The Rangers' 4–3 overtime victory masked a game where the Oilers controlled the underlying metrics—and left real points on the ice through a combination of their own underperformance and I. Shesterkin's excellence.

Edmonton generated 4.32 expected goals to New York's 3.89, a 53–47 advantage. The Oilers also won the high-danger chances, 6–3, and held 51% of shot attempts (36–35 on goal). By every structural measure, this was an Oilers game. By the scoreboard, it wasn't.

The breakdown: Edmonton underperformed their xG by 1.32 goals—they received chances worth 4.32 goals and converted only three. Meanwhile, I. Shesterkin stopped 1.3 goals above his expected mark, posting a 0.917 save percentage despite Edmonton's high-quality assault. On the other end, S. Skinner recorded 31 saves on 35 shots (0.886 SV%).

The Oilers' rush game became their undoing. They scored just twice on 34 rush attempts. The Rangers, by contrast, buried four on 32 attempts, and when they crashed the net for rebounds, they generated four opportunities (though converted none).

J.T. Miller ended it with a wrist shot from the slot at 02:49 of overtime, converting the single chance that mattered most.

Three Stars

• #25 D. Nurse (EDM) — 2G in 21:00
• #4 B. Schneider (NYR) — 1G, 1A in 20:48
• #96 J. Walman (EDM) — 2A in 18:42

Final: NYR 4 – EDM 3 (OT)

The model gives Edmonton a 55% win probability in a projected 2.4–2.6 game. That edge rests on clean underlying play, but one variable clouds the forecast: regression.

Edmonton controls the shot equation convincingly. They own 52.0% expected goals versus New York's 48.0%, and nearly identical Corsi supremacy at 50.0% to 49.0%. The Oilers are winning possession and shot quality consistently. Yet their 0.982 PDO—a luck-adjusted metric of actual goals versus expected—reveals they're underperforming their shot talent. When PDO regresses upward, Edmonton's already-slim edge widens.

New York's 1.009 PDO is the counterpoint: the Rangers are running hotter than underlying numbers suggest.

Goaltending adds noise. S. Skinner and J. Quick are both on back-to-backs, creating a coin flip: either a backup appears, or the starter plays on fumes. Whichever team avoids the tired-goalie trap gains a meaningful edge.

Offensively, the talent distribution is lopsided. Zach Hyman leads Edmonton's offense at 1.23 xG/60, translating into 14 goals and 13 assists on a sharp 13.0% shooting rate with one high-danger goal. Connor McDavid's 0.92 xG/60 understates his role: 25 goals and 33 assists with five high-danger tallies represent the engine that will force New York into reactive defense.

New York's attack profiles differently. Adam Sykora operates at 0.91 xG/60 with a remarkable 20.0% shooting rate (3 goals in limited looks), indicating shot quality mastery. Gabe Perreault's 0.79 xG/60 and 13.1% shooting pace yield 8 goals and 2 high-danger tallies—he's finding premium locations despite lower creation volume.

Defensively, both teams have liabilities. Jake Walman is Edmonton's weak link; Vincent Iorio is New York's. The model assumes neither team fully exploits these avenues, but if either side executes the transition game, these names become the first pressure points.

The model's lean is sound: Edmonton's possession edge, even discounting for PDO regression, should prevail. The 55% probability already prices in regress-to-mean luck. Watch for tired goaltending to accelerate that margin.