Buffalo won 3–2 in overtime against all evidence. The Senators controlled 67% of expected goals (3.84–1.88) and 51% of shot attempts (26–25), generating five high-danger chances to Buffalo's one. This was a heist.

The Sabres generated higher-quality looks than their shot volume suggested, but the Senators' dominance was complete. Buffalo's victory turned on one moment: Bowen Byram's snap shot from the high slot at 00:31 of overtime.

A. Lyon was the reason this worked. The Buffalo goaltender made 24 saves on 26 shots (0.923 SV%) with +1.84 goals above expected, stopping all five high-danger chances he faced. L. Ullmark finished –1.12 on expected: 22 saves on 25 shots (0.880 SV%) despite facing less work. The Senators had quality—0.061 xG per shot versus Buffalo's 0.046—but couldn't convert it. Ottawa scored twice off 32 rush chances; Buffalo converted three times in 23 attempts, with one of those conversions deciding the game in overtime.

The difference wasn't intensity or territory. It was execution against profligacy.

The model projects a BUF 2.3 - 2.7 OTT game with the Senators at 57% win probability.

Ottawa's edge runs through the underlying numbers. The Senators own 56.0% of 5v5 expected goals — a six-point gap over Buffalo's 50.0%. That advantage extends to Corsi, where Ottawa drives 53.0% of shot attempts, establishing territorial control throughout the game.

Buffalo's risk is regression. The Sabres are running at 1.020 PDO — unsustainably hot — while Ottawa sits at 0.993, suggesting room for variance to cut the other way. The closer the Sabres' luck moves toward average, the wider Ottawa's underlying advantage becomes.

Goaltending adds complexity. Both L. Ullmark and U. Luukkonen are on back-to-backs, creating uncertainty about who actually takes the ice or whether fatigue becomes a factor for either starter.

The offensive threats shape the tactical approach. Jason Zucker carries 1.07 expected goals per 60 minutes for Buffalo, backed by 14 goals and 14 assists; his 15.1% shooting rate has buoyed six high-danger chances into the scoresheet. Josh Doan offers another dimension at 1.01 xG/60 with 15 goals and 18 assists, though his 11.5% shooting is more sustainable.

Ottawa counters with Brady Tkachuk, who leads at 1.19 xG/60 — the highest individual volume in this matchup — while managing just 11 goals and 20 assists. His 7.5% shooting rate signals he's generating chances the scoresheet hasn't yet validated. Dylan Cozens operates at 1.00 xG/60 with 13 goals and 15 assists, converting at 9.5%.

Buffalo's defensive weak link is Jack Quinn. Ottawa's is Nick Jensen. Neither team generates meaningful offense from high-danger situations — both at 0% HD goal rate, 0% rebound goal rate — which actually favors the team defending better underlying chances.

This is a game where the numbers point decisively in one direction, but regression and fatigue introduce enough noise to respect Buffalo's floor. The Senators are the lean.