Goalkeepers have no business existing in soccer. Maintaining an entire class of players whose job is to snuff the joy out of a game is absurd, and strapping on gloves to play a sport that prohibits the use of the hands is blasphemy. In defiance of not just strikers, but more or less everyone, the keeper believes there is too much scoring in soccer. They will gladly blot out the most beautiful art that the opposing team can produce, without so much as an apology. They are not sorry, and given the chance, they will do it again. There is no better position in all of sports.
In the 147 years since the position was codified, soccer has never really figured out what to do with the goalkeeper. They are the subjects of both intense scrutiny and an atmosphere of outsider disdain, like the result of some unethical American football experiment to combine a quarterback and a field goal kicker. In the best cases their services are almost unnecessary. In the worst, their mistakes play out under a glaring spotlight where they are punished at an unforgiving rate. Analysis of the position is somehow both unremitting and aggressively superficial. Pundits are confident declaring when a save was spectacular and when You’ve Got To Do Better Than That! but for the most part their conclusions don’t go much deeper than what you can already see on the screen. Further evidence or examination is all too rare.
The position’s function, and whether or not it’s being performed well, has never been easily explained. Even before the last few decades, a goalkeeper’s job description was sprawling. Obviously, stop shots, but also communicate threats, organize the defensive shape, position set pieces, intercept crosses, gobble-up through balls, manage rebounds and scrambles, distribute from the back, and whatever else a specific system demands. Quantifying all of that has always been difficult, and modern changes have complicated as much as they have clarified.
First came the addition of the back-pass rule in 1992 (and its throw-in addendum in 1997). The bylaw barring keepers from picking up the ball when it was played back to them from their own team’s feet was an effort to curtail rampant time wasting and make the game more entertaining. Like basketball’s three-point line and the 24-second clock before it, it was (after an awkward introduction) hugely successful. And like those rules, it also fundamentally altered the sport. The pace of the game picked up, teams were given real incentive to press defenses, and keepers found themselves playing the ball with their feet more than ever before. Goalkeepers with trustworthy footskills suddenly became a valuable commodity.
With the baseline outfield skill level of goalkeepers rising, coaches began to rethink what was possible from the position. The patron saints of this trend, managers like Pep Guardiola and athletes capable of playing sweeper-keeper like Manuel Neuer, pushed the boundaries and expectations for everyone. The ability to build and support the attack with your feet, and to patrol the defensive space behind the back line have grown from unique bonuses to yet more basic requirements for a top level keeper. At the same time, and inseparable from this evolution, came the arrival of advanced analytics, but thus far the sea change has been slow to unlock the Rosetta Stones of goalkeeping analysis.
As with every major sport in the 21st century, soccer is being shaped by our ability to collect and analyze data. But the fluid nature of the game means that it’s even harder to break down into raw numbers than something more regimented like baseball, and the transition has not been easy. What started as “Here’s a quirky stat about how far this player has run during the game; isn’t that fun?” has gradually made way for more actionable location heat maps, passing vectors, and complex metrics like expected goals (xG). These ideas have run into the same growing pains and pushback from purist curmudgeons that have been encountered everywhere else, and they are far from settled science. The proponents of xG, for example, readily admit that it’s an imperfect measure that’s still being refined, and there are a whole list of competing methods and algorithms that go into its determinations. But as far as they still have to go, these advances have already been major steps in our ability to understand the game.
It’s not that these innovations have left behind goalkeepers entirely—xG has useful applications, as do other emerging metrics; people running systems like Guardiola’s are finally interested in information like a goalie’s passing success rate—but as with everything else in the game, netminders have sometimes seemed like an inconvenient afterthought.
For an illustration of how hard it is to quantify goalkeeping and how, now as ever, the position is treated as an outlier, we can use the example of the gold-standard American goalkeeping performance: Tim Howard at the 2014 World Cup against Belgium.
Howard’s cornucopia of saves, allegedly 15 of them, were a FIFA World Cup record. To begin with, though, there was not even consensus on how many saves Howard had made. FIFA initially listed 16 and then corrected the number to 15. Deadspin’s Billy Haisley reported seeing numbers ranging from 12 to 17 after the match.
This is not the easiest thing to confirm, as there is no mention of the accomplishment on the official FIFA page that lists the tournament scoring records. In fact, goalkeeping statistics, even the mundane, fairly useless ones, are hard to come by. If you visit an advanced statistics site WhoScored?, you’ll find a page listing for Tim Howard, but quickly notice that there are no dedicated categories for goalkeeping metrics, and even basic numbers like saves or crosses caught/punched aren’t readily available. MLS and ESPN versions will at least get you saves, but not much beyond that. Some sites, like the Premier League’s, are a little better, but even fantasy scoring systems for keepers are remarkably simplistic. (Squawka is the exception to the rule and has some slick features.)
Even when those numbers are accessible, the reactive nature of much of what goalkeepers do renders many of them all but meaningless for comparisons. A goalkeeper will never have an equivalent statline to something like Kobe Bryant’s final game, where he scored 60 points on 22 of 50 shooting, because a goalkeeper doesn’t have the agency to make a decision equal to chucking up 50 shots. Besides the fact that he played wonderfully, the reason that Howard was able to make 15 saves was that Belgium took a whopping 38 shots. This could only happen because the U.S. team was seriously outclassed by the Belgians, who had one of the most impressive attacking units in the tournament. This inability to affect many of the game’s circumstances makes stats like saves or goals against even less helpful than looking at a baseball pitcher’s wins, and a goalkeeper’s save percentage, when you can find it, isn’t much better.
Finally, in part because the lack of technical attention all around, the average viewer tends to be lacking, even with just eye tests, at determining the quality of any individual save. Several of the point blank kick saves that Howard made were top class—set up by precise positioning and timing, and executed with solid technique—but earned shoulder shrugs by appearing less than spectacular. Some of the lauded, more acrobatic saves he made late in the second half were closer to routine. In these cases announcers largely fall back on exclamations and effusive praise rather than any detailing of technique that would help audiences differentiate. Again, tools like xG may be able to help improve this situation, but they only touch on a portion of the job and are hardly a panacea.
This last weekend, Wolverhampton’s Rui Patrício made a beautiful leaping save to deny a long distance strike from Man City’s Raheem Sterling.
The shot was from an well outside the box and came from the normal run of play, while Sterling was relatively well defended. Depending on methodology, the xG number on such a shot is likely to be very low—shots from this general location score at a rate of something like 2 percent. However, Sterling hit it perfectly, swerving the ball severely with the outside of his foot into the top corner with pace, and any keeper in the world would be happy to have saved it. (It’s worth adding that the extent of the technical analysis provided by the booth was a commentator pointing out that Patrício almost “went with the wrong hand there,” which is such a persistent misconception that it’s become a trope, and was completely wrong in this case). We’re left without much explanation of why this is a good save, except that when people see it, it tends to make them yell superlatives, which I suppose is as good as most.
In the face of all this, we’re going to try to help. Goalkeepers certainly deserve better, and even if they don’t they aren’t going anywhere, regardless, so we’re launching this periodic column to highlight the position, give preternatural stops and heartbreaking blunders the attention they deserve, and hopefully provide at least a little bit of technical analysis that’s so sorely missing from the conversation. The timing and format will be fluid; depending on how things go we may be back frequently with montages of diving saves, check in every few months with a discussion on the finer points of how someone wound up nutmegged, or blog daily poetry about Rogério Ceni. Above all, we’ll be rooting for 0-0. Welcome to None Shall Pass, Deadspin’s home for words about goalkeeping.
I’ll leave you with two saves from this week that your life will be richer for seeing.
The first is Aston Villa’s Ørjan Nyland coming up with an unreasonable double save against Reading in the EFL Championship second division:
The second is U.S. National Team keeper Ashlyn Harris stoning all-universe scorer Sam Kerr one-on-one in a losing effort for the Orlando Pride vs. the Chicago Red Stars in the NWSL.
If you have tips, questions, topics you’d like covered, or you find video of someone in a beer league game doing a scorpion kick save, send them to [email protected].