The Bruins won 4–3 in overtime despite losing the expected-goals battle 2.95–6.02. Buffalo controlled two-thirds of the shot quality. This was a win written by goaltending, not play.

J. Korpisalo (BOS) saved 3.0 goals above expected on 37 of 40 shots (0.925 SV%). That's elite performance under pressure. A. Lyon (BUF) conceded 1.05 more than expected on just 18 saves (0.818 SV%), a mismatch between workload and execution. Buffalo generated 10 high-danger chances to Boston's 2 — a 5-to-1 advantage — yet finished a goal behind. The per-shot quality gap was stark: Buffalo's xG per shot (0.087) nearly doubled Boston's (0.050). Puck luck reversed the outcome.

Boston found their edge on the rush, converting 3 of 27 chances while Buffalo scored 2 of 30. Thin margins, but Korpisalo's save rate was the difference.

Marat Khusnutdinov's snap shot from the high slot at 02:07 of overtime decided it.

Three Stars

• R. Dahlin (BUF) — 1G, 1A in 25:49
• D. Pastrnak (BOS) — 1G, 1A in 20:17
• A. Tuch (BUF) — 1G, 1A in 19:19

The model projects a 2.6–2.6 game with the Sabres at 50% win probability. This is a true coin flip.

Buffalo arrives with better form: 4-4-2 over their last 10 games. Boston has imploded in the same window, posting an -8 goal differential. But the underlying play is similarly tight. The Sabres own the 5v5 battle at 50.0% expected goals percentage versus Boston's 47.0%, but both teams sit at 49.0% Corsi%, meaning shot volume is equal. That xG edge is the only genuine separation.

Both teams are running hot. Boston's PDO of 1.023 and Buffalo's 1.020 indicate both are outperforming their underlying metrics; regression is coming for both, marginally sooner for Boston.

Players to Watch

Buffalo

Jason Zucker: 1.07 xG/60, 14G–14A, 6 high-danger goals, 15.1% shooting

Josh Doan: 1.01 xG/60, 15G–18A, 2 high-danger goals, 11.5% shooting

Boston

Viktor Arvidsson: 1.04 xG/60, 18G–24A, 2 high-danger goals, 13.0% shooting

Tanner Jeannot: 0.75 xG/60, 6G–14A, 2 high-danger goals, 8.3% shooting

Zucker and Doan operate at nearly identical shot-quality rates; Arvidsson closes gaps through volume and sits at 18 goals despite just two high-danger conversions. Jeannot occupies a smaller role. Buffalo's xG% edge suggests marginal offensive efficiency, but not margin of victory. Both teams' PDO means regression is coming; the statistical case for Buffalo is modest. The 50% projection reflects what it actually is: no clear winner. Treat this as permission to ignore consensus and watch the play.