The model projects 2.6 goals for each team, with Pittsburgh at 51% to win—as close to a pick'em as the numbers offer. This game has no strong conviction either way.
Recent momentum tilts toward the Penguins, who are 4-2-4 in their last ten. But Utah's season-long underlying numbers cut deeper: they own 53.0% Corsi and 52.0% expected goals share at 5v5, suggesting territory and shot volume they've earned. Pittsburgh counters with 51.0% xG and 50.0% Corsi that bends closer to even.
A. Silovs starts for Utah on a back-to-back—fatigue matters in net. K. Vejmelka does the same for Pittsburgh, which means both teams face the same fraying resource. Neither side gets a rest advantage here.
Dylan Guenther is Utah's offensive engine: 1.02 expected goals per 60 minutes, 28 goals, 15 assists, and a 16.8% shooting rate anchored by three high-danger chances. Daniil But generates 1.01 xG/60 but has fired only 2 goals on 2 assists with a 5.6% shooting rate—the talent is there, the finish is missing.
Pittsburgh's Filip Hallander is likely out, recalled from an AHL conditioning loan pending further evaluation. That removes a 1.28 xG/60 weapon. Rickard Rakell absorbs that loss with 1.01 xG/60, 13 goals, 13 assists, and 3 high-danger goals at 12.1% shooting. He's been consistent; Guenther just has more ceiling.
Both offenses struggle to convert high-danger chances into goals—neither team has scored on rebounds or high-danger opportunities at rates you'd expect from their shot quality. Utah's defensive weak spot is Maveric Lamoureux. Pittsburgh's is Caleb Jones. In a game this tight, those weak links will see pressure.
The PDO edge to Pittsburgh (1.012 vs 1.003) hints at small-sample noise playing into their numbers rather than deep talent, but it's noise that's won them games lately.
This isn't a blowout waiting to happen. Utah should control pace and territory. Pittsburgh should survive on the back of recent execution. The Penguins' narrow win probability reflects that—they're favored, but not by much. Take the points and expect a grind.
