Pittsburgh's expected-goals advantage was massive—63% to 37%—and it lost anyway. The Penguins held 56% of shot attempts and 10 high-danger chances to Columbus's 7. They still finished 4-3 in a shootout. When control is that lopsided and a team loses, the goaltender won.

E. Merzlikins (CBJ) made 30 saves and overperformed his expected goals by 2.53. He stopped 90% of high-danger chances. A. Silovs (PIT) made 23 saves at an 0.885 save percentage, underperforming his model by 2.5 goals.

The Penguins wasted their possession. They created 7 rebound opportunities and converted none. Pittsburgh was lethal on the rush—3 goals on 38 chances—but not lethal enough. Columbus answered with 2 goals on 24 rush attempts, and that discipline cost Pittsburgh the game.

The shootout hinge was Charlie Coyle. S. Crosby (PIT) had already earned 1 goal and 1 assist in 21:57. Z. Werenski (CBJ) collected 1 assist in 29:18, setting up what mattered most. K. Letang (PIT) added 1 assist in 25:55 but couldn't push Pittsburgh over the line.

This was not a lucky win. It was a heist.

The model projects a CBJ 2.6–2.7 game with the Penguins at 51% win probability. This is about as close to a pick'em as you'll find.

On the underlying play at evens, Columbus has a slim edge. Their 52.0% xG% and 51.0% Corsi% beat Pittsburgh's 51.0% xG% and 50.0% Corsi%, but marginally. The gap widens only when you account for PDO. Columbus sits at 1.003—results tracking expectation. Pittsburgh is at 1.012, meaning the Penguins have run hotter than their play should allow; both goaltenders and favorable bounces have buoyed them.

Both S. Skinner and E. Merzlikins are on back-to-backs. Workload management could send a backup in, or ask a starter to push through fatigue. Either way, goaltending is a wildcard.

Boone Jenner is Columbus's fulcrum. He's hunting at 0.98 xG/60, with 11G and 21A on a 9.6% shooting rate, anchored by 2 high-danger goals. That's high volume from dangerous ice with above-average finish. Mathieu Olivier (0.89 xG/60, 12G, 9A) is shooting at 13.3% but has only 1 high-danger goal, implying his 12 goals have come on lower-danger chances—a sign of luck or opportunism rather than consistent pressure. Pittsburgh fields Rickard Rakell: 1.01 xG/60, 13G, 13A, and 3 high-danger goals at 12.1%. He's generating similar xG/60 to Olivier but finishing at a higher rate, particularly from in close.

Filip Hallander was recalled from his conditioning loan for evaluation and is out. Even when active, his 1.28 xG/60, 0 goals, and 3 assists mark him as a facilitator without finish.

Both teams have failed to convert rebound chances or high-danger opportunities outside their primary threats. Columbus and Pittsburgh both sit at 0% on high-danger and rebound goals, suggesting neither team is manufacturing second chances or feeding off hustle plays. Defensively, Conor Garland is Columbus's vulnerability; Caleb Jones is Pittsburgh's. Whoever breaks that secondary-scoring stasis will likely win.

The 51% split is right. This is even money. Execution matters more than scheme.