5-3: Final (MIN). 1.54-2.29: xG (MIN-TOR). 0.49: MIN goals combined xG. -3.46: Campbell GSAx Minnesota scored five goals on the road on shots worth a combined 0.49 xG . The model graded none of the five Minnesota goals above 0.12 xG. Toronto outshot the Frost 30 to 25 and outchanced them 2.29 xG to 1.54 , but Minnesota's PP went 2-for-3 on two shots, Lee Stecklein scored twice, and Kristen Campbell finished at -3.46 GSAx in a rough save-process night. Five goals, none above 0.12 xG Lee Stecklein 's first goal at P2 4:41 was a 0.09 xG wrister. Michela Cava made it 2-1 at P2 10:57 on a 0.11 chance. Stecklein scored again at P2 12:59 on a 0.04 xG power-play shot . Sophie Jaques put Minnesota back ahead in the third on a 0.08 xG attempt. The PP capper from Mellissa Channell-Watkins at P3 18:49 , a 0.12 xG wrister, sealed the 5-3 final. The Frost converted at a rate more than 3x what the chance quality predicted; on a different night this is a 1-2 game with the same shot mix. Toronto's biggest chance didn't go The single best scoring chance of the game for either team was Daryl Watts 's 0.23 xG attempt at P3 7:04 . It didn't go. Julia Gosling 's 0.18 PP shot at P2 16:22 also stayed out. Toronto's three goals came on shots worth 0.11, 0.11 (PP), and 0.05 xG, which is mid-quality conversion of mid-quality chances. The Sceptres had the better chance profile (6 HD shots to 3) but couldn't break through Maddie Rooney's six high-danger saves. Hayley Scamurra opened the scoring in the first; the Toronto goals after that landed in a 27-second window in the second period to tie it 3-3. Goaltending, decomposed Maddie Rooney faced 30 shots worth 2.29 xG and allowed three for -0.71 GSAx . Save-process below expectation. None of the three goals against her was a high-danger chance: 0.11, 0.11, 0.05. She turned aside Watts's 0.23 third-period look and Gosling's 0.18 PP shot, and the win condition for Minnesota was that Rooney kept Toronto's bigger chances out while the Frost converted the smaller ones at the other end. Kristen Campbell stopped 20 of 25 live shots worth 1.54 xG, finishing at -3.46 GSAx . Five goals on shots worth a combined 0.44 xG is the save-process line that says the variance went sharply against her on a small workload. None of the five Minnesota goals was an in-tight high-danger chance; the model expected roughly 1.4 goals on the shot mix she faced. The realized five lands well outside what the chance quality predicted. The game in cumulative xG The scoring timeline P1 7:11 · Hayley Scamurra · xG 0.13 · MIN 0, TOR 1. P2 4:41 · Lee Stecklein · xG 0.11 · MIN 1, TOR 1. P2 10:57 · Michela Cava · xG 0.12 · MIN 2, TOR 1. P2 12:59 · Lee Stecklein (PP) · xG 0.04 · MIN 3, TOR 1. P2 16:56 · Savannah Harmon (PP) · xG 0.12 · MIN 3, TOR 2. P2 17:23 · Allie Munroe · xG 0.05 · MIN 3, TOR 3. P3 13:47 · Sophie Jaques · xG 0.10 · MIN 4, TOR 3. P3 18:49 · Mellissa Channell-Watkins (PP) · xG 0.11 · MIN 5, TOR 3 Top scoring chances TOR P3 7:04 · Daryl Watts default · xG 0.27. TOR P2 16:22 · Julia Gosling default · xG 0.20. TOR P3 0:48 · Renata Fast default · xG 0.14. TOR P1 14:41 · Emma Woods default · xG 0.14. TOR P1 7:11 · Hayley Scamurra default · xG 0.13 (goal). TOR P2 16:56 · Savannah Harmon default · xG 0.12 (goal). MIN P2 10:57 · Michela Cava default · xG 0.12 (goal). MIN P1 9:45 · Liz Schepers default · xG 0.12. MIN P2 4:41 · Lee Stecklein default · xG 0.11 (goal). MIN P3 18:49 · Mellissa Channell-Watkins default · xG 0.11 (goal). MIN P2 10:49 · Taylor Heise default · xG 0.11. MIN P3 13:47 · Sophie Jaques default · xG 0.10 (goal) What this read isn't Not a process win for Minnesota. Toronto outshot, outchanced, and won Corsi. The Frost won by converting at a rate three times what the chance quality predicted, and by getting steady goaltending against Toronto's bigger looks.. Not a single-game referendum on Kristen Campbell . A -3.46 GSAx night on 25 shots is a rare-tail outcome. Her save-process line over many games is much closer to expectation; this game was an outlier.. Not a sign Toronto's offense is broken. The Sceptres generated 2.09 xG and six high-danger chances. Watts's 0.23 third-period chance and Gosling's 0.18 PP shot didn't go; on a different night either of them produces a different scoreboard.. Not a model failure. Five goals on 0.44 xG is well past the median outcome for a 25-shot night, but it's a possible draw from the distribution. The model's per-shot probabilities are unchanged; the realized count was a tail event. 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. Overtime shots are included.. 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/298 .