Colorado controlled this game but lost anyway. The Avalanche held 52% of shot attempts and matched Vegas in expected goals (4.36 to 4.35), yet underperformed their xG by 2.36 goals while Vegas underperformed by just 1.35—a chasm that explained the OT result.

The numbers crystallized in high-danger chances: Colorado generated seven, Vegas six. Both teams converted zero. That tie should have favored Colorado's volume, but J. Eichel punished it at 01:19 of overtime.

The goaltending gap was decisive. C. Hart stopped 30 of 32 shots (0.938 SV%, +2.36 vs expected) and was flawless on high-danger opportunities—7-for-7 when the outcome hung in the balance. M. Blackwood made 26 of 29 (0.897 SV%, +1.35), but even his above-expected performance couldn't overcome Colorado's wastefulness. The Avalanche couldn't convert, and Vegas only needed to convert once.

Colorado generated the chances and controlled the game. They just didn't finish. In a contest where both sides squandered high-danger looks, execution in overtime is everything.

The model projects a 2.4 - 2.7 Colorado victory, giving the Avalanche 55% win probability. The edge is real but modest, rooted in process: Colorado's territorial control is genuine, but masked by unsustainable offensive efficiency that will regress.

The Avalanche dominate possession. They control 57.0% of Corsi and out-expected-goal Vegas 57.0% to 54.0%—a 3-5 point gap in pure driving play. Expect the puck to camp in the Golden Knights' end.

The PDO story is the hedge. Colorado sits at 1.021, overperforming their actual chances through unsustainable shooting and save percentages. Vegas sits at 0.983, running cold and due to regress upward. The Avalanche can't count on repeating their current offensive output; the Golden Knights have margin to improve. Variance typically filters in over multiple games, but a single contest can flip the narrative.

Both teams navigate goaltending uncertainty. M. Blackwood and A. Hill are both on back-to-backs, raising the possibility of backup starts or sputtering starters.

Colorado's offensive load runs through Artturi Lehkonen, whose 1.11 xG/60 and 18 goals with 22 assists makes him the primary engine. His 6 hard-chance goals represent conversion well above league average. Ross Colton (1.10 xG/60, 7G 15A) generates similar chances but converts at only 4.9%—leaving room for positive regression.

Brandon Saad is marked inactive for Vegas. Tomas Hertl fills the offensive role at 0.88 xG/60 with 5 goals, 17 assists, operating at 4.3% shooting. Both teams struggle to score from high-danger chances despite decent volume—a coaching imprint or personnel restraint. Rasmus Andersson anchors Vegas' defense as the weak link; Nazem Kadri does the same for Colorado.

The Avalanche lean is defensible. Their territorial edge is durable, and the model prices in only partial PDO regression. But variance cuts both ways: Colorado's overperforming, Vegas' underperforming, and one game is enough to flip the script. Expect both teams to trend closer to true talent by the final buzzer.