The Oilers ran the numbers and lost anyway. Minnesota beat Edmonton 7–3 despite trailing in expected goals 3.77 to 5.22—a 58–42 split in Edmonton's favor at 5v5. The Oilers took 42 shot attempts to Minnesota's 29 and generated nine high-danger chances to three. The Wild simply finished theirs.

J. Wallstedt was the difference. The Minnesota goaltender stopped 39 of 42 shots (0.929 save percentage), running 2.22 goals better than expectation. Edmonton's goalies collapsed: T. Jarry recorded 15 saves on 20 shots (0.750 save percentage, -2.40 vs expected), while C. Ingram saved 7 of 9 (-0.83). Minnesota overperformed their expected goals by 3.23; Edmonton underperformed theirs by 2.22. When goaltending variance stretches like this, the numbers stop predicting wins.

Brock Faber's wrist shot from the high slot at 14:17 of the third period marked the turning point, but the structural damage ran deeper. Minnesota scored twice from low-danger spots Oilers goalies should have stopped. Edmonton was efficient off the rush—three goals on 39 chances—yet Minnesota was more lethal: five goals on 28. The Oilers crashed the net harder, generating 11 rebound opportunities and converting zero. Minnesota created eight and buried one.

The Three Stars reflected Minnesota's dominance:

• B. Faber (MIN)—1G, 1A in 25:50
• Q. Hughes (MIN)—1G, 1A in 25:14
• J. Eriksson Ek (MIN)—1G, 1A in 17:01

Edmonton controlled possession. Minnesota controlled the game.

The model projects a Minnesota 2.5 to 2.6-goal game, with Edmonton at 52 percent win probability. This is about as close to a coin flip as the numbers get.

Expected goals tell you why. Both teams sit at 52.0 percent xG share at five-on-five, meaning neither outfit has built a meaningful shot-quality advantage. Edmonton holds a slight edge in Corsis (50.0 percent vs. 48.0 percent), but that's not enough margin to matter when underlying quality is identical. What breaks the tie is luck.

Edmonton's PDO sits at 0.982—running cold. The Oilers are due for positive regression on the score-effects and goaltending fronts, which narrows a model line that looks tight already. Minnesota's PDO is 1.007, suggesting they're banking on continuing to outshoot their expected outcomes.

C. Ingram and F. Gustavsson both face back-to-back scheduling. Either could feature a backup start or arrive fatigued. That's a real variable when the underlying talent is this balanced.

Minnesota's attack hinges on Matt Boldy, who carries 0.89 expected goals per 60 minutes and has converted at a 13.7 percent rate (22 goals, 18 assists, 5 high-danger goals). Robby Fabbri sits at a similar 0.88 xG/60 but has managed just 2 goals and 3 assists (1 high-danger) on a 6.9 percent shooting percentage. That gap between talent and results is noise waiting to normalize.

Edmonton counters with Zach Hyman at 1.23 xG/60, a level of shot quality that ranks elite. Hyman carries 14 goals and 13 assists but only 1 high-danger goal, suggesting he's beaten his xG on overall execution. Connor McDavid remains the constant: 0.92 xG/60, 25 goals, 33 assists, 5 high-danger goals, 13.7 percent shooting. His direct impact is obvious.

Both teams sit at 0 percent in high-danger and rebound-goal rates. Minnesota's defensive weak link is Matt Kiersted; Edmonton's is Jake Walman. Corsi and xG alignment means this truly is a toss-up. Model probability—52 percent Edmonton—is barely an edge. Take that as a warning: whichever team executes cleaner over 60 minutes wins. That's not a prediction. That's the data saying it genuinely doesn't know.