The Islanders stole one. They lost the expected-goals battle 3.25–5.23 and won anyway.
Minnesota's 62% share of expected goals (5.23–3.25) should have won this game. Instead, New York stole it 4–3 in overtime—a disparity so wide it suggests the scoreboard got lost on the way to the rink.
The Wild dominated possession and shot quality. They controlled 56% of shot attempts (36–28), generated nine high-danger chances against four, and averaged 0.080 xG per shot; the Islanders managed just 0.054. Minnesota's depth of play—their ability to create genuine scoring opportunities from every angle—was substantially better than the scoreline suggests.
But the game was decided by variance: I. Sorokin stopped 33 of 36 shots with a save percentage 2.23 goals better than expected, while F. Gustavsson faced 27 shots and allowed three more than his xG model predicted. When one goaltender steals two-plus goals and the other surrenders them, analytics loses.
The turning point was Simon Holmstrom's backhander from the slot at 1:34 of overtime. It wasn't the result of sustained pressure or a series of high-danger setups—it was the kind of goal that comes when a team plays worse but stays dangerous enough to capitalize when the moment arrives.
Q. Hughes recorded three assists in 30:37 for the Wild, orchestrating much of their possession dominance. S. Holmstrom finished with two goals and an assist for the Islanders, the second-line catalyst who doesn't pile up shot attempts the way Minnesota does but shows up when the moment matters. T. DeAngelo (2A in 17:55) anchored the Islanders' blue line.
This was New York's game to lose and Minnesota's game to fail to win—a reversal so complete it will haunt Minnesota's film review.
