By the numbers: How Michigan's penalty kill changed over the last two seasons
Breaking down the five largest statistical changes, both positive and negative
Welcome to Fresh Ice, a data-driven college hockey newsletter. Each week, I’ll be publishing fresh (no pun intended) content that focuses on educating and explaining all aspects of the game. Haven’t subscribed yet?
Last week, I introduced this penalty kill tracking project and broke down the tape on both the 2018-19 system and the 2019-20 system. The response and feedback has been incredibly positive, so I just wanted to say a quick thank you to everyone who has supported this project so far!
Today, I’m going back into the tracked data to talk about the top 10 largest numerical changes from year to year — the five most positive changes and the five most negative. To be clear, I mean positive and negative in the numerical sense, not the value sense. Some of the negative numerical changes have a positive impact in terms of the penalty kill’s success and vice versa.
As a quick recap, here are the basic numbers from each season. The 27-game sample is the games I was able to track and the 36-game data is the entire season. All of this data comes from 5v4 — it does not include any power play opportunities, shots, goals or anything else at 5v3 or 4v3.
I’m not deeply digging into this (yet) but something that stands out just in those numbers is the change in power play opportunities for the opponent. Mel Pearson talks a lot about wanting Michigan to be one of the least-penalized teams in college hockey, and the Wolverines lived up to that this past year. This past season, Michigan ranked seventh in the nation in penalty kill attempts. In 2018-19, the Wolverines came in at No. 20. That’s obviously not something that really has a bearing on the performance of a penalty kill, but it is an interesting data point.
I started this tracking project with the intention to see what changed between the two seasons, and I didn’t quite find what I expected to. Given that the penalty kill improved by almost six percent overall, I expected to see things like fewer shots and more blocks. Spoiler alert: That isn’t what happened.
Top 5 Positive Changes
Again, just to be clear, these are the top five positive changes in a purely numerical sense. As you’ll see, some have a positive impact on the kill’s success, and some have a negative impact. Because I tracked 173 penalty minutes in 2018-19 and 160 penalty minutes in 2019-20, I’ll show the change on a per 60 minutes basis.
Opponent zone entries: Michigan’s opponents entered the offensive zone almost 14 more times per 60 minutes in 2019-20 than they did in 2018-19. This number also includes failed entries, which increased by 4.60. What this tells us is the Wolverines were getting the puck out of their defensive zone much more frequently and forcing the opponent to come at them and try to get into the offensive zone. This makes sense with Kris Mayotte’s overall approach to penalty killing, which is heavily focused on getting the puck out of the zone and getting fresh bodies on the ice whenever possible.
Defensive zone exits: This is the other side of the coin to increasing opponent zone entries. If the opponent is entering the zone much more, Michigan must have been getting the puck out of the zone much more, and it was. As I just talked about, this fits with Mayotte’s goal of keeping penalty kill shifts short to get fresh legs on the ice as frequently as possible.
Faceoff losses: Michigan lost almost nine more faceoffs per 60 minutes this season than it did last year — and faceoff percentage on the penalty decreased by 10% from year to year. Even more interesting is that the Wolverines’ 2019-20 faceoff percentage went down from 46.5% in the first half of the season to 37.3% in the second half of the year. Adding intrigue is that Michigan’s faceoff percentage in all situations was 49.3%, which ranked No. 39 in the nation — certainly not excellent, but not abysmal either. To be honest, I have no idea what led to this, and I am super interested in what the cause might be.
Shots: Going into this project, I expected to see fewer shots because Michigan’s penalty kill was much improved overall. Instead, this suggests that one of the biggest reasons the Wolverines’ PK improved was the performance of Strauss Mann in net, and not necessarily anything to do with the system they were running.
Opponent zone entries (dump-in): As I broke down last Friday, Michigan’s change in forecheck systems led to an increase in the amount of times its opponents had to dump the puck into the offensive zone. This wasn’t unexpected, but I was a little surprised to see how large this increase was.
Top 5 Negative Changes
Misses: Any shot attempt that hit the post or went wide of the net counts as a miss, and this change really stands out. Michigan’s opponents missed the net eight fewer times per 60 minutes, while also taking over eight more shots on goal. I don’t have any sort of explanation for this, and it could be entirely coincidental that misses decreased by almost exactly the same amount that shots on goal increased. But regardless, it’s an interesting stat that may be worth looking into further.
Hits: I was pretty generous with what I identified as a hit when I was tracking, both because I was somewhat using it as a proxy for aggressive play and because laying big hits on the PK is basically a cardinal sin, so if I didn’t include bumps and other contact, there would be basically nothing to track. With that said, I didn’t expect to see this number decrease as much as it did. Off the eye, Michigan’s PK was much more aggressive this season than it was in 2018-19, so I expected to see a bit more physical contact, rather than less.
Opponent zone exits: This measures how many times the opponent left its own defensive zone. As I’ll talk about with shorthanded misses, the Wolverines appeared to shift focus from pressing for shorthanded offense at every opportunity to just dumping the puck down and getting off the ice, which explains why this decreased.
Misses (shorthanded): Something I heard said a few times in the first half of the year was that Michigan wanted to score more shorthanded goals than they gave up on penalty kill, but it seemed like that emphasis shifted as the year went along. The emphasis on pressing for shorthanded offense seemed to decrease in favor of out-changing the opponent and getting fresh bodies on the ice. Michigan also took fewer shots on goal while shorthanded, though that changed less than misses.
Faceoff wins: With faceoff losses increasing and faceoff percentage decreasing by 10%, it makes sense that faceoff wins would also decrease by a pretty good amount. Again, I have no idea what caused this, but I would be fascinated to find an answer. I didn’t track faceoff data at the player level when I started this project, but I just might be intrigued enough to go back and take a look at that.
In Conclusion
Obviously, none of these numbers on their own — or even all together — explain a penalty kill’s success or lack thereof. But I thought it was really interesting to get down to the granular, event-specific data and see what changed the most from year to year. I certainly didn’t see what I expected to, and it both helped uphold conclusions I drew from the film breakdowns and also give me a lot more questions and ideas to dig into going forward.
Speaking of that, if you have a question you think I can answer with my dataset, I would love to know about it! You can reply to this email and I’ll see it or you can let me know on Twitter.
I hope you enjoyed the third installment of this series! Still haven’t subscribed?