Colorado won 4–1, but the underlying numbers suggested a closer contest. Expected goals split 3.48–3.22 in the Avalanche's favor—a two-goal lead on paper that became a three-goal margin on the scoreboard.

The explanation: goaltending variance. The Islanders underperformed their expected goals by 2.22; S. Wedgewood (COL) overperformed by the same margin. I. Sorokin (NYI) finished with 24 saves on 27 shots (0.889 SV%), while Wedgewood stopped 28 of 29 (0.966 SV%). Colorado's net-minder stole this game.

Colorado controlled shot attempts at 49%, holding 28 shots to New York's 29. High-danger chances went 5–2 in the Avalanche's favor. The hosts were lethal in transition, scoring two goals on 23 rush chances—an efficiency rate most teams cannot sustain. New York created six rebound opportunities and converted none.

The Islanders underperformed their expected goals while Colorado's goaltending refused to bend. That gap determined the outcome.

The model projects a 2.4–3.0 game favoring Colorado at 60% win probability. That lean is built on territorial dominance, but the Avalanche carry a specific fragility into this matchup.

Colorado controls the underlying game by a significant margin. The Avalanche generate 57.0% xG at 5v5 compared to the Islanders' 48.0%, translating to command of the most dangerous areas. That edge extends to shot attempts: Colorado's 57.0% Corsi means they dictate pace and positioning, forcing the Islanders into reactive play.

The catch: Colorado's 1.021 PDO suggests fortune has been working in their favor. The Islanders sit at 0.993, implying they've been unlucky. PDO regression is rarely dramatic, but it's one of the few measurable forces that can compress a model's confidence. If the Avalanche cool slightly while the Islanders regress toward average, the game tightens.

Goaltending adds another angle. D. Rittich faces a back-to-back; monitor whether Colorado deploys their backup or leans on a fatigued starter. Either scenario introduces risk that the model's 60% projection already prices in, but fatigue matters most in close games.

Offensively, the Islanders rely on Anders Lee's efficiency. Lee generates 1.20 xG per 60 minutes with 9 goals and 20 assists on only 6.4% shooting—he's creating chances but finishing at a conservative rate relative to his opportunity. That shooting percentage offers upside if variance swings his way, but it also signals unsustainability in his favor. Bo Horvat carries 18 goals on 12.8% shooting, a more durable marker of scoring prowess.

Colorado counters with Artturi Lehkonen, who generates 1.11 xG per 60 with 18 goals, 22 assists, and 6 hard-area goals. Lehkonen's 14.9% shooting rate marks him as a finisher who habitually converts; that's the harder nut to play against. Ross Colton, generating 1.10 xG per 60 with 7 goals and 15 assists, contributes more in volume than finishing, but his shot-creation rate keeps Colorado dangerous.

The Islanders' 0% hard-area and rebound-goal rates—alongside Alexander Romanov's defensive vulnerabilities—suggest their back end will absorb sustained pressure. Colorado's parallel gap at net-front dominance seems less concerning given their overall volume; they generate enough that secondary chances cascade into scoring opportunities.

The model's 60% edge reflects a real quality gap. Colorado's PDO is the only meaningful brake on what might otherwise be a stronger lean.