clock menu more-arrow no yes mobile

Filed under:

D.C. United’s move to a 352, explained

A look at why Ben Olsen changed it up, and whether it’s a one-off or something more permanent

Xavier Dussaq / D.C. United / Courtesy of Major League Soccer

In pursuit of better results and a better performance, Ben Olsen opted to move D.C. United out of their normal formation against FC Cincinnati. The 4231 United has largely called on in recent years, most recently at a disappointing MLS is Back Tournament, was scrapped. The flexible 433 United started the year in? Nope, not that either. Olsen prepared for United’s first non-tournament regular season game in months by going with a back three for the first time since last year.

Why? Well, there’s a lot to talk about, so let’s dig in:

What are we talking about, exactly?

Olsen has played with a back three before, but this was not the 3421 from last year. United had two distinct shapes against Cincinnati: a 3412 in attack, and a 343 without the ball. Let’s start with the former, which you can see here on the right (or, if you’re on mobile, just below this paragraph):

Julian Gressel and Kevin Paredes pushed up higher than the central midfield pairing on a regular basis, but what’s really interesting here is how the front three were more structured than the trio of Wayne Rooney, Luciano Acosta, and Paul Arriola/Ulises Segura were in last year’s 3421. In 2019, United’s attack were given a lot of freedom to interchange; on Friday, the jobs were mostly quite defined, particularly on the ball. Segura played right of center, Ola Kamara played left of center, and Yamil Asad took up a starting point as the link between midfield and the forwards.

We have to say “nominally,” though, because most of Asad’s actions in the game came out to the right. This was down to another United tactical choice: overload the right in order to get Gressel more time on the ball (because defenders had too much on their plate to double up on him) and more options nearby (because there were more white shirts within 15 yards of him).

Asad wasn’t really involved often enough for this to work perfectly, but he did nearly score once — on a pass from Gressel — and in the first half, two-thirds of his actions with and without the ball came distinctly right of center. The overload had an even more extreme impact on Segura’s involvement, as this chart of all of his first half actions shows:

This also gave us an idea of how Kamara might thrive in this set-up. If United overloads their right, they’ll necessarily end up dragging more opposing players to that side. That means more room and fewer potential markers for Kamara in that left-center forward role. With Kamara being a threat predominantly because of his off-the-ball movement, it’s a good idea to help him out in this manner. Again, it didn’t really come off, as United were a bit too predictable once they got into the final third, but that’s a problem that can be solved with a little training, a couple more games, and above all better execution.

What about when they didn’t have the ball?

United went into a 343 without the ball, with Asad moving up a bit and the forwards dropping back a few yards to form a solid line of three:

The arrows here note something United did depending on Cincinnati’s movement: if the high press was really on, the front three moved up as a line, and Russell Canouse stepped higher to provide a layer underneath. This makes it more difficult for the opponent to bypass your first layer and find a central player, which can be a real problem. If you’re not forcing a turnover or a negative pass with your pressure, you at least want the opponent thumping a long ball or playing out to someone with few options on the touchline.

United did a good job of this for most of the game, and a particularly good job in the first half. That structure is what allowed this to happen:

United effectively forced Cincinnati’s attacks to be restricted to one man chasing a hopeful long ball down, and as a result the home team spent virtually no time in the Black-and-Red’s defensive third in the first half. It was a well-conceived, well-executed gameplan from the perspective of what you do without the ball.

“If you look at the first half, I thought that our structure caused them some real issues, and we turned them over in some very good spots,” said Olsen, who was correct. United surprised Cincinnati with the change, of course, but they also put on a good show of how this system is supposed to work against a team that tries to build from the back.

When the press wasn’t on, or in the second half when Cincinnati overloaded their central midfield by getting a forward or wingback (or sometimes both) into the middle to change the passing angles, United dropped this same 343 structure deeper, getting compact but never being forced to turn their 343 into a 541. That’s often a problem for teams that aren’t familiar with how to play out of this set-up: the wingbacks have a lot of tough decisions to make about how to balance their roles at both ends of the field, and eventually you end up so deep that you can’t break out effectively.

To a certain extent, the intention to make sure this remained a back three is clear simply from the players selected. Paredes, a versatile attacking midfield player, is not who you send out as a wingback if you’re looking to defend first, especially with Joseph Mora available. On the right, United didn’t have a ton of options due to the injuries to Oniel Fisher and Chris Odoi-Atsem, but Gressel is clearly an attack-first player in that role. United wanted to create more danger, and put more attack-minded players on the field in pursuit of that.

Why make the change?

Let’s start with why any coach makes a formation change. Usually it’s one of three reasons:

  • It’s a better way to get your best players into roles where they can be at their best
  • It’s a better way to trouble your opponent
  • You just need to change something because what you were doing before wasn’t really working

For United, it’s all of the above to some extent. This formation put Gressel in the role he was best at with Atlanta United, and it also emphasized pressure (something Asad has been less able to do in earlier appearances than he was here). It also hid a major weakness United would have had trying to play 433 or 4231 right now: they’re short on right backs. With Fisher and Odoi-Atsem injured, and Felipe Martins suspended, Olsen’s options were to play a formation that didn’t have a right back; play Mohammed Abu as his ball-winner and move Canouse to right back; or play Gressel as a right back.

This formation may or may not be the best way to maximize what this group can do in a vacuum, but when you factor in the realities Olsen was looking at (no true right backs, the next-best option being needed in midfield, and a star attacker only available to play 30 minutes), it’s hard to argue with.

The third reason above was also, pretty obviously, a factor. United had issues putting the 433 into practice outside of the preseason finale against Philadelphia, and they were unconvincing playing the old standard 4231 down in Orlando. Coaches and players said it wasn’t good enough, fans agreed, and we saw a real change. Simple, right?

That second bullet point is the interesting one, though, because this switch gets into a specific tactical need that United doesn’t see much in MLS. Under Jaap Stam, FC Cincinnati tried an open 433 — a formation we see all the time in this league — and got clobbered. Stam took a look at the tape, decided this was not a problem that would be solved short-term, and changed to a low-block 532. The soccer wasn’t pretty, but the results improved.

There was some mystery about whether he’d carry that into regular MLS play after having a couple of weeks to work on a new gameplan, and Olsen had to gamble on unknowns: Was Cincy going to play with one central forward, or two? Were they going to try to control the game, or were they going to sit deep and play for counters?

Olsen made his bet on the conservative 532 by switching to a back three, and from a defensive perspective, he got it right. Playing a back four against a 532 means your fullbacks have plenty of room to push up, but it also means that if you get caught on the counter, you’re suddenly 2v2, which is extremely risky.

Moving to a 3412 gave United an extra man in a central position to help defuse those long, direct passes into space for Cincinnati’s two forwards to chase down. That was how the home team played for the entire 90 minutes: it was all about looking for one killer ball that caught United as they stepped up, or as they were disorganized.

(As an aside, United did give up one big chance this way, with Jurgen Locadia hitting the post after being found all alone just as DCU tried to step high, but that came after D.C. responded to Cincinnati’s aforementioned midfield overload by moving into a 4231.)

Not only did this switch make sense in order to maintain the extra man against a front two, but it also gave United more players to press with, and to attack with. Olsen used that on a forward and asked his wingbacks to make sure they were more “wing” than “back.” As such, D.C. often verged on playing a 3232, flooding Cincinnati’s defensive third with numbers.

On the pressing side, United used their extra man really well. While Cincinnati had 56% of the possession, they struggled terribly at advancing the ball into United’s half, and ended up playing a ton of safe passes between their three center backs. That’s good!

But the attacking side was not so good. United had numbers forward, but they simply couldn’t get into positions to work any passing triangles or diamonds fast enough. This stagnant look has been there for a while now, arguably showing up in moments against Colorado in the opener, and then becoming more apparent in the win over Inter Miami. At MLS is Back, it was nearly ever-present, and in Cincinnati it came up again on a regular basis.

Incidentally, this is another big reason Paredes has had the chance to break through. While United’s more experienced players have attacked at a very methodical pace for unclear reasons, Paredes does everything with urgency. He covers ground with and without the ball by really bursting forward, putting opposing players under immediate pressure in one way or another. United has needed that contrast to the slower approach the rest of the attack seems to be taking right now, and hopefully the rest of the group will start to sharpen up and follow the youngster’s lead.

How does this work if everyone’s available?

Most likely, the major change is Edison Flores starting ahead of Segura as a forward. Segura could also replace Paredes at left wingback, while Mora will also certainly be in contention over there as well. We could also see Flores take over for Asad between the midfield and forward pairing, but in this system Asad’s ability to press and win the ball might be worth keeping in support of the forward duo. Flores has played as a second forward in Mexico, and he’s certainly got the goal scoring credentials to end up being Kamara’s forward partner rather than the guy we’re expecting to create first and score second.

Honestly though, that’s about it as far as big changes go. Gressel is a Best XI level wingback, and we’re still some ways away from Paul Arriola being ready to suit up. Donovan Pines, as by far United’s fastest defender, is a must-start in a system that necessarily gives the center backs more ground to cover than they normally have. Canouse and Junior Moreno, who incidentally looked far more like his normal self than he has all season, will have Felipe to compete with for two central midfield roles, but that’s no surprise. Kamara, provided he stays fit, will remain the other forward, and will probably see overloads forming to one side so that he can get loose on the other.

Are we going to see this formation again?

Yes, and possibly even on Tuesday. I say possibly because the New England Revolution are a team that are somewhat between playing a 4231 and a 442. Defensively, they are 100% a 4231 team. With Carles Gil out, it’s Gustavo Bou playing the no. 10 role, and unless he gets hurt between now and kickoff, he’ll be in that space underneath Adam Buksa when United is in possession on Tuesday.

However, going forward, the Revs give Bou (a player they think of as a “nine-and-a-half”) a ton of freedom, and he is ultimately more of a second forward than anything else. As such, New England’s attacking look often sees Bou cheat forward after the ball turns over, and Bruce Arena wants his team getting the ball to his attackers in a hurry. Speaking to media yesterday, Olsen pointed out that the Revolution push their wingers up high as well, and they can take on a look that is closer to a 424 than anything else.

It’s a complicated situation for United, both because the Revs are sort of a two-forward team and sort of not, and because they have different players available for tomorrow’s game. With Felipe back from suspension, the option of playing Canouse at right back (I know, none of us like it, but he’s a perfectly viable right back) exists. If Olsen’s read on the Revs is that they’re more 4231 than 442, we may well see a back four for the Black-and-Red. If he thinks of Bou more like a forward, though, the 3412 or some other version of a back three may end up being preferred.

The issue is further muddied by the fact that United arguably played their best 25 minutes of 2020 at the start of this Cincinnati game, and there’s less tape of United playing that way than there is of them playing the 4231 or 433. There’s also just the fact that the Revs see those formations on a regular basis; teams are in general going to be more comfortable against formations they play 25-30 times a year than they are against the different looks that make up the balance.

Even if the Revs aren’t a team that Olsen views as vulnerable against the 3412, he was very clear that he is going to use it during the course of this season. Philadelphia, who have shown a strong allegiance to a 442 diamond and are on the schedule for this coming weekend, are a good example. The New York Red Bulls, who United will play next week and then again the week after that, have played a 4222 every game this year.

In other words, there’s a concrete reason to expect United to play a back three again a few times this year, and now that they brought in Axel Sjöberg, they have enough center backs to play this way without one injury completely ruining their plans. Don’t be surprised if we see Olsen toggling between the two defensive set-ups throughout the rest of this season.