Florida held 51% of shot attempts and still lost. Montreal won 3–2 in OT not because of superior possession or even better chances—expected goals ran 4.14–4.39, nearly even—but because the Panthers couldn't convert high-quality opportunities while S. Montembeault was immaculate.

Florida went 0 for 6 on high-danger chances. Montreal scored three on ten. S. Montembeault stopped 2.1 goals above expected and posted 100% on high-danger shots. The Panthers underperformed their xG by 2.14 goals, twice as much as Montreal's 1.39.

Nick Suzuki ended it at 3:24 of overtime, finishing a game where Montreal's transition game was lethal: three goals on 33 rush attempts. Neither team capitalized on rebound chances (six created, zero converted each). The open nature of the game—16 high-danger chances combined—masked a stark shooting reality: Florida's two goals came from low-danger areas, the kind a goalie has to have. Montreal's quality chances found the net.

The models see this as a genuine coin flip: a projected 2.5–2.7 game with Florida at 53% to win. The teams are separated by marginal edges in the places that matter most.

Florida's one structural advantage sits in possession. The Panthers drive 52.0% Corsi at 5v5 versus Montreal's 49.0%, meaning territorial control tilts toward the visiting team. That possession advantage hasn't yet materialized into expected goals—Florida sits at 49.0% xG% to Montreal's 48.0%—but teams that own the puck eventually own the scoreboard.

The Canadiens' principal vulnerability is sustainability. Montreal runs a 1.018 PDO, overperforming the shooting rates their chances should generate. Florida, by contrast, sits at 0.989, underperforming. Regression cuts both ways—if Montreal regresses down and Florida regresses up, the margin widens.

The rosters explain why this stays close despite Florida's structural edge. Cole Caufield carries 33 goals and 0.97 xG per 60 for Montreal, but his 20.0% shooting rate is unsustainable. Oliver Kapanen (19 goals, 0.83 xG/60) provides secondary production, though his 17.8% shooting rate is equally inflated.

Florida counters with Sam Bennett at 0.97 xG/60—matching Caufield's chance creation—while scoring at a more realistic 13.1% clip. His 18 goals include 6 high-danger conversions, suggesting more consistent finish than Montreal's top scorer. Brad Marchand (0.93 xG/60, 17 goals, 19.5% shooting) hints at upside; he's been beaten by variance rather than riding it.

The structural difference cuts deeper when you examine shot quality. Caufield's 1 high-danger goal against 33 total signals most damage arrives from range rather than close. Kapanen's 4 high-danger goals in 19 shows more inside presence. Bennett's 6 in 18 tilts decisively toward high-probability chances—Florida is, genuinely, more dangerous when it shoots.

What this truly hangs on: Does Florida's possession advantage compound, or does Montreal's elite efficiency on a small sample prove durable? The model, refusing to pick a side at 53%, is saying both outcomes are equally probable.