Canucks October Deep Dive

It’s been a unique October for the Vancouver Canucks. If you had told me back in August that they’d end the month with no less than four rookies in their lineup, Ryan Miller would be posting some of the best goaltending numbers in the league, and the Anaheim Ducks would be in the lead for the Auston Matthews sweepstakes, I would have been questioning the efficacy of your prescription medications, but here we are. 

Let’s dig into the first month of the 2015-16 season after the jump. 

Advertisement - Continue Reading Below

The Competition

The Canucks exited October with a respectable 14 points (5 wins, 2 losses, and 4 OT losses). They benefited from a pretty sheltered schedule in the first month, with only 6 games against teams currently sitting in the playoff picture (Montreal (W), Washington (L), LA (W), Dallas (OTL), Arizona (W), St. Louis (L)).

November will get more challenging with 14 games on the slate, including a long 7 game Eastern Conference road trip. 

The Goaltending

With Markstrom injured to start the season, it was up to Ryan Miller to carry the lions share of the starts for October, and he was superb. In fact, the .923 SV% he posted in October was his best start since the 2011-12 season (October 2011 SV% 0.930).  

Advertisement - Continue Reading Below

sv% october

When assessing goalie performance, what we’re really looking to assess is how often your goalie put your team in a position to win. We can calculate that using quality starts, which is the percentage of the times the goalie posts a save percentage higher than the league average, which was 0.910 during October 2016: 

Quality Start % Oct

As we can see, Miller posted a quality start 7 out of the 10 times so far this year, which is exactly what you hope to see from your starting goalie. 

Goalie analytics have been advanced lately, with metrics like High Danger Save Percentage and Adjusted Save Percentage getting some love amongst the analytically inclined in the hockey community. From my perspective, the most interesting new goalie stat out there is Nick Mercadente’s Adjusted Goals Saved Above Average per 60 minutes. Basically, this stat calculates the average save percentage for goalies by danger area (high, medium and low danger) then calculates how many more, or less, goals the goaltender in question will save, per 60 minutes of ice time. The benefit of this stat is that it takes into consideration whether or not a goaltender faces easier or tougher shots behind their defense, versus an average goaltender. Until he comes up with a more succinct name, we’ll call this MERC/60. 

merc60

Advertisement - Continue Reading Below

It remains to be seen whether Miller can continue this pace throughout the remainder of the year, but you’ve got to give credit where credit is due. Miller provided the club with full value in the month of October. 

Forwards

One of the emerging areas of advanced analytics that has received a ton of interest from NHL teams of late is zone entries, first pioneered by Eric Tulsky, now with the Carolina Hurricanes front office. What Tulsky found was that when a team is able to carry the puck into the offensive zone with control they are able to generate more shot attempts than dumping and chasing, thus increasing the teams shot attempt differential, which correlates to more wins. Recognizing the value of a player having the ability to gain the zone, a number of teams employing individuals to track zone entries, and our very own JD Jerk will be tracking all 82 Canucks games for us. 

forward ZE oct

The first thing that jumps off the page is just how good the Canucks top two centers have been so far this year, with Henrik, Horvat and Vrbata leading the way for the Canucks. McCann has also been really impressive so far as a rookie, with his neutral zone play being very impressive for a so early in his NHL career. 

On the topic of McCann, he led the team in even strength points per 60 minutes, helping him to cement his spot on the roster by leading the team in goals: 

forward p60

Advertisement - Continue Reading Below

We can see this more clearly when looking at on-ice shooting percentage, where McCann and Prust have clearly enjoyed their fair share of luck so far while Sven Beartschi has had the type of luck that sends you straight to the press box: 

OI SH%

So what do we make of this? While clearly McCann won’t be able to keep his white hot pace up, neither should we expect Vrbata and Baertschi to continue to suffer from sub 3% on-ice shooting percentage at even strength. Apart from the really exceptional or terrible players, on-ice shooting percentage tends to gravitate to the 9% league average over the long-term, so we should expect things to regress on either end of the graph. 

However, shot attempt differentials tells a somewhat worrying story, with only the first line (Henrik, Daniel, and Vrbata) and Sven Baertschi managing to managing exceed the 50% mark, while scoring chance differential tells an even more worrying story. If the Canucks expect to win consistently this season, they’ll need to see a significant improvement in both areas. 

oct f cf scf

Advertisement - Continue Reading Below

Defense

d ZE oct

Interesting trends emerge for defensemen zone entries, with a far lower percentage of controlled entries from the blue-line. Interestingly, Sbisa has clearly had his fair share of criticism for his failing in the eyes of advanced stats, and we see those same trends emerge with zone entries. On the other end of the spectrum, Yannick Weber had the lowest percentage of failed zone entries, chalking up yet another reason why we’d like to see more of him. 

In terms of shot attempt differential, the Canucks blueliners got their teeth kicked in at even strength in October, with the exception of Yannick Weber who only played in 5 games: 

oct d cf

Team Level

October
Record 5-2-4
ES CF 46.8 (26th)
ES SH% 7.5% (16th)
ES SV% 93.4% (9th)
   
PP CF/60 108.7 (4th)
PP GF/60 7.5 (15th)
PP SH% 14% (15th)
   
PK CA/60 80.9 (3rd)
PK GA/60 3.2 (2nd)
PK SV% 92.7% (2nd)

Advertisement - Continue Reading Below

As we saw last year, strong work on special teams can cover up weaknesses at even strength, and we’re seeing that play out early for the Canucks.In October, the Canucks were a very poor possession team, bailed out by exceptional goaltending from Ryan Miller and sublime work on the penalty kill.

Conclusion

This Canucks team is not a contender. Miller is unlikely to continue to perform at the level we saw in October and their poor possession numbers are likely to catch up with them. In many other years, this would mean they likely have no shot at a spot in the postseason, but that doesn’t mean this team is not a hell of a lot of fun to watch.

It’s been pretty amazing to watch what is unfolding with the young players right in front of our eyes. In the space of 12 months, Bo Horvat has gone from “will he make it past 9 games?” to the team’s undisputed second line center, and one of the team’s leaders. 

We all expected Jake Virtanen to get a long look, but no one expected Jared McCann to force his way onto the roster like he has. Sure, he’s benefited from some unsustainable numbers, and his possession game leaves much to be desired, but let’s not forget Horvat himself played a sub 44% possession game for this first three months in the NHL. McCann’s shooting percentage will cool, and his defensive game will improve. He’s earned a job as an NHLer. 

Advertisement - Continue Reading Below

Then there is Ben Hutton, who has gone from playoff scratch in Utica to one of the better performing Canucks blueliners. 

This year isn’t really about whether the 2015-16 Canucks are a playoff team, in fact they’re probably not. It’s about whether or not they’re better than four other teams from the Pacific division, which they very well may be. 



  • Possej Syhr

    money puck, finally someone speaking some sense around here. The Canucks are not a playoff team, but a mirage created by goalie voodoo.

    They have been hard to watch, getting their teeth kicked in by the opposition night after night on the possession scoreboard. I’d much rather the Canucks be play driving, ice tilting, puck possession mofos like these teams that obviously get analytics:

    Columbus (2-10-0) – CF 50.9% (12th best CF)

    Toronto (2-7-2) – CF 51.9% (7th best CF)

    Buffalo (5-7-0) – CF 52.2% (6th best CF)

    Carolina(5-7-0) – CF 54.4% (3rd best CF)

    Slow starts, but clearly the possession numbers under the hood show playoff teams.

    Statistically yours,

    Possej

    • Canuck4Life20

      Clearly those teams get analytics, just not hockey.

      Which is why they’re all sub-.500 and will be out of the playoff race by US Thanksgiving.

      How long before the hoi polloi in Toronto start calling for Babcock’s head? Five more games?

      Fortunately, quality goaltending solves most problems, and that’s why Benning spent the Deniro on a proven No. 1.

      Where would the Canucks be with Lack (.852 SV%, 3.60 GAA)? Crickets.

  • Larionov18

    “This year isn’t really about whether the 2015-16 Canucks are a playoff team, in fact they’re probably not. It’s about whether or not they’re better than four other teams from the Pacific division, which they very well may be.”

    Hang on – if they beat out four other teams from the Pacific division, then they ARE a playoff team.

    Let’s not ignore score effects in this analysis – the Canucks have, refreshingly, had the lead a lot early in the season. This leads to the dreaded ‘score effect’ wherein the trailing team pours on the offense to try to even up the game. They shoot from everywhere and take risks that leave themselves open to defensive lapses in hopes of getting a favorable bounce.
    Writers on this site are quick to point out these effects when it aligns with their narrative, but ignore it when it doesn’t. I would much rather see the Canucks start the season this way, than the alternative where they have great corsi numbers as a result of playing catch up all the time.

    The fact is the Canucks have been in every game, and have yet to lose by more than a goal. Four of their losses have come in overtime which, with the implementation of 3v3 play, is more of a circus act than a hockey game. So far so good, I say!

    • bigdaddykane

      17 likes, 0 dislikes on this comment. There is a thing called score adjusted corsi, which takes into account score effects. We are still under 50% in this.

      • Possej Syhr

        Way to catch people slippin jim,

        Excellent point. Score adjusted Corsi of 50% or greater separates the wheat from the chaff.

        Wheat: Carolina, Toronto, Buffalo

        Chaff: Pittsburgh, Detroit, Anaheim

        Statistically yours,

        Possej

      • Possej Syhr

        That’s fine. If score-adjusted corsi exists, why not use it?

        Presumably, given the fact that the Canucks have spent far more time leading than they have trailing, the score-adjusted corsi would paint a much rosier picture of things. Maybe not over 50%, but better than the raw corsi statistics.

        • Larionov18

          It’s 48.8%. It’s a number that is concerning. Given the relatively small sample, it’s possible we are quite a bit better than that, but yea, teams with a 48.8% SA Corsi are almost never contenders in the playoffs. I am quite shocked that readers of this blog don’t understand this fact.

          If you want to make the argument that Vancouver’s true talent SA Corsi is above 50, fine, but don’t come at me with this garbage that it is irrelevant. Year after year, once a reliable sample is available, it is a very good predictor for which teams near the top of the standings are good and which teams are lucky.

          Now, as moneypuck mentioned, Vancouver’s special teams is very good and will mask some of our problems at 5v5. Are we a playoff team? Sure, probably in the Pacific this year, since it is absolute garbage. But we’re not a contender.

  • Larionov18

    Bah, fancy stats.

    Yes, they’re useful, but most fancy stats are just a way of getting a larger sample size than goals for/against. Their predictive value has some merit, but the most predictive statistic for who will go deep/win the Stanley Cup is still goals for/against ratio.

    And tied for 3rd in the league is pretty good in my books.

    Canucks usually suck as a fancy stats team and every year the fancy statisticians are amazed at how they make it to the playoffs year after year (let’s pretend we never had a coach Torts).

    Maybe they regress, but considering we’re mid-league for shooting percentage, it doesn’t seem very likely. Miller’s been hot in goal, but he’s posting numbers that are sustainable if you look at his record (especially if the Canucks continue to limit quality shots over quantity of shots as they appear to be doing so far). Yes, I know “scoring chances” go against us too, but that’s not exactly an objective number.

    On the whole, I’m happy with the team. We have three rookies doing well and every reason to expect a sharp learning curve from them. Some vets are still trying to find their stride, but on the whole they’ve been very competitive every game, and we’ve lost twice, only twice in regulation.

    Last year the naysayers say the Canucks early returns couldn’t be sustained. IMO they’re doing better overall this year.

    Should be an exciting season!

    • andyg

      >> Bah, fancy stats.

      Now, now. Just because money puck does a poor job of statistical analysis doesn’t mean we throw the baby out with the bath.

      “Raw” corsi puts the Canucks at 47.73%; 26th in the league. Score adjusted corsi compensates for score effects and puts the Canucks at 50.0%; 16th in the league. Score adjusted fenwick (usually considered the “gold standard” for correlation to future results) puts the Canucks at 52.7%; 6th in the league. All numbers from http://www.puckon.net, includes numbers from last night’s game.

      Goal differential is often considered the best proxy for the “true talent” level of a team but it’s usually not until the end of the season that enough goals collect to make it a useful measure. According to nhl.com (which gives the Canucks an extra +1 for winning in a shoot-out) the Canucks are +10; tied for 3rd in the league. At this point in the season the goal differential is based on 50 to 60 events while corsi is based on over 1000. You’d expect the goal differential numbers to move towards the corsi numbers, not the other way around.

      Even strength shooting percentage is 12th in the league, even strength save percentage is 13th, PDO is 11th. (courtesy of naturalstattrick.com)

      I’d expect some of those more positive indicators to moderate back to the pack, but nothing here screams, “unsustainable”. I wouldn’t expect the Canucks to stay at the top of the Pacific; the Kings probably have better underlying numbers and I’d worry about the Canucks combination of age and youth holding up over a full 82 game schedule. The Canucks numbers are also based on a pretty easy schedule so far.

      I’d say that the Canucks are what we think they are: one of the better teams in the worst division in the NHL. At this point they’re a solid pick for a playoff spot with a best case scenario of being a speed bump on the Central Division’s path to the Stanley Cup Final.

      • Possej Syhr

        D,

        Thanks for this. As all of us analytic acolytes (analytes for short) can attest, advanced stats are always advancing and it just takes some digging to mine the data we need to make ourselves look as smart as we know we are.

        Statistically yours,

        Possej

  • wojohowitz

    Integration of new players always results in some dysfunction as the rookies and newbies try to figure out where they should be on the ice in each situation with and without the puck and that affects the overall performance of whole team. As the season progresses the mistakes happen less often as the group becomes a real team. Even the coach is on a learning curve with his TOI management which has been less than stellar. In last nights game no forward had 17 minutes of icetime. The article suggests the team will perform worse as the season progresses when logic would dictate they will perform better as the integration continues.

    The Canucks has an excellent month of October being close in all games with all losses by only one goal and will only get better barring injuries. Rather than missing the playoffs they could win the division.

  • wojohowitz

    Why is Canucks Army still drawing conclusions from corsi when using 11-game sample sizes?

    This is why people who disagree with the newer statistical/analytical approaches and trends in hockey always have ammunition to fire back at those they disagree with. It’s not like they need any encouragement but it makes it so much easier for them when people who should (and most likely do) know better continue to use statistics inappropriately.

    Nobody who, prior to the season, had an opinion on whether or not the Canucks would make the playoffs should be changing their mind or drawing new conclusions based on only 11 games worth of possession data.

  • bigdaddykane

    The Sedins are the only real concern for me. They look rather ordinary 5 on 5, and slower than Murray baron in OT. I hope they’re due for some puck luck. This team goes as far as Hank and Dank take them

  • Possej Syhr

    Thanks, I’ve been Jonesing for some statistical analysis this season! The analysis largely lines up with my own opinion from watching the games. This team is not a contender and will succeed based on special teams play and strong goaltending.

    Nice to see that the Sedins and Vrbata are a bit unlucky thus far. Hopefully their possession and finishing improves.

    Baertschi’s possession numbers are impressive, but not that surprising as whenever he is on the ice he seems to be controlling the puck and making plays happen. Nice to see him get a point for his efforts last night and hope it’s a sign of good things to come.

    As for McCann, not only is he leading the team in points per 60, he is leading all rookies in the league:

    http://www.tsn.ca/mcdavid-might-be-even-better-than-we-thought-1.384720

    What a surprise he and Hutton have been! Great to see Virtanen with a nice goal last night.

    This team may not be great, but it is fun to watch and vastly more interesting then the past two seasons.

  • bigdaddykane

    I am really hoping that Markstrom comes in and is worthy of a shot at an NHL starting goaltender position. If that becomes the case this club could see the Luongo-Schneider strength that carried this club for years.

    Some of the most successful teams in history have rode great goaltenders and Benning signed Miller to give the Canucks that same opportunity.

  • Dirty30

    What was the recent metric bandied about in the Province? 10 points in the first 10 games or you’re pretty much statistically out of the play-off race? Calgary one game away from digging a hole they likely cannot climb out of?

    So statistically, the Canucks are in contention by virtue of not being out of contention.

    Yes, Miller is playing well — wasn’t that the reason for the 6 million dollar contract?

    Sedins having a slow start? Well if they’re not having to pull every game winner out of their asses they can manage their efforts better. BTW: both of them are throwing punches this year … If that doesn’t say ‘contender’ then what does?

    Now the rookies are adding speed, skill and enthusiasm to the mix — sure I miss KB3’s wit, but I’ll take a play-making frowner like Hutton any day! He can smile for his grad picture.

    Yes, stats say everything returns to the mean, but stats said last year Bo should still be playing in Junior.

    Meh, I have enough to be depressed about in my life — Go Canucks!

  • Dirty30

    I don’t disagree with the point that the Canucks aren’t a particularly good possession team and they’re relying heavily on a seemingly rejuvenated Miller. But your analysis seems — as many others have pointed out — focused on a small sample size and missing the bigger picture. The Canucks have a much more fickle fan base than the Leafs or Habs and don’t have the deep pockets of those clubs so years of tanking isn’t really an option since it can lead to a deep spiral of not being able to afford any decent FAs in the future. You also have to remain competitive to placate the vets you do have.

    Yet in spite of this context the Canucks are icing a team that is a quarter rookies and 2nd year players! I cannot recall the last time we came close to that — maybe in the Sedins early years?

    What was the narrative over the summer? Aging team, no real succession or replacement plan for the Sedins, questionable trades, etc. I’m astounded that even with playing all these young players we still have a prospect pipeline that’s far stronger than 2 years ago and actual results on the ice.

    So perhaps you might consider writing an article that doesn’t just look at metrics related to possession or wins and losses but evaluates the PCS model in light of the developmental trajectory taken by the Canucks. The road taken here has made me for one seriously re-evaluate players like Prust and Dorsett who I’ve always seen as total plugs. But if you’re going to spend money on complementary players, I’d rather have ones with proven work ethic and stability (i.e. Higgins, Hansen, Burrows, Prust, Dorsett) than reclamation projects (Booth, to some extent Matthias and Ballard) or be difficult teammates (Kesler).