Dallas dominated by the numbers in ways the scoreboard barely reflected. The 3-2 final masked a 65-35 expected-goal split — the Stars generated 5.18 xG to Nashville's 2.80, the kind of gap that usually yields a three-or four-goal margin, not a single one.

The Predators had the shot volume (47% of attempts, 25 shots on goal to Dallas' 28) but not the danger. Nashville generated just 0.049 xG per shot; Dallas created three times that rate at 0.096. High-danger chances went 3–9 in Dallas' favor.

Here's where the evening fractured from the scoreboard: C. DeSmith stopped 23 of 25 shots (0.920 SV%), beating his expected save percentage by 0.80 goals — solid work. But J. Annunen, Nashville's goaltender, stopped 25 of 28 and finished 2.18 goals above expected, effectively stealing back what Dallas' dominant underlying play should have delivered. Annunen was near-perfect on high-danger attempts, saving all nine.

Dallas got their chances. They went 0 for 9 on high-danger looks — a brutal conversion rate that won't repeat. They were lethal off the rush, though, converting three of 24 chances. M. Rantanen's wrist shot from the slot in the third period exemplified Dallas' efficiency on simpler plays.

The Stars' depth offensive attack shone in their top names: J. Robertson collected two assists in nearly 25 minutes of ice time, while T. Harley chipped in one assist from the blue line.

Nashville created four rebound opportunities and converted one. Dallas created five and converted none — marginal differences, but they add up when your goaltender isn't bailing you out.

The underlying dominance eventually mattered, even if barely.

The model sees Dallas and Nashville as dead even: the Predators at 50% win probability, the projected score a flat 2.5-2.5. This is about as close to a pick'em as you'll find.

The problem is form and fortune. Nashville sits underwater at -7 goal differential over their last 10, while Dallas carries a PDO of 1.024—the Stars are running hot, and regression is coming.

On the ice, Nashville owns territorial advantage. The Predators drive Corsi at 50.0%, Dallas at 47.0%. But expected goals tell a different story: Dallas leads 51.0% xG% to 49.0%. Nashville's PDO of 0.992 sits closer to sustainable; the team is generating but not converting at the rate underlying play suggests.

Watch Jason Robertson for Dallas. He carries 0.96 xG/60, has 25G 21A with two high-danger goals, and is shooting at 12.4%. Michael Bunting operates at 0.83 xG/60 with 8G 14A on a 7.6% shooting rate—secondary, but still a threat.

Nashville counters with Zachary L'Heureux at 0.85 xG/60 with 4G, one high-danger goal, and a 16.0% shooting rate. His low output despite strong underlying numbers suggests a regression brewing his way. Filip Forsberg, meanwhile, sits at trend: 0.80 xG/60, 23G 17A, four high-danger goals, and 13.6% shooting.

A wrinkle: neither team shows depth scoring from high-danger chances. Both list 0% HD goal rate and 0% rebound goal rate. In a close contest, that's a liability. Dallas's defensive weak link is Tyler Myers; Nashville leans on Jonathan Marchessault.

The projection lands here because neither side owns an obvious edge. Dallas has puck-luck regression hanging over it; Nashville has recent form working against it. Robertson could tilt Dallas's way. Forsberg and an L'Heureux correction could tilt it Nashville's direction. The Predators' territorial control doesn't guarantee an expected-goals win.

This is genuinely a coin flip.