The Islanders beat the Lightning 3-2 in a shootout despite getting dominated in expected goals—controlling just 36% of the xG at 2.85 to Tampa's 5.06. This wasn't a close game that happened to break against the Lightning; it was a mugging.

Tampa controlled 64% of expected goals at 5v5, generating the type of shot quality and volume that typically swamps opponents. The Lightning were relentless. They just weren't effective.

I. Sorokin won this game. He stopped 3.06 goals above expected with a 0.941 SV%, including a perfect 9-for-9 on high-danger chances. J. Johansson, defeated, posted a 0.882 on 32 saves—a respectable performance buried under an avalanche of Tampa chances.

Tampa's shootout loss looked worse in context. The Lightning went 0-for-9 on high-danger chances—a misfire rate that's typically noise, but damning when combined with overwhelming volume. They couldn't finish. The Islanders were more efficient on the rush, scoring 2 of 17 chances to Tampa's 2 of 28—a distinction that shouldn't have mattered at this expected-goal deficit, yet did.

The shootout's turning point: Emil Heineman's wrist shot from the high slot, putting the Islanders over. This was a reprieve for New York, not a referendum.

The model sees Tampa at 53% to win—essentially a pick'em: Lightning by 2.7 expected goals to 2.5. On paper, that should favor the road team. In reality, it reflects a game where neither team can afford mistakes.

Tampa owns the underlying play. At even strength, the Lightning are generating 54.0% of expected goals to the Islanders' 48.0%, and controlling 53.0% of shot attempts to 49.0%. That's the kind of territorial dominance that wins most nights. But they're running hot: a 1.019 PDO means luck has inflated their actual performance relative to the quality of chances they're creating and allowing. Regression is overdue. The Islanders sit at 0.993—they're slightly underperforming what they should generate, which cuts the other way.

Both starting goalies are on back-to-backs. D. Rittich (Tampa) and J. Johansson (New York) could be rested or exhausted. Tired goaltending tends to decide close games, and this one will be decided by marginal moments. Watch whether either team opts for a backup.

Brandon Hagel is Tampa's engine: 1.10 expected goals per 60 minutes, 23 goals and 26 assists, 2 high-danger goals. His 15.8% shooting rate is elite and sustainable. Anthony Cirelli operates differently—0.95 xG/60 but 17.0% shooting and 5 high-danger goals, which means he finishes at an exceptional rate on the looks he gets.

New York's best matchup advantage is Anders Lee. At 1.20 xG/60, he leads both rosters in shot quality. Only 9 goals to Hagel's 23, but 20 assists and just 6.4% shooting suggests variance will eventually favor him. Bo Horvat, at 0.87 xG/60 with 18 goals and 12.8% shooting, is their secondary threat and has been efficient.

The metrics reinforce the toss-up. Tampa should win on talent and shot quality—factors that matter over seasons or series. In a single game, especially with compromised goaltending and one team running hot, neither edge is large enough to bet against the other. The model is right: flip a coin.