• Boudreau: Offensive Zone Faceoff Loss Witch

    by  • April 16, 2014 • Hockey • 8 Comments

    I wrote a couple of days ago about Bruce Boudreau and the fantastic results that his teams achieve following 5v5 offensive zone faceoff losses. Here’s the key paragraph:

    Basically, every Boudreau team is above average after losing offensive zone faceoffs – most are well above average – until Defence Wins Boudreau comes to town. If you think about what it means installing the trap, it was probably tied to other defensive changes, like retreating towards the neutral zone once possession was lost in the offensive zone – like, say, after an offensive zone faceoff loss. Sure enough, there it is.

    Then, the other day, I wrote about the impact that coaching changes involving Randy Carlyle have had in Anaheim and in Toronto. In the course of doing that, I put together this table:

    The change in results following offensive zone faceoff losses is mindblowing. I’ve got a database for this stuff from 2007-08 through 2012-13. The average Corsi% in the 21 seconds following offensive zone faceoff losses is about 54.8%. Boudreau’s teams (save and except his Blue Period, when he trapped) have routinely blown it away.

    Just to give you a sense of the spread, only 16 of the 180 teams in my database have posted seasons with a Corsi% following OZ losses that’s greater than 61%. Only eight teams have beat 64% – three of them were Boudreau teams. Three! At the time of writing, I didn’t have results for 2013-14. I went ahead and calculated them for this year for Anaheim. Sure enough, Boudreau rides again: 64.8%.

    It’s amazing. I wanted to dig a little deeper into it, as I’m occasionally prone to doing. What I’ve done is just graph how many shot attempts 100 offensive zone faceoff losses would generate for the 2010-11 Carlyle Ducks (who were a little below average in terms of the post-OZ loss Corsi% but not terrible), the 2012-13 Boudreau Ducks and the 2013-14 Boudreau Ducks.

    As you can see, through the first four or five seconds, all three teams are pretty similar. From that point forward, the Boudreau Ducks explode. If you ignore what happens in the first five seconds, Carlyle’s team generated 14.7 shot attempts per 100 faceoff losses from seconds six through 21. For Boudreau, it’s been 23.2 and 25.1 in 2012-13 and 2013-14 respectively.

    This pretty much has to be a tactical thing. The Ducks turned around the instant that Boudreau replaced Carlyle. What’s more, the effect is not insubstantial. Boudreau’s Ducks generated 28.5 shot attempts per 100 lost offensive zone faceoffs this year. The NHL average from 2007-13 was 19.6. The average team lost 525 offensive zone faceoffs per 82 games. So we’re talking about an extra 47 shot attempts that the Ducks generate relative to what an average team would generate. If half of those are shots and 8% go in, there’s two extra goals for.

    Of course, that’s only half the equation. Who do you think gives up fewer shot attempts in the 21 seconds following offensive zone faceoff losses? The buttoned down, defence first Randy Carlyle or the damn the torpedoes, offence first Randy Carlyle?

    Yeah – it’s the defensive coach who gives up more shot attempts in this situation. As we all expected. Boudreau’s edge relative to the league isn’t as big here – the Ducks had a good run last year – but there’s another couple of shot attempts that he tips the scales in the Ducks’ direction.

    I’m inclined to give Boudreau a lot of the credit for this because he’s got a track record and he seems to be able to make it happen instantaneously. It’s a very valuable skill – buying three goal difference outside of the salary cap should be worth an awful lot of money to teams. And it’s one, tiny, fragmentary part of the game.

    I’m inclined to think that he’s a very good coach. It’s pretty cool that we’re getting to the part where we can use the data that the league gathers to illustrate a coach’s impact.

    Email Tyler Dellow at tyler@mc79hockey.com


    8 Responses to Boudreau: Offensive Zone Faceoff Loss Witch

    1. Locky
      April 16, 2014 at

      Typo in the last sentence before the last graph mate. Good stuff though!

    2. Murat
      April 16, 2014 at

      This is so neat.

      I assume that the next step is to then look at video following an offensive zone loss for any Boudreau with him and without him at the helm. The answers would probably be found in the movement of his x’s and o’s, which is really cool.

      The biggest problem in evaluating the impact of coaching has always been separating that impact from all the noise that surrounds it. Jack Adams trophy nominees are generally coaches of teams with surprisingly good records but no one cares if their teams got outshot, if a goalie played out of his mind, or they got all of the puck luck – there is so much noise here that the Jack Adams stories are almost not worth following.

      The way forward is to isolate game states, as you have done, and (painstakingly) compare how different coaches approach them. Hockey is fluid, sure, but faceoff situations like you’ve identified repeat over and over again. Surely there are other comparable, standardize-able game states to dig into.

      I’m not sure this adds to your post, so much as summarizes the value in it and kind of cheerleads a bit about the era we’re getting to. Combine that with Corsi on TV the other day and it’s a pretty cool time to be a hockey fan.

      • Murat
        April 16, 2014 at

        correction for sentence 2: for any Boudreau team*

    3. Danton
      April 16, 2014 at

      This is really great work. I’m curious – why the 21 second time frame following a face-off? Not disputing it, just wondering what’s behind that.

      • Babaganoosh13
        April 16, 2014 at

        I would assume that eventually a game would get to a point where the face-off that started it would basically be almost meaningless.

        • Danton
          April 16, 2014 at

          Yeah, that makes sense, I just wonder why we’re talking about 21 seconds (as opposed to 15 or 25 or whatever).

    4. Tyler Dellow
      April 16, 2014 at

      I use 21 seconds because my research has shown that that’s how long it takes for the game to get to the point that the effect of the location/result of an OZ faceoff loss to wash out.

      • Simon Lamarche
        April 16, 2014 at

        This is very interesting.

        While your work on faceoffs showed that teams get back to an average (50%) Corsi% a bit after 21 seconds and it seems to be the case for Carlyle’s teams, Boudreau’s teams still haven’t come back down to below 65% by that point. So basically, while 21 seconds is the average time it takes, it looks like that also depends on systems.

    Leave a Reply

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