Dallas won despite fewer shots—31 on goal for Seattle versus 25 for Dallas—a sign of dominance through efficiency. Expected goals bore this out: 3.82–2.88 for Dallas, a 57–43 advantage. The Stars generated 0.071 xG per shot to Seattle's 0.050. Dallas was lethal off the rush, converting 2 goals on 29 chances, while Seattle's 3 rebound opportunities produced nothing.

Both teams underperformed their expected goals against stellar goaltending. Dallas generated 3.82 xG but scored 2; Seattle generated 2.88 but scored 1. C. DeSmith allowed just 1 goal on 31 shots (0.968 SV%, +1.88 above expected), but the high-danger disparity was stark: Dallas generated 6, Seattle 1. M. Murray stopped 23 of 25 (0.920 SV%, +1.82 vs expected) but couldn't overcome the quality advantage. Tyler Seguin's backhander from the doorstep (19:26, P1) sealed it.

Dallas is the model's lean at 56% win probability, with the projection landing at 2.2–2.5 goals scored. The edge stems from superior underlying play—the Stars generate 51.0% expected goals to Seattle's 46.0%—but that advantage comes partially baked into an inflated PDO of 1.024. PDO regression is the tax on Dallas's outperformance.

The Stars' possession metrics support the model. At 5v5, they hold a 47.0%–45.0% Corsi advantage, reflecting both shot volume and quality control. Seattle, by contrast, is underwater at -7 in goal differential over its last 10 games, a sign the Kraken are generating more chances against than for.

A close contest would mean Dallas's PDO regression reverses some of the underlying advantage. Seattle's 1.011 is more sustainable, though still above neutral. When unsustainable scoring environments collapse—and they do—Dallas loses a margin that looks comfortable today.

The Stars get a rest edge; they come in on one day off to Seattle's three. More consequential is J. Oettinger's schedule: he's on the back end of a back-to-back and may not start. A tired Oettinger or a backup goalie significantly narrows the Stars' margin.

Shane Wright drives Seattle's offense at 0.77 expected goals per 60 minutes, posting eight goals, ten assists, and four high-danger chances on a 10.4% shooting rate—elite conversion that outpaces even his strong shot quality. Berkly Catton ranks second at 0.68 xG/60, with six goals, eight assists, and five high-danger chances at 8.8% shooting.

Jason Robertson anchors Dallas at 0.96 xG/60, 25 goals, 21 assists, and just two high-danger goals, signaling he's earned his totals through volume rather than variance. Michael Bunting provides secondary scoring at 0.83 xG/60, eight goals, 14 assists, and a 7.6% shooting rate that underperforms relative to his shot attempts.

The model is right to favor Dallas. The underlying play is real. But PDO regression and a potentially compromised Oettinger keep this from being a lock.