Winnipeg's 5-2 scoreline misleads. The Jets held 55% of 5-on-5 expected goals (3.17-2.59), insufficient to justify a three-goal margin without scoring luck. E. Merzlikins explained the gap: a 0.833 save percentage on 30 shots, 1.83 goals below expectation.

Winnipeg was clinical off the rush, converting 5 times on 29 chances. Columbus managed 2 goals on 22 rush attempts, an efficient rate that might have tightened the game if Merzlikins had held standard.

Logan Stanley's slap shot from the top of the circles (06:23, P3) provided the turning point in the third period, cementing a lead Winnipeg controlled to the end.

The model projects a 2.6–2.5 game favoring Columbus, assigning the Blue Jackets 52% win probability. This is as close to a pick'em as you'll find.

Columbus owns the underlying play. At 5v5, they sit at 52.0% xG% with 51.0% Corsi%, against Winnipeg's 48.0% xG% and 49.0% Corsi%. Both teams carry PDO near even (1.003 for Columbus, 0.998 for Winnipeg). The shot-quality edge belongs to the Jackets, but barely.

Boone Jenner is the driver for Columbus. He's running 0.98 xG/60 with 11G 21A at a 9.6% shooting rate. The catch: only 2 of those goals come from high-danger areas, which suggests his finishing is efficient on cleaner chances. Mathieu Olivier is the variance play—0.89 xG/60, 12G 9A at 13.3% shooting with just 1 HD goal. That shooting percentage is unsustainable and masks thinner underlying production.

Winnipeg's offense splits between volume and finishing. Kyle Connor leads with 25G 30A at 0.83 xG/60 and 13.0% shooting (2 HD goals). Nino Niederreiter operates at lower volume—0.87 xG/60, 6G 6A—but has clustered production in high-danger areas with 3 HD goals. If Niederreiter's net-front efficiency holds, the Jets can keep this close.

The structural constraint: Columbus has generated 0% of its goals from high-danger and rebound situations, suggesting an offensive profile built on clean looks. Conor Garland is their defensive weak link. Winnipeg faces the same profile—0% HD goal rate, 0% rebound goal rate—with Colin Miller as the exposed defender.

This is what a genuine coin flip looks like. Columbus should win on process, and the model reflects that with a 52% edge. But Olivier's shooting rate will regress, and Niederreiter's net-front production could sustain the gap. Two equally plausible outcomes, both compatible with the underlying data. The model's call is probably right.