0-4: Final (OTT). 2.22-1.11: xG (OTT-MTL). 5-4: HD chances (OTT-MTL). 3-1 MTL: Walter Cup Champions The 2026 PWHL season ended in Ottawa. Montréal won 4-0 and lifted the Walter Cup on the road, finishing the best-of-five 3 games to 1 . The line on the scoreboard tells one story. The underlying numbers tell another. Ottawa had 2.22 expected goals (xG) , the chance-weighted value of a team's shots, to Montréal's 1.11 . The Charge outshot the Victoire 23 to 16 , won 54 percent of the shot attempts, and took 71 percent of the faceoffs. They got shut out. Ann-Renée Desbiens stopped all 23 shots for a +2.22 GSAx performance. Abby Roque scored twice. One of those was shorthanded, on the back of a sequence that turned an Ottawa power play into the moment the game broke. Ottawa controlled the play. The scoreboard didn't move. Through forty minutes this looked like a tight, low-event game with Ottawa carrying a little more of it. Abby Roque got Montréal on the board at P2 3:49 , finishing a Marie-Philip Poulin setup with Laura Stacey also credited, on a 5-on-5 chance the model graded at 0.05 xG . After two periods Montréal led 1-0 on a half-goal of expected offence. The third period is where the gap opened, and where it closed: Ottawa took 13 shots in the period to Montréal's 4 , generated 1.82 xG to Montréal's 0.39 , and conceded three goals without scoring. 5-6: P1 shots (OTT-MTL). 5-6: P2 shots (OTT-MTL). 13-4: P3 shots (OTT-MTL) The 47 seconds that decided it Marie-Philip Poulin took a minor at P3 8:11 with Montréal up 1-0. Ottawa had a power play and a chance to push the game back to even. The Charge generated their two biggest looks of the night inside that man-advantage. Gabbie Hughes got a 0.38 xG shot at P3 9:19 . Brianne Jenner followed one second later with a 0.46 xG chance, the single biggest of the game. Desbiens stopped both. Thirty-nine seconds later, before Poulin's penalty had expired, Abby Roque got loose at P3 9:58 and scored shorthanded to make it 2-0. The model graded her chance at just 0.11 xG . Ottawa generated 0.99 xG of power-play offence across three opportunities and went 0-for-3 . Montréal generated one shorthanded shot worth 0.11 xG, and scored. That is the swing the game was decided on. What followed felt like an emptying-out. Maggie Flaherty scored at P3 13:54 off a Jessica DiGirolamo feed, Dara Greig also assisting, to make it 3-0 on a 0.16 xG shot. Lina Ljungblom finished it at P3 15:44 , unassisted, on a 0.10 xG chance. Ottawa pulled Gwyneth Philips with just under four minutes left, down four. Montréal did not need to add to the lead, and didn't. Win probability: an even game until Roque scored shorthanded The live win-probability model below tells the same story in a different shape. The trace stays near 50 percent for Ottawa through the first period and a chunk of the second. Roque's opening goal drops it to the low thirties. Then Ottawa pulls back up into the forties on the power play in the third, and falls off a cliff when the second Montréal goal goes in shorthanded. From there it monotonically declines. There is no late spike, no comeback chance, no missed slam-dunk that would have changed the curve. By the time the goalie comes out, the game has been gone for five minutes. Desbiens, the wall Ann-Renée Desbiens faced 23 shots worth 2.22 xG and allowed none. That is a +2.22 GSAx shutout in a Cup-clinching game. GSAx, goals saved above expected, compares the goals a goalie allowed to the summed xG of the shots she faced. A +2.22 line means she saved roughly two goals more than an average goaltender would have on the same workload. In a 4-0 game, that gap was the difference between a shutout and a competitive game that could have looked very different in the third. Across the four-game series Desbiens started every game for Montréal and her body of work was the most durable Montréal advantage. Game 4 was the cleanest version of it: a high-shot night, a high-xG night, no goals through. The two biggest stops were the Hughes and Jenner power-play looks at P3 9:19 and P3 9:20 that totalled 0.84 xG and became the buildup to Roque's shorthanded goal at the other end. Rory Guilday also tested her twice late, with chances at P3 17:09 and P3 17:43 totalling 0.33 xG. None of it dented the shutout. Philips and the model Gwyneth Philips stopped 12 of 16 and finished at -2.89 GSAx . The model expected about one goal on her workload; she allowed four. That is a heavy line for a single game, and it is not the line of a goalie playing badly on most of the shots. Three of the four Montréal goals graded at 0.16 xG or below : the kind of mid-range looks that beat any goalie a small fraction of the time. The fourth was the shorthanded Roque breakaway, also priced low. The model is telling us this is the bad end of the distribution, not a structural collapse, and that the same workload played a hundred times produces a 4-0 result very rarely. That framing matters because it is also true that on this night, in this game, the Charge needed at least one of those four to be saved. They generated 2.22 xG and got zero goals on the other side. With average goaltending both ways, the same chance count would have produced something like a one-goal Ottawa win. The chance count was not the problem. The finishing was, on one end, and so was the saving, on the other. The scoring timeline P2 3:49 · Abby Roque · xG 0.05 · OTT 0, MTL 1. P3 9:58 · Abby Roque (SH) · xG 0.11 · OTT 0, MTL 2. P3 13:54 · Maggie Flaherty · xG 0.16 · OTT 0, MTL 3. P3 15:44 · Lina Ljungblom · xG 0.10 · OTT 0, MTL 4 Top scoring chances OTT P3 9:20 · Brianne Jenner default · xG 0.46. OTT P3 9:19 · Gabbie Hughes default · xG 0.38. OTT P3 17:43 · Rory Guilday default · xG 0.18. OTT P3 17:09 · Rory Guilday default · xG 0.16. OTT P3 13:55 · Sarah Wozniewicz default · xG 0.14. OTT P2 3:16 · Jocelyne Larocque default · xG 0.13. MTL P2 9:05 · Catherine Dubois default · xG 0.17. MTL P3 13:54 · Maggie Flaherty default · xG 0.16 (goal). MTL P3 9:58 · Abby Roque default · xG 0.11 (goal). MTL P2 16:09 · Catherine Dubois default · xG 0.10. MTL P3 15:44 · Lina Ljungblom default · xG 0.10 (goal). MTL P1 10:17 · Marie-Philip Poulin default · xG 0.10 What this read isn't Not a 'Montréal got lucky' argument. Montréal won the series 3-1. Three of the four games were one-goal, low-xG affairs; the fourth was this one. A team that posts +2.22 GSAx in a Cup-clincher on the road has done more than survive variance. The point of the model is to say where the chances were, not to overrule who won.. Not a model failure. A 2.22 xG night producing zero goals against an above-replacement goalie is on the low end of expected, not anomalous. The same applies to four goals on 1.11 xG against a goalie having a rough night. Both pieces fit comfortably inside the distribution the model is built on.. Not a power-play indictment. Ottawa went 0-for-3 for 0.99 xG. That is 0.33 xG per opportunity , slightly above the league average. The power play generated chances; it did not finish them, and on one of them it conceded a shorthanded breakaway. The shorthanded goal is the kind of event that adds variance to any single-game read.. Not a one-game series. Anything in this post is about Game 4 specifically. The series had its own arc: a Montréal blowout in Game 2, an Ottawa late-winner in Game 3, two close games at Place Bell and a split in Ottawa. The cumulative story across four games is closer than the 3-1 scoreline suggests and tighter than this Game 4 shot-quality gap implies. 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.. Win probability : the in-game model behind the trace updates after every event from score, time remaining, and game state. The shot map and cumulative-xG charts read straight from the play-by-play.. Numbers regenerate on every build. A PBP re-sync or xG recalibration reshapes every figure and every chart 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/321 . The 2026 Walter Cup bracket, including the four-game series view, lives at /playoffs/bracket .