This is about as close to a pick'em as hockey gets. The model projects a 2.4-2.7 expected goal range with Dallas at 55% win probability.
Dallas owns the underlying play by expected goals: 51.0% xG compared to New York's 48.0%. But the Islanders claim the possession edge with 49.0% Corsi% versus Dallas's 47.0%—a reminder that shot volume and shot quality don't always align. The Stars are generating better chances despite fewer touches.
The red flag for Dallas is sustainability. Their PDO sits at 1.024, well above the league mean, suggesting some regression is coming. When a team's luck breaks—when bounces stop going their way and goaltenders stop stealing games—that 55% projection evaporates fast. The Islanders, meanwhile, are running a 0.993 PDO, closer to neutral.
I. Sorokin enters on the second leg of a back-to-back, which either means a tired starter or a backup getting the net. Either way, it's a wrinkle.
At even strength, both teams generated 0% of their goals from high-danger situations and 0% on rebounds—a statistical oddity that suggests depth scoring or defensive execution has been unusually clean. That can't last, and whoever corrects first wins the game.
The Stars lean on Jason Robertson, who has 25 goals and 21 assists while maintaining a 0.96 xG per 60 minutes at 5v5. At 12.4% shooting, he's converting at a rate that's elevated but not unsustainable. Michael Bunting (0.83 xG/60, 8 goals) gives Dallas a secondary scoring layer with an impressive 7.6% shooting rate.
For New York, Anders Lee remains the catalyst: 1.20 xG per 60, 9 goals, 20 assists, and 2 high-danger goals despite a modest 6.4% shooting rate—meaning he's still leaving points on the board. Bo Horvat (0.87 xG/60, 18 goals, 12.8% shooting) is running hotter, with 3 high-danger goals already, and is the Islanders' most dangerous bet to cash in first.
Alexander Romanov is New York's defensive weak link. Tyler Myers is Dallas's. Whichever side exploits first wins.
The model is right. This is a coin flip with the coin still spinning.
