One of the funny things about sports is that things that are counted draw an inordinate amount of attention. Much to the aggravation of the analytics community, this has been the case with faceoffs in hockey. A huge amount of attention is placed on winning faceoffs; comparatively little is placed on what happens afterwards. That, I suspect, is due to a dearth of information about what happens afterwards.
Advertisement
The immediate aftermath of a faceoff is particularly interesting because it’s one of the rare occasions on which a coach can really dictate what happens in hockey. He knows where all the pieces are going to be when the puck is dropped, enabling him to put into place a strategy for how things will be done. Hockey’s organized chaos most of the time but this is a rare opportunity to really set out what will happen with knowledge of where players will be. If you pre-scout well, you should have a good idea what the opposition might throw at you too.
The frequency of 5-on-5 offensive zone faceoff wins has increased over the past few years. On a per 82-game basis, it’s gone about 520 per year at the start of the decade to just under 600 last year, a slight step back from 2016-17. In order to look into this a bit, I took a look at what happened in the 10 seconds after a 5-on-5 offensive zone faceoff win since 2010-11. Ten seconds is an arbitrary figure, but you need some sort of a cutoff and that seems as good as any.
The rate at which goals are scored in the 10 seconds after an offensive zone win hasn’t really moved all that much over the past decade, hovering around 5.0 GF/60. This past season, teams scored 4.9 GF/60 after an offensive zone win. Between 2010-11 and 2016-17, GA/60 fluctuated between 0.32 GA/60 and 0.46 GA/60. That number took a bit of a jump this past season, to 0.58 GA/60.
Intriguingly, the rate at which teams generate shot attempts has trended up, moving from 157.7 CF/60 in 2010-11 to 184.1 CF/60 in 2017-18. The rate at which teams allow shot attempts in these 10 seconds has been pretty constant, with a low of 5.3 CA/60 in 2012-13 and a high of 6.8 CA/60 in 2011-12. Last year, they allowed 6.6 CA/60. So, at the risk of stating the obvious, a team winning an offensive zone faceoff has a hell of advantage in the next 10 seconds. Since 2010-11, they’ve been +4.5 goal differential per 60 minutes in those ten seconds.
Advertisement
While per 60 metrics are something with which we’re all pretty comfortable, I’m going to do something a little different here to try to make the differences between coaches who’ve had success and those who haven’t easier to grasp. The average team is winning about 600 5-on-5 offensive zone faceoffs per year at the moment. Thirty-six coaches have been on the bench for at least 1,200 5-on-5 offensive zone faceoff wins since 2010-11. Here’s how those coaches stack up in goal difference per 600 offensive zone faceoff wins in that time. As always, the red line represents league average.
There’s all sorts of fascinating stuff in here. As always, it’s the extremes that are most interesting. Peter Laviolette and Michel Therrien stand head and shoulders above the rest. I’ve actually written about Laviolette and offensive zone faceoff wins before. As we’ll see, his team had another big year. Laviolette’s teams have been a three-goal difference a year better than the league average team in these 10 seconds. If that’s due to his coaching, he’s adding a point in the standings over the average coach. That can’t be easy to do. Peter DeBoer stands at the top of the rest of the pack. Jack Capuano catches my eye here, too, for reasons that I’ll get into below. Mike Babcock’s numbers here are pretty average although, as I’ll discuss below, there’s been an interesting uptick lately.
At the other end of the scale, there are some remarkably successful coaches. Barry Trotz’s presence is interesting because Laviolette took over Nashville and the Predators started putting up good numbers. I’m surprised by Todd McLellan’s presence at the bad end of the list, given that McLellan has a long history of power play success, which seems to call for the same sort of thing in terms of coaching to set up high-value opportunities. The absolute stunner for me is Joel Quenneville. He’d be on any short list of the best coaches of the past 20 years, but his teams are persistently terrible in this situation, despite him having coached two Stanley Cup winners in this window. It’s bizarre.
For the sake of completeness, I’ve broken this list out based on goals scored and allowed.
I’ll use this year’s data as a jumping off point to get into some individual coaches in a little more detail.
Edmonton had a lot of stuff go historically wrong last year, but this basically defies belief. The difference between the Oilers and the Red Wings is the same as the difference between the Red Wings and the Sabres, who were in the middle of the league. Edmonton didn’t score a single goal within 10 seconds of a 5-on-5 OZW all year, which is incredible, given that they have 5-on-5 offensive machine Connor McDavid. They somehow conspired to allow four goals in this situation, the worst number of the decade. They scored none; the only other team to do this since 2010-11 was the 2012-13 San Jose Sharks, who played just 48 games.
Watching the goals that the Oilers allowed, they all seemed to start from losing pucks in that grey area between the forwards and the defencemen on an offensive zone faceoff win. Then the race was on and it frequently ended poorly.
Now, these things will happen. And it’s certainly possible that a team could just have some bad luck here and give up some goals. There’s another interesting bit of information that suggests that this happened a lot to the Oilers though. The rate at which the Oilers allowed shot attempts here was more than double the league average for this situation. (Montreal was also horrifically bad at this, for some reason.) The problem with allowing shots in this window is that they’re going to be dangerous ones – off the rush, frequently in an odd man situation. Since 2010-11, teams have an .897 save percentage on shots allowed within ten seconds of an offensive zone win, which reflects that.
Advertisement
In the aggregate, Todd McLellan’s teams have been nothing to write home about here, as can be seen above. From 2013-14 through 2016-17 though, they were actually pretty good. This happened after a year in San Jose in which his teams didn’t score a single goal, so you sort of wonder if McLellan went away in the summer and reconsidered his approach. Last year was a horrific aberration when compared to the previous four seasons. I’m a coaching apologist by nature, so take that into account, but I’d be shocked if McLellan and his staff can’t sort this out for 2018-19.
The Blackhawks were awful *again* in this situation, which is pretty much what they’ve been for the past decade. It’s baffling to me that they’ve struggled so much here, given that they’re laden with talent and that Joel Quenneville is a hell of a coach. I wondered whether the Blackhawks were maybe doing something that reduced their chances of scoring in the first ten seconds but upped their chances later but if you expand the window to twenty seconds, Quenneville’s teams are still below average. Last year was, obviously, particularly bad.
It kind of looks like the Blackhawks are aware of their struggles here and have tried to take steps to deal with it. Their shot rate after winning a 5-on-5 OZW has spiked over the past two years – their shot attempt rate is up nearly 30 per cent in the past two years compared to the preceding six years. It hasn’t really paid off in results though. Chicago’s a team that I think has made some investment in analytics and has the capacity to do a smart best practices study here. With the Blackhawks being a bubble playoff team, this might be an area where investing some time could have real payoff.
Speaking of candidates for best coaches of the last twenty years and teams with smart analytics groups, Mike Babcock’s had an interesting progression. In the 2010-11 through 2014-15 seasons, Detroit generated 140.2 CF/60 and 6.6 goal difference per 60. The Red Wings weren’t moving the needle in this situation. Since coming to Toronto, the Maple Leafs have taken 193.8 CF/60 and generated 9.8 GD/60. Even the terrible 2015-16 Maple Leafs were good at this, after being awful at it in 2014-15. You can reasonably surmise that Babcock and, quite possibly, the Maple Leafs’ smart analytics group have something to do with this.
We don’t talk about the Florida Panthers all that much here but their presence on this list is notable. Assistant coach Jack Capuano’s teams tended to be good in this situation, as you can see in the first chart. Head coach Bob Boughner did his apprenticeship on an NHL bench with Pete DeBoer, another coach with a long history of success here. Watching the Panthers score goals, I was struck by how they created space with passing and movement from the defencemen.
Meanwhile, elsewhere in Florida, the Lightning were pretty good at scoring goals from OZW too. Again, what stands out to me is the movement amongst the Tampa players when the puck is dropped but before they’re shooting. The movement does two things – it pulls the defence around, creating the possibility of blown coverage and gets the goalie moving too.
Finally, let’s check in on Nashville and Peter Laviolette, the king of generating goal difference from offensive zone faceoff wins. There’s something interesting happening here. When I looked at the Predators in 2017, I marveled at how they generated drives down the boards, followed by passes into the slot. Looking at their goals this year, that wasn’t really how they scored them. Plays from the point seemed much more important.
Advertisement
Looking at the data, the way in which Nashville’s offence worked after 5-on-5 OZW changed last season. In 2016-17, Nashville’s defencemen generated 49.7 per cent of their shot attempts in the window I’m looking at, sixth lowest in the NHL. Last year, they generated 58.3 per cent of the shot attempts, good for 12th in the NHL. The range is pretty big here, something that seems to me like it would be useful for coaches doing their pre-game preparation to know. San Jose’s defencemen – hello Brent Burns – generated 68 per cent of their shot attempts after OZW. For the Rangers, they generated just 47.9 per cent of the shot attempts.
Watching the goals that the Predators scored last year, there was a lot of east/west movement of the puck amongst the defencemen. If there’s a formula here, “puck movement and screens” seems like a handy shorthand description of it.
When the player and puck tracking comes online, there’s a real opportunity for teams that are set up to deal with it to turn that data into knowledge that helps win hockey games. You’ll be able to build entire profiles of what teams and players do in these situations and really see what works and what doesn’t, as well as know exactly what you’re likely to see from an opposing team without the current constraints of needing data like that cited in this post that a lot of teams probably can’t easily access and then needing to watch a lot of video.
Even with that, there’s probably a lot that clever teams can do now to squeeze more out of these opportunities. We’ve seen coaches get radically more efficient with power plays. This strikes me as another area that’s ripe to be improved upon by those teams and coaches who have struggled.
(Main photo: Peter Laviolette. Credit: John Russell/NHLI via Getty Images)