Game Context
DC United came into this game with its playoff hopes fading after a crushing defeat in Philadelphia. With four games left, this was one of two home games and so theoretically a huge opportunity. Except not only is Columbus one of the best teams in the league, Audi Field has been anything but a fortress. In fact, DC came into this game with one more road point than home point (17 vs. 16).
For Columbus, they were already in the playoffs. Their goal now is to try to catch Inter Miami for the Supporters’ Shield and to that end they really needed 3 points at Audi Field. I was sure they would rest starters after playing 120 minutes against Club América in the Campeones cup the previous Wednesday, plus they have an arguably even bigger game coming up on this Wednesday against Inter Miami. But, alas, they started a full strength lineup, so they must really care about the Supporter’s Shield (and playoff seeding).
According to the MLS power rankings, this was the #1 team (Columbus) against #21 (DC United).
Formation
The season is almost over and we’ve all seen what the players can do, so of course Troy Lesesne rolled out what I believe is a new arrangement of players. In his defense David Schnegg still isn’t available, so he was stuck trying to make DC’s typical 5-2-3 work. Alex Bono was in goal, of course, with Matti Pelota, Lucas Bartlett, and Chistopher McVey as the three central defenders. Aaron Herrera played at right wingback and, in a new wrinkle, Jared Stroud played at left wingback. Mateusz Klich and Martín Rodríguez played the double pivot in central midfield, while Ted Ku-Dipietro and Dominique Badji flanked Christian Benteke in attack.
As has frequently been the case, the Klich/Rodríguez midfield was pretty much run over. They were unable to reliably progress the ball, and while they were occasionally capable of forcing mistakes out of Columbus in possession, just as often Columbus cut through them.
Down a goal and pressing, at about 78 minutes I think the formation changed. Pedro Santos subbed on for Christopher McVey, and if I’m not mistaken at that point DC played with Bartlett and Peltola as the two central defenders in a back four with Santos and Herrera as fullbacks. Possibly it was a 4-3-3 shape, with Dájome and Klich playing aggressively forward ahead of Boris Enow as the lone defensive midfielder. This did the job of increasing variance, generating more chances for DC and also giving Columbus a couple chances too.
Goal Breakdowns
30’ 1-0 Christian Benteke
The Crew have DC pinned back in their own box and were passing it side to side when Max Arfsten misplays the ball. Herrera pounces, tapping it between two onrushing Columbus players to Ku-Dipietro. Ku-Dipietro touches the ball past Malte Amundsen, then outruns him to the point he gets between Amundsen in the ball and he’s off to the races. There’s only one Crew player left between Ku-Dipietro and the goalkeeper, central defender Rudy Camacho, and Benteke is rushing down the far side of the field. I was under the impression defenders are supposed to take away the pass in this situation and leave the goalkeeper to deal with the player with the ball, but I’ve also seen them try to collapse on the player with the ball and hope someone else can catch up in time if there’s a pass. For some reason, Camacho does neither of these things and instead sort of floated between Ku-Dipietro and Benteke. Ku-Dipietro has the presence of mind—no small feat after carrying the ball almost the entire field—to pick his head up, see the situation, and pass the ball past the Crew defender to Benteke, who hits it in stride. The placement isn’t great, almost right at goalkeeper Patrick Schulte, but Benteke hits it with enough power that it glances off Schulte’s leg and then into the goal.
54’ 1-1 Cucho Hernandez
The Crew are passing it around DC’s box. As the ball reaches the right side, Badji incorrectly (I think?) rushes into a double team that leaves defender Steven Moreira wide open. Moreira uses this time to look around and pick out an entry pass that finds Cucho Hernandez a few yards outside the box. Cucho traps the ball but doesn’t really seem to have his body positioned to shoot. You’ve probably seen this play at this point, but if you haven’t, it’s hard to express how non-threatening this looks. Then Cucho just…kicks it in? This is a thing you can do if you’re Cucho, apparently. It’s a hard enough shot that Bono doesn’t have time to take a step back, but it’s lofted enough that it goes over him and drops under the bar. It’s an extremely tiny window, but he hits it perfectly. It’s the sort of play you expect from the best player on the arguably the best team in the league. I do think Badji could have done slightly better on defense earlier in the play, but really the failure that led to this goal was Cucho’s representation not insisting on a transfer after having a great season last year the way I expected them to do. He’s too good for MLS.
57’ 2-1 Jacen Russell-Rowe
The Crew once again have the ball near DC’s box, but elect to play it back toward midfield, I think intending to pull DC players outward. A defender sends the ball wide to Cucho near the right sideline, closer to midfield than the enplane. Cucho has time to look up and see that Jacen Russell-Rowe has gotten himself back onside and is making a good run between McVey and Bartlett. Cucho hits a deep cross that flies just over Bartlett and directly on to Russell-Rowe's diving header. It wasn’t a gimme—Russell-Rowe did very well to hit it into the upper corner—but only an inch-perfect ball from Cucho would have reached him. Alas, the ball was perfect. Maybe McVey could have been a little more alert to this, or maybe Bono could have come out to catch or punch it away? But honestly, this seems like another "what can you do, Cucho should have transferred to a big 4 league by now, sheesh” situation.
83’ 2-2 Christian Benteke
Boris Enow gets the ball in DC’s half and is lightly pressured, so he passes it wide to Herrera. Herrera pushes aggressively down the sideline, drawing in defenders, before squaring it back to Enow in the middle of the field. Enow quickly plays an entry pass to Jacob Murrell standing about ten yards out of the box with his back to goal. Murrell flicks it inside, possibly trying to get it to Benteke, but if so it's behind him. Gabriel Pirani cuts in, touches the ball once, then plays a very short pass that catches a Crew defender flat-footed and puts Benteke in on goal. Benteke’s finish is clinical. It’s an outstanding team goal. Benteke gets the glory but Murrell, Pirani, and Enow, all substitutes, did most of the work to make it possible (plus Aaron Herrera, but we expect this from him).
Expected Goals
DC United won the expected goals battle 2.53 to 1.53 on the back of Christian Benteke’s 1.52 xG. .88 of Benteke’s xG was earned on the two goals. On the other side, Cucho’s goal was 0.05 xG and Russell-Rowe’s was 0.12, so it was actually their shots that didn’t go in that brought them most of their xG. It was a rare game in which DC “won” both halves in terms of xG, 0.96-0.51 in the first and 1.57-1.02 in the second.
Player Ratings
Alex Bono - 5 - One very dangerous giveaway on a mid-range pass, but otherwise made some good saves. I don’t think he could do much on the goals.
Lucas Bartlett - 6 - He had a crucial intervention near the end of the game to save what seemed like a bad situation.
Christopher McVey - 5 - Mostly fine, though he started a Columbus breakaway with a bad giveaway while pressed up high. Maybe could have done more to contest their second goal.
Matti Peltola - 8 - I've been critical of him at times this season but he was immense in this game. Fotmob confirms the eye test, noting he won 6 out of 8 ground duels, which is both more duels and a better winning percentage than the other starters. Oh, and he was 37 for 41 passing. Centerbacks get a lot of safe passes, but this was more and better passing than the other CBs.
Aaron Herrera - 7 - He seemed a little off to start the game, but he grew into it. He always plays with fire even when he’s tired at the end of a game, and this game was no different. His aggression helped set up the second goal.
Jared Stroud - 4 - He didn’t look good on the left. No one looks good on DC United’s left, I’m realizing. More on that in the next section. He did have one really nice left-footed cross to Benteke in the second half.
Mateusz Klich - 5 - Baseline Klich game: great set piece delivery, annoying (to the other team) gamesmanship, some important forward drive to possession, but also part of a midfield that got run over.
Martín Rodríguez - 4 - He shuttles the ball well enough but he’s no Darlington Nagbe, that’s for sure. He seems to defend decently when he’s in the right place, but he doesn’t cover enough ground to get into enough duels, let alone win them, hence we always lose the midfield battle.
Ted Ku-Dipietro - 6 - He spent the game mostly doing “bad KDP” things: losing the ball instantly when it came to him, mostly, with a dash of shooting directly at the keeper. But his assist! That was the kind of play that has won him some notoriety among the national media: skill and hustle coming together to make a dominating play against the starting defenders for one of the league’s best teams.
Dominque Badji - 4 - He’s okay as a second forward, but in this formation I’m starting to feel he provides Pirani-level defense and Stroud-level offense, which isn’t a great combination. He could have been subbed earlier for my tastes. His defense got more limited in the second half.
Christian Benteke - 8 - Well, he doesn't create chances out of thin air like Cucho, but he's a dominant forward and once again he showed why, netting 2 goals on about 4 or 5 solid chances.
Substitutes
Gabriel Pirani - 7 - The commentators complained he was playing high instead of sitting deep in midfield helping the team possess the ball, and it’s true he doesn’t get many touches playing high. But the goal showed why Lesesne wants him near the goal. He’s not a string-pulling long passer, but he’s very good with short passing, and in this game his short passing earned him an assist.
Andrew Murrell - 5 - Great flick to Pirani in the buildup to the second goal, but on a different play he missed the fact Benteke was absolutely wide open.
Boris Enow - 6 - Maybe his best game? Not saying much, obviously. He still is kind of rough, but he made some important tackles and it was his entry pass to Murrell that started the dangerous part of the buildup to the second goal. Hopefully he’ll be even better next season after a full preseason.
Cristian Dájome - 5 - Didn’t make an impact in the short time he was on the field.
Pedro Santos - 3 - In theory he should be a good offensive option when we are trailing, but his main impact seemed to be getting caught ball watching and allowing his man to be played in on goal. Luckily Bartlett saved him from being at fault for a game-losing goal.
DC United’s Right-Wing Extremism
Something that’s been apparent throughout the season is that DC United attacks much more frequently down the right side of the field.
I say it’s been apparent, but as a stats nerd my first question is always: can we match up this impression with statistics? Unfortunately, it’s hard to get good statistics on this from the public sites. My first idea was to use touches per 90. Surely I’d see more touches from the right-sided players (in the Columbus game, Ku-Dipietro, Herrera, and Klich) than the left-sided players (Badji, Stroud, and Rodríguez). So I go to FBref and scaled the touch numbers up to 90 minutes for players who subbed out:
Right side: 19 (KDP), 60 (Herrera), 52 (Klich)
Left side: 22 (Badji), 46 (Stroud), 65 (Rodríguez)
Two things to note here. First is that boy, our attacking players don’t see much of the ball. Benteke had 42 touches, at least. But for the Crew, Cucho had 55 touches and Diego Rossi had 54. But the Crew had way more possession and so had 699 touches overall to DC’s 526.
Anyway, I was expecting more of a disparity: Klich to have more touches than Rodríguez and KDP to have more touches than Badji. Maybe it was just this game? But looking at the stats for the season, vaguely right-sided players don’t seem too different from mainly left-sided players in touches per 90 minutes. Herrera has 63.1 to Pedro Santos’ 61.9, for example. Until this past game Jared Stroud has played mostly on the right, but he’s at 41.4 touches, which (this really surprised me) is actually less than Gabriel Pirani’s 41.6. Ku-Dipietro and Badji seem like very different players (and Badji has only played just over two games worth of minutes) but are at 33.1 and 32.7 touches respectively.
Okay, so if the touches don’t show it clearly, maybe this is all just made up? Well, not so fast. WhoScored has heatmaps for teams overall in a game and the individual players, and DC looks clearly right-sided against Columbus (as the home team, DC is shown attacking from left to right):
That’s just one game though. What about against Philadelphia (on the road, so attacking from right to left)?
Right-sided again. What about an actual win, like DC’s 2-1 win against Chicago (going right to left):
Massively right-sided. One last one, the 3-1 win in Toronto (right to left again):
Okay, okay, so we’re not crazy, it’s really happening. So…why?
So why is this happening? I think there’s two possible reasons:
Troy Lesesne and the rest of the coaching staff explicitly tell the players to use the right side as their primary option.
If the players on the right are better, they’ll have more success, which will translate into more memorable plays in our minds and “heat” in the heatmaps (but apparently not touches).
These are definitely not mutually exclusive: maybe the players on the right are better and as a result the coaching staff tells the players to prefer that side. But they could also be orthogonal. Left back is a big issue for DC United, but they aren’t alone. It’s probably the weakest position across MLS.
We don’t have access to coaching instructions, but I think there’s one unambiguous indication of what they want. When Alex Bono hits goal kicks long, where does he hit them? Unfortunately I don’t think there’s a stats site that can prove this (the MLS site used to have a passing arrow “chalkboard” showing origin and endpoint but doesn’t seem to this season) but just relying on my memory from Audi Field, I think they always go to the right side.
Assuming I have that correct, I think that’s indicative of intent. Sure, on goal kicks all ten field players squeeze into a little patch of the field, but by biasing it to the right it makes it more likely DC’s right-sided players end up on the ball.
But what about the other theory? Are the players on the right side better? In two cases, surely the answer is yes: Aaron Herrera was an all-star, and Mateusz Klich is a DP midfielder who leads the team in shot-creating actions. When it comes to the other players, guys like Stroud, Dájome, Ku-Dipietro, and so on, it’s a lot harder to say, though. Stroud has a ton of assists, but is that because the ball is always on his side or is the ball always on his side because Stroud is so good?
The reason I started down this path, however, is that Jared Stroud played at left wingback against Columbus. I haven’t gone back and checked every single game but I think it’s his first significant time on the left this season. In an extremely rare moment of informed DC commentary on MLS’s official podcast Extra Time this week, Andrew Wiebe even attributed Stroud’s move to the left as Lesesne trying to get more production out of the left side of the attack. So…did it work?
The first heat map was from the Columbus game, so DC still was right-sided. We can also look at his stats: 40 touches in 77 minutes, which is fine. 80% passing accuracy, which is much higher than is usual for him, perhaps because he was playing farther back than normal. 0.4 expected assists on a single key pass, his excellent cross to Benteke. That was also his only shot-creating action. All this matches what I saw, which is that he had one great pass but otherwise didn’t make much of an impact.
Well, I don’t think any of us really thought Stroud was the MVP of the right side. What about Herrera and Klich? DC has had to play without them a few times. What happened then?
The most recent game like that was DC’s 1-1 draw with NYCFC on September 15th. Klich was suspended, so Boris Enow replaced him in midfield, playing on the left while Rodríguez shaded to the right. How did the overall team heatmap look? This was at home, so going left to right:
This favors the right side, especially near the box, but to me this is noticeably more balanced the heatmaps I showed earlier. This is despite Stroud and Herrera playing on the right, although I think Herrera had more defensive responsibilities in this game. Dájome was in front of him on the right, and NYCFC actually favored attacking on their left (DC’s right) instead of going down the Santos/Pirani side.
Now let’s go back to DC’s 2-1 win Chicago on September 8th. I actually included the heatmap above as an example of right-sidedness. It’s quite pronounced. But Herrera wasn’t on the field for this game! The right-sided players were Stroud, Klich, Dájome, and McVey.
So let’s jump to August 25th, DC’s 3-4 loss to FC Dallas. Herrera came in as a halftime sub in this game, but it’s most notable because David Schnegg was playing at left back. DC was at home, so the map is left to right:
A bit of right-side bias, but I’d say more balanced overall.
Now let’s go back to July 18th, a 3-2 win over Minnesota United. Stroud, Herrera, and Klich all played…but Klich played on the left side of midfield with Rodríguez on the right. What happened?
Just a bit more on the right, but pretty balanced.
Anyway, my conclusion is not a shocking one. At least when we have the ball, Klich is by far our best midfielder, so the attack leans toward his side. Aaron Herrera produces a smaller but still visible effect. The coaching staff directs goal kicks to that side but also influences this by choosing to play Klich there so that our best attackers can combine with each other. David Schnegg was the front office’s answer to creating balance in the force attack and it probably would have worked and DC would be over the playoff line if not for his injury. For next season, I think this points to the need to get better midfield play as a key priority, plus getting a backup left back. But more on that after the season ends.
Standings
DC United ends up in 12th place, tied on points with 11th place Atlanta United and three points behind Toronto and Philadelphia in 8th and 9th, respectively. 6th place is now out of reach, and worse, so is finishing ahead of the Red Bulls seeing as they’re currently in 6th. Charlotte is in 7th place with 42 points. Technically DC can get to 43 points if they win all three of their remaining games. I wouldn’t be surprised if they are actually mathematically eliminated from 7th already, but I’m not going to bother checking.
In the standings that are perhaps more relevant to this DC United season’s place in history, Benteke is now three goals ahead of the pack in the race for the Golden Boot with three games to play. It’s definitely not in the bag yet, but he’s getting close.
Coming up
The whole season has come down to three games. First, a Wednesday game against Nashville, 13th and one point behind DC in the standings. Nashville had a putrid stretch through the middle of the season before showing signs of life by beating Atlanta, beating Chicago, and then drawing Cincinnati. However, they just lost 1-0 to 14th place New England, so they haven’t put things totally back together.
Next, DC takes on the aforementioned New England Revolution in Boston. New England has been very bad and have to play on the road in Houston on Wednesday. They could already be mathematically eliminated from the playoffs by their game against DC, and even if they’re not, their last two games are against Columbus and Inter Miami. They’re done. That should mean they have nothing to play for and will crumble…but you never know in this league.
Finally, DC plays its “decision day” final game at Audi Field against Charlotte, who are clinging to 7th having just gotten a draw against Inter Miami. Let’s hope the game still means something for DC’s playoffs hopes when it happens.
I believe the main reason for using Stroud there was to rest Santos for the midweek game. Then get Schnegg there by the NE game.
Appreciate your articles and thoughts Matt! Keep up the great work. I'm not sure if you meant it this way but get tired of this player is to good for MLS narrative. Cucho is a top player that has played in the Premier League but that doesn't mean MLS is beneath him and he never acts that way. MLS needs more players like him if we all want the level to raise! He is getting paid, doing great and appears to be enjoying his time in MLS!