Colorado's 7–2 victory looked more decisive than the underlying play deserved. The Avalanche controlled expected goals 4.60–2.66 (63–37%), a margin that doesn't justify a five-goal scoreline.

The gap came down to goaltending and finishing. J. Dobes surrendered 2.4 goals beyond expectation — well below the workload of 36 shots he faced. Colorado's M. Blackwood, meanwhile, stopped all six high-danger chances, outperforming his xG by 0.66.

Denver was lethal on the rush: four goals on 32 chances. Montreal was effective too—two goals on 23—but at a lower rate. Most striking: Montreal was 0-for-6 on high-danger chances, a drought unlikely to repeat.

Colorado's shot volume (36–23) and high-danger count (8–6) showed sustained pressure, yet the 4.60 xG tells a different story than the 7–2 scoreline. The Avalanche outperformed their expected goals by 2.4 — whether elite finishing or variance, it's the difference between a blowout and a competitive game.

Final: MTL 2 – COL 7

Data: NHL API, MoneyPuck xG model

Colorado enters at 60% win probability, with the projected scoreline centered around 3.0 goals for the Avalanche and 2.4 for Montreal. The advantage rests on one foundation: Colorado controls the territorial battle.

At 5v5, the Avalanche own 57.0% xG and 57.0% Corsi against Montreal's 48.0% xG and 49.0% Corsi. Those gaps—nine points in xG, eight in Corsi—show Colorado generates better chances from anywhere on the ice and tilts play into Montreal's end with consistency. This is the game's largest edge.

The caveat is mutual: both teams are running hot. Colorado's PDO sits at 1.021, Montreal's at 1.018. PDO regresses, which means both teams' actual results have benefited from luck. Colorado's momentum—9-0-1 in their last 10—masks underlying regression waiting to happen. Montreal's slump at -7 goal differential over the same stretch carries the opposite implication: their luck ran worse than their process deserved.

Recent form and underlying process are in tension here. Colorado is hotter and deeper in the hole on PDO. Once luck normalizes, their edge shrinks but doesn't disappear.

The goaltending matchup introduces fatigue. S. Wedgewood takes a back-to-back for Colorado; S. Montembeault for Montreal. Both are live starters, though both carry the standard second-game wear.

Colorado's offensive edge flows through Artturi Lehkonen and Ross Colton. Lehkonen generates 1.11 xG per 60 with 18G 22A (6 HD goals, 14.9% shooting)—a profile of volume and finishing. Colton carries similar volume at 1.10 xG per 60 but has just 7G 15A; his 4.9% shooting rate flags him as underperforming his chances.

Montreal counters with Cole Caufield and Oliver Kapanen. Caufield (33G 16A, 0.97 xG/60, 1 HD goal, 20.0% shooting) is the heavier threat—his goal total is real, though his limited HD production suggests his damage comes elsewhere. Kapanen (19G 12A, 0.83 xG/60, 4 HD goals, 17.8% shooting) operates at lower volume and sits on elevated shooting luck.

The model's lean is grounded. Colorado's process edge is substantial, and both teams' PDO suggests room for regression that currently favors the Avalanche. Once that evens out, so does the game—but for now, Colorado owns the matchup.