Who were Ryan Kesler’s best linemates last season?

Sometimes when I get curious enough about something to investigate it, digging up statistics and putting together charts, the answer turns out to be the obvious one. Fortunately, it can also turn up some other interesting information along the way.

Here’s the question I had: which wingers were most effective with Ryan Kesler last season? One of the big questions coming into this season is who should play on the second line with Kesler, once he returns too early? David Booth seems to have his spot all sewn up, but there are many competitors for the opposite wing, including Chris Higgins, Mason Raymond, Jannik Hansen, Zack Kassian, and Nicklas Jensen. Heck, if Shane Doan signs with the Canucks, you can add him and Alex Burrows to that list.

David Booth and Chris Higgins were Kesler’s most common linemates last season, but were they his most effective linemates? To get the answer, I did some WOWY (With Or Without You) analysis to see how Kesler performed with and without various linemates. In this case, the answer appears to be pretty definitively “yes.”

I used the tools available at hockeyanalysis.com to put all of this together and I looked at seven different linemates that Kesler had throughout last season. Of those seven, Cody Hodgson and Daniel Sedin played significantly fewer minutes with Kesler than the others, which limits the effectiveness of this type of analysis, but I kept them in just in case something interesting could be gleaned.

I started off with Corsi, a statistic that uses all shots for and against, including missed and blocked shots, to measure possession.

While you can click on the chart above to view the details, I have isolated the most important parts in the following chart:

This chart shows the Corsi rate per 60 minutes for seven Canucks when the played with and without Kesler, as well as when Kesler played without them. The yellow section of the chart is the difference between the Corsi rate for when the two players were together and when they were apart.

The only player who had a better Corsi rate when he played without Kesler was Burrows, which makes sense. When Burrows isn’t playing with Kesler, he’s playing with the Sedins, who not only drive possession better than anyone else on the Canucks, but also start significantly more often in the offensive zone.

It’s important to keep in mind what this means for Kesler’s other linemates: when Higgins, Raymond, Hansen, and Booth were not on the second line, they were generally bumped down to the third line, which would negatively affect their Corsi. What’s astounding is how big a change it was for Booth and Higgins when they played without Kesler, though that’s also simply an indication of how well they possessed the puck when they were together.

When with Kesler, Booth and Higgins had very strong Corsi rates and the difference is clearly seen in Kesler’s Corsi without those two players. Hodgson and Daniel also boasted strong Corsi rates when with Kesler, though they played more limited minutes with him, which is a fairly significant sample size issue.

Still, Hodgson’s Corsi with Kesler is an indication that playing Hodgson on Kesler’s wing wasn’t quite the disaster we may have initially thought. It certainly didn’t pass the eye test, but this might indicate that it would have been worth some further exploration. Again, it’s tough to say given the small sample size.

It’s interesting to note that Raymond, Hansen, and Burrows all posted similar Corsi rates with Kesler and that Kesler was significantly better off without them on his wings. There is often a push to reunite Frick and Frack and have Kesler and Burrows anchor the second line, but this suggests that Burrows wouldn’t be much of an improvement over Raymond on Kesler’s wing. That’s a sobering thought.

I also did WOWY analysis using goals.

This is where sample size really becomes an issue for Hansen, Burrows, Hodgson, and Daniel. Only 7 goals were scored either for or against when Hansen and Kesler were together, 10 when it was Burrows and Kesler, 6 with Hodgson and Kesler, and 5 with Daniel and Kesler. I have included them on the chart above, but removed them from the chart below:

The interesting one for me here is Booth. Some of Booth’s detractors suggest that he artificially inflates his Corsi rate with low-percentage shots. Essentially, this would mean that when he is on the ice, the overall quality of shots would be lower and the team’s on-ice shooting percentage would likewise be lower.

While this is far from definitive proof of that theory, when Kesler was on the ice without Booth, the Canucks scored at a higher rate than when he was on the ice with Booth. The same is true of goal prevention and, as you can see from the chart above, Kesler’s overall +/- per 60 minutes of ice time was better without Booth than it was with him.

This is not true at all with Higgins: when together, Kesler and Higgins were significantly better in terms of goal-scoring than when they were apart. As for Raymond, the difference is minor.

It’s still hard to ignore Booth’s impressive Corsi rate, but I am now a little bit more willing to be persuaded by Booth’s detractors when it comes to on-ice shooting percentage. It will be interesting to see how well Booth performs without Kesler in the lineup if the season starts before Kesler is able to return.

In any case, Booth and Higgins are still the best options for the second line, unless one of Nicklas Jensen or Zack Kassian is able to step up in training camp, if training camp occurs.

Tags: , , , , , , , , , ,

13 comments

  1. ArtemChubarov
    September 6, 2012

    Solid first attempt sir.

    While you did well to contextualize the assertion, your point about Hodgson/Kes doesn’t strike me as convincing.

    In particular, because of the extreme way the Canucks sheltered Hodgson, Kesler played more difficult minutes this past season than he did in 2010-11. Presumably, in games where Hodgson and Kesler were deployed together, Kesler played more “offensively oriented” minutes than he did otherwise. The face-off adj. numbers bear this out too.

    Also, while you’re right that the Canucks scored at a higher rate with Kesler sans Booth than they did with both of them the margin is .1 goals/60 (2.15 goals/60 with Booth + Kes, 2.25 goals/60 with Kes and no Booth). I’d suggest to you that a margin that slim is analytically meaningless.

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

      I agree with you on the Hodgson point. It’s still something I found interesting.

      Fair enough on Booth. I’m still trying to figure out a way to see if Booth does drive on-ice shooting percentage down as some people claim. Do you know if anyone has done that type of analysis?

      VN:F [1.9.16_1159]
      Rating: 0 (from 0 votes)
      • David Johnson
        September 6, 2012

        I actually just did a study players influencing their line mates shooting percentages (see http://hockeyanalysis.com/2012/09/03/relative-importance-of-a-players-impact-on-teammate-shooting-percentage/). Over the past 5 seasons Booth’s line mates shot at a rate of 8.906% without Booth and 8.632% with Booth so in theory Booth had a -0.274 impact on his line mates shooting percentages. The actual negative impact is actually not that bad for Booth, but relative to other top NHLers who have held steady top 6-9 jobs over the past 5 seasons he is in the lower tier of players. In the study I did he ranked 165th of 223 forwards considered.

        VA:F [1.9.16_1159]
        Rating: 0 (from 0 votes)
        • Daniel Wagner
          September 6, 2012

          Very interesting. I think I’ll need a bit more time to absorb everything in your post, but it’s interesting how small a gap that is, yet how it still puts him in the bottom half of that list.

          VN:F [1.9.16_1159]
          Rating: 0 (from 0 votes)
          • David Johnson
            September 7, 2012

            The reason why he is so low despite the small gap is because I only considered players whose line mates took 1000 shots while he was on the ice in the past 5 seasons. This results in a selection bias towards good players. The median of the players studied as a +0.262 impact on shooting percentage so essentially Booth is 0.536 below the median player in the study.

            VA:F [1.9.16_1159]
            Rating: 0 (from 0 votes)
            • Harrison Mooney
              September 7, 2012

              You’re all nerrrrrrrrrrrrrrds

              VN:F [1.9.16_1159]
              Rating: +11 (from 11 votes)
              • John in Marpole
                September 7, 2012

                Too. Much. Math.

                VA:F [1.9.16_1159]
                Rating: 0 (from 0 votes)
  2. ArtemChubarov
    September 6, 2012

    From stats.hockeyanalysis.com of NHL players who’ve played at least 2000 even-strength minutes over the past 3 seasons, Booth is 415 out of 423 in on-ice shooting%: http://stats.hockeyanalysis.com/ratings.php?db=200912&sit=5v5&type=goals&teamid=0&pos=skaters&minutes=2000&disp=1

    Here’s what I wrote on Booth being a sh% outlier earlier this summer: http://canucksarmy.com/2012/5/16/might-david-booths-style-of-play-suppress-on-ice-shooting-percentage

    Quick summary: there’s suggestive evidence that Booth’s “style” of play suppresses on-ice sh% somewhat, but the trend is relatively new and the sample size isn’t large enough to be definitive. Booth’s personal shooting% and his ES scoring rate (which, I think is more important when projecting future output) has been consistently strong, and rebounded last season in a big way.

    VA:F [1.9.16_1159]
    Rating: +1 (from 1 vote)
  3. Kagee
    September 6, 2012

    Yeah the Booth Kesler Higgins line was by far the best 2nd line we have had in a while, I know AV wants Higgins sturdy self on the 3rd, but the all USA line of BKH should be re-united full time whenever this upcoming starts.

    Unless AV wants to get fired. :D

    OR we get Versteeg in the Luongo deal OR some miraculous way Phoenix can’t get a good re-negotiated deal with their arena lease and then Doan signs in Vancouver!

    VA:F [1.9.16_1159]
    Rating: 0 (from 0 votes)
  4. Tim
    September 7, 2012

    I can’t help wondering if the following quote doesn’t apply…

    “Statistics are like a bikini. What they reveal is suggestive, but what they conceal is vital.”

    VA:F [1.9.16_1159]
    Rating: +1 (from 1 vote)
    • akidd
      September 8, 2012

      tim, that’s a direct quote from the honorable ron maclean. he dropped that gem when he was commentating the lpga canadian open a couple of weeks ago. believe it or not. i’m starting to think that ron is the victim of a botched lobotomy. either that or he’s taking cringe-comedy lessons from ricky gervais.

      VA:F [1.9.16_1159]
      Rating: 0 (from 0 votes)
      • Tim
        September 9, 2012

        akidd, I think your theory about a botched lobotomy makes a lot of sense. It might account for the fact that Ron didn’t give proper credit to the person who authored that quote, Aaron Levenstein, a professor of business administration in the US. Only in a Ricky Gervais script could Ron McClean be considered a journalist.

        VA:F [1.9.16_1159]
        Rating: 0 (from 0 votes)
        • akidd
          September 9, 2012

          to be fair it was a paraphrase. i think he said, “concealed the important parts.” and for anyone that didn’t see what happened, a 15 yo, became the youngest winner ever of an lpga event and presumably there were many young girls watching and probably getting a little too inspired for ron’s liking. so he figured it was time to reaffirm what was really important in the world.

          VA:F [1.9.16_1159]
          Rating: 0 (from 0 votes)