Open and shut
Back-fours, back-fives, and the question of space. Why Arsenal thrived against PSV and struggled in familiar ways against Manchester United — and what can be done about it
“A Trauma-Informed Theory of Coaching.”
It sounds like a solid book proposal — or at least a dissertation topic — doesn’t it?
Please don’t answer that.
Just a few weeks ago, we wrote at length about how Arsenal’s 2023 summer overhaul could be interpreted as a direct, and successful, response to the previous season’s crushing 4-1 loss at the Etihad. Mikel Arteta saw Erling Haaland and Kevin de Bruyne galloping through his beleaguered team, skipping past a press that struggled to contain space against elite opposition, all ahead of a jumpy-at-times keeper, and said: Fuck that. Never again.
The response carried considerable conviction — and expense. Declan Rice, Kai Havertz, Jurriën Timber, and David Raya arrived, and with them, a great deal of space-eating, cross-claiming, and athleticism. Arsenal haven’t lost to City since.
Arteta is not the only manager who could respond to the questioning of his tactics as The Joker would: Do you wanna know how I got these scars?
Antonio Conte is another. After years of success in Italy, he joined Chelsea in 2016. Though widely known for his back-three Juventus teams — we’ll use “back-three” and “back-five” interchangeably in this newsletter — he initially rolled out a back-four at Stamford Bridge. That all changed after Arsenal ran riot in September.
Arsenal stormed to a 3-0 lead by halftime, a scoreline that held until the end. It was Conte’s heaviest defeat since his time at Siena six years prior.
“It was a shock, for sure,” he said. “Usually, my teams are not used to conceding three goals in the first half, so it was a shock, and a shock I tried to transfer to my players. I don’t want this.”
Conte didn’t wait until the final whistle to make a change that would prove momentous. Ten minutes into the second half, he brought on Marcos Alonso and, with him, a shift to a 3-4-3 that Chelsea hadn’t trained with the personnel at the time. In truth, though, it was never far from Conte’s mind.
The results were immediate. Chelsea rattled off thirteen consecutive wins, ultimately setting a then-Premier League record for most victories in a single season (30). By May, Conte had a Premier League title to his name.
The shift rippled through the league. By the end of his debut campaign, 17 of the 20 Premier League teams had used a back-three system at least once, with total usage rising from 4.5% to 21% in just one season. That included Le Professeur himself.
“It’s the first time in 20 years I’ve played it. It shows that, even at my age, you can change,” Arsène Wenger said after deploying a back three against Middlesbrough in April 2017.
Such brilliant tacticians wouldn’t have turned this way without ample evidence. Not only were the winning percentages generally higher at the time, it also proved to be a serious obstacle for the top positional sides of the time — embodied by Pep Guardiola. Conte, of course, has had Pep’s number over the years, winning four of their seven meetings.
Guardiola’s thoughts drifted to solving this issue when previewing the 2018/19 season.
“We found a lot of problems when attacking a 5-4-1 when the striker is marking our holding midfielder and 10 players are defending so deep. We are thinking about how you could attack against that system.”
Three years later, Guardiola’s City fell to Chelsea in the Champions League final, a game which Pep famously overthought, starting no holding midfielder.
Thomas Tuchel employed a back-three.
🎢 Ups and downs
In the latest chapter of Arsenal’s challenging season, the side became the first in Champions League history to score 7+ goals away from home in a knockout stage match. Then, a few days later, they faced “the worst team, maybe, in the history of Manchester United” — their manager’s words, not mine — and drew 1-1. If not for heroics from David Raya and Declan Rice, it could have been even worse.
Then there was the dead-rubbery second leg against PSV. Raheem Sterling was named man of the match; his batterymates across the front were Mikel Merino and Kieran Tierney, of course.
Real Madrid await. (Gah, I’m looking forward to that.) In the meantime, I’ve tried my best to retain my curiosity about this period. As much as it can get chalked up to injuries and competition levels and familiar issues, there are little details in play that are worth investigating. The big question: aside from the obvious, why did the results against PSV and United differ so much?
There’s the normal blah blah blah, but Arsenal’s ongoing struggles against physical, mid-to-deep Premier League blocks are true, and they are well-documented, not least of which here.
Rúben Amorim, to his credit, has frustrated Arsenal before.
He did this first as the opposition manager in that dreadful Euros knockout two years ago, and again this season by knocking Arsenal out of the FA Cup. Given that these games also coincided with the Saliba, Tomiyasu, and Jesus injuries, I’d like to formally retract any previous nods to his “rizz” in early press conferences, and fully commit to the idea that he is some kind of cursed raven portending doom upon our flock. Don’t be fooled by the hoodies.
When contrasting the relative performances against PSV and Manchester United, it’s not hard to see why the games had different fortunes. In the first, more competitive match in the tie, PSV played open, pressed high, and left gaps. United bunkered. And yes, the arms race in the Premier League is worth remembering, too: Amorim, having a brutal year in charge, was the best manager in Portugal; Zirkzee and Højlund were two of the best players in Italy; Bruno is one of the best in the world; Casemiro, whatever his issues now, was a pillar at Real Madrid; De Ligt was a two-leg starter for the Bayern team that knocked Arsenal out of the Champions League. Even the memeable teams in this league have plenty of quality.
But it’s also worth considering how Arsenal’s attack functions against back-four and back-five teams.
Arsenal have faltered against disciplined four-backs — Villa, Newcastle, Fulham, Porto, the like.
They also failed to score against five-backs like Atalanta, Inter, and … Newcastle again, who switched to a back-five in the last meeting.
But the recent performances in those back-five matchups — particularly West Ham’s five-back, that brutal Newcastle second leg, and Manchester United (twice) — feel particularly underwhelming, especially considering West Ham and United’s struggles overall. For whatever I said earlier, Manchester United cannot really run or pass, two characteristics that are typically important in this sport.
Comparing the first PSV leg to Manchester United, Arsenal had:
213 fewer passes
18% less possession
12 fewer crosses
Lower average passes per possession
…and six more goals.
What was the cause of the different results? And how much had to do with the back-five formation?
🎯 Enter PSV
In search for evidence of the trauma-informed theory of coaching, PSV coach Peter Bosz also applies. He noted last season’s battering as a key moment that helped propel the team to an Eredivisie title.
“I learned a lot from Arsenal. Afterwards, I thought, ‘We played pretty well… but we lost 4-0?’ In the build-up, we did well and we had our position game, but as soon as we got to their box it was over. How is that possible?,” Bosz says.
“Me and my coaches studied them. What is it that they do differently to us? The answer is that they are outstanding in the opposition box but also their own. They get a lot of players behind the ball as soon as possible. They do it with 10 or 11 but we only did it with six or seven and then the distances are bigger. It’s the transition.
“We showed the players and, in the games after, we started doing it with 10 or 11 like Arsenal, staying compact in attack and defence.”
The problems continued this year.
Arsenal’s ability to attack a team like PSV made it so.
🧨 Attacking a back-four
In an ideal world, you have players who simply line up 1v1 against opponents and destroy them. This is generally what works for Arsenal from a defensive perspective: the reason the press can be so ferocious is because you fundamentally trust Saliba and Gabriel to win their battles, which frees up an audacity by everybody else.
In attack, Arsenal are shorter on this quality. Saka has it, Havertz often has it, and Nwaneri is certainly trending that way. But generally speaking, Arsenal are heavier on thoughtful, creative associators than outright attacking juggernauts; we’ve got more Modrics than Vinis, in other words. That means you need to add another element, especially with all the injuries.
Let’s go back to that goal that helped provoke Conte to change his formation.
What do we see here?
Arsenal using central access (Iwobi) and quick passes
Forwards moving the backline around
…and, crucially: a zippy full-back (Bellerín) marked by a winger (Hazard)
…and that insight will show up again and again in the clips to come. When facing a back-four system — particularly one that presses — your full-back is often tracked by the opposing winger. If you can safely bet on an advantage between your full-back as an attacker and their winger as a defender, that is one of your best means of opportunity; your full-back will be able to provide numbers to the wide areas and pounce from there.
This has played an enormous role of late.
In the first leg against PSV, Arsenal generally set up in a look that will be familiar: the left-back (Lewis-Skelly) inverted, and we mostly saw a 3-2-5 formation in build-up, with each zone covered across the pitch.
PSV generally move and track individual markers, instead of camping in a zone. Early on, you could see the sheer amount of rotations and activity we saw in response.
“What I loved about this game was the freedom, the expression and the courage to just go a little bit crazy,” said Adrian Clarke. “It was mayhem in the movement, and I’ve been banging on about this all season. I’ve especially cranked up my moans and groans lately because, without a striker, the only way you’re going to disrupt the opposition is to cause chaos.”
In all, there were three notable aspects of the game-plan that produced dividends.
First, fluidity. Here, without a pass even being made, you can see Partey dropping down to right-back (really RCB), Timber pushing up to right-wing, Nwaneri pinching in, and Ødegaard floating every which way. With this mostly man-to-man look, you can drag around players at will.
Second, goalkeeper as a +1. This is almost universally the case with Arsenal, but it goes to show how a goalkeeper’s comfort on the ball is not just about helping build-up. It is a transaction: one more player added to build-up is one more player shifted to attack.
Below, you’ll see how PSV have a relatively passive front-line of press, but are aggressively man-to-man in other matchups. Because Raya is there, the buildup is basically a 6v4. If you take out the Timber matchup, it’s still a 5v3 — which should be solvable by the players on the pitch.
And so … our third point, right-back as winger.
In that same sequence, look at Timber counting the numbers that the team has in build-up. Then, knowing that Noa Lang will follow him wherever he goes, he clears out.
Facing a similarly Man-to-Man City (😘) last season, this was a big source of the advantage that was ultimately gained.
This, likewise, is for a few reasons.
By pushing Ben White up, Phil Foden was reduced to wing-back shifts, which impacts his ability to get into dangerous positions as an attacker.
Again: if you have full-backs who are capable in wide areas, you can gamble that they’ll attack better than the opposition winger can defend.
Your right-winger can get more central.
In all, it’s easier to push numbers on the back-line.
That expensive Giant Killer that Arteta bought in the Summer of 2023 also works against sides like PSV.
This all came to fruition in that record-breaking Champions League match.
You can see how that previous sequence worked itself out. Nwaneri is pinched in and closer to goal, Ødegaard can float and thread things through, build-up is easy, and five attackers are still occupying the backline.
This translates directly to immediate occupation in the box. By the time of the cross, there are five players there.
That one didn’t fully come off. Others did.
➡️ How full-backs broke the seal
The goal that started the melee began with, you guessed it, a full-back. Both of them, in fact.
Lewis-Skelly got passed into trouble and wriggled his way out, as he does. In keeping a back-four system, PSV is essentially trading numbers in the back for pressure like this. It pays off in the Eredivisie, but it didn’t here.
Now that Trossard has it, there are individual matchups everywhere. With three PSV defenders essentially discarded from the play, the wide area is a 2v2, and the box is a 5v5. Rice peeks out to give Trossard an option.
Trossard has been uncharacteristically off as a ball-striker — which may coincide with fatigue from an every-week role — and he also had a few more bad back passes against United. But in this one, he was phenomenal at this kind of wide playmaking that sets everything into motion. It’s so necessary in a team with an absent Saka and an oft-struggling Ødegaard.
Rice spins and delivers it on a plate with his left foot.
The important matchup there, to prove our point, is Timber. As PSV has a back-four and a man-to-man scheme, he’s got a simple 1v1 battle to win — and it’s against a winger, not a defender: his old friend Noa Lang.
He vaulted over him.
This was not the only time that full-backs were used to exploit the space in the backline.
The second goal, likewise, was after Lewis-Skelly dipped into the half-space to receive and deliver it to his teenage pal.
Calafiori has also had a huge role in proceedings. As Arteta said:
“He certainly has that capacity to flow in areas that, for a defender, is very unusual, that was something that really caught my eye. The other one is that he is very unpredictable, sometimes chaotic, and that can be very difficult as well from the position. That eye for goal, especially from open play, I didn’t have that understanding of the things he could do, but you see it day-to-day in training that he is capable of creating things.”
Here, Calafiori was essentially marked by Perišić, who had to track him into the half-space after a one-two. To put it simply: in a back-five, there’d be another CB waiting in that gap.
Calafiori then beats Perišić 1v1 with some little moves, then fires in a ball for Trossard to chip over Benítez, who had a mare.
Here, interestingly, it happened too. With Zinchenko installed at LCM, Merino dropped to false-nine (basically the Ødegaard spot), and Timber went up to the striker position. This threw off some of the responsibilities, and Gabriel took notice by delivering an aggressive ball into Merino, centrally.
And now, with one pass to Ødegaard, that back-four is overloaded with five attackers — all without the striker even being up there. Ødegaard scored here; one more blocker in the middle and that’s unlikely to have happened.
…and who can forget the Calafiori striker run here?
In the second leg, we saw the sheer ask of a winger marking an Arsenal full-back. Here, Bakayoko has to go all the way inside to track his reception.
And watch that talented winger here, trying to stay within a few steps of the swashbuckler.
Aside from putting pressure on the winger to defend, the essential idea is to stretch and bend that backline, and specifically elongate the gap between the full-back and the ball-side CB. By stretching, or “decompacting,” this line, the space opens up.
We’ve also talked a lot about timing switches to interior runs. It’s one of the low-key things that Ødegaard is great at: making a run right when the ball heads out to Saka, to give Saka a second or two to receive. On the left, that has been a little less present and refined.
Against PSV, with their wingers man-marking the full-backs, Ben White could supply that moment for Sterling. White makes the run in a deliberative way so Kiwior can see it. Kiwior does.
And now the winger is arriving at the scene on a switch, instead of defending Sterling in a settled way. This gives him the moment he needs:
He rips around the edge and hammers in the cross that Rice was not allowed to celebrate by the powerful forces of football decorum.
All these kinds of things were tried against the back-four of Nottingham Forest.
Why didn’t it result in a goal then? I didn’t see a big reason to revisit tactics after that one. That performance was structurally solid and energetic, given the circumstances. Forest are just good, big, and disciplined, and there wasn’t enough pure attacking quality to get a goal in. It comes with the time.
👉 Three other factors
A few other factors played into the difference between the two performances.
1. Pressing as playmaking. Arsenal’s press is often their best creator, but it’s nullified when teams surrender possession and go long. Against PSV, Arsenal could engage in high turnovers.
Timber had one of his best pressing performances.
2. Aerial duels. Arsenal won 17/23 (73.9%) of aerial duels, the top mark of the season. We often have an aerial advantage in the Premier League, but this was even more pronounced.
3. A temperamental sense of directness.
Something I’ll continue to shout from the rooftops: if you believe that your team is fundamentally better than your opponent, you should do everything in your power to increase live-ball time in which they can prove it. I was happy to see free-kicks taken quickly against PSV, for example. This led to a great cross with Gabriel in the box:
There were other, dangerous counters, too.
The opposition understands this all. For example, Arsenal have played 25 games across all competitions where they have been fouled 10 or fewer times. They haven’t lost any of them. In contrast, in the 17 games where they’ve been fouled more than 10 times, they’ve dropped points (or lost on penalties) in 11. There are correlation-versus-causation discussions to be had, but the trend is worth noting.
This sense of directness also appears in possession numbers. It may be statistical happenstance, but Arsenal tend to perform better when they are more measured in possession. Higher possession percentages do not necessarily correlate to better results.
And, of course, there was the simple matter of finishing overperformance. Arsenal scored 7 goals on 15 shots in the first leg of PSV, recording the biggest xG overperformance of the year. Perhaps we should keep our findings in check on this one.
🧐 Breaking a back-five
Arteta has been clear that breaking a low block is a shared responsibility:
“Yes, for sure. That’s the responsibility of all of us, and we don’t want to put that into the minds. That’s fully the initiative that everybody has to take, especially when we defend and attack against so much density. We’ve done it in the past, we’ve done it recently, but we have to do it consistently.”
Initially, Arsenal handled United’s early press well. Raya got involved in the buildup, and Arsenal spread out in a wide 4-2 structure to stretch the first line. Ødegaard and Merino floated on top, finding space against United’s midfield, which lacks the athleticism to press and recover quickly.
It worked. In the first half, at least, United lost confidence in their press, scaled it back, and the game turned into a deep, attritional battle.
Once in settled possession, Arsenal’s shape was familiar but with a small tweak: less emphasis on five-channel occupation. With five defenders already in position, simply filling the attacking lanes wasn’t as fruitful. Instead, movement to drag defenders out of shape became more of the focus.
This is where Arsenal have struggled. The team lacks that final acceleration into attacking spaces. Liverpool, for example, excel in these situations because they can sustain patient possession in the second phase before launching a high-speed attack. Arsenal, particularly before the Martinelli return, don’t have that blistering transition.
Once play moves forward, we see some of the reasons these back-fives can be tough to break down.
The first is that wide areas, and underlaps, aren’t as readily exploitable.
The simplest way to look at it is this. In that freeze-frame when Sterling beat PSV to the end-line for the cut-back cross, he was able to get a step because a wing-back wasn’t already there — it was a winger tracking back from high, and switching off White. Facing United, a real defender would be waiting in the vacant black circle:
In a similar situation for Nwaneri against United, his defender was already set, with Eriksen providing support and the half-space congested with bodies. The same moves that found space against PSV were running into more traffic against United. It was no surprise that Nwaneri struggled more.
It was just so crowded, and the support had support had support. There are just a lot of bodies to get through.
This also goes to show why, anecdotally, you’ll see more shots blocked when facing these blocks, for reasons that are likely obvious. There are just a lot more bodies in the way, and CBs are generally better (and bigger) shot-blockers than full-backs and wingers.
On the other hand, the first leg against PSV had only three shots blocked (and only five open play crosses).
So what do we do about it?
🚧 Working the problem
So we’ve covered some of the challenges when facing a back-five. Among them:
There’s more wide support in general.
It’s harder to generate situations of dynamic full-backs taking on wingers who are tracking back.
CBs sitting in the half-spaces make underlaps/overlaps less effective.
There are more bodies in the way of shots.
With little clipped balls behind the line, there are more players to mop up.
But what are some of the vulnerabilities?
The middle is less supported.
The CBs in the half-spaces may be slow and bulkier compared to their opponents (dynamic 10s or wingers).
If a back-five doesn’t have speed in the wing-back positions, it can be hard for them to generate attack. Especially if there’s only one striker, it can be hard to get out of being pinned.
After all, the 4-4-2 block is what Arsenal employ, and we certainly have a case as the best defensive side in the world. If you see a shape like this, it is almost impossible to penetrate centrally. We remember it too well against Porto.
Then, if you combine that with wingers who take their defensive responsibilities seriously, you have all the advantages of a back-five and a back-four in one shape.
A back-five is typically less reinforced through the middle.
The main vulnerability in this shape, in theory at least, is in the middle. If you can stretch it out horizontally through wide play, then vertically through pinning by a striker, this gap should open. Bruno and Casemiro are a bit like Jorginho in that they’re vulnerable defensively if the game gets huge, but they are so smart and disciplined in times like this.
Still, we saw shades of the right kind of moves here early. You want to commit players, beat one, and then hope that the house of cards falls down.
Ødegaard went to drive between them, laid off a pass, found a dead-end (which is fine — mayhem is an acceptable compromise), and then Merino got the shot off. This is what opportunities can look like.
This kind of central run is how you plunge the back-five into disorder. It’s what worked against Sporting in that demolition — here’s Ødegaard doing something similar to draw a penalty while Sporting were still resetting.
Here’s Timber doing the same from deep. If you beat one man, another has to cover, and the cascading gaps create opportunity.
Late in that one, Timber kept going to create a late shot for Nwaneri. Notice how frenetic it feels in a moment — and how the defenders have to jump out of position to compensate.
Arsenal did a solid job of doing this in the first half of the Manchester United match.
Likewise, the eventual goal didn’t come from some magical structural advantage. Timber took it wide; for all intents and purposes, this was a 3v4 or a 3v5. Casemiro dropped centrally as the line was pushed back.
And Timber simply beat his man. Unexpected idea meets expected quality. This creates a chain effect where other players have to collapse to cover, and that kind of mayhem is where all the goals come from.
In that one, Rice did find one of the structural openings: when things collapse, they can get awfully flat with a back-five, and there can be soft zones ahead of that. Still, it’s hard to shoot through — that’s why you need great ball-striking.
Other than direct take-ons, you have to get creative with moving around the line. I’d usually aim for a #9 who is occupying the CB, and two #10s who are forcing the wide CBs into tough decisions. That worked here against Crystal Palace, when Merino moved to pull Lacroix out of position, which opened up that right channel. Jesus immediately noticed and got ball-side.
Goal. The speed with which Jesus can rip out the edge is important here.
The other idea is to be more aggressive with wide overloads. Simple overlaps and underlaps usually don’t cut it against these shapes because there are sufficient numbers in the opponent's backline. There is one other big element to discuss.
Here’s how the scoring started against Sporting. After a throw-in, there was a pretty aggressive 4v4 on the outside. Timber slipped behind — that’s one thing. But in the box, you see the other.
There are two snappy players, who are at a numerical disadvantage, who both beat their opponent to the ball.
Those two are so key. Bolstering their ranks with another runner in the summer is key, too. While Merino has been about as good as expected at many of his responsibilities at #9 — specifically, leaping, target man play, and some of the quick connective work — that kind of darting, immediate, behind-the-line run is not something he is offering in the range of Havertz (or Martinelli).
You can see it here, too. The front three were outnumbered, but Havertz and Martinelli got there.
…and crucially, Saka is able to gather that scooped ball, which is difficult to wrangle.
That crew is just great in these situations. I’m sorry to make you sad again.
‼️ Other contrasts
When things look genuinely lethargic for Arsenal, it’s often for a simple reason: a lack of adequate activity in the box.
In some of the more dispiriting games, with Timber over at left-back and Partey inverting at right-back, nobody was out wide to provide width for Saka’s inevitable double-teams. So Havertz would swing wide, as he did at Fulham. The problem? Nobody was filling his seat. Here was the result.
Here’s an even starker example.
It is no surprise that Arsenal went 3-of-19 crossing on that day. There weren’t human beings to receive those crosses.
Things started to change in the final stanza of that game when caution was thrown to the wind and players lined up at the goal-mouth. Here’s what it looked like before a Martinelli cross that would have won it, had he not been marginally offside in an earlier phase.
In comparison to those sparse earlier screenshots against Fulham, a settled-play goal often looks like this.
Against PSV, we saw more of that kind of thing. Earlier rotations helped, but a lot of it was a pure commitment to send runners forward. I think Arsenal still have a way to go with the consistency and timing of these attacking runs (sadly, Big Ange is a master of this at Tottenham), but you make your own luck when you push players forward like this.
Here’s one I loved — three runners sprinting through the middle.
More specifically, that felt like a called play for Nwaneri’s trademark shot, where he slinks behind Ødegaard and gets it off. As more teams become aware of his abilities in this zone, you’ll have to get more creative and complex about getting opportunities for him here.
Here, too. Nwaneri does a cut-back, and it’s Rice near-post, Trossard far-post (eventually), and Ødegaard cleaning up at the spot.
The other thing that happened was just a willingness to lose the ball and battle for second balls up here.
Which leads me to another point. The Athletic had a wonderful piece — English football is besotted with second balls – but how important are they? — which should provoke some interesting internal discussion.
We’ve seen our share of injuries and red cards this year. With skeleton squads, intensity has often been turned down to preserve players. That said, with how this team has been built, you’d still like to see that second-ball number higher.
My general feeling is that Arsenal have gone a little too aerial from deep. It’s a great option for a reason: you have the players for it, you are good at creating winnable situations, and you want to create a little chaos — which is hard otherwise.
Merino’s been great on aerials and can do even better. There’s a stat out there that demonstrates “first contact percentage,” or how often a player makes first contact on a launch or a corner kick; I don’t have access to it at the moment, but I imagine he’d fare quite well.
His duel created the situation for that first Zinchenko goal.
The problem with a potential over-reliance on launches, though, is that it adds a bit of variance to the game — and you’re just not going to win them all. After Merino had done so well, he didn’t get this one. Within two passes, it was a goal the other way.
I think the days of ideologically “playing out the back” or “going long” should be over for Arsenal. A team at this level should be good at both, and do both as the situation requires. I prefer a situation in which there is a slightly higher preference for playing out the back, and long balls are used when there are more ready advantages (numerically or qualitatively).
👊 So what are the takeaways?
A key strategic difference when attacking a five versus a four is choosing when to play direct (vertical) and when to shift the play (horizontal). Vertical progression (quick forward passes, through-balls, runs in behind) can be effective against a back-four because there are fewer defenders to cover depth. A single well-timed run behind the line or a splitting pass between two defenders can result in a clear chance if the opponent’s back-line is stretched or caught pushing up.
However, against a back-five, a purely vertical attack will often hit a wall — the spare CB provides an extra cover player to mop up through-balls, and the wing-backs drop off to track runners. Simply trying to thread balls straight through a 5-4-1 low block is low-percentage, as Pep has observed.
If you pull the wing-backs out, though, you perhaps can drive through the midfield. You’ll always hear calls to “make the pitch big” when attacking, which is important against any shape, but especially here.
There are a lot of similarities in facing a tough back-four and back-five. The characteristics and intent of the players matter more than any numbered formation. But some of the takeaways from this:
Against a back-four, pure passing/overlaps/underlaps can create more advantages, and clips behind the line and quick switches are more likely to result in overloads.
It’s also easier to isolate a full-back on a winger and look to win on quality against a back-four. This may be one of the primary reasons for hope against Real Madrid.
Immediately throwing that full-back up high has compounding effects. Ødegaard can drop and thread play, and then, once the line is broken, numbers can join the box more quickly. There was for sure some statistical noise to that 7-1, but that sense of commitment to box occupation is the biggest thing I’d like to take away from that performance.
Against a back-five, more patience can be required. You have to keep probing for the moment to attack centrally, which is more likely to open up if you do your job. You can increase your odds with a pinning striker and a #10 floating on the same vertical line.
Against a back-five, it seems more important to gain a commitment through an immediate dribble to start a cascading effect.
Because your striker is likely to be outnumbered, having one who can hit those quick, darting, poke-in goals is even more important against a back-five.
Ball-striking, smart runs, and audacious dribbles can win in any situation. It’s especially the case here.
The other option is same-footed wingers (i.e. right-footers on the right wing, and vice versa); because full-backs don’t provide as much of an advantage wide, you can use them to drive and dribble centrally, while the wingers stay on the touchline, sprinting and hammering crosses.
I was drawn back to this BreakingTheLines post, which covered some of Pep’s attempts to break through Newcastle’s 5-4-1 in 2021. I remembered this: Ferran Torres proactively trying to bait the line back, with KDB lurking in the spot just below. This puts the CBs in a tough decision of whether to push forward or not.
Could Martinelli play that role, with Ødegaard floating central?
🔥 Final, scattered thoughts
As I said, a big motivator behind this piece was to retain curiosity about some of the little details of the season and not chalk everything up to other factors, as tempting (and true) as that is these days. The reality is that most of the difference between the results at PSV and United came down to physicality, finishing variance, and space.
But it did once again drive home the importance of full-backs in the current system. So many opportunities were routed through them. Against back-fours, this is through wide combination play, underlaps, overlaps, and winger isolations. Against back-fives, first by stretching the opponent’s shape, then driving at it as a dribbler.
Moving forward, you’d like somebody on the left who can approximate some of what Saka does: take on one or two defenders on their own, and win the spot. Less should hinge on the full-backs.
The current Arsenal squad may be better suited to frustrate top sides than it is to wallop more overmatched, defensive ones. PSV fell into a unique category: they play more like Manchester City than Newcastle, but the qualitative difference is much higher. It’s not a coincidence that Arsenal went 5-1 against City and 7-1 against PSV.
The other thing that stood out was the importance of that final darting move and the tap-in. If so much of the play is taking place in the wide areas, you need somebody to stick a leg on the end of a cut-back and put it in. Havertz is sorely missed. To replicate that quality, the best imitator on the squad isn’t Merino — it’s Martinelli. He is extraordinarily quick at reacting to balls in the box like a striker. It’s been a failure across the board that he’s only been able to generate 1.75 shots per 90, especially when his ball-striking looks good again.
I’ve joked that this is how I want to line up, particularly at the Bernabéu, but of course, I’m not really joking.
After all the frustrations against physical blocks this year, it would be poetic to turn into the most overpowered version of it.
Do I find a run of Martinelli at #9 likely? I don’t know, I don’t think so. But I know this back-six (plus Raya) can beat anybody in the world. The front-four needs some fresh ideas.
Some other notes:
Here’s another thing I found interesting, via PFF: Nwaneri is in the highest tier for opening his body with the ball.
Source: PFF He’s also getting tackled at some of the lowest rates in the league, while his finishing stats are near the top. Timber has really stepped up his attacking output of late, but White may still have him beat in the wide-wing work. Anything that pushes Nwaneri in and gets him those turn-and-shoots is good by me.
It will also be interesting to watch the placement of his losses moving forward. Outside of pure wide take-ons, where he’s so secure right now, he’s a little less delicate with it in central areas than Arsenal players typically are. That can lead to chances — or counters the other way. Will it be trained out of him, or will he continue to let it rip? (I vote let it rip.)
His defensive intensity and positioning are improving every day. The biggest delta right now is his quick interplay — he’s still forming associations with players like Timber, Ødegaard, Merino, and Rice, and the timings are a little off. It’s impacting him as both a receiver and a sharp creator. There’s potential to unlock still.At some point a few months ago, I said Arteta’s primary challenge moving forward was turning Declan Rice into a player who is universally viewed as a top-10 player in the world. I thought that could take longer than it has, but ever since 2025 rolled in, he’s been that. He’s a horse, yes, but more accurately: a unicorn.
I’d rank the potential usage of full-backs in midfield as follows: Zinchenko, Lewis-Skelly, Calafiori, Timber, White. With previous editions of Arsenal, Zinchenko was most sorely needed in the deeper areas. With the creative slowdown, I don’t think his usage in midfield should be a passing fad. The thing to remember is the defensive block—against PSV, for instance, you wouldn’t have wanted him to play LCM, because then you have a defensive pivot of Jorginho and Zinchenko. He may be best just spelling Ødegaard for the last 20.
Calafiori may be one of the most unique players in the world.
So happy to see Lewis-Skelly’s call-up to the England squad. There have been some injuries that sped that timeline up, but when you see the ridiculous quality on the U21 team, you’ll remember how meaningful it is to bypass that step. I can’t imagine a more eventful first season than this — getting a yellow before his debut, scoring a huge goal against City, all the bullshit (loaded) punditry, and some huge learning moments. One of the more interesting little learnings for me? He’s fast. I was watching all the players track back at full speed, and he showed a gear I didn’t really see in his youth tape. I think we’ll see him a good amount at both left-back and midfield next year.
We talk about needing wingers and strikers and midfielders a lot. One quality the club hasn’t quite valued enough, to me, is “spray passing” — those whips, curls, lifts, whatever, from anywhere on the pitch: this can be from CBs, FBs, midfielders, or wingers. It’s a quality I’d really like to see added this summer, most likely from the wing position. Nico Williams does have that in his locker. It’s also a reason why Adam Wharton could be so attractive.
Saka on free-kicks when he gets back, please. Until then, Rice.
More quick restarts, too.
Much content is to come. I’ll do a deep dive on new sporting director Andrea Berta, and may even finally rename this thing. I’m also working on a longer opposition report on Real Madrid before that tie, before getting started on some of the rankings for summer targets.
The Chelsea game is up tomorrow, and all possible results are on the table. Arsenal are structurally and defensively superior, and marginally better suited to face an outfit like Chelsea these days (with space and the ability to press), but we all know about the issues with firepower. It is what it is.
I’ve shared my frustrations with January. Whatever the challenges of the year, there’s been plenty to cherish, too. Rice has ascended to the highest level. Raya has proven himself to be one of the best keepers in the world. The back-two is impervious as ever, with Gabriel especially taking another step, and new full-backs are bedding in nicely. Hale End has spawned two geniuses at the exact right time. The structure and level of intensity have been worth celebrating, led by one of the best managers in the world, who has kept us in second despite everything. The final additions to the squad may be difficult and expensive, but at least they’re obvious.
In Chelsea, Fulham, and Everton, we’ve got a run of three straight matches that will test some of these principles to the hilt — and specifically, test the problem-solving of the remaining team to find goals. It’ll be difficult and illuminating.
Then, a Champions League quarter-final against Real Madrid.
What a world.
“And you, to whom adversity has dealt the final blow
With smiling bastards lying to you everywhere you go
Turn to, and put out all your strength of arm and heart and brain
And like the Mary Ellen Carter, rise again.”
— Stan Rogers, The Mary Ellen Carter
“Then, if you combine that with wingers who take their defensive responsibilities seriously, you have all the advantages of a back-five and a back-four in one shape”
I mean, fair enough
Fantastic, as usual. If you’re looking for name suggestions, “Trattoria Berta” sounds nice.