The Canucks won 2–1 despite losing the expected-goals battle, 3.42–1.65. New Jersey controlled 67% of the xG; this was a heist.

Vancouver's victory defied every underlying metric. The Devils dominated 63% of shot attempts, generated six high-danger chances to one, and carried a superior chance quality of 0.057 xG per shot versus 0.041. By the numbers, New Jersey owned this game.

But the score favors the Canucks. That's because T. Demko was exceptional.

The Vancouver goaltender stopped 25 of 26 shots—a 0.962 save percentage—and robbed 2.42 goals above expectation. On high-danger chances alone, Demko was 6-for-6, an improbable shutout within a shutout.

New Jersey's finishing was the culprit. The Devils went 0-for-6 on high-danger chances, squandered three rebound opportunities, and underperformed their 3.42 xG by 2.42 goals. That gap doesn't close often.

Vancouver converted on the power play—2 goals on 6 power-play shots, an absurd 0.60 xG. The Canucks capitalized where the Devils couldn't, a microcosm of the lopsided outcome.

J. Markstrom, New Jersey's goaltender, faced a different sort of game: 13 saves on 15 shots. When your team generates this much volume and quality, your goaltender's performance becomes secondary.

The Canucks stole this. The Devils earned it. The xG said so, the shot chart confirmed it, the chance quality proved it. Vancouver simply refused to lose, riding Demko's wall and opportunistic power-play execution to a road win that shouldn't have happened.

New Jersey's underlying metrics paint a convincing edge. The model projects a 2.4–2.8 goal margin with 57% win probability in favor of the Devils. The numbers support it: at 5v5, New Jersey controls 50.0% of expected goals versus Vancouver's 44.0%, a meaningful gap. Corsi backs it up—51.0% to 46.0%—showing territorial dominance. The Devils spend more time generating from high-danger areas and dictate where play develops.

That's the baseline case. Both teams' recent form muddies it.

Vancouver is collapsing: minus-14 over their last 10 games. New Jersey isn't far behind at minus-7. Form matters when teams are spiraling. Neither side has footing, and momentum swings can override projection models. But here's where PDO regression becomes relevant.

Vancouver's PDO sits at 0.974; New Jersey's at 0.971. Both are running cold by design—converting below what their shot quality and volume warrant. Both are due for positive regression. If Vancouver corrects upward, they close the gap. If New Jersey regresses into an already-favorable spot, the Devils' advantage amplifies. The model doesn't yet price in that bifurcation fully.

One tactical variable: J. Allen is on a back-to-back, so monitor for a backup appearance or a fatigued starter. Shot-stopping variance could swing expectations if the underlying play doesn't unfold as cleanly as the metrics suggest.

The forward depth tells the story. Filip Chytil carries Vancouver's offense at 1.03 expected goals per 60 minutes with 3 goals and a 15.0% shooting rate—high volume, high luck. Drew O'Connor (13 goals, 11.8% shooting, 0.85 xG/60) is a complementary scorer operating on unsustainable efficiency.

New Jersey counters with Timo Meier: 1.08 xG/60, 13 goals, 15 assists, only 6.7% shooting. That profile suggests sustainable excellence, not variance. Jack Hughes (18 goals, 0.88 xG/60, 11.5% shooting) generates at volume and finds finishers. The Devils' forwards profile as drivers and creators, not lucky finishers banking on PDO luck.

Defensively, Victor Mancini represents Vancouver's weak link. Johnathan Kovacevic is New Jersey's liability. Neither is elite, but isolation plays and traffic in front should favor the Devils' ability to generate secondary scoring.

The Devils' lean rests on durable superiority: better underlying metrics, deeper forward talent, a goaltender advantage. Vancouver can regress positively and still be outplayed—exactly what the model projects.