Detroit stole this. Toronto controlled 57% of expected goals yet lost 2-1 in overtime—a heist that rested almost entirely on J. Gibson's shoulders.

The Maple Leafs dominated shot quality. They generated 0.074 xG per shot compared to Detroit's 0.044, creating 10 high-danger chances to the Red Wings' 5. Yet Toronto managed just one goal while Detroit capitalized on its opportunities more efficiently, including 2 goals on 31 rush chances.

The explanation lies in variance. J. Gibson posted a 0.968 SV%, stopping 3.54 goals above expected. J. Woll saved 1.38 goals above expected with a 0.951 SV%—strong, but insufficient when his team was inefficient. Toronto underperformed its xG by 3.54 goals; Detroit underperformed by just 1.38. That gap decided the game.

Gibson faced 10 high-danger chances and stopped 9, an elite performance on Detroit's most vulnerable possessions. The Red Wings took 41 shots to Toronto's 31, with inferior shot quality but superior finishing. Dylan Larkin's goal from the slot at 03:08 of overtime came on one of Detroit's five high-danger chances. Toronto's 10 high-danger opportunities yielded just one goal for the night.

Toronto crashed the net and generated rebound opportunities (5 attempts, 1 converted), yet one goal proved insufficient. Detroit created 4 rebound chances with no conversions, relying on Gibson's brilliance and sharper execution on its limited chances.

The model projects a 2.5-2.5 game with Detroit at 51% to win. This is as genuine a coin flip as you'll find — both teams possess genuine edges that cancel out.

Detroit owns the territorial edge: 49.0% Corsi and 49.0% expected goals, suggesting sustained puck possession and shot generation. Toronto sits at 45.0% Corsi and 46.0% xG, a significant gap in volume. But the fine print matters. Detroit's PDO is 0.985, indicating the Red Wings have been slightly unlucky; Toronto's is 1.002, suggesting the Maple Leafs are riding minor variance. Over time, that reverses.

The edge that tips the scale exists in rest. Detroit arrives on three days off; Toronto on one. This becomes acute at the goaltending position. J. Woll, on a back-to-back, may cede the crease to a backup or start fatigued — a small margin in a game this tight.

Detroit's Alex DeBrincat drives expected goals at 0.88 per 60 and shoots at 9.0%, posting 17 goals and 28 assists — including two from high-danger areas. Marco Kasper mirrors the profile (0.81 xG/60, 7 goals, 6.1% shooting) with one high-danger conversion. Toronto's Auston Matthews leads the creation department at 1.10 xG/60 with 17 goals and 13 assists, executing at 10.6%. John Tavares (0.91 xG/60, 15 goals and 24 assists, 12.4% shooting) has been more efficient with limited opportunity, three of his goals coming from prime scoring areas.

Neither team has scored from high-danger situations or rebounds at meaningful rates — both sit at 0% — which compresses the margin. Axel Sandin-Pellikka represents Detroit's defensive liability; Bo Groulx anchors Toronto's. In a low-variance matchup, these gaps matter less than rest and execution. The model's skepticism is justified. One team leaves with an extra point because a fresher unit simply executed when it mattered.