• The Fourth Line

    by  • October 15, 2013 • Hockey • 14 Comments

    Whenever a whipping boy is selected by the smarter parts of the online Oiler fanbase, I’ve generally been entirely willing to jump aboard the bandwagon. Ryan Whitney? I coined “Barbaro.” Nik Khabibulin? I led the charge. There’s an emerging trend in dumping on the fourth line though that I don’t think I’m entirely on board with.

    Let’s be clear: the fourth line looks horrible. All three of them have Corsi%’s below the Petrell Line of 40%. The game isn’t about getting the most Corsis though – it’s about scoring more goals than the other team. What would this fourth line cost the Oilers if they kept them together for 82 games? Turns out that we can come up with an estimate through the application of math.

    Let’s look at the 5v5 shot attempts for the fourth line and we’ll contrast those with what we’d expect from an average fourth line:

    Screen Shot 2013-10-15 at 1.41.10 AM

    The average fourth line Corsis at about 47.1%. So there’s a big gap there between the average fourth line and what the Oilers have got from their group so far. Here’s what it would look like over 82 games:

    Screen Shot 2013-10-15 at 1.51.47 AM

    Shot attempts aren’t shots though and shots aren’t goals. In 2011-12, fourth liners shot 6.24% and posted a .923 save percentage. About 50% of shot attempts are actual shots. I suspect that figure may be lower for fourth liners but we’ll work with that. So what if we convert our Corsi data into shots and then into goals using those figures?

    Screen Shot 2013-10-15 at 2.00.02 AM

    You can see that an average fourth line, over 82 games, would be about 4 goal difference better than the Oilers fourth line. Using the standard math of three goal difference to a point, you would figure that the fourth line being so much worse than the average fourth line costs the Oilers about a point.

    Fourth lines are special though. Goals for and against by fourth lines tend to be scored at low leverage moments in hockey games. Luke Gazdic scored his goal against Winnipeg very early in the game. Will Acton scored tonight in garbage time in Washington. The Oilers have played 18 periods so far this year. 22% of the 5v5 minutes played by the Oilers fourth line has come in the third period against Vancouver and the third period against Washington. Garbage time.

    The unspoken assumption behind converting goal difference into standings points is that over team, a team of X ability scores a proportionate amount of meaningful and meaningless goals. This falls apart with fourth lines, which tend to play a disproportionate amount of their minutes during times when a game is out of reach one way or another. In other words, because of when Acton/Gazdic/Brown will play their minutes if they’re kept together all year, the difference between them and the average fourth line is less meaningful than it appears – some of the extra goals that they’ll allow will be meaningless. When you allow for that, the gap between the Oilers fourth line and the hypothetical average fourth line becomes smaller.

    So to me, the criticisms as currently framed kind of silly. Would I like the Oilers to have better players on the fourth line than they currently have? Yes. I’d like them to have better players on every line than they currently have. “Is the fourth line costing them significantly at 5v5 relative to a better fourth line?” is a better question to ask. I doubt that the answer is yes. “Can the Oilers play three lines enough to make up for having a poor fourth line that they don’t play a ton?” is a better question to ask and one that I haven’t heard seriously discussed by anybody.

    There was a moment when Dallas Eakins was talking about RNH’s return and how much he’d play and he said something like “We don’t want to push him into deficit.” It struck me as being a very physiological comment, like he was talking in terms of how much a player has him in his bodywise. The consensus about him seems to be that he’s a physical fitness nut. I’d bet he’s turned his mind to this issue.

    Two more points. One, it wouldn’t stun me at all if, at some point in the season, the Oilers added a fourth liner who can kill penalties to give Eakins a little more flexibility with his roster. Gazdic is bigger than Mike Brown and possibly a better hockey player, although that remains to be seen. He seems to me to make fewer outright goofy plays. Gazdic’s also a MacT property whereas Brown isn’t. MacT’s got no history of a fourth line like the one that the Oilers have now and I wouldn’t expect it to be like this all year long.

    Two, I think people are drawing conclusions on Will Acton a little early, possibly because of his last name. I haven’t thought he’s looked that bad – I was watching for him during the Toronto game – and I think a large chunk of his poor Corsi% is tied up in who he’s on the ice with. Take this:

    That’s one shift but it’s hard to really tie anything to Acton there despite there being a number of SAA. Gazdic and Brown do a series of odd things in there which are likely to related to lack of skill/being uncomfortable at the NHL level in Gazdic’s case (skipping easy passes to just club the puck out) and I don’t know what with Brown – passing to Gazdic on a 3 on 2 at the end of a shift that they’ve barely survived against Kessel rather than gaining the red line and chipping the puck in for a change.

    I’m not saying that Will Acton’s a player – I’m undecided – only that those are tough circumstances in which to look like an NHLer. He hasn’t looked nearly as out of place to me as his wingers do, although I’m open to changing my mind as the season progresses.

    In any event, the fourth line’s unlikely to have a major impact on where the Oilers finish this year. To the extent that they’re discussed, it should be in the context of how the makeup of the fourth line affects the penalty kill and imposes demands on other lines in terms of ice time. They aren’t as good as the average fourth line but then it probably doesn’t matter that much.

    Email Tyler Dellow at tyler@mc79hockey.com

    About

    14 Responses to The Fourth Line

    1. October 15, 2013 at

      I largely agree with what you’ve said here about the impact of the fourth line. They’re awful, and it sure would be nice if the Oilers made better use of those roster spots, but they don’t play enough to make a huge difference.

      But given that it would be nice to see better players, I’m surprised you’re so open about the decision to keep Acton around. It’s not often that a guy is a complete non-factor offensively in the AHL in his mid-20s only to become a solid NHLer. If some team suddenly decided that Chris VandeVelde was a reasonable bet for NHL minutes, I wouldn’t need to see him play a shift before being awfully skeptical. If he then proceeded to put up awful results in his first few games, I wouldn’t be quite so generous as you about giving time before passing judgment.

      • Tyler Dellow
        October 15, 2013 at

        Scott -

        Vandevelde looked like a horrible hockey player to me. Acton hasn’t. I also thought Vandevelde had better wingers – Petrell and Eager or whoever was on teh fourth line, bad as they are, aren’t as bad at playing hockey as these guys.

        • October 15, 2013 at

          I picked on VandeVelde because he and Acton are very similar players offensively (and by role) at the AHL level. I think the AHL data is relevant in helping us draw conclusions, and the AHL data for Acton isn’t usually “NHL player” stuff. I’d be quite surprised if he was able to deliver as a quality NHL player based on that.

          Point taken that you’ve liked Acton better than VandeVelde, though I’m not really seeing it. He seems like he’s either not able to process the game quite quickly enough or that he doesn’t have the footspeed to get to where he needs to be quickly enough. Admittedly, it’s just a few games, his linemates have been awful, and my eyes do fool me sometimes. I’ll be sure to pay closer attention to him tonight.

    2. Woodguy
      October 15, 2013 at

      . “Can the Oilers play three lines enough to make up for having a poor fourth line that they don’t play a ton?” is a better question to ask and one that I haven’t heard seriously discussed by anybody.

      Wasn’t this concept widely discussed as a key reason for the Leaf’s collapse against BOS last year?

      Grabbo’s line and Bozak’s line were gassed and the 4th was stapled to the bench and bad things kept happening?

      How do you measure if 25 minutes of Taylor Hall is actually less effective than 21 minutes due to exhaustion?

      It would be interesting to figure out.

      Also,

      In a league where success or failure falls of the regular season can be a matter of 1-2 points in the standings, doesn’t it behoove the management to try for every incremental advantage possible?

      Oilers were one MAB brainfart away from winning the Cup when they made the playoffs by 3 points. LAK made it by 5 when they won.

      Just making the playoffs this year would be of immense benefit for the Oiler organization, players and the sanity of the long suffering fans.

      When you look at the 4th lines for the past 3 Cup champions we see this for TOI in the playoffs: (using 5v5 toi/gm)

      CHI 2013 – Stalberg, Kruger, Frolik (10:08-10:23)
      LAK 2012 – Richardson, Fraser, Nolan (7:36-8:00)
      BOS 2011 – Seguin/Thornton, Campbell, Paille (7:59-9:32)
      CHI 2010 – Brower, Madden, Eager (8:30-9:00)

      I know Pat has been beating the drum that when you have 4th line that can play like CHI’s did last year it gives your coach the option of playing them during meaningful minutes.

      It also gives the coach the option of using them if other lines are having tough nights.

      As Pat has pointed out, when CHI had to hold the lead to win the Cup Quenville went Kruger-Toews-Kruger in terms of lines that saw the ice.

      CHI’s 4th line was going and it gave the coach the option to use them.

      During the regular season guys like Frolik and Krueger also had CF% at 55% or so.

      So when 89 gets healthy they will “probably” have a top 9 of:

      LW 4-57-6
      C 93-89-27
      RW 83-14-64

      That leaves 26 and 94 available for the 4th.

      Will putting a 4th that looks like 94-26-puncher (Oilers want one, so one is there), be better than puncher-Action-puncher?

      If its yes, then don’t you have to do it to gain whatever incremental advantage is available?

      If it turns out that running your best horses for 24+ minutes results in lower quality minutes, then a 4th that doesn’t bleed becomes even more of a necessity.

      • darin
        October 15, 2013 at

        Yes. This post is exactly right.

        I wonder how such a ‘objective data’ proponent as Eakins is, will continue to defend bad hockey players. San Jose appears to have given up on face punching in favour of more balanced minutes for 4 good lines. I don’t think we should say “wait till the playoffs”, I think we should appreciate dominant hockey and assume we can figure it out in the playoffs if we are dominant in the regular season.

    3. MAttM
      October 15, 2013 at

      The nice thing about the 4th line problem is how fixable it seems. If there’s a relatively simple way to add 1W to the team’s expected total for the year, why ignore it? Just because it’s not the biggest reason they’re struggling doesn’t mean it’s not worth fixing. I mean, trading an asset to fix it would probably be dumb for the reasons you’ve outlined, but just sliding a couple of better players that you already have on the 50 man list to get a 1W improvement? Why the hell not?

      I think the reason people are so annoyed about this is not that it’s a huge problem. It’s that it’s a small problem that could be fairly easily corrected.

    4. Pingback: Spectors Hockey | NHL Blog Beat – October 15, 2013.

    5. Saj
      October 15, 2013 at

      Nice job taking a run at quantifying what the deficient fourth line means for the team’s point total. A few thoughts:

      1) I’d imagine there’s a high standard deviation to that “3 goals to 1 point” rule of thumb you refer to. Conceivably, all four goals could turn wins into losses here, which could make the impact as high as 8 points.

      2) This one decision (a 4th line from hell) may not turn out to have a huge impact, but what does it say about the team’s process? A good management team would be taking every advantage it can, and all those “expected points” would accumulate, leading to a better team. What other personnel decisions were made on the basis of “toughness over ability as a hockey player”? Was Grabbo not signed because not tough enough? Was Gilbert not signed because considered too soft?

      3) Finally, your point is a good one that 4th line impacts may tend to come at garbage time, but I also think there may be some disadvantages also not captured by the above analysis, such as powerplays granted to the other team (through both poor positioning and poor decision-making) and putting the other three lines in worse zone start position.

      • Triumph
        October 16, 2013 at

        A goal can’t turn a loss into a win. It can only turn a loss into a tie or a tie into a win. So they’d be worth, at most, 6 points. Even so, this method of analysis is fruitless, because like Tyler said, 4th line goals tend to come in extremely low-leverage spots – it’s more likely that they’d be worth 0 points than 6.

    6. Andrew
      October 16, 2013 at

      Thanks! Glad someone else realizes this. 4th line play is never the main reason for a team losing. In the Oilers case especially, they have 3 lines that can play top 6 minutes. You typically want a breakdown of 20-17-15-8 discounting special teams play. The Oilers could go with something like 20-18-18-4 and have just as much success provided guys in the top 9 stay healthy enough to allow it. I see no reason for 4th line guys to play more than 6 mins in games that are not blowouts. In that case, or like the Wsh game with a game the next night, it might be wise to give them more minutes in the 3rd of a game out of reach. I appreciate some stats to prove my opinion to be somewhat valid. Nevertheless I don’t see a reason to need both Brown and Gazdic, one of them should be plenty possibly having a guy who could pk a bit on the other wing.

      • dawgbone
        October 17, 2013 at

        It’s an 82 game season that’s compressed more because of the Olympics.

        Having proper depth in your lineup is very important, so having guys you can only trust for 4-5 minutes per game puts you in a hole right off the bat.

        Further to that, the Oilers are using a bunch of guys in PK roles who have never done it before and they are bleeding really badly there.

        Sure, if you can hide a 4th line for 5 minutes a night for 82 games they might not hurt you too much (as a line) over the course of the season. The problem is your other 3 lines have to pick up the slack and they might be the ones that end up costing you.

        • Andrew
          October 17, 2013 at

          But my point is that the top 3 lines can handle it. I think the main worry should be on the back end and in net at this point.

          • dawgbone
            October 17, 2013 at

            For how long can they handle it?

            There’s a good reason why the consistently good teams (especially in years they make deep playoff runs), have 4th lines they can rely on.

            And the backend isn’t as much of a problem as is the play of all 5 skaters in the defensive zone. It’s not just the blueline that has to make defensive plays, there are 3 forwards who do as well.

            And again, instead of having face punchers on the 4th, why not guys who can at least kill penalties?

    7. Pingback: Islanders News: Clutterbuck on hitting; Strome patient; Spano film | All New York Islanders

    Leave a Reply

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