Throughballs!

Salacious details of soccer’s SEXIEST PASS revealed! Jürgen Klopp EXPOSED!

Every soccer game has a moment when a bad bounce or light breeze might have rewritten the morning papers, maybe even changed the course of a season. In Liverpool’s first Champions League match against RB Leipzig, it happened just after halftime.

Forty-sixth minute. Still a 0-0 game. Some clumsy buildup play from Leipzig draws Liverpool’s press up the wing, but the ball shakes loose to Dani Olmo between the lines for a three-on-three break. He dribbles in from the left toward retreating center backs while Christopher Nkunku loops around to the right channel, behind Jordan Henderson. As Olmo steams into the channel, Nkunku knifes back inward and all the passer has to is sidefoot the ball into a 15-yard gap between Henderson and his partner. Textbook throughball. Nkunku beats his defender to the ball, sends the goalkeeper sprawling, and chips it over Alisson’s shoulder at an empty net—until the keeper’s goodbye wave serendipitously slaps the shot away.

Another inch, a few milliseconds here or there, and Liverpool’s season might have crumbled. Instead they won 2-0 and still have a chance to claim another Champions League trophy, if only they can answer a question that’s been nagging at them all season: what’s up with all these throughballs?

Passes that split the last line used to be something good teams did to bad ones. In 2018-19, by Statsbomb’s count, Liverpool allowed 0.89 throughballs per league match, better than about two-thirds of Europe. Today they’re giving up 1.88, the worst rate of any defense in any top league in FBref’s four years of data. The why of it is pretty straightforward: unlike Bayern Munich, whose defensive collapse is more complicated, a glance at Liverpool’s injury list will make you wonder how they field a back line at all. I’m more interested in the how.

How do you create a good throughball—or stop one? What are the steamy secrets of soccer’s sexiest pass, and how can Jürgen Klopp resist?

The View

To play a throughball, you’ve got to see the throughball. Sounds dumb, I know, but it matters. Almost all throughballs are set up by a carry or a pass, and two particular kinds of set-up action are more effective than the others: short forward carries and short sideways passes (where “short” means anything under 15 yards).

Why those two actions? I can’t prove this to you with on-ball data alone, but I think it has to do with the passer’s field of view.

A couple notes on data before we dig into this. First: What is a throughball? If you’re dealing with event data, it’s whatever the provider says it is. Actual human event loggers are given some kind of written definition and told to flag any pass that looks like it meets it. This can get messy. Definitions vary across providers and may be inconsistently applied because soccer is primal chaos cosplaying as an organized sport and it’s not always crystal clear when the ball has gone “through” a “line.” Unfortunately, since throughballs only happen a couple times a game, you can’t really get a big enough sample to draw firm conclusions about a question as narrow as why Liverpool is giving up more than most teams in 2021. Instead, for the rest of this letter, we’ll be working with five seasons of Premier League data to try to get at some more general truths about throughballs.

Second thing: What makes a throughball good? You might say it’s good if it’s completed, and that’d be fair enough, but we’re going to go a step further and look at the goal probability of the whole possession after the throughball. In the table above, I can’t tell you why throughball attempts after long forward carries have a higher completion rate but a lower possession xG than after short forward carries, but I can tell you which of those two numbers matters more. As far as I’m concerned, a throughball is good if and only if it’s likely to produce a goal.

Anyway, back to the table. Most of the time it doesn’t make much difference whether the action before the throughball is long or short, forward or sideways: the difference between most of these possession xG averages is so small that it probably doesn’t mean anything. But short sideways passes and short forward carries seem to set up significantly better throughballs, and like I said, I think it’s about players’ ability to see the pitch.

A passer who’s carrying the ball forward can get a good look at the situation in front of him and pick the right moment and angle to release the ball. But if he’s receiving a pass before attempting his throughball, it’s better that it come to him sideways, not forward. A forward pass forces our guy to receive, turn, get a good look at the field, and then play the throughball. That’s a lot of time for the defense to get organized. A backward pass gives the throughball guy an optimal view of the pitch, but it moves the play farther away from the back line and makes the pass harder; worse, there’s a good chance a defender is now chasing the back pass, putting our throughball passer under pressure and cutting down his angle.

A sideways pass avoids the drawbacks of a forward or backward one. Our throughball passer can receive a diagonal pass on the half turn, with a body shape that gives him a good view of both the pass he’s collecting and the one he’s about to play, allowing a quicker turnaround on the ball than if he’d had to, you know, actually turn around.

Even better, a sideways pass that travels across the face of a defender forces him to turn to follow the path of the ball, which can disorient him for valuable fractions of a second. A runner who was in front of a center back before the sideways pass may suddenly find himself behind him, out of view, which is the perfect moment to cut in behind for a throughball.

The Angle

As for the throughball itself, it’s best if it travels toward goal (obviously) but not straight toward goal. Most throughball attempts are aimed pretty close to the imaginary line connecting the passer to the center of goal, but the wider a pass strays from that center line, the more likely the throughball is to lead to a goal.

Let me explain what you’re looking at here, because it may not be obvious. This is a variation on Eliot McKinley’s popular “pass sonar” viz, which shows how often passes travel at different angles: longer bars mean more pass attempts in that direction, while yellower bars mean that passes in that direction are more likely to produce goals.

But unlike most sonars, where the top of the viz represents the endline, up on this one points from the passer toward the center of goal. The difference between this “progressive angle” and the more familiar “endline angle” depends on where our passer is standing. If he happens to be in the middle of the field, in line with the penalty spot, the progressive angle and the endline angle will be the same. But if he’s over on the wing, the zero progressive angle will point away from the endline, because passing to the corner flag isn’t progress. True north on our progressive angle sonar always points to goal.

The reason this matters is that a defender who’s trying to get goalside—between the ball and the net he’s protecting—will position himself along that zero progressive angle line. He’ll be less concerned with stopping passes that aren’t aimed north on our viz, since they’re not straight toward goal. But that’s exactly why these angled passes are most dangerous. A throughball attempt aimed at goal will create a goal less than 9% of the time. One that’s angled 30 degrees away from goal in either direction has a 13% chance of winding up in the back of the net. That’s a big difference!

A quick clip will make it easy to see why throughballs away from goal are more dangerous. Tyler Adams could have tried to thread the needle here at a zero progressive angle, along the purple arrow, in hopes of reaching Yussuf Poulsen in the central channel. But that passing lane is tiny: there are two defenders between the ball and the goal, and even if the pass beat them the keeper would probably get to it first. Instead Adams aims his throughball away from goal, through a near channel that he’s got a more generous angle on, and puts Hwang Hee-chan in on goal with an option to shoot or square it.

Of course it helps that Liverpool’s center backs are on totally different wavelengths in that clip, which brings us to ...

The Line

The easiest defensive line to break is one that’s not really a line at all, or at least not a straight one.

Offside traps are a risky game: defenses like Liverpool and Bayern love to use them to compress the playing area and defend high, but if you mistime your step up you wind up with an attacker in on goal while the defense is moving the wrong way. That’s pretty much a worst-case scenario as far as throughballs go.

Instead of trying to beat a whole line, it’s better to only have to beat part of it. Look at how much leeway Adams has in that earlier gif against Liverpool’s ragged stoppage-time defense, with the center backs several body lengths apart. To create holes in a back line like that, attacks will often try to draw one defender out with third-man passing zig zags or dummy drops, tricks that were a little too easy to do to Joe Gomez even back when Liverpool’s squad was at full strength. But there’s another way to catch a defense in disarray: beat it in transition.

RB Salzburg’s Jesse Marsch is far from the first coach to preach chance creation through high pressing and direct transitions, but his old New York Red Bull sides were so dogmatic about it that he’s always the guy I associate with the strategy. Any time, anywhere his MLS team won the ball, their first look was into the channel between the opposing center backs. “The goal is to get to the penalty spot, to get to goal as fast as possible,” Marsch explained in a recent webinar. Circulating or going wide would only waste time and let the defense regroup. Klopp would know, actually: back when Marsch’s Salzburg still had Erling Håland, their hyper-direct transition game gave Liverpool a 4-3 scare in the Champions League.

Like Marsch, Southampton’s Ralph Hassenhütl worked under Ralf Rangnick in the Red Bull system and plays a similar game (close enough to land in the same Aggressive Style gang as both Leipzig and Liverpool). When Southampton beat Liverpool in January, the gameplan was all lightning buildups and even lightning-er transitions to catch the defense out. When the dust settled, FBref counted five throughballs against Liverpool in a single game, the most they’ve conceded since that 7-2 catastrophe to Aston Villa.

So is the best throughball always the fastest one? Not necessarily. Premier League data suggests that throughballs are most dangerous during the first couple seconds of a possession, probably off of high turnovers, but the time-danger curve is U-shaped: after the defense has had a moment to get organized, throughball effectiveness swoons until very long possessions take the time to pull the defense apart with passing patterns, creating new holes to play through. The shape of this curve probably also owes a lot to the fact that teams that can string together really long possessions tend to be really good teams: Southampton’s pretty good at throughballs, but Manchester City is better.

The Runners

Not going to beat this last point to death because you’re probably as throughballed out as I am right now, but so much of a throughball’s effectiveness depends not on the passer but on the runner at the other end of it. His skulking around before the play to make the defense lose track of him; his timing, angle, and quickness when he breaks for goal; his first touch to cushion the pass, his strength to hold off last-ditch challenges and stay upright, his eye for finding the keeper’s weak spot and snapping off a shot: all this stuff matters at least as much as the quality of the throughball itself.

Better yet, make that runners, plural. If you’ve ever seen a nicely timed move with teammates shooting two different channels at once—and if you think you haven’t, just scroll back up and watch that Tyler Adams play again—you can picture how it heightens the threat. Instead of shooting, the guy who received the throughball (and who is now most likely receiving extremely close personal attention from the goalkeeper) can square it to the second runner, who’s looking at an open net. Even if the throughball receiver doesn’t pass, his shot may be made a little easier by the hesitation his teammate’s run induces in the keeper or defense. As with most things in soccer, throughballs are better with company.

I assumed that second-runner phenomenon would show up in the data, but when I ran the xG for guys who shot immediately after receiving a throughball compared to teammates who didn’t receive the throughball but did take a shot some time in the next two or three actions, the average possession xG favored first-time shooters, 0.28 to 0.22. Which seemed weird until I remembered that xG models typically use a key pass’s throughball flag to boost the probability of a resulting shot, but that flag doesn’t carry over to a shot two actions later if the guy who received the throughball passes to someone else.

Sure enough, the actual scoring rate on second runners’ shots appears to be much higher than the xG model predicts, producing a goal 34% of the time on just 0.22 xG. The throughball and square pass have cleared the defense out of the way; the model just doesn’t know it. Are little quirks like that a problem for some of the possession xG analysis in this letter? Yeah, might be. But hey, it’s a good sign for the power of teamwork!

The Conclusion

So just to recap, a good throughball should be:

• Very early or very late in the possession to disorganize the back line
• Set up by a short forward carry or sideways pass to give the passer a view
• Angled away from goal to avoid goalside defenders
• Supported by a second runner for a better shot
• Preferably attempted against Liverpool’s fourth-string center backs

Is Klopp going to get his throughball sitch sorted out in time to make another deep Champions League run? Maybe, maybe not, but at least now you’ve got some stuff to watch for when Leipzig comes to Anfield next week. ❧

Thanks as always to Eliot McKinley, Tyler Richardett, Matthias Kullowatz, Jamon Moore and the rest of the American Soccer Analysis crew for helping me be less dumb about this stuff.

Image: Roy Lichtenstein, Crying Girl