C. Hellebuyck played 1.22 goals above expectation, yet Winnipeg lost in overtime. The Jets' goalie made 24 of 28 shots (0.857 SV%), keeping his team within reach against Dallas's overwhelming underlying advantage. The Stars controlled 63 percent of expected goals at 5v5 (5.22 to 3.02) and prevailed 4-3 after extra time.

The gulf in shot quality was stark. Dallas averaged 0.089 expected goals per attempt against Winnipeg's 0.054. The Stars generated 10 high-danger chances to the Jets' 5 and were clinical on the rush—3 goals on 29 chances. Both teams crashed the net identically: 3 rebound opportunities each, 1 converted per side. But the underlying advantage lay entirely with Dallas.

Winnipeg had a single advantage: a dominant power play. The Jets scored twice on 14 shots (0.84 xG). At even strength, they were outmatched.

Thomas Harley's backhander from the doorstep at 02:05 of overtime ended it. J. Oettinger (24/27, 0.889 SV%) managed the game in Dallas's crease, but C. Hellebuyck's performance made this competitive.

The model projects a 2.3-to-2.6 goal spread in Dallas's favor, with the Stars holding 55% win probability. This is as close to a pick'em as you'll find.

Dallas owns the underlying play. Their 51.0% 5v5 expected goals percentage dwarfs Winnipeg's 48.0%, suggesting the Stars should sustain pressure and generate better chances. But here's the catch: Dallas's PDO sits at 1.024, which means they're running hot on conversion luck. Expect some regression there.

Winnipeg's Corsi dominance—49.0% to Dallas's 47.0%—suggests the Jets control play volume at even strength. Yet xG favors the Stars, indicating Dallas capitalizes on what they do generate. That's the tension. Winnipeg gets more shot attempts but Dallas gets the higher-quality ones.

C. DeSmith's back-to-back situation adds opacity. Watch whether Dallas rolls with him or dips into reserves. A tired starter or a fresh backup could swing momentum either way.

The Players to Watch

Nino Niederreiter leads Winnipeg's charge. He's at 0.87 xG per 60 minutes with 6 goals and 6 assists, including 3 high-danger chances. His 8.6% shooting rate is sustainable, meaning he's not simply running hot—he's converting opportunities at a reasonable clip. Kyle Connor's 13.0% shooting is the real outlier; his 25 goals and 30 assists come on just 0.83 xG/60, suggesting he's beating the model on finish quality.

Jason Robertson leads Dallas offensively. At 0.96 xG/60, he's generating more scoring chances than anyone else on the ice. His 25 goals and 21 assists come with only 2 high-danger markers, implying his xG overestimates his actual scoring. Michael Bunting provides secondary scoring at 0.83 xG/60, though his 8 goals and 14 assists reflect a lower shooting rate (7.6%) that could inflate if his luck normalizes.

The Decision

Dallas has the analytics edge—better five-on-five shot quality and possession efficiency by expected goals. Winnipeg has volume. Dallas's PDO regression risk is real, but the Stars have earned their 55% win probability. With DeSmith in question, Dallas's depth in net becomes the deciding variable. If they go with their starter, edge Dallas. If they rotate, Winnipeg gets closer to fair value.

This is genuinely uncertain. Neither team has overwhelmed the other's fundamental case.