Forest for the trees

Taking a step back to see what we can glean from the opening win against NFFC, including: the new shape, Havertz and Rice, how to break down a block, and the limits of tactics

TL;DR: I was sad about Timber, so I wrote a couple thousand words on the Nottingham Forest win to distract myself. Enjoy.

---

🔥 It’s the players, dummy

In the first half of Arsenal’s season-opening 2-1 win against Nottingham Forest, Martinelli was feeling himself.

After contributing a comp-reel back-heel assist to Eddie Nketiah, he left Serge Aurier on the ground, the latter looking like a pitiable old man trying to flag a bus:

The battle wasn’t over, however.

In the second half, Martinelli dropped low to offer his teammates an outlet for progression. Pride in tow, Aurier cleaned him out and stepped over him:

Thirteen minutes later, Martinelli decided to return the compliment:

This battle was one feature of a bristling contest that included exciting debuts, unfamiliar patterns, moments of brilliance, familiar reasons for pause, heartbreaking injuries, and an all-too-nervy ending.

But I’m not even focusing on this skirmish to shine a light on Martinelli’s play, which would be easy to do. I’m focusing on it for another reason: the players. The wizardry, the wounded pride, the frustration, the unexpectedness, everything. It’s all so profoundly player-driven, and so too was the game. As it should be.

This was true of all three goals, starting with the moment of genius we already mentioned. Poor Aurier:

It was true, too, of Saka’s long, bending rip past a crowded box — which was not exactly the product of some clever tactical machination:

It was even true of Forest’s goal, in which a subbed-on and shot-out-of-a-cannon Anthony Elanga simply ran past the leggy legs of Arsenal. There was ample coverage, but two players just did better than six:

Last week, I talked about how I like to review an article by coach Juanma Lillo once every couple weeks as a slap in the face — reminding myself not to overcomplicate the simple or to cocoon my nonsense in fancy terminology (at least, as much as I can).

In that article, he said this:

“We, the managers, have too much influence. It’s unbearable. We have our own ideas and we say that we espouse them to help people to understand the game. Bullshit! It should be for the players to understand the game as they understand it.”

Complimenting James Maddison — our new hated rival, who, in fairness, is looking like a made-to-order #10 in the Angeball system — he said this:

“He’s an authentic footballer, more of a street product than an academy product. He’s daring. He has nerve. A manager can tell him to do A, but if he believes B is the right option, he’ll do B. I just love this. Any idea that comes into James Maddison’s head is 100 times better than any idea you will find at any coaching conference.”

Something I’ve enjoyed noticing lately is how, against stubbornly disciplined blocks, almost every goal is scored by somebody defying a teammate’s nudge to do something safer, smarter, more “correct.”

Ødegaard is a delightfully extreme example of this. Always communicating, usually gesticulating, he can be found imploring his teammates to make the right pass from the gameplan, all while opting for the weirdest possible idea when the ball is his. His on-ball prowess manipulates flow and expectation, resulting in goals.

That’s because if you do the expected, high-percentage thing against a rugged, low block, there is a high chance that the defenders will be on your page, as well. If you want to score, the freeze-frame of the goal is likely to include a shot of your very smart teammates suggesting you do something else, like so:

Or like so:

To paraphrase Lillo, “Any idea that comes into [Bukayo Saka’s] head is 100 times better than any idea you will find at any coaching conference.”

Martinelli’s back-heel. Nketiah’s glorious blast. Yes, Saka’s claustrophobic curler. The shoving and long-shots and all that. For as much as we talk about tactics — and we’re about to talk about tactics, don’t worry — we should never forget who we pay to see.

---

💎 The boxy diamond

Many were surprised when the lineup was announced. There were two reasons.

First, Gabriel was missing for the first time in 73 Premier League games. And second, the personnel seemed to point to a new shape.

But as a keen reader of this newsletter, you may not have been so surprised. Writing back in mid-June, two weeks before Havertz signed, we outlined the first way we’d like to see him used in progression, particularly against low blocks. Here was that:

As circumstances arose, we also suggested playing Havertz up front in a sturdy, 4-2-4 long-ball formation. That’s what you saw against City:

That first look is ultimately what was used against Nottingham Forest in the season opener: the three-back, the right-inverted Thomas, the pushing Rice, and the more central Havertz.

(As we all know, suggesting formations before they happen is what really matters in life, and I should be showered in adulation — not to mention, newsletter subscribers.)

It evolved a bit, and seemed to have even earlier triggers for Havertz to float as the game progressed, but let’s examine how it worked in practice.

With a Wyscout PPDA (passes per defensive action) of 33.21, Forest showed almost no pressure in the first phase of build-up. That means that, statistically, they allowed 33 passes before an intervention. Arsenal was able to progress up to the half-line unimpeded, as many of their opponents will this year.

Partey was generally the primary pivot, but in earlier build-up, Rice had the option to drop down and help as he saw fit, and he’d often drop all the way to the CB level for a carry.

Here, he was transitioning out of helping the ball up after a recovery, so a familiar 3-2 is played, with Havertz in the standard left-8 position. Swap out the names and what starting positions the players are transitioning from, and this could have been any game last year, perhaps with more tilt the other way:

But as the ball moves up, this is when things change. Rice motions to Havertz that it’s time to modulate — Thomas doesn’t need his help anymore — and signals a rotation:

…and there’s your diamond, or whatever you want to call it:

If we’re going by numbers, it might be most accurate to call it a 3-1-3-3 — or a 3-4-3, or even a 3-1-6. My official position is “I don’t give a fuck.”

What is important is profiles and attributes over specific positions. And the complementing skillsets of Havertz and Rice were a key reason that Arteta has looked to bring them together:

The entire tactical remodelling of Arteta’s midfield had been predicated on the German playing alongside Rice.

Havertz has been earmarked to replace Granit Xhaka in the left No 8 role but he has played most of his football at Bayer Leverkusen and Chelsea as a No 10 or in a withdrawn striker role. It is a bold move but it was considered feasible because Arteta sees Rice as having the athleticism and ball-winning ability needed to complement Havertz and Martin Odegaard either side of him.

On paper, this makes Declan Rice the #6 and Thomas Partey the right-back. But in practice, Partey often became the lone pivot — and Rice became a hybrid in possession: 75% touch-heavy controller (think Zinchenko) and 25% box-crashing 8 (think Xhaka). Another way of saying that is just a true box-to-box midfielder.

This freed up Havertz to be something entirely new: a floating, off-ball #10, with a base position that leaned more central in advanced possession.

What do you get out of all of this? An extra attacker. Here’s an oversimplified version of what that looked like — 6 front-line attackers instead of 5:

The promise of this formation showed up in fits and starts throughout, and the objective was fairly straightforward: by cramming numbers centrally in the box, you give the opponent’s bus a more difficult choice of where to park. They can stay compact to match all their marks, and risk giving the wingers (Saka/Martinelli) space to create wide, or they can stretch out to the wingers and risk leaving gaps in the box.

Here’s potentially the most obvious example of its benefits. Saka is doubled out wide, stretching the formation. Little gaps appear, and because Rice is crashing the party from high, he is lost in the sauce. A Saka cross goes in his direction:

And after a deflection, he gets a free shot on goal which is narrowly bounced out:

---

🐶🍚 Havertz and Rice

As you see above, Havertz is up there attracting attention in what would be Xhaka’s spot in the lineup; Rice is up there as almost an aggro-Zinchenko.

In any case, without this formation tweak, Rice is likely marked. His shots were great all day, with all three on target.

Said Arteta:

"He probably should have scored two goals today, so more of the same. He’s not thinking, he’s just acting and I’m really pleased with what he’s done."

Meanwhile, Rice’s pass-map shared some similarities to those of the Ukrainian. It was a little lateral but overall, looked like an expansive, second-phase controlling #8, just pushed forward a bit. You always got the sense that this kind of all-communication, all-action role is what Rice lives for:

Rice: Passmap

Havertz had an excellent game out-of-possession, and was a key to keeping the ball pinned without active menace from Forest. He was, as always, pristine at finding little pockets of space to generate moments, was involved throughout, and looked natural. One thing which hadn’t crossed my mind was his usefulness as a closeout #9: when defensive subs come on, they can just blast it long to him, and he can bring it down aerially and pass it to a sprinting winger in behind.

Havertz: Received Passes

He was, however, overly demonstrative in two ways. First, when demo’ing the new look, the entire lineup was a little more mechanical in their rotations, requiring active communication (like we saw above with Rice telling Havertz to rotate) instead of looking like true automatisms. This caused the couple seconds of delay that can dull the cutting edge of the team, and was typical of some of the struggles in the stretch run. This especially impacts Havertz.

On this day, there was a particular reason Arteta wanted to be quick:

“We spoke that we wanted a very high rhythm game today, because we knew the pitch was going to get dry and it was really hot, and it’s going to be slower.”

Second, he had a couple opportunities where he’d get the ball, then think, instead of acting more decisively:

At Chelsea last year, this was almost unavoidable. He’d have to take extra touches because there weren’t any creative runs happening around him. As he adjusts to a new reality, he’s still putting in good performances. As we saw in the late-game dash, he has genuine speed, and it’ll be nice to see him use it — both more frequently, and more immediately.

This formation has one more benefit. By often having more of a “3-back” look than a “2-back” in more advanced possession, the CB’s can be a little bit more free to push forward, with the opposite-pitch 8 dropping back to support the counter.

This led to White overlaps, despite his work as an RCB, not an RB. If we’re looking to do an apples-to-apples comparison to last year, this is the equivalent of Gabriel doing overlaps around Martinelli and Xhaka last year:

And here’s Tomiyasu doing sprints through the half-space, setting up a narrowly-missed chance for Martinelli:

---

👈👉 The left and the right

Everything was not perfect, however, and I’m sure Arteta would have liked to have left with more than two goals. Some of the clues lie (lay?) in the pass map:

These maps should always be viewed with a hearty helping of context, as I think people can often seek perfect visual balance over optimum game performance.

For instance: if I, Billy Carpenter, were to start at left-back against Bukayo Saka, why should the gameplan be balanced? Shouldn’t they just feed the ball to Saka to destroy me 100 times in a row? Don’t answer that.

That said, we can look at it to understand a few things. For one, you’ll see how Rice’s (welcome) desire to drop and receive the ball on the backline dropped his circle all the way down behind Partey’s. You’ll also see how the left side sags below the right side, and how Nketiah would lean a little right (potentially as a result).

But if you look at them as four-player pods:

• White/Partey/Ødegaard/Saka = proven world-elite combinations with dozens and dozens of starts together

• Timber/Rice/Havertz/Martinelli = 75% brand-new players

As one might expect, most of the imperfections on that side weren’t about structure, but about the simpler things. Namely, timing and shared anticipation.

Below, Rice thought Martinelli would be cutting in behind, but Gabi ran inside:

Pure, unadulterated “new connection” things.

For that reason and more, I’m incredibly patient. Do I think some combination of Rice/Havertz will provide more support to Martinelli than Xhaka did? I do. The speed and fluidity of Havertz, in particular, should work to get Martinelli more central and into his happy places more often. Do I think there may be some adjustment period before we get there? Also, yes.

We should be patient for another reason. The two great lubricators of the left-sided machine, Zinchenko and Jesus, didn’t play. While Rice and Havertz settle in, some combination of the former Man City tandem can work to create the decisive passes and unexpected moments that were too rare.

If this formation continues its use, Tomiyasu or Kiwior will likely have to actively improve their high support of Martinelli, overlapping and underlapping with more frequency.

It’s a work in progress, but the key word is progress.

---

🛑 Back to the block

Which brings us back to original topic, which we’ve hovered around, but will now address in a little more detail, though less detail that it deserves: the art of breaking down a low-block. Arsenal fans can be frustrated when facing such blocks, which puts them in league with, well, the fan of pretty much every top-tier side.

If you’re frustrated, that’s because they’re frustrating. Lillo, the beautifully clear-spoken coach whom we opened with, had this to say in the same World Cup piece:

Whatever block Morocco had, it was a lot of players working in conjunction with each other, paying an incredible desire not to open up spaces.

It is getting harder and harder to overcome this approach. Teams can move a whole line — midfield line, defensive line — from one side of the pitch to the other almost quicker than the ball can travel. You have to have a lot of quality to beat it.

On Saturday, Arsenal fared OK, but with some notes. They were able to create a general, incessant thrust of pressure towards their opponent, but it often lacked the final moments of incision. See:

I compared the first and second halves, and when I tried to find evidence of the team taking their foot off the accelerator, I didn’t find too much. The tempo (number of passes per minute) and threat were broadly similar. What was different was the passing quality; whereas Arsenal completed 10.17 passes per possession in the first half, that number dropped to 5.36 in the second half. A higher number is not always a good thing — Arsenal have a sweet spot between possession and directness — but the second half was full of some sloppy directness, punctuated by some passes that just looked a bit tired.

Back to the block. Your job as an attacker is to get the defenders out of position. The defenders are not inclined to do so. A patient, zonal block with great athletes can teeter-totter between sides of the pitch in milliseconds. That means the gaps narrow, so the passes and subsequent actions have to be immediate.

Your window of opportunity grows as you gain numbers, which is the primary reason why Arteta is attacking such shapes with six players. By making rotations a little more fluid and unexpected, with multiple players attacking different half-spaces at will, the coverage becomes more complex and liable for indecision or error. By further sending runners from deep, like we saw with the White overlaps, the numbers are more likely to be on Arsenal’s side.

But no matter what — with this amount of coaching, training, tactics, size, and physicality in the league — if the passes or runs take a moment too long, or there’s a lack of coordination, you’re unlikely to find joy. As we’ve covered elsewhere, that popped up a bit on Saturday.

The other option is for some dazzling work on the dribble to yank players around. Once the block gets “set,” however, this becomes difficult, stressful work — and reminds us of why Gabriel Jesus is so special.

But the best option is to attack quickly, and decisively, before the block is set.

We rightfully gave all the credit to the players. But tactically, all three goals were a result of post-corner routines.

On the first, Martinelli tracked down a cleared header, and immediately turned around to pressure the box — knowing that this was a fragile opportunity to find the Forest team out of shape:

On the second, Saliba sprinted to the corner, dispossessed Gibbs-White after a heavy touch, and Saka went running into the middle to get a shot off a less-settled block:

The same numbers and forward-lean that created the first two goals then resulted in the post-corner Forest goal that made it 2-1. Attacking these chaotic moments is vital.

If you can’t handle me at my “Awoniyi near-post tap-in,” you don’t deserve me at my “Saka far-post screamer.”

---

🔥 Final thoughts

All my thoughts are with Jurriën Timber. It is just genuinely so sad. As you could probably tell by these newsletters, I couldn’t love him more. Here’s to hoping he comes back better than ever.

I was going to shoe-horn some thoughts on the future of the defensive line into this post, but it’s already too long, so I think I’ll save that for later this week. There are many other topics — Nketiah, Ødegaard, Gabriel, Tomiyasu among them — that I’d expect to cover more in the weeks to come.

Back to this game. My general thought is that if you were encouraged by it, you were correct. If you thought it lacked some incision and polish, you were right too. But structurally, I think the approach is more than sound, and will only gain steam upon further uses. The idea that Areta is sacrificing some short-term comfort for some long-term potential has become a weekly chorus of this newsletter, and I repeat it now. Just because something is not perfectly clicking now does not mean it won’t in the future; in fact, that kind of possessional dominance and functional superiority is especially encouraging when you think of how much potential there is left in the tank.

In truth, I’m less concerned about the forward lean or the dynamics of the attacking line — I think there are enough talented players, and enough ideas, to make that go bang. The open questions are what happens when a team more forcefully attacks the spot behind the inverter (Partey in this case, who still looks pretty quick these days, but a bit slow) that it becomes too difficult to cover back with so many players forward. The team already played on the edge of that last year. With another player up, it raises the burden, and the loss of Timber may hurt. But we still have Saliba.

Thus brings us to Declan Rice. All of these moves emanate from his decision to sign with Arsenal, including the sturdy block against Man City, and the ability to use Havertz as a bonus attacker against Forest. I watched more than a few expensive squads this week — Bayern, Liverpool, and Chelsea among them — and found myself thinking how much better they’d look with Rice in the lineup. Too bad.

However sad the Timber news, we are still very fortunate — and still much improved. That is because of the vision of Edu, Arteta, and a wider support system. But at the heart of it all is the player.

To paraphrase Rudyard Kipling, we can tactic without making tactics our master; we can data without making data our aim.

What a joy it is to watch these guys play football every week.

Happy grilling.

❤