Early signs point to Connor Carrick being a good piece of Leafs’ future

As much as we’ve enjoyed this late-season version of the Leafs with their new dose of youth, the fact remains that this is a team sitting thirtieth in a league of thirty. Whether this iteration of the roster would operate at that level if given the whole year is probably debatable, but as Mike Babcock has even alluded to recently, you have to be careful about buying too much into this kind of short-term success we’ve seen in the last week or so.

But even with all that in mind, we can still take look at some of these kids – a few now approaching the 15-game mark – to see how they’ve been performing to date, both production-wise and in terms of pushing play in the right direction. As we all know, these waves can, in fact, die, so it’s good to look at who’s been for real.

Perhaps the best story of this Leafs youth movement since the deadline is that of Connor Carrick. When Toronto pulled the trigger on the Winnik trade last month, most of the early response was “Yes, another draft pick”, referring to the second-rounder they picked up, but we then quickly learned that the Leafs were pretty high on Carrick and pushed to have him in the deal. He was no throw-in.

A glance at Carrick’s career numbers at first probably suggested to many of us he’s a project-type prospect, as he’s plugged away for most of the last two years at the AHL level after breaking in with the Capitals in 2013. And as we’ve come to learn with someone like T.J. Brennan, even with strong offensive numbers, breaking out of the AHL can be impossible for some, especially defencemen. But that’s more of a cautionary tale, and considering Carrick is still a youngster, his future is obviously more wide open. He’s actually a month younger than Morgan Rielly. 

Considering all of this, it’s been somewhat surprising to see Carrick get such a good look with the big club right away. But the looks of things early on, it was the right call. 

Now that we’ve had some time to evaluate the trade, it seems the Leafs were smart to key in on Carrick, as he was sort of a victim of circumstance with the Caps, a team in prime contention and loaded with depth on the blue-line for the long term. With a rebuilding team in Toronto he can immediately factor into their plans, and so far he’s shown why with Babcock and his crew giving him plenty of opportunity already. 

In his audition with the Leafs (that will take him through the rest of the schedule), Carrick has already nabbed a pair of goals and assists through twelve games. While those obviously aren’t eye-popping numbers, it’s still an impressive 27-point pace over an entire schedule for a guy who’s tasked with about 15-16 minutes per night. 


Carrick’s underlying numbers have been strong as well, as he’s been clipping along at 55% in terms of score-adjusted Corsi For, also good for +2.3% relative to his teammates when he isn’t out there. Keep in mind he’s spent the majority of his time at even-strength alongside Gardiner, perhaps the Leafs’ best play-driver, but his numbers overall have looked fine. 

Again, it’s early, and these numbers are through 15 games this season (including three with the Capitals), but it’s clear so far that Carrick is no anchor, or at least he hasn’t shown to be yet. At the very least, it’s hard to say he isn’t a strong upgrade on the blue-line compared someone like Roman Polak, a player who has almost zero offensive upside and carries the unfortunate tag of “defensive defenceman” (usually just a nicer name for someone who gets hemmed in their own zone all night). Carrick has some nice puck skills, and as his experience in the NHL grows, his reads and overall offensive instincts should only get better.

Given where the Capitals and Leafs are in terms of contention and organizational depth, the deal that sent Carrick to Toronto last month should be looked upon as a move that made sense for both clubs. And who knows, maybe Lamoriello finds another way to add to the defence corps and push him down the lineup before next season. But for now, management, the coaching staff, and fans alike should all be pleased with how he’s turned out so far given his place on the roster.

Numbers from Corsica.Hockey and Hockey-Reference.

  • JimMc

    I hope you’re right about Carrick, but using any form of CORSI to evaluate a player is foolish. Here is a list of the top point scorers this year along with their league wide CORSI ranking (assuming > 20 GP)

    Player Pts Rank Corsi Rank (>20 GP)
    Kane 1 145
    Benn 2 68
    Crosby 3 87
    Karlson 4 205
    Kuznetsov 5 183
    Seguin 6 53
    Gaudreau 7 377
    Pavelski 8 57
    Kopitar 9 9

    Let’s look at the top players in the league by Corsi:

    1) Nick Shore
    2) Milan Lucic
    3) FRANK CORRADO!!!
    4) Drew Doughty
    5) Tyler Toffoli
    6) MARC ARCABELLO!!!
    7) Dustin Brown
    8) Pavel Datsuk
    9) Anze Kopitar

    The leafs should be really proud having 2 players in the top nine. Notice however that 6 of those players play for LA and 2 for Toronto.

    Corsi is not a player stat it’s a coaching STYLE stat!

    Corsi has nothing to do with possession. Shooting the puck and keeping the puck are not the same thing!

    Corsi is a proxy of a proxy of what we’re trying to ascertain.

    • Benjamin

      Putting your team in a position to take shots and restricting the amount of shots against is far and away the number one priority for a defenceman.

      Corsi isn’t that useful for wingers, and it’s just part of the story for centres, but it tells you most of what you need to know about defencemen.

      • Benjamin

        Wow that’s so awesome! Frank Corrado is better than Norris trophy winner Erik Karlsson! A few more waiver wire pick ups like that and we’ll win the cup for sure!

        Maybe the number one priority of a defenseman is not to give a crap about shots that miss the net or shots that are easy to save. Maybe the number one priority should be to help your team get more goals than they give up.

        Some of the best defensemen in the league rank relatively low in Corsi, like weber, petriangelo, subban, Karlsson, Keith, Seabrook etc….

        The death knell of analytics in hockey will be served on the altar of bad statistics, Corsi being chief among them. And that’s a shame because analytics have a lot to offer!

        • Benjamin

          “Wow that’s so awesome! Frank Corrado is better than Norris trophy winner Erik Karlsson! A few more waiver wire pick ups like that and we’ll win the cup for sure!”

          I checked the strawman’s pulse and, don’t worry, he’s dead.

          “Maybe the number one priority of a defenseman is not to give a crap about shots that miss the net or shots that are easy to save. Maybe the number one priority should be to help your team get more goals than they give up.”

          So Fenwick > Corsi, that’s what this is about? Sure, who cares.

          “Some of the best defensemen in the league rank relatively low in Corsi, like weber, petriangelo, subban, Karlsson, Keith, Seabrook etc….”

          The number of things that I disagree with in this statement force me to conclude that your definition of ‘best/top/quality defenceman’ is quite different than mine. Cheers.

        • TGT23

          CF/60 Among D with 900 minutes (127 players):

          Subban – 14th
          Karlsson – 20th
          Webber – 24th
          Pietroangelo – 26th
          Kieth – 43rd
          Seabrook – 47th

          How is being in the top 25% (the first 4) ranking low, relatively?

          Further, I think we all know that guys like Corrado get protected, so his number needs to be taken with a grain of salt, as does Carrick, but they are encouraging.

      • JimMc

        The best players in the league don’t have the best Corsi and the players with the best Corsi aren’t the best in the league. A players Corsi depends on the playing style of the coaching staff.

        Why use a proxy like Corsi in situations where you have enough data to use the actual metrics? It’s basic modelling. Proxies should only be used rarely when there is absolutely no other choice due to data deficiencies.

        So….yep

        • Benjamin

          “Why use a proxy like Corsi in situations where you have enough data to use the actual metrics?”

          What are these actual metrics for a defenceman, to your mind?

          • JimMc

            For a start a corsi like percentile that ignores shots that miss the net and shots that are blocked! As these have zero chance of ever going in. That’s an especially grievous penalty to apply to the defensemen that actually blocked the shot! He blocks the shot thereby preventing a possible goal but gets a corsi penalty? Does that make sense to you?

            How about actual possession ie time with the puck.

            Percentage of shots blocked to shots taken while on the ice. Team Save and shooting percentage differentials of a player when on versus when off the ice. Take away giveaway differentials. Zone entries, controlled entries. Points, assists, goals.

            Time on ice is one of the simplest but it only compares players on the same team. So compound metrics that group players with similar profiles can be used to extend all of the above analyses further.

            Metadata such as text analytics that analyze how players are perceived in the regular and social media and scouting reports.

            Just don’t use Corsi heavens sake! It is not a player stat!

            And it’s important for those who endorse analytics to renounce Corsi before the opponents of analytics figure out how to renounce it themselves and use this to discredit the entire field.

            Corsi is not possession!

          • Benjamin

            “For a start a corsi like percentile that ignores shots that miss the net and shots that are blocked! As these have zero chance of ever going in. That’s an especially grievous penalty to apply to the defensemen that actually blocked the shot! He blocks the shot thereby preventing a possible goal but gets a corsi penalty? Does that make sense to you?”

            Yes, Fenwick>Corsi, we’ve done this already.

            “How about actual possession ie time with the puck.”

            Not measured, at least publicly.

            “Percentage of shots blocked to shots taken while on the ice. Team Save and shooting percentage differentials of a player when on versus when off the ice. Take away giveaway differentials. Zone entries, controlled entries. Points, assists, goals.”

            First one doesn’t make sense because those are two positive traits. Why do you want a ratio between them? Second one sucks. Third, fourth and fifth ones are fine. I don’t really care about points for a defenceman, very PP and team dependent generally.

            “Time on ice is one of the simplest but it only compares players on the same team. So compound metrics that group players with similar profiles can be used to extend all of the above analyses further.

            Metadata such as text analytics that analyze how players are perceived in the regular and social media and scouting reports.”

            What?

            In all your talk about ideal metrics, that range from non existent/outright awful, to ok, you didn’t mention goal differentials. So I’ve got to go back to my conclusion that we evaluate defencemen very differently.

          • Benjamin

            Ok shooting and save percentages are just as important as shots. Goals=shooting percentage times shots ON GOAL. Some players/hockey systems sacrifice shot attempt frequency for shot quality and vice versa. As long as the percent increase in shooting percentage is greater than the percent decrease in shots on goal, more goals are scored. This is why some players/teams have consistently higher/ lower than average percentages. So these measure don’t suck!

            Secondly Fenwick includes shots that miss the net. But shots that miss can’t ever go in so they are not relevant. Plus they are easy for players to manipulate. Hello there Kadri!

            Third it takes just as much effort to measure corsi as it does to measure possession. But if you cant measure possession don’t just call Corsi possession because it isn’t!

            Fourth I meant shots blocked to shots faced in the relation.

          • Benjamin

            “Ok shooting and save percentages are just as important as shots.”

            Not for defencemen and, in terms of save %, not for any skater.

            “Plus [shot rates] are easy for players to manipulate.”

            This is also wrong, individual shot rates typically stay very consistent season-to-season. Much more consistent than sh %, for example.

            “Third it takes just as much effort to measure corsi as it does to measure possession. But if you cant measure possession don’t just call Corsi possession because it isn’t!”

            No it doesn’t. You’re talking about a continuous measure that switches rapidly and exists in varying degrees vs a discrete measure that is relatively night and day.

            And people feel comfortable using Corsi as a proxy for possession because it just makes sense. If your team gets a shot attempt, it means you won the faceoff/gained possession, successfully moved it to a shooting area and voluntarily gave up possession when you took your shot. Get another shot attempt and it means you regained possession and did it all over again. It’s super intuitive man.

            “Fourth I meant shots blocked to shots faced in the relation.”

            Fine, still super weird. Player sv%? Probably useless.

          • JimMc

            What part of the equation goals = shots × shooting percentage do you not understand? It’s the same for team stats as it is for defensemen and forwards.

            And no corsi doesn’t make sense as a proxy for possession. A shot that misses the net or gets blocked results in a turn over 50 percent of the time. Good players and good teams can win at hockey without winning at Corsi necessarily. Otherwise nick shore and Milan lucic would be the best players in the league and and Johnny hockey wouldn’t be in the NHL let alone one of its top players!

          • Benjamin

            “What part of the equation goals = shots × shooting percentage do you not understand? It’s the same for team stats as it is for defensemen and forwards.”

            Again, this is where we clearly differ in what we think a defenceman’s job is supposed to be. For instance, I don’t want my defencemen habitually trying to pinch into high sh% areas and leaving the team susceptible to a counterattack. Sh % is not a remotely useful measure for a defenceman because a defenceman does not have the same role as a centre or a winger.

            “And no corsi doesn’t make sense as a proxy for possession. A shot that misses the net or gets blocked results in a turn over 50 percent of the time. Good players and good teams can win at hockey without winning at Corsi necessarily. Otherwise nick shore and Milan lucic would be the best players in the league and and Johnny hockey wouldn’t be in the NHL let alone one of its top players!”

            Alright, another angle. A ‘shot attempt for’ happens in the other team’s end, right? A ‘shot attempt against’ happens in your team’s end, right? So in order to take more shots than the other guys, on average, you’re probably in the other team’s end more often, right? Because your team has control of the puck more often, right?

  • Benjamin

    Babcock already warned us to avoid reading too much into players performance at this time of the year.

    And given Carrick couldn’t crack the capitals, that suggests on a “good team” he would quickly fall to a depth option.

  • Gary Empey

    Connor Carrick has fit in with Gardiner nicely.

    Here is his profile from “Hockey’s Future

    Talent Analysis

    Carrick lacks size (5’11, 185 pounds), but he more than makes up for it with his skating and offensive ability. He shows great confidence carrying the puck and attacks the offensive zone, jumping into the rush and creating opportunities. Though he is undersized for a defenseman, he shows competency in his own zone despite facing physical challenges. His shot and ability to quarterback the power play are what make him most valuable and he has the skill set to be very good as an offensive defenseman on the NHL level.
    Future

    Carrick was dealt by the Washington Capitals along with Brooks Laich and second round pick to the Toronto Maple Leafs in exchange for Daniel Winnick and a fifth round pick.

  • Benjamin

    If you have a meaningful statistic like points or goals in which one team had 6 of the top 9 players that team would be the best in the league by a huge margin. LA is good but not near the best.

    And if you had a team with 2 of the top 9 players in the league in a meaningful stat then that team wouldn’t be dead last in the league. But the leafs have that in Corsi and are the worst team. So that tells me that corsi is not a meaningful stat!

    Death to the Corsi stat!

  • magesticRAGE

    What about tracking the average of how long it takes a defenseman to exit their zone, by a pass or controlled. As far as I’m concerned, this is the primary purpose of being a defenseman. Forwards generally are the scoring players, which are fed the puck from their defensemen.
    If there was a metric to combine zone exit times and preventing zone entries, that would be helpful.

    • Benjamin

      There used to be a few bloggers who tracked this, however they’ve all been snapped up by NHL teams and their work is no longer available to the general public.

  • JimMc

    @mst

    Jay has done you a favour by sharing the laws of analytics. You should read them and thank him. The very simple explanation is that a team directing more SAG than their opponents will 1) Have possessed the puck more, 2) will score more often, 3) will win more games.
    It is clear that you seem to feel the definition of a defenseman is one that creates the most offense. I disagree. I think a top defenseman first is elite at defense and secondly then is a higher end contributor offensively. CF, Shots at 5 on 5, fewer give aways vs takeaways can all factor into that. It is for this reason that I continually dispute Subban or less so Karlsson as the top d man in the league and therefore Norris candidates. If there were a reverse version of the Selke, say an award for the most offensive defenseman than I would vote for them…otherwise they (in particular P.K.) are too dangerous to have out there when you are trying to protect your own end.