The case for control
How Arsenal lost its identity in the second half at Anfield, and how it can get it back
Arsenal started the second half at Anfield in a way that offered some reassurance to an anxious lot. With a 2-1 lead at the boss lair of this particular level, they came out of the gates energetic and threatening.
Gradually, cracks started to appear. The fluid passing and clinical finishing wasn’t fully coming together. After a chance conceded, then a penalty, then another chance, Arsenal regained enough composure to deny a further barrage for a spell, but the control still felt a little shaky.
At 65’, Rob Holding took the ball near the half-line, and had options for some standard Arsenal build-up.
He plays it out to Gabriel, who was magnificent all day in one of the most difficult assignments imaginable. Thanks to his role marking Salah 1v1, Gabriel had a pretty wide home base. Salah does the expected midgrade sprint towards the big defender:
…and here, Gabi has options aplenty. Ramsdale is available for a pass, Zinchenko is sprinting into his happy place, and Holding is giving him an angle. Xhaka is also pushing out to the touchline, but is probably not the best option:
Instead of playing it to one of the more straightforward options, Gabriel looks vertical. He takes a touch to his left, surveys the options, and tries to boot it long. Salah easily blocks it out, restarting the play and forcing a crowded throw-in:
It’s a small moment, but an indicative one. As the game wore on, and the need for confident control only heightened, the team started to play against type, opting for long-balls, crowded midfield dribbles, and hurried clearances in increasing numbers.
After the play, a frustrated Gabriel could be seen passionately making his case to the team upfield. (If I had to hazard guess, he wanted Xhaka to make himself available sooner, but again, that’s just a guess.)
The camera then cut to Arteta — who was either making the universal gesture for calma, or motioning to keep the ball on the ground when passing:
Either interpretation would have done.
There’s a recent study by a pair of Stanford researchers that examines the effectiveness of communication by leaders. In it, they found that leaders are nine times more likely to be criticized for undercommunicating than overcommunicating.
Francis Flynn, one of the researchers, explained it as such:
“Overcommunication may be seen as annoying and a nuisance, but it’s not seen as a damning flaw for a leader, partly because a leader’s overcommunication is seen as an attempt to benefit you, even if it is misguided, as opposed to an attempt to undermine you or simply ignore you.”
Put another way by organizational psychologist Adam Grant:
“It’s better to overcommunicate and sound redundant than to undercommunicate and seem unclear and uncaring.”
Arteta always errs on the side of redundancy. As cagey as he can be about injuries, he can be entertainingly disciplined about repeating the same messages to convey his expectations.
He is even repetitive about his repetitiveness. After all, “clarity” was one of the three characters in that famous drawing.
We can all think of a few words that seem to come up every time he makes himself available to answer questions. One of those words is control.
Last March, after escaping with a 3-2 win against Watford, Arteta raised his issues with the team’s sense of control:
“We have to show that maturity, understanding and capacity to control the game how we wanted to. At 3-1, after, we had to make 300,000 passes in the opposition half and when they have the right moment to come at us, then we can attack them. We didn’t do that. The game was open and you had the feeling the game was open right until the end.”
The word came up again in his post-match analysis at Anfield:
“We lost control and started to give a lot of balls away. The game became open … The lesson is 'Stick to what we've done the first half’. Play with our personalities and that's the way we have to continue to play. If we do that, we'll win a lot of games.”
Guardiola is frequently preaching the gospel of possession-as-defending. But as we’ve seen against Bayern, and indeed against us, control does not necessarily mean possession: it means imposing your desired style of game on the opponent.
They say the test of a first-rate intelligence is being able to hold two opposing ideas in the mind at the same time and retain the ability to function. While I can’t claim a membership in the first-rate intelligence club, after Sunday, there are two lingering thoughts:
Any team who is good enough to be top of the table in April, and good enough to be up 2-0, is good enough to turn that into a win.
It’s fucking Anfield.
Those two thoughts can help us understand some of the underlying statistics, many of which don’t fall neatly in line with the story of the season thus far. While many of the individual efforts were pristine — Martinelli, Gabriel, Ramsdale, take a bow — and there was no shortage of effort when it came to direct dueling, many of the control stats fell adrift from the standard.
Here’s a quick chart to compare:
While the possession number can’t be considered a failure, particularly when considering the environment, opponent, and game state (i.e. an early lead), many of the other stats point to decision-making by the team, either through necessity, fatigue, or design. The lateral game was 100 passes lower than the season average; they completed 130 more lateral passes last week against Leeds.
The amount of clearances is also of interest. While we all love a good goal-line clearance, the list of top-clearing teams tends to have a negative relationship with “place in the league table”:
Now, this is a bit like saying that wet streets cause rain, but regardless: bad teams aren’t in control; bad teams are attacked a lot; bad teams have to clear a lot. A clearance, after all, is another word for “giving the ball back,” and as a general rule, you don’t want to see your team clearing the ball at rates higher than the relegation dwellers.
In sum, the stats tell the same story: a team that played with an uncharacteristic amount of verticality, directness, and impatience.
The first half was not impervious.
At 14’, Holding corralled a loose ball. Despite having a little space, a couple options, and Xhaka asking him to keep the ball with the calma signal, he nonetheless feels the pressure and hoofs it over the top:
…and a couple seconds later, VVD has secured the ball without challenge, and is starting an attack the other way:
His reasons for concern were not completely ill-founded. He wasn’t a particular target of the press, but had a ball-dominant role up the middle, with Gabriel often far out wide. When he did step up to do some confident press-resistance, it got intercepted:
Later in the game was a more frustrating example. With Núñez now in, Holding is in a matchup with a bigger athletic delta. Liverpool looks to isolate the two, then Henderson pings the ball into space for a running contest:
…but Holding takes a vastly superior angle and easily beats Haaland Lite to the ball by a few steps. Here’s where my frustration rolls in: instead of gathering the ball, or hitting it off to one of the available options (including Ramsdale), he just hits it out, again giving possession back:
In the final minutes, a meek header is on its way to Holding, and he is largely unpressured and has options surrounding him. He plays it off his chest and then boots it right back to Konaté:
Saliba’s recovery speed and defensive peak were missed against Liverpool, of course, but Gakpo ultimately wasn’t a nightmare matchup for Holdini. In the lead-up to the first goal, White was sucked upfield, and Holding was meant to cover some difficult combinations with light help. He was perhaps a little slow-footed, but did fine, and the ball still required a couple of fortunate deflections to go in. The penalty was, I don’t know, a little clumsy and a lot unfortunate.
It was perhaps in this area, however — the area of control — that Saliba’s absence loomed largest. Thanks to Holding’s penchant for long-balls, some impatient clearances, and a few misplaced passes, he wasn’t able to offer the same type of possessional dominance that Arsenal imposes this year. His 9 clearances matched Saliba’s season-high. His 15 ball losses were more than Saliba has had … in any game of his career, across leagues.
This had a few knock-on effects, and one of them was that it dulled the in-possession games of his colleagues in back. Holding, White, and Gabriel only exchanged 19 passes on the day. This is down from 43 against Leeds, and 59 in Saliba’s last game against Fulham. White had the fewest touches of any player on the day, which in turn made it difficult for Ødegaard, Partey, and Saka to get involved to the degrees we expect.
This is equal measures surprising and unsurprising. On one hand, White played further back in a flat-3, providing more of an outlet (as we covered in the preview, and saw in his outlet ball to Saka that led to the first goal). On the other hand, Liverpool is a superior opponent with a superior press, and as we’ll see next, White didn’t have the best game.
While directness may have been the order of the day in the early phases, some calmer periods of possession would have been preferred when it turned into a basketball game.
Holding was not alone, and perhaps not even one of those most responsible for the loss of control in the second half. It was a team effort.
At 48’, Xhaka received it in standard build-up and had Martinelli on the touchline (which isn’t usually a preferred option for Arteta) and Zinchenko cutting into space. Instead, he tried to dribble through pressure, keeping it long enough for Konaté to enter the frame and kick it away, setting off a good transition opportunity for Liverpool. He had a couple of these overdribbles on the day:
A little later, Arsenal gets it in the exact kind of spot they’d want. Zinchenko switches the play, which allows them to play around a narrow Liverpool press with ease. Holding gets it out to White in space, but White immediately hits it long to a double-covered Saka, who had no shot at it. Possession lost:
Four minutes later, White receives it on the touchline after a throw-in. Liverpool uses aggressive field-leans in these situations, and can be vulnerable to quick switches of play. Though he has some tiki-taka options around him, which might open up such a possibility, he boots it directly to Alisson instead. Possession lost:
In the sequence leading up to the equalizer, Saka gathered a loose ball. Instead of passing it back to White or Partey, he turned back upfield, and was eventually triple-marked and dispossessed:
Martinelli was not immune, trying some heroic dribbles before getting squeezed:
Some of Zinchenko and Partey’s signature line-breaking passes were intercepted, as well; it can be argued that they were exactly what the team needed, or, given game state, that they should have settled down, and opted for death by 300,000 passes instead. Regardless, it didn’t work out. Even Ramsdale booted one into the stands!
🔥 In conclusion
It’s difficult for we mortals to understand the difficulties of playing at a place like Anfield. Despite all their struggles, Liverpool have only lost once in the league there all year, and had conceded precisely 0 goals at home in 2023 before this one. Man City lost there 1-0, those bottlers.
It can also be difficult to disentangle when a team loses control from when an opponent takes it away. Liverpool certainly deserve credit for their performance, and Klopp deserves credit for his tactical tweaks, pulling TAA into the midfield to regain control, and subbing the right players at the right time. They looked pretty vintage.
Still, I can be sanguine about taking 4 of 6 points from Liverpool, and optimistic about the season moving forward, and still put this at the midpoint of “Liverpool taking control” and “Arsenal ceding it.” Arsenal looked every bit the better team in the opening period, but this game was bound to ebb and flow. The trouble started when Arsenal were caught in-between identities: are we the dominant, controlling, top-of-the-league side who can impose our style of play on anyone, even Liverpool? Or are we a team on the wrong side of momentum, holding a tenuous lead at Anfield, and need to hoof it long as much as possible?
Wyscout measures a long pass as a “ground pass longer than 45 meters or a high pass longer than 25 meters.” By that measure, Arsenal is the least long-passing team in the league (at 7.16%). Without sufficient numbers forward, particular physical advantages to exploit, or, frankly, a lot of experience in that style, it’s not a recipe for success at the moment. Liverpool has long baited opponents into playing this way, and with Alisson usually leading the league in how far he acts away from goal (as we saw in one particularly bummer of an example), the margins are slim.
Much has been made of the subs. When Firmino came on for Fabinho, I think there was a case for responding with a Jorginho sub to help fill the vacated midfield with the 300,000 passes that Arteta seeks, but I don’t feel that strongly about it.
On Kiwior, it’s an easy decision to disagree with, particularly after the fact, and most of us can say we Disagreed With It at The Time Because We Are Very Smart. What was less justifiable was not immediately pairing the Kiwior sub with the Tierney sub; if we’re pulling out Ødegaard for a centre-half, waving the white flag on midfield and possession control, and looking to max-defend and perhaps spring the counter, then that’s a game that maximizes both Zinchenko’s weaknesses and Tierney’s strengths. Similarly, fresh legs for Saka were probably in order given that game state — be they from ESR, Vieira, or Reiss. It may or may not have changed anything, but it has a better logical throughline.
In general, though, I’m not in love with this particular 5-back for two reasons: one, it’s asking Arsenal to play against its identity. Two, a 5-back does not guarantee defensive control (take a look at Tottenham), and relies on a good amount of mutual understanding among the backline. I’m sure a back-three of Saliba, Holding, and Gabriel have that; I’m not sure if Holding, Kiwior, and Gabriel are at that level of mindmeld quite yet. But again, easy to say in hindsight.
What can be improved for next time? First, I’d make a passionate case against playing at Anfield again this year 🙃. Second, when the game is feeling nervy and unsettled, this team should do what they do best: dominate possession, cycle the ball through, manipulate defenders, regain confidence, and pick the spots to attack (instead of picking every spot to attack). The lack of lateral passes, particularly against an opponent who can be vulnerable to being moved around, had a significant impact on the proceedings.
While some absences were undoubtedly felt — notably Saliba and Tomiyasu in this case, but perhaps an underrated shout goes to Eddie — I also believe this team was fully capable of winning this game with Holding in back. For all the concerns about his ball-playing ability from the fanbase, Rob Holding may be the one who needs to improve his appraisal of Rob Holding; there were more than a few times when he came down with a ball that he was fully capable of passing, but instead, he booted it in haste. He’s a wonderful person and a capable player; with him, this team is still good enough to win the title. Most of the first half showed that.
While he was still a player, Arteta gave an interview to FourFourTwo in which he outlined some of his philosophy:
“Stick to your passing game and keep the ball in their half. If you start forcing the play and giving the ball away you give them the opportunity to hit you on the counter attack. Arsene Wenger is so calm. He tells us to keep playing our football until the last minute of the game, and that has paid off.”
He’s been nothing if not consistent.