The Problem With Possession

Too much passing is bad. But what this article presupposes is .... what if it's good?

I have a completely unsubstantiated theory for you: US Soccer is experiencing the same progression of soccer history as everyone else—it’s just on a 10-year delay.

I grew up on Long Island, and from about 1996 to 2002, if you had a British accent and owned a pair of wrinkled Umbro shorts, you could pretty easily convince Suffolk or Nassau County parents to give you $50 an hour to teach their kids how to do diving headers. I’m not kidding; there was this scam called “NOGA” -- I can’t remember what the acronym means; in fact, it might not even be an acronym! -- where they tried to turn every activity into a game with some catchy name and then spent a ton of time teaching 10-year-olds ... how to cross  and then head the ball? I remember one of the trainers telling my parents I’d “scored a peach of a goal”. It was by far the worst coaching I have ever received ... and yet this fascination with the British game came directly on the heels of England failing to qualify for the 1994 World Cup, thanks in no small part to the faulty coaching methods being espoused by Charles Hughes and English FA.

Since then, the soccer world has moved through a variety of phases. First, there was the obsession with possession, then came counter-attacking, then came counter-pressing. We’re firmly in the age of counter-pressing right now, but the peak hysteria has cooled off a bit, as top teams now employ their presses more strategically and selectively, rather than trying to turn the full 90 minutes into a Metallica concert.

Every preceding style is both an evolution from and a refutation of what came before it. The phrase “tiki-taka” used to be a loving onomatopoeia for the never-before-seen quickness at which Spain and Barcelona passed the ball. Since then, though, it’s morphed into a pejorative. Take it from the guy who invented it: “I hate tiki-taka”, Pep Guardiola said while coaching Bayern Munich. “Tiki-taka means passing the ball for the sake of it, with no clear intention. And it's pointless. Don't believe what people say. Barça didn't do tiki-taka! It's completely made up! Don't believe a word of it!”

He’s mad! And well, there really isn’t anything more maddening than watching a team dominate possession but fail to turn it into any end reward. Most teams now are more focusing on finding ways to control space rather than the ball, but remnants still exist. There was Spain at the World Cup, completing 1008 passes to Russia’s 192 ... and losing the match. There’s Maurizio Sarri’s Chelsea this season, hoovering up the third-highest average possession total in all of Europe, but scoring just the sixth-most goals in England. And then there’s ... the United States Men’s National Team? On Tuesday, in their first game with a full complement of players under new manager Gregg Berhalter, they secured 66 percent of the ball, but managed just six total shots. For reference, Ligue 1’s Caen take the fewest shots per game of the 98 teams in Europe’s top five leagues ... and they’re still averaging nine attempts per 90 minutes.

If you wanted to subscribe to my theory and wildly overreact to one match, you could say: The USMNT is entering the possession-obsession stage of things—10 years too late. I am not going to do that -- it’s way too early to say anything, really, about the USMNT -- but I want to take a look at the other side of the coin. Because even if you can’t turn your passes into shots into goals, all that possession does at least keep the ball away from the other team. Russia never scored from open play, Chelsea only concede more chances than five other teams in Europe, and Ecuador registered a solitary chance -- a big ol’ “1” -- the other night in Orlando.

Possession as defense isn’t necessarily something new -- it’s a core component of Guardiola’s teams -- but I do wonder it can’t be used more aggressively after a team takes the lead. Reader-question time -- here’s Noah:

A team, we’ll call them Definitely Not Liverpool, is up 2-1 in the 75th minute of the Champions League final. They have a goal kick. What, if anything, prevents them from simply passing the ball amongst themselves for the remaining quarter hour? Soccer fields are enormous and it’s not hard to think that 11 “attackers" (if you include the goalkeeper) could pretty easily keep the ball away from 10 defenders, given that the opposing keeper is rendered more or less useless in this scenario.

You see teams sort of do this where they clearly don’t have any intention of attacking the goal but then they lose the ball because Dani Alves gets bored and decides to try to thread a 60-yard diagonal pass through eight defenders because yolo and also he’s late for his tattoo appointment. I’m talking about a very formalized game of keep-away. Extended further, why not go into keep-away as soon as you go up a goal. Get the ball back as fast as possible and keep it for the rest of the game. Aside from the obvious lack of sportsmanship and a dramatic increase in mean glares from Gianni Infantino’s box — it’s so much harder to do crime when your fanbase is bored — what’s the downside? Plus, think of the possession percentage advantage! (One argument against is that if you did turn the ball over, it would likely be in a dangerous position?)

This person, it may shock you to hear, is someone I would consider a “good friend”. Anyway, aggro-rudeness aside, the legend of tiki-taka suggests that it’s borne out of the “rondo” -- Barcelona’s glorified pre-game version of keep-away. (Listen. We did this to warm up in college, and our coach was no tactical mastermind; he also made us play full-field 1-on-1. I refuse to believe that tons of other clubs weren’t doing the “rondo”! And no, I don’t care if I sound like Sean Dyche!) Anyway, what Noah is suggesting is basically just a full-field rondo, where the end goal isn’t a goal, but to keep the ball from the other team. No shots, because shooting is the quickest way to ensure that you lose the ball.

I’ve written before about how it seems like teams, on average, don’t quite know how to handle taking a lead. Here are two paragraphs of PREMIUM CONTENT:

When it’s 0-0, the two sides typically have the same incentives -- attack enough to win, but don’t attack too much that we give up a goal and lose. Once it’s 1-0, though, things start to change. Theoretically, the team that’s up a goal can now sit back slightly, harden up its defense, and then attack more selectively into what, theoretically, should be an undermanned defense that has pushed players forward. That, broadly, shows up in research about different scorelines (or “game states” as they’re called): When a team goes up a goal, it tends to take fewer shots, and vice versa. But teams also tend to convert their chances at higher rates when they’re winning, presumably because there are fewer defenders in the way and because counter-attacks are an efficient way to score.

However, some other research has found that, despite the increased conversion rates, teams actually perform worse than you’d expect -- both given pre-game gambling lines and universal in-game win-probabilities based on the score -- when they’re winning, compared to what you’d expect when the game is tied. Think of it like an NFL coach who punts on every fourth-down; a lot of teams play it too safe when they take the lead.

I wouldn’t advise any team to try to play keep-away for 85 minutes after going up 1-0 in the first five. I mean, I obviously do advise that; it would be incredible to watch, and I’m going to dream about Ajax doing this to Juventus for as long as my unconscious allows. But, as Chris Anderson and David Sally write in The Numbers Game, the second goal in a game is actually the most valuable in terms of affecting the outcome, so playing toward that makes the most sense.

In fact, I wonder if managers don’t misread the incentives here. The downside of not playing defensive enough when you’re at 0-0 is that you give up a goal and lose. The downside of not playing defensive enough when you’re up 1-0 is that you give up a goal and draw. *Reads Michael Lewis once* You might say it’s the Endowment Effect -- coaches and players aren’t scared of losing the possibility of three points until they actually have the three points in their possession.

But let’s say a team gets that second goal, or let’s say, as Noah says, that it’s the 75th minute of a one-goal game between two teams of either roughly equivalent ability. Typically, the goal kick would be the toughest position to keep possession from -- the opponent can push its attackers up to the edge of the penalty area and force a long ball or a dangerous chipped pass to the midfielders behind the attackers. However, that rule is going to change soon and passes will be allowed to be completed within the penalty are. Our thought experiment continues apace.

There have been plenty of studies about how possession doesn’t have a strong correlation with success, and then there have been a succession of remixes about how we’re quantifying possession all wrong. The best teams have the most skilled players and so they tend to have more possession on the whole, but those numbers don’t say much on a game-by-game basis. However, all of this research at least agrees on one thing: when a team goes down a goal, it will typically go on to have more possession, and when a team takes the lead, the opposite happens. A 2016 piece from Statsbomb also found that teams that are winning tend to attempt longer passes, while losing teams attempt shorter ones. This is all despite the fact that possession serves as as a successful defensive strategy.

Maybe this ties in with why teams aren’t as good at winning when they’re, you know, winning as they should be? A limited analysis of 24 Premier League games from the 2001-02 season produced the following takeaway:

It was interesting to note that both the successful and unsuccessful teams kept the ball for longer periods when they were losing compared to winning. This may be explained by a greater effort to regain possession by a team when losing i.e. needing a goal to avoid defeat compared to when winning. ... Perhaps the most pertinent finding was that the successful teams had longer possessions than the unsuccessful teams when in winning situations. This suggests that possession should not only be investigated with respect to how goals are scored and created but how possession strategy is related to preventing the opposition from scoring.

My hunch is that something is off. If the possession profile of a team changes based on the scoreline, then it seems like there’s some kind of strategic opportunity in there. Maybe it’s that teams should make a concerted effort to keep more possession once they take the lead -- fewer opportunities for your opponent and given their increasing desperation, perhaps all the wild chasing would tire them out even more? Or maybe it’s that the aggressive play of the trailing team makes it especially difficult to hang onto the ball, and so the leaders should suddenly start playing a way more direct style once they go up a goal in order to get that incredibly valuable second.

Whatever the answer, I’d love to see a coach and a team try to use the effects of the change in the scoreline to their advantage. And why not Berhalter’s USMNT? American soccer doesn’t have to be in the thrall of a delayed history. After all, they did decide to ban children from doing diving headers.