Desparate Times? Revisiting the Three-Man Defense

Back in May, when D.C. United were going into a game at Toronto FC after having seen their 7-game unbeaten streak snapped in a spectacular 5-3 loss to the San Jose Earthquakes, there was some discussion in this space about returning to the formation United used in its most recent trophy-winning years, the 3-5-2.

Some of us - well, I - called the idea crazy. Why would we expose our lack of healthy natural center backs (at the time, the lineup would have been Russell, McDonald, Woolard) by changing to a formation that really requires three of them, instead of two, to be successful? Why should we move our notoriously aggressive and attacking wingers into wingback positions, with more defensive responsibility? Mostly, it seemed like a loss on the road to the league's best team (then and now) was a pretty small reason to be trying to retool our entire system - but now, especially after Jamaica deservedly beat the U.S. using a three-man back line, maybe it's worth thinking about again.

Especially since United has been a bit more experimental with their formations of late. Not too very long ago, we were a team that was wedded to the 4-1-3-2, but now it appears that the Black-and-Red are a flat 4-4-2 in defense, with the outside midfielders taking extremely high positions in attack. Except that sometimes we shift into a 4-3-3 when we're holding the ball. And really, the midfield has been fluid all season, with the D.C. formation ranging from 4-1-3-2, to 4-3-1-2, to 4-2-3-1, to 4-4-1-1 to 2-3-5 (okay, maybe not, but it would be awesome if we did).

So, coming off a stretch where United dropped from the Supporters' Shield race to (for the moment) outside of the playoff positions, let's talk it out. Since we have a break for World Cup qualifiers (and since I don't really want to think on that at the moment) let's step squarely outside the box and figure out whether the 3-5-2 is even worth talking about at any point during the rest of this 2012 season.

Basics of the Formation

Just to set some things out from the beginning, I'm not talking about the same 3-5-2 that United actually ran in 2007 and through the U.S. Open Cup final loss to the Seattle Sounders in 2009 (a game that I have referred to as the death of the 3-5-2 in MLS). That formation involved Fred and Chris Pontius playing in nominally defensive wide roles, which, as Chest Rockwell put it, is "balling out of control." It was beautiful when it worked, but truly, it was too beautiful to live (especially with Tommy Soehn coaching it).

The 3-5-2 as it needs to be played to be successful in a mature league doesn't just need three center backs to anchor it, it needs players that can serve as competent defenders in the wingback spots (so, essentially "not Fred"). Yes, the wingbacks employed with a 3-man back line are responsible for generating width in the attack, but they're also critical in defending width that the opponent might provide and to provide outside cover for the defenders. Beyond that, the only thing that's really necessary is at least one dedicated central defensive midfielder. Under Peter Nowak, United typically played a 3-4-1-2 variety, with two D-mids sitting in behind Christian Gomez in the midfield. Of course, as Kenny Daglish proved with Liverpool a couple years ago, you can do a lot of interesting things with the other positions and actually have some (at least isolated) success in the process.

Applying the 3-5-2 to United

Basically, the 3-5-2 is really useful if you have a solid corps of center backs and a good spine and if you're also blessed with speed, work rate and defensive ability on the outside. Earlier this season, I would have taken issue with the last of those, but today, I'm less sure.

One of my biggest criticisms of the 3-5-2 suggestions earlier this year was that it simply didn't fit our personnel. We were short, not long, on center backs, and the formation would force our outside attackers into too much of a defensive posture, neutralizing their attacking prowess, but things have changed since early May. Since then, Dejan Jakovic, Emiliano Dudar and Ethan White have all regained fitness, joining Brandon McDonald in the platoon. At that point earlier in the season, our starting outside midfielders were Chris Pontius and Andy Najar. Arguably, neither of them should be playing that position today. Najar has been called our right back of the future by none other than Ben Olsen, and Pontius has shown a lot of promise playing at forward. As far as Andy goes, he plays the fullback position like a wingback anyway, so moving him explicitly to wingback might actually give him more freedom to move forward. (I'm pretty sure Patrick Nyarko and Gonzalo Segares each just woke up sweating from horrible flashbacks/nightmares.)

With United now having four natural center backs (plus Woolard, who surprised at the position, and Russell, who can fill in a pinch) and with Nick DeLeon and Chris Korb both having the speed, work rate and defensive ability to play the left wingback position, and Andy Najar already playing wingback as a fullback, my biggest worries about this system is at least alleviated.

Flavors of 3-5-2

If you bothered to click on those Liverpool links to the indispensable Zonal Marking, you saw that Peter Nowak's 3-4-1-2 is not the only flavor of the 3-5-2 formation that's out there. King Kenny managed to get wins againstperennial tough out Stoke City and eventual European champions Chelsea using, respectively, 3-4-2-1 and 3-1-3-1-1 (call it the 3-5-2 diamond if you want) formations. If these twists on the 3-5-2 can get results - even on a one-off basis - in the Premier League, why shouldn't we be able to find a version of the system that could generate wins in MLS?

The Classic: 3-4-1-2

This is probably the most common version of the 3-5-2, due to its defensive stability and its ability to control possession. This all comes from the two defensive midfielders screening the defense, which clogs up the central zone on the field and forces opponents to break you down on the outside.

The issue is that this can easily turn into what's known as a broken formation, with very little connection between defense and attack. If you're not careful, you can end up with seven guys playing defense and clearing the ball up to three guys playing offense. This probably means you're not playing possession football, and instead you're playing on the counter, for which this formation can actually provide a pretty solid base.

The image below is how I imagine United would probably line up if they used this variation. The biggest drawback is having to drop two of Dwayne De Rosario, Branko Boskovic, Hamdi Salihi and Chris Pontius, which I don't love.

A Twist: King Kenny's Diamond 3-5-1-1

You want to talk about trying to hoard possession, how about putting four men in the central zone. But do that with a hold up man and dedicated wide players. That's what Liverpool did when they beat Chelsea in February 2011. It's a really interesting formation, but it's one that in no way fits United's strengths (No Boskovic, no Pontius, no Salihi; DeLeon out of position), so I'm just going to dismiss it out of hand. And it's my column, so I can do that.

Midfield Madness: 3-4-2-1

This variety also gets four guys in the central midfield zone, but does it in more of a Sampsonian 3-6-1 than with something that could really be called a 3-5-2. This should be a safe formation defensively, as you have two D-mids screening the defense, plus another two more attack-minded midfielders patrolling the central zone. Again, though, you have the potential to have a broken formation, especially with the lone striker potentially ending up on an island.

From a D.C. United perspective, though, a big, big problem with this formation is that is leaves practically no room for Chris Pontius, which basically makes it dead on arrival - Partyboy has been one of our two best players this year, so a system that omits him isn't really worth considering. Nonetheless, here's a potential lineup:

Shifting Forward: 3-3-2-2

This is a variation I haven't actually seen anywhere, but it seems to me to fit United's personnel quite well. The deeper part of the midfield is not unlike the 3-3-1-3 that Marcelo Bielsa used when he managed Chile to surprising heights in the last World Cup, with only one defensive midfielder located centrally to screen the defense. This formation should prove more stable in possession, though, as two more advanced midfielders also occupy the center of the field. Put a classic two-striker combination up top, and there you have it.

Looking at United's roster, our two most clinical strikers (considering DeRo as a midfielder, which he'd probably need to be in this formation) - Pontius and Salihi - are at their most effective working off of a target player in the mold of Maicon Santos and Lionard Pajoy. This formation takes that into account and gets our best attacking players on the field in positions where they can be effective. Check it:

Since I Mentioned It: 3-3-1-3

It's no secret that I'm a big fan of the tactics of Marcelo "Loco" Bielsa, who is now heading up Athletic Bilbao in Spain's La Liga. His base formation (which was shifted often depending on the opposition) when he managed Chile's World Cup run was a 3-3-1-3. The big virtue of this formation, next to others featuring 3-man defenses, is the width. In other formations, all the width - all of it - has to come from the wingbacks. In the 3-3-1-3, though, you've got wide forwards in front of those wingbacks, so you've got more tools to put opposing outside defenders on their heels. The downside: you have to get those outside forwards from somewhere, and to get that width, you sacrifice numbers in the center.

This could be an interesting look for DCU, especially given our outside attacking talent. Honestly, I'd take Najar and DeLeon or Korb and Pontius against almost any fullback in the league on any day. The problem, though, is in the center of the park. When United run the 4-4-2 diamond, we've usually got an outside midfielder or two pinching in to help make numbers and keep possession. In the 3-3-1-3, there's nobody to pinch in; the D-mid and attacking mid are all on their own, and if they can't effectively keep and move the ball, then the system is going to be very, very reliant on the counter.

That doesn't mean I didn't mock up a lineup sheet, though.

Circling Back

So, there are a number of formations United could conceivably trot out utilizing a three-man back line. Some of them might even work. But are any of them enough to take us back to a tactic that hasn't worked in MLS since the Sounders invented the league in 2009? It would mean ditching the only base formation we've played all year and going to a thoroughly different system.

I, for one, am not yet desperate enough to really push for this just yet.

But it would be interesting in future years to see if the 3-5-2 (or some other 3-x-x) formation can make a comeback. Juventus used it at times to go unbeaten in Serie A last season, and Italy experimented with it at Euro 2012 this summer; Manchester City even toyed with it at the start of their 2012-13 campaign.

One of Bielsa's core tenets when devising formations is to have an "extra man" in defense - if the other team is running two strikers at you, have a 3-man back line; if they're running out a lone striker or a 3-man attack, then use 4 in the back. The idea is to have one, and only one, extra defender - that way, you have the numbers, but you don't have anybody standing around doing nothing, taking up a slot that could be better used elsewhere on the field. It would be a lot of fun - if inevitably infuriating - to see that principle put into practice in MLS. It would take an incredibly smart, disciplined team to do it, and I'm not convinced United is that team.

Moreover, I'm still not convinced we have the right personnel, particularly further down the depth chart, to make this work. This entire post - all 2200+ words of it - was inspired principally by Andy Najar's shift to fullback. He was out in United's last game, the 1-0 loss at Real Salt Lake. If we were running one of these systems, I suppose we could shift Korb to the right and drop the defensively adequate and hard-working DeLeon to left wingback, but that leaves us with no other top-level cover at those positions. (There is enough of a difference between the attacking and running responsibilities of wingback and fullback that I don't think Robbie Russell and Daniel Woolard will really be able to make the transition, even when healthy.)

So, those of you who have made it this far, what's your take? Would it be worth United's exploring a three-man back line, and which flavor of the system do you think best suits the Black-and-Red?