Detroit won 2–1 in overtime, but the underlying numbers say Tampa Bay deserved better. The Lightning controlled 56% of expected goals at five-on-five while losing the game—a margin so significant that it constitutes not variance but theft.

Expected goals finished 3.89–3.10 in Tampa's favor. Detroit held 49% of shot attempts (31–32 on goal). High-danger chances split 6–4 to Tampa. None of this produced a Lightning win. J. Gibson made that impossible.

Gibson finished with 31 saves on 32 shots and a 0.969 save percentage, posting a +2.89 performance above expected. He faced six high-danger chances and stopped all six. By the numbers, he prevented nearly three goals beyond what models predict—the difference between a loss and a win.

A. Vasilevskiy made 29 saves on 31 shots and finished +1.10 above expected. His night was competent. The problem was not goaltending but finishing: Tampa converted zero of its six high-danger chances. They outplayed Detroit across the board and scored once. That imbalance doesn't happen twice.

The turning point arrived at 3:36 of overtime. Dylan Larkin took a snap shot from the high slot. It was all Detroit needed.

This was a game Tampa won through the first sixty minutes and lost in one. J. Gibson made that loss inevitable.

This is a genuine pick'em—the model projects expected goals of 2.6 for Tampa against 2.5 for Detroit, yet Detroit sits at 48% win probability.

Tampa Bay controls the underlying play. The Lightning run a 54.0% expected goals share at 5v5 against Detroit's 49.0%. They drive territorial pressure at 53.0% Corsi. But Tampa's 1.019 PDO signals a team converting luck at unsustainable rates. Regression is coming.

Detroit's edge lies in restraint. The Red Wings operate at 0.985 PDO, well below league mean. Their expected play doesn't approach Tampa's, but bad luck has masked their true strength. If that flips, even partially, Detroit's competitive.

The offensive matchups favor Tampa. Brandon Hagel (23G, 1.10 xG/60) generates high-danger chances at an elite rate, and his 15.8% shooting amplifies the result. Anthony Cirelli (15G, 17.0% shooting) is similarly efficient. Detroit counters with Alex DeBrincat (17G, 0.88 xG/60) as a volume-based threat and Marco Kasper (7G) converting at 6.1%, generating at only 0.81 xG/60—well below Cirelli or Hagel.

The deciding variable: Does Tampa's regression happen tonight? The Lightning own the underlying play by a clear margin. Victor Hedman represents their defensive weak link; Axel Sandin-Pellikka serves the same role for Detroit. The team that leaks fewer soft chances wins.

• Tampa's underlying edge is clear—54.0% xG share and 53.0% Corsi.
• But Tampa's luck is running out—1.019 PDO signals regression is imminent.
• Detroit sits at 0.985 PDO, due for a positive variance event.

Data: MoneyPuck 5v5 models, NHL API