Vancouver Canucks Player Usage Charts for 2011-12 season

Advanced statistics in hockey aren’t everyone’s cup of tea. Also, not everyone likes tea in the first place, so when you offer them a cup of tea and they politely, yet firmly, decline and you keep shoving tea in their face, it’s rather rude. And yet, I continue to talk about advanced statistics on PITB, trusting that the tea-drinkers will appreciate a nice rooibos tea while everyone else will ignore it completely.

But what if I promised a pretty picture that might make advanced statistics a bit more clear? Or, in my increasingly forced tea metaphor, what if I added a bunch of high-fructose corn syrup to green tea but still pretended it was healthy by putting ginseng in it?

Robert Vollman of Hockey Abstract has released the 2011-12 Player Usage Charts, which take three of the most common and useful advanced statistics and put them into a handy-dandy chart that makes it easy to see at a glance how a player was used and how well they performed in their role. I’ve taken a look at these charts in regards to the Stanley Cup Finalists over at Backhand Shelf; now let’s take a look at the chart for the Canucks and see what can be gleaned from it.

I recommend downloading the PDF of the Player Usage Charts as it has a more complete explanation of the charts as well as analysis of each team. I know that these charts will be my first stop when the Canucks acquire any new players this offseason.

First, a quick explanation of the elements of the chart:

  • All statistics are at even-strength.
  • The horizontal axis represents zone starts: the further to the right, the more often the player is starting in the offensive zone.
  • The vertical axis represents quality of competition: the higher up the chart, the more often the player is facing the opponent’s best players.
  • The bubbles represent Relative Corsi, which is a puck possession statistic that uses shot attempts. The blue bubbles indicate a positive Relative Corsi, the white bubbles indicate negative Relative Corsi. At it’s most basic, a big blue bubble is good, a big white bubble is bad.
  • Defencemen are in purple.
  • Players with an asterisk did not play a full season with the Canucks.

(HockeyAbstract.com)

One of the first things I noticed when looking at the Canucks’ chart as compared to those of other teams is how small the spread of quality of competition is. Over the last few seasons, Alain Vigneault has moved away from hard line-matching to deploying his lines according to zone starts. Instead of sending checking players like Malhotra and Lapierre out against the opposition’s top forward lines, he buries them with defensive zone starts instead.

Context matters here, which is why I like these charts. Malhotra and Lapierre, as well as Volpatti and Weise to a certain extent, have a legitimate excuse for their poor puck possession numbers and their big white bubbles on the chart. Those four players started in the defensive zone more than pretty much any other player in the NHL. Given that kind of deployment, it would be astonishing if they didn’t have negative Corsi numbers.

As we start to get more towards the middle of the chart, however, there’s less excuse for Alberts, Rome, and Ballard, who are all clustered near each other. While they did start more often in the defensive zone, it wasn’t to the extreme degree of the forwards and they faced weaker competition. Comparing them to Gragnani does make me a bit wary, however: Gragnani’s positive Corsi is somewhat in proportion to his higher number of offensive zone starts and he faced similar competition to Alberts, Rome, and Ballard.

As for the rest of the defencemen, the duo of Hamhuis and Bieksa come out looking very good, facing the toughest competition among defencemen and pushing puck possession in a positive direction. Tanev had a strong season as well: he wasn’t the least bit sheltered and pulled his own weight. Salo and Edler, on the other hand, don’t come out of this as strong. Salo’s puck possession numbers were not good and Edler was barely in the positives despite starting more often in the offensive zone. As Harrison suggested, finding a partner for Edler will be one of the major challenges of the offseason.

On to the forwards, and Chris Higgins is a good player to start with: he faced the toughest competition among Canucks’ forwards (other than Pahlsson, whose numbers are skewed by his time in Columbus) and still managed to keep his puck possession in the positives. Hansen’s white bubble isn’t a huge concern considering his zone starts and quality of competition and he posted a career-high in goals, assists, and points.

The Sedins obviously benefit from Malhotra’s defensive zone starts, starting nearly 80% of their shifts in the offensive zone. Burrows benefits also, of course, and all three took advantage with the best Corsi numbers on the team.

Hodgson’s white bubble is conspicuous. The only other forward with a white bubble near him is Raymond (you can barely make it out behind the “m” in his name). While Hodgson wasn’t sheltered all season (and wasn’t with the Sabres after the trade), he was sheltered throughout January and February, seeing a big boost in offensive zone starts. Prior to that, however, he was deployed in a pretty normal fashion and, in terms of puck possession at least, struggled.

Kassian, on the other hand, was sheltered, though the chart more represents his usage in Buffalo than in Vancouver. He did put up a positive Corsi, but it was against weaker competition.

One player to keep an eye on is Ebbett, who played just 18 games for the Canucks due to injury, but moved the puck forward in his limited minutes while starting more often in the defensive zone. Personally, I’d like to see him re-signed as a 13th forward option who can slot in and out of the lineup as necessary.

The one other observation is how well Samuelsson did in Florida. The chart barely represents his time with the Canucks this season as he was traded early on, but it does show that he faced tough competition, started more often in the defensive zone, and still pushed puck possession forward for the Panthers. Sturm was also a positive puck possession player for the Panthers, albeit in slightly easier minutes.

Tags: , ,

40 comments

  1. Cam Charron
    June 5, 2012

    First thing I noticed is that Hodgson and Salo are the only players to the right of 50% that have a white circle…

    VA:F [1.9.16_1159]
    Rating: 0 (from 0 votes)
  2. Chris the Curmudgeon
    June 5, 2012

    The first thing I noticed was that supposedly “sheltered” Hodgson actually faced higher quality opposition than the Sedin twins, Booth and Raymond. I guess when you have an agenda to advance you can be selective in your interpretations of data eh Daniel?

    VA:F [1.9.16_1159]
    Rating: -17 (from 23 votes)
    • bergberg
      June 5, 2012

      You know this includes his time in Buffalo right?

      VA:F [1.9.16_1159]
      Rating: +12 (from 14 votes)
    • Schneider's Teeth
      June 5, 2012

      The only person I can see with an obvious agenda who is selectively interpreting stats around here is you.

      VA:F [1.9.16_1159]
      Rating: +13 (from 17 votes)
  3. James W.
    June 5, 2012

    I think the most telling thing here is that Rome, Ballard, and Alberts all basically performed the same, regardless of cap hit. It’s the best justification as I’ve seen for getting rid of Ballard, beyond the ‘Watch him play’ argument.

    As well, I think it shows that Rome/Alberts are not top 6 guys – they’re fine fill ins, but that’s about it.

    I think the D-core of Vancouver needs some work – if Salo is done, and Rome/Alberts are depth defencemen, then we’re looking at Hamhuis, Bieksa, Edler, Tanev, and Gragnani. I hope we can snag Justin Schultz, but even then we have 3 defencemen with little NHL experience, and who knows if any of them can play with Edler.

    VA:F [1.9.16_1159]
    Rating: +5 (from 7 votes)
  4. MB13
    June 5, 2012

    Based on the above data and the “Canucks would’ve beat the Kings had D. Sedin played”, when’s the pretend parade?

    VA:F [1.9.16_1159]
    Rating: -16 (from 20 votes)
    • Knight of Cydonia
      June 5, 2012

      June 30th???: http://commonground.ca/about/walkforpeace/

      VA:F [1.9.16_1159]
      Rating: +3 (from 3 votes)
    • iceman
      June 5, 2012

      Right in your head. Where else?

      VA:F [1.9.16_1159]
      Rating: +4 (from 6 votes)
    • Brent
      June 5, 2012

      Even with Daniel healthy, I don’t think we would have beaten LA. Maybe if we had played San Jose first and got playing at peak efficiency before, then MAYBE.

      VA:F [1.9.16_1159]
      Rating: +3 (from 3 votes)
  5. Bejoyed
    June 5, 2012

    Excellent compilation, thanks Daniel.
    I agree that finding Edler a suitable partner is the main hurdle for the offseason, no easy task!

    VA:F [1.9.16_1159]
    Rating: +6 (from 6 votes)
    • Rituro
      June 5, 2012

      If recent history has taught us anything, the answer — as always — lies in a big trade with Florida. Brian Campbell, anyone?

      [/halfjoking]

      VA:F [1.9.16_1159]
      Rating: +4 (from 4 votes)
  6. Dane
    June 5, 2012

    Currently drinking Coffee. This entire post is lost on me.

    VA:F [1.9.16_1159]
    Rating: +6 (from 6 votes)
  7. akidd
    June 5, 2012

    nice chart but too bad it’s all predicated on the corsi. I wonder if this corsi evaluation has sunk in for the players yet? are they taking low% shots just to work the corsi? it wouldn’t surprise me. in a competitive world it’s adapt or perish. they have to be aware of it. dump the puck in the corner or flip a weak wristshot in the direction of the net? one gets you a blue circle the other doesn’t. the observer influences the observed again. obsurd.

    at least with plus/minus the players’ aims and the team’s are the same.

    there’s lots of good stuff with the advanced stats but my guess is that in five years’ time statisticians will look back and chuckle at the corsi.

    VA:F [1.9.16_1159]
    Rating: +3 (from 3 votes)
    • akidd
      June 5, 2012

      is there a stat that combines corsi events and shooting %?

      VA:F [1.9.16_1159]
      Rating: +1 (from 1 vote)
      • akidd
        June 5, 2012

        oops. maybe that just equals plus/minus. what’s so bad about plus/minus again?

        VA:F [1.9.16_1159]
        Rating: +2 (from 2 votes)
        • akidd
          June 5, 2012

          a boxing analogy just popped into my head and as it seems i’m not interrupting anyone here(“ping,” said the pin,) let me muse that corsi is like punches thrown and landed and plus/minus is a total of knockdowns, standing 8 counts and ko’s . and when it’s a good sign that a boxer can get into a postion to throw a lot of punches it doesn’t always mean it is an effective overall strategy. foreman would’ve had a big blue circle and ali would’ve had a big white one.

          nice chart for the zone starts though and q of c. have no clue of the accuracy of the latter stat but the chart looked good cross-referenced like that.

          VA:F [1.9.16_1159]
          Rating: +1 (from 1 vote)
        • Daniel Wagner
          June 5, 2012

          The issue with plus/minus is that it is influenced by a lot of things outside an individual player’s control: his teammates on the ice, his goaltender, the opposing goaltender, bad luck, good luck, etc. Shooting percentage tends to fluctuate wildly from year to year, so it’s tough to rely on it.

          For instance, Ryan Kesler’s career high in goals last season coincided with a career high in shooting percentage. He went from a 15.8 shooting percentage last season to a 9.9 shooting percentage this season. It seems unlikely that either one of those is indicative of his “true talent.” Some of that is luck, some of it is shot selection, and some of it is opportunity, but the fact is that shooting percentage and plus/minus are not reliable indicators of future performance.

          VN:F [1.9.16_1159]
          Rating: +2 (from 2 votes)
          • akidd
            June 5, 2012

            i don’t think shooting percentage is all luck though. kesler had a new linemate this year who suppresses shooting % for example. and corsi relies on linemates as well.

            both corsi and plus/minus are flawed imho but over a longer period relative plus/minus could be pretty valuable indicator. whereas guys like booth could go a full career riding the corsi train and no one would be the wiser… if they didn’t watch.

            VA:F [1.9.16_1159]
            Rating: 0 (from 0 votes)
            • Daniel Wagner
              June 5, 2012

              It’s not entirely luck, no, but I also wouldn’t put the blame on David Booth. I think there’s a huge number of factors that roll into changes in shooting percentage and they’re very difficult to delineate. One of the ways that people have tried to do this is with WOWY (With Or Without You), where they’ll separate a player’s CORSI with or without a certain linemate. Here’s a look at this type of analysis in regards to Ovechkin: http://www.mc79hockey.com/?p=3400

              I have no idea what you mean by “relative plus/minus” though.

              VN:F [1.9.16_1159]
              Rating: 0 (from 0 votes)
    • MB13
      June 5, 2012

      You know what’s a great advanced stat? Goals.

      I don’t get Corsi – all I know is that when I took a quick peak at the hockey abstract document I noticed that the players I already knew were really good had big blue circles and the players I already knew that weren’t good had a hollow one. What’s the point of that?

      VA:F [1.9.16_1159]
      Rating: +1 (from 3 votes)
      • Daniel Wagner
        June 5, 2012

        I definitely understand this criticism: Corsi often just confirms what we already know from watching the game to a certain extent. The really good players tend to have a really good Corsi and the really bad players tend to have a really bad one. Where it gets interesting is the in-between, which is where teams might be able to gain an advantage by looking at advanced statistics. Everyone knows that the Sedins are good hockey players who control the puck well and create good scoring chances. That’s not revelatory. But if you can scout out players for your third and fourth line that are able to either suppress scoring chances for the opposition or create them for themselves, then you’ve got something worthwhile.

        Good players sometimes end up on bad teams where they don’t score a lot of goals and have trouble preventing them, giving them a terrible plus/minus and low point totals, but their Corsi relative to their teammates might be surprisingly good and on a better team they might get better results. Isn’t that worth pursuing?

        VN:F [1.9.16_1159]
        Rating: +6 (from 6 votes)
        • Jyrki Lumme
          June 6, 2012

          Given the need for an defensive partner for Edler it would be good to see a chart like this on the 2012 d-man UFAs. Inspiration for future article Daniel?

          VA:F [1.9.16_1159]
          Rating: 0 (from 0 votes)
      • tom selleck's moustache
        June 6, 2012

        Never saw “Moneyball” I take it? It’s about trying to get an advantage over other teams; everybody has access to the basic stats of goals/assists. If it was really that simple, then any schmuck off the street could run a team.

        It’s also about taking a more methodical approach towards building a team and not relying on the subjective/touchy-feely assessments of “he’s got a nice stride/good looking shot” which is arguably just a lot of guessing and praying.

        VA:F [1.9.16_1159]
        Rating: +2 (from 2 votes)
        • MB13
          June 6, 2012

          Yes – but moneyball at the time was revolutionary. I just don’t think corsi is all that revolutionary.

          The point of advanced statistics is to mine data to find hidden indicators of undervalued talents. When looking at the Canucks corsi, very little jumped at as being hidden in terms of what we did or didn’t know about the players.

          In other words – moneyball statistics have essentially become more important than the big 3 offensive statistics of the past (avg, RBI, HR) whereas I still view the G, A, +/- as more influential than corsi.

          I hope the Canucks have (actually I don’t because I stopped rooting for them for personal reasons) something more advanced than just plain corsi. corsi to me probably mimics the w/l ratios of teams. Ie teams with a good corsi have a good Winning %. A stat I would find useful is why a team with a poor corsi has a good winning % and vice versa. That was moneyball – why does a team full of guys that have (relatively) low batting averages, can’t steal, etc win? Oakland knew because that is how they built the team.

          VA:F [1.9.16_1159]
          Rating: -1 (from 1 vote)
    • Daniel Wagner
      June 5, 2012

      It’s definitely possible that Corsi won’t last; it was one of the earliest attempts at an advanced statistic to track puck possession, so something else might come along that is better.

      I don’t think it’ll go the direction you’re suggesting, however, because teams tend to keep their own proprietary statistics that they don’t release publicly. Vigneault talks about tracking scoring chances and has mentioned players being “plus” or “minus” in scoring chances, which doesn’t line up with their Corsi numbers or the types of scoring chances tracked by the blogosphere, so the Canucks are obviously tracking something in a way that we don’t have access to. We use Corsi and similar statistics, because that’s all we’ve got right now and it’s been shown to correlate with winning over the long haul.

      Players, however, know that their coaches and managers are tracking other things. It’s unlikely that the statistics on scoring chances will treat a weak wristshot towards the net the same way that Corsi does. They might, however, see it as better than a dump-in as it creates an offensive zone faceoff. I don’t know. I wish I could pick Vigneault’s brain on such things.

      Here’s a blogpost that might be an indication of the direction advanced stats is headed in the future: http://www.coppernblue.com/2010/11/17/1818041/the-touches-stat-the-next-horizon-in-advanced-stats

      VN:F [1.9.16_1159]
      Rating: +1 (from 1 vote)
      • akidd
        June 5, 2012

        thanks for the link, daniel. a good read. the ‘touch’ stats seem very pertinant to me. especially the 2nd version. it seems like the right road to travel. very specific and non-generalized. ‘battle percentage’ in particular has a ton of potential.

        whew, much better than the path of generalizing.

        VA:F [1.9.16_1159]
        Rating: +1 (from 1 vote)
  8. Brent
    June 5, 2012

    Have always liked Ebbet. Not sure why he didn’t play more in the LA seris. Too soon from his injury?

    Hope they keep him.

    I guess I still have hope for Ballard, I think he was playing better when he came back from the inury, but yes, the cap hit is considerable. Also Rome, he is a steady defenseman. Sure he needs to work on the timing of his hits.

    Still miss Samulson, he could have helped against LA. Can we get him back at the trade deadline?

    VA:F [1.9.16_1159]
    Rating: +3 (from 3 votes)
  9. Gimmic
    June 5, 2012

    Man I love Tanev. I’m not sure if he’s technically considered a rookie anymore, but I still think of hims as one. If you compare his #’s to rookie defensman this yr, he played tougher mins than all but 3 of them. And he came out with better underlying stats. This kids a gem, I see him as a legitimate shut down guy in 1-2 yrs, if not now.

    Other notes:

    - Good bye Keith Ballard. $4.2M of below normal stats is unacceptable. He’s paid like a top 4 d-man, but is playing bottom 6 mins of guys who normally make around $1.6M. If Gillis can’t offload him this summer (like he should have last summer), I’ll be angry.

    - Salo struggled, a lot. Edlers numbers are pretty similar to last season, so I woudln’t say he dragged Edler down, but he definitely didn’t help him. If you plotted Jason Garrison on this chart he’d have better numbers than bot Hamhuis and Bieksa (albeit with higher o-zone starts). Gillis needs to sign him in July, ASAP.

    - Did I mention Tanev is just the best? :)

    VA:F [1.9.16_1159]
    Rating: +1 (from 5 votes)
  10. Lenny
    June 5, 2012

    I want to see a big blue bubble on the top left end of our chart. THAT player would be god like.

    VA:F [1.9.16_1159]
    Rating: +1 (from 1 vote)
    • Daniel Wagner
      June 5, 2012

      Download the PDF and take a look at the St. Louis Blues. David Backes is exactly what you’re describing. Remember when Gillis signed him to an offer sheet? It was for a good reason.

      VN:F [1.9.16_1159]
      Rating: +2 (from 2 votes)
      • J21 (@Jyrki21)
        June 5, 2012

        Or look at the Red Wings — pretty impressive how much they’re tilted that way (and to an extent, the NJ Devils). Also Marcel Goc on Florida really stood out.

        VA:F [1.9.16_1159]
        Rating: 0 (from 0 votes)
      • Lenny
        June 5, 2012

        St Louis’s chart’s scale is different from ours. % of offensive zone by Backes is about 45%. Compare to the rest of his team, yes he is a beast, but on our chart’s equivalent, he is around Higgins and Samuelsson. On the Canucks chart, the far left is 10%. I want to see a blue bubble on the top left of the CANUCKS chart.

        Still Backes is good.

        VA:F [1.9.16_1159]
        Rating: +1 (from 1 vote)
        • Daniel Wagner
          June 5, 2012

          Fair enough. :)

          VN:F [1.9.16_1159]
          Rating: +1 (from 1 vote)
  11. Lenny
    June 5, 2012

    Also, Luke Schenn’s numbers are not very good. He is somewhere in the middle with a white bubble. Better think twice before trading Lu for him.

    Having said that, he is a big boy who averages3 to 4 hits per games. The qualitative benefits cannot be measured by advanced stats.

    VA:F [1.9.16_1159]
    Rating: +1 (from 1 vote)
  12. Dan J
    June 5, 2012

    The stats leave me wanting. For instance, what is the threshold for quality of competition? I certainly wouldn’t want to line up against a fresh 3rd liner on a routine basis compared to a winded 1st or 2nd liner, more so when I am tired and caught out on an offside on an errant pass.
    Each player has a role and expectation and only a few each year are afforded the opportunity to exceed it and lesser still surpass the expectations imposed upon them.

    VA:F [1.9.16_1159]
    Rating: 0 (from 0 votes)
  13. Locky
    June 6, 2012

    Sulzer jumps out to me. Middling QoC, middling zone starts and +ve Corsi. Very good for a 7th d-man, especially one of his cap hit. Disappointed we let him go to Buffalo and not an Alberts.

    Data may be skewed by the tear he went on when paired with Ehrhoff. But that only doubly disappoints me that Ehrhoff is gone. Which makes me sad about Edler this season. Which means I need alcohol.

    VA:F [1.9.16_1159]
    Rating: +1 (from 1 vote)
  14. Locky
    June 6, 2012

    Also from viewing the Oilers chart, someone needs to draw the Flames twitter manager’s attention to the location, colour and size of Ales Hemsky’s representation.

    VA:F [1.9.16_1159]
    Rating: 0 (from 0 votes)
  15. Gabe
    June 6, 2012

    Surely you mean “increasingly strained tea metaphor.”

    VA:F [1.9.16_1159]
    Rating: +2 (from 2 votes)
    • Daniel Wagner
      June 6, 2012

      Brilliant.

      VN:F [1.9.16_1159]
      Rating: 0 (from 0 votes)
  16. Aaron
    June 6, 2012

    The follow is a chart of Vancouver Canucks Stanley Cup wins:

    There, very interesting! Now bring on October. I need my Canucks fix!

    VA:F [1.9.16_1159]
    Rating: +1 (from 1 vote)