck. The Rangers have embraced the latter this season, but neither of the former very well. Both can be aptly illustrated by the usage this season of 24-year old Niagara-native, Adam Clendening.
Despite his young age, Clendening is already something of a journeyman in the NHL. The Rangers, who signed him to a one-year/$600,000 contract on July 1st, are his sixth team in three seasons since he was originally drafted by the Chicago Blackhawks in the second round, 36th overall, in 2011. The Blackhawks traded him to the Vancouver Canucks in January of 2015 for Swedish defenseman Gustav Forsling. Six months later the Canucks would deal him to the Pittsburgh Penguins as part of the Nick Bonino/Brandon Sutter trade. After playing just nine games, the Penguins would flip him again to the Anaheim Ducks in a larger trade for Carl Hagelin. The Ducks, who he never actually played for, shortly placed him on waivers where he was claimed by the Edmonton Oilers who he played 20 games for last season. That lead to this past summer where he entered free agency for the first time in his career and decided to spend some time on Broadway.
Clendening’s rocky path in the NHL thus far has limited him to just 56 games total in which he’s scored a total of 12 points. On the surface, if you accounted for nothing else, it would be easy to see why teams haven’t been impressed with him enough to keep him around for more than a cup of coffee before jettisoning him off to the next stop on his Mike Sillinger-like path. But there is more beneath the surface once analytics are factored into the equation. It’s here where a keen and proactive front office (not to mention head coach) could plausibly reap untapped rewards should they put Clendening in a position to capitalize on what that data seems to indicate could be there.
Among all Rangers defensemen this season, according to Corsica.hockey, Clendening leads the pack by a wide margin in numerous analytics categories including Corsi-for percentage (63.43), Corsi-for/60 (62.24), Fenwick-for percentage (63.04), Fenwick-for/60 (42.47), Goals-for percentage (83.33), and Goals-for/60 (3.66). In fact, though only playing in seven games thus far, he leads all defenseman in the league with a 63.43 CF% and is second league-wide with a 63.04 FF%. While both this season’s games played and his career games played are, by definition, a small sample size, the indication within that sample is promising. The kid has offensive game. Surely there are limitations to what he can provide at the NHL level, but as the data indicates, the potential for offense is not among them. The reason that offense hasn’t translated to more easily accepted production yet is probably due to the fact he has not been given a consistent length of time to prove it can. Instead, he’s been injected into the line-up after long stretches of not playing, only to be pulled back out after a single game. As if that type of treatment is fair and sufficient in giving a player in his position an opportunity to produce and impress his coach.
The Rangers offense may be league-leading, but that offense is mostly riding a very high PDO. They lead the league with a 103.81 PDO rate with Chicago (103.06), Washington (102.73), and Minnesota (102.40) each following suit respectively. Joe Fortunato of BlueshirtBanter spoke on much of this recently:
Here’s the Rangers rolling PDO average inclusive of all of last year and this year.
There is an argument to be made that the Rangers with this offense will always shoot hot because of their speed/odd man rushes, coupled with Henrik Lunqvist’s elite status that literally spells “higher than average PDO.” The timing of this year’s jump, however, says a lot.
The PDO jumping in November as the Rangers’ possession was beginning to dwindle makes sense, since that was when the Rangers started losing possession but still winning games.
The danger behind not understanding the importance of this is that PDO rates that far above 100 tend to come back to earth, and when they do—when the scoring “dries up”—if the roster composition isn’t reliable, those same teams can suddenly run into trouble winning games. Sound familiar? It should. The month of November has begun to reveal that trend in a similar manner it did last season.
Like last season, the Rangers aren’t a bad team, but they are a flawed on. This is an incredibly important distinction to keep in mind when talking about the league’s Cup contending clubs. One of their biggest flaws—one that’s been systematically exploited in recent games—is their propensity to bleed shots and shot attempts against. For example, in the last five games, dating back to the 6-1 loss to Pittsburgh on November 23rd, they’ve been collectively outshot 158 to 119 and their 5-on-5 CF% average over that span is a horrific 37.2%. If that’s too small of a sample size, also understand that through the first quarter of the NHL season they also own the fourth-worst team CF% (46.33), ahead of the 20th place Detroit Red Wings (46.16), the 27th place New York Islanders (46.07), and the 30th place Arizona Coyotes (45.29). That is bad company to keep and things don’t fare any better in the FF% column. There they own the fifth-worst percentage (48.06) ahead of the Buffalo Sabres (48.11), the Islanders (48.07), the Red Wings (46.16), and the Coyotes (45.29). Again, bad company to keep.
The defense is especially to blame here. Only Clendening currently sits above the 50% threshold in CF% this season, with Ryan McDonagh a distant second at 47.22, and only Clendening and McDonagh are on the positive side of the margin in FF%, too. It’s a bit of a gong show that could really stand to be addressed either through coaching or the front office. But, playing Clendening more now could help to improve those figures while a better long-term solution is determined.
I’m not suggesting that Clendening should become a regular player tomorrow. I understand he was brought in as a depth player, but a depth player with his kind of potential should be given more opportunity, not less, to contribute. He should be given permission to play in more than 28% of games this season, especially in spot relief of veterans like Kevin Klein and Dan Girardi, both of whom struggle mightily with individual possession performance.
Clendening is in seventh defenseman hell right now when he can arguably offer the team more offensive production and improved possession numbers overall, albeit in a managed role. While he is a far more effective NHL player, the role Keith Yandle played with the Rangers last season where much of his ice time was intentionally designed around getting more starts in the offensive zone and fewer in the defensive zone is the same type of role Clendening could fill this season. In fact, in the few games Clendening has played this season, this is already something the coaching staff has employed. His even strength offensive zone start percentage (oZS%) is 68.4%. That means he begins 68.4% of his shifts on the ice in the offensive zone, where his talents are better suited for success.
The only way any of this can happen, however, is if he’s actually given an opportunity to play more than one game a month. The only way he can fulfill the promise he brings is to play in more games where one mistake isn’t deemed one too many, immediately costing him the chance to build on his efforts game-to-game.
Growing older by the day, spending this amount of time in the press box is the last place he belongs. All the while as the Rangers continue to ignore science in favor of superstition and hockey’s versions of alchemy through unmeasurable categories like “grit”, “toughness”, and “leadership”. They may each have some inherent value, but that value is not quantifiable, and as such, it should take a backseat to values that are.