Arsenal’s Premiere Screening
What Gabriel’s goal can help teach us about Arsenal’s screening, the rest of the PL, and whether this play might be too good, actually
So a quick recap for those who 1) don’t watch Arsenal 2) don’t have twitter or 3) have a better algorithm than I do:
Last weekend Arsenal handily beat Crystal Palace 5–0, after scoring twice in the four games before this to much derision. The first couple of these goals were set pieces, incredible news for me, before game-state created three transition-y goals to round out the win. Importantly, these set piece goals came from Gabriel, and marked his first corner goal(s) since March. Also importantly, the first of these goals especially has garnered a whole load of engagement (thank you PL twitter account) regarding the use of little Leo Trossard as a screener for two of Palace’s core defensive unit to create the space for Gabriel’s thumping, crashing, etc., header. Naturally this prompts the question of Why? Why has Gabriel not scored from a corner in nearly a year? Why does anyone give a shit about 5ft8 Leandro Trossard not going for a header? Why is this not a standard play across the league if it’s so interesting?
Well, you’ve come to the right place, but please stop asking so many questions. This all links together, don’t worry, but we’ll break it down.
Gabi’s Goalscoring Gap
Prior to this goal chasm, Gabriel had been a consistent Arsenal set-piece scorer under Arteta and Jover’s tutelage, and it doesn’t take an especially deep look to see why: he’s immense aerially. There’s a reason his relationship with Saliba is so good, and that’s his complimentary dawg to Saliba’s feline nonchalance. Battles in the air or on the floor are his domain, and whilst he actually doesn’t have a crazy success rate, sitting at 62.7% per fbref, I would put money on him winning the vital ones. Since his arrival, Gabriel has scored at least three goals in every single season and had plenty of close chances that make him a natural target for our routines. Until this season.
Side note: I wouldn’t be too secure on success rate judgements. Zinchenko has a 99th percentile aerial success rate simply because he picks and chooses his battles. Gabriel has 2.89 aerials won p90 and that’s enough for 84th percentile, which is great for a player not in a deep block side.
This season — as I covered extensively before — we haven’t really been approaching corners in the same way. Part of this comes from volume considerations. Arsenal are getting two more corners every single game than last year, and the season before that, and so naturally these opportunities are actually less vital on a ‘per set piece’ basis. These can now become parts of a wider gameplan within both the 90 minutes and the season, rather than an immediate “oh shit okay we’ve got a corner, let’s do something”. This logic carries when you consider Everton, who are the second most successful set-piece side in the league this year, have 4.71 corners per game. Smaller volume suggests more emphasis on scoring from each one, and leaves fewer opportunities for testing — compare the ingenuity from Brentford too for further argument. The Athletic’s latest ranking of our ten best routines shows a whole bunch of interesting movements and tactics, but a key feature here is that you’ll see the most innovation in previous seasons, where we had more focus on surprise and less on enforcing learned patterns — I’d argue as a result of the reduced volume.
This fits into our wider concept of domination and control too. A lot of focus in the first half of this season has been on the front post, and dominating the notorious ‘first contact’. This fixation is logical within our game-model, as shots from here typically will go either in, or out of play, preventing transitions back (most of the time, Forest clearly taught us a lesson on this opening day), with any clearances from here likely returning to Saka/Nelli. There’s also a sense of Pavlovian behavioural training from Jover and co here. A key trait AFC use is ‘cat and mousing’, wherein sudden switches in our corner targets allow us pretty decent chances as they operate against working theories from opposition defences — aiming front post for a half and going deeper second (see Rice vs Utd, Kai’s scuff vs Sevilla H)
This leaves scoring opportunities few and far between for someone like Gabriel, whose role has primarily been to secure the front post, or engage in fairly static duels at the back post and hence become reliant on front-post touches — which we’ve not benefitted from quite yet. Brighton saw what can happen if we do, but even this required five bodies at the front post and Brighton to get the touch. Rarely are we seeing the committed crashing that Gabriel is notorious for, and teams are now so clog-focused on the front-post to ensure they can get the first contact, that we also haven’t seen goals like Leicester and Brentford A for a while either.
Previous Arsenal strategies also heavily favoured Gabriel, and then Gabriel and Saliba, as they were essentially the only tall players on the team — which I’ll admit isn’t mind-blowing analysis. We now have five players over 6ft starting weekly, and White’s heading seems to be really improving too (his goal vs Bmouth, flicked header vs Brighton, stunning finish past Ramsdale in The Bobbins) so threat can be spread nicely rather than laid upon Gab’s angry Brazilian shoulders.
In sum: we don’t need to let Gabriel do everything now, and this has reflected with us scoring 15 set piece goals last season compared to 12 already this time round (plus I’m fairly sure Gab’s second is an OG in stats-speak, so we’re at 13). Tough on the FPL owners, but great for Mikel.
Little Leo’s unLocking
Anyway, this was before, and now we have… well, now. Analyses of the Palace routine are everywhere, look to the potshot podcast account for a nice one, but the boost for Gabriel here is that Palace took the front-post bait, secured by a rushing Saliba, and Kai secured the separation far post as well. This is another tactic we focus on heavily, spacing out players from both sides so that we create clearer shots for our potential receivers — often there will between three and five players out of the ~twenty in the box between the posts. Gabriel is then allowed to return to his older tactic of crasher and destroys the Palace CB — doing so again for the second goal.
A lovely return for box-crasher Gabriel, who I’d mourned the loss of on potshotpod (yes I’m plugging it again), and yet all the focus is on Trossard for turning his back?
The screening tactic is very visually intuitive, and it looks every more effective when you show someone as ikkle as Leo can pull it off on two very good Palace CBs. I’ll briefly talk about its importance relative to Arsenal, but the beef of this article is a bit later.
Trossard’s work is novel for Arsenal this season, and should be celebrated as it goes against a lot of our core traits:
Where we normally clear the full central zone, and favour the near post, Trossard secures space in the goalkeeper’s potential claim zone — amplified by the work of White, notorious GK shithouse that he is. Gabriel’s header ‘aim’ is restricted, sure, but his arrival zone is so totally free of distraction that it no longer matters. This all comes with the benefit that in the VAR era, GK fouls are a minefield for disallowal potential (White vs Leicester), and so securing the claim space without actually focusing on the goalkeeper is a neat way around the issue.
Using Trossard this way is also a neat method of organically producing more insecurity in defensive setups, as now even the smaller players prevent real threat. Whereas previously we get 8 players in the box and try and smash it at ’em well ‘ard, kinda like Liverpool but without Trent’s immaculate whips to the penalty spot, this now demonstrates an adaptability we haven’t particularly shown before now. I’m hopeful this will create more space at the near post again as teams adjust for the new techniques, or give Kai another chance to scuff the ball into the ground at the back post if teams become more Gab-focused in man-marking orientation again.
Looking immediately forward, there is some potential of this in upcoming games — Rice notably sat as a counter plug on Saka-side corners in the first half, and will be doing so again for Forest, LFC, and WHU, meaning the average height in the box could tank again — but I would be shocked to see Rice as a taker again for that reason. Martinelli can spank a near-post whip fairly well, so I’d guess we’d see Saliba + Gab back there with Jesus and Havertz returning to back-post floaters, but I’ve been wrong before and I’m not paid 50k a year like Jover to try and correct that.
Barclays’ Blocking Bias?
To your third question then, thanks for being so patient. Here we’ll try and relate this to how Arsenal can learn from everyone around them — how they differ from current blocking techniques, whether it’s viable, and really whether they even should imitate.
A recent tweet by Michael Caley of DoublePivotPod and stats nerd fame suggested that screens and blocks are actually *too* effective in current form, and that soon the league will need to intervene to stop games being dominated by them. As part of this deep dive. I’ll be testing whether this is true, as effectiveness is a core part of the questioning above.
Special mention to https://twitter.com/Ewansfooty who’s just kicking off doing some daily looks at set-plays across the leagues as well.
So, I’ve now watched basically every corner scored in the PL this season. Whilst this is probably not perfect, I feel it gives us a nice view of what actually works, and also means I don’t get into wake up, corner analysis, repeat. This gave us a core group of teams that screen ‘best’: Everton and Brentford, the answers that will surprise no one, along with West Ham, and Luton a bit too.
Everton primarily focus on this due to their back-post corner bias, which in many ways necessitates screens. By this I mean that back-posters will need to pass down through/across the 6-yard, and most keepers worth their salt are likely to be aggressive with floaters in this region given the opportunity — hence Everton regularly place at least one man in front of the keeper, especially on Dwight McNeil’s side. (Data per @From_the_wing)
They also require unimpeded runs to the back post, which requires screeners to prevent the space being attacked being full of freely-moving CBs eager to keep a clean sheet bonus, and freaked-out GKs looking to not piss off bonus-loving CBs.
Take a look at their goal vs Brentford, for an example. Here, Doucouré occupies Flekken centrally, and once it’s clear the claim is not on for him, Flekken retreats. Doucouré meanwhile activates ‘annoying dickhead mode’, and forces his back into two Brentford markers who have been sucked into the space created by their vacating keeper. This does the beautiful thing of giving Tarkowski a clean 1v1 and hence free header from all of four yards, at the back-post again.
We also have the below versus Burnley, in another example of people bullying James Trafford. Everton have clearly identified Trafford’s box domination as an issue, and place a two-man screen to prevent any hope of collection, of Doucoure (Everton’s Ben White) and Godfrey. By the time Onana is up and heading, Trafford is met not only with a ball flying away from him, but a wall that necessitates him jumping about one second earlier than he’d like. If that doesn’t seem like much, go try a high jump one second early and let me know what your local athletics department’s concussion protocol is like.
Can Arsenal inherently learn anything from this then? This is likely why Rice has been brought in as a taker honestly, to give us an improved deep inswinger which Trossard and Martinelli have so far struggled to provide. We’ve actually only scored one goal from the LHS corners this season (Saliba from Trossard vs Burnley, a Trafford bully-session again), and hence blocking has served very little purpose outside of the confines of a front-post squash and a goalkeeper baby-sit at best. Rice therefore gives us the ‘McNeil’ type deliveries that allow for intricacies in screening, and as a result we maybe see White’s role grow in increasing prominence, and a ‘guard-rail’ screen developing across the back end of the 6-yard box, like Trossard/Doucoure+Godfrey.
Side note: Everton also have an example of the ‘screens either side of the header’ that Gabriel’s goal resembled — vs Tottenham — but Andre Gomes’ goal comes from the second play which is a whole different thing that Everton are good at.
Brentford’s best example is this creation of back-post space, just last week vs Forest. Again, the player doesn’t make an attempt for the ball, but the blindside screening is so effective for ensuring on-rushers’ freedom. This didn’t result in a goal, mind you, but it basically did considering it’s a goal-line clearance, and it’s also a shot at Iñaki Caña alumnus Matt Turner.
For Arsenal? We’ve already seen something kind of like this, below:
Whilst the locale is different, the movements and counter movements are roughly the same, and with Kai’s increased size and strength at the back-post, a screen here is likely and would be welcome in unlocking the predatory movement of Gabby J and Martinelli; I frankly think our attackers are quite wasted in our current corner setups so this would be a massive help for unlocking them from just ‘nothing’ roles.
Other teams benefit excellently from this, and its unsurprisingly those blessed with physical mountains. West Ham are great at blocking centrally, allowing Aguerd to do his best Gabriel impression, and boosted by JWP doing his best JWP impression. Luton notably also do this vs Arsenal in that game, for their first goal. The screeners ensure Rice and Gabriel cannot engage with the runner behind all the chaos, and Arsenal are totally ill-equipped to deal with it.
This feels like something Arsenal aren’t particularly well-suited to learn from, given the lack of serious powerful headerers like Luton, or pinpoint central deliveries from JWP, combined with the total aversion to outswingers. Again, perhaps this delivery ‘problem’ is something Rice lets us benefit from, but this feels like a bunch of optimistic projection right now. Defensively it’s suggestive of a need for allowing Raya to be the most aggressive in his starting position, to ensure this sort of blocking is limited to the far reaches of the box rather than right underneath the crossbar.
Does Caley have a point then? If the best set-piece team in the country isn’t really utilising screens as much as the second-through-fourth best, then perhaps they are OP? (OP meaning overpowered, for the dads). Well, partially. These routines are still great works of deception, and to place their success purely as a result of a guy standing looking the wrong way is a bit harsh. Everton and Brentford’s, for example, are primarily a result of splitting decoy runners, and more often than not the space they create is a result of a runner charging front-post to ensure these zonal markers can’t retreat (Lewis-Potter in Mee’s goal vs Forest, Branthwaite for Tarkowski vs Brentford).
Combined with this is that the ‘Gianni Vio play’ is still immensely popular, and has been the play for about four-to-five years now, as noted on the latest Tifo Football podcast*. This naturally attracts players to the front post — Man Utd notably have a ~four-man cover there every single game — in hopes of securing that first contact, and preventing the flick into all the space behind. Teams are being punished for being overly fearful of this play through screening (Dendonker vs Utd). Screening perhaps isn’t generally OP, perhaps it’s just OP right now in the boom following Vio.
*For the uninitiated, this is the ‘flick front post, header back post’ one.
Even with this, teams are still all individualised in approach. Bournemouth for example are quite individually focused due to the quality of Senesi, and rotate through a wealth of options most weeks. Liverpool meanwhile use VVD as a core part of their work, using smart counter-movements from their high-quality finishers to allow him to act as both scorer and provider (often this will be hitting VVD deeper due to the outswinger favouritism, with a Jota/Salah rotating through to the front post on the markers blindside). Counter-movements are Spurs’ bag too. As it stands, screening is seemingly a tactic most used by the set-piece nerds in the PL world. I’d agree with Caley that in the future we’ll likely have to move to stop these — especially since the advent of ‘cloggier’ press-focused games — in order to stop the sport becoming NFL-ised, but right now all you’d be doing is docking Everton another 10 points a year — and feel free to tell Goodison that’s coming.
So, finally, back to Arsenal. Should screening be ‘our thing’ now? I mean, it already kinda was, albeit with a front-post-squeeze bias. What we can learn from these sides is the better usage of our smarter forward players in relation to the giants. Jesus, Martinelli, and Trossard (and to a weird extent Kai) all have value in off-ball movement which is currently, mostly, unrealised in-game, and a more active engagement with central screening could set them up well for intricate movements that we’ve slightly lost this year. I’m not saying to become Everton or Brentford, because even I’d kick off a bit at that, but it’s another weapon for us to arm ourselves with, especially with eight chances of this a game. Jover and the boys have established we have the crash bang wallop approach down. Now perhaps let’s show them we never lost the artistry too.