The Maple Leafs won 4-3 in a game the underlying metrics nearly called even. Toronto held just 47% of shot attempts—23 shots on goal to Buffalo's 26—yet overperformed its expected goals by 1.15. The scoreline exceeded what the underlying play suggested.

Expected goals were close: 2.85–2.65 in Toronto's favor, a 52–48 split. High-danger chances nearly matched at 4–3. But goaltending and efficiency separated the teams.

U. Luukkonen (BUF) underperformed by 1.2 goals, with Buffalo scoring two low-danger goals the opposing goalie should have stopped. Toronto's rush attack was lethal: four goals on 21 rush chances. Buffalo, on the same rush workload, scored twice. C. Primeau (TOR) saved 23 of 26.

John Tavares settled it at 01:28 of overtime with a wrist shot from the high slot.

The model projects Buffalo at 52% win probability in a 2.6–2.5 game. This is as close to a toss-up as hockey gets.

Buffalo owns the underlying play. The Sabres control 50.0% of 5v5 expected goals against 46.0% for Toronto and generate 49.0% Corsi against Toronto's 45.0%, establishing territorial dominance. But that edge comes with a catch: Buffalo's 1.020 PDO is running hot, and regression is inevitable. Toronto sits at 1.002, closer to the line.

The Maple Leafs are -6 in goal differential over their last 10 games, which tracks their season-long struggle to convert chances into wins. Yet the models still give them a puncher's chance, largely because the underlying metrics are closer than the recent results suggest.

Goaltending uncertainty adds another layer. A. Stolarz is on a back-to-back, as is A. Lyon. Either starter could be fresh or fatigued—or the teams could pivot to backups entirely. That's a variable neither team fully controls.

The skill gap sits with Buffalo's forwards. Jason Zucker leads at 1.07 expected goals per 60 minutes, backed by 14 goals and 14 assists with 6 high-danger goals at a 15.1% shooting percentage. Josh Doan runs 1.01 xG/60 with 15 goals and 18 assists but only 2 high-danger goals, his 11.5% shooting rate a sign that volume isn't converting at an elite clip. Both threaten, but neither represents the finisher Toronto will counter with.

Auston Matthews carries 1.10 xG/60, the game's highest rate, with 17 goals and 13 assists—though he's managed just 1 high-danger goal at 10.6% shooting, a hint that elite chances aren't finding the back of the net at the frequency his underlying play merits. John Tavares offers a secondary threat at 0.91 xG/60 with 15 goals and 24 assists, including 3 high-danger tallies at 12.4%.

The Sabres will rely on dominating the second chance—they're at 0% rebound-goal rate and 0% high-danger-goal rate, meaning they're either converting on the first opportunity or not at all. Jack Quinn represents their defensive weak link. Toronto mirrors that dependency: 0% HD-goal rate, 0% rebound-goal rate, with Bo Groulx as the defensive liability.

This matchup lives in the margins. Buffalo has the meta advantage; Toronto has the star power and resilience. The probability split reflects genuine uncertainty, not a model hedging its bets.