On one hand, D.C. United is facing a predictable opponent on Sunday. The Colorado Rapids are still a primarily cautious, conservative team that plays at a slow tempo primarily for defensive purposes. They'll play out of a 4231, a couple players will be at least slightly out of position, and they'll be tough to score headers on. There has been talk that the Rapids are changed this year, but on these points they haven't taken a new approach.
On the other hand, Pablo Mastroeni's "everyone gets to play!" approach is allied with a far more abstract definition of who can and can't play a given position. Are you a center back? Well, you can probably play right back, they're near each other. Or maybe you're a deep-lying playmaker...well, let's try to push you way up the field. Or out to the wing. You have to pass in all of those spots, right? Strikers are wingers, left is right, up is down. Even when the Rapids make a move towards improving themselves as an attacking threat - mostly via pushing their fullbacks up more this season - they dilute the impact by playing the most predictable player in MLS (that'd be former United left-footer Marc Burch) at one of those positions and giving the other to a player making his debut while also playing out of position (in this case, Mekeil Williams).
Still, I don't want to be too hard on Mastroeni, who had to swap out both of his week 1 fullbacks due to injury (I'm not sure what's wrong with Eric Miller, but apparently Marlon Hairston's hamstring was tweaked ahead of LA's visit to Dick's Sporting Goods Park last week). It's just that the odd outcome - left back Williams playing his first MLS game on the opposite flank, and Burch starting an MLS game in 2016 - is so Rapids.
So while the players that start may be hard to predict, the 4231 formation is not:
Zac MacMath will start in goal, and you have to feel for the guy. The Philadelphia Union could have just developed him, but United bamboozled them on draft day into handing over some allocation money to get a player they could have had at #2 (Andre Blake). United got Steve Birnbaum with the next pick, while the Union signed roughly half of the world's goalkeepers, then loaned MacMath to the Rapids for the season. Colorado barely acknowledged MacMath all season (in part because they had Clint Irwin), and then seemed disinterested in exercising the buy option on him, but in the end they opted to sign him and trade Irwin.
All's well that ends well, right? Sure, MacMath could have ended up at a better club (there are 18 or 19 in MLS), but at least he's the man somewhere, right? Someone finally showed some faith in him! Well, not quite. In any case, MacMath is an all-around adequate MLS goalkeeper. He's a bit small for a GK in this league, though, and he's not great on crosses. The Rapids cover for that with some big options at center back, so that flaw doesn't often show itself.
Speaking of those big dudes, Mastroeni appears to have chosen Jared Watts - who last year was more of a defensive midfielder - as a partner for second-year player Axel Sjoberg. That means no spot for Bobby Burling, whose 22 starts last year count as a lot on a team that endlessly rotates players for reasons that are usually not clear. Sjoberg, at 6'7" and over 200 lbs, is a giant. Hitting crosses in his vicinity is a waste of time. However, like many huge defenders, he's not the fastest and has little lateral quickness. You beat the big Swede on the ground. Watts seems to be playing in a constant rage this season, so United should see if they can pick up some free kicks from the Wake Forest product.
Assuming Hairston is back in contention, Mastroeni will have to consider changing his fullbacks again. Williams seemed comfortable enough at right back last week, though the Galaxy's "ugh, do we really have to play this game?" attitude certainly helped him out. If Hairston can go, it would surprise me if he wasn't back in at right back, which would mean a move to left back for Williams. If Burch starts, United should focus their attack on him all game long. He's simply lacking in the speed, agility, and quickness of thought to play as a fullback in this league. In fact, even if Hairston (and Miller, for that matter) return, United is going to want to focus on getting the ball onto the wing and into space.
It appears that Mastroeni has settled on Micheal Azira (yes, "Micheal") and Sam Cronin in the engine room. It's an odd fit, since Cronin is probably better receiving the ball from the back for than Azira, while the Ugandan is a ball hawk capable of covering tons of ground. That's not how they've been used in either game this season, though. Cronin functions as more of a midfield metronome ahead of Azira, who has reined in his tendency to charge around since he's supposed to be the anchor here.
Ahead of them will be Dillon Powers, who would be best playing the #8 role Cronin has. He was even told he'd get that spot after a trial at Reading in England didn't work out due to a coaching change over there. This being the Rapids, he naturally is back playing a position he's not really suited to. That doesn't mean Powers is to be ignored, though. He's still the craftiest player Colorado will start, and he's the best passer on the team bar none. United has to shut him down, or he'll be a major factor in this game.
The wings are far less clear right now in part due to the (rather harsh) suspension of Shkelzen Gashi, Colorado's big DP signing in the offseason. Gashi had started both games so far opposite Dominique Badji, with frequent position swapping a feature of the Colorado attack. I suspect that Badji will play on the right, since Taylor Kemp has had issues with speed in the past. Badji will be the fastest player either team can call on for this game, but beyond that he's pretty one-dimensional. Still, if you're only going to have one way to beat a team, speed is a good one.
On the left - they'll still switch a few times throughout, but this group will end up on the left more often - I think Luis Solignac has a slight edge. The Argentine was signed as a forward, but Mastroeni began using him on the wing after only a handful of starts up top. Like most strikers playing on the wing, his instinct is to drift inside and go to goal rather than spread things out and set up others.
At home, Mastroeni would consider Dillon Serna more strongly, but on the road I think Conor Doyle is more likely to get a run. United fans know why: Doyle is smart and committed defensively, and he's very willing to sacrifice for the team. Serna's lack of time in Colorado has baffled people for years, though, and if he does play United will want to remember what happened the last time he appeared at RFK Stadium.
What about Marco Pappa, Colorado's other big acquisition this winter? Pappa was involved in some kind of unresolved domestic incident that ended with him suffering a stab wound that kept him off the field for nearly the entire preseason. Colorado has gotten him on the field in two straight games, but only for an 11 minute cameo in San Jose and 23 better minutes against LA (during which he scored the game-winner in the 95th minute). I suspect that Pappa is only fit to play 35-45 minutes, tops, and MLS teams largely prefer to use players like that off the bench rather than to start them and then make a halftime sub.
As such, Pappa isn't on the graphic, but he is a major threat for United. The Guatemala international is a long-range shooting threat, a creative passer in and around the box, and generally skews towards making audacious plays for his team's attack. On the other hand, if he enters a game that's tied, United should absolutely look to attack down his flank. Pappa is good at a lot of things, but he's never been very aware on the defensive side of the ball.
Up front, Kevin Doyle will lead the line. Last season, the prototypical #9 had to play as an attacking midfielder numerous times because Mastroeni does things like that. Doyle's game is what you'd expect from a Premier League target man. He's good at playing simply, he's physically strong and enjoys that side of the game, and he's going to look to knock the ball down before quickly running in behind the defense. He's not the fastest player on the planet, but his runs off the ball are sharp enough that a brief letdown by United's center backs could be very costly.
Off the bench, I've probably mentioned the players that would come in. Pappa seems a lock to come in at the hour mark (or earlier if Colorado is losing), while one of the non-starting wingers from Solignac, Doyle, and Serna will probably get in as well. If the Rapids need to make a change in the middle, former Philadelphia homegrown player Zach Pfeffer is an option. Mastroeni could also bring in Burling and move Watts into defensive midfield if need be.