Lost in Lens
An overlong investigation into the questions facing Arsenal at the moment, including: what made Lens so tough, the changes in build-up and advanced passing, Havertz, tempo, box defending, and mistakes
I hadn’t caught too much of Lens this year, so before the game, I looked to get reacclimated. I was an admirer of their exploits in the last campaign. They earned their Champions League place by finishing second in Ligue 1, just a point short of PSG. Their electric crowd helped propel a 17-1-1 home record, and head coach Franck Haise shepherded a defensive shape that surrendered only 29 goals in 38 matches.
But there’s been turnover since then, with Loïs Openda (22 goals) moving on to Leipzig, and Seko Fofana off to Al-Nassr — joining a squad with Sadio Mané, Aymeric Laporte, Marcelo Brozovic, Otávio, and some guy named Cristiano Ronaldo.
The team sought to reload in the form of record-signing Elye Wahi and former PSG prospect Andy Diouf, but it was always going to be a tall order to repeat last season’s accomplishments. That assumption has prevailed thus far. Through 7 games, they’ve gained 7 points, good for 15th in the league.
And so my initial look through their stats wasn’t too concerning. While the underlying picture was more affirmative than their ranking would suggest — with a positive differential based on xG, and other stuff like that — their numbers elsewhere looked solidly midtable.
But as I dug in more, my eyebrow raised. As a decently high-possession side, they spent very little time with the ball near their own goal. Their 44.7 touches per 90 in their own penalty area were the lowest in the league. So, too, were their touches in the defensive third overall: they notch 33 fewer touches than anyone in the league.
Why is this interesting? Look up the pitch. Despite not being a tremendously high-pressing team, here are their ranks in the league:
Second in passes to the final third (44.7/90)
Second in passes to the penalty area (11.9/90)
Second in crosses to the penalty area (3.29)
Second in progressive passes (56.7)
Third in shots (16.3/90)
Third in key passes (12/90)
What this means is that they’ve managed to secure a lot of balls in open play and turn them into incisive attacks with velocity. The direct speed of their attacks averaged 1.98m/sec, the third fastest in France’s top-flight.
They’ve just been missing their shots. Their 26.3% on-target percentage is the lowest in the league.
In so many words, it’s a dangerous side, table be damned.
In so many other words: Hey, the Arsenal! Welcome back to the motherfucking Champions League. There are no easy games.
These factors, ultimately, transpired on Tuesday from Lens.
A disciplined but aggressive mid-to-low block, uproarious crowd support, and incisive advanced passing formed a potent cocktail. The only thing that didn’t take place was the whole “missing a lot of shots” thing.
In truth, their semi-pressy 5-4-1 is almost exactly how I’d defend us. Whereas an Everton can be inherently negative, this Lens vibe can turn into genuine attack with more ease — doubling wingers, yes, but aggressively jumping the ball in the midfield, and actually sending runners forward when the ball is won.
A slightly higher block like this is, simply, closer to the opponent goal.
It was a stellar performance from them. It was led, in my mind at least, by Kevin Danso — the physical, intelligent beast of a centre-back who kept Arsenal’s attack in check for much of the night, and always seemed to have position.
“So, Billy, Lens are better than their numbers suggest, and much of their stylings reek of the untapped potential, which they managed to tap against Arsenal. Cool, bro. I’d still like us to win.”
While this all hopefully explains why Arsenal went full-strength — again, it’s the motherfucking Champions League — this is all dull comfort to a team with aspirations of winning the Big Stuff, and who will have to learn to beat teams like Lens to do it. We also can’t cry travel schedule, or fatigue, or injuries — or act like the issues we saw from Arsenal were aberrational.
From our watchtower, there is no mistaking a few trends:
Arsenal still make mistakes
Arsenal have trouble generating threat when Saka is out (or managing himself)
Arsenal aren’t reliably creating chances to the level of their ambitions
Arsenal can struggle to adjust their attacking speed
Mistakes, unreliable chance creation, and difficulty in attacking at different tempos: these are the hallmarks of a disappointing Arsenal night.
But it's been a busy, bewildering time since we spoke last, and this isn’t the only game we have to discuss. The depth guys put in a great shift at Brentford in the Carabao. The starters looked their best in a 4-0 win against Bournemouth, and my brain still hasn't managed to eject the Havertz song from its hollow depths. And now we’ve been humbled in the Champions League.
Football is confusing. Arsenal have lost, finally, and the winding road leaves enough room for anyone to interpret according to their priors. The doom-mongers are awake; the sunshine-pumpers are at a whisper. Let’s investigate.
🤔 How do Rice and Raya impact build-up?
In this year’s preseason, a lot of the intrigue was in the pivot. Not only was Declan Rice signed for a club-record fee of £105m, the manager was getting experimental with the midfield.
In all, Arteta deployed around 15 different pairings during the pre-competitive slate. Kiwior played within the block as a true Zinchenko impersonator; Tierney even played the role of inverted full-back while Partey was the one playing on the byline on the other side. Plus, a bunch of other stuff.
For example, here was an ESR/Jorginho double-pivot against Barcelona:
It’s been a bit more straightforward since then.
By and large, Rice has run the show as the deepest midfielder. Against Lens, his 112 passes were the second-most of his career across competitions. His passmap looked like he tried to paint the pitch blue:
Against Brentford in the midweek Carabao match, Jorginho served the role of a true single pivot, shepherding play on his lonesome:
In the Community Shield, Partey and Rice were a fairly traditional double-pivot; Partey then swung to right-back while Zinchenko regained fitness. Since then, and after Partey’s injury, Zinchenko has reigned supreme.
The Bournemouth match was particularly interesting in this regard. After taking over the job from Gary O’Neil, Andoni Iraola has led a press that has been at a similar intensity to Arsenal’s, statistically speaking. Judging by the imperfect measure of PPDA (passes per defensive action), it has outstripped the likes of Liverpool, Newcastle, and Manchester United.
It’s hard to cleanly describe what took place. But due to that press, the broadstrokes were this:
Raya joined the build-up crew as a left CB
Ødegaard often formed a true double-pivot with Rice
This left Zinchenko to start his view as a left-back, but roam up the pitch as he saw fit, almost as a “diamond 8” midfielder
The very short version is that Raya and Ødegaard actively joined build-up in expansive ways to ensure there was always an overload there.
It could start like this:
…before Raya would fill the role as the LCB and Ødegaard would join the pivot, and Zinchenko would float around, like this:
This solved a lot of things at once. For one, a deeper Ødegaard got more touches and was able to orchestrate play much more forcefully than he did against Tottenham, when he was shadow-marked and turned into an ancillary figure:
Meanwhile, Rice was surrounded with help, and wouldn’t be overburdened to cut the lines himself. Zinchenko could float around as an opportunistic, helpful box-to-box type player. It was the best of all worlds.
The vision could be seen early. Here, Raya has it deep, but Zinchenko is asking him to slow things down until he generates commitments on behalf of Bournemouth:
Raya has no problem doing that. He waits, the press floats over to their new assignments, and Zinchenko is the extra man. Raya hits him, and it’s off to the races:
If you changed the kits, that could very well have been a perfectly-executed sequence by De Zerbi’s Brighton. (Brighton, Brighton, Brighton, I know). What may be underappreciated about the Italian’s tactics is not how confidently his sides play in the first third, but how he’s able to reliably generate space and time for his carriers, like we see above — not just for the press-buster himself (in this case, Zinchenko), but also for the attacker who waits (Saka).
In all, Zinchenko had an interesting, advanced game:
By having so many players playing confidently deep, and then creating a huge gap between that line and the attacking line, there’s an acre in the middle of the pitch, and it ends in a more advantageous 1v1 out wide — because the help is sprinting to get back. Here’s an extreme look at how it can start for the Seagulls:
Thanks to these tactics, Brighton were able to create more shots off of take-ons than anyone last year. Their wingers are consistently in positions to succeed.
Against Bournemouth, Zinchenko was manipulating things left and right, and Arsenal’s extra man always found room to push forward, and then hit it out to a more isolated attacker. Here’s Rice doing just that:
The Bournemouth game looked really good to me. I was cranky about Jesus out wide against Tottenham, but after Nelson and Smith Rowe both played (great) against Brentford and looked gassed near the end, I was good with the move back to LW here, so long as Nketiah played expansively — which he did, as he also did earlier in the week. Ødegaard orchestrated play from everywhere, and there were some truly beautiful team moments and support from the away crowd. Raya led the team in progressive passing distance, hit 24/24 in the medium-length, and attempted as many passes (48) as he ever has in the Premier League. It felt like a good template to use against pressing teams moving forward.
As a result, you can see how clean the “in possession” visual looks here. This graphic doesn’t tell us everything, but I can imagine Arteta was pleased with the balance:
But Arsenal found tougher sledding against Lens.
Why was that?
What made Lens so tough?
For one, the French side boasts some great defenders, and the crowd was uproarious.
But the bigger reason was that they had intention. They largely refused to get baited, often allowing play to get into the middle third before swarming like a press. From there, they weren’t just blocking lanes; they were adding pressure. If there was a moment to pounce, they’d all jump up and press in unison. If it was time to regroup, they got all the way back.
A lot of Arsenal sequences would start like this:
From there, you can see how they’d jump together when a certain condition was satisfied. It could turn into a higher press on a moment’s notice:
This makes things less easy to predict and manipulate, and generally kept Lens out of the situations that so many teams find themselves in: pressing constantly but without particular conviction or intensity. That can be the worst of both worlds. Chelsea last year comes to mind; Bournemouth got caught “in the middle” a bit on Saturday, too.
But while Arsenal’s rest defending has been incredible, and the midfield trio looks like a restless, formidable unit out-of-possession, Arsenal are again not looking quite as vibrant or sharp these days in attack. This would be easier to live with if they avoided mistakes, which they haven’t been able to do. Said Danso:
“We gave it our all, Arsenal kept us really deep. We kept our concentration, they didn’t really. It was from our mistake they scored a really brilliant goal from Gabriel Jesus but we got one back early. At home we know how strong we are. We will take the three points. we did our best and luckily we won the game.”
So we’ll continue to talk about it until the Bournemouth game looks more like a rule than an exception.
What’s different in attack?
👉 Teams have adjusted
The most obvious thing to note is that the first exciting campaign is always the easiest, as teams adjust to a new level:
The most underreported change is the one that sounds like bad teenage poetry: this year, last year has already happened. Unlike the previous go-round, teams have now watched a full season of Arsenal’s title-challenging ways, and a reputation is fully conjured. There are no more first impressions.
This adjustment has been extreme.
Arsenal now face the lowest lines of any top-5 team in Europe:
Even against Lens, who felt relatively press-y, their defensive line height was 40.5 — which is lower than Luton plays every week. The 21.9 PPDA they registered against Arsenal would be the lowest-intensity press in the Premier League. Of course there are game state and matchup factors there, but still: the fact that the game felt relatively open to us shows how far our expectations have shifted.
👉 Trouble with tempo
The second thing is how Arsenal go about managing tempo. After Spurs, we talked about Arsenal’s invisible dial:
Since the beginning of last season, Arsenal have shown that they have a general tempo in which they’re most productive. Speaking anecdotally, that always seems to be when they average around 5 or 6 passes per possession. When it’s fewer than that, they’re usually running too hot — pushing the ball up the pitch regardless of the opportunity ahead of them, and losing it too often. When it’s more, they look lethargic, and likely to make a mistake.
Arsenal averaged 3.47 passes per possession against Tottenham. That was too fast, often rushing passes that weren’t on, and letting the game devolve into basketball.
Against Bournemouth, they were perfectly in the aforementioned window: with 5.6 passes per possession.
Facing Lens, they went all the way up to 7.39 per possession. That’s higher than any game since the beginning of last season — aside from a similar one versus Nottingham Forest last year. Like Forest, Lens had something to do with it, but this relative sleepiness always seems to creep into the disappointing performances.
I promised to never talk about Europa again, but if you can brace yourself for a moment and walk down memory lane with me — you’ll remember a search for targets of ire (It’s Sambi! It’s Vieira!) in the early stages. Then, Arteta started fielding more full-strength lineups, which, unfortunately, helped us understand that “dialing it down a touch when playing midweek in Europe” was a teamwide affliction, and one that seemed to recur on Tuesday. Their performances against Lens and Sporting CP bore some similarities. Hopefully it doesn’t continue to happen.
👉 Passing has changed
And finally, the passing. We can mourn a missed chance here and there, but ultimately, Arsenal need to create more reliable mechanisms for chance creation.
Below, I’ve grabbed relevant samples from four positions, comparing last year’s output to this year. Again, this should be caveated with the difference in opponent strategy. Still, you’ll see a bunch of interesting stuff, which I’ve highlighted in orange:
Some notes:
Ødegaard has indeed been creating less this year; believe it or not, he’s actually been shooting less, too. He needs to move around and orchestrate from deep more consistently, and when he’s shadow-marked, others need to immediately drop into the lanes he creates (instead of parking, like Nketiah often has). You know the back-post cross that he hit to Jesus/Havertz against Bournemouth, that led to the Saka goal? I have no idea why that isn’t a steady part of the repertoire.
Rice has largely played a highly-competent, high-percentage, more lateral version of Partey’s role. I think Rice is a superior overall player to Partey already, and depending on the status of Partey’s legs, the gap may be large. But Rice could learn some more risk-tolerance from his predecessor. After all, if it goes wrong, he’s even better at blotting out the ensuing counter. I loved the urgency and forward thrust of his play late against Lens.
As a bit of a control group, Zinchenko offers interesting context to all of this. He is also passing forward 5 fewer times per 90, and going lateral more. You can decide how you interpret that. These lateral passes are, to some degree, inescapable when a team is defended like this: Rodri pounds them in at prodigious levels.
The biggest difference is from Xhaka to Havertz, which doesn’t look good here. I’ll get into his out-of-possession characteristics in a second — I don’t know how many people expected him to be Joelinton in the block, but here we are — but Havertz has been even more off-ball than expected, accepting 14 fewer passes per 90 than Xhaka (though it’s closer in league play). General progression hasn’t stalled because of him, but even with more output, that’s still too low; Xhaka had turned down his involvement already. The ball needs to hover around that half-space more if chances are going to happen. I thought he was pretty good against Bournemouth, too timid elsewhere, and have seen too many adjustment periods to believe “not working yet” is the same as “not going to work.” Havertz himself can improve; the play and dynamics around him can improve; he can rotate out and/or fill in other positions; other players can get more chances to make their mark in the spot. Look at that, we can hope for all four, without clutching to a dogma!
While I’d expect a few more touches for Havertz, I’m not sure Arteta is surprised by the numbers in this table: he would just hope they’re offset by goals and pressing contributions. The latter has happened. The former, not so much, and as I’ve said elsewhere, it’s fair to judge him by output.
I will also say this, which is wider than discussions about him: people can have trouble seeing the value of any form of positioning that isn’t “running to the carrier and asking for the ball at your feet.” There is all kinds of responsibility in positional play: holding space, pinning CB’s, holding width, occupying defenders in the middle, opening lanes, etc. Sometimes the issue lies on the player; sometimes it’s on a teammate for not exploiting the space that he provides. It’s all about overall makeup; you can’t have all 5 of the front-line staying put.
But in all, the so-called “control” and general progression has been good — but the final bit of creativity is missing. Here, you can see that Arsenal have the highest % of touches in the final third of any team in the EPL:
But as a result of all of this, Arsenal have only created five goals from open play in the league. I don’t mean to alarm you, but that’s fewer than Nottingham Forest, and the same as Burnley and Wolves:
That’s right. Arsenal have as many penalties as open-play goals. Ugh.
In many ways, these are the goals of a Champions League team: by all means necessary. Of the 23 goals in the UCL on Tuesday, only three did not come in transition or dead ball situations. But Arsenal don’t play in the UCL every game, and one has to ask how long this is sustainable. More reliable opportunities must be created.
In the first game of the season, Arsenal faced the back-five of Nottingham Forest. To get numerical superiority, Arteta debuted a new look: a 3-diamond-3 with Partey at the base, Havertz as a roaming CAM, and Rice joining the front 6:
The initial results were mixed but damn interesting, and with Partey back, perhaps we’ll see that again the next time we face a back-five.
⚡️A focus on moments
So hopefully we’ve established that there are structural things to iron out. From there, it’s a game of moments, particularly in the UCL. So let’s do a speedround of those.
In the first Big Moment, Jesus scored in a picture-perfect finish:
The important thing here is that Saka didn’t just intercept the ball. He had the wherewithal to pass it to Jesus on his very first touch. As Danso inferred, if Arsenal had ultimately prevailed, Lens would be able to say “if we didn’t make that one mistake, it could have turned out differently.” That’s always the case. Mistakes happen. It’s about making fewer than the opponent, and making them pay for theirs.
The next Big Moment was a Big Mistake.
I wouldn’t classify this as a particularly bad read or idea — it was just Raya mis-hitting the ball, which he doesn’t often do. There was plenty of space to play to Tomiyasu but it still ran away from him; Tomiyasu was also a little slow in recognition of where it was headed, and got jumped:
From there, we see the risks of “keeper as LCB.” The actual CB’s are stretched, Raya is rushing back, and a simple ball loss throws things into chaos. Both Rice and Tomiyasu lean inside, and pull up their runs a bit. One of them could have tracked Thomasson with more determination.
It was a sick finish for the score:
My view of Raya is the exact same as it was before the game. I’m bummed on a personal level for Ramsdale, as I can imagine most supporters are. But if Arteta judges Raya to be the superior option, as I do, even if it’s marginal — then his role shouldn’t be impacted by a single performance, or even a couple. Goals are going to go in against either of them and, in a high-variance position, there’s no value in ping-ponging between keepers every time a ball hits the net.
Next, we cover Tomiyasu. I was so happy to see him slotted at RCB (CCB) against Brentford, as I think that’s an ideal use of his unique abilities. His athleticism seems to be all the way back, and his last-line defending in that one was pretty epic. I was still probably lower than many for his performance in that one, as the weighting of his passes was off with too much frequency: he had 10 low losses on the day, whereas Saliba casually put in a 100% passing day against Bournemouth and I hardly saw it mentioned anywhere.
Nevertheless, Tomiyasu is looking better as a depth CB every day. I still would have started him at RB in this one, as Arteta did, but there are limitations. These final moments in the packed boxes require some high-level playmaking, and Tomiyasu isn’t in the highest tier in situations like this. Rice was visibly pissed that he didn’t see his late arrival:
I miss Timber :(
Meanwhile, before the second Lens goal, Tomiyasu lost the ball on an ill-faded dribbling voyage. Because of the rotation, Vieira was out at left-wing. He’s normally been pretty good at offering work-rate and picking up these rotations, but this change-of-possession and rotation happened pretty quickly, and he was a touch slow. Havertz did fine to track the runner and slow this down, but Zinchenko cheated up and then back and was a step behind his man before the winning cross:
The first thing I’ll say is that I’ve found Zinchenko to take his defensive responsibilities seriously this year, and he was a huge reason for the domination against Bournemouth — he stepped up and intercepted possession on six different occasions.
Through the pre-season, we saw the spot behind the inverted full-back to be a huge weakpoint in the shape, no matter who was in it (aside from Timber) — that included Tierney and Tomi, who had some embarrassing errors that led to chances. In the competitive season, it’s been less of an issue — until this moment. I haven’t really seen much of a structural indication that Havertz can’t cover the Zinchenko side against most opponents.
But that wasn’t the end of it. Arsenal still had plenty of numbers in the box for the cross. This shouldn’t really be a problem:
But the cross and finish were perfect, and highlighted a recurring Arsenal issue: aside from Gabi XL and Tomiyasu, the last-line defending can be meek, and players can still fall onto their youthy instincts of avoiding the ball.
This screengrab of our all-world CB isn’t all that flattering:
He had time to spin around. As I hinted, his batterymate often (though not always) provides a good example of what to do in these situations: act like a keeper; crowd the space with small steps; put hands behind the back; shade to the far post; twist the knee at the end to avoid a nutmeg; but don’t twist the whole body to make it small; be willing to take it on the chin:
From Son’s goal on the Maddison cross, to Palhinha’s goal on Saka, it’s a specific way in which Arsenal can improve.
Meanwhile, there was the urgent (if harried) last-line defending of Lens to keep Nelson’s chance out in the dying embers:
🔥 Final thoughts (on mistakes)
Here’s one of the primary questions on the mind of any Arsenal supporter: why so many mistakes?
As much as we’d like to chalk it off to certain players, youth, or inexperience, the truth is harder to dismiss. Jorginho made the clunker against Tottenham, and over the course of the past year, almost everybody has joined in.
There’s a growing body of research about the psychology of top athletes, and I won’t pretend to know most of it. One recent project took years to complete, and looked to compile information about the developmental predictors of the world’s “super-elite” athletes. Called The Great British Medalists Project: A Review of Current Knowledge on the Development of the World's Best Sporting Talent, it reviewed and evaluated research on the personality traits of top athletes, concluding that “super-elite athletes are conscientious, optimistic, hopeful, and perfectionist.”
I like that, and the last word was an interesting inclusion.
Perfectionism often has a cost:
Even though sport might encourage perfectionistic thinking, athletes who become preoccupied with perfection may be vulnerable to motivation, well-being, and performance difficulties. This contradiction has been termed the “perfectionism paradox” by Flett and Hewitt (2005, 2014).
The “healthier” form of perfectionism involves striving — striving to master a craft, maintaining high (but reasonable) standards for oneself, and being self-oriented in this pursuit. The less healthy version is called “perfectionist concerns,” in which the individual becomes more preoccupied with external factors: unrealistically high standards, fear of a negative crowd reaction, too much focus on mistakes, looking for parental approval, the like. This form is “associated with maladaptive patterns of motivation and emotion/well-being and unrelated to performance.” In short, it can lead to anxiety and burnout.
It’s a difficult balance to strike, and it’s not always one or the other.
But if I were to put on my best Arteta costume — Armani zip-up, and all that — I’d talk about the maladaptive aspects of perfectionism, and how, on the other side, it can be harnessed for good. Above all, I’d hope to guide them towards worrying less about the approval of the crowd, and more on the mastery of the craft itself.
But then again, I’m a guy on the internet, so listen to my bullshit at your own risk.
Let’s wrap it up with some final thoughts:
There was a situation last week where I felt I had enough information to comment: Saka was down, White was begging the physios to come on the pitch, but they never did, and Saka limped around at right-back for about eight minutes. I felt that should have been taken more seriously and he should have been subbed immediately. Otherwise, I acknowledge the information asymmetry that teams have over fans when it comes to player health. It sure feels like he should indeed play a fuck-ton, but rest a bit more; it also doesn’t feel like the Champions League is the time to do it.
What we’re continuing to discover every week is that the front-5 are pretty sensitive to the interlocking dynamics of the other players on the line, and there are rarely like-for-like subs. Trossard/Havertz/Jesus/Ødegaard/Saka would have likely worked out; once Saka went out, there wasn’t enough 1v1 pressure on the wings by Vieira and Trossard, so Jesus tried to compensate by swinging right a lot. It feels like there always needs to be at least one dribbling threat out wide at all times, and some form of combined alchemy.
That said, I was kind of flabbergasted by Vieira’s shorter cameo upon rewatch of the Bournemouth game. Almost every action felt like it could have been an assist. He also drew the foul that led to the White goal. He wasn’t as good against Lens, but I wouldn’t write him off at the RW completely; whereas you’re in the dishwasher in the midfield, the wing can offer you a better view of the proceedings. Müller and Messi always like camping out there to observe.
For all our bullshit, it feels like the fans are in general agreement about the approach against Man City: if possible, a Rice/Partey double-pivot, with Havertz up front and Jesus at right-wing, assuming Saka is out. That all sounds good by me. The most interesting choice may be at left-back, where Arteta can go for a progresser (Zinchenko) or defender (Kiwior, Tomiyasu). If a double-pivot is chosen, I’d probably lean toward the latter. Things will change based on injury updates.
ESR is making a case for an upcoming league start with every passing week.
I like the Saliba/Gabriel duo very much, thank you.
Man, those kits were so nice. Elite stuff.
Wow, that took a while.
I hope you enjoyed. If you did, pass it around.
Be good.
❤️
I love you Billy, never stop. Do you think Partey/Zinchenko/Martinelli would unlock Havertz better than Rice/Zinchenko/Trossard have? I feel like Havertz, Rice, Trossard and Zinchenko all operate best in the left half spaces and crowd each other out. The left side felt lackluster even as the right side died down without Saka
Thanks for the insight, as always. Do you think another midfielder who offers a better presence in all the phases (e.g. Matteus Nunes) should be targeted, or do you think in the long term Havertz will be able to gel in midfield?