clock menu more-arrow no yes

Filed under:

Fancy stats agree: Stefan Frei is having a landmark season

New, 28 comments

If there’s any justice, Frei will be leading the discussion for Goalkeeper of the Year, writes our guest columnist from American Soccer Analysis.

Max Aquino / Sounder at Heart

A conditional draft pick was all that it took to obtain Stefan Frei. It seems like robbery now, particularly when you follow that draft pick’s meandering journey around four different teams before winding up as Columbus’ 2015 first-round selection of Jamaican defender Sergio Campbell. At the time, nobody could have known that for the price of a throwaway SuperDraft pick, Toronto had traded Seattle the man who would almost single-handedly deliver an MLS Cup to the Pacific Northwest at the Canadian team’s expense.

To say that Stefan Frei is a popular figure amongst Sounders’ fans would be an exercise of understatement in the extreme. Sounders fans have always enjoyed excellent goalkeepers. Kasey Keller, Marcus Hahnemann, and Michael Gspurning have all stood between the sticks at CenturyLink, and while all have performed admirably, in 2018 Stefan Frei has set himself apart from that pack and the league as a whole.

To be a goalkeeper is to be judged. Somehow this position seemingly manages to be the easiest for armchair observers to evaluate and the most difficult for analysts to quantify. Choosing the league’s best goalkeeper seems like it should be easy. If we zoom out far enough, the position would appear to be rather binary. Stop the ball from going in: good. Let the ball go in: bad.

However, we all know it’s not that simple. Saves don’t tell us the whole story, nor do clean sheets. A goalkeeper’s numbers can be betrayed by an absent-minded centerback, or an ineffectual tactical plan. Indeed, a goalkeeper can be made to look prolific when they are merely adequate. A goalkeeper that saves 50% of 100 shots has more saves than a goalkeeper that saves 80% of 50. In that same vein, is a goalkeeper with a save percentage of 75% better than a goalkeeper with a save percentage of 60%? Intuitively it would seem so, but what if Goalkeeper A has mostly saved shots from outside the box while Goalkeeper B has come up big more often when the shooter is within the penalty area? These are tough questions to answer; without context it’s nearly impossible to look at a spreadsheet of statistics and point to the best keeper.

So... then what? The eyeball test is always an option, but I don’t have to tell you how swayed our eyes and minds are by the outliers. Stefan Frei has made 100 saves so far this season. If I close my eyes I can think of a handful. The really good ones. The highlight reel stuff. Likewise, without even digging, I can think of four or five mistakes he’s made over the years. Those tend to be glaring. Even more so than the saves, those rare moments — indecision, a lack of concentration, a misread, or a bad bounce — stick in my mind in perfect HD clarity. It’s odd that we tend to think of goalkeepers in these terms. We remember the outstanding on either side of the coin, the outliers, and then shrug at the vast majority of their body of work.

One of the metrics that’s starting to gain a lot of traction this season, is GA-xGA. For those who aren’t familiar with the concept of xG, I’m going to attempt a nice “tl;dr” style explanation in hopes that you don’t immediately hit the back button on your browser the second you eyeball that lowercase x. Expected goals can simply be explained thusly:

Using historical data we can calculate the likelihood of a goal being scored from a shot based on its location and the type of shot (headers have a lower xG than kicked shots). Furthermore, you’ll see the term “post-shot xG” below. This is just an expected goals model that takes a shot’s placement in the goalmouth into account as well.

In order to to apply this same data to goalkeeping/defensive performance, we can determine from an opponent’s xG how many goals a team could expect to allow — expected goals against. So far this season, Seattle has an xGA of 45.4, while in reality they have only allowed 32 goals. That’s great news for Sounders fans, and it may give us a some insight to how well Stefan Frei is doing his job.

Traditional metrics can only get us so far. The eye test can provide some direction, but it’s often unreliable. Advanced analytics get us closer to our goal, but not quite all the way there. It would seem our best course of action would be to use all of the data that’s available to us. Let’s take a look:

Stats via @analysisevolved and @etmckinley

It doesn’t take much more than a brief glance at this table to start banking some points for Stefan Frei. He’s tied for the league’s best GAA, leads the league in save percentage, and absolutely crushes the competition in regards to GA-xGA.

If there’s any data here that might temper our praise a bit, it’s only that the Sounders defensively have done a decent job of limiting high-leverage shots, and shots within the six-yard box and penalty area specifically. All this means is that if we’re being as critical and reductive as possible, it’s true that Frei could have faced a more difficult barrage of shots. We can use xGA here to get a good sense of that. Evan Bush ranks highest here, and that’s no surprise given Montreal’s early season dysfunction. Tim Howard, Andrew Tarbell, Joe Bendik, and David Bingham’s inclusion in the top five is also no surprise given defensive awkwardness each has dealt with. Attinella sports the lowest xGA of anyone on our list, but that has more to do with the fact he has 700-800 less minutes than most other first-choice keepers on the list.

So yeah, with a worse defense in front of him, it’s likely that Frei would have had a more difficult season, but divorcing goalkeepers from their backline feels disingenuous. A lot of analysts like to praise goalkeepers for their ability to “organize a defense”, and I think there’s probably something to that even if we don’t know much about the details of that beyond when we see a GK set a wall or point out potential runners in the lead up to a corner kick. Even if we accept that Frei hasn’t had the hardest GK assignment in the league this year, his -11.74 xG-xGA means he’s handling whatever assignments come his way more efficiently than anybody else; and really, isn’t that all you can ask?

If you feel like I’m leaning on that particular metric heavily, you’re not wrong. Given all the data that we have access to, I personally think that GA-xGA gives us the best clue in discovering what it is we actually want to know, which is how well a GK does at preventing goals. Frei has conceded a number of goals this season that nobody would ever blame him for. Conversely he’s also saved shots that most other GKs would have let slip past them. Sometimes a GK can be completely exposed, and I think it’s what we see a GK do in these situations that can really help us determine their relative value. Having a huge hand in the Sounders outperforming their xGA metric by almost 12 goals (so far) suggests that Frei is a very valuable player indeed.

So how does Stefan Frei compare to GK’s of recent history?

Visualization via @thedummyrun

Not badly at all. Tim Melia is an excellent goalkeeper, and I would certainly expect him to continue to put up solid numbers, but it’s important to recognize that Frei has played about 4,000 more minutes since 2014. Of GKs with comparable minutes (Rimando, Robles, Bingham, Bush), Frei sitting solidly with the second-best save efficiency is more than respectable. Also if you look at the right margin of this graphic, you can see that other than an unusual bump in 2015, Frei is steadily trending upwards and this year sits alone atop the pile. If this season’s numbers hold, Frei’s performance will rank as one of the best statistical seasons since GA-xGA data became available in 2011.

Frei has long been on the fringes of the annual “Goalkeeper of the Year” conversation. This year (barring a late-season collapse), he should not just politely interrupt but start yelling over everybody else.

You’ll notice I’ve glossed over the two big mistakes. That was not an oversight, but rather an honest assessment of how heavily they weigh in my appraisal of Frei’s season. Goalkeepers are going to drop shots they should have saved, get hung out to dry by defenders, or suffer a momentary lapse in concentration. Points will be lost on the back of a GK, and that’s going to be true for Stefan Frei and for every future GK until the polar ice caps melt. When you only look at the outliers you get an incomplete picture of a goalkeeper. Hopefully this fills in those blanks a little more. While it’s only natural and understandable to feel frustration after seeing your team take an L for an unbelievable mistake, I think it’s impossible to deny that compared to other keepers around the league, Seattle simply could not have asked for a better one this season.