Raya? But whya?

A scouting report on the audacious Arsenal target. But also: how Arteta is looking to take a weapon that has been used against him and turn it back at his future Champions League foes

There is an under-appreciated advantage to becoming a top side, especially if you happen to use some of the more “meta” tactics of the age.

The only catch is that it doesn’t initially appear as an advantage. It appears as a goal — against you.

Here’s the syrupy phrase that gets to the heart of the matter: “You don’t lose. You win, or you learn.” And as a top side, one learns, often in the most direct and painful way possible, what causes the biggest headaches for a top side.

---

Arsenal presses in a man-to-man scheme with some hybrid elements. They are bold, energetic, and effective — ranking second in the league for tackles in the attacking third (with 118, behind only Leeds “DGAF” United). With the additions of Kai Havertz and Declan Rice, as well as the feisty and front-foot Jurriën Timber, they look, at least on paper, to be even better this year.

Though Arteta offers some truly interesting little tweaks, including some new seesaw pressure stuff that could dial up the intensity even more, the style shares similarities with the dominating philosophies of many of the top teams in Europe.

Here’s one incomplete way of proving that point: the top-20 “most intense” pressing sides in the European Top-5, judged by passes per defensive action (PPDA), is a list that includes Arsenal, Liverpool, Milan, Napoli, Barcelona, Bayern, Brighton, Dortmund, Chelsea, and Manchester City. Sounds pretty Champions Leaguey to me.

Put simpler: this shit is popular because it works. But how do you beat it?

---

Last year, Arsenal faced a common recipe. This recipe called for less-than-committed build-up play, loads of long-balls, roaming 10’s, and a hearty helping of manufactured transitions. We’ve covered a bit of this lately.

Against Manchester City, Pep changed tactics, going to a 4-2-4 double-pivot build-up, with Gündoğan joining as a second #6. This stretched Xhaka and Partey, and Haaland was able to win first balls, unlocking KDB on the second:

Sporting CP pulled the press up with some perfunctory build-up play, before going long to Pedro Gonçalves, who looked to rip Jorginho across the pitch. Look how far the next line of Arsenal defenders are on a play like this:

Even Brighton, the building-out-the-backiest of teams who build out the back, would do a lot of deep and seemingly committed first-phase stuff, before ferrying the ball over to Levi Colwill, who would wait for Mitoma to find the right amount of isolation, then backspin-blast it over the top. You’ll also note Enciso cutting across:

What is less-appreciated about De Zerbi’s tactics isn’t the 4-2-4, it’s the 4-2-[HUGE GAP]-4, which stretches the press as high as possible, giving attackers time to go 1v1 once it’s broken, and midfielders space to dribble as needed. Brighton created more shots off of take-ons than any other Premier League team by some margin, and a huge reason why is how they fabricate space.

Which brings us closer to the topic of the day.

We’ve written about Raya’s ridiculousness before, both before and after the game, and saw it come into clear focus when Ivan Toney ruined Arsenal’s day — winning 12/14 of his aerial duels, and positively bullying Saliba. The monster starlet actually looked his age for once, playing the ball instead of muscling for position, and winning neither:

Saliba has seemingly improved in the air since. Toney is a bit like Steroid Jesus — Gabriel Jesus, that is — in a lot of ways: all of the pure intelligence, savvy, and muscling instincts, with a little more size, jumping and strength. As many aerial duels as he won there, he generally doesn’t just try to win the duel, statistically — he tries to win the situation. It often results in the right kind of bouncing ball.

But the other star of the show was the person offering all that delivery: David Raya. In addition to whipping all those long-balls, he also threw in a couple moments like this:

Meanwhile, there’s been an interesting through-line to much of Arsenal’s summer business. Kai Havertz, the first signing, was #4 amongst all non-defenders in the Premier League in winning aerial duels. Declan Rice is one of the best recovery and second-ball guys in the world.

Two other factors loom.

• The first we covered last week in Anatomy of a goal-kick. Last year, the team loosely committed to a distribution strategy of deep build-up against weaker opponents, and long-balls against more evenly-matched foes. They went long more than most of their top rivals, and did so to (mostly) great effect against Barcelona. By showing both, defenders risked getting caught in the middle.

• The second is a clue hidden in the games of those two midfield additions: the biggest question-mark for each is in their early build-up, where Rice’s back-to-goal game can be a little conservative and Havertz can be a little gangly. Hmm. Perhaps they can make do against weaker sides, and skip the phase as needed against the fiercer pressers?

Meanwhile, this year’s build-up has featured 13 different pivot pairings by my count, including work by Kiwior, Havertz, Tierney, and ESR:

This raises other questions. Is the pivot getting more modular? Are they introducing the concept of a “false pivot,” meaning a player who occupies defenders centrally, but mostly does quick recycling?

It sure looks like Arteta is looking to build a team that not only excels in modern build-up from the back, as Timber and Kiwior dramatically raise the floor of the backline ballplayers, but who also have enormous potential going long — so he can take a weapon that has been used against him, and turn it back at his future foes in the Champions League, ensuring maximum unpredictability.

Backlines everywhere are increasingly exposed and stretched man-to-man, and Arteta knows as well as anybody how to ruin their days. If so, Havertz and Rice continue to reinforce their schematic fits. Jesus forcing second-balls to Ødegaard and Rice, who playmake from there? Havertz doing flick-headers to Martinelli doing Mbeumo runs in the blind spot? Sounds good to me.

It may be less surprising than it seems, then, that Arsenal have reignited interest in one of the better long-passing keepers you’ll ever see.

---

Enter David Raya

If you’re enough of a dork to be reading this, you’re probably familiar with the contours of his career, but here’s the TLDR.

Raya is a Spanish goalkeeper who likes collecting figurines and started playing a lot of outfield futsal while training at his primary position. In his younger years, he joined Rovers, and went on loan to Southport, before returning to the Riversiders to back up the elder Jason Steele. After relegation, he took over as the top choice, and helped lead a campaign that won them promotion back to the Championship.

He eventually moved to Brentford, where he was a cut above other ball-playing GK’s in the Championship by a long shot; speaking of long shots, an error in the play-off helped solidify their position there for another year, but he pushed them to earn promotion the next time up.

Brentford have come to rely on his abilities: when a collision forced him out last year, the Bees lost 9 of 14 league matches, and gave up 29 goals. When he came back, they let in 18 goals across their remaining 15 matches. He’s become renowned for his ball-playing abilities, but isn’t particularly tall, and hasn’t won significant minutes for Spain yet.

He’s plenty experienced, starting 264 games over his career, and has long had the eye of top managers: Tuchel is interested; Klopp said he could wear the number 10; and Arsenal has previously bid for his services. Last year, he made the most saves and had the most touches of all Premier League keepers.

But if anything needs context, it’s GK stats. I feel pretty comfortable with his game, so for this, I rewatched fewer full-90s than I usually do (just two), but then watched a huge swath of his distributions going back to the Championship, as well as all of his conceded goals and some other key runs of play.

Throughout, I was more interested in understanding him on his own merits than in comparison to Ramsdale. My confidence interval is probably wider with keepers than other positions, as there’s all kinds of miniscule stuff that only a real keeper would know. Read this, sure, but conclude for thyself.

---

Passing

Raya is one of the best ball-playing goalkeepers in the world. Armed with a huge array of short balls, long balls, curvers, drop-kicks, bowls, and volleys, his technique is repeatable, urgent, and confident.

Below, you can see what he does best — claim the ball, and drop-kick it without hesitation, setting off a counter the other way:

On goal-kicks, he generally pushes it long these days, but going back to the Championship, he has long (and successful) experience as a deep controller, which he attributes to a simple insight: keepers are pressed less directly than centre-backs, so they should fare better.

Below is how that looks over the last year. You’ll see only successes in the own third (77/77), and more variance when moving forward:

In general, a lot of his distribution lately has gone long and to the left, where Toney looks to win position and get second balls to his midfield and over-the-top flickers to runners like Mbeumo:

Are his long-passing stats inflated by the presence of Toney? Probably. But as we covered before, Toney doesn’t necessarily try to win every duel himself (48.2% this year), but exploit the situation the best he can. The other point in Raya’s court is that his long-balls just pass the eye test and then some. They bend and swirl and back-spin at will, and he can do them with either foot without issue.

I’ve read some light concerns about his short-passing game, but I’m not sure I share them. We should never confuse a player’s current responsibilities with a player’s total capabilities. Over the last 5 years, he is a 98.8% short passer, and though there are definitely a couple examples of improper weighting or miscommunication with the backline, I offer two bits of context. First, I’m not particularly enamored with Brentford’s build-up players. And second, every keeper with this amount of touches will have a couple of those, and this is well within any reasonable proportion for his role. I’d expect a couple more if he came to Arsenal. He profiles similar, if slightly above, Ramsdale in the short stuff.

That said, he’s had a gap in his aggressive first-third passing CV, so there could be an adjustment period.

As a sweeper, he positions himself quite high, and one of the more likable aspects of his game is his anticipation: when his team needs a bail-out, or a bouncing ball or save heads his way, he is not getting the ball then deciding what to do — he’s already decided.

Look how quickly this near-steal turns into an attack the other way:

It’s almost Partey-like in how his first touch can be a through-ball.

His throws are quick and ambitious. Overall, he sometimes pushes a little too fast, and every single ball is not strictly perfect, but I have few notes on his game here. He may not be quite as swaggering as Ederson or Onana, but I’m not sure he’s any less skilled.

---

Shot-stopping

Almost every keeper stat is situation-dependent. Last year, he faced the most shots on target, had the most saves, and posted the highest save percentage. The baseline is good, but we must evaluate further.

For however noisy keeper stats can be, there’s one I like best by a margin: PSxG+/-, or post-shot expected goal plus/minus. This is, in effect, a tally of how many goals the keeper saved (or didn’t) based on how much we should expect them to save it.

For years, Raya has performed about as expected, if a smidge below. Last year, he saved five more goals than expected, good for fourth in the league. He’s generally been on an upward trend here:

In my watching, he has shown a quick set of reflexes and reliably good positioning, especially last year, with quick little steps allowing him to recover towards the ball, and a couple limitations that are worth discussing when we’re through with our compliments:

This and following graphics from Wyscout

I’ve seen a stat circulating that a high percentage of his goals conceded are from out of the box, which is a good example of the kind of stat I hate. For example: if a keeper only let in four total goals in the season, and one of them was from out of the box, his tally would be 25%. If another keeper let in one hundred goals, and 20 of them were from out the box, it’d be 20%. Who is the better keeper?

Brentford, who packs the box with giants as much as anybody, simply provokes a lot of long blasts: their average shot length is the furthest from the goal in the league (18.4 yards). I looked through all of them and while he couldn’t get fully Courtois-heroic on the goals, they were all bangers that most guys wouldn’t have saved.

In the top-right of the below graphic, you’ll see the two charts that provide a simple, encouraging picture: a lot of shots faced, not a lot of goals conceded.

He had a few of these, including some against Arsenal last year:

These saves are increasing in frequency but not too regular of an occurrence. He is athletic, though not completely domineering, in 1v1 situations, and conceded a lot more in the past on those. It seems to be an area that is improving.

He’s not without issue, though. For one, he is not the upgrade on penalties you were hoping for:

Which may correspond with the big one: at 6-0 or 6-1, he is short for an elite keeper. He compensates in a few ways, not least of which with bursty athleticism, quick steps, and a mature feel for angles. Another way is by sweeping up crosses — he is second in the league in both crosses swept (52) and percentage (8.7%). He never really looks small on those.

The big problem is that he is good enough to get his hand on the ball but not long enough to actually secure it, which results in a lot of second opportunities. Brentford led the league in shots allowed that create another shot with 74. This is partially due to the kind of shots they concede, and the actual concession numbers are more modest (9 rebound goals in total, good for 7th-most). Raya frantically tries to clean up the bounce, but it can lead to some nervy moments.

Here’s the ugliest version, for your records:

Oy.

Elsewhere, he can have good judgment on coming for crosses, but if he stays back and the cross is good, he can feel really exposed (and small) in the middle of the net. Going through all his concessions this year, this has been the most reliable way to score on him:

Witness enough football banter, and you’ll notice that there are a few rites of passage for a fan of any club: they are uniquely aggrieved about their treatment by referees. They all think they are cursed. And, except in a few rare cases, they think their striker misses too many chances, and their keeper is prone to some howlers.

Raya had a few errors to his game in the Championship days, and had a rough one against Newcastle, but by and large, he’s ironed most of that out — and looks pretty consistent.

If you have a similar appraisal of his passing ability as I do, and merely see him as an average shot-stopper, you nonetheless have something pretty elite. If you have a top-four Premier League-shot stopper (or better), and one of the best ball-players anywhere, then it gets well and truly exciting.

---

🐻 The bear case

Time for me to put on my hater hat:

• Goalkeeping is communication by another name. Passing, scheme, shot-blocking, corners, free-kicks: it’s all a team sport. Any new keeper carries some risk of settling in. Brentford’s system is not always similar to Arsenal’s, and there could be some awkward moments that lead to significant questions.

• There is legitimate concern that Brentford’s band of giants have insulated him from being exploited more on crosses — I’m a little wary about corners. There is, also, a worry that Toney is the tide that lifts all boats.

• His shot-stopping stepped up another level last year. If he reverts to previous form, any perceived gap over Ramsdale will narrow considerably.

• …and finally, and most obviously: you could wind up angering a lifer, unsettling team dynamics, and racking up some opportunity cost in devoting this money to a position that is already in a good place. If Ramsdale winds up accelerating his trajectory, you may just create a confusing, expensive situation that could have been avoided.

OK, let’s wrap it up with some of my bullshit subjectivity.

---

The final verdict

OK, now I’ll let you know my rather embarrassing starting point. I have vague recollections of posting, then quickly deleting, something about how “Raya is the third-best keeper in the world for me” last year. I spent the early weeks of summer praying Spurs wouldn’t get him. I’m not sure I even totally believed the first claim at the time, and that episode will give you a sense of my relationship with take culture: fully human and susceptible to the pressures, but ultimately, grossed out and keen to not add more shit coals to the shit furnace. I wonder if you can relate.

My keeper opinions should be held in some doubt, so take everything as such. I don’t necessarily think he’s the third-best keeper in the world, especially after going through this and attuning myself to some of the picadillos of his game, but I also don’t think it’s an outlandish opinion, either. I do think there are two keepers I like above all — Alisson and Maignan — followed by a group of 8 or so who are brilliant but carry a wart or two. I think Raya is firmly in that group. I don’t think Ramsdale is there, at least not yet.

This is no knock against the latter. I feel a profound sense of loyalty and gratitude towards him, and think he’s quite good. If this falls through, or he’s otherwise the first-choice for the rest of the year, consider me happy.

But the best reason to be excited for Ramsdale is for his performance relative to age. He is very good already, occasionally sloppy and impulsive, and young for a keeper; if you believe he keeps up that trajectory, he’ll be in that highest tier in no time. Speaking personally, I think either of these scenarios carries an equal chance of coming to fruition: a) he turns into a bonafide star, or b) he continues on as a distinctly above-average keeper with a persistent, and sometimes-frustrating “vibes” element to his game.

Raya is a player in his prime who has already demonstrated most of Ramsdale’s less-certain potential. They offer similar shot-stopping peaks; Raya is clear now, for my money, but Ramsdale may offer higher potential. But I think Raya is ultimately a step above where Ramsdale can get with his best quality: his feet. He’s wild, man.

As such, my entire being urges me to say the company lines: competition is good; no spot is guaranteed; Raya is a market opportunity, and is not signed as an upgrade on Ramsdale, but on Turner; he’ll have to earn the number one role. Perhaps the biggest? This removes one of the only remaining “doomsday injuries” from the squad. And that’s all true, but not the whole story.

What is also true is that I think Raya genuinely should have £30-40m+ value on an open market, especially for teams starting over at keeper. Considering the Ramsdale/Turner tandem, though, I probably wouldn’t have the balls to make this move if the budgets were mine to keep. If the mindset is less thrifty, and more “fuck it, we need the strongest player everywhere to try and win the fucking Champions League,” calculations change.

That is another characteristic of Arteta’s moves this summer: a devilish impatience to get Phase 4 moving. In the form of dialogue:

Standard-issue human: “You know, Rice hasn’t been properly coached on some of the nuanced aspects of the toughest position in world football. I think we could phase him in over a year before he fully takes ov-

Arteta: “Now.”

Human: “OK, next up is Jurriën Timber. Most of his relevant experience is at CB, and he’s never really been a ball-dominant midfield--

Arteta: “Now.”

Human: “Havertz was a striker. We could ease him in across the lin—

Arteta: “Left attacking midfielder. Now.”

This logic, then, seems to extend to the question at hand. Should we continue to put our eggs in the Ramsdale basket, or see how the year goes, then look for competition?

“Now.”

That’s why I’d err to the simplest explanation. I think they’re interested in Raya because they think he’s a level better than Ramsdale, as I do, and they had expected his ultimate sale price to be higher. When it actually started looking more viable, they started moving.

You may think to yourself: this still seems like a lot of money for marginal gains.

I consider him more of a genuine upgrade, but even so: marginal gains have compounding effects that reverberate throughout the pitch. At this level, marginal gains are all you have left.