Columbus won 4-3 in a shootout despite trailing badly in expected goals: Montreal 3.89, Columbus 2.79. The Canadiens controlled 58% of xG at 5v5, and Columbus managed only 42% of shot attempts (23 to 32). This was a heist.

Columbus overperformed their xG by 1.21 goals while Montreal underperformed by 0.89. J. Greaves stopped 29 of 32 shots (0.906 SV%) for Columbus; J. Dobes made 20 saves on 23 shots (0.870 SV%) for Montreal. Greaves was the difference—nearly a full goal better than expected.

High-danger chances split evenly at 7-7, but both teams found their scoring elsewhere: Columbus went 3-for-20 on the rush, while Montreal did the same, 3-for-27. That symmetry masks Montreal's structural advantage—they generated 6 rebound opportunities and converted zero, suggesting they crashed the net but couldn't finish the cleanup.

Kirill Marchenko's wrist shot from the slot decided it in the shootout, completing a game where Columbus's finishing—not their play—carried the result.

Columbus enters this matchup as a lean favorite. The model projects a 2.4-to-2.8-goal advantage for the Blue Jackets, who carry 56% win probability—a modest but meaningful edge.

That edge rests on superior underlying play. Columbus controls 52.0% of expected goals at 5v5, outpacing Montreal's 48.0%. The gap widens slightly in raw shot attempts, where the Blue Jackets claim 51.0% of Corsis to Montreal's 49.0%. These aren't dominant advantages, but they compound: teams that sustain shot advantages usually win marginal contests.

The divergence that matters lies in regression. Montreal's PDO stands at 1.018, a figure that reflects unsustainable offensive luck. Columbus sits at 1.003, closer to the long-run mean. When Montreal's PDO normalizes—and it will—the Canadiens' actual win rate will fall below the underlying metrics.

Cole Caufield fuels Montreal's attack. The 33-goal winger generates 0.97 expected goals per 60 minutes and converts at a 20.0% rate; that efficiency is borrowed time. Oliver Kapanen reinforces it, with 19 goals off 0.83 xG/60 and 17.8% shooting. Both are running hot.

Boone Jenner anchors Columbus at center, where 11 goals and 21 assists on 0.98 xG/60 translate to elite playmaking at 5v5. Mathieu Olivier (12 goals, 0.89 xG/60) provides secondary scoring depth.

Defensively, Mike Matheson represents the liability Montreal must shore up. Conor Garland's weaknesses invite Montreal to attack. The tactical clarity is sharp: if Montreal can neutralize Jenner and starve Columbus's 5v5 volume generation, regression plays out painlessly. If Columbus wins the play-balance battle as these numbers suggest, Montreal's luck simply runs out. Regression is mechanical, not punitive. The Canadiens were never this good. The Blue Jackets just have to play their game and wait.