When Arsene Wenger shifted to a back three late last season, there were some pundits who called it a move born out of desperation. Some of these were the same pundits who had long clamoured that Wenger’s Arsenal sides followed the same script every year (fair) and always succumbed to the same weaknesses (somewhat fair), overlooking the blatant hypocrisy of this ‘desperate’ narrative position in favour of sensationalism and laziness, as punditry is wont to do. Some others, meanwhile, viewed the switch as a copycat move, of merely doing what others had already done, sufficiently long after they had done it to not be so much an inspired move as it was a reluctant one.
Perhaps it was a copycat move. Last season, Chelsea cantered their way to the Premier League title behind Antonio Conte’s 3-4-3 system – the canter only began once they had adopted it. But they did not invent it. Chelsea, it could be said, copied it from Barcelona, whose play-from-the-back, flood-the-midfield style that proved so successful under Pep Guardiola started with a hybrid three-at-the back system, at least when in possession of the ball. Prior to that, Roberto Martinez and Wigan Athletic somehow got there before everyone else. Perhaps we are all now living in Wigan Athletic’s shadow.
Either way, Wenger made the change. And having stuck with the formation thus far this season, the shift looks to have now taken root going forward. Despite the permanent urgency of Premier League football in the minds of the pundits, fans and - Arsenal’s excepted - majority shareholders, any systemic change needs time to bed in, yet with half a calendar year now behind it, there has been enough Arsenal football under the 3-4-3 to appraise its results so far and its potential going forward.
Wenger himself cites one of the reasons for a shift to a back three as being the previous back four being vulnerable to vertical balls down the centre of the pitch. And while he did so without citing certain individuals, the playing style of his first choice centre back was self-evidently a factor in this decision.
Looking into Arsenal’s back line
Long one of the Premier League’s best centre halves, Laurent Koscielny is one of the more mobile players at his position in the league, and still is, even as he ages and suffers injuries with increasing regularity. This mobility, combined with his composure on the ball, allows Koscielny to push up, gamble, and intercept, to step beyond the back line into the mid-range area when circumstances permit or require it. Moreover, Koscielny, quite good in the air, would often step up to attack the high hard ones coming at him.
However, no one wins all the interceptions nor all the headers. And when Koscielny is caught out gambling, his centre-back partner would be even more caught out than he was. The problem with having one centre-back wanting to push up the pitch to anticipate play (particularly when it comes to aerial long balls) is that, if he is unsuccessful, the other centre-back is now isolated. And when that other is Per Mertesacker, this invariably creates a scoring chance for the opposition.
Mertesacker’s anticipation, vision, and timing of those little disruptive tackles around the edge of box have always been exquisite. But, he has the mobility of a bag of spanners. Mertesacker caught in space by opponents running at speed is a problem. It matters not if he knows what to do when he hasn’t the foot speed to actually do it. Playing him alongside Koscielny, as was the case for a long time, often led to this mismatch being exposed.
Playing either Shkodran Mustafi or Gabriel Paulista alongside Koscielny was a theoretical fix to this problem. Both are more mobile, as well as being more skilled and comfortable on the ball than the German, thus potentially giving the team two Koscielnys, which would have been enough of a foundation for any back four. However, Mustafi struggled at times in his first season of English football, alternating good stretches of play with error-filled streaks, losing his position too often, and overcommitting to low-percentage tackles. While Gabriel struggled in all his seasons and just wasn’t good enough, eventually being sold to La Liga outfit Valencia.
In addition to this, when playing as more conventional full-backs in the back four, Nacho Monreal and Hector Bellerin still had to provide most of the team’s width going forward, especially when it came to getting to the by-line, due to Wenger’s continued love of inside forwards. For them to be able to push so far forward and still not allow opposing wingers or inside forwards to get in behind them as they ventured up the pitch relied upon the central defenders being able to come across and cover without leaving space of their own. This, too, did not suit the leaden-footed, and the centre midfielders did also not do enough to cover the space opened up.
Seemingly, then, Wenger’s logic is that a back three protects against that by having more cover behind any venturing defender, be they a gambling centre-half or an attacking full-back. An extra man in the middle of the park means less space for opponents to counter attack into, an oft-exposed weakness of Arsenal’s, especially against the better opposition. In this respect, occasional positional error notwithstanding, it has worked.
Conversely, however, having a back three confers extra playmaking responsibilities onto the defensive line. With one less midfielder in the centre of the park, the attacking possession must start more from the very back. And while Arsenal’s centre halves are secure enough at passing the ball along the back, they are, the occasional needle-threading through ball from Mustafi excepted (at the cost of many intercepted ones), not especially good at igniting incisive passes to the attacking unit, nor bringing it forward themselves.
Arsenal’s Key Problems
In the most recent game against Swansea City, Koscielny occasionally forayed forward and tried to make some plays from the back. But all too often, he then sprayed the ball into the car park. Arsenal’s defenders are neither especially willing nor able to get long balls into the box, nor to fire shots up from deep, and while both of these things are in keeping with Arsenal’s overall philosophy, they also speak to Arsenal’s main problems scoring the ball – stagnation.
When attacking, Arsenal have always been somewhat negatable by flooding the area around the box and denying the through passes their bevy of inside forwards want to thrive upon. The 3-5-2 somewhat counters this by providing extra width via the wing backs, but when playing three at the back, there is one less person around the edge of the box to pull the defenders around. If one of the three at the back could step up in these situations, effectively acting as a centre midfielder when attacking before dropping back to defend, this would help. But none of Arsenal’s current centre half crop have shown enough ability in this area.
Mohamed Elneny, a career centre midfielder with a willingness to defend more recently used of late as a centre half in the 3-4-3, wants to be that playmaking centre half, and theoretically could be. A precise, keen passer, Elneny’s career to date has involved lurking between the half way line and the edge of the box when attacking, refeeding the offense when it cannot penetrate the back line, and tackling tough in the middle of the park. Yet, in his time at centre half, he has too often been exposed at the back; his natural impulses to push up the field and be a midfield playmaker are overexposing him at the back, and he has been a central defender in name only. A Mascherano-esque full-time conversion is not coming.
This is not to say that Gabriel should not have been retained. There is no argument for that; Gabriel was not a Premier League calibre player. However, as the right sided player of the back three, what Gabriel did do was enjoy the freedom to push forward, just as Monreal currently does on the left. Had he been able to defend adequately without resorting to fouling (or sagging too far off in order to avoid fouling), he might have made it. Mustafi should be the right sort of player, the player Gabriel was expected to be, the commandeering and fearless centre half who combines determination with ball skills and shores up the middle of the park, while still pushing on with the ball and feeding the attackers. But this has yet to materialise for more than a month at a time.
Calum Chambers and Rob Holding – Chambers in particular could be fit for the role going forward, considering the brief but genuine promise he showed amidst has way-too-brief stint as a defensive central midfielder before his loan to Middlesbrough – have shown promise, but both have ways to go in all facets of the game. Monreal’s abilities going forward help in this area when he plays at the left sided centre half, yet his deficiencies in the air aggravate what is already one of Arsenal’s problems in conceding goals from set pieces. None, then, are quite right. And while the trio of Mustafi, Koscielny and Monreal with Sead Kolasinac and Hector Bellerin as wing backs is the best incumbent five-man unit for the system, injuries have meant we haven’t seen it yet.
Suiting the Arsenal way
The three-at-the-back system itself is fine. Inasmuch as any one system itself cannot be held accountable for an inconsistency and lack of resolve in pressure situations, it stands up to scrutiny, and is theoretically the answer to some of the team’s longstanding limitations. The results thus far are also fine in the sense that, ability to occasionally self-immolate notwithstanding (last 20 minutes of the Watford game, all 90 minutes versus Liverpool, etc), eight wins out of their last nine and being in a points tie for fourth in the league is par for the Arsenal course, and comparable with both their own form under the previous 4-2-3-1 and the current form of their most immediate peers.
The personnel, however, do not all yet fit. A 3-4-2-1 or a 3-4-3 or a 3-5-2 or a 3-2-2-2-1 or whatever it is Arsenal roll out most nights relies upon a central defensive unit that are both mobile and tall, while confident enough to play the ball. It is a big ask, certainly, to find a new-era Frank de Boer or two. If players like this were easy to find, John Stones would not have gone for £47.5 million. But having that style of player, along with a Pep Guardiola or Sergio Busquets-esque central midfielders in front of them, is the key to drive the forward via this set-up.
Be it a back three with wing backs or a back four with full backs, any defence needs cover from the midfielders in front of it. In Arsenal’s early season struggles defensively, this lack of cover has been costly. Granit Xhaka, who could be the key to the success of the formation going forward, has all the tools – rugged defensively, willing and able to fire away from deep, and capable of splendidly incisive passes, still continues to let all that good work down with sloppy errors and being too easily hounded off the ball when in possession. In theory, he and Aaron Ramsey, with lashings of Santi Cazorla, Jack Wilshere, Elneny. and the very occasional turn for Francis Coquelin are a solid combination of pieces to fit within the system. In theory, Monreal and Mustafi work to the left and right respectively of Koscielny, who could then sit further back, gambling when necessary, knowing the protection is there, letting the other two take it in turns to push on. But through individual struggles and injuries, plus less well-fitting squad depth, this theory has yet to come to pass.
Good Options Going Forward
On the plus side, the starting wing backs are right for the role. Kolasinac has been excellent, a tank going down the left flank who also holds his own defensively, while despite being shakier defensively and still liable to cross into the side netting, Bellerin’s speed and runs open up the right side of the pitch more. Doing so allows Aaron Ramsey to do what he does best – play the ball around the box, run on, tackle back where needed – and despite their usual inefficiencies with chance, Arsenal are showing enough going forward, at least now that the trio of Alexandre Lacazette, Mesut Ozil, and Alexis Sanchez are finally being played together,
As a unit, Arsenal have been getting enough goals in the 3-5-2. Their 19 goals in the 10 Premier League games so far is not a patch on what Manchester City are doing, but it will do, if they can keep it clean at the back. Arsenal still capitulate in a way unbefitting of the country’s best teams. They throw away one goal leads as often as they overturn them. They make sloppy errors in the centre of the pitch, and they have a knack for conceding ridiculous goals. These things hold them back. But the 3-4-3 doesn’t.
It is not quite there, yet. While the squad has better depth than ever before in recent memory, there is still a big gap between the choice starting 11 and everyone behind them, and as a squad with somewhere between the third and sixth most talent in the country, the margins are too small to survive multiple injuries. That said, the tactical evolution that had to come eventually is here now. And while it is hard to find strong causative links between the implementation of the new system and any discernible Arsenal improvements, it has not made them any noticeably worse. The long game continues.