3-2: Final OT (OTT). 2.32-1.05: xG (VAN-OTT). 66%: Vancouver Corsi. 2.86x: Ottawa finishing vs xG This was a road win that ran almost entirely on finishing variance. Vancouver out-shot Ottawa 36 to 17 , owned 66% of shot attempts at 5-on-5, and produced more chance quality (2.32 xG to 1.05, 6-3 in high-danger looks). The Charge scored anyway. Three Ottawa goals on combined 0.24 xG , a per-shot finishing rate of 2.86 times expected . Sarah Wozniewicz tied the game with 54 seconds left in regulation; Rebecca Leslie won it 1:59 into overtime. Three goals on 0.22 combined xG The Charge's three scoring shots were 0.11, 0.05, and 0.06 xG . None was a true high-danger one-timer; Fanuza Kadirova 's 5-on-4 wrister at P2 19:25 was the highest-quality look of the three and the model still rated it as a roughly one-in-ten chance. Sarah Wozniewicz 's tying goal at P3 19:06 was a 0.05 xG attempt and Rebecca Leslie 's OT winner was 0.06 xG. Three goals on a quarter of expectation is the kind of finishing run that doesn't repeat on demand; the model says these three shots collectively score about one-fourth of a goal across many PWHL seasons. They produced three. Goaltending, decomposed Gwyneth Philips stopped 34 of 36 live shots worth 2.32 xG, finishing at +0.32 GSAx . The save process did most of the heavy lifting. Ottawa's shot-creation profile didn't keep pace at the other end, so Philips's slightly above-average line was the structural piece that gave the Charge a chance for the finishing run to mean something. Kristen Campbell faced 17 live shots worth 1.05 xG and conceded three. GSAx came out at -1.95 . All three Ottawa goals were under 0.12 xG; that GSAx number reads as variance landing on her side rather than a save-process problem. The save process underneath was much closer to expected than the headline number suggests. The game in cumulative xG The scoring timeline P2 19:25 · Fanuza Kadirova (PP) · xG 0.09 · OTT 1, VAN 0. P3 9:50 · Sarah Nurse · xG 0.09 · OTT 1, VAN 1. P3 14:38 · Sophie Jaques · xG 0.07 · OTT 1, VAN 2. P3 19:06 · Sarah Wozniewicz · xG 0.06 · OTT 2, VAN 2. P4 2:59 · Rebecca Leslie (OT) · xG 0.10 · OTT 3, VAN 2 Vancouver's looks above 0.07 xG The Goldeneyes' chances the model graded above the 7% scoring-probability threshold. Two of them went in (Nurse and Jaques in the third); the rest are the chances Philips turned away. Plenty of generation, but the high-danger ceiling held below where it needed to be to put the game out of reach. P1 10:22 · Mellissa Channell-Watkins · xG 0.27. P3 6:59 · Mannon McMahon · xG 0.17. P1 0:47 · Sarah Nurse · xG 0.11. P4 2:31 · Claire Thompson · xG 0.09. P1 1:12 · Sarah Nurse · xG 0.09. P2 13:33 · Michelle Karvinen · xG 0.09 What this read isn't Not a claim Ottawa was the better team. They weren't. Vancouver dominated the chance-creation profile by every measure the model produces. The result reflected a finishing run, not the underlying read on the run of play.. Not a single-game indictment of Kristen Campbell . Three goals on shots all under 0.12 xG is mostly variance landing on her side. The save process was clean.. Not a claim Ottawa's finishing rate is repeatable. Three goals on a team xG around 1 is roughly triple the game's expected finish, the kind of single-game run that fades hard on rolling samples. The chance-creation profile is what to keep over many games; this conversion isn't a forecast.. Not a claim Vancouver collapsed in the third. The Goldeneyes generated chances throughout the third and gave up two goals on combined sub-0.10 xG; that is finishing variance landing in the wrong window, not a structural breakdown. 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/75 .