The Senators controlled 58% of expected goals at 5v5, yet lost 2–1. A heist requires a thief.

D. Kuemper was that thief. The Kings goaltender stopped 27 of 28 shots (0.964 SV%) and outperformed his expected save rate by 2.66 goals. More damning for Ottawa: he allowed zero of six high-danger chances. The Senators created a 1–6 edge in those chances and converted exactly none.

Ottawa's broader shot geography told the same story. The Senators took 28 shots to Los Angeles' 22, yet surrendered the game. The Kings' 42% expected-goals share masked a fundamental imbalance—L.A. was being badly outplayed.

The goal distribution exposed why shooting 58% xG sometimes isn't enough. L. Meriläinen (OTT) posted a 0.909 SV%, beating his expected rate by 0.67 goals, but that surplus evaporated against D. Kuemper's wall. The Senators underperformed their xG by 2.66 goals, a deficit Meriläinen could not bridge.

Ottawa's inability to finish was the night's central fact. They created six rebound opportunities and converted one. High-danger chances hit at zero percent. Rebound control meant nothing without conversion.

The turning point arrived in the third period: Brandt Clarke's slap shot from the point at 13:50. That goal, not goaltending heroics, shifted momentum when it mattered most.

The model sees a pick'em. It projects the game at 2.4–2.4 goals, with Ottawa favored at 51% win probability—close enough to a coin flip that narrative doesn't matter.

This near-parity masks an underlying truth: Ottawa owns the underlying play. The Senators rank 56.0% in 5v5 xG and 53.0% in Corsi%, while Los Angeles sits at 52.0% and 52.0% respectively. That's meaningful control, but luck narrows the margin. Ottawa's PDO of 0.993 suggests regression upward; Los Angeles at 1.002 runs roughly to expectation.

Brady Tkachuk (1.19 xG/60, 11G 20A, 7.5% shooting, 2 high-danger goals) and Dylan Cozens (1.00 xG/60, 13G 15A, 9.5% shooting, 1 high-danger goal) power Ottawa's attack. Tkachuk generates the most chances on the sheet and converts at 7.5%; Cozens sustains an elite 9.5% shooting rate despite lower volume. Los Angeles counters with Alex Laferriere (0.87 xG/60, 16G 18A, 9.9% shooting, 2 high-danger goals), who scores more often than his chances suggest—a sign that finishing, not volume, carries his load. Jared Wright (0.83 xG/60, 0G 4A, 0.0% shooting) generates looks without capitalizing.

Both teams show identical inefficiency at the margins. Nick Jensen (OTT) and Adrian Kempe (LAK) are defensive weak links; the model's narrow projection suggests neither side will crack decisively.

If Ottawa's process advantage converts to results, they escape with a win. If Los Angeles' superior efficiency on fewer chances finds daylight, the Kings keep pace. The model can't resolve the tie because the data doesn't support one. Trust the coin flip.