The Islanders stole one. They lost the expected-goals battle 2.92–4.03 and won anyway.
The Islanders stole this one. Down 58–42 in expected goals, Tampa Bay had no business losing 2–0 to New York—yet here we are.
The Lightning controlled everything but the scoreboard. They held 63% of shot attempts (32–19), created more high-danger chances (6–4), and finished with 4.03 expected goals to the Islanders' 2.92. But I. Sorokin didn't care. He stopped 32 of 32, running up a +4.03 gap between his saves and what the xG model predicted. When a goalie beats his expected performance by that margin against a team that dominated the game, that's the difference between a blowout loss and a shutout.
The turning point came at 19:58 of the third: Casey Cizikas's backhander from the wing.
The Lightning's execution crumbled where it mattered most. They went 0-for-6 on high-danger chances—a brutal conversion rate that won't repeat. They also created six rebound opportunities and converted none, showing they could generate chances but couldn't finish them. J. Johansson wasn't the problem; he actually outperformed his expected goals by 1.92. But his team couldn't give him a chance to steal it back.
This was a variance game. The Islanders underperformed their xG by 0.92 goals—a rounding error. The Lightning underperformed theirs by 4.03. That gap, more than any single play, tells you what happened: one team capitalized on its chances, the other didn't.
Sorokin put on a clinic on high-danger chances, turning away all six. That's not luck; that's a wall showing up when the game is on the line.
