These last few weeks, I’ve been in writer’s block. During the Black Friday game against Detroit, I began to feel ill. I went downstairs after the first media timeout of the third period, where I ate a late lunch before heading home. The rest of the weekend, I was pretty much knocked out. I ended up writing a good chunk of an article but didn’t like it and never finished it.
The following weekend, I was dealing with a death in the family, and I spent my “free time” trying to catch up, including watching four games that Sunday.
I am at a point in my life where I am truly happy. This is all a hobby for me, and it’s one I enjoy. I’m very lucky that I have the ability to attend games in different arenas, have so many intelligent and talented friends in the sport, and to have a hint of a voice and influence myself. But it makes me feel guilty, almost as if I am letting others (and myself) down when I can’t fully complete ideas or pieces of content on a regular basis.
I think the best cure is to simply write down all of the incomplete ideas that I’ve been thinking about and share them with the world. So here is a smorgasbord a bite-sized content.
Why good teams don’t rely on the rush
On Friday, Meghan Chayka of Stathletes tweeted out this chart.
I think it creates more questions than answers, even in the content provided. Is this 5v5, even strength, all situations? Regardless, I think this passes the smell test on how teams score goals. Vegas and Dallas are super rush heavy while the Bruins are cycle and forecheck heavy.
The first thing I thought about was Corey Sznajder’s presentation from the Seattle Hockey Analytics Conference last December. Corey wrote an article on his Substack, but the quick summary of his project was that he tracked additional context for every 5v5 goal in the 2021-22 season. It was the deepest study I have seen in that regard.
In the 2021-22 season, the Vegas Golden Knights missed the playoffs. They were one of the most rush-reliant teams that season. As Corey summarized:
There was nothing wrong with how they played per se, it’s just that you can see that their offense was somewhat one-dimensional even with how good they were off the rush…
This is how you wear teams down, but what kept them from being in the next tier with teams like Colorado, Minnesota and St. Louis was the lack of sustained offense. They didn’t follow-up their next shifts if they didn’t score on the initial chance and a lot of their offense came from point shots where you’re relying on a rebound or deflection.
This also brought me back to an article by Jen Lute Costella, who I think is one of the greatest minds we’ve had in this community. In an article on shot generation and puck recoveries, she writes:
Teams that excel at shot generation do so through puck recovery in the offensive zone even after they have taken a shot, after a passing turnover and after direct turnovers…
Recovering the puck in the offensive zone and generating shots showed a 61.46% correlation in this sample. Obviously, this is a small sample size, but the fact that logic supports a relationship between recovering loose pucks in the offensive zone and generating shots lends credence to the findings.
I think when we create silos for stats, we do ourselves a disservice. Creating many goals off the rush isn’t bad. Having a high percentage of your goals come from the rush isn’t necessarily bad either. It may mean that the quality of your rush shots is towards the top of the league, and you are finishing on such shots. You want to have the tool, the ability to effectively attack off the rush, in your tool kit. But a good carpenter doesn’t just use their hammer.
The game of hockey is interconnected, and we need to think about how each event affects the next when we try to draw conclusions from the game.
Goalie breakouts are a necessary evil
A few years ago, there were many more tracking projects going on in the hockey analytics sphere. One breed of projects was to look into which exit styles worked best. It was a fun time in the community where we were moving into blending tactics and numbers. Ryan Stimson was the leader in this in my opinion. In an article from 2017, he found that goalie breakouts were the worst in terms of the net shot differential over the next 10 seconds. Others and I ultimately found similar results in our tracking.
From the data, you would ultimately conclude that having the goalie play the puck, regardless of skill, is a bad idea. However, many coaches would tell you that they love a goalie that can play the puck well. It adds a tool to their tool kit (see the theme?). When a goalie plays the puck effectively, they beat at least one forechecker, which should aid the skaters in their effort to move up ice.
Similar to my last blurb on rush shots, when we isolate these metrics, we may be missing context. I hypothesize that goalie breakouts are caused by conditions that force them to come out of their net and assist their skaters. This may be that an opposing forechecker beat the first skater back. Or the goalie may sense that the first skater back doesn’t have the proper support and intends to aid that skater by changing the manpower.
I think we may move into an era of tactical, and even player development, focused analytics once again, and I feel these things need to be considered this time around. Personally, I believe that given the acceptance and interest of coaches and scouts this time around, collaboration will bring us enhanced insights.
Benching Poitras in the third is fine
It feels that the Poitras game management discussion skyrocketed this week with a planned healthy scratch on Thursday against the Sabres, and being benched in the third period on Saturday with the Bruins down a center.
I think the first reaction, and rightfully so, from Bruins fans is that this is not what was promised to them. The frustrations of the last decade plus under Bruce Cassidy and Claude Julien was that they had no patience for the kids. Jim Montgomery was supposed to be more of a “players coach” who would allow young players to make mistakes and learn from them.
Ultimately, I think we’ve seen some of that out of Montgomery. For example, a little over a month ago, he gave Mason Lohrei free reigns in his NHL debut against the Toronto Maple Leafs, including letting him see the ice in overtime. Lohrei was having an extremely great game relative to his baseline, and sensing the confidence, Montgomery followed through and let him play.
With Matthew Poitras, he has struggled relative to his baseline lately. His puck management has been poor, which led to a goal against versus the Coyotes, and the game has been more physically taxing for him. Putting Poitras in a position where he is not best suited to succeed, like defending a lead in the third, at a time where he is not playing with confidence, isn’t best for his growth.
Does that mean he should spend the entire third period on the bench? Personally, I think it’s overkill. I’d like to see him get a couple of shifts starting in the offensive zone or against the opposing third or fourth line. But I do think it’s best to limit the chances he has to make major mistakes in order to manage the confidence of a 19-year-old. Plus, at the end of the day, the NHL is not a development league. The Bruins are winning, and Jim Montgomery is paid to win hockey games.
We probably haven’t seen the best of Pastrnak this season
Throughout the first 26 games, I’ve found myself feeling like David Pastrnak hasn’t found his game. Sure, that feels strange to say about a player who is third in the league in points, has scored a couple of ridiculous penalty shot goals, and leads the Bruins in GAR and xGAR. But we’re also talking about a superstar who has shown capable of consistently showing up to make a difference shift after shift.
One of the things that has stuck out to me, is the change in his transition metrics. Last season at 5v5, he controlled 61% of his entries and led the team in controlled entries per hour with 21. This season, his controlled entry percentage is down to 52% and he’s only controlled 15 entries per hour. This doesn’t translate directly to his on-ice stats as he is by far the biggest driver of offense, but I think this speaks to him playing a less individual style of play. This may be linked to Brad Marchand subbing in for most of the time David Krejci took up last season, so it doesn’t worry me much.
What makes me feel potentially “bullish” is that his unblocked shooting percentage is quite low relative to his baseline. Heading into this season, he had a career unblocked shooting percentage of 10.3% at all situations. This season, he’s only shooting 8.8%. This season, he’s taken an extra unblocked shot attempt per hour, but that unblocked shot attempt has missed the net. If we start to see some of those missed attempts get on net, we are due to see his goal scoring rate go even higher.