• Pyrrhic Faceoff Wins and Refining Problems

    by  • July 15, 2014 • Hockey • 3 Comments

    A few months back, I wrote a pair of posts about what I referred to as the Oilers’ Pyhrric faceoff wins in the offensive zone. There were a few different observations tied up in there, which I can summarize like so:

    1) Sam Gagner and Ryan Nugent-Hopkins won a lot more OZ faceoffs after December 7, 2013;

    2) Despite the Oilers winning more OZ draws with those guys on the ice, they both did much worse following OZ faceoffs (whether won or lost) in terms of Corsi%;

    3) They really struggled following OZ faceoff losses;

    4) But they declined after winning OZ faceoffs too.

    I had a paragraph that touched on precisely what happened following those OZ wins. This was the key portion.

    There’s a decline in Corsi%, enough to make the Oilers a pretty ineffectual team when they won an offensive zone faceoff at 5v5 over the last 51 games of the year – that 68.6% figure stinks when you’re starting with the puck in the opposition’s end. The interesting thing is the why – the rate at which they allow shot attempts following offensive zone faceoff wins actually declined, from 0.323 shot attempts per OZ faceoff win to 0.292. The rate at which they generated shots following OZ faceoff wins plummeted, from 0.848 to 0.646.

    At the time that I wrote this my thinking (which I didn’t divulge but which a lot of commenters guessed) was that this whole thing was tied up with the Oilers doing something to try and help Gagner and RNH win some more faceoffs. I’ve generated some additional data now that I think sheds more light on the issue.

    As I said at the time:

    The drop for Gordon is insignificant – it’s one Corsi event going the other way. The drop for Gagner and RNH is massive. All of the decline is focused around them, which is consistent with my theory that the Oilers were doing something to try and help them out and it backfired badly.

    As I’ve thought about this some more and looked at some data in different ways, I’ve kind of refined what I’m thinking about this. When I’m thinking about problems like this, I like to take a look at it in what I call a longitudinal way – taking a look at the timing of things to see if there’s a specific time point post-faceoff at which things change.

    As I noted above, the big change in the Oilers’ post-OZW Corsi% after December 7 lay in a decline in the volume of shot attempts that they generated. They actually allowed fewer shot attempts post-December 7 but they generated so many fewer that the Corsi% fell. Let’s take a look at their shot generation second-by-second post faceoff win. I’ve split this into three groups: overall, when Gagner/RNH took the draw and when someone other than Gagner/RNH (generally, this is Gordon) took the draw.

    Here’s what overall looks like:

    Screen Shot 2014-07-15 at 1.36.38 PM

    I’ve included the NHL average because it gives us something to compare things to. You can see that through December 7, the Oilers were pretty good: better than the league average. After December 7? Ugh. They went from generating 84.8 shot attempts per 100 offensive zone faceoff wins – very good! – to generating just 64.6 shot attempts per 100 offensive zone faceoff wins – very bad!

    Screen Shot 2014-07-15 at 1.37.12 PM

    It’s hard to attribute much blame for this decline to the draws that weren’t taken by Gagner or RNH. There was a decline in terms of shot volume there but it’s not huge – less than 10%. That could easily be a mix of factors, such as fewer non-Gagner/RNH draws that had star forwards on the ice who generate shots (remember – Gagner missed the first month of the year) or chance. I don’t think that this is the source of the problem. For that, we’ll look at what happened following Gagner/RNH draws.

    Screen Shot 2014-07-15 at 1.37.29 PM

    Well. This is a little bit different. Through December 7, the Oilers are generating piles of shot attempts when Gagner and RNH won an offensive zone faceoff. After December 7, they become a tire fire.

    Taken by itself, this doesn’t help us isolate the problem sufficiently. It suggests that the trouble arose when Gagner or RNH won the faceoff but it doesn’t give us a very good sense of what, precisely, went on. In order to dig into that, we need some more granular data about what happened after Gagner/RNH won an OZ faceoff. I’ve generated this data by re-watching the offensive zone faceoff wins that they were on the ice for and tracking when the puck entered/exited a given zone on the ice and how it did so.

    First things first. Every offensive zone faceoff win generates a possession in the offensive zone. You don’t need to carry it in or retrieve a dump-in: you’re already there with the puck. That makes these possessions sort of special, in that they should be shot attempt heavy possessions that contain the bulk of your shot attempts in the 37 seconds following an offensive zone faceoff win.

    That is, in fact, what the data shows. The Oilers generated 66 shot attempts following 68 offensive zone faceoff wins by Gagner/RNH through December 7; 71.2% were on the initial possession, before the puck was cleared. After December 7, they generated 124 shot attempts from 200 offensive zone faceoff wins by Gagner/RNH; 79.8% of those were on the initial possession.

    The weird thing, or the sign that there’s a problem, is that the rate at which they generated shot attempts on the initial possession itself plummeted: the Oilers fell from generating .691 shot attempts per Gagner/RNH OZW to .495 shot attempts per Gagner/RNH OZW. That’s nearly a 30% drop.

    It may be that this is nothing; chance. It’s a pretty big fall though and what it suggests to me is that it’s worth examining these initial possessions in closer detail, to see whether the Oilers had changed what they did in an order to help Gagner/RNH win more draws with an unintended consequence of a few seconds down the road, having guys in spots on the ice that were less helpful, driving down the shot totals.

    Just as an aside on this – I’m not sure that hockey coaches can really say “If A then B but if C then D but if E then F” when instructing their players. The game’s too fluid. I’ve been re-reading Jonathan Wilson’s fine book about Brian Clough lately and he spends a lot of time talking about how Clough, one of England’s greatest soccer managers, managed his teams.

    Clough – almost to a fault, in Wilson’s view – didn’t really focus on “If A then B but if C then…” type stuff with his players. Instead, he paid attention to what soccer people call “the shape” – having his team maintain a formation – and having players who were good at doing what he wanted within the shape – forwards who’d turn quickly and run at defenders, defenders who could play the ball, that sort of thing.

    I mention this just because hockey’s just soccer on speed in a lot of ways. The challenge a coach has here is not, I think, in telling players “If A then B but if C then…” but figuring out the best “shape” in terms of the shape that results in the most players naturally being in favourable positions as the initial possession from the OZW plays out. I think it’s fair to wonder if the Oilers were getting that right as the year wore on. There’s a definite red flag here to me that a significant portion of the decline was tied to what happened on the initial possession.

    Of note: if the Oilers had kept generating shot attempts on the initial possession off RNH/Gagner OZW as they had through December 7, it’s likely worth another 1.4 goals or so. That’s also likely an underestimate – more shot attempts means more faceoffs means more shot attempts – it’s a virtuous circle.

    I’ll cut this post off here. Tomorrow, we’ll talk about what happens after the puck leaves the offensive zone.

    Email Tyler Dellow at tyler@mc79hockey.com


    3 Responses to Pyrrhic Faceoff Wins and Refining Problems

    1. Fraser
      July 15, 2014 at

      Love this series!

      The graphs tell an interesting story. It looks to my eye that the first 6 seconds are pretty constant pre and post Dec 7th. This suggests to me that whatever this “different strategy” is it seems to be putting players out of position if the set play falls apart. Any thoughts on why this could be?

    2. jvuc
      July 16, 2014 at

      Eakins also changed his defensive strategy from the swarm to a more traditional system. It is hard to isolate what the numbers mean

    3. Mark
      July 16, 2014 at

      Excellent article, as always. I have a question though:

      “if the Oilers had kept generating shot attempts on the initial possession off RNH/Gagner OZW as they had through December 7, it’s likely worth another 1.4 goals or so”

      In order to generate shots at the rate they did through December 7th then they would go back to winning a lower percentage of faceoffs. Does the 1.4 goal estimate take that into consideration?

    Leave a Reply

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