The Best and Worst NHL Teams for Goalies: Boston Helps Most, Philadelphia Hurts Most

How we rated each team

Team save % is not the answer. A team with a great goalie has a great save %, and that says nothing about the team. The question is whether a team makes the same goalie better or worse.

So we compared goalies with themselves. For each franchise, we took every goalie who played there and somewhere else, and asked how their save % at that team compared with their save % everywhere else. Our model does this for all goalies and teams at once. It accounts for each goalie’s own level and yearly form, so a team gets no credit for signing a good goalie and no blame for signing a bad one.

The data is every NHL regular-season goalie game since 1955-56 (127,409 of them), from the NHL’s game logs. Ratings are in points of save %: 1 point is .001, so +4 turns a .905 goalie into a .909 goalie. Each rating has an error of about ±2 points, so only the top and bottom of the table stand clearly apart from average.

The ranking: 2016-17 to 2025-26

The spread between the best and worst teams for goalies is about 9 points of save %. For a starter facing 1,500 shots, that is roughly 14 goals a season. It is real, but it is smaller than one season of luck, which moves a starter’s save % by about 7.5 points either way.

Boston: the one team that stayed on top

Of the 11 goalies who played for Boston and elsewhere since 2005, 8 stopped more pucks as Bruins, by 7.3 points on average (90% range 1.4 to 12.4). Linus Ullmark was 19.3 points above league average in Boston, where he won the 2023 Vezina, and 2.1 above on his other teams (Buffalo and Ottawa). Tim Thomas was 18 points better in Boston than elsewhere, Jaroslav Halak 6 and Chad Johnson 20. Boston was 2nd in 2005-16 and 1st in 2016-26. No other team stayed in the top two across both decades.

A good team is not the same as a good team for goalies

These ratings measure how a team’s goalies do per shot, compared with the same goalies elsewhere. That is different from being a good team. A team can control the puck, outshoot everyone and still leave its goalie facing breakaways and backdoor passes. Another can spend most nights in its own end and still give its goalie shots that are seen cleanly.

To check, we compared the goalie rating with shot share, the team’s share of all shots in its games. Shot share measures how much the skaters control play, and the goalie has little to do with it. Over 2016-17 to 2025-26, the two correlate at only 0.22 across the 32 teams (90% range 0.00 to 0.45). Good teams lean slightly toward being good for goalies, but most of the goalie rating is something else.

Philadelphia is the extreme case. It has outshot its opponents slightly over the decade, 10th in the league, and still ranks last for goalies. Its goalie environment cost about 13 goals a season, roughly two wins. The Islanders are the reverse: 24th in shot share, yet their goalies saved about 10 goals a season more than the same goalies would on an average team.

Edmonton and Winnipeg, season by season

Edmonton has been a good team for most of the McDavid era and below average for goalies in every one of those seasons. From 2019-20 to 2024-25 it never finished outside the top 12 in points % and never better than 20th for goalies. Winnipeg has been mid-table for most of the decade (12th in points % overall), yet it was top 7 for goalies in every season but one, and 1st in both 2023-24 and 2024-25. In 2024-25 it also had the best points % in the league.

In goals, the gap is large. Winnipeg’s goalie environment was worth about 13 goals a season in 2023-24 and 2024-25. Edmonton’s cost about 4 goals a season on average, and almost 7 in 2025-26. Put the same goalie on each team and that is a swing of more than 15 goals.

Edmonton: a goalie graveyard for 20 years

Oilers fans have watched a long line of goalies arrive, struggle and then play well somewhere else. The data backs them up. Since 2005, the same goalies stopped 4.1 points less in Edmonton than on their other teams (90% range −8.4 to 0.0), and only 6 of 17 did better as Oilers. Edmonton’s rating has been below zero entering every season since 2006, through bad Oilers teams and good ones.

Dubnyk is the clearest case. He left Edmonton after a terrible 2013-14 and went from 22.8 points below league average that season to 21.4 above in Minnesota the next. Brossoit was 23 points better on his other teams, and Campbell 26.

It was not always like this. In 1983-93, with Grant Fuhr and Bill Ranford in net, Edmonton ranked 7th of 24 as a team for goalies. It fell to 28th of 30 in 2005-16 and has recovered only to 23rd. In the expected-goals era (2017-18 on), Edmonton is below average on both parts: it allows slightly harder shots than average (−1.8 points), and its goalies fall short of xG (−1.0).

Winnipeg yes, Florida and Carolina no

Winnipeg is a good team for goalies, but only since it became the Jets. The franchise’s Atlanta years (1999-2011) were bad: 30th of 30 in the 1993-2005 window and 23rd in 2005-16. As the Jets, it climbed to 5th in 2016-26, and its rating entering 2025-26 (+6.1) was the best in the league. Eric Comrie was 28 points better in Winnipeg than elsewhere, Laurent Brossoit 17 and Michael Hutchinson 15. Connor Hellebuyck has only ever played for Winnipeg, so he tells us nothing either way. In the xG era, Winnipeg’s goalies beat xG by more than any other team’s (+3.1 points).

Florida is not a good team for goalies on this measure: −0.3 in 2016-26, and only 9 of 20 goalies did better there since 2005. Sergei Bobrovsky has been 7.2 points worse in Florida than on his other teams, even with two Stanley Cups there. Chris Driedger, Anthony Stolarz and Alex Lyon did well there; Jacob Markstrom, Tim Thomas and Sam Montembeault did not.

Carolina is −0.7, and 8 of 17 goalies did better there. Alex Nedeljkovic was 22 points better as a Hurricane, Scott Darling 33 worse and Antti Raanta 12 worse. Carolina may look like a good team for goalies because it allows so few shots, which keeps goals against down. But the shots that get through are hard: per unblocked shot, Carolina allows the third-toughest shots in the league. Its goalies beat xG by 1.4 points, and that roughly cancels out.

History: teams rise and fall for goalies

We fitted the same model separately for five eras. Being good for goalies turns out to be a phase, not part of a franchise’s identity.

• Philadelphia has the famous “goalie graveyard” reputation, and right now it earns it. It was not always so: the Flyers were 3rd of 21 in 1967-83 (the Bernie Parent years) and 1st of 24 in 1983-93 (the Ron Hextall years). Since 2016 they are last. Bobrovsky, Ilya Bryzgalov, Brian Elliott, Petr Mrazek and Anthony Stolarz all did worse there than elsewhere. The exception is recent: Dan Vladar was 10.2 points above average in 2025-26, his first Flyers season.
• The Islanders were the best team for goalies of 1967-83 (Billy Smith and Chico Resch on the dynasty teams), then 25th and 27th for two decades. Their rating turned positive right after 2018-19, Barry Trotz’s first season as coach. Robin Lehner was 17 points better there and Semyon Varlamov 7.
• Montreal was 2nd in 1967-83 and stayed in the top seven until 2016. It has been 24th since.
• Detroit was last in 1967-83, near average from 1993 to 2016, and 30th in 2016-26. Ville Husso, Alex Nedeljkovic and Cam Talbot all did worse as Red Wings.
• Tampa Bay was 29th and 30th from 1993 to 2016, a graveyard before Andrei Vasilevskiy. Dwayne Roloson, Dan Ellis and Marc Denis were each 17 to 25 points worse there.

Across all teams, ratings from 2005-16 and 2016-26 correlate at −0.12 (90% range −0.42 to +0.16). A team’s goalie rating from one decade tells you nothing about the next. Within a decade it holds up better: the lasting part of a team’s effect is about as big as its year-to-year swings (3.2 vs 3.1 points in 2016-26). The likeliest reason is that the rating follows the coach, the system and the defence corps, and all of those turn over.

Why some teams help their goalies

Since 2017-18 we can split each team’s effect in two using our expected-goals (xG) model. Shot quality is how hard the team’s unblocked shots against are, by location, type and rebounds. Goalie results beyond xG is how the same goalies did against those shots at this team compared with other teams. Both are in points per unblocked shot; higher is better for the goalie.

Shot quality is the steadier of the two: a team’s shot quality repeats at 0.55 year to year, while its goalie results beyond xG repeat at 0.27. Minnesota allows the easiest shots in the league.

The part beyond xG is harder to explain, and it is not small: 2.5 points of spread between teams, about as large as the goalies’ own spread in this model. xG models see where a shot came from, but not the screen, the pass before it or how fast the rush was. Defences differ on all three. Goalie coaching could matter too, but this data cannot separate it from defence.

Using it: fantasy and betting

When a goalie joins a new team, about 60% of that team’s current rating shows up in the goalie’s save % (our holdout test from 1980 on). Recent seasons matter more than franchise history. The full breakdown of good-to-bad and bad-to-good moves is in what happens when a goalie changes teams.

For betting, we took each team’s rating as of the start of the season and checked it against 2024-25 and 2025-26 closing moneylines. Backing the teams with the biggest goalie-rating edge in a game lost 6.7% (1,308 bets; 90% range −11.5% to −2.1%). Betting against them lost 1.1%, and that result flipped from −4.8% to +4.2% between the two seasons. On totals, overs between two bad teams for goalies and unders between two good ones each lost 4.1%. The market already prices this.

What this cannot tell you

• Ratings are rough. About ±2 points of error per team means teams 8th to 25th are hard to tell apart. Only the ends of the table are solid.
• Franchise goalies are invisible. Henrik Lundqvist, Andrei Vasilevskiy, Connor Hellebuyck and Carey Price played most of their careers for one team, so they add little to the comparison. The ratings lean on backups and goalies who moved.
• Teams choose their goalies. The model accounts for goalie level, but a team that keeps a backup because he fits the system will look a little better than it is.
• Before 1983-84, shot counts are less reliable, so read the 1967-83 ratings loosely.

Show the work

Frequently asked questions

More in Speed, faceoffs, lines and goalies

• Verdict: Promising

  Do Faster NHL Teams Win More? What NHL EDGE Speed Data Says

• Verdict: Myth

  Does Playoff Hockey Reward Hitters Over Speed? We Tested Five Postseasons

• Verdict: Mixed

  NHL Faceoff Matchups: Ignore the Head-to-Head, Watch the Hands

• Verdict: Mixed

  NHL DFS Stacking Works. “Chemistry” Is Mostly Noise.

• Verdict: Mixed

  How Much of a Goalie’s Save Percentage Is the Goalie?

• Verdict: Mixed

  When a Good Goalie Joins a Bad Team for Goalies, Save % Drops About .005. The Team Explains Part of It, and Luck Plays a Part

All Prop Lab research →

Keep reading

• Good goalie, bad team for goalies →What happens to save % when a goalie moves between good and bad teams for goalies, and whether it is bettable.
• Is save percentage a team stat? →What 70 years of goalie moves say about goalie skill, team defence and luck.
• Betting NHL teams on back-to-backs →How much tired teams lose, the backup-goalie split, and whether the price already knows.
• NHL Prop Lab: hockey betting edges and myths →The full hockey research hub: what survived testing, what failed, and what we are tracking live.

Sources and method

1. NHL stats (NHL.com)
2. NHL API reference (community documentation)
3. The Odds API (historical NHL moneylines, puck lines and totals)
4. Empirical Bayes shrinkage
5. Responsible Gambling Council (Canada)
6. ConnexOntario (1-866-531-2600)
7. National Council on Problem Gambling (1-800-GAMBLER)

Research and analysis, not betting advice. Past results do not guarantee future returns; bet only what you can afford to lose.