As a part of my presentation last weekend at the Vancouver Hockey Analytics Conference, I debuted a new power play chart that I feel presents a good view of each team's power play performance and success as the season has progressed. It will take quite a bit of explaining so let's dive in.
From past analysis, we know that getting shot attempts off on the power play is a decent predictor of future success. Unfortunately, that predictive ability has waned in recent years, to the point where goal scoring efficiency is maybe just as useful for analysis. Through my tracking this season, I have established that the time a team spends in formation in the offensive zone is hugely important (more on this next week). With these charts, we can visualize all of this and more.
Let's start with the Toronto Maple Leafs. Take a glance at this, and then I will explain what you are seeing.
The Leafs were the only of my six tracked teams that I didn't stick with for the full season. In the interest of eliminating from my sample games in which Mark Arcobello and Colin Greening qualified as prominent power play performers, I stopped tracking the Leafs after 54 games, just prior to the trade deadline. That said, here is what you are seeing in the chart above:
The bottom half of the chart features 5-game rolling average shot attempt per 60 minutes numbers (yellow solid line) and goals per 60 minutes (red solid line) over the course of the season, all at 5-on-4. The top half of the chart displays the proportion of total 5-on-4 team for the team spent in each of four situations:
Regrouping: Time spent outside of the offensive zone. You want to minimize this as much as possible.
OZone Out of Formation: Time spent in the offensive zone not on the rush and not in formation. You probably want to limit this as well as it suggests that entries aren't being maximizing to put players quickly in position to succeed.
OZone In Formation: Time spent in the offensive zone in a strict recognizable formation. Teams register shots and goals at far higher rates when in formation than when not.
Rush: Time spent on the rush (an approximation)
The dotted lines in all cases represent league (six-team) averages for the full season based on my data. Note that games in which the team receives zero 5-on-4 opportunities are excluded, which is the reason why the game numbers don't quite go up to 82 despite the entire season being tracked for five teams.
So what does this tell us about Toronto at 5-on-4 this season? Despite minimal talent, Mike Babcock implemented a structured system that revolved around three features. The first is the drop-pass entry. Because of that lack of skill, and the fact that drop passes tend to take longer to execute than regular entries, the Leafs spent a larger than average proportion of time regrouping, represented in red. Nazem Kadri, Tyler Bozak and P.A. Parenteau were especially important to these entries, so the period in which Bozak was hurt was something of a struggle on that front.
The second feature was a give-and-go entry play from the primary puck carrier to one of the wingers that produced consistent rush attempts. As the chart shows, this play led to an above average number of seconds on the rush.
Finally, the Leafs played a structured 1-3-1 system with a focus on clean puck movement on the perimeter and screened outside shots (slappers from the point or wristers from the half-wall). The prominence of the blue area in the chart shows that the Leafs did a good job setting up in formation and you can see that in those rare periods in which the team failed to do so, their shot rate tumbled -- this will be an ongoing theme.
The Canadiens power play started the season out pretty well playing a traditional overload formation with Alex Semin on the half-wall. The problem is that their shot numbers and entry success -- particularly in games (approximately) three to 10 -- didn't live up to their goal scoring, and the goal numbers soon tailed off. Struggling to turn entries into formation time or generate much in the way of rush opportunities devastated the team's units all year long, and it's no surprise that with injuries to Brendan Gallagher, P.K. Subban, Jeff Petry, and others late in the year led to a downward spiral in zone time, shots and goals.
I noted the stretches that Gallagher missed in particular because they seemed to line up with periods in which the team was especially woeful with the man advantage. Obviously the undersized winger plays a big role in front of the net, but I'm not sure why that would affect entries and formation time as much as it seemed to.
The Tampa Bay Lightning power play -- a woeful unit all year thanks to a lack of structure and inconsistent personnel and role choices -- recently devised an alteration in philosophy, catalyzed by the injury to star scorer Steven Stamkos. As Prashanth Iyer wrote about in his great playoff preview series here, around the 65 game mark, Tampa changed its power play ideology from one of finesse and shot quality to one of screens and shot volume. As one would expect, their shot rate increased dramatically even as their formation time was stagnant. Jason Garrison and Nikita Kucherov are the team's primary one-timer weapons with Brian Boyle, Ryan Callahan and Alex Killorn providing the screens (a feature the team rarely employed earlier in the season). Once again we see the correlation between time spent in formation and shot rate.
I wrote about the rise of the Ghost previously, but this chart shows the immediate uptick in both shot attempts and goals he catalyzed. Prior to his arrival, on all counts the team was at a seasonal low, and it can't be understated how important his emergence was for that unit. Lately, you can see that the team's drop-pass focused entry scheme has struggled at times, though its escalating shot rate in the last few games is promising facing a top-ranked Capitals penalty kill (and going head-to-head with their towering man advantage).
And now the gold standard. Once again you can see the importance of spending time in formation for the Capitals, an uber-talented yet super-structured team up a man. The Caps came a hair short of leading the league in 5-on-4 GF60 for the fourth straight year, failing to do so because of a poor finish and several significant slumps in the second half. Around the 30 game mark, this team was firing on all cylinders, with an entry efficiency I doubt any other team this year managed to replicate even over a short period, but as the year sagged on there were issues, especially when John Carlson went down to a couple of significant injuries. The Capitals will need to regain their entry polish -- not only entering the zone successfully but efficiently transferring those entries into formation time -- to go deep in the post-season. This video is a good guide to what you will see on entries from the Capitals in the playoffs. If they look as crisp as this, the team's in good shape.
Here's a study in contrast. As I mentioned in Vancouver, the Islanders don't place a priority on getting set up in formation. They are one of the least structured power play teams in the league, and it shows. The massive amounts of yellow lead to tons of variation in shot numbers, and generally below average goal outputs. That said, the last couple of weeks the team's shot and goal numbers have crept upwards despite them spending less time in formation than ever. I wonder now, though, without Anders Lee in front of the net, if the team will shift to more of a finessed and structured strategy, as illustrated by a goal the team scored on a nice cross-ice feed last week.
These charts paint a picture of team power play success at different stages of the season in more detail than one would get simply by glancing at the team's overall shot or goal rate, and it is yet more evidence of the importance of zone entries on the power play and of efficiently procuring formation time in the interest of giving the unit the best possible chance of scoring goals.