Dallas overwhelmed Pittsburgh in underlying metrics—high-danger chances 8-4, expected goals 3.90-3.65—yet needed a shootout to separate from a Penguins team that rode T. Jarry's stellar play and scored on the rush.

The Stars controlled quality despite being outshot 29-23. Dallas generated 0.071 expected goals per shot versus Pittsburgh's 0.053—a gap that should have buried the Penguins. Both teams badly underperformed their models: Dallas by 0.90 goals, Pittsburgh by 1.65. But T. Jarry saved 1.9 goals above expected, and Pittsburgh scored on the rush (2 goals from 27 attempts), staying dangerous despite inferior chances.

Dallas wasted all 8 high-danger attempts, then only broke through when Mikko Rantanen's shootout snap shot from the high slot finally solved the Penguins.

Goaltending. T. Jarry (PIT) made 21 of 23 saves at 0.913 SV%, 1.9 goals above expected—the main reason Pittsburgh nearly escaped. J. Oettinger (DAL) posted 27 of 29 at 0.931 SV%, 1.65 above expected, and was excellent in his own right.

Dallas owns a 52% win probability, according to the model, in what projects as a 2.5–2.6 game. In other words: flip a coin.

The Stars earned that toss-up by posting a 7–1–2 record in their last ten, while Pittsburgh has cooled. But Dallas's run may not hold up to scrutiny. The Stars' PDO of 1.024 sits well above sustainable levels, suggesting some regression ahead. Pittsburgh's luck metric sits at 1.012, marginally healthier. Both teams generate the same expected-goal share at 5v5—51.0% each—but the Penguins own the territorial edge: 50.0% Corsi to Dallas's 47.0%. That's not a dominance margin, but it's the kind of structural advantage that compounds over time.

Goaltending becomes critical. J. Oettinger faces a back-to-back schedule, which means either a backup start or a starter running on fumes—watch the pre-game skate for answers.

Pittsburgh's offense centers on Rickard Rakell, who generates 1.01 expected goals per 60 minutes at 5v5 and has converted at 12.1% (13 goals, 13 assists). Filip Hallander, previously a 1.28 xG/60 threat, is unavailable due to injury, removing a secondary playmaker.

Dallas counters with Jason Robertson, who has scored 25 goals and 21 assists despite an 0.96 expected-goal rate per 60 minutes and just 2 high-danger goals. Michael Bunting, with 8 goals and 14 assists, generates only 0.83 expected goals per 60 minutes on a 7.6% shooting rate, making him a complementary threat at best.

Defensively, Dallas will probe Pittsburgh's weak link in Caleb Jones, while the Penguins should target Tyler Myers on the other end.

The model favors Dallas slightly because its recent surge is real, even if the underlying metrics don't justify it fully. The Penguins hold the structural edge in possession but trail in momentum. This is as close as the league gets to a genuine coin flip.