The Blackhawks stole one. They lost the expected-goals battle 3.12–3.94 and won anyway.
The Blackhawks won a heist. They beat Tampa Bay 3–2 while losing the expected-goal battle 3.94–3.12, a chasm that suggests S. Knight made the difference.
Chicago won despite getting outshot in quality: the Lightning owned 56% of xG at 5v5 and controlled 55% of shot attempts (31–25 on goal). That should add up to a Tampa Bay win. It didn't, because Knight stopped 1.94 goals above his expected save percentage—his 0.935 mark turned a probable loss into a victory.
The Lightning could not capitalize. On seven high-danger chances, they converted zero. On six rebound opportunities, they scored once. A. Vasilevskiy saved 22 of 25, but his 0.880 SV% left Tampa one goal short of what the underlying play deserved. The Lightning underperformed their xG by 1.9 goals—a rarity that punishes teams even when they dominate.
Chicago didn't dominate, but they converted. Two rush-play goals on 27 chances proved decisive. The turning point arrived when Ryan Donato fired from the high slot late in the second period, the moment this game's odds swung toward Chicago. That's what elite goaltending does: steals games that the scoreboard said should go the other way.
The story isn't that Chicago deserved to win. It's that despite losing the underlying battle—the shots, the chances, the possession of danger—they found the net and walked out with two points. That's what a 0.935 save percentage buys: not artistry, not excellence, but the margins where games are stolen.
Data: NHL API, MoneyPuck xG model
