5-1: Final (OTT). 29-24: Shots (OTT-NYS). 2.56-2.19: xG (OTT-NYS). 3-1: Special-teams goals The xG ledger says this was close. Ottawa finished at 2.56 against New York's 2.19 , an edge of 0.40 expected goals, and held a 7-5 advantage in high-danger chances . Corsi was 51.9% for Ottawa, 48.1% for New York. Under those inputs, the model's point expectation is something like a narrow regulation win or a coin-flip OT game. Instead the scoreboard read 5-1. That four-goal gap was not built by five-on-five play. Where the four-goal gap came from Ottawa took 5 minor penalties to New York's 4 . On paper, the Sirens were the team with the special-teams opportunity ( 5 power-play tries to Ottawa's 4 ). On the scoreboard, special teams moved the other way. Ottawa went 1-for-4 on the power play and, more decisively, scored 2 shorthanded goals while killing penalties. New York's power play converted 1-for-5 . Net special-teams goal differential was OTT +2, NYS -2 . Subtract that from the final and you are back to a one-goal game at five-on-five. Two shorthanded goals in six minutes The second period carried the swing. At P2 7:54 , with Ottawa skating three against four (Guilday in the box from the coincidental roughings, Leslie for holding), Jocelyne Larocque beat Osborne from the slot on a pass from Brianne Jenner . The model priced the shot at 0.21 xG , a mid-slot chance the average PWHL shooter converts about one time in five. Ottawa converted it while down two skaters. Six minutes later, at P2 14:08 , Peyton Hemp scored again on the penalty kill ( 0.15 xG ), this time on a standard four-on-five, assisted by Alexa Vasko . The two shorthanded goals combined for roughly 0.36 xG , or about a third of an expected goal. Ottawa scored two. New York's power play generated 0.57 xG across all five opportunities and converted once. The penalty kill outscored the power play. The finishing gap Ottawa scored five goals on 2.56 xG. New York scored one on 2.19. The Charge's per-shot conversion was 0.172 goals per shot on 29 shots, which on an 2.56 expected-goals budget is a substantial overperformance. Most of that gap sits on the goaltender's line, not on the shooters. The model thought the five chances Ottawa converted were worth 0.64 xG combined . Osborne stopping three of those five would have kept this a one-goal game. Goaltending, decomposed Gwyneth Philips faced 24 shots worth 2.19 xG and gave up one. Her GSAx for the night was +1.19 , which puts her about an expected goal above the average PWHL goalie handed the same workload. New York generated five chances the model graded at 0.10 xG or higher; Philips stopped all but the Levis power-play goal. Kayle Osborne faced 29 shots worth 2.56 xG and allowed five. That is a -2.44 GSAx night, roughly 2.7 goals below what the model would expect a league-average goalie to concede on that shot profile. Two of the five against came on shorthanded plays she had no reason to expect a clean look at, but the other three were from locations the model priced below 0.20 xG. Combined, the two goalies accounted for a swing of nearly +3.62 expected goals in Ottawa's favour before any of the special-teams scoring even mattered. Jenner and Larocque touched four of the five Brianne Jenner had 1G and 2A. Jocelyne Larocque had 1G and 1A. Between them they were on the scoresheet for four of Ottawa's five goals: Jenner's primary assist on the Leslie tying goal, her primary assist on the Larocque shorthanded goal, Larocque's goal itself, Larocque's assist on the Jenner power-play goal in the third. The only Ottawa marker without either name was Ronja Savolainen 's goal in the third, set up by Emily Clark and Gabbie Hughes . Assist clustering like that is not a model finding, but it is a clean way to read a game where special teams overwhelmed the run of play. Corsi versus xG, at even strength Ottawa held 51.9% of Corsi and 54.7% of shots , both narrow edges. At five-on-five the underlying numbers were close to even. Ottawa's shot-quality edge shows up in the high-danger column ( 7 to 5 ) and in the faceoff split ( 58.5% for Ottawa, with a 14-of-23 advantage in the defensive zone that kept the Sirens' offensive-zone starts short). Those are real wins, but they are worth tenths of expected goals, not four. The game in cumulative xG The scoring timeline P1 11:44 · Paetyn Levis (PP) · xG 0.17 · OTT 0, NYS 1. P1 18:59 · Rebecca Leslie · xG 0.13 · OTT 1, NYS 1. P2 7:54 · Jocelyne Larocque (SH) · xG 0.22 · OTT 2, NYS 1. P2 14:08 · Peyton Hemp (SH) · xG 0.15 · OTT 3, NYS 1. P3 3:15 · Ronja Savolainen · xG 0.03 · OTT 4, NYS 1. P3 12:28 · Brianne Jenner (PP) · xG 0.13 · OTT 5, NYS 1 New York's best non-goals The chances Philips turned aside. Every entry is a shot the model graded at 0.07 xG or higher (about a 7% average-shooter conversion rate). The list clusters in the first period, when the run of play was closest. P1 9:17 · Maddi Wheeler · xG 0.18. P1 14:23 · Casey O'Brien · xG 0.17. P1 10:23 · Sarah Fillier · xG 0.16. P3 7:03 · Anne Cherkowski · xG 0.13. P3 15:44 · Paetyn Levis · xG 0.13. P2 2:49 · Maddi Wheeler · xG 0.12 The penalty timeline P1 10:10 · OTT · Stephanie Markowski · Delay of Game (2 min). P1 13:33 · OTT · Peyton Hemp · Holding (2 min). P2 6:16 · NYS · Emmy Fecteau · Roughing (2 min). P2 6:16 · OTT · Rory Guilday · Roughing (2 min). P2 7:11 · OTT · Rebecca Leslie · Holding (2 min). P2 10:06 · NYS · Sarah Fillier · Slashing (2 min). P2 13:31 · OTT · Rory Guilday · Interference (2 min). P2 18:32 · NYS · Denisa Křížová · Ob-Tripping (2 min). P3 11:08 · NYS · Denisa Křížová · Roughing (2 min) What this read isn't Not a claim Ottawa controlled play. The xG edge was 0.40, the Corsi split was within two points of even, and New York had more power-play time. Five-on-five, this was a close game.. Not a shot-quality story for the shorthanded goals. Two shorthanded goals on 0.36 combined xG is unusual by volume, not by finishing. A 0.21 slot shot and a 0.15 mid-range chance are the kinds of looks league-average shooters should convert occasionally. The story is that Ottawa generated 4-on-5 chances at all, and that Osborne didn't stop the mid-probability ones.. Not a narrative about momentum or travel. This post doesn't reference either. If five of the six goals had come on grade-A looks, a momentum read might carry weight. They didn't, so it doesn't.. Not a referee story. Ottawa took more penalties than New York and still won the special-teams battle. That framing is not available in either direction. Methodology xG model : PWHL-calibrated logistic regression on 16,709 non-EN shots. 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 and goals are excluded.. Shorthanded attribution : strength states 3v4 and 4v5 at goal time are classified as Ottawa-shorthanded. The 3v4 marker appears during the second-period coincidental-roughing window plus the overlapping Leslie minor.. 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/111 .