With the PWHL now back in full swing after the Olympic break, it’s a good time to look back at the first half of the season and where teams may want to go from here. Let’s look at the net, exploring PWHL goaltending performance this season.

Aerin Frankel (Boston Fleet) and Gwyneth Philips (Ottawa Charge) have returned from having contributed to the US team’s record-breaking 352 minutes and 17 seconds without allowing a goal, the longest such streak in Olympic hockey history. Frankel and Canada’s Ann-Renée Desbiens (Montréal Victoire) both made 30+ saves in the gold medal game.

Goaltending was also the story in both the 2024 and 2025 PWHL Walter Cup finals, and with just 4 playoff spots available, it’s going to be important if teams want to have a shot at the playoffs this year.

Looking deeper than traditional statistics

To look at the PWHL goaltending performance so far this year, I wanted to dig a little deeper than just goals-against average (GAA) and save percentage (SV%). One way I decided to do this is to calculate separate save percentages for when the goalie’s team is ahead, and when they are trailing.

To do this, I used Lars Skytte’s web scraping tool for PWHL play-by-play data, available here. For goalies that had faced at least 100 shots by the Olympic break in the 2025–2026 season, I have plotted their SV% when their team is behind against their SV% when their team is ahead (data are for all games this season before the Olympic break).

For reference, I have also plotted the league average SV% when ahead as a vertical line and the league average SV% when behind as a horizontal line, dividing the plot into four quadrants. The league average SV% is higher when the goalie’s team is ahead compared to when they are trailing (0.932 vs 0.923).

Goalies with above-average performance

In the top right quadrant, we find goalies that are above-average, whether they are leading or trailing. Not many surprises here: Desbiens and Frankel are both exceptional, and Nicole Hensley (Minnesota Frost) has had a fantastic year so far. Emerance Maschmeyer (Vancouver Goldeneyes) is the only goalie in this quadrant whose team has a losing record, but despite the Goldeneyes’ early-season struggles, she’s been a key piece in 4 of the team’s 6 wins.

PWHL Goaltenders with differing performances based on game state

Moving clockwise to the lower right quadrant, we find goalies that perform above average when their team is ahead, but below average when their team is behind. This is the quadrant I’m calling “streak city” – goalies that are solidly in this quadrant tend to match how their team is performing. They may not be as able to rebound from a low, but they’re unlikely to give up a lead. Kayle Osborne (New York Sirens) and Kristen Campbell (Vancouver Goldeneyes) are the current members of this quadrant, though Osborne is very close to making it into the top right quadrant.

Goalies in this quadrant are unlikely to be the make-or-break factor in a win or a loss, so teams may want to focus on offence to ensure their goalies can get into their preferred position of defending a lead.

In the top left, we find goalies that perform better than average when they are behind, and below average when they are ahead. Philips is the only member of this quadrant, which I’ve taken to calling the “learning from their mistakes” or “motivated by spite” quadrant. The opposite of streak city, goalies in this quadrant are able to keep their team in it when trailing, but are more likely to give up a lead.

The Ottawa Charge were undefeated in overtime and shootouts before the Olympic break, so Philips’ ability to hang in and give her team a chance to get there is a benefit for them. However, the 3-2-1-0 points system with regulation wins as the tiebreaker may end up biting them when it comes down to making the playoffs. In the second half of the season, Ottawa may look to improve their defence when leading to make sure they’re not taking more games to overtime than they need to.

PWHL Goalies with below-average performance

Continuing clockwise to the bottom left quadrant, we arrive at goalies who are struggling a little bit. It says something about the quality of goaltending in the league that every goalie in this quadrant is close to the edge of another; no goalie is much below average in all situations.

This also seems to be a good moment to remind ourselves that this plot has not accounted for shot quality, and these goalies may be the victims of their defence’s positioning. These goalies’ teams may need to look to shoring up defence for the second half of the season if they want to have a good chance at making the playoffs, especially Toronto and Seattle, who find both their main goalies in this quadrant.

Data considerations

With all that said, this analysis method does have some caveats. A save when trailing by 3 goals is treated the same as a save when trailing by 1 goal, despite those situations being very different. Since we use save percentage here, we have all of the usual drawbacks of that statistic (namely, not accounting for shot quality).

It also ignores situations where the game is tied, which vary in intensity from the first shots of the first period to moments in overtime with the game on the line. Multiple teams have fantastic, Olympic-calibre goalies on their roster, and it shows in the data. But PWHL teams can’t win games on goaltending performance alone, and teams with goalies that need more defensive support will need to ensure that their team composition and play structure give them the best chance they can get just to make the playoffs, let alone succeed once they get there.