Three at the back has been effective for Manchester United, but it can’t be the only option

What a difference two weeks makes. A fortnight ago Manchester United were worried. The club sat two points above the relegation zone, hadn’t won away from Old Trafford in over 200 days, and were preparing to face Liverpool followed by four straight away games.

Heading into that match, Ole Gunnar Solskjaer had been tactically stubborn, refusing to change from his 4-2-3-1 formation despite injuries leaving his squad ill equipped to play it. That prompted our own Nathan Heintschel to write this excellent piece about why United should switch to a back three.

It came down to needing something different to spark a good run in the team, and given the players that were available, it was probably for the best. Solskjaer had experimented with three at the back towards the end of last season, but given the players, and the players unavailable due to injury, it almost always ended in disaster. This season is a different story.

Solskjaer obliged, but, like everything else for him this season, it nearly blew up in his face before it even had a chance to get off the ground. Axel Tuanzebe was injured in the warmup and Solskjaer was forced to replace him with Marcos Rojo. This was hardly the game you’d like Rojo to be playing in.

Luckily, everything else worked. The switch to wing backs allowed Aaron Wan-Bissaka and Ashley Young to push higher up, pressing Liverpool’s fullbacks higher up the field and away from the danger area. The team were able to dictate play and should have escaped as 1-0 winners if not for a brain fart from — who else — Marcos Rojo.

Solskjaer’s 4-2-3-1 was designed around the buildup play coming from central midfield, or more specifically, Paul Pogba. While the results don’t suggest it, it was working. Playing from a deeper position, Pogba led the team in expected assists (xA). He’s now fifth, albeit in 200+ fewer minutes, and still leads the team in xA/90 (0.25), Key Passes/90 (2.60), and xGBuildup/90 (0.36). [James Garner has an xGBuildup of 0.65, but he’s only played five minutes in the league.] Pogba’s numbers aren’t higher simply because the finishing and final ball in front of him has been poor.

The concept behind this was simple. Solskjaer moved Pogba into a deeper position because he needed to fix a defense that was awful last year. That wasn’t just because of the defenders, but because under José Mourinho the team sat back and invited pressure and under both managers the midfield offered the back four zero protection.

United were going to defend higher up the pitch. The front four/five were going to press like maniacs. They’d trap the opposition, and while they weren’t going to steal the ball, they were going to force the other team to give it away. Who would be waiting in midfield to recover it? Pogba. This is why Pogba led the team in recoveries in every game he played. It’s also why Fred leads the team in recoveries in every game he plays. It’s the design of the system.

The key in the clip below is that as soon as Pogba recovers the ball, he turns possession into attack right away, playing the quick pass that gets the team into an attacking position.

Without Pogba, that fell apart. Whether it was Nemanja Matić, Scott McTominay, or Fred, United’s midfield became very passive, always playing the safe ball rather than playing it forward. That culminated in the Newcastle match where United’s attack was anemic, mostly thanks to the midfield spending 90 minutes making lateral passes.

Without Pogba, United don’t have the creativity to build up from the midfield. That’s where the back three comes into play. Harry Maguire and Tuanzebe are both very good with the ball at their feet. Victor Lindelöf isn’t bad either. By switching to a back three and pushing the wing backs high up the field, the responsibility of playing line breaking passes went to Maguire, Young, and Wan-Bissaka. They bypassed the midfield.

The idea worked against Liverpool, and Solskjaer ran it back four days later against Partizan in the Europa League. Once again it worked, sort of. United came away with their first away win since March and kept a clean sheet, but going forward they only managed one shot on target and that came from a penalty. They conceded a few more chances than normal, which is expected with a heavily rotated side, but like always the chances they conceded weren’t of great quality. At the end of the day, they controlled most of the match until the final 10 minutes when Solskjaer called off the press.

Heading into the match against Norwich all signs were pointing towards Solskjaer rolling with this three at the back formation. It was almost a surprise when he reverted to the old 4-2-3-1. The result was the best attacking performance United had all year. They took 21 shots, putting 11 of them on target. That’s the most they’ve had in a game without Paul Pogba in the four years since they signed him.

When United travelled to Stamford Bridge in the League Cup this week, it was three at the back again. Again United controlled play. They pressed high up the field and Chelsea couldn’t mount any attack in the first half. But again the team mustered only two shots on target, one from a penalty, and one from Marcus Rashford’s sensational free kick.

Chelsea took control for a bit in the second half and equalized. Needing a second goal, Solskjaer switched to the 4-2-3-1. Immediately United returned to being the dominant side and were much more attacking. They just couldn’t put their shots on target.

The two and a half games playing with three at the back have been very good for United because they’ve shown they can do it. They don’t lose a step defensively, which is important for a team that’s built on defense (they lead the Premier League in xGA - which is crucial if you’re struggling to score goals). It also gives them tactical flexibility. The fact that they can line up for a match in multiple ways makes it that much harder to prepare for them.

Unfortunately, it still doesn’t solve their offensive woes. That’s not to say it can’t do it eventually. Brandon Williams has been a revelation on the left side, but his end product in the final third isn’t there yet. Same with Wan-Bissaka on the right.

This is where I need to say: that’s totally OK! Williams is 19 and Wan-Bissaka is 21. They will develop that end product that will make them that much more dangerous going forward. Don’t worry. It’s not going to happen overnight, but it’ll happen.

United’s offensive blow up with the 4-2-3-1 last Sunday came for a few reasons. The fact that Norwich are the second worst defense in the league (last in xGA second from bottom in goals conceded but only because Southampton conceded nine last week) is certainly a factor.

But the most important reason was the return of Anthony Martial. When United play with a back three, Marcus Rashford and Daniel James tuck more inside, a position that works for Rashford but not so much for James. With Martial in, it allows Rashford to play as a second forward on the left, and James to play more as a right winger. The best positions for both.

With United having two wide players, Andreas Pereira can move inside and play as a number 10. This is by far his most comfortable position, and his two best games this year have been against Liverpool and Norwich when he played centrally.

But we should also be careful not to mistake Pereira’s “best game” and “best position” for him being actually good at the position. If you look around the other number 10s in the league, Pereira is pretty underwhelming.

Maybe that’ll change with time, and I hope it does, but even against Norwich his numbers weren’t that great. Against Norwich, United had a non-penalty xG of 2.99. Playing in a central creative role, Pereira’s xA was 0.17, fourth highest among United players. More indicting was his xGChain (0.05) — fourth lowest — and his xGBuildup (0.02) the lowest on the team. United took 21 shots despite Pereira at no. 10, not because of him.

For the 4-2-3-1 to work, this has to be the minimum Pereira gives you. If he’s any worse, United are going to keep struggling.

The choice of formation ultimately comes down to the question of who United have available. At the moment, barring one or two changes, United can really only play Solskjaer’s preferred formation when they have their first choice squad available. The loss of Martial caused too many pieces to move around and too many players to come in that aren’t creative enough to make the 4-2-3-1 work.

Results aside, the best thing to come out of the past few weeks has been Fred and McTominay proving that they can work as a midfield pair, and not be a detriment defensively. Thus, if/when Paul Pogba returns, Solskjaer can keep that Fred/McTominay base and play Pogba either as a number 10, or as the furthest forward in a three man midfield. Until then, United can only play the 4-2-3-1 when they have Pereira, James, Rashford, and Martial.

The back three has emerged as a good change of pace for United. It’s an excellent formation to go to in cup games when the squad is going to be rotated. Unfortunately, United just aren’t able to generate enough of an attack from it. And until that changes, the back three will be a good formation, but it can’t be the formation.