Colorado escaped the scoreboard by overperforming its expected goals by 3.60 — a historic margin that separated an 8-4 victory from what the numbers suggested was a much tighter affair. The underlying play told a different story: the Avalanche dominated 5v5 expected goals 4.40–3.12, but the final margin more than doubled that advantage.

Shot volume favored Colorado decisively — 42 attempts to 30, with expected goals matching that 58–42 split. The Avalanche also dominated the high-danger chances category, 7–4, the kind of edge that typically predicts the score. But Colorado's finishing was elite; they scored 3 goals from low-danger spots, the kind of bounces that require defensive breakdowns or exceptional luck.

New Jersey's goaltending collapse made the margin worse than the play. J. Markstrom faced the consequences of Colorado's efficiency, surrendering 3.60 more goals than expected on a 42-shot workload. S. Wedgewood, by contrast, underperformed his expected goals by 0.88, keeping the deficit manageable.

The Avalanche showed their precision in other ways. They generated 0.067 expected goals per shot — a quality metric that exceeds New Jersey's 0.050 — proving volume and placement both mattered. Colorado was particularly dangerous off the rush, converting 7 goals on 35 rush chances, where speed and spacing eliminate defensive structure.

The turning point arrived early in the second when Victor Olofsson's wrist shot from the top of the circles found the net at 17:38. That moment summarized the night: Colorado was sharper, more decisive, and more willing to attack high-danger areas. By the final buzzer, they had turned a convincing process into a lopsided scoreline.

Colorado is favored here, but modestly and conditionally. The model projects a 2.4-2.9 game with the Avalanche at 59% win probability—an advantage rooted in control and mean reversion, not dominance.

The Avalanche own the play. They generate 57.0% of expected goals at 5v5 against New Jersey's 50.0%, a seven-point gap that shapes the flow of possession. Corsi reinforces the pattern: the Avalanche at 57.0% to the Devils' 51.0%, another edge in the same direction. Territorial control this stark usually predicts wins, but hockey reserves the right to surprise.

The complication is PDO. Colorado's 1.021 signals a team running hot, converting chances at rates that don't persist. New Jersey's 0.971 signals the inverse—they're running cold with positive regression waiting. When lucky teams face unlucky ones, variance often has other plans. If Colorado's finishing regresses and New Jersey's improves, the Avalanche's seven-point xG advantage shrinks to nothing.

The forward matchup favors Colorado in less obvious ways. Timo Meier drives New Jersey at 1.08 expected goals per 60 but shoots only 6.7%, producing 13 goals and 15 assists—a gap suggesting his underlying creation outpaces his scoring. Jack Hughes operates at lower volume (0.88 xG/60) but at an 11.5% shooting rate, banking 18 goals and 17 assists. New Jersey's game plan leans on these two, hoping their streaky conversions hold.

Colorado counters with depth. Artturi Lehkonen sits at 1.11 expected goals per 60 and shoots an elite 14.9%, totaling 18 goals and 22 assists—elite at both ends. Ross Colton mirrors the volume at 1.10 xG/60 but at a thinner 4.9%, posting 7 goals from 15 assists. Colorado's edge isn't star power; it's that both wings generate and convert reliably, while New Jersey depends on two inconsistent finishers.

Defensively, both carry chinks. New Jersey's Johnathan Kovacevic is marked weak, a target Colorado's forward depth should attack. Colorado's Nazem Kadri bears the same burden. These gaps don't swing outcomes alone, but they offer tactical entry points.

The model's edge stands on possession and regression momentum. Colorado should win on the play. Whether they do depends on which PDO converges first.