4-1: Final (MTL). 1.62-2.17: xG ex-EN (MTL-SEA). 7-7: HD (MTL-SEA). +1.17: Desbiens GSAx This game looked close on the surface but the chance-quality split favoured Montreal across both sides of the special-teams ledger. Seattle held 51% Corsi at 5-on-5 and won the shot count 25 to 23, but the high-danger battle came in tied at 7-7 and Montreal's in-play xG (1.62) sat above Seattle's (2.17) once the empty-netter is stripped. The decisive piece was Laura Stacey 's shorthanded goal in the second period; Seattle's 4-opportunity power play produced 0 goals and conceded one. The shorthanded goal that flipped the special-teams ledger Seattle drew a Montreal penalty midway through the second and went to the 5-on-4. The Torrent power play produced its volume but didn't break through; what broke through was a Montreal counter-rush. Laura Stacey took a feed and beat Hannah Murphy on the short side at P2 13:02. The model priced the shot at 0.10 xG, the kind of mid-quality look that scores roughly one time in ten. With Montreal already up a goal on the Catherine Dubois 5-on-5 wrister at P2 4:17, the shorthanded marker pushed the lead to two and effectively closed the chance-quality argument for the rest of regulation. A power-play opportunity that ends in a goal against is the largest single swing the special-teams ledger can produce. Montreal answered Seattle's tying-attempt goal in 64 seconds Alex Carpenter pulled Seattle within one at P3 7:56 on a 0.10 xG wrister, and 64 seconds later Skylar Irving restored the two-goal lead at P3 9:00 on a 0.06 xG tip. Laura Stacey added the empty-netter at P3 19:27 to set the final. Three of the four Montreal goals were on shots in the 0.06-0.12 xG range; finishing on mid-quality looks did the work in front of Ann-Renée Desbiens 's clean save line. Goaltending, decomposed Ann-Renée Desbiens stopped 24 of 25 live shots worth 2.17 xG, finishing at +1.17 GSAx . The only Seattle goal was Carpenter's 0.10 xG attempt; the seven high-danger chances Seattle generated mostly ran into clean saves at the doorstep. Hannah Murphy stopped 19 of 22 live shots worth 1.62 xG, finishing at -1.38 GSAx . The empty net is excluded from the line. Three regulation goals against on shots all under 0.12 xG is mostly variance landing on her side of the ledger; the headline GSAx reads worse than the underlying save process was. The game in cumulative xG The scoring timeline P2 4:17 · Catherine Dubois · xG 0.11 · MTL 1, SEA 0. P2 13:02 · Laura Stacey (SH) · xG 0.10 · MTL 2, SEA 0. P3 7:56 · Alex Carpenter · xG 0.10 · MTL 2, SEA 1. P3 9:00 · Skylar Irving · xG 0.07 · MTL 3, SEA 1. P3 19:27 · Laura Stacey (EN) · xG 0.00 · MTL 4, SEA 1 Seattle's looks above 0.07 xG The Torrent's chances the model graded above the 7% scoring-probability threshold. Carpenter's tying-attempt wrister sits in the middle; the rest are the chances Desbiens turned away. Plenty of generation, but no single look the model considered near-coin-flip to score. P2 10:22 · Jenna Buglioni · xG 0.26. P3 14:25 · Alex Carpenter · xG 0.21. P3 11:23 · Danielle Serdachny · xG 0.20. P3 12:20 · Julia Gosling · xG 0.18. P2 12:41 · Danielle Serdachny · xG 0.15. P2 10:21 · Jenna Buglioni · xG 0.14 What this read isn't Not a claim Seattle deserved better. They produced more shots and slightly better Corsi, but the shorthanded goal against on a power play was the largest single swing of the night and the chance-quality split otherwise leaned Montreal.. Not a single-game indictment of Hannah Murphy . Three regulation goals on shots under 0.12 xG is mostly finishing variance; the save process underneath the GSAx number was much closer to expected than the score line suggests.. Not a claim Seattle's PP is broken. Four power-play opportunities yielding zero goals AND conceding a shorthanded marker is a bad single-game outcome, but that conversion-against rate doesn't repeat. The right diagnostic is the rolling profile across the next several games.. Not a model failure. A near-even underlying chance-quality game decided by a special-teams swing of two goals is well within the variance the model's probability outputs allow on small samples. Methodology xG model : PWHL-calibrated logistic regression on 16,709 non-EN shots. Same fit as the methodology post . Features: distance, angle, rebound (≤3 s), penalty shot, power play, overtime, scorer-tagged quality chance, seven shot-type dummies. Empty-net shots are priced outside the main model (MoneyPuck / NST convention).. GSAx : summed xG on shots faced minus goals allowed, computed per-goalie from the PBP feed. Empty-net shots and goals are excluded from both sides of the ledger.. Numbers regenerate on every build. A PBP re-sync or xG recalibration reshapes every figure in this post on the next static export. Full play-by-play, shift overlays, and the per-player game log for this game are at /games/82 .