Is There a Case to be Made for Jake DeBrusk?
Bruins nation is conflicted about this soon-to-be UFA.
Boston’s most notable Unrestricted Free Agent has had an interesting career wearing the Spoked B. Drafted 14th in the notorious 2015 draft by Boston, DeBrusk has had good and bad years with the team. Perhaps it’s because of that history that many Bruins fans seem to be on the fence about whether or not they want the winger to stick around. It also doesn’t help that, as of June 21, 2024, the 27-year-old has yet to sign a deal with GM Don Sweeney.
I myself am one of those conflicted fans. I was thrilled when he finally seemed to find his groove on the top line with Patrice Bergeron and Brad Marchand. The JD we saw playing in 2023-2024 was the most elite I think he’s ever been. He played consistently at this high level for well over a year after being moved up to the top line in early 2023. But with the departure of veteran players like Bergeron and Krejci, DeBrusk seemed to have struggled to find his niche again. As I’ve said many times on the podcast, everyone on a hockey team has a job or a role. DeBrusk seemed to struggle to find what his job was in this new iteration of the Boston Bruins. As such, his scoring luck seemed to dry up and he was back on the third line with nary a goal in sight.
Many folks that I’ve spoken with postulated that it was the pressure of the trade deadline that seemed to hinder a full performance. With the Bruins determined to make it past the first round in the playoffs, it likely meant that Don would do whatever it took to get there. Perhaps lucky for DeBrusk, the limited cap space left over from a summer where the cap only went up by one million dollars, DeBrusk was still a Bruin after 5pm on March 8th. Seemingly over night, a massive weight was lifted from his shoulders and he finished the month of March with some of his best stats of the season.
Not knowing what to think of him, I decided to go to my trusty stats.
Player Usage
I’ve been a fan of the player usage model since I first learned about it. If you need a refresher, you can read a previous post I’ve written about it here. This year I made a few updates and tweaks to my model, namely adjusting how I weighted the percentile values for players who were traded during the season.
I was quite surprised when I ran Jake DeBrusk’s player usage numbers. Likely because his goal scoring was down and I didn’t hear his name quite so often, but I had no idea what kind of jump he had in time on ice.
Team Use
Let’s break these numbers down. If you take a look at the values JD has under QoTF and QoTD, these are his player usage percentile values comparing him to his teammates. What do these mean?
First of all, you can see his progression among forwards on the team over the last 3 seasons. He has inched his way up in percentile ranking in QoTF as he has grown and developed on the team. That’s great to see and something you would expect from a player like DeBrusk.
What really caught my eye was the fact that his QoTD was above the 50th percentile. Defensemen always play more minutes than forwards; it’s the nature of the position when there are only six of them compared to twelve forwards. The fact that DeBrusk was essentially used more than most of the defensemen is pretty impressive. There could be a few reasons for this. The first that comes to mind is the fact that the team had a large rotation of defensemen. This could have lowered the time on ice for quite a few of the bottom-level defensive players.
Sure enough, my guess at why JD was ranked so much higher than a good number of defensemen proves to be correct. JD still fell below the true regular defensemen: Hampus Lindholm, Charlie McAvoy, and Brandon Carlo. Nothing surprising to see here.
Player Usage x Peers
Looking at player usage on the team isn’t enough. This just shows how valued he is on the Bruins’ roster. I pulled data from across the league to see where Jake ranked.
I started off by filtering down players to either left or right wingers and then by age. Specifically, I searched for players who were between the ages of 26 and 28, since Jake is 27 as of this post.
I was pleasantly surprised to see Jake still highly ranked among his peers in both offensive and defensive categories.
Player Usage x Goals
Ok, so Jake DeBrusk gets used a lot. And he plays a lot of “tough minutes”. So what? He didn’t have the stellar goal scoring season in 2024 as he did in 2023. Remember when he scored two goals on a broken leg at Fenway?
But what if I told you he out-performed expectations based on how he was used?
I pulled in a bunch of other individual stats for players in my data set to see if there was something else worth looking at. I happened to notice that there were occasions where notable players whose names have been in the free agency rumor mill had high goal counts but fairly low player usage stats. What if those players are just really efficient? Is there some kind of link between goals and player usage?
At first blush it seems obvious. Of course, if you play more minutes you have more opportunity to take shots and score goals. But if I control for that, can I find a way to measure a player’s goal scoring efficiency as it relates to player usage and tough minutes? Yes!
Fun with Regressions
My first experiment was to run a simple linear regression to see if there was a good fit between goals and tough minutes.
There is indeed a fairly good fit between goals scored and tough minutes. Plus the regression showed statistical significance! And the co-efficient for tough minutes indicates that for every additional 0.0913 to the tough minutes variable means an increase in goals. But what about controlling for the amount of time and tough minutes?
Fun with Fancier Regressions
This terrifying thing shows just how much each variable contributes to the predicted number of goals. For example, a positive coefficient for total points suggests that more total points lead to a higher predicted goal count, while a negative coefficient for toughest minutes suggests that more time on the ice during tough situations may slightly reduce predicted goals. I was really excited to see there results of this regression because for a large number of players in my data set, the prediction was spot on.
Keep in mind there are a LOT more variables out there that have an effect on predicted goals per player. And sometimes players will simply out perform expectation.
So how did Jake stack up? In the last two seasons, he has out performed his predicted goals based on player usage. And when it comes to overall individual expected goals (ixG), he wasn’t too far off either.
Limitations: Keep in mind that these predictions are based on the assumptions and limitations of your regression model. They represent estimates and may not perfectly reflect actual performance due to other unaccounted-for variables, random variability, or changes in player performance over time.
Does this mean that Jake DeBrusk should stay in Boston? I don’t know the answer to that and only time will tell. What I will say, with more conviction, is that he is better than we give him credit for.