Despite a 37–28 shot advantage, New York fell to Pittsburgh 4–3. The underlying numbers were nearly a coin flip—xG favored Pittsburgh 3.79–3.63—but the Penguins' finishing and Jarry's goaltending made the difference.

Expected goals barely moved the needle, yet high-danger chances told a harsher story. Pittsburgh dominated 9–1, a gap that forecasted the result more accurately than raw shot volume. This is what separates a shot and a scoring chance: the Islanders fired from distance; the Penguins shot when it mattered.

Jarry's 0.919 save percentage (+0.63 versus expected) was crucial. Sorokin faced a more efficient attack: 24 saves on 28 shots (0.857 SV%), but the Penguins' accuracy from dangerous ice wore him down.

Pittsburgh weaponized the rush. Three goals came in transition on just 24 chances—a lethal rate. The power play added two more on ten attempts, generating 1.19 expected goals. New York matched Pittsburgh's three rush goals but needed 35 chances to get there, underscoring the gulf in efficiency.

The turning point came early: Justin Brazeau's backhander from the doorstep at 14:21 of the third period gave Pittsburgh breathing room.

Three Stars

• #71 E. Malkin — 3A in 18:11
• #67 R. Rakell — 1G, 1A in 20:01
• #87 S. Crosby — 1G, 1A in 19:30

The model projects Pittsburgh favored at 55% win probability—essentially a pick'em. The underlying play belongs to the Penguins, who control 51.0% expected goals at 5v5 versus the Islanders' 48.0%. But possession metrics (50.0% Corsi% for PIT, 49.0% for NYI) reveal how thin that edge is.

Pittsburgh is the hotter team by points pace, winning 6-1-1 over their last ten games. New York has struggled, sitting at -6 goal differential in the same span, a cumulative signal that luck hasn't masked deeper problems—their PDO of 0.993 trails Pittsburgh's 1.012, meaning the Penguins are both playing better and getting the bounces.

The clearest tactical edge belongs to New York in the rest department. The Islanders enter on seven days without playing; Pittsburgh gets just one. That recovery gap could matter, especially if goaltender A. Silovs shows fatigue from back-to-back duty.

Anders Lee carries the Islanders' offensive load at 1.20 expected goals per 60 minutes, backed by 9 goals and 20 assists. His 6.4% shooting rate is pedestrian—he's due for positive regression if he stays healthy. Bo Horvat, at 0.87 xG/60 with 18 goals and 3 high-danger chances, has converted opportunities at a 12.8% rate, a harbinger of his actual value outpacing raw chance creation. Pittsburgh will need to neutralize both to limit seepage.

On the flip side, Rickard Rakell generates 1.01 xG/60 and backs that with 13 goals, 13 assists, and a 12.1% conversion rate—a genuine indicator that he's consistently positioned to score. Filip Hallander's 1.28 xG/60 would otherwise be worth watching, but he's likely unavailable. That's a net loss for Pittsburgh's upside.

Defensively, neither team has distinguished itself in high-danger or rebound scoring, a weakness both will want to address. The Penguins' vulnerable link is Caleb Jones; the Islanders lean on Alexander Romanov as a potential leak.

Pittsburgh controls the underlying play, but that edge is marginal—close enough that a well-rested Islanders team could disrupt the script. The model sees it as a coin flip. The numbers agree.