The Islanders won 2–1 despite being outchanced, because I. Sorokin played exceptionally well while the Rangers squandered their opportunities. The Rangers created a slight edge in expected goals (2.67–2.59) and dominated high-danger chances (6–2), but scored only once, underperforming their xG by 1.67 goals. Sorokin, by contrast, saved 1.67 goals above expected, stopping all 6 high-danger shots at 100% SV.

The Rangers' inefficiency was brutal. They generated 0 goals from 6 high-danger chances—a conversion rate that won't repeat. They also created 5 rebound opportunities and converted none. Meanwhile, the Islanders scored twice from low-danger spots, the kind of luck that requires an opposing goaltender to fail. J. Quick (19/21, 0.905 SV%) actually played fine; Sorokin (20/21, 0.952 SV%) simply played better.

Mika Zibanejad's slap shot from the top of the circles in the third period gave the Rangers their one goal at 2:48, but it wasn't enough. The Islanders controlled the result through a combination of Sorokin's brilliance and New York's inability to finish despite multiple clear-cut chances.

The shot-attempt battle finished even at 21–21. The high-danger-chance disparity—6–2 in the Rangers' favor—suggested they should have won this game. That they didn't speaks to Sorokin's performance and the Rangers' execution at the finish line.

The model projects a 2.4–2.6 goal game favoring New York, with the Rangers at 53% win probability—as pure a coin flip as projection modeling produces.

The underlying metrics explain the stalemate. Season 5v5, both teams sport identical 48.0% expected goal shares and 49.0% Corsi percentages. The only separation is luck: the Rangers' PDO sits at 1.009 against the Islanders' 0.993, meaning New York's goals-for and goals-against have both run slightly in their favor. Strip that away and this is a dead-even matchup.

Recent form tilts Islanders' way. They're 5–4–1 over their last 10, riding genuine momentum. The Rangers, meanwhile, are underwater at minus-19 goal differential over the same span—a sign that their lucky PDO isn't translating to wins when the details matter.

Anders Lee (1.20 xG per 60, nine goals, 20 assists, 6.4% shooting) drives the Islanders' expected output; Bo Horvat (0.87 xG per 60, 18 goals, 12.8% shooting) converts that into goals at an efficient rate. Adam Sykora (0.91 xG per 60, 20.0% shooting) and Gabe Perreault (0.79 xG per 60, eight goals, 13.1% shooting) shoulder the Rangers' offensive load.

Defensive exposure exists for both. The Islanders lean on Alexander Romanov at the back; the Rangers on Vincent Iorio. Both have gaps—not atypical for playoff-contention teams, but worth tracking when margins are this thin.

The verdict sits exactly where the math lands: neither team commands a structural edge. The Islanders have momentum, the Rangers have run-of-play equality and a favorable PDO. One will shade out the other, but the projection models can't cleanly pick it. You shouldn't pretend to, either.