High-danger chances were tied at 2–2, yet the Golden Knights won 4–1 — the extra goals came from places the Kraken defense should have cleaned up.

The xG told the real story: Vegas controlled 67% of five-on-five expected goals (3.29–1.62), a decisive edge that mirrored the final scoreline. The Golden Knights didn't just out-shoot Seattle (26–23 on goal, 53% of attempts) — they created far more dangerous looks. Vegas generated 0.051 xG per shot versus Seattle's 0.039, pure quality over volume.

But the scoreboard exaggerated Vegas's dominance at the highest level. Two Golden Knights goals came from low-danger areas — the kind opposing goalies are supposed to stop. Two more came off the rush (29 rush attempts yielded 2 goals), where Seattle's structure broke down rather than Vegas executing a systematic pattern.

C. Hart made this look easy, posting a 0.957 save percentage against 23 shots — 0.62 goals better than expected. N. Kokko's 0.846 mark represented a 0.71-goal shortfall, a troubling gap when you're already being out-chanced by that margin. Rebounding didn't decide it: Vegas created four rebound chances and converted one, Seattle created six and also converted one. The difference was ruthlessness — Vegas punished the looks it got.

Vegas overperformed its xG by 0.71 goals — not a small variance when 3.29 xG yielded 4 actual goals. That's execution meeting opportunity.

The model has Vegas at 59% win probability in a projected 2.1–2.6 game—a meaningful edge that's grounded in the two-week form disparity and a deeper structural advantage that PDO regression could amplify.

Vegas is the hotter team by record, going 6–1–3 over their last ten. Seattle has deteriorated to minus-14 goal differential in the same span, a gap that forecasts poorly for the Kraken's chances.

The underlying play edges further toward Vegas. At 5v5, the Golden Knights own 54.0% expected goals against Seattle's 46.0%, and they drive territorial control at 52.0% Corsi% versus 45.0%. This isn't luck—it's systematic dominance in how both teams play the game.

But here's where the edge sharpens. Vegas's PDO of 0.983 signals they're running cold relative to their underlying performance. Positive regression is mathematized, not promised, yet the gap between play and results is the kind of mismatch models exploit. Seattle's PDO sits at 1.011—closer to neutral—which means they're not getting lucky; they're simply being outplayed.

The goaltending angle adds friction. Both C. Hart and N. Kokko are on back-to-backs, creating ambiguity around lineup deployment. If Seattle plays a backup, the model's edge widens. If Vegas does, it narrows. This is noise the model prices in but can't eliminate.

The individuals who matter most tell the same story. Shane Wright drives Seattle's offense at 0.77 expected goals per 60 minutes with 8 goals and 10 assists, anchored by 4 high-danger chances—his 10.4% shooting rate is sustainable. Berkly Catton complements him at 0.68 xG/60 with 6 goals and 8 assists. Vegas counters through depth: Brandon Saad operates at 0.91 xG/60 with only 3 goals (a 4.1% shoot rate that screams positive regression), and Tomas Hertl at 0.88 xG/60 has cashed 5 goals on 17 assists with just 1 high-danger goal, leaving room for offensive upside. Hertl's efficiency on volume is the kind of variance that resolves favorably.