The Blue Jackets beat the Panthers 4–2, but the underlying metrics flipped the script: Florida generated 4.52 expected goals to Columbus's 4.14, a 52–48 edge that usually predicts a Panthers win. The goaltending decided it.

J. Greaves was the difference. He saved 2.52 goals above expected—the margin between a 2–2 tie and a 4–2 Blue Jackets win. His 0.929 save percentage on 28 shots turned what should've been a tight affair into a rout.

The game hinged on shot quality. Columbus generated 0.086 expected goals per shot; Florida managed 0.059. That's not precision—that's execution. The Blue Jackets were relentless on the rush, converting twice on 29 chances. The Panthers matched that conversion rate (2 on 33 attempts) but couldn't exploit their territorial advantage.

The real indictment: Florida landed 10 high-danger chances and buried zero of them. Columbus had six and buried at least two—the ones M. Olivier created from the slot in the third period. Florida then scored twice from low-danger spots, plays where you expect the opposing goalie to hold the line. Greaves didn't.

D. Tarasov played well for Florida—1.14 goals above expected—but faced shots of lower quality. He stopped what he should have. Greaves stopped what he shouldn't have.

Shift attempts split 28–28. I. Provorov orchestrated the win with 1 goal and 2 assists. A. Ekblad recorded 2 assists despite the Panthers' offensive drought. This was less about tactics than variance. Florida created the better game. Greaves simply played the better game.

The model gives Columbus a 54% win probability with Florida's expected goal margin between 2.5 and 2.7. That's as close to a pick'em as you'll find. The underlying metrics nearly explain it—Columbus owns a 52.0% expected goal share at 5v5 versus 49.0%—but recent form and goaltending concerns muddy the picture enough that the odds barely distinguish these teams.

Columbus has momentum with an 8-1-1 record over their last 10 games. Florida has faltered, posting a minus-11 goal differential in the same span. That suggests freefall, but the underlying patterns tell a subtler story. Columbus's shot volume edge (51.0% Corsi versus Florida's 52.0%) is razor-thin. The real gap lives in finishing: Columbus's PDO of 1.003 against Florida's 0.989 reveals that Columbus is converting at a sustainable rate while Florida is running cold on the other end.

The expected goal numbers confirm it. Columbus generates chances at a 52.0% rate. That xG% advantage translates directly into the model's slight edge. Florida's PDO slump will eventually correct, but not in one game.

Goaltending is volatile. J. Greaves and S. Bobrovsky are both starting on a back-to-back, raising durability questions. Whoever looks fresher in the second period could swing a close result.

For the matchups: Sam Bennett anchors Florida with 0.97 expected goals per 60 and 18 goals on a 13.1% shooting rate. Brad Marchand, at 0.93 xG/60, is shooting 19.5%—unsustainably high. Boone Jenner slightly edges Bennett at 0.98 xG/60 with 11 goals and 21 assists. Mathieu Olivier (0.89 xG/60, 12 goals) is the secondary threat.

This is a game where models and margins align. Columbus's xG advantage is real and recent performance reflects it. Florida's cold streak is real but driven partly by finishing luck. Take the projection: Columbus favored, but barely.