Washington's 2-1 win defied the possession count. Buffalo held 58% of shot attempts (30–22) and led 4.09–3.89 in expected goals, yet scored only once. The Sabres underperformed their xG by 3.09 goals and went 0 for 6 on high-danger chances—a collapse that C. Lindgren's brilliance masked.

Lindgren was flawless. He stopped 29 of 30 shots (0.967 SV%), saving 3.09 goals above expected. On high-danger chances, he was perfect: 6 for 6. U. Luukkonen posted a 0.909 SV% and gave back 1.89 goals above expected on 20 saves. The difference was the chasm between high-danger goal conversion.

Washington also underperformed their xG by 1.89. Yet J. Chychrun's snap shot from the high slot in the third period proved decisive—the moment that separated two teams with similar inefficiency. Both teams couldn't convert on high-danger chances (Washington went 0 for 9; Buffalo, 0 for 6), but Washington's goaltending turned margin-of-error hockey into a result.

Buffalo dominated statistically and still lost because they couldn't finish. Lindgren stole the win.

This game defies conviction. The model projects a 2.6–2.7 goal range with Buffalo at 52% win probability—essentially a pick'em. When the analytics say flip a coin, trust the coin.

Buffalo arrives hotter by points pace: 8-1-1 over their last 10. Washington brings recency momentum too, but the underlying five-on-five numbers reveal why the model shrugs. Both teams sit at 50.0% expected goal share and 49.0% Corsi%. They are, at the level that predicts wins, identical.

Both are also running hot in PDO. Buffalo at 1.020, Washington at 1.018—luck that historically regresses. The question isn't whether both cool down, but whether it happens tonight or next week.

Goaltending introduces the first lever: L. Thompson faces a back-to-back. Watch for a backup start or a starter operating on fumes.

Where Buffalo's edge emerges is in elite scoring talent. Jason Zucker (1.07 xG/60, 14G 14A) generates chances at a higher clip than Washington's leaders. His six high-danger goals anchor real production. Josh Doan (1.01 xG/60, 15G 18A) represents secondary elite scoring; together, they generate looks Washington's depth can't replicate.

Alex Ovechkin (0.97 xG/60, 19G 17A) remains Washington's engine but runs slightly behind Zucker's per-60 rate. Anthony Beauvillier (0.89 xG/60, 14G 9A) carries 10.4% shooting, the kind of outlier that corrects—his 14 goals likely peak. That asymmetry matters.

Defensively, the weak points are mapped: Declan Chisholm for Washington, Jack Quinn for Buffalo. Buffalo's advantage in scoring depth suggests better armor around their liability; Washington's Beauvillier uptick masks a thinner middle-six in absolute terms.

The reason this reads as a true coin flip isn't absence of edges—it's balance. Expected goals mirror. Possession metrics mirror. The skill difference in scoring is real but modest. Regression risk hits both teams equally. When models refuse to separate teams this cleanly, they're usually right to refuse.