At the start of the offseason, the Michigan hockey team’s weakness was clear as day. Two-year starting goaltender Erik Portillo — who only sat three games in that span — left for the pros. Backup goalie Tyler Shea also departed, hitting the portal and eventually transferring to Brown, leaving only Noah West as the program’s sole hope in net.
But now, West has some help — really good help. Enter grad transfer goaltender Jacob Barczewski, who signed with Michigan out of the portal from Canisius in April. If his former coaches are any indication, he could become a core piece of next season’s puzzle for the Wolverines.
“He’s a light-minded kid, but when it comes to goaltending, he is very serious about his craft,” Canisius coach Trevor Large told The Michigan Daily. “He’s an excellent teammate. (We) got to know that through the recruitment process and it couldn’t be more accurate. Spending four years with him, he was exactly what we could have hoped.”
While that endearing personality will certainly help him win over teammates, Michigan also needs Barczewski to win games. He was the Golden Griffins’ starter for their run to the Atlantic Hockey Championship and NCAA Tournament, playing well throughout. Even on the wrong end of a 9-2 NCAA Tournament loss to Minnesota that ended his season, Barczewski finished the postseason with a .932 save percentage.
But that miracle run wasn’t a fluke. Sure enough, Barczewski is pretty good at winning. He was a talented junior goalie who played in crowded goalie rooms during his recruiting process. Compared to what he accomplished in his four years with Canisius, Barczewski flew under the radar. Large compared him to Charles Williams, the first Mike Richter Award and Hobey Baker Award finalist from Canisius.
Across his four years with Canisius, Barczewski’s stats show plenty of potential — a career .917 save percentage and 2.60 goals against in the NCAA. But stats can also be deceiving. There’s a drastic difference in competition between the Big Ten and Atlantic Hockey Association. Tactically, some teams also play a style that can hang their goaltenders out to dry. Still, the Golden Griffins did their part to defend in front of Barczewski.
“In terms of the way we defend, I think we’re difficult to get in and around the net and get multiple scoring chances,” Large said. “That’s something that I think our team is very, very good at. Shots will be from the outside. If the shots are coming from areas that aren’t dangerous, (Barczewski) will make those saves and more.
“He has the ability to make the big save, but he’s very structurally sound. So it’s a combo of our structure and his structure go very well (together) because we’re not letting up a lot of second chance opportunities.”
Luckily, other stats can better account for these differences. Namely, the metric of goalie win shares (GWS). Used in some NHL discourse for more than a decade, goalie point shares take into account conditions like shots against per minute and league averages to estimate the contributions of a goaltender. But the NCAA doesn’t assign points for wins in non-conference play, so substituting wins for points can convert this stat to a usable metric for college hockey.
Using goalie win shares, we can assess that Barczewski contributed around three wins a season for Canisius each of the past two seasons. This doesn’t mean he actually took on a team 1v20, but rather that his individual contributions added up to that number. In 2021-22, he ranked 14th of all NCAA goalies in win shares, while he finished 11th in 2022-23.
Most of that caliber of goalie is the homegrown type, one who stays at a school and acclimates to systems across their career. That makes picking Barczewski out of the portal a rare coup. Compared to the other goaltending options available for Michigan in the transfer portal, Barczewski held the highest win share total.
But readjustments can be tricky, especially for goaltenders who played to stud levels with their previous program. Looking at recent transfer goalies and their most recent seasons before transferring, only three fit a similar mold as Barczewski. Two of those goalies went to North Dakota — Zach Driscoll and Drew DeRidder — and Eric Dop transferred to Boston College in 2021. All played their grad years with a new program, and all saw varying degrees of success.
Overall, Driscoll and DeRidder sank compared to their previous contributions, whereas Dop improved ever so slightly. When examining grad transfer goalies, it’s important not to dismiss any regression from previous seasons.
Barczewski, however, is confident that he can repeat his success.
“I expect to repeat what I’m doing here,” Barczewski told MLive in May. “I know it is a step up, a good league, good shooters, a lot of skill. I understand it, but I don’t expect anything less than what I have been doing and what I’ve been producing.”
To be fair, it’s easy to amass those kinds of contributions as a starting goaltender, and the raw GWS metric can be deceiving. A goalie who plays 30 games is in an invariably better position to perform better than a goalie who plays 10. For example, the 2020-21 season looks like a down year for Barczewski. In reality, he just didn’t get many games due to COVID-19 schedule constraints. Adding another lens — that is, dividing win shares by games played (GP) — can compensate for this.
Viewing goalie win shares per game played (GWS/GP), 2020-21 was actually Barczewski’s best season. That followed a trend for many goaltenders that season thanks to a shorter schedule. Ranking 10th in the NCAA with .105 win shares per game, Barczewski even finished higher than the stalwart Strauss Mann, who was Michigan’s high water mark for goaltending the past few seasons.
Barczewski also finished 12th in 2021-22 and eighth in the NCAA in 2022-23. For reference, Portillo was 13th and 22nd in those seasons, respectively.
There’s another takeaway from this: With a capable tandem partner, Barczewski might play in fewer games and approach those focused 2020-21 numbers.
Overuse can be a detriment to any goalie, but especially one in college hockey. Playing 60 minutes back-to-back nights can wear on a starting goaltender, especially one who faces a high volume of shots. Barczewski fits that bill well, playing in 32 of the Golden Griffins’ 42 games last season.
But Barczewski was also remarkably consistent in the face of this overuse. To analyze this, let’s turn to another goaltending stat: quality starts. For the uninitiated, these are starts in which a goaltender performs higher than the league average save percentage, or eclipses .850 in a game with fewer than 20 shots faced.
For 2022-23, that average save percentage rounds up to .908. Against that comparison, Barczewski put up 23 quality starts out of 32, good for 72% of his games. Fairly high, but what about the remaining 28%?
The majority of those remaining starts are those deemed really bad starts (RBS), which represent games in which a goalie fails to reach an .850 save percentage. Barczewski had five — around 16% of his games.
Compared to the top 15 starting goalies in GWS/GP who played at least 20 games, Barczewski ranks seventh in non-RBS games. He earned the third-most quality starts of the field.
The abundance of really bad performances compared to merely mediocre ones could also suggest that Barczewski faced too many games. His off nights weren’t just mediocre — they were complete letdowns in which his usually sound play almost entirely gave out.
Putting him on a proverbial pitch count could yield positive results, helping him achieve a higher percentage of quality starts. Of his really bad starts, three of them came on the second night of a back-to-back; one of them ended his team's season in the NCAA Tournament after a long postseason run.
But even when he is leaned on heavily — exemplified in Canisius’ postseason run — Barczewski can also deliver. Of eight postseason games, he put up seven quality starts including two shutouts, one of which came in the AHA Championship Game. His worst outing until losing to Minnesota was a .931 outing against Niagara in the conference semifinals. Needless to say, his head hurt from standing on it so much.
Having a capable backup, or a 1A-1B tandem, disperses that burden — and that’s what Michigan figures to capitalize on. West, the other fist of the presumptive one-two punch, was a capable freshman starter at Robert Morris and made the Atlantic All-Rookie team in 2020-21. Then a program closure forced him to find another school, ultimately winding up at Michigan. But he could hardly find starts because of Portillo’s control of the net.
West proved serviceable in three starts last season, actually ranking third in win shares per game played. But three games doesn’t offer a large enough sample size to really show any effectiveness. That's a large part of why there’s so much mystery behind how West might perform when given a starting role. That made Barczewski’s pickup not just a good roster move, but an essential one.
Staying at Michigan after a dalliance with the transfer portal likely means there’s some way to get West in more games, which could prove fruitful if he can post similar numbers. As far as Michigan coach Brandon Naurato has said, the plan next season is to start with a tandem system. Then, he plans to let a starter earn the job.
“(West and Barczewski) are gonna fight for the starting job,” Naurato told Gulo Gulo Hockey in May. “I told both of them I would probably plan on some kind of split for the first half, and if someone takes it — great.”
Instead of leaning on one bellcow, sharing will be the name of Michigan’s goaltending game next season. But as much as the competition and lighter workload can help both goalies, there are also flaws in that plan — for instance, picking the right starter when the postseason comes around.
Naurato’s gambit is that the situation will resolve itself. Either West will earn the spot after two seasons on the bench, or Barczewski will rise to a higher level of competition.
That plan seems to be Michigan’s best hope, and the early statistical indicators show that it could theoretically work. Barczewski and West must both live up to the numbers if it is going to work.
Because if they don’t, Michigan might have no options to fix its greatest flaw.