On Feb. 7, in the most important game of the Premier League season, Liverpool‘s Alisson, generally regarded as one of the best keepers in the world, handed a game, and potentially the league title, to Manchester City.
First, he sent a pass directly to City’s Phil Foden, who assisted for Ilkay Gundogan for the go-ahead goal in the 73rd minute; just three minutes later, he did almost exactly the same thing, only with Bernardo Silva as the thief and Raheem Sterling as the scorer.
It got worse this past weekend. In Germany, Bayer Leverkusen‘s Niklas Lomb fielded a simple back pass… or tried to, anyway. He whiffed, and a charging Florian Niederlechner put Augsburg ahead; the match ended in a 1-1 draw, and Leverkusen fell five points back from a Champions League spot. The same thing had happened to Marseille just two days earlier: keeper Steve Mandanda whiffed, Nantes‘ Ludovic Blas scored and a favored Marseille dropped points.
These moments brought to mind a panel at the 2019 Sloan Sports Analytics Conference, in which Daryl Morey, current Philadelphia 76ers president and soccer-curious nerd, said to StatsBomb CEO Ted Knutson, “[passing the ball back to the goalkeeper] is absolute insanity.”
He explained his position. “Let’s say you assume you gain some small advantage of possession. You basically then need to connect like 50 passes to the front to get to an efficient spot. But then it gets even worse: the managers are choosing the goalies based on… their feet skills.”
To Morey, the purpose of a goalkeeper was getting muddled. “I would want a guy that can stop a ball that some other goalie cannot stop… The risk of passing it back is so, so high relative to the reward that it doesn’t make any sense.” To him, the downside — a turnover in an extremely vulnerable position — was not worth the marginal upside.
Anecdotally, this seems obvious. The disasters are disastrous. But if teams led by the best, most successful managers in the sport keep playing the ball to their keeper’s feet, is there really a problem?
Playing out from the back is why teams press
Let’s begin exploring this by laying out some basic data:
– On average, teams in Europe’s Big 5 leagues this season average 94.8 possessions per match and begin 7.5% of them, about 7.1 per game, in the attacking third. They score about once every 32 times on these, or about 0.03 goals per possession. If they start a possession here because of a ball recovery, defensive action, dispossession or take-on, the odds increase to 0.04 goals per possession, or one in 23.
– Teams start 38.8% of their possessions, 36.8 per game, in the middle third and score once every 61 times, or about 0.02 goals per possession.
– Teams start 53.7% of their possessions, 50.9 per game, in their defending third and score once every 89 times, or 0.01 goals per possession.
This data alone tells you why teams are getting more aggressive at chasing possession in advanced areas. You’re almost three times more likely to score when you start the ball in the attacking third, and in a sport in which teams average about 1.4 goals per game, improving your odds of scoring however slightly can make a massive difference.
It’s no coincidence that the 14 teams from the “Big Five” leagues averaging at least two goals per match all start at least 7.7% of their possessions in the attacking third, and five are over 9% (AC Milan, Bayern Munich, Inter Milan, Manchester City, PSG: three of these teams are top of their respective leagues, too). Of the 21 teams averaging under a goal per match, only two start more than 8% of their possessions in the attacking third (Nimes and poor Eibar, one of the best pressing and worst finishing teams in the known universe).
For most non-Eibar teams, part of pressing high is also creating a high defensive line so you can confine opponents to a smaller area of the pitch. Once they’re locked into a tighter space, pressing after you’ve lost the ball is more likely to succeed. But building a high line generally requires quite a bit of passing in the back, and yes, the goalkeeper sometimes gets involved. As Gab Marcotti wrote after Alisson’s dueling howlers, “they don’t do it because they like to needlessly have a giggle or show off their tekkers mid-game. They do it because their managers believe it gives them an edge in terms of breaking the press or enabling them to play a higher line or whatever, and that edge translates in goals and chances created.
Now, Morey’s rant was directed specifically at the act of a field player having the option of making a lower-percentage pass forward, but instead choosing to pass the ball to the goalkeeper. It does seem that the more you can build your line without getting the keeper involved, the better it is for you: teams whose keepers average more than 30 pass attempts per game start 7.2% of their possessions in the attacking third, while their opponents are at 7.9%.
This makes some logistical sense: a pressing team’s defensive line should end up near midfield and if your keeper is seeing that much action, the ball’s probably not close to where it should be. Only a few of Europe’s best teams — basically, much of the Bundesliga’s top tier (Bayern Munich, Eintracht Frankfurt, Wolfsburg, Borussia Monchengladbach and, yes, Bayer Leverkusen) and Serie A’s Lazio — get their keepers involved this much. Others like Atalanta (22.9 GK pass attempts per match), PSG (21.2) and Manchester United (19.2) don’t want their keepers’ feet touching the ball any more than it has to, but still generate solid pressure on the other end.
And now, a special word about West Ham
What would Morey do if he ran a club?
“My football club would be the ‘launch-and-squish’ football club,” he told the Sloan audience. “Launch [the ball] and press and squish, keep them in their end and look for turnovers. It seems like that style is coming when I watch.” He and David Moyes might get along well, then, as Moyes’ West Ham United squad is the closest thing to a successful “launch-and-squish” team in the Big 5 at the moment.
West Ham’s stats stick out like a sore thumb:
– Of the 24 Big 5 teams currently averaging at least 1.8 points per match, they are the only one with a possession rate under 50% (and they’re at an extremely low 41.8%).
– Only two of these teams have their keepers pass the ball (a) under 25 times per match and (b) forward at least 70% of the time: Diego Simeone’s patented defense-and-counter-attacking Atletico Madrid and West Ham.
– While 20 of these 24 teams have their keeper send under 40% of his kicks to the opponent’s half of the field, West Ham does so 59% of the time.
– In part because of this, opponents are starting only 6.1% of their possessions in the attacking third; only six of these 24 teams can top that. (Two of them, admittedly, are Bayern and Eintracht, two of the more keeper-heavy passing teams in Europe.)
– Their possession rate is low because of all the 50/50 balls downfield, but when they do establish possession, they advance the ball 24.0 meters per possession, sixth in the Premier League.
Moyes has never minded the idea of maintaining possession, but first and foremost he wants the ball out of his team’s half as quickly as possible. And this combination is working staggeringly well this season: the Hammers are fourth in the Premier League, ahead of four of England‘s big six clubs, and only four points out of second.
Of course, they are far more “launch” than “squish.” They’re starting only 6.2% of possessions in the attacking third, and only 34.7% in the middle third. They allow 17.7 passes per defensive action, fourth-lowest in the league. They are as high as they are in the table because of their extreme effectiveness in the midfield — they average a goal every 43.5 possessions beginning in the middle third, and they’re led in scoring by a central midfielder, the brilliant Tomas Soucek (who’s also second on the team in ball recoveries).
To maintain a top-four spot over the season’s final 13 league matches, West Ham might have to create a few more easy scoring opportunities for itself. But the verticality that the Hammers employ might be key to eventually breaking the possession game’s stranglehold.
The real threat of playing out from the back is…
The thing most likely to convince the richest clubs to change their possession ways probably won’t be a goalkeeper miscue, no matter how much of a howler it may be. Instead, it might be that counter-attacks — “squish-and-launch,” if you will — are getting better and better.
The most vulnerable moments for a rich, high-quality, high-possession team comes in the seconds after they lose the ball and their counter-press fails. Last season, Pep Guardiola’s Manchester City possessed the ball 66.3% of the time — a decimal point behind Barcelona‘s 66.4% for highest among the Big 5 — and the ball spent so much time at the other end that opponents could only average 0.07 shots per possession, lowest in the Big 5. But the shots opponents earned were likely to spring from quality counter-attacks and were of extremely high quality — 0.14 xG per shot, fifth-highest in the Big 5.
This meant that while City spent most of the year looking like the same dominant team that had won the Premier League the two previous seasons, the Sky Blues stepped on enough random land mines that their point total fell to 81 points, far behind champions Liverpool.
This year, Liverpool and defending Champions League winner Bayern Munich are suffering from their own issues in that regard.
Liverpool are still mostly the “high-upside” team we’ve seen in recent seasons — their heavy pressing allows them to still start 10.2% of their possessions in the attacking third (second in the Big 5) and possess the ball 63.9% of the time (also second). But their patently absurd injury problems at center-back — starters Virgil Van Dijk and Joe Gomez suffered long-term injuries, and virtually every replacement has gone the way of Spinal Tap drummers as well — have created two different issues. First, they’ve helped create random disasters in building from the back; then, when these disasters occur, this frail back line has struggled to prevent opponents from creating good chances. Opponents are starting only 5.0% of possessions in Liverpool’s defending third (third-best in the Big 5), but are averaging a goal every 17.2 such possessions (fourth-worst).
Yes, Alisson’s case of the yips against City contributed to those numbers, but it was an issue long before that.
Bayern’s issues have come about in a slightly different way. Like Liverpool, the eight-time defending Bundesliga champions spend most of the time looking like Big, Bad Bayern — they start 12.2% of their possessions in the attacking third (first in the Big 5), average 2.8 goals per match (first) and possess the ball 62.0% of the time (sixth). But their defensive timing has been randomly off.
Opponents begin a normal amount of their possessions in the middle third overall (32.4%, six percentage points lower than average), but Bayern allows a goal every 35.7 such possessions. In a 3-2 loss to Borussia Monchengladbach in January, all three goals they allowed came from counter-attacks following midfield turnovers. In a 3-3 draw against relegation-threatened Arminia Bielefeld in mid-February, two goals came from quick vertical-thrust situations against their high defensive line, one from a long throw-in and one on a quick move from midfield.
Manchester City’s issues last year proved temporary: personnel changes and manager Pep Guardiola’s endless tinkering have led to enough of a solution that City is up to seventh in the Premier League in xG allowed per shot, while still allowing the fewest shots per possession. We could soon find that Bayern’s issues are equally short-term. After all, they’ve dealt with massive fixture congestion and constant shuffling of personnel; they also lost midfielder and occasional tactical fouler Thiago to Liverpool in the offseason. (Then again, they’re about to lose defender David Alaba this coming offseason as well.)
Liverpool’s issues could settle when they actually have their top two central defenders. Bayern could stabilize with fewer two-match weeks and steadier lineups. But these champions’ struggles have at least hinted at the way forward for teams looking to make up the current gaps in both talent and possession numbers.
If a team figures out how to combine the “launching” of a West Ham, the “squishing” of an Eibar and some proper finishing, we might see the recipe for a post-possession future, one with less goalkeeper-related comedy.