Ottawa's 7-2 win over Boston masks an expected-goals game that favored the Senators only marginally. The xG split ran 3.15–2.84 — close enough that the Bruins' scoreboard deficit tells a story the underlying numbers don't support. Ottawa outperformed their model by 3.85 goals; Boston undershot theirs by 0.84. This is the scoreboard disagreeing sharply with process.

The Senators controlled shot quality, not volume. They held 45% of shot attempts (23–28 on goal) and generated 0.062 expected goals per shot versus Boston's 0.045 — a meaningful edge in shot danger. High-danger chances favored Ottawa 6–3. That concentration mattered: the Senators finished 50% on HD chances (3 of 6), while Boston converted nothing on their rebound opportunities.

The goaltending split was the game's largest variable. L. Meriläinen (OTT) posted a 0.929 save percentage — 0.84 goals better than expected. J. Swayman (BOS) recorded 16 saves on 23 attempts for a 0.696 mark, performing 3.85 goals worse than his workload suggested.

Ottawa's power play was the other separator: 4 goals on 8 shots, a 0.50 expected-goals rate that no penalty kill can withstand. In transition, Ottawa scored 6 rush-goal opportunities out of 25 attempts; Boston managed 2 from 26 chances — identical volume, opposite conversion.

Viktor Arvidsson's slap shot from the top of the circles at 19:50 of the third period proved to be the turning point.

The model projects a 2.3–2.8 goal game with the Senators at 59% win probability. That edge sits squarely on Ottawa's underlying play advantage.

The Senators own the 5v5 shot quality battle, carrying 56.0% expected goals to Boston's 47.0%. They press their territorial advantage with 53.0% Corsi, meaning they'll spend the night in Boston's zone. The Bruins countered with only 49.0% Corsi and 47.0% xG—a recipe for reactive hockey.

Boston's win expectancy leans on variance working their way. The Bruins are running a 1.023 PDO, which means their actual results are outpacing their underlying metrics. Regression lurks. Ottawa sits at 0.993 PDO, holding closer to luck-neutral, making them the model's cleaner read.

Three players set the tone for each side:

• Viktor Arvidsson (BOS) generates 1.04 xG/60 with 18G 24A and a 13.0% shooting rate. He's in rhythm.
• Brady Tkachuk (OTT) is Ottawa's engine at 1.19 xG/60, anchoring 11G 20A with a 7.5% rate. He's the leverage play in this matchup.
• Tanner Jeannot (BOS) carries 0.75 xG/60 with 6G 14A and an 8.3% rate—depth scoring on volatility.
• Dylan Cozens (OTT) chips in at 1.00 xG/60 with 13G 15A and a 9.5% rate, offering secondary punch.
• Arvidsson has 2 high-danger goals; Tkachuk mirrors that with 2.
• Jeannot and Cozens each sit at 2 and 1 high-danger goals respectively.

The tactical imbalance surfaces in Boston's offensive construction. The Bruins generate 0% of goals from high-danger looks and 0% from rebounds—a sign of one-and-done chances rather than sustained pressure. That efficiency gap (Arvidsson's 13.0% rate v. Tkachuk's 7.5%) matters when puck luck regresses; Boston will need conversion to hold water.

Defensively, Boston's weak link is Elias Lindholm. Ottawa's is Nick Jensen. Both teams register 0% high-danger and 0% rebound rates on the defensive side—meaning their leaks come from perimeter sequences, not scrambles.

The model's 59% lean makes sense: Ottawa owns play, holds luck closer to neutral, and edges Boston's depth scoring with Tkachuk's center gravity. Regression hits Boston's PDO; the Senators stay tethered to their underlying advantage. That's a durable edge in a 2.3–2.8 goal game.