• Shift Lengths and Faceoff Choices

    by  • October 9, 2013 • Hockey • 4 Comments

    The 5v5 TOI against New Jersey for Edmonton’s forwards was pretty amazing. Here’s a summary of the ice time, with the number of 5v5 shifts, broken down into a couple of different lengths. It’s a pretty interesting breakdown because it tells us a little bit about how the Oilers achieved such a wide spread of ice times. It wasn’t really about the coach putting certain players out more; it was more to do with certain players taking/being told to take much longer shifts than other players.

    Screen Shot 2013-10-09 at 12.44.59 PM

    Of the 233 5v5 shifts taken by Oiler forwards, 35 of them were more than a minute in length. That seems like an awful lot in today’s NHL – I’ve written about this before, as has Gabe Desjardins, and the average shift length in the NHL has been dropping over the years. You don’t have Phil Esposito wandering around for 3:30 and then stopping by the bench for a smoke before heading back out anymore.

    Gabe came up with a pretty interesting finding a few years back: good things didn’t happen as shifts got longer. Basically as shifts got longer, they were more likely to result in a worse Corsi%, regardless of whether the player was one who tended to take long shifts.

    Screen Shot 2013-10-09 at 1.31.47 PM

    What you see in that table are sixth different groups of players, sorted by how long their average shifts are. Every group of players did worst once they’d been on the ice for more than a minute. There’s a bit of selection bias here I think – if the coach wants players out on the ice for less than a minute and you don’t change when the puck is in your end, you’re not necessarily capturing the impact of a player being fatigued – you might just be capturing the impact of a team being stuck in its own end, which is something that happens from time to time in hockey.

    This is a question that you could answer with data and expertise in terms of physiology I think. Anecdotally, I tend to think that long shifts are made possible by gliding. There’s no reason that a team that was so inclined couldn’t collect data on its players with a tag in the jersey or something and see how their performance changes as their shift gets longer – whether there’s more gliding and a reduced top speed, for example. Arsenal have done something similar in the Premier League and it’s shaped how Arsene Wenger manages the team as well as given him a stronger case to make when dealing with players unhappy with being substituted.

    There’s an interesting passage from The Numbers Game: Why Everything You Know About Soccer Is Wrong that came to mind as I considered this:

    …another study coauthored by Carling that looked at players’ work-rates after a teammate had been sent off shows that soccer players are expert at pacing themselves and operating at less than full capacity.

    As a result, fatigue may not show up in their average workrate: it may not be visible from the dugout. Instead, it shows up when they try to go from the 90 percent capacity they are operating at to the 95 percent they need to stretch for a tackle or to leap for a header. Early in the game, they can do this easily. Later, when they are pacing themselves, reaching the required capacity is not longer possible.

    Eakins has talked a lot about wanting guys going hard all the time and upping the intensity and pace of the game – long shifts seem inconsistent with that. Soccer’s the sort of game where you have bursts of effort and then effectively take your breaks on the pitch. Hockey isn’t – burst of effort, go to the bench.

    All of that said, one game is one game and it’s hard to get much out of it – for all we know, it just so happened that Hall/Eberle/Perron caught shifts where they were on the ice for a while. The game was also a bit of a perfect storm for this too, with the Oilers trailing and then having two days off. The contrast between the long shifts that RNH/Hall/Eberle/Perron took and the long shifts that the rest of the top six didn’t take is pretty jarring though. It’s worth watching as the season progresses.

    * * *

    One other thing that caught my eye, from a “Things That Might Be Good For Will Acton’s Career” perspective, is how Eakins dealt with defensive zone faceoffs when shorthanded. Here are the forwards he had on the ice for them:

    Acton-Gordon
    Gordon-RNH
    Acton-Gordon
    Acton-Perron

    In Vancouver:
    Acton-Gordon
    Acton-Gordon
    Acton-Gordon
    Gordon-Hemsky (Acton in penalty box)
    Acton-Gordon
    Acton-Gordon
    Acton-(likely) Gordon (Scoresheet says Perron, who was in the penalty box)

    Against Winnipeg:
    Gordon-Hall
    Gordon-Hall
    Acton-Hall
    Acton-Hall
    Acton-Gordon
    Acton-Gordon
    Acton-Gordon
    Acton-Gordon
    Acton-Gordon
    Acton-Gordon

    There’s a theme emerging here I think. Eakins virtually always has two centres on the ice for a faceoff on the PK and (as I’ll maybe show tomorrow), there’s a pattern where one of them goes to the bench as soon as the puck as cleared. It’s an interesting tactic – one of the things that it does is give the first centre who goes in to take the faceoff a lot of room to push the rules to their absolute limit – if he gets thrown out, there’s another centre there to take the draw. The Oilers are 14-12 so far in faceoffs while shorthanded. It’s early but, at the very least, they aren’t getting killed.

    How does this affect Will Acton and Mark Arcobello? Well, Acton’s a LH C who kills penalties and Arcobello isn’t. Sam Gagner isn’t really either. If the Oilers get to the point where they’re making a choice between those two guys, Acton has a skill set that the coach seems to value that won’t be present elsewhere. I’m not sure that Arcobello, who has played 12 seconds of PK time, does.

    Email Tyler Dellow at tyler@mc79hockey.com

    About

    4 Responses to Shift Lengths and Faceoff Choices

    1. SKINNY65
      October 9, 2013 at

      I’ve no idea how much of what you are showing is from Eakins actually saying stay out longer or not; but the one thing I do know from years of watching the Oilers play is that Hemsky is almost always the first player off the ice. It’s actually driven me crazy as I see him head to the bench far earlier then this linemates and potentially miss chances because Stortini or Brown comes on for him.
      I’m sure he’s doing exactly what a player is supposed to do, you come off when you START to get tired, not when you are tired, and as frustrating as it is on occasion, I’ve thought of it a lot when the idiot scribes talk about Hemsky being a ‘selfish player’ or not a team guy. But I’m sure that the MSM would just say that proves he doesn’t care enough about the Oilers and just wants off the ice.
      I was shocked to see so many long shifts for Nuge though. If that repeats that is obviously a coaching change.

    2. October 10, 2013 at

      This doesn’t impact your larger point, which I think is solid, but there are some faceoffs missing from your data. I remembered Hall and Hemsky being out for a DZ draw in the Vancouver game because it struck me as being kind of dumb (the reason, which makes sense, is that Gordon and Acton were both out for the previous faceoff and couldn’t get off the ice). That FO happened at 10:44 of the 1st. Not sure how you’re collecting the data, but I figure having the time-stamp might help.

    3. Tyler Dellow
      October 10, 2013 at

      Thanks Scott – I was just scanning down the play-by-play files, so I guess I missed it.

    4. Henry
      October 10, 2013 at

      Not sure if it’s relevant, but I caught several Barons games over the past three years when they came to town. Two seasons ago Arcobello killed penalties with Hunter Tremblay and seemed reasonably effective. Last year he didn’t kill as much though he was #1 centre at that time.

      Its anecdotal evidence, I know but Arcobello has at least tried the PK before. Acton may be useful for this but when Gagner gets back, it could be that Arcobello challenges strongly for his spot. I bet Eakins experiments with him on PK some in the next few games. His FO numbers are surprisingly good thus far.

    Leave a Reply

    Your email address will not be published. Required fields are marked *