The Bruins beat the Maple Leafs 5-3, but the underlying numbers said the game was closer to even — and favored Toronto. Expected goals split 4.65–4.19 in Toronto's favor.

The Maple Leafs controlled play: 46% of shot attempts, 33–39 on goal, and 8–6 on high-danger chances. They should have won. Instead, Boston's goaltending held while Toronto's did not.

The variance was enormous. The Maple Leafs underperformed their expected goals by 1.65 goals. The Bruins overperformed by 0.81. That swing is the entire game.

J. Swayman (BOS) stopped 1.65 goals above expected, perfect on eight high-danger shots. The Maple Leafs went 0 for 8 on their highest-probability chances. A. Stolarz (TOR) gave up 1.96 goals more than expected, well below his workload — a sign of a collapsed defense. D. Hildeby (TOR) looked solid in relief, but the damage was done.

The Bruins' rush attack was lethal: four goals on 34 chances. The Maple Leafs got three on 31 chances — similar rates, different results. Toronto crashed the net effectively, creating seven rebound opportunities and converting zero. That pattern repeated all night.

Fraser Minten's wrist shot from the slot in the third period was the turning point. One goal separated the teams at that moment, and Minten sealed it.

The model projects a 2.5-2.6 goal game with the Maple Leafs at 51% win probability. This is a pick'em.

Boston dominates possession at 49.0% Corsi, which correlates to 47.0% expected goals. Territory and shot creation both favor the Bruins. Toronto musters 46.0% xG% while conceding 45.0% Corsi—a structural disadvantage that forces them to win from smaller windows. Boston's path is volume; Toronto's is efficiency.

The Bruins' underlying metrics warrant caution. A 1.023 PDO means they're outscoring their play; regression is typically next. Toronto's 1.002 sits at equilibrium, which is what earned play looks like.

J. Korpisalo enters on a back-to-back, a variable that shapes the goaltending dimension.

Viktor Arvidsson is Boston's primary offensive vehicle. At 1.04 expected goals per 60 minutes, he's posting 18 goals and 24 assists with 2 high-danger chances and a 13.0% shooting rate—proof of both volume and precision. Tanner Jeannot operates at 0.75 xG/60, lower in creation but sharper in conversion: 6 goals on an 8.3% shooting clip, a sign he's found his spot in the lineup.

Auston Matthews leads Toronto's offense at 1.10 xG/60, the game's highest. He's compiled 17 goals and 13 assists with 1 high-danger chance; the assists suggest his line is functioning. John Tavares anchors at 0.91 xG/60 with 15 goals, 24 assists, 3 high-danger chances, and a 12.4% shooting rate—the reliable secondary option.

Elias Lindholm represents Boston's defensive weakness. Bo Groulx is Toronto's. Both become tactical targets.

The core tension is this: Boston owns the territorial game while Toronto thrives on efficiency. Neither has cracked the code to dominate both. The Leafs can steal this by staying patient and forcing Boston into low-probability attempts. Boston can win by suffocating Toronto into rushed decisions. Both paths hold equal weight in the model's projection, which explains the coin flip.