Film breakdown: Michigan's 2018-19 penalty kill
Film breakdown: Michigan's 2018-19 penalty kill
Diving into the numbers and the tape to break down the system from the 2018-19 season
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?
Monday, I announced my Michigan hockey penalty kill tracking project and laid out the plan for sharing my data. Today, we’ll break down the penalty kill system from the 2018-19 season to get a better sense of what was going on. Friday, I’ll share the same breakdown of the 2019-20 season.
Penalty kill data can be fairly noisy because of the small sample size — Michigan only spent a total of 275 minutes shorthanded as compared to over 2,000 total minutes played — but there are still ways to evaluate performance. For this project, I’m combining manual tracking with film breakdowns to present a comprehensive look.
The basic numbers
I’m working with a 27-game sample here and I only tracked data at 5v4 to keep the comparisons as simple as possible. In the chart below, you’ll see a simple breakdown of the basic numbers both from my tracked 27-game sample and from all 36 games. All of the data in this chart is based on 5v4 play. It does not include any shots, goals or power play opportunities at 5v3.
These basic, underlying numbers present an at-a-glance explanation of Michigan’s penalty kill. It didn’t give up a ton of shots, but the save percentage wasn’t great. This tracks with Michigan’s overall PK performance in the 2018-19 season, in which its 78.5% kill rate left the Wolverines 46th of 60 teams in the country. It also tracks with Michigan’s overall goaltending in 2018-19, which finished 59th in the nation with a .887 combined save percentage.
The tape (and more numbers!)
To go beyond just the basic numbers, let’s take a look at the underlying systems Michigan used in the offensive zone, neutral zone and defensive zone.
The offensive zone
The goal of forechecking in the offensive zone on the penalty kill is essentially the same as at even strength — that is, force a turnover and create offense — but with one major difference: It’s riskier because the numbers are already uneven, so the odds of an outnumbered rush the other way after a failed forecheck are much higher. For this reason, offensive zone penalty kill forechecks are typically more focused on slowing down the opponent than truly forcing turnovers.
According to my tracking, Michigan sent at least one forward to forecheck in the offensive zone about 31% of the time. For these purposes, I only counted it as a forecheck if they made an actual attempt to pressure the puck carrier — that means that situations in which a Michigan skater simply swung through the offensive zone and retreated to the neutral zone without pressuring the opponent didn’t count.
On 62 of the 76 total offensive zone forechecks, the Wolverines sent two forwards into the offensive zone to apply pressure.
In the first clip below, forward Dakota Raabe carries the puck into the offensive zone but loses it around the blueline. Ohio State recovers in the corner, and Raabe goes to work on the forecheck. As Quinn Hughes shades back into the neutral zone, forward Garrett Van Wyhe holds his position higher in the zone to support Raabe’s work down low. Their goal is to force the Buckeyes toward one side of the ice for their breakout, and as the puck changes sides, the Michigan forwards are in hot pursuit.
In the second clip above, Hughes clears the puck down into the far end of the offensive zone. As the puck rims around to the right corner, forward Adam Winborg is on his way there to meet it. Forward Luke Morgan, as Van Wyhe did in the first clip, holds his position higher in the zone to support Winborg and help push Minnesota to one side of the ice. When the puck moves across the zone, Morgan chases and forces a cross-ice outlet pass, where a Michigan skater is waiting to break up the entry.
These clips are somewhat exaggerated examples of this forecheck — it wasn’t typically this aggressive — but I find that exaggerated examples make the underlying concept more clear. This forecheck is generally known as same-side press, which gets its name from two forecheckers forcing opponent to one side of the ice.
Less commonly — only 14 of the 72 times — the Wolverines sent just one forward into the offensive zone while the other three skaters hung back in the neutral zone to break up the entry attempt.
The neutral zone
A well-executed neutral zone forecheck can frustrate an opponent into failed zone entries, which both wastes time on the kill and, obviously, prevents scoring chances for the power play. The Wolverines used two main setups for their neutral zone forecheck: a 1-1-2 (or same-side press) look (43.2% of the time) and a 1-3 look (53.8% of the time) with one active forechecker and three retreating skaters.
Below, the first three clips show the 1-1-2 look with varying degrees of success. In the first clip, the two forecheckers push Minnesota to the near side of the ice but the Gophers are able to enter the zone with minimal trouble and create a scoring chance off the rush, and the same happens in the second clip. In the third clip, they’re more successful at stopping what Minnesota is trying to do and Michigan disrupts the passing after the entry.
When running same-side press, the goal is for the first two forecheckers to force the opponent to one side of the ice. This puts more pressure on the opponent because they have less space to work with and creates a more difficult entry if it’s successful, but it also can be beat with cross-ice passing.
The final four clips in the video show the 1-3 look. It’s immediately evident that the 1-3 forecheck puts much less pressure on the opponent, but it’s also more successful at disrupting and preventing entries with the three skaters across the blueline. In the fourth clip, Penn State is forced to dump the puck into the zone, and in the fifth, they get caught in a battle on the wall right after the entry. In the sixth clip, Michigan once again forces a dump-in, and in the last clip, the puck is picked off as Minnesota enters the zone.
The defensive zone
When set up in the defensive zone, Michigan mostly stuck to a diamond formation, using that look 80.3% of the time. In a diamond, there’s one penalty killer at the net front, two in the middle of the zone, and one in the high slot.
This is a pretty standard look on the penalty kill that cuts off passing lanes and puts penalty killers in position to block shots. It can be vulnerable to passes across the slot from the point to a shooter on the backdoor, but those shots tend to be low-percentage — though by no means is that a guarantee, as you’ll see from the goals Michigan gave up. It’s also a flexible formation; the penalty killers can squeeze closer together to take away the middle of the ice or play a bit wider to put more pressure on perimeter.
In this formation, you can see that at any one time, typically only one skater moves up to pressure the opposing puck handler. Adam Winborg and Griffin Luce moving back and forth as Minnesota passes in the first clip shows this clearly. To avoid getting caught with two penalty killers out of formation, which opens up passing and shooting lanes, as Winborg or Luce moves up to pressure, the other quickly recovers back.
Each skater has a general assignment, and if the power play switches positions, the penalty killers follow suit. The best example of this is in the fourth clip as MINN15 and MINN17 switch positions.
In general, Michigan didn’t exert a ton of pressure in the defensive zone. As you can see, they’d send one penalty killer to pressure when the moment was right, but it didn’t happen every time. You can also see that the shots they gave up were largely one-timers from the perimeter, which forced the goalies — either Strauss Mann or Hayden Lavigne — to get over for the save. Perhaps if they’d pressured the far man on the power play a bit more, those shots would’ve been easier saves or less likely to come through in the first place.
The goals against
Michigan gave up 16 goals at 5v4 in the 27 games I tracked and 24 5v4 goals on the season. Here’s six of them.
The first goal is just a straight up defensive breakdown. The penalty killers got caught chasing the puck in the corner and no one was paying attention to Taro Hirose — not exactly the kind of player you want to leave uncovered — in the slot.
But in the other clips, some trends emerge. These few clips shouldn’t be mistaken for a representative sample of every goal Michigan gave up on the penalty kill, but it gives at least a bit of an idea. The Wolverines were vulnerable to quick passing and one-timers, particularly on the far side. These are situations where you really need your goalie to be on his game and ready to bail you out which, as we discussed at the top of this piece, was not quite what Michigan got from its goaltending in 2018-19.
I hope you enjoyed the first installment of this series! A similar breakdown of the 2019-20 season will be in your inboxes at 9 a.m. on Friday. If you haven’t signed up yet, I’d love to have you join us!