clock menu more-arrow no yes mobile

Filed under:

The Philipp Grupower of Positive Regression

A look at Philipp Grubauer’s outlier season — and why nobody needs to worry about it

If you buy something from an SB Nation link, Vox Media may earn a commission. See our ethics statement.

San Jose Sharks v Seattle Kraken Photo by Abbie Parr/2022 NHLI via Getty Images

Philipp Grubauer had a season to forget in 2021-22. So did the guys in the goalie room with him — Chris Driedger and occasionally Joey Daccord. The Kraken, as a whole, finished 24th in goals against, despite allowing fewer shots than all but three teams in the NHL. There were a whole host of other issues in their inaugural season — special teams and goal-scoring near the forefront — but the disappointing year in net hurts a bit more because that’s where Seattle was supposed to be good. Here’s what Dom Luszczyszyn had to say in his season preview:

Seattle is different [from Vegas] in the sense that good goaltending will not be a surprise, it’s an expectation. The Kraken made a very savvy move at the draft nabbing Chris Driedger from Florida and only added to those riches in free agency by adding Vezina trophy finalist Philipp Grubauer to the tandem. Together, they’re expected to be the third-best tandem in hockey with Grubauer grading out as the 13th most efficient goalie and Driedger in seventh.

As we inch closer to the 2022-23 season and are inundated with previews and predictions, you’ll likely encounter a general vibe that the Kraken should be better in large part because the goaltending should be better. And it’s true, it should be better — even though they’re bringing back the same starter and replacing his backup with Martin Jones. But why should we expect it to be better? It’s all about regression to the mean.

Regression towards the mean is a statistical concept that tells us that if we encounter one outlier data point, our next data point is more likely to be closer to the average of the data rather than the outlier point. And whoa boy did our dearest Grubert have himself an outlier data point last season.

A chart showing goalies’ save percentages in a single season compared to their career averages. Grubauer’s 2021-22 is far below his norm

Philipp Grubauer came into the year with a .920 career save percentage through 214 games of NHL action. Looking at only goalies with a well-established NHL career (200+ games), Grubauer’s drop from his career .920 save percentage down to a season save percentage of .899 was the seventh largest drop-off in the last 30 years. Four of the players with a bigger dip (Jonas Hiller, Grant Fuhr, Miikka Kiprusoff, and Jimmy Howard) were all at the end of their careers and retired the following offseason. That makes Grubauer’s year one of the wildest statistical oddities in recent hockey memory. (Hang the banner?)

Knowing just how significantly different this past year was for Grubauer, it’s a safe assumption that he’ll improve at least a little in his second season in Seattle. An .889 save percentage is not a repeatable feat for an NHL-caliber goaltender. But just how much should we expect him to improve? Will he get back to that .920 he came into the year with? Let’s look at the most comparable outliers to Grubauer’s 2021-22 (aside from the guys that retired immediately after) and see how they did in their following season.

A chart showing the biggest dropoffs in season save percentage relative to a goalie’s career save percentage prior to the season. Nearly every goalie bounces back closer to their career average following a down year.

Every single one of them aside from Niklas Bäckström (no, you’re thinking of Nicklas, this one is the goalie) improved the following season, by an average of .018 points. And while Bäckström is included here since he didn’t immediately retire following his down year, it should be noted that he only appeared in four games the following season before moving back to Finland to play in the SM-Liiga.

A handful of these guys weren’t quite retirement age, but they were fast approaching it. Isolating the players in Grubauer’s age range (we’ll take ages 29-31) and history bodes even more well for our boy. These players averaged an improvement of .022 points the following season. None of them aside from Cam Talbot got all the way back to career average, but every one of them ended up much closer towards it.

The best comparisons we can make to Grubauer from this chart are probably those of Evgeni Nabokov and José Théodore. Théodore was a former Vezina & Hart Trophy winner who even received a couple Hart Trophy votes in the season immediately prior to his down year. After the 2005 lockout, he came back to Montreal and posted a dismal .881 save percentage before being traded to Colorado at the trade deadline. He didn’t find his footing right away in his first full season in Colorado, as the chart above shows, but he did still improve. He’d go on to play another nine seasons after that, finishing his career with a .909 save percentage and being one of just three players in hockey history with a Hart Trophy to not be elected to the Hall of Fame.

Nabokov might be an even better comparison. A former Calder Trophy winner, Nabokov also received Vezina votes in three separate seasons before his down year, also the 2005-06 season that followed the lockout. He came back extremely strong for the Sharks after that year though, getting right back to his career average numbers and posting 7 shutouts in 2006-07. He’d go on to start 77 games in 2007-08, finishing second in Vezina voting, and ended his 14-year career with a .911 save percentage.

Philipp Grubauer’s 2021-22 season was downright bad. But history tells us that it’s not entirely unique, and in all likelihood it won’t repeat itself. And if the goaltending is going to get that much better in 2022-23, then the ceiling for the Kraken just got way higher.