5-6 OT: Final (SEA). 26-42: Shots (SEA-VAN). 1.52-3.44: xG (SEA-VAN). 6-12: High-danger (SEA-VAN) Every underlying indicator in this game pointed one way. Vancouver held 60.4% of Corsi , generated 3.44 xG to Seattle's 1.52 , took 42 shots to 26 , and doubled Seattle 12 to 6 in high-danger chances. The only metric that did not agree with the rest was the scoreboard, and then only for about fifty-seven minutes. Seattle led 2-0 after one, 2-1 after two, and 5-3 with 3:53 left in regulation. Vancouver scored three unanswered. The process and the result arrived at the same place from opposite directions. Seattle built the lead on two low-xG finishes Seattle's first two goals combined for about 0.33 xG. Hilary Knight opened scoring at 13:11 of the first on the game's highest-xG shot by either team ( 0.31 xG ), assisted by Danielle Serdachny and Julia Gosling . Then Anna Wilgren scored from distance at 18:58 on a shot the model priced at 0.02 xG , the kind of long-range look the average PWHL shooter converts about one time in fifty, assisted by Cayla Barnes and Hilary Knight . The Knight goal was a legitimate slot conversion. The Wilgren goal was a point shot that found a hole. Either way the Torrent were up 2-0 on 9 shots against 10. Vancouver's second and third periods The Goldeneyes took over territorially from the second period on. Over the final forty minutes of regulation plus overtime, Vancouver outshot Seattle 32 to 17 and won Corsi by a wider margin than the game-level number suggests. Seattle's first-period shot edge was the one stretch where the two teams were close to even. Once the Goldeneyes' volume game started producing chances, Vancouver's per-shot xG ( 0.082 , compared with Seattle's 0.058 ) was close enough to Seattle's that their extra attempt count translated into an xG gap of nearly a full expected goal. Jenn Gardiner scored four Jenn Gardiner 's four goals came on shots the model priced at 0.65 xG combined . The first was a 0.14 xG tap at the buzzer of the second period to cut the deficit to 2-1. The second was a 0.06 xG rebound at 9:28 of the third, assisted by Abby Boreen and Hannah Miller , that gave Vancouver a short-lived 3-2 lead. The third was a 0.28 xG look at 18:11, assisted by Michelle Karvinen and Anna Meixner , that cut a two-goal Seattle lead in half. The fourth was a 0.17 xG wrist shot 52 seconds into 3-on-3 overtime to win it, assisted by Hannah Miller and Sophie Jaques . The shot quality ran from low-probability to decent, the aggregate sits around 0.65 xG, and Gardiner finished every one. The empty net and the comeback Vancouver pulled Maschmeyer trailing 4-3 with roughly four minutes left in regulation. At P3 16:07 , Danielle Serdachny hit the empty net to make it 5-3 (assisted by Cayla Barnes and Aneta Tejralová ). Most coaches would not pull the goalie again after conceding an empty-net goal with 3:53 left and the deficit now at two. Vancouver did. Two minutes later Jenn Gardiner scored at 6-on-5 to make it 5-4, and with two seconds left in regulation Hannah Miller tied the game at 6-on-5, assisted by Sarah Nurse and Sophie Jaques . The tying shot was priced at 0.14 xG . The sequence of decisions that produced those two goals started with committing to the empty net past the point where most coaches re-insert the goalie. That decision cost a goal and produced two. Goaltending, decomposed Carly Jackson faced 42 shots worth 3.44 xG and gave up six. Her GSAx for the night was -2.56 , about two-and-a-half goals below what a league-average goalie would be expected to concede on that shot profile. None of the six goals came on a shot the model priced above 0.30 xG. The issue was volume plus finishing: every Vancouver shot that beat her was a look the model says most goalies would have held. Emerance Maschmeyer faced a lighter workload, 25 shots worth 1.52 xG (empty-net shots excluded), and allowed four non-EN goals. Her GSAx was -2.48 , comparable to Jackson's on a shorter shift. Neither goalie had a good night. The difference was that Vancouver's skaters produced enough offence to cover Maschmeyer's lapses and Seattle's did not cover Jackson's. The combined GSAx across the two netminders is within a tenth of symmetric, which is another way of saying the goaltending did not decide this game. The volume imbalance did. Special teams were a non-event Both teams went 0-for on the power play. Seattle had 2 opportunities and generated zero shots on them. Vancouver had 3 and generated 0.37 xG on 5 shots . Every goal in this game was scored at 5-on-5 or at 6-on-5 with an extra attacker. That is unusual with five minors called, and it means the game's structure was determined entirely by run-of-play minutes. The xG gap is a five-on-five xG gap, unmediated by power-play accumulation on either side. Corsi versus xG Vancouver's 60.4% Corsi and 61.8% shot share were matched by a per-shot quality roughly in line with Seattle's. Typically when Corsi and xG agree at this scale, the team behind in attempts is also behind in quality per attempt, compounding the edge. That did not happen here: Seattle's chances were just as good per-shot as Vancouver's. They simply did not produce enough of them. The high-danger split ( 12-6 ) maps onto the volume difference rather than a Seattle defensive collapse in the slot; Vancouver generated more of everything. The game in cumulative xG The scoring timeline P1 13:11 · Hilary Knight · xG 0.31 · SEA 1, VAN 0. P1 18:58 · Anna Wilgren · xG 0.02 · SEA 2, VAN 0. P2 19:05 · Jenn Gardiner · xG 0.14 · SEA 2, VAN 1. P3 3:04 · Hannah Miller · xG 0.07 · SEA 2, VAN 2. P3 9:28 · Jenn Gardiner · xG 0.06 · SEA 2, VAN 3. P3 10:07 · Cayla Barnes · xG 0.16 · SEA 3, VAN 3. P3 11:09 · Anna Wilgren · xG 0.04 · SEA 4, VAN 3. P3 16:07 · Danielle Serdachny (EN) · xG 0.00 · SEA 5, VAN 3. P3 18:11 · Jenn Gardiner · xG 0.28 · SEA 5, VAN 4. P3 19:58 · Hannah Miller · xG 0.14 · SEA 5, VAN 5. P4 0:52 · Jenn Gardiner · xG 0.17 · SEA 5, VAN 6 Vancouver's best non-goals The chances Jackson stopped. Every entry is a shot the model graded at 0.07 xG or higher. The list is longer than Seattle's despite a lower conversion rate, because Vancouver's shot count was larger and spread across more mid-quality looks. P2 17:33 · Anna Segedi · xG 0.26. P3 15:24 · Madison Samoskevich · xG 0.22. P3 18:10 · Michelle Karvinen · xG 0.17. P2 15:27 · Sophie Jaques · xG 0.13. P2 11:29 · Tereza Vanišová · xG 0.11. P2 5:10 · Sarah Nurse · xG 0.10 Seattle's best non-goals P1 9:20 · Alex Carpenter · xG 0.12. P1 2:37 · Aneta Tejralová · xG 0.11. P3 13:36 · Theresa Schafzahl · xG 0.11. P2 1:48 · Julia Gosling · xG 0.09. P2 19:53 · Danielle Serdachny · xG 0.08 The penalty timeline P2 4:58 · SEA · Hilary Knight · Holding (2 min). P2 19:38 · SEA · Cayla Barnes · Roughing (2 min). P2 19:38 · VAN · Sarah Nurse · Slashing (2 min). P3 6:08 · VAN · Sarah Nurse · Hooking (2 min). P3 6:41 · SEA · Cayla Barnes · Holding (2 min) What this read isn't Not a claim Seattle played poorly. Seattle's per-shot xG (0.058) was competitive with Vancouver's (0.082). The Torrent just got outshot 42-26 across sixty-five minutes.. Not a goaltending story in isolation. Jackson was under expectation and so was Maschmeyer. Combined GSAx is within a tenth of zero. The decisive variable was shot volume, not save percentage.. Not a goaltending-independent Gardiner story. Four Gardiner goals on 0.65 combined xG averages around 0.16 xG per shot, inside the distribution of chances most forwards get most nights. Gardiner finishing four of them on one night is a plausible tail outcome that sits on top of a team-level xG gap of 1.92 that already argued Vancouver was the better team.. Not a coaching verdict on the early goalie pull. Pulling with four minutes down one, and re-pulling after conceding the empty-netter, is a defensible expected-value decision under the standard win-probability models. It happened to work. The result does not vindicate the process, and a worse result would not have invalidated it. 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 or less), 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 (goalie pulled) are excluded on both sides of the ledger.. 6v5 attribution : the tying goal at P3 19:58 was scored with Maschmeyer on the bench for an extra attacker. The strength state is tagged 5v5 in the feed because the source doesn't expose the extra-attacker indicator separately; the post labels it 6v5 based on the goalie-pull sequence.. 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/110 .