New York won 5–4 in a shootout despite being dominated on expected goals, 2.60 to 4.90. The Islanders engineered this heist through goaltending and rush scoring — a win built on variance favoring goaltender efficiency.

Vegas controlled possession and shot quality. The Golden Knights averaged 0.079 xG per shot against the Islanders' 0.049. New York compensated by attacking in transition: four goals on 29 rush chances beat Vegas's two on 27.

Emil Heineman's wrist shot from the high slot in the shootout proved decisive.

Goaltending was the difference. I. Sorokin stopped 33 of 37 shots (0.892 SV%, +0.90 vs expected), surpassing his workload. C. Hart (VGK) moved the opposite way: 23 saves on 27 shots (0.852 SV%, −1.40 vs expected). Hart surrendered 1.4 more goals than expected.

The underlying counts confirmed Vegas's edge: 37 shot attempts to New York's 27. High-danger chances extended that gap to 9–3. New York created three rebound opportunities (converting none) while Vegas created eight (converting two).

B. Horvat (NYI) led scorers with two goals in 22:19. J. Eichel and M. Stone (VGK) each notched two assists in similar ice time (24:30 and 22:32, respectively).

The model projects a 2.6-2.4 Vegas win with the Golden Knights at 53% win probability. That qualifies as a genuine coin flip—there's no clarity in the numbers.

Vegas owns the underlying play. The Golden Knights control 54.0% of expected goals at 5v5, compared to the Islanders' 48.0%. Possession metrics amplify the advantage: Vegas drives the Corsi at 52.0% versus New York's 49.0%. Territorial dominance usually compounds into goals, but the Knights are running ice-cold. Their 0.983 PDO lags well behind neutral; the Islanders' 0.993 sits closer to normal. That imbalance suggests Vegas is the side due for positive regression.

The Islanders' goaltending introduces a wildcard. C. Hart is on a back-to-back, which typically means either a backup enters or a tired starter gets the workload. Models can't price that uncertainty in.

Brandon Saad (0.91 xG/60, 3G 6A, 1 HD goal, 4.1% shooting) is marked inactive, removing a layer from Vegas' attack. That leaves Tomas Hertl as the primary engine—0.88 xG/60, 5G 17A, 1 HD goal, 4.3% shooting. For New York, Anders Lee anchors the offense (1.20 xG/60, 9G 20A, 2 HD goals, 6.4% shooting), while Bo Horvat supplies volume (0.87 xG/60, 18G 10A, 3 HD goals, 12.8% shooting). Horvat's 12.8% shooting rate is inflated; regression is waiting.

Defensively, gaps matter. Rasmus Andersson is Vegas' weak link. Alexander Romanov fills the same role for New York.

Players to Watch

• Tomas Hertl (VGK): 0.88 xG/60, 5G 17A, 1 HD goal, 4.3% shooting
• Anders Lee (NYI): 1.20 xG/60, 9G 20A, 2 HD goals, 6.4% shooting
• Bo Horvat (NYI): 0.87 xG/60, 18G 10A, 3 HD goals, 12.8% shooting

Vegas holds the edge in underlying play and should benefit from PDO bounce. But goaltending uncertainty and New York's scoring depth—particularly Lee—narrow the margin. Vegas at a slim favorite.