Calgary's 5-4 victory was a heist: New Jersey dominated with 58% expected goals (4.95–3.59) but couldn't finish. The Devils created the better chances and controlled play, yet the Flames escaped with the win.

The scoreline inverts the underlying quality entirely. New Jersey's power play looked lethal—two goals on 26 shots, 2.74 expected goals—but the even-strength game belonged to Calgary. The Flames squeezed better looks from worse volume: 0.084 expected goals per shot versus New Jersey's 0.063. Both teams scored three on the rush, but Calgary's came from 19 chances; New Jersey needed 35.

Goaltending split the difference. D. Wolf stopped 28 of 32, overperforming his model by 0.95 goals. J. Markstrom faced 29 shots and made 24, underperforming by 1.41—a two-goal variance that explains the final score. In a game of 15 high-danger chances combined, Calgary found the margin in shooting efficiency, not opportunity.

The turning point came on Luke Hughes's wrist shot from the high slot late in the first (P3, 19:56). But the real story was Calgary's discipline in a wide-open game: they weathered New Jersey's possession advantage and converted their chances at a higher rate.

The model projects a 2.4–2.7 game with New Jersey at 54% win probability. This is about as close to even money as you'll find.

New Jersey owns the underlying play. The Devils' 50.0% 5v5 expected goals share dwarfs Calgary's 47.0%. Their Corsi numbers (51.0% vs 49.0%) reinforce it—they generate possession advantages and direct more shot attempts toward Calgary's net. This isn't margin; it's a systematic gap in control.

Both teams are running cold at the PDO line. New Jersey sits at 0.971; Calgary at 0.985. When a team's actual goal rate diverges from its expected rate this sharply, regression is the default forecast. Each team has unsustained luck working against it, which means each has something to gain if the variance corrects.

The star matchup belongs to New Jersey's Timo Meier, who generates 1.08 expected goals per 60 minutes at 5v5—elite volume—and has converted it into 13 goals and 15 assists. He's potted 3 high-danger chances, a marker of finishing quality. Calgary's Blake Coleman is the Flames' nearest equivalent, generating 0.89 xG/60 and posting 16 goals with 12 assists. Coleman's shot rate sits at 11.4%, slightly elevated, suggesting some regression risk.

Matt Coronato represents Calgary's secondary firepower. Though his 0.80 xG/60 trails Coleman's, Coronato has 10 goals and 16 assists on a 6.6% shot rate—a number that screams underperformance waiting to correct. Jack Hughes anchors New Jersey's depth, pairing 0.88 xG/60 with 18 goals and 17 assists. He's accrued just 1 high-danger goal, meaning much of his production flows from high-volume play and efficiency outside the home plate.

Tactically, both teams generate a startling void: neither has scored a single high-danger or rebound goal this season. That's not an edge—it's a symmetry. It suggests both are missing some portion of their opportunity cleanup. Calgary's defensive perimeter carries risk with Coronato in the mix; New Jersey's with Johnathan Kovacevic.

The verdict stands. New Jersey's process dominance tips the needle to 54% win probability, but Calgary's underperforming PDO and Coronato's shooting regression create enough tail risk to call this a genuine toss. Bet the model, but hedge the outcome.