The Penguins won 5–2, yet the Panthers held the advantage in expected goals, 4.07–3.57. This scoreline looked closer in the underlying data than the final tally.

Goaltending made the difference. A. Silovs stopped 29 of 31 shots (0.935 SV%, +2.07 above expected), while D. Tarasov surrendered 5 goals on 23 shots (0.783 SV%, −1.43 below expected). When one goalie needs to have those, and one exceeds his workload that badly, the outcome is set.

High-danger chances finished 7–7, so neither team dominated the dangerous areas. But Pittsburgh generated far more quality per shot (0.081 xG/shot vs. Florida's 0.057), and it showed in rush play: the Penguins scored 4 goals on 26 rush chances while the Panthers managed 2 on 38. Pittsburgh made their volume matter.

S. Crosby led with 1 goal and 2 assists in 17:34. B. Rust added 1 goal and 1 assist in 18:50. R. Rakell scored twice in 18:16, anchoring Pittsburgh's finish.

The Panthers wasted their chance at the workload advantage. Florida couldn't convert on high-danger chances (0/7) and cashed in twice from low-danger areas, the kind of variance that seldom repeats.

The model gives Pittsburgh 54% win probability in a projected 2.5–2.7 goal game. That is as close to a toss-up as preseason mathematics allows.

Florida has been a mess lately, outscored by 8 goals over their last ten games. But the season-long 5v5 numbers tell a closer story. At even strength, the Panthers generate 49.0% expected goals against Pittsburgh's 51.0%—negligible separation. Pittsburgh's 50.0% shot share looks sharper on paper, but Florida's 52.0% Corsi means they're still controlling possession. The real signal: Pittsburgh's 1.012 PDO says the Penguins have caught all the lucky bounces, while Florida's 0.989 PDO suggests they've been snakebitten.

Both starter situations are clouded. A. Silovs and D. Tarasov are each on a back-to-back, inviting either tired performances or unexpected bench appearances.

Sam Bennett is Florida's motor: 18 goals, 19 assists, 0.97 expected goals per 60 minutes, six high-danger chances, and 13.1% shooting. Brad Marchand carries 0.93 xG/60 and 19.5% shooting, but hasn't played in 30 days—likely unavailable.

For Pittsburgh, Rickard Rakell is the primary weapon: 1.01 xG/60, 13 goals, 13 assists, three high-danger chances, 12.1% shooting. Filip Hallander's 1.28 expected goals per 60 is elite-level, but he's been recalled from conditioning and is undergoing evaluation, making him unavailable.

Both teams are oddities on the scoresheet: neither has recorded a high-danger goal or rebound goal in the tracked sample, suggesting either sample noise or a tactical wrinkle worth monitoring. Florida's defensive weak link is Marchand (when available); Pittsburgh's is Caleb Jones.

Without clarity on the injury front, this truly is a coin flip.