The Ducks won a game the numbers said they shouldn't have. Expected goals favored Montreal 4.25 to 3.51—a margin suggesting the Canadiens generated genuinely more threatening chances. Yet Anaheim stole a 6-5 shootout with ruthless opportunism on low-danger chances and rush offense.

The Ducks held 54% of shot attempts (33–28 on goal), but shot volume masked inferior chance quality. They converted twice from low-danger spots—areas that demand nothing of a goalie. S. Montembeault in Montreal's net faced 33 shots and fell 1.49 goals short of expected performance, a penalty no one survives. Yet L. Dostal for Anaheim allowed 0.75 more goals than expected; his team still won, variance incarnate.

Anaheim proved lethal off the rush, scoring 3 times on 22 chances—a rate that punishes defensive breakdowns in tight games. Montreal countered with 5 rush goals on 38 attempts, a conversion that looked efficient until the scoreboard betrayed them. Where Anaheim collapsed: they managed 0 goals on 5 high-danger chances. S. Montembeault's 100% save rate on those shots was virtuosic—wasted.

C. Kreider orchestrated the upset with 1G, 3A in 17:40, setting Anaheim's opportunistic offense in motion. L. Hutson answered for Montreal with 1G, 2A in 25:51; C. Caufield chipped in 2G, 1A in 19:48. The turning point arrived in the shootout: Alex Killorn's snap shot from the slot.

Anaheim's expected goals deficit proves that scoreboards and underlying metrics don't always align—especially when your team overperforms xG by 2.49 goals and your opponent by only 0.75. The Ducks won the numbers lost.

The model projects a 2.5–2.9 goal game with Anaheim at 56% win probability. The Ducks own the underlying play — 52.0% xG and 52.0% Corsi translate to sustained territorial advantage.

The gap is real but contained. Montreal's 48.0% xG and 49.0% Corsi show the Canadiens competing at even strength — they're not dominated, just methodically bested on volume and efficiency.

This is where PDO regression matters. The Ducks run at 0.981 — converting below their underlying shot quality, positioning them for positive regression. Montreal sits at 1.018, running hot, expecting regression downward. The model's 56% lean incorporates this edge but doesn't overcommit; it's calibrated confidence, not certainty.

V. Husso starting on a back-to-back is the variable. A well-rested backup tips Montreal's way; a fatigued starter accelerates Anaheim's advantage.

The Player Angle

Cole Caufield drives Montreal's offense at 0.97 xG/60, with 33 goals backing a 20.0% shooting rate. Yet he's converted just 1 high-danger chance all season — his production has outpaced chance quality. Oliver Kapanen adds 19 goals at 0.83 xG/60, benefiting from a hot 17.8% shooting percentage and 4 high-danger goals.

Beckett Sennecke represents Anaheim's depth edge: 1.03 xG/60, 20 goals, and 8 high-danger goals demonstrate elite chance generation paired with elite finish. Cutter Gauthier (0.99 xG/60, 25 goals) reads differently — his 2 high-danger goals against a 12.9% shooting rate signal he's punching above his underlying performance, another regression candidate.

Mike Matheson is Montreal's defensive weak link; Mason McTavish fills that role for Anaheim.

The Verdict

Anaheim's structural advantage — superior shot creation and luck regressing toward the mean — is the story. The 56% model probability reflects a team slightly better in a competitive matchup, with PDO offering marginal reinforcement.