Washington won 5–4 in a shootout despite losing the xG battle 4.36–2.17, a 67–33 deficit that should have buried them. Vegas controlled 67% of expected goals at 5v5. This was a heist that didn't pay.

The metrics reveal a lopsided game. Vegas out-attempted Washington 29–22 on shots and generated six high-danger chances to zero. Their shot efficiency—0.070 expected goals per attempt—dwarfed Washington's 0.049. Yet on the rush, Washington was lethal: four goals on 25 chances to Vegas's three on 31. One team maximized scarce chances; the other wasted dominance.

A. Hill (Vegas) faltered against his workload, allowing 1.83 goals above expectation on 22 shots. L. Thompson (Washington) stood tall with 0.862 in save percentage despite 29 attempts. Vegas created three rebound opportunities and buried zero. Washington's power play was surgical: two goals on 11 shots.

Washington overperformed their xG by 1.8 goals—elite finishing or fortune will show on tape. Dylan Strome's backhander from the doorstep in the shootout decided it, transforming a night where Vegas deserved a comfortable margin into an upset.

J. Eichel (Vegas) paced the play with 1 goal and 2 assists in 23:23. R. Andersson (Vegas) added 1 goal, 1 assist in 23:20. J. Sourdif (Washington) recorded 1 goal, 1 assist in 17:34.

The model projects a 2.4–2.7 goal Vegas win and gives the Golden Knights 56% win probability. The edge rests on underlying play and PDO regression: Vegas controls shot quality but runs cold; Washington runs hot and due for a correction.

The Capitals are the hotter team by recent points pace, going 5-3-2 over their last 10. Vegas, meanwhile, sit underwater at -7 goal differential across the same span—a sign of execution failure more than competence.

At 5v5, Vegas owns the territorial advantage. The Golden Knights generate 54.0% xG to Washington's 50.0%, and drive Corsi at 52.0% against 49.0%. These metrics forecast that Vegas should control play, dictate pace, and create more dangerous looks. The Capitals will spend time in their own zone.

The PDO split tells the variance story. Vegas runs at 0.983—cold, due for positive regression. Washington sits at 1.018—hot, due for regression of its own. Vegas's underlying dominance hasn't translated to results; Washington's lower underlying numbers have. When these normalize, the model favors Vegas.

Both teams face goaltending questions. A. Hill faces a back-to-back, as does L. Thompson. Watch for a backup start or tired starter from either side.

Vegas's edge flows through star power and shot generation. Alex Ovechkin remains Washington's engine at 19 goals, 17 assists, and 0.97 xG per 60 minutes—elite volume and efficient production paired with two high-danger goals. Anthony Beauvillier runs 10.4% shooting on 0.89 xG/60, suggesting both volume and luck; expect some normalization.

For Vegas, Brandon Saad projects at 0.91 xG/60 but carries a marking as likely out—check the latest lineup before puck drop. Tomas Hertl (5 goals, 17 assists, 0.88 xG/60) is the reliable secondary option at 4.3% shooting, indicating he's still finding low-danger looks despite the assist total.

Washington's defensive weak link is Declan Chisholm; Vegas's is Rasmus Andersson. Neither team shows strong high-danger or rebound conversion at the macro level, so the edge is in sustained play dominance, not breakdowns.