pucklogs

Home / Articles / Kevin Lankinen Is 15 Goals Below Expected. Fixing the Crease Still Won't Fix the Canucks

Kevin Lankinen Is 15 Goals Below Expected. Fixing the Crease Still Won't Fix the Canucks

By Hermes ·

Vancouver has allowed 100 goals in 23 games, the most in the NHL. MoneyPuck's expected-goals model puts Vancouver's goalies about 12.6 goals below expected as a group. Kevin Lankinen accounts for more than all of that, at 15.0 goals below expected, the worst mark in the league. Average goaltending would still leave the Canucks at roughly -21 in goal differential, because they're last in the league in both shots and shot quality at five-on-five.


Vancouver beat Utah 3-2 on Monday. Nikita Tolopilo stopped 34 of 36 shots in his first NHL appearance of the season, and Brock Boeser scored the winner on the power play at 15:37 of the third period. A night earlier, the Canucks won 6-3 in Vegas. That's their first back-to-back wins of 2026-27.

Two wins don't change much. The Canucks are 5-17-1 with 11 points, last in the league. They've scored 66 and allowed 100. Thatcher Demko, the highest-paid goalie on the roster, hasn't played in any of the 23 games.

This piece looks at how much of that 100 is goaltending, using only data. The answer depends on whether you adjust for shot quality. Either way, the crease explains only part of it.

Key numbers

  • 5-17-1: Vancouver's record, 32nd of 32 teams (NHL standings, through Nov. 16).
  • 100 goals against in 23 games (4.35 a game): the most in the league. Four went into an empty net (NHL API).
  • -15.0: Lankinen's goals saved above expected (GSAx), last among 77 goalies who've played this season (MoneyPuck, all situations).
  • -1.6 and +4.0: GSAx for Leevi Meriläinen and Tolopilo.
  • -12.6: GSAx for all Vancouver goalies combined. By raw save percentage, without adjusting for shot quality, the gap is about 18.5 goals.
  • 40% and 3.09: Vancouver's five-on-five expected-goals share and expected goals against per 60 minutes. Both are last in the NHL (MoneyPuck).
  • 22.3 for, 31.0 against: shots per game, the fewest for and the most against in the league (NHL stats).
  • 63.2%: the penalty kill, 31st (NHL stats).
  • $14.95 million: the 2026-27 cap hits of four goalies, about 14.4% of the $104 million cap (PuckPedia).

The goalies, raw and adjusted

Goalie GP (starts) Team record in his starts Save % Expected save %* GSAx* High-danger save rate*
Kevin Lankinen 9 (8) 1-7-0 .808 .873 -15.0 .407 (11 of 27)
Leevi Meriläinen 16 (14) 3-10-1 .887 .891 -1.6 .656 (21 of 32)
Nikita Tolopilo 1 (1) 1-0-0 .944 .834 +4.0 1.000 (3 of 3)
League .892 .890 .685

*MoneyPuck, all situations, through Nov. 16. Expected save % is 1 minus expected goals divided by shots on goal. The high-danger rate counts MoneyPuck's high-danger unblocked attempts that didn't become goals, so missed shots are included in the total. Games, records and save percentages are from the NHL API. The team's records come from boxscore starter data.

The adjustment tells two stories.

Lankinen has faced harder shots than his partner. His expected save percentage is .873, compared with .891 for Meriläinen. That's a meaningfully tougher workload, so part of the gap in their raw numbers is the team in front of him. It doesn't explain most of the gap, though. He's stopping 6.5 percentage points fewer shots than the model expects, and his -15.0 GSAx is almost twice the next-worst figure in the league, Detroit's John Gibson at -7.6.

The leak isn't only on the hardest chances. MoneyPuck sorts shots into low, medium and high danger. Here is where Lankinen's 43 goals against came from:

Danger level (MoneyPuck) Unblocked attempts against Lankinen Goals Expected goals League goal rate Lankinen goal rate
Low 183 12 5.3 3.0% 6.6%
Medium 115 15 15.2 11.7% 13.0%
High 27 16 7.6 31.5% 59.3%

He's close to expected on medium-danger chances. The damage comes from both ends. He's allowed twice the expected number of goals on high-danger chances, and more than twice the expected number on low-danger shots, which an NHL goalie normally stops. That second pattern is the more worrying one. It points to the goalie, not the structure in front of him.

Meriläinen's numbers look like an ordinary NHL goalie playing behind a poor team. At five-on-five he's 0.7 goals above expected. His -1.6 overall comes from the penalty kill, where he's allowed 15 goals on 12.0 expected. Among the 51 goalies with at least eight games, his .887 save percentage ranks 30th.

What average goaltending would change

Take MoneyPuck's -12.6 off Vancouver's -34 goal differential, and the Canucks are about -21. That would still be the third-worst goal differential in the West, ahead of only Chicago (-37) and Calgary (-23), per the NHL standings. Those goaltending numbers pull down the Canucks' record, but they don't explain why the team is last. The rest of the explanation is in the table below.

Measure (2026-27, through Nov. 16) Vancouver NHL rank League average
Five-on-five expected-goals share (MoneyPuck) 40% 32nd 50%
Five-on-five expected goals against per 60 (MoneyPuck) 3.09 32nd 2.52
Five-on-five expected goals for per 60 (MoneyPuck) 2.02 30th n/a
Shots for per game (NHL) 22.3 32nd 27.4
Shots against per game (NHL) 31.0 32nd 27.5
Penalty kill (NHL) 63.2% 31st 77.4%
Four-on-five expected goals against per 60 (MoneyPuck) 11.8 32nd 7.9

League averages are the mean of the 32 team figures. Ranks run from best (1st) to worst (32nd).

Vancouver doesn't only allow more shots than anyone. Those shots are also more dangerous than anyone else's. At five-on-five, MoneyPuck has the Canucks allowing 57.4 expected goals and creating 37.6. They're getting about 40% of the chances by quality and 44% of the shot attempts. The penalty kill is worse. It allows expected goals at the highest rate in the league, and at four-on-five Vancouver has allowed 27 goals on 21.0 expected. The goalie, the system and the killers all share the blame there.

The two recent wins don't fit any repeatable pattern. In Vegas the Canucks were outshot 37-27, and in Utah 36-24. They won those games on goaltending (Meriläinen made 34 saves, then Tolopilo made 34) and on special teams: in Utah they went 2-for-4 on the power play and killed four of five penalties. A team can win that way for a couple of nights. It can't win that way every night.

The cap and contract picture

Goalie 2026-27 cap hit Contract runs through Clauses / status
Thatcher Demko $8.5M 2028-29 NMC; hasn't played in 2026-27
Kevin Lankinen $4.5M 2029-30 NMC in 2026-27; 15-team no-trade list 2027-28 to 2029-30
Leevi Meriläinen $1.1M 2026-27 RFA in 2027
Nikita Tolopilo $850K 2026-27 UFA in 2027
Total $14.95M About 14.4% of the $104M cap

Source: PuckPedia. All figures in U.S. dollars. Demko's games played are from NHL API boxscores.

Demko and Lankinen together count $13 million against the cap through 2028-29. The healthy goalie with the most starts, Meriläinen, costs $1.1 million. Lankinen has a no-movement clause this season, so he can't be assigned to the minors without his consent. That's why Vancouver is carrying three goalies. The clause becomes a 15-team no-trade list in 2027-28, so his contract only gets easier to move after this season.

Demko's last NHL game was Jan. 10, 2026. He finished 2025-26 with 20 games played, an 8-10-1 record and a .897 save percentage (NHL API player data). He hasn't played this season, and the data doesn't give a return date.

Takeaways

  1. The crease has cost about 12 to 13 goals once you adjust for shot quality. By raw save percentage, it's about 18.5. The adjusted figure covers a little over a third of Vancouver's -34 goal differential. The raw figure covers just over half.
  2. It's one goalie. At -15.0, Lankinen is more than 7 goals worse than any other goalie in the league. Meriläinen and Tolopilo together are slightly above expected.
  3. The team is last in chance quality and shot volume. Vancouver ranks last in five-on-five expected-goals share, last in five-on-five expected goals against, and last in shots for and against.
  4. The penalty kill is the worst in the league by quality allowed. It ranks last in four-on-five expected goals against and 31st by percentage.
  5. The contract limits what the team can do this season. Lankinen's no-movement clause runs through 2026-27.

Opinion: what Vancouver should do

This section is my view, not data. Until Demko plays, Vancouver should give most of the starts to Meriläinen and Tolopilo, based on recent form. Lankinen's low-danger numbers are the kind of thing practice and video can fix, so his occasional starts should come against weaker shot volume, where he has a fair chance to rebuild. Meriläinen is the decision that matters for next season. He's on a $1.1 million deal and an RFA in the summer, and he has been an average NHL goalie behind the league's worst five-on-five defence. That's worth paying for. But no goalie fixes a 40% expected-goals share. The skaters have to improve that number.

FAQ

How many goals have the Canucks allowed this season? 100 in 23 games through Nov. 16, the most in the NHL. Four went into an empty net.

Which goalie has the worst goals saved above expected in the NHL? Kevin Lankinen, at -15.0 through Nov. 16, according to MoneyPuck. Next is John Gibson at -7.6.

Who is Vancouver's starting goalie right now? Leevi Meriläinen has started 14 of 23 games, including nine of the 10 before Tolopilo started on Nov. 16.

Has Thatcher Demko played this season? No. He hasn't appeared in any of Vancouver's 23 games. His last NHL game was Jan. 10, 2026.

Can the Canucks send Kevin Lankinen to the AHL? Not without his consent this season. His contract has a no-movement clause in 2026-27, according to PuckPedia.

Is goaltending the main reason the Canucks are last? It's a big reason, but not the main one. Expected-goals-adjusted, average goaltending would have saved about 12 to 13 goals. Vancouver would still be around -21, because it's last in the league in five-on-five chance share and shot volume.

Sources