Montreal's 7-3 victory masks a stark underlying story: Colorado controlled 57% of expected goals, yet Montreal turned a 2.88–3.79 xG deficit into a 7-3 scoreline. This wasn't dominance in open play — it was clinical execution in the spots that mattered.

The gap between xG and result reveals two truths. J. Dobes (MTL) stopped 26 of 29 shots (0.897 SV%), overperforming his expected outcome by 0.79 goals. S. Wedgewood (COL) faced 28 shots, allowed 21, and underperformed his xG by 4.12 — a chasm that explains the lopsided final. Montreal, meanwhile, finished 4 for 4 on high-danger chances; Colorado generated 9 but converted only 2 of them.

The rush proved decisive. Montreal scored 6 times off 29 rush attempts; Colorado scored twice from 27. That efficiency margin — and Juraj Slafkovský's snap shot from the doorstep in the third period — gave Montreal breathing room that xG suggests they shouldn't have had.

Montreal created 3 rebound opportunities and capitalized on 1. Colorado did the same: 3 rebounds, 1 goal. The battle at the net was even; the results were not.

This was variance tilting heavily in Montreal's favor — they overperformed their xG by 4.12 goals — but it didn't emerge from nowhere. High-danger finishing and rush execution are repeatable skills, not randomness. Whether this performance was elite execution or fortune likely won't resolve until these teams play again.

Final: COL 3 – MTL 7

Colorado should win this game. The model projects 56% win probability for the Avalanche and a 2.8-2.5 scoreline in their favor, built on territorial dominance that figures to persist.

The Avalanche own the underlying play at 57.0% expected goals at 5v5, compared to Montreal's 48.0%. Their Corsi edge of 57.0% to 49.0% tells the same story: Colorado will spend the game in Montreal's end and control the territorial flow.

Both teams are running hot on PDO—the Avalanche at 1.021, the Canadiens at 1.018—which introduces a complication. PDO regression is real. Both teams face normalization headwinds that will likely erode some of the model's baseline confidence.

Rest tilts toward Colorado substantially. Montreal arrives on one day of recovery; the Avalanche have four. That gap compounds through sixty minutes. J. Dobes is working a back-to-back, a factor to monitor for Montreal's goaltending. Colorado carries no corresponding liability in goal.

Artturi Lehkonen drives Colorado's offense at 1.11 expected goals per sixty minutes, posting 18 goals, 22 assists, and 6 high-danger goals on a 14.9% shooting rate. Ross Colton contributes 1.10 xG/60 with 7 goals, 15 assists, and 1 high-danger goal on a 4.9% shooting rate.

Montreal counters with Cole Caufield, who carries 33 goals and 16 assists at 0.97 xG/60, though only 1 high-danger goal alongside a 20.0% shooting rate. Oliver Kapanen provides 0.83 xG/60 support with 19 goals, 12 assists, and 4 high-danger goals at a 17.8% shooting rate.

Nazem Kadri functions as Colorado's defensive weakness; Mike Matheson plays that role for Montreal. Both teams' respective liabilities will face opposing forwards seeking to exploit mismatches.

Colorado's territorial command, rest advantage, and goaltending setup create the edge. PDO regression introduces uncertainty, but not enough to flip the Avalanche's lean.