The Blue Jackets built a superior shot profile and lost anyway. Columbus posted 4.01 expected goals to Florida's 3.49, held equal high-danger chances at 7 apiece, yet coughed up the game to S. Bobrovsky's wall.

The goaltending split told the story. S. Bobrovsky (FLA) stopped 30 of 31 shots for a 0.968 save percentage, exceeding his model by 3.01 goals—the sort of performance that flips close games. E. Merzlikins (CBJ) posted 19 of 21 for 0.905, which would be respectable had the margin not swung on Bobrovsky's dominance. Against high-danger chances, Bobrovsky was perfect: 7 for 7, while Merzlikins faced identical volume yet surrendered two.

Columbus wasted its offensive advantage through concrete numbers. The Blue Jackets underperformed their xG by 3.0, failing to score on 7 high-danger opportunities—a conversion rate that won't recur. Florida's power play, conversely, punished efficiency: 2 goals on 10 shots (1.28 xG), while Columbus generated zero goals from 4 rebound chances.

The Panthers controlled just 40% of shot attempts (21–31), yet Florida's execution overwhelmed Columbus's volume advantage. S. Reinhart's wrist shot from a sharp angle at 02:20 of OT sealed the outcome, capping a 1G–1A performance in 17:52.

The Three Stars:

• #13 S. Reinhart (FLA) — 1G, 1A in 17:52
• #17 M. Marchment (CBJ) — 1A in 20:19
• #19 A. Fantilli (CBJ) — 1G in 19:31

The model sees this as pick'em territory: a projected 2.6–2.6 game with Columbus at exactly 50% win probability. The analytics are split accordingly.

Columbus owns the underlying play. At 5v5, the Blue Jackets generate 52.0% expected goals to Florida's 49.0%, a meaningful edge in shot quality. But the Panthers counter with 52.0% Corsi share, suggesting they control possession. That split—stronger shots for CBJ, more volume for FLA—explains why this feels like a genuine coin flip rather than a clear favorite.

Recent form tips toward Columbus. The Blue Jackets have gone 6-1-3 in their last ten, compiling points at a pace that suggests sustained play rather than variance. Florida sits at -7 goal differential over the same span—submarine territory. One team is trending up; the other is treading water. Yet the models still refuse to break for either side, which means Florida's underlying structure—that 52.0% Corsi, that 0.989 PDO—leaves room for mean reversion.

Where the goals come from will matter more than possession counts. Boone Jenner generates 0.98 expected goals per 60 at 5v5 with 11 goals and 21 assists on the season; his 9.6% shooting suggests room for regression, but his 2 high-danger goals indicate he's getting dangerous looks. Mathieu Olivier operates lower at 0.89 xG/60 but shoots 13.3%, well above league average, supporting 12 goals in a bottom-six role.

Florida counters with Sam Bennett at 0.97 xG/60, nearly Jenner's rate, and Bennett has 18 goals to show for it with 6 high-danger tallies. Brad Marchand rounds out the watch list at 0.93 xG/60 and a blistering 19.5% shooting mark, though that efficiency can't sustain. The Panthers have talented shot-makers. They've just stopped finishing at the rate their underlying metrics suggest they should.

Both teams show 0% rates on high-danger and rebound goals—either a data anomaly or a sign neither has figured out the dirty-area game. Columbus flags Conor Garland as a defensive liability; Florida's weakness circles back to Marchand, whose offensive production masks a vulnerability in transition defense.

The Corsi split, the xG split, the goal differential—none of it screams dominance. Columbus has the slight edge in shot quality and recent form. Florida has the slight edge in possession and, with PDO at 0.989, has room to regress favorably. That's why the model calls it 50-50.