6-5: Final (MIN). 4.86: Combined xG. 3/3: MIN power play. -6.14: Combined GSAx Eleven goals went in. The model's combined expected-goals total for both teams was 4.86 . That gap is the analytical signature of the entire night: a finishing run on both sides of the rink against goalies who could not get out of the way. Emerance Maschmeyer and Nicole Hensley combined for -6.14 GSAx ; the average PWHL game generates near zero combined GSAx by construction. The 6-5 final reads close because it was close. The path to it was anything but quiet. Minnesota's power play won a game they otherwise lost At even strength, Vancouver outscored Minnesota 4 to 3 . The Frost lost the 5-on-5 battle by a goal and won the special-teams ledger by two: 3-for-3 on the power play against Vancouver's 1-for-3. Three opportunities, three goals. Kelly Pannek accounted for two of those personally (P3 4:17 and P3 13:44, both on the 5-on-4); Taylor Heise had the other in the second period. The model graded the three Minnesota power-play goals at 0.06, 0.14, and 0.15 xG; combined about 0.35 xG of expectation against three actual goals. That is the kind of conversion run that does not repeat, and on this night it was the entire margin. Two goals in 7 seconds At P1 17:39, Katy Knoll put MIN ahead. 7 seconds later , at P1 17:46, Tereza Vanišová answered for VAN. Back-to-back goals inside a half-minute window are mostly an artifact of post-goal faceoff variance and quick line matchups, not a structural breakdown either way; both shots were graded above 0.15 xG, so the chances were real, but the rapid sequence is the kind of single-game noise the season-long process numbers smooth out. Goaltending, decomposed Emerance Maschmeyer faced 35 shots worth 2.91 xG and conceded six. That is -3.09 GSAx , the kind of single-game number that lives in the bottom 1% of starts. None of the goals was particularly egregious in isolation: the highest-xG shot Minnesota scored on was Katy Knoll 's 0.27 first-period tip, the rest were below 0.16. Six of those finding the net is where the GSAx comes from. On the other end, Nicole Hensley faced 25 shots worth 1.95 xG and conceded five. Her GSAx was -3.05 , essentially the same shape: a workload she was supposed to keep around two goals against, conceding five. Vancouver's biggest scoring chance of the night was Sarah Nurse 's opening tip at 0.29 xG; the other four scoring shots were below 0.16. This was not a game where either goalie was facing a wave of high-danger looks. It was a game where mid-quality chances kept finding space, in both creases. The game in cumulative xG The scoring timeline P1 1:18 · Sarah Nurse · xG 0.31 · MIN 0, VAN 1. P1 9:05 · Grace Zumwinkle · xG 0.13 · MIN 1, VAN 1. P1 17:39 · Katy Knoll · xG 0.30 · MIN 2, VAN 1. P1 17:46 · Tereza Vanišová · xG 0.13 · MIN 2, VAN 2. P1 18:19 · Sophie Jaques (PP) · xG 0.03 · MIN 2, VAN 3. P2 3:30 · Jenn Gardiner · xG 0.12 · MIN 2, VAN 4. P2 12:07 · Mae Batherson · xG 0.11 · MIN 3, VAN 4. P2 13:25 · Taylor Heise (PP) · xG 0.05 · MIN 4, VAN 4. P3 4:17 · Kelly Pannek (PP) · xG 0.16 · MIN 5, VAN 4. P3 13:44 · Kelly Pannek (PP) · xG 0.12 · MIN 6, VAN 4. P3 19:07 · Izzy Daniel · xG 0.01 · MIN 6, VAN 5 What this read isn't Not a claim either goalie played poorly in any structural way. Eleven goals on five combined xG is finishing variance landing simultaneously in both nets. The right diagnostic for both goaltenders is the rolling save profile across the next several starts, not this single shared outlier night.. Not a claim Minnesota's 3-for-3 power play is repeatable. Three power-play goals on 0.35 xG is roughly nine times expected. The 5-on-4 unit is good, but the conversion rate this game is closer to a casino number than a season-long forecast.. Not a claim Vancouver was unlucky to lose. They won the 5-on-5 battle by a goal and the special-teams battle went two the other way. That is a -1 swing from one of the harder areas to defend, not a bad-luck story.. Not a model failure. An eleven-goal game with combined xG of 5.0 is well inside the variance that the model's probability outputs allow over a sample of this size. Hot games happen; this was one. 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/98 .