Time and space
On "the next five actions," the hidden layers of football athleticism, what happened against Liverpool, Arsenal’s new midfield geometry, box arrivals, fresh defensive tweaks, and the Return of the Kai
Alexander Fleming was known for many things as a researcher, but tidiness was not one of them.
In 1928, while working at St Mary’s Hospital Medical School in Paddington, he was studying staphylococci, bacteria responsible for many of the wound infections he saw firsthand in World War I. His was one of those small, cluttered bacteriology labs typical of British medical research between the wars.
But he was not the reclusive shut-in one might associate with a leading mind. As covered in The Man and the Myth, Fleming enjoyed chess, snooker, and cards, and was “always ready for a gossip.”
If no one dropped in to see him, he would wander into the main laboratory for a chat. This usually involved Fleming planting himself in front of the fire, with his hands in his pockets and a cigarette dangling from his lips, while he looked more or less into space.
(It was at that point in the book that I wondered if the author was describing Fleming, or my father.)
On rare occasions … he would utter a few words on the marriage of a colleague, the state of the stock market or some current scientific scandal. But, one morning in the autumn of 1928, he had something to show anyone who was interested. This was a culture plate — one that he had rescued from being discarded. It showed a blob of mould, the growth of which clearly disagreed with the staphylococci that covered the rest of the plate.
Fleming’s untidiness, that dopish, relatable humanity that still exists in those we call geniuses, would prove fortuitous. He often stacked his Petri dishes or left them uncovered for longer than today’s standards would allow. The late summer of 1928 was no exception. While he was away on holiday, the story goes, a breeze blew through the open windows and allowed airborne spores to settle. Returning to inspect the dishes, one plate made him pause and, according to legend, say his famous line: “That’s funny.”
He showed this culture plate to his labmates to little fanfare or reaction, but he had his own suspicions. He saw something potentially significant: the fungus was producing a substance capable of inhibiting a common pathogenic organism.
Spending months calling it “the inhibitor” or, my preference, “mould juice,” Fleming eventually settled on its formal name: penicillin.
Penicillin, in other words, was the byproduct of a mess and the man who made it.
😤 Mistakes, and what happens next
“You can’t keep blaming yourself. Just blame yourself once and move on.”
— Homer Simpson
The evolution of Gabriel as a defender hasn’t really been about technical or physical characteristics. It’s mostly been about modulating his aggression and ironing out mistakes.
In 2022, Arsenal were comfortably outplaying Fulham with a lineup that often featured an Elneny–Tierney double pivot. Gabriel then accepted a difficult delivery from Saka before succumbing to the pressure of Aleksandar Mitrović.
He was stripped of the ball, and Fulham went ahead.
Gabriel promptly pulled his shirt over his face.
“After I lose the ball, my head is down,” Gabriel said after the game. “But I look at my brothers, they say, ‘Gabby, let’s go, let’s go!’, and I put my head up.”
Arteta also singled out the roar of the home crowd.
“The reaction of the stadium straight away gave support and belief. Then the player has to show courage to continue playing.”
“This is part of football, mistakes are part of it and it’s going to continue to happen, but it's about how you respond to that and I'm really pleased. It's not a coincidence with Gabi because the way he's changed his mentality and the way he approaches every single day is different to a year or two ago.”
In the eighty-sixth minute, he soared in the six-yard box and then swatted home the rebound. Arsenal won the game 2-1 to stay atop the table.
Mistakes have been much fewer and farther between in the years since, but it wasn’t his last time singing the redemption song.
Earlier this year, he was responsible for conceding a Nick Woltemade header after losing touch contact with him, then trying in vain to draw a foul on a defensive corner.
He rectified his error at 96’.
Pancaked on all sides, he rose above.
Against Bournemouth, he had an early howler, delivering a belated Christmas present for his compatriot Evanilson.
But we had our history to fall back upon. My pal Alex on Twitter reflected on Gabriel’s previous exploits and immediately said: “Don’t worry, gooners. Goal incoming.”
The inevitable only took six more minutes, thanks in large part to Madueke.
Now that’s what I call mould juice.
I guess it’s unsurprising for Gabriel to help find the net in any situation. His goal contributions (G+A per 90) are higher than whichever forward you feel like bantering today: Matheus Cunha, Jørgen Strand Larsen, Jarrod Bowen, Beto, Igor Jesus, Liam Delap, Florian Wirtz, etc. It’s even with Viktor Gyökeres, and when I originally researched last week, it was ahead of Sesko and Mateta. (I also think Gabriel might have been short-changed an assist against Palace for Eze’s goal.)
Errors are a big deal in this sport. Virtually every goal is preceded by some kind of error, and it’s pervasive enough for The Athletic to devote an entire edition of the Alternative Premier League Table to the topic.
Liverpool’s only real sight of goal was the result of a mistake. Saliba thought Raya was gonna be closer to the goalmouth, and rifled a pass to him with the appropriate pace for that distance. But Raya was coming out.
Opta classifies the bigger miscues, and when I pulled the numbers after Matchweek 20, it called into focus just how bankable these chances are.
6.6% of shots come from errors (319/4,848)
16.3% of goals come from errors (91/558)
Finishing efficiency is tripled after errors: 28.5% vs 10.3%.
In other words, about 1 in 6 goals are scored off clear errors, even though only about 1 in 15 shots come from these errors. One glance back up at Evanilson’s goal, that 0.73 xG chance gifted from the heavens, can help explain why. (Now that I look at it, Gabriel is also beating Evanilson in G+A/90).
So you can look at this pretty simply. A key part of any game plan should be forcing errors from your opponent. Then, on your side, there are two parts: first, prevention (avoid mistakes if at all possible). Second, treatment (bounce back as quickly as you can when you do).
Luís Campos, the football advisor for PSG who has as strong a claim as anyone to being the best sporting director in the world, has recently done us the immeasurable kindness of sharing some of his thinking publicly.
What becomes abundantly clear is that he takes a long, careful look at a player’s psychological makeup. That’s an area prone to quackery and bias, so he applies a systematic lens to it. Here’s a slide he’s shared on the challenges faced by the players he evaluates.
You’ll see cues in there to both avoid the mistake (by making good decisions, filtering noise, scanning), and rebound from it (through emotional control and resilience).
From there, he refines it into a four-step score that helps quantify these “cognitive indicators.”
That fourth box adds context to what we’re discussing today, namely:
How are the next five actions?
How is the body language?
A nice recent example is this Piero Hincapié sprint against Villa. He emptied the tank to arrive in the box in transition, just in time to pratfall:
…but he got right back up again to pressure the ball and help reset the press.
After giving himself a second to clap his hands in frustration, he was the one who ultimately regained possession off a Gabriel header.
…and that possession, and Hincapié’s left-sided width-holding, resulted in the Trossard goal.
I’ll always think the practice of meditation and mindfulness is helpful for these moments, because they give us the tools to live in the present instead of regretting the past or worrying about the future. Martin Zubimendi credits chess.
“It’s the concentration more than anything. It’s a game in which you can’t give away pieces, you can’t give away any squares, and you have to be focused all the time. In football, it’s the same.
“In the Premier League, you get distracted for a second and they (the other team) already score, which happened to me the other day (for Tottenham forward Richarlison’s goal in the north London derby).”
But as one noodles around for interesting research on how high-performers can get their minds right after an error, one inevitably will come across the stuff from Carol Dweck. In books like Mindset, her research suggests that the crucial difference is not whether mistakes happen, but whether they are treated as evidence of limitation or as information for the next attempt.
I also came across some background from Solomon and Becker, who laid out five principles of performance excellence in athletes. The fifth principle was resilience, that word we saw in the Campos slides.
Simply stated, resilience is the ability to remain composed, confident, and consistent in the face of errors. A resilient athlete is one who can let go of errors and return to the present; s/he uses the error as an opportunity to learn and improve.
They recommended a four-step process for dealing with mistakes.
A = Acknowledge the error and the frustration it has caused
R = Review the play and determine how and why the error occurred
S = Strategize a plan to make the necessary corrections for the future
E = Execute and prepare for the next play
The ARSE method.
Contrary to what your brain is telling you right now, I did not just make that up.
🤔 What is football athleticism?
Athleticism feels easy to read. That’s especially true in basketball, where nothing is subtle: it’s a game of enormous humans with superhero muscles jumping out of the gym. (For reference, Virgil van Dijk is smaller than the average NBA player.) Many come to know the sport’s imagery before the sport itself. Stamped on shoes across the world is a spring-loaded Jordan silhouette suspended in midair.
This can lead to a visceral adjustment period for newcomers.
“Back in France I was super athletic,” said Zaccharie Risacher about his pre-NBA career. “And now I’m just a regular dude.”
It’s tough out there.
Anyway, here’s the best basketball player in the world.
When he’s not tending to his horses, Nikola Jokić is winning three consecutive MVP awards and making seven All-NBA teams, all while looking fairly unimpressed by his exploits. His lack of obvious physical bona fides, other than size (which is important, to be fair), caused him to be overlooked by scouts, going 41st overall in the draft.
Why?
“Jokic has the lowest vertical jump of any NBA player we’ve ever assessed,” said Dr. Marcus Elliott, founder and director of the Peak Performance Project (P3), which has assessed two-thirds of the NBA at the time of the interview.
He’s now regarded as the greatest steal in draft history. In a game last week, he had 56 points, 16 rebounds, and 15 assists. If you don’t follow basketball, it’s like if a basketball player had 56 points, 16 rebounds, and 15 assists in a game. I’m not good at analogies.
The scouts weren’t dummies. They were just human, and valuing the wrong things. What’d they miss?
“His movement quality is really, really high,” said Dr. Elliott.
Jokic is what they call a “Kinematic Mover”, basically like a Swiss Army knife. “He has the right answer for every job from a movement standpoint.”
He does display superlative core strength, balance, and coordination. But more than that, he’s got a Rodri brain, doing a bit of everything, racking up touches on the ball, patiently guiding the whole picture, forcing the game to a tempo of his choosing, showing what we’d call pausa. He always seems to be a step ahead in processing information, and never wastes a movement.
Jokic is not alone in being undervalued. Another future MVP, James Harden, was widely seen as an unexceptional athlete: not a high-flyer, fairly languid, no beach muscles. But what he did have was exceptional.
[Harden’s] traditional performance numbers – the ones that get discussed at the Combine – were unremarkable. His metrics related to Deceleration, however, were beyond elite. Eccentric Force generated during the CMJ, the rate at which he generated this force, the kinematic values associated with his eccentric phase movement, and forces produced during change of direction testing…all broke our models. James has some elite athletic tools, they just aren’t the ones that we were accustomed to looking for.
The same is true of Luka Dončić, the NBA’s current leading scorer. “Our athleticism metrics have him off the charts and nobody thinks he’s an athlete,” Elliott said. “Luka jumps lower than your average NBA guard. He doesn’t jump very high, he doesn’t run very fast. But he stops better than almost any other athlete we’ve ever assessed.”
These benchmarks help clarify their superpowers.
We may have a fundamental misunderstanding of what athleticism even is, and more specifically, what traits are actually most important for a particular game. The reality is that players like Jokić don’t succeed despite their athleticism, but because of it. It’s our definition of athleticism that is the problem.
This complexity led Ben Taylor of Thinking Basketball to create this view of “basketball athleticism.”
Among other things, I love it when cognitive elements aren’t separated from our appraisals of athleticism. If somebody sees the game especially quickly, they are fast.
In the NBA, great athleticism can hide behind “bad” physical markers. Football is no different.
👉 Deceleration
One thing was easy to confirm after diving into more research: deceleration is awfully important in football, too.
Picture a front-foot player in a modern pressing system, and how many of the responsibilities revolve around the simple act of stopping.
N’Golo Kanté, who was known back at Suresnes as chouchou, or “little pet,” was underrated in his youth for his famously short frame, but it wasn’t long before he was universally celebrated as a cheat code of a player. I remember seeing a YouTube comment that said, “The best double-pivot is N’golo and Kanté.”
You’ll hear him get ample credit for his relentlessness and work-rate, but one truly special characteristic was his stopping power, which was helped along by his height. Look how there is no bobble, chopping, or extra steps here. He just … stops.
Controlled deceleration helps you defend closer to players, because when you rush up to them, you can be confident of your inertia not bringing you all the way through.
There’s nuance, especially with and without the ball. Somebody like Noni Madueke is sublime at controlling the speed of a dribble or a carry, toying the defenders into heavy steps. As a defender, if he gets his big body going, it can take more force to slow it down.
He’ll update some of his angles and be good to go; I’m pretty bullish on his prospects out-of-possession. Martinelli offers the flipped equation, as he’s more rigid when changing direction with the ball, but whip-fast without.
Somebody who has (had?) both is Sadio Mané. In one simple press-win, look how many different changes of direction there are: stops and starts and jinks and sprints.
If he were a millisecond later, or slightly clumsier in stopping, there’s no goal for Salah.
The stopping-and-starting, combined with world-class levels of balance and dexterity with a ball, was a superpower of Andrés Iniesta.
“What happens is that Andrés brakes,” said Lorenzo Buenaventura. “That’s the key, the most important thing. People say: ‘Look how quick he is!’ No, no, that’s not the point. It’s not about speed, about how fast he goes; what it’s really about is how he stops and when, then, how he gets moving again.”
I think that’s a good enough excuse for me to post this clip, yeah?
That clip shows the symbiotic relationship between physical traits and quality on the ball, which produces a kind of one-hand-washing-the-other effect.
A player like Manuel Ugarte has plenty of physical ability, but not always the time to use it because he’s always chasing the ball.
On the other hand, plenty of “technical” teams have been promoted to the Premier League in recent years and been overwhelmed; this year’s group has fared better because players like Granit Xhaka or Anton Stach give them the physical base to express their technical quality.
Regardless, the importance of stopping like Iniesta shows up in the data. Tracking data shows that players perform far more high-intensity decelerations than all-out sprints during matches, and those decelerations actually show up more often in games that teams win than in games they draw or lose. So it’s much more than just a defensive or injury-prevention thing. It clearly helps you score goals and emerge victorious.
It’s taxing, of course. Slowing your body down quickly places more force on muscles and joints than accelerating. Peak braking forces can reach several times a player’s bodyweight (~5.9x at peak, according to that last study). This, then, produces a wider range of outcomes, everything from tackles to stumbles to injuries, which just ramps up the importance.
👉 Bending and curving your runs
Deceleration then meshes together with other attributes. Here, Kanté beats Sterling to the spot, then closes out Silva, then curves a run around to win the ball near the box.
The player who probably comes closest to showing that ability now plays for Arsenal.
When he started at right-back against Brighton, I was worried Declan Rice was going to get sucked out to the wings too much, and leave the middle open for transition. No worries. He just did both.
To eat up that ground and quicken up in the medium distances, all while running on curved lines, is rare.
👉 Speed and agility
Like Kanté, though, Rice isn’t necessarily winning a 100m dash. Being good at angling and turning these runs is not the same thing as being fast in a straight line.
Studies consistently find weak links between straight-line top speed and change-of-direction performance. This makes sense when you consider how it may be related to balance and where your gravity lies. Tall players often have powerful, long strides. The height of Olympic sprinters, for example, is on a steady trajectory upward.
And here’s the fastest recorded sprint in Premier League history, by the 6’4” Micky Van de Ven, which didn’t look particularly difficult for him.
Force–velocity profiling shows that players reach top speed in different ways. Some accelerate quickly (via force), some build speed gradually, and some peak later in a run. Two players can post the same sprint time but arrive there through very different mechanics, with real consequences for fatigue, injury risk, and which spaces they are actually best at attacking.
The layers of this are seen in Viktor Gyökeres, who doesn’t possess the most immediate burst or stopping power, but builds speed over medium-to-long distances. Out-of-possession, this helps him generate pressures and rushed long-balls for opposing goalkeepers, but it doesn’t turn into advanced tackles, which require more stopping power and agility. In possession, this means he looks best when running the channels.
Somebody like Madueke is rare because he’s quick at every phase of a run, whether that be short, medium, or long.
The mixture of height and agility is pretty unique. Most agile on-ball players tend to be closer to the ground, able to zag through traffic rather than just run past it. One study of elite youth players summed it up this way:
Elite players demonstrated superior performance in cardiorespiratory fitness, countermovement jump (CMJ), 10-metre sprint, and agility compared to sub-elite and non-elite players.
Pep Guardiola has always been careful to give our earlier darling, Iniesta, his due on these qualities.
“Even in a midfield where he’s surrounded by countless players, he chooses the right path every time. He knows where and when, always. And then he has this very unique ability to pull away. He pulls out, then brakes, then pulls out again, then brakes again. There are very few players like him.”
👉 Efficient, anticipatory movement
Efficient movers buy themselves time. Sergio Busquets made the same point from a defensive angle.
“Defensively, positioning is everything. More than stopping the counterattack when it happens you want to prevent it from starting. It’s more about being tactically astute than physically dominant. You need to know exactly where to be on the pitch at all times.”
We have one of those Spaniards to ourselves now. Before Martín Zubimendi signed, I went looking for clips of him being run past in transition to understand how much risk there might be, and I couldn’t really find any. As I wrote at the time:
Zubimendi not in the highest tier of length and bursty athleticism? True. How often do you actually feel it, even against players like Doué, Mbappé, or Neto? Very little.
His game features gobs of moments like this, where, based on angles, he makes a probabilistic guess on where the ball is going to go, then uses that extra step to win.
Watch again: as soon as Eze cuts off the in-ball, Zubimendi sneaks a step to the obvious pass. He’s always sneaking steps, showing “cognitive athleticism,” because you cannot separate mental factors from physical ones.
Readers of the newsletter will also know how often I harp on the idea of decision fatigue. Studies on the topic show that players can maintain their running output even as their decision-making and technical execution quietly deteriorate. Per one:
Mental fatigue had an unclear effect on most physical-performance variables but impaired most technical-performance variables.
Mentally tired players can forget their technique, and I do often wonder if Arteta’s system is unusually mentally taxing. There isn’t much dividing up mental and physical responsibilities. There’s everybody lock in, every phase, every moment of every game.
👉 Symmetry
There are also interesting effects of asymmetry to all of this. If a player is a little stronger or more powerful on one leg than the other, even by about five percent, they tend to jump lower, sprint slower, and stop less cleanly. Is it any coincidence, then, that players like Kevin Schade and Antoine Semenyo, who are among the most two-footed of players, are also some of the best aerial presences at wing?
That makes ambipedality go well beyond kicking the ball. It can be a marker of wider athletic capacity. Look back at Zubimendi, who is a balanced, anticipatory, two-footed player, and is miraculously great at headers despite being shorter and an unexceptional jumper. Ambipedality means balance and control, too.
It also means that the direction you see a player change can have a real effect on how good they look doing it. In other words, you can get false positives about how quick or balanced a player is if you’ve only seen them go in one direction.
This all points to a simple mismatch between how athleticism is often discussed and how the game of football actually works. Timo Werner and Adama Traoré will catch our eye. Many players do not look so fast or powerful, yet keep arriving early, staying upright, and making clean decisions under pressure. We tend to call those players “technical” or “intelligent,” when it may just be athleticism expressed differently.
Ultimately, it boils down to a simple question: do you have the capacity to manufacture more time and space for yourself, and deny it to your opponent? In sum:
Rapid, controlled deceleration generates time for players. Defenders will overrun them, regularly providing an extra beat. The player’s own defensive pressure can also be closer to their mark.
Sheer comfort on the ball can’t be discounted from all this. If a player is chasing their touches, they won’t have time to show off their physical qualities.
The appearance of core strength and balance is often the result of technique. Still, physical strength helps players absorb contact without losing body shape, and have more control and stability (time) when changing direction.
Running style (force vs speed) determines whether a player wins the first step, the middle of the run, or the longer exit, shaping where they can actually create separation. This helps match players to roles and play-styles.
Symmetry between legs makes braking and turning more reliable in both directions, so players do not leak time. It also has wider athletic implications.
“Cognitive athleticism” is like a multiplier on all other characteristics.
There are all kinds of other little topics (leaping, leaning, intensity, endurance) that we can investigate more later.
Another point. Instead of wishing away a player’s trade-offs, maybe we should come to appreciate them. A “taller David Raya” is probably not as agile as David Raya. A quicker Gundogan or Jorginho probably never evolves to be Gundogan or Jorginho; a taller Kanté can’t stop so well and is unbalanced more easily. A blazing-fast Saka may feel less pressure to make so many good decisions.
It’s all about space and time.
🎯 Pressure points
Now, mercifully, it’s time to talk about Arsenal in the Arsenal newsletter.
The Arsenal haven’t lost in more than a month, but that hasn’t stopped the #discourse. I say that without critique: it’s been a lot to unpack. Among other matches, a dreadful performance against Wolves resulted in a 2-1 win; an uproarious, domineering attacking performance against Palace was level 1-1 at full-time; the xG gods repaid debts at home against Villa in electrifying, 4-1 fashion, that still had some nuance under the surface; there were stumbles in a tough win away at Bournemouth; and most recently, plenty of consternation after a goalless draw against Liverpool, and plenty of untidy, joyful, shithousing merriment away at Portsmouth.
First, I’ll dive into some of the things I see contributing to the tougher moments, and then outline a few promising developments.
👉 Where you lose the ball
In an ideal world, you progress the ball into a final-third cage that looks like this. By back-footing the opponent and retaining ideal rest structure, it’s almost impossible for the opponent to play out. You’re able to attack again and again.
With such a structure, losing the ball in a central area can turn into a positive. In the early days of the season, Nwaneri went to dribble and combine early, but a superhuman Rice cut across, a rest shape stopped any outlet, and the ball was won. Saka then scored.
That’s a “ball loss” and a “ball in the back of the net” separated by about seven seconds. It’s the heart of gegenpressing.
Ideally, there’d be an even more central structure to the midfield pressing here, but Rice makes up for it with his (wait for it) stop-start athleticism.
That’s how you want to lose the ball.
These are not the ball losses that have been plaguing Arsenal of late.
This is the kind of ball loss in build-up that has led to some compounding issues.
We’ll get back to that one.
Ultimately, it hasn’t been confined to any player or situation. Another example:
And another.
And another, from Gyökeres. 😬
And, more strangely, this from Zubimendi.
A lot of these have been from new connections. Some of it is less obvious. Arsenal lost the ball 26 times in low areas against Bournemouth, which was even with the Aston Villa loss for most of the season. I’d usually rush to credit the well-coached Cherries side, but watching it back, a lot of it was unforced, and it didn’t look especially tactical. Our players were just fucking up.
It’s been a really busy period of matches, and some mental fatigue would be understandable. But there’s another thing. Since Ødegaard’s return, Arsenal have looked to keep the ball on the ground more. There were eight matches before December in which >10% of Arsenal’s passes went long, and there have been zero since. So that means more building out the back, and more pressure to perform in these situations. I think that’s a welcome change, but one with an adjustment period.
It also reminds us of why the location of a ball loss matters. If you lose it up the pitch, the distances are closer. You squeeze into a counterpress, and it may even turn into a goal. If you lose it in a sloppy giveaway in build-up, you are unlikely to have any structural cover, with the team splayed out across the pitch. Often, all you can do is delay the ball carrier as best you can and fall back into your lower block.
When you fall back into such a block, your job is not really to press or intervene. It’s to move as a unit, block lanes, and be patient. The opponent is counting on you to step out and create a gap, and you mustn’t take the bait.
That’s how you end up in low-energy situations like this. This period against Wolves was an especially bleh example, where patience almost manifested as disinterest.
And that was what turned the game against Liverpool.
We can see how the amount of defensive actions fell off a cliff in the second half.
And this shows how the possession story was a tale of two halves.
Liverpool had 66% possession from 46’-75’. This resulted in 2 shots, 0 of which were on target. In their entire possession-dominant second half, they had zero shots by anyone other than Dominik Szoboszlai. It was a classic case of “defending with the ball.” And it was a brilliant idea, because Liverpool needed to preserve their legs.
It was less about Arsenal’s defensive intensity, per se, than the ability to show it. The ball kept going back to Liverpool in situations that should have been managed.
Now we return to the real case study. Arsenal easily found some gaps in build-up and got Lewis-Skelly some space. But he and Zubimendi were on different pages about what happened next, and promptly delivered the ball back to the opponent.
That pass was made at 76:00. You can delay the carrier if you’re near the ball, but otherwise, everybody else’s role is just to sprint back and fall into a lower midblock, and be patient from there. That’s what Arsenal did. They didn’t get another kick of the ball until 78:06.
Losses like that turn into sitting back, corners, shots, and minutes without the ball. It also turns into a silent killer of teams that want the ball: clearances.
Could Rice have controlled this? Probably not. But whacking it results in another minute, or two, or five without possession.
It’s what happened against Wolves, too.
This is a subtle thing that is missed when Calafiori is gone. He is always aiming clearances, and seeking to play out instead of just getting rid of it.
I think that was the major story of the second half. In reality, it’s very difficult to nab the ball from a team like Liverpool if they’ve shoved you down the pitch, and don’t seem particularly interested in taking attempts on goal. They can just pass it around, waiting for you to jump out.
But it wasn’t the only story.
The other story was about winning the duels in the middle third. Here were the first half recoveries:
And here’s the dramatic difference in the second half.
This came in stark contrast to the 4-1 win against Villa, which gave me some flashbacks to a 4-1 win against Newcastle a couple of years back. In both, a Gabriel goal opened the seal, and then it was a press-fest from there. Look at all the high central wins in the second half:
It shows how getting pinned in can compound, as the pinning team can have better structure for regains in loose-ball situations.
Liverpool competed well. Arsenal weren’t actually too passive in the bouncing balls, they were just often second-best in the second-half.
I’d attribute a large chunk of the problem to cheap giveaways. This was also correlated to two changes currently in process: a less anchored, more rotational midfield, and a desire to attack more directly.
👉 Little out-of-possession vulnerabilities
All in all, you can’t hang your defensive head too far when holding a team like Liverpool to zero shots on target. But you also can’t neatly separate the phases, and there were some frustrating aspects of that performance that we hope to see get ironed out.
Elsewhere, you’re always reminded that the price of the brick always goes up in this league. With so much money and so many analysts poring over every second of footage, if one team finds a little gap in your approach, you can bet that everybody else will know about it.
Last time, we talked about how Chelsea found some ways of exploiting Arsenal’s “over-responsibility” in the press. Declan Rice had double coverage, and Chelsea would reset and play it wide-right, splitting the CBs incredibly wide to elongate the space.
These moments forced players like Hincapié to jump up (from LCB), Rice to commit, and the other high midfielder (Eze) here to switch over and cover the other free midfielder.
But it’s too hard to do all this immediately, and cover runners behind your back, all at once, particularly with new teammates who are learning the press. This led to the early Zubimendi yellow that established the tenor of the game.
Maresca went back to that well many times.
It’s a touch different, but we can see how many of the outlines existed in Liverpool’s best build-up attack against Arsenal, which resulted in Bradley’s shot off the crossbar. Get your CBs wide, get 3-4 players on the touchline, drag the LCB up, free a FB+CM in the blindspot, and go go go.
It’s not the only place Arsenal have been a bit vulnerable, at least relatively speaking.
We’ll come back to this screenshot later, but for now, you’ll see how both Frimpong (right-wing) and Gakpo (left-wing) have followed the full-backs into the half-space, essentially man-marking. Over the last couple of years, it’s more likely that this was handled zonally.
Jon Mackenzie of The Athletic has talked about how clubs are revising some of their out-of-possession approaches in this fashion.
Teams are finding ways of exploiting the typical 4-4-2 zonal look, specifically by drawing out the front-two from the midfielders, then dropping somebody in their cover-shadow, then giving them players to combine with. The master at gaining this central access is Unai Emery and his never-ending glut of midfield occupation.
This is the kind of central access that may only happen a few times a season against Arsenal. Now, it’s a bit more regular.
And the problem, again, is that none of this happens in a vacuum. Once Emery shows everybody a handy case study, teams like Wolves can use some of the same moves.
This leads to the front-two wanting to stay more compact, which increases the chances of sustained possession for the opposition.
As far as I can see, there are a couple of causes to this:
Gyökeres has been doing well pressing up front, but not as well at tracking players in his blind-spot and quickly changing direction.
Jesus and Ødegaard, usually so impervious in these situations, have been awfully cautious since returning from injury. This leads to a conservatism in their jumps that has been getting exploited. They can seem especially reluctant to go from zonal to what Jon calls the “between phase,” where you blast up and start pressing again.
The wingers can get caught marking no one. It’s subtly complex, and Eze/Madueke are adjusting to their responsibilities.
What do we take away?
I think the application levels of Jesus/Ødegaard will naturally improve as they get more comfortable with their health. Same, too, with the newcomers. They have a big portfolio.
I think the wingers can pinch in a bit more and help with Rogers types.
It can be kept in proportion, as we’re still the best defensive side around.
If you want my simpler opinion on it, it’s this: Save us, Kai Havertz.
👉 Attacking the box on crosses
First, let us dispel the notion that Arsenal don’t provide service.
Gradient grades players throughout the league on dozens of inputs. Per their rankings:
The highest graded winger at open play crossing is Bukayo Saka.
The highest graded central midfielder at open play crossing is Declan Rice.
The second-highest through-ball passer at attacking midfield is Martin Ødegaard.
Not to mention all the set-piece opportunities.
Heeere’s Declan Rice:
But readers’ll know I have a long-term, though intermittent, frustration with how we attack these situations. These are the ones that stick with people:
And here.
Nobody produces 0 xG chances quite like us.
Aside from individual performances, I’m most frustrated with some of the haphazardness with which players can arrive, especially considering how great our players can be at delivering cutbacks.
I once put together an oversimplified version of what zones should be covered on a cutback.
These runs aren’t role-specific. What matters is just that they’re filled.
Madueke is so great at getting to the byline, and he was in a clear cutback situation here. I also think the running responsibilities are clear. Timber can jink into Van Dijk’s blindspot, then sprint nearpost; Eze can arrive (slowly) near the pen spot; Martinelli can clean up at the back post.
…and here’s how it ended. Timber slipped near the six-yard box, Eze arrived at the spot, and Martinelli did too.
A cross, unrequited.
When the job is done properly, the defensive line is bulldozed forward, and space opens up in their blindspots. A perfect example was Rice’s second goal against Bournemouth.
All three goals against Bournemouth had some elements of this: good box occupation and a pass against flow.
This chance against Pompey didn’t quite make it in, but it had all the right ingredients, including a block-splitting central dribble from our striker.
See.
Arteta discussed this topic after Liverpool.
“Well obviously we wanted to win the game … We were much closer to that with the amount of situations and the spaces that we arrived into. Normally you have to pick someone in the box and it’s a goal.”
“The movement was there, the arrival was there, the players were there. There are two or three that they come across and we need to occupy a space that we didn’t occupy to score the goal. But in many other occasions, the situation was there.”
Both of those last two goal permutations (Eze late arrivals, and Rice near the top of the box) have so many goals in them, especially considering the quality of cutbacks that Saka, White, and Madueke can provide. The other runs just have to reliably clear out the space for them, and we need to be greedier about drawing them up.
👉 The half-spaces
For much of the game, we saw the full-backs pinning the half-space, bringing their markers with them. This effectively made Liverpool defend in a back-six.
This is Arteta trying to open up space for his most incisive passers (Ødegaard, Zubimendi, Trossard, Saka) to do what they do best. It can be tough to progress through the middle against teams like Liverpool, so the thinking goes: might as well have some full-backs occupy them and then have some others pass it in behind.
In practice, there were a lot of rotations, but this was employed a lot. And it should also be said that the dynamics generally looked fine in the first half, and this fullback-pinning thing wasn’t the reason the game got away from Arsenal in the second half.
Still, there are a couple of issues here, some of them obvious.
The first is just what happens as play progresses. Saliba kept joining high, which I really loved in this case. It was accepted that Liverpool were playing a strikerless formation, and by filling a zone usually plugged by the likes of Ødegaard, it gave us an extra man in the box.
But still, you want to generate opportunities for your most incisive players. For all their qualities, Timber and Hincapié can both be chaotic from a final action perspective. Timber has a lot of layers, and for all his smarts and savvy and lockdown qualities and tight control elsewhere, he can almost be a fullbacking Darwin Núñez in the box: popping up in the right spots but looking rushed and imprecise when he does.
There’s another problem with this use of the half-spaces. Let’s tie the room together by going back to Pep talking about Iniesta. In his book, he says the latter taught him how important it was to dribble against central defenders:
“We had to attack in such a way as to get the ball to Andrés and Leo so that they could attack the central defenders and that opened them up. When we managed that, we knew that we would win the game because Leo scored goals and Andrés generated everything else: dribbling, numerical superiority, the ability to unbalance the game, the final pass, both to the outside and filtered through the middle.”
With a front-three of Hincapié, Gyökeres, Timber, that’s not really happening, unless Timber pulls off something nice.
This also gets into two other interlocking topics. For one, it’s painfully easy to predict a starring Martinelli performance versus a middling one.
When play leaned left against Bournemouth in static positions, it was a diamond of Rice, Hincapié, Martinelli, and Gyökeres. It’s nice for Martinelli to arrive centrally, but what about when he’s pinning? Who is dribbling at the shape to unsettle it? What is the philosophy of attack here, outside of transition?
On the season, this has continued the rightward drift of all dribbling, despite Trossard’s heroics. And most of it is on the full-backs.
There are multiple options to solve this, and with so many competitions a-brewing, we can always be horses-for-courses. But I’ve really liked some of the recent dynamics I’ve seen on the left with Eberechi Eze, even if it hasn’t turned into individual production.
Facing Palace’s back-five, Calafiori shepherded things from deep, Nørgaard was in the lone-6, and the pockets were occupied by two players who expected to receive in them: Merino and Eze. There were dribblers all over the place, from LCB to striker.
And we can compare that slow stuff we saw on the left with Martinelli and Rice to this diamond. Doesn’t this fundamentally just feel more threatening?
When he was off the ball, Eze was often giving Martinelli room to cook, assuring that doubling was hard and making the CB trio suspicious of shots. Most of the credit for Martinelli’s performance belongs to Martinelli, but I also don’t think it’s a coincidence that he did so well against Palace and Portsmouth with Eze by his side.
The dynamics, the mismatch, and the speed with which the ball was delivered to Martinelli, all contributed to a dominant performance. He created five chances in all.
…and it’s not the only thing to be intrigued by with Eze’s positioning. By pinching in from the left, he was able to strike diagonally between the lines to receive through-balls from the other advanced midfielder, Merino. It’s safe to say this kind of thing was not happening when Timber and Hincapié were in these positions.
Meanwhile, there have been some tweaks out-of-possession to better platform certain players. This is interesting because it’s the kind of thing that might have felt non-negotiable in previous years.
👉 In possession 🤝 out-of-possession
Facing Portsmouth, the middle four all moved a spot from the usual configuration. I’d show you a picture, but the camera angles were too close to get all the players in one shot. Here it is, drawn up:
Eze, the LCM, moved up to the higher role.
Merino, the DM, moved from the right of the double-pivot to the left.
Jesus, the striker, pressed from the right.
Nwaneri, the RCM (Ødegaard’s role), defended from the pivot.
This makes a lot of sense on a fundamental level.
If the RCM (Nwaneri) is going to be an all-action role, receiving deep and all over the pitch, it jives to correlate those on-ball activities with what he will then do off-ball. If he (or another player) loses the ball in build-up, he won’t have to travel all the way up the pitch to lead the press.
The same is true for Eze. This higher position helps him poach and look for balls in transition, without having to track all the way back to Rice’s usual responsibilities in the LCM.
This helped Nwaneri have a really nice, active all-phase role.
What this also meant is that both the LCM and RCM pockets have a proper threat in them. In moments like this, Eze was already high, and could serve as a backboard as Martinelli sprints in behind.
See, again, how quickly he can get out of the block on a regain. You can hit it up to him blind, and he’ll figure out what to do.
(Most of this can be replicated by him at LW with Hincapié playing, or with Calafiori leaning wide).
As the subs came on, we saw Ødegaard embrace Nwaneri’s same responsibilities, something we’ve very rarely seen over the years.
Here he is double-pivoting in build-up:
And here he is double-pivoting in the press as well.
What’s nice there, too, is that Havertz can serve as 75% a typical striker and 25% a right-10, providing some running and action for the right-winger as needed, which suits his skillset perfectly.
This isn’t necessarily something that projects well against all opponents. The risk is that an opponent can isolate somebody like Rogers on Ødegaard in space. But against more overmatched, low-block opponents, Arsenal can be overkilly from a defensive point of view, and anyway, we rarely actually get into a mid-block shape. In those situations, we don’t routinely lean enough into the firepower we have. This is one route to achieve it.
By having a lone-6 (Zubimendi or Rice), and swinging one of those two out to full-back, or just resting a DM, you can then place “Twin-10s” in the places they can do the most damage: Ødegaard all-action, Eze high and transitional. Combine that with proper wingers and a proper striker, and you have dribbles, incisive passes, speed, shots, height, and relentlessness.
I also have something to whisper. It’s early days, but I’m sensing a shared understanding between Eze and Ødegaard.
This play was the perfect embodiment of everything I was looking for in those two roles. Whatever position he’s coming from, if Eze can swoop into the D, he offers such reliable shot mechanics.
Regardless of who is playing, we’ve seen that this team has more potential to offer in terms of high-central combination play.
The secret there? Plenty of humans are in threatening places. Instead of getting outside of the block, I would like to see Ødegaard drifting behind the front-two pressers more, drive up, and have 1-2 more players to combine with.
Here he is doing it for La Real with Merino.
This, one of the most satisfying goals of the season, came from the #6 smelling blood, and the right “10” defending as a left pivot, ripping the ball, and driving it up centrally.
That is fundamentally a goal that the midfield would not have scored last year. The same was true for the Jesus goal, in which the #6 (Zubimendi) sprinted up the left side of the pitch to carry the ball up, delivering it in transition.
Much of what we’re watching is Arsenal bedding in new players while rebuilding the midfield’s mechanics. This period has already borne fruit. Zubimendi looks so untethered and all-action, Ødegaard has hinted at his best, and Rice has unlocked a new level. It’s also led to some slight jerkiness as other players learn to interact with that swirling trio. Eventually, one hopes, the gears mesh.
It’s been a fixture-rich period. It’s also been an information-rich one, one that reaffirmed some of the special qualities of the group, showed us a lot of lineups and tactical permutations deployed to production, provided some new hopes, and confirmed a few suspicions. We’ve missed Calafiori on that left, I must say.
I gave some extra weight to the rough edges today because they’re more interesting to discuss. But the obvious truth, one that we all logically know, is that the Arsenal are in a strong position, regardless of what happens today. The Jesus goal felt like a tonic, and without putting too much pressure on it, Kai’s return feels like something important for this team to unlock its final form.
“Obviously we missed him a lot,” said Arteta. “A year is a long time in football but as you said, the way he’s come back, he looks really fit, really confident. He’s such an intelligent player that straight away he is someone who connects with the players really naturally, and you can tell how much they love him as well, because every time he was in and around the box, everyone was jumping on their seats on the bench. So yeah, we are so happy to have him, he’s going to help us so much, and now we have to keep him fit.”
Elsewhere, there’s been a lot of learning to do. As matches pile up, adjustments are made, new experiments are tried and tested and discarded, and new connections formed, things will inevitably feel cluttered at times. All you can do is return to the lab, pause, and look at the mess with a bit of curiosity before cleaning the bench.
Billy, that first section on athleticism is one of your most insightful pieces of writing - an incredibly high bar. Any tracking data providers reading should be paying you to demonstrate it with their data. I can’t thank and praise you enough for it.
A deep, rich work of art and empirical facts. I must first give credit to author who put their experience, expertise, perspective, philosophy, and observation down in a coherent and impactful manner.
Excellent work from you, Billy Carpenter. You nailed this one, as always.
Great as Trossard is, it is true that even he has his limitations. He selects his moments very carefully due to his physical deficiencies and protects the team by simply recycling if he cannot make progress. But as you have said, Eze provides us with much more dynamism on that side. I hope that we see more of him.
You spoke on cheap giveaways/technical insecurities. For example, the miscommunication between Skelly and Zubimendi vs Liverpool. The first instance of this was vs Manchester United - our first game of the season. I chalked it down to fitness but the theme has reoccurred multiple times across different phases of the season.
Despite being a team willing to defend, I find that clearing the ball mindlessly scarcely ever works in our favor.
Hopefully, it is just the team forming new patterns, as you have said.
I may have to read this masterpiece two, three, or four more times to fully understand it.
In the meanwhile, do you mind recommending books.? I'd love to notice these things both as they unfold and after a rewatch.
Again, wonderful work. Perhaps, the best compliment I can pay it is that reading it just once is a tantamount to the sin of blasphemy.