The Bruins won 4-3 in a game that xG says they had no business winning. Vegas controlled play decisively, with 62% of expected goals (5.44 to Boston's 3.37). The Bruins were the worse team on ice; they just had a goaltender playing out of his mind.

J. Korpisalo stopped 30 of 33 shots, bailing out Boston with 2.44 goals of overperformance. Across the ice, A. Schmid faced fewer high-danger chances—the Golden Knights hoarded 11 to Boston's 8—but still allowed the result to slip away, finishing -0.63 against his xG.

The through-line was goaltending, not mechanics. Vegas generated higher-quality looks (0.080 xG per shot versus Boston's 0.058). That precision didn't translate to the scoresheet. The Bruins, meanwhile, were lethal in transition: three goals on 24 rush chances to Vegas's two on 30. Boston also found success in low-danger areas, converting two goals the opposing goalie should have had.

Pavel Dorofeyev's snap shot from a sharp angle (P3, 17:25) stands as the turning point, emblematic of the push Vegas couldn't complete despite the statistical edge.

The Golden Knights underperformed their xG by 2.44 goals. Boston escaped with a heist, winning despite 47% of shot attempts and 38% of expected goals. That's not sustainable, and Vegas's analytics will sting afterward.

Vegas edges Boston in underlying play—but barely. The model projects 2.6-2.4 Golden Knights with 54% win probability, about as close to a pick'em as you'll find.

The Golden Knights own the territorial game. They generate 54.0% expected goals at 5v5 and drive 52.0% of shot attempts. Boston's puck luck runs hotter: their 1.023 PDO means they're overperforming their chances, while Vegas' 0.983 PDO signals regression incoming—exactly the variance that flips close games.

J. Swayman's workload affects Boston's ceiling. On a back-to-back, he may start fresh or give way to a backup.

Vegas' offense hinges on Brandon Saad and Tomas Hertl. Saad (0.91 xG/60, 3G 6A, 1 HD goal, 4.1% shooting) generates volume without consistent finishing. Hertl (0.88 xG/60, 5G 17A, 1 HD goal, 4.3% shooting) shoots at only 4.3%, a low rate that signals conversion upside. Boston counters with Viktor Arvidsson (1.04 xG/60, 18G 24A, 2 HD goals, 13.0% shooting)—finishing at 13.0%, he's the Bruins' lifeline. Tanner Jeannot (0.75 xG/60, 6G 14A, 2 HD goals, 8.3% shooting) outperforms his underlying metrics.

Rasmus Andersson is Vegas' defensive weak link; Elias Lindholm is Boston's. Both teams fail to generate hard-contact or rebound goals—neither tolerates sloppy transitions.

That 54% win probability means luck decides it.