DC United Season Preview Part 3: Defense

Goal: Give Up Fewer Goals

Here at All-Points Bulletin we’re launching this season’s coverage with a mammoth four part season preview. Part one looked at the team overall, focusing on roster construction decisions, and part two went deep, really deep, into the team’s goalkeeping.

Today we’re going to talk about the team’s defense…or lack thereof. If the defense was so bad, what constitutes a “good” defense? How can we distinguish between good and bad defensive players? Do the team’s moves at least make sense on paper?

Next time we’ll wrap things up with a look at prospects for DC’s attack. Make sure to subscribe to catch that as well as other preseason articles. Nothing is guaranteed in life, but I’m expecting at least one other article before DC’s season begins on February 22.

How many goals against is too many?

Before we dig into the roster details, I want to start by considering what success would look like for DC United’s defense in 2025. One thing we can say for sure: giving up 70 goals in a 34 game season is not a success. That’s unfortunate, because that’s exactly what happened in 2024. DC’s offense, strapped to Benteke’s back as he marched to the Golden Boot, wasn’t incredible, but it was good enough to get the team into the playoffs. It was the defense that left them on the outside looking in.

This raises the question. 70 goals is definitely too many, but what is the correct number to give up?

Zero, obviously. Are they stupid? Why let the other team score at all?

That’s a joke, of course, but it’s worth noting that an all-out bunker that conceded 0 goals and filled the season with 0-0 ties would result in 34 points. That’s six fewer than DC earned in 2024, well outside the playoff line.

Okay, so what’s a better target? Philadelphia Union used to be the yardstick I used for DC, but they finished three points lower and fired their longtime coach last year, so what about…ugh…Red Bull New York? I know, I know, but bear with me. Despite a history of combined villainy and incompetence going back to the days of Cheatin’ Bob, their 2024 team played with a similar game model to DC, finished 7th in the East, and ended up making a Cinderella run to MLS Cup.

They scored 55 goals during the 2024 regular season. DC scored 52. Not a huge difference. What about on the defensive side? Well, DC gave up 70, as we mentioned. The Red Bulls gave up…50. So despite a similar offensive output, the Red Bulls has a goal differential of +5 while DC had -18.

So is 50 goals allowed the target? Despite this enormous gulf in goal difference, the Red Bulls only ended up with seven more points than DC. If you rank by goals allowed, it depends a bit on how you rank ties, but basically their defense was about 15th out of 29. Pretty mediocre! Well, they did finish 7th in the East and 15th in the Supporter’s Shield standings. Literally middle of the road. Maybe we should be setting the bar higher? Or, wait, this is goals allowed…setting it lower? You know what I mean.

This turns out to be one of those times where just stack-ranking teams can be deceptive. The Red Bulls might be 15th having given up 50 goals, but FC Cincinnati ranks 6th with 48 goals allowed. There’s a lot of bunching. So let’s just look at the big picture and apply some groupings instead:

Using the groupings in that picture, I propose that DC United should aspire to upgrade itself out of “bad”, skip over “below average”, and take up residence in “decent”. When you see it laid out like this, it doesn’t feel like setting our sights too low. It means giving up 20 fewer goals this season and joining what is effectively most of the top half of defenses.

Sports pundits like to say “defense wins championships” and it might be true in other sports (probably not, but let’s not get sidetracked), but in 2024 Inter Miami and LA Galaxy won the two main trophies with “decent” defenses (49 and 50 goals allowed respectively).

What about the Seattle Sounders and Charlotte FC? They had great defenses, the best in MLS, but they only ended up with the #4 and #5 seeds. Their defensive prowess even continued in the playoffs against the league’s best teams. Seattle gave up just three goals in four games while Charlotte gave up three goals in three games. Awesome! Except Seattle lost in the conference finals and Charlotte didn’t even make it out of the first round.

The fact is that in soccer, defense and offense aren’t nearly as distinct as they are in American football, baseball, or even basketball. Most teams can, if they exert themselves, give up fewer goals by playing more conservatively. The problem is they’ll score fewer goals, too. And whether it’s the regular season’s three points per game won or the fact that draws are settled by coin-flip penalty shootouts in the playoffs, soccer is intentionally designed to reward offense. This is in turn why the vast majority of MLS Designated Players are attackers.

For his part, Troy Lesesne often makes comments that indicate he prefers offense over defense. Ally Mackay and even Dave Kasper before him made comments to the same effect. As a fan, I’m happy to support this. Yes, we want to win, but all things being equal, it’s more fun to win 4-2 than 2-1.

So it’s settled. We won’t ask for greatness. DC United just needs a decent defense. And all they have to do to get there is go from 70 goals allowed to 50.

So…now what?

Two Explanations

When I started working on the research for this post, I already had two main explanations in mind for DC’s poor defense in 2024. These aren’t any kind of brilliant insight; they’ll be familiar to fans who follow the team closely:

1. The contracts Ally Mackay inherited, plus some bad luck, meant the team played the entire 2024 with only one real fullback, Aaron Herrera

2. DC’s midfielders, in particular its central midfielders, were either unwilling or unable to cover enough ground when defending, leaving the back line exposed

I kind of wish I could promise that these theories were totally wrong and I’m about to tell you the secret truth. Alas, after a lot of research, I think these explanations continue to make sense. I did add on a third explanation, however!

Since it turns out these “explanations” align with position groups, we’ll examine each one in tandem with a look at how 2024 players performed and the changes the front office made for 2025.

But there’s one more thing to discuss before we start. I mentioned doing a lot of research. You might wonder…what research is there to do? Soccer is a complex game of movement, position, and skill. How can we evaluate DC United’s defensive problems beyond just watching the games and coming up with hot takes?

I guess I could have refreshed my memory, firing up MLS Season Pass to relive the terrible 0-5 loss to Orlando or the humiliating 0-4 collapse against DC’s nemesis Philadelphia just when the playoffs seemed in sight. Then at least my hot takes would be theoretically better informed.

Well, I’m the guy who wouldn’t even write game reviews in 2024 if DC didn’t at least tie, so no, I definitely didn’t do that.

Advanced Statistics to the Rescue?

Regular readers know I lean pretty hard on advanced stats whenever I can. Ideally these would shed some light on DC United’s specific issues in 2024 and maybe even the specific players that were the problem. Unfortunately, the publicly available statistics for defense are not very useful.

For example, you can look at a team’s expected goals allowed and see how that diverges from actual goals allowed. There’s perhaps something we can learn from this, and I’ll get into it later when we talk about the central midfielders, but it’s pretty limited.

Then there are defensive counting stats, like tackles, interceptions, and so on. I won’t go into detail here, but suffice to say, when you compare them across MLS teams, these don’t correlate with what we’d expect. DC United tends to be above average in defensive counting stats because they play with a high tempo and its offense turns the ball over a lot, so the other team gets more possessions in a game than average. A stat like “percent of tackles won” isn’t a counting stat and so theoretically corrects for this, but it still doesn’t correlate. Personally, I think these stats are too subjective and just too limited in what they capture. A failed tackle isn’t ideal, but it’s still a better defensive outcome than just letting the opposing player run through your midfield unopposed.

That isn’t to say there aren’t useful advanced stats. At least at the squad level, we know there are stats like “[opposing] passes per defensive action” that evaluates pressure applied to opponents. MLS-employed analysts sometimes cite stats like that, but as far as I know the data is proprietary to sports analytics companies like Opta and not publicly available through sites like FBref.

If You Want a Job Done Right…

I really wanted something that was at least vaguely objective to evaluate 2024 defenders, so I resorted to creating my own statistic: goal blame. Here’s how I did it. I went back through all 70 goals scored on DC in 2024 and for each one, I watched carefully and decided who I thought was at fault. If you read my game reviews in 2024, I was doing a similar exercise in prose where I tried to assign blame for goals conceded. So I did this again, but for all the goals conceded this time, not just the games I covered here (draws and wins), and I put the data in a spreadsheet instead of just writing.

Some of you are coming up with lots more questions about this metric, but others are just saying “Fine, just shut up and show me the numbers”. Originally I went on here to explain things in detail for those of you in that first group, but the example images made an already long post too long to fit in people’s email. I’ve moved the detailed explanation to its own post that I just stealth-published, so if you’re interested, you can feast on the gory details as well as the many caveats and limitations.

That allows us to immediately move on. I’ll show you the overall table sorted by goals blamed per 90 minutes played, then we’ll go through players by positional unit, using my novel metric to shed light where it can.

Fullbacks

We’ll start with fullbacks. Here’s the roster, with player age on opening data in parentheses. I know DC might actually have played more games with wingbacks than fullbacks, but bear with me.

2024: Aaron Herrera (26), Mohanad Jeahze (26), Conner Antley (28), David Schnegg (25)

Out of position: Pedro Santos (35), Cristian Dajome (30), Matti Peltola (21), Jared Stroud (27)

2025: Aaron Herrera (27), David Schnegg (26), Derek Dodson (26), Conner Antley (29)

As 2024’s training camp opened, DC United had just two fullbacks on the roster. Aaron Herrera was a new arrival and coming off a bad-fit season in Montreal, but he had proven himself a capable MLS starter at right back for Real Salt Lake, most notably with a 10-assist season in 2021.

Mohanad Jeahze was theoretically a capable MLS starter. DC had paid an $800,000 transfer fee for him at the beginning of 2023 after all! Surely they wouldn’t have been so wrong! Then again, that money was paid by the Rooney-influenced version of DC’s front office that signed Ravel Morrison to a guaranteed contract. I don’t know. I remember Jeahze looking…plausible…in the four games’ worth of minutes he played in 2023 before a terrible ankle injury ended his season.

With no backups for either Herrera or Jeahze, Mackay signed USL veteran Conner Antley during training camp. Antley had no MLS minutes at age 28, but he’d played both centerback and fullback in USL, so he was at least a useful utility player.

The good news here was that Aaron Herrera played great from the start, going to the MLS All-Star game after three assists in his first ten games. Antley looked better than expected as well, earning an assist in his debut as a left back. He wasn’t going to ever join Herrera and Benteke on the All-Star roster, and in fact DC wouldn’t even want him to be starting 30 games, but he looked like a solid backup.

The bad news was that although Jeahze played a bit in preseason, some combination of injury and off-the-field issues meant that he would end up playing just a single minute for DC in 2024 before—as far as we can tell, since Troy Lesesne seems to believe in keeping locker room drama firmly under wraps—being kicked off the team. DC would go on to use its 2024 contract buyout to get his contract off the books. That might have promoted Antley to starter, but he tore his ACL after only 7 appearances.

That left the team having to play players out of position, limping until the midseason transfer window when DC paid a nearly $2 million transfer fee for Austrian left back David Schnegg. Schnegg wasn’t perfect, but looked promising in Leagues Cup and in one regular season appearance, then succumbed to a calf injury that kept him out until a sub appearance in the final game of the season.

This combination of a thin roster, poor choices from the old front office, and bad injury luck meant that all season Lesesne had to choose between playing with players improvising at left back, typically either Pedro Santos and Cristian Dájome, and occasionally both when Herrera wasn’t available. Or he had to use a three man backline that (as we’ll get to soon) exposed weaknesses in DC’s centerback corps and often still left either Santos or Dájome struggling nearly as much at wingback.

What’s the impact of playing attacking players out of position at fullback? In qualitative terms, at least we can say that Santos and Dájome didn’t sulk. They threw themselves into their assigned roles and played hard. But effort only gets you so far. Santos in particular had lost several steps since his prime and most MLS attackers were able to easily outrun him to the endline. Dájome was quicker, but not left-footed and usually less effective going forward. And both of them made critical mistakes in marking and positioning, the kinds of mistakes that enraged fans but which you have to expect from someone who finds themselves playing major minutes at a position for the first time as a professional.

In my Goals Blamed statistic, Santos was third on the team in absolute terms with 5.6 goals blamed on him. His 0.28 goals blamed per ninety minutes was lower than a number of defenders, but he benefitted from having played some of his 2024 minutes much higher up the field (I haven’t broken out goals blamed by position played, it’s just too time consuming to do that since positions are adjusted during games and I don’t trust the statistical services to notate it correctly).

I was surprised that Dájome only ended up with 2.6 goals blamed, 0.12 goals blamed per 90 minutes, but he also played higher at times.

Matti Peltola often plays left back for Finland’s national team, but Troy Lesesne only rarely used him there. I’m not sure how trustworthy this is, but FBref has him starting at left back from DC United only once in 2024, in a 2-4 loss to Montreal where I blamed him for 4/5 of a goal (a misplayed pass to Birnbaum for their first goal) and he got ejected after a second yellow card at the end of the first half. So…I guess Lesesne was right not to play him there.

We’ll wait until the final part of this preview series to discuss the offense in depth, but it’s worth mentioning here that Lesesne wants wingbacks and even fullbacks to help advance the ball, get forward, and swing in crosses. We know what it’s supposed to look like because Aaron Herrera did this when he wasn’t stuck playing out of position himself as a centerback. In modern soccer, sometimes one fullback stays at home and the other is given license to go forward, but in his short time on the field David Schnegg was crashing forward just like Herrera.

Even though you’d think Dajome and Santos would actually have superior offensive skills to a career fullback, in practice they never looked even half as effective as Herrera, and they both struggled (Santos especially) to get forward enough to contribute to the attack but still get back to help out the defense.

So what’s the plan for 2025?

In his first press conference, Troy Lesesne said he expected to vary the formation between a four man backline and three centerbacks as the situation demands, but in another answer he referred to a four man backline as the default, so it’s probably safe to say he wants to play with two true fullbacks.

The good news here is that, while you never know since the team is secretive in the off-season, David Schnegg seems to be healthy. He posted an instagram video of himself working out energetically as training camp began and played a half in DC’s first preseason game. If DC United can get 25+ games with Schnegg and Herrera both playing at their natural fullback positions, right away I think it’s reasonable to expect a big step forward in 2025. Yes, David Schnegg’s goals blamed per 90 in 2024 makes him look pretty bad, but that was just 120 minutes. I’m willing to give him a pass and remain hopeful for 2025.

But injuries happen, not to mention crazy Jeahze-like situations. What if that happens again? Herrera missed the first preseason game with what we can only hope was a very minor injury, for example.

Over the off-season, DC signed Derek Dodson, a career USL right back technically coming from Minnesota United but with no actual MLS minutes. Conner Antley has also been seen in team videos of training camp. Lesesne said he’d be at full fitness in a month, a timeline that would make him ready for opening day.

Two clear starters and two USL-level backups makes fullback a solid position group by DC United’s standards. I’m not actually sure Antley is going to be considered a backup fullback; the team might prefer him as a backup centerback. We’ll see. If so, new central midfielder Hosei Kijima is supposed to be able to fill in at right back as well. And then Matti Peltola remains an option, though I think the team wants him to focus on central midfield.

Central Midfielders

2024: Mateusz Klich (33), Matti Peltola (21), Russell Canouse (28), Jackson Hopkins (19), Jeremy Garay (20), Boris Enow (23),

2025: Matti Peltola (22), Boris Enow (23), Hosei Kijima (22), Jackson Hopkins (20)

Although Troy Lesesne varied between four man backlines (with two fullbacks) and three centerback lines (assisted by two wingbacks), a consistent feature was two central midfielders occupying the middle of the field in front of the defensive line. Orientations varied a little bit, but in general they played in a horizontal line and seemed to have equivalent responsibilities, even though often one player (e.g. Klich) was more “offensive” while another (e.g. Peltola) was more “defensive”.

We already talked about him a bit in the first part of this series, but I think we have to start with Mateusz Klich.

Splitting up defense and offense into separate preview articles is a bit of a problem for this discussion. Since I’m about to complain a lot about him, I’ll say that Klich provided important qualities going forward. There’s a reason he played nearly every minute he was available.

And yet there’s also reason why Mackay shipped him to Atlanta. To do it justice, Lesesne’s system needs all of its midfielders to run hard and cover ground against the ball. It’s not a coincidence that the Supporter’s Shield-winning New York Red Bulls in 2018, running a similar albeit even more frenetic system than Lesesne’s, had an aging Bradley Wright Phillips in front of a bunch of young players. At the time, the joke was that except for BWP, players would get shipped out before they hit 30, like an MLS Logan’s Run. And in the engine room was an 18-year-old Tyler Adams, already showing the world-class athleticism that allows him to hold down the center of the field by himself.

There’s no way to know for sure, but I suspect as 2024 training camp opened, Lesesne and Mackay were imagining Matti Peltola and Russell Canouse in midfield with Klich playing a role higher up the field. Peltola and Canouse would be youth and experience, and in the previous year Canouse had improved a lot on his linking play.

But it wasn’t meant to be. Canouse ended up out with essentially season-ending surgery. Klich played basically the entire season in central midfield, and while he played an important offensive role, he just didn’t have the legs to cover a lot of ground or the defensive chops to effectively tackle when he did get to opposing players in time. I don’t totally trust “tackle” stats, but Klich’s tackle percentage was 47.5% on the season, putting him down among the attacking players, whereas players like Peltola, Herrera, and Bartlett were all in the 60s.

But Klich was never alone in central midfield. Unfortunately, no one else ended up looking all that good either.

I thought Matti Peltola looked very bright in his first few games, but his performance seemed to fall off while still in the early part of the season. He ended up benched at times in the middle of the season for Jackson Hopkins, then had to deputize at centerback. He was fairly ball-secure at the price of playing very conservatively. Lesesne praised his passing at one point, but I didn’t see much evidence Peltola was able to really help the offense as the team struggled with ball progression.

He’s quite young, so the fact he showed some flashes of defensive energy and some good deep-lying distribution makes me tempted to end on an optimistic note. There’s a lot of reason to hope he’ll improve in 2025, especially if he can focus on central midfield. My concern is with his athleticism. He’s no Tyler Adams, that’s for sure. In fact, he looks…slow. Slow to accelerate and maybe even slow at his top speed. That’s a big problem for a ball-winner in any system and makes me worried his ceiling isn’t as high as we all hoped when he arrived.

Jackson Hopkins struck me as a very young, very raw version of Klich. On his good days, he offered a lot more going forward than Peltola. On his not-so-good days, of which there were many, he didn’t provide much offense and his defense was if anything worse than Klich’s. Like Peltola, Hopkins looked slow to me. That he’s still struggling with ankle issues that look to keep him out for the first part of 2025 isn’t encouraging either, but I’m trying to keep an open mind.

After my gripes about Peltola and Hopkins’ speed, Ally Mackay’s midseason addition Boris Enow seemed a bit more promising. He looked fast, aggressive, but positionally he was all over the place. He was thrown in immediately, but only ended up playing just over three games’ worth of minutes, so he never had time to really adjust, so the jury is still out. As with Peltola and Hopkins, he’s young enough there’s still a chance (maybe not very likely, admittedly) of him taking a major stride forward this season.

I’ve got to come clean here and admit that when I did the “goals blamed” exercise, I expected to see Klich and Hopkins at fault for a lot of goals. I was sure they’d look a lot worse than Peltola. Instead, their rate of goals blamed per 90 (0.12 and 0.11 respectively) was down near Lucas Bartlett’s. In absolute terms, Klich gave up 3.6 and Hopkins 1.2. Doesn’t seem so bad! The reason I expected more was because their mistakes were so memorable. When a centerback is a bit slow to mark someone, I marked them as blamed, but we know this happens. Everyone gets beat sometimes. Hopkins and Klich ended up getting blamed for fractions of goals where they weren’t running hard to get back, and that’s really frustrating to watch because it seems so preventable.

But the goals blamed metric is very forgiving of midfielders. When Klich or Hopkins (or Peltola, who was guilty of this a couple times too) fails to rush back, they got a sliver of blame, but often a central defender also had to make some kind of mistake for a goal to be scored. Peltola maybe got the worst of both worlds by racking up goal blame improvising in central defense while also earning some “jogging back” demerits in midfield.

The goals blamed metric also doesn’t account for all the opposing attacks that could have been snuffed out before they even got started through better midfield play. I’m going to still trust my eyes in lieu of stats that the midfielders were a drag on DC’s defensive performance (this also goes for more advanced players like Stroud, Ku-Dipietro, and Pirani, though their impact on the defense is even more diffuse).

I mentioned at the advanced stat “expected goals against” isn’t tremendously useful, but for what it’s worth, only San Jose had a bigger discrepeancy between goals allowed and expected goals allowed. DC United gave up 70 goals, but expected goals allowed was only 51.3. That’s an 18.7 goal difference! What do we make of this? The fact that San Jose, the worst team in the league by a mile, was also bad points to the fact that bad teams generally give up more goals than expected, so this could just be another way of saying DC was bad. Still, objectively worse teams like the Chicago Fire and the Revolution only had differences of 10.2 and 11.2 respectively.

Possible explanations:

1. Alex Bono was particularly bad and failed to stop shots that other goalkeepers would have stopped. See the first part of the preview series for more on this.

2. Expected goals are based on the location of the shot, not necessarily the position of the defenders. If I’m right that DC’s midfield was often overrun, attacking players had more time than usual to shoot from a given position.

3. Lesesne’s system expects the sort of maximum effort best provided by young legs, but often saw Klich (33), Santos (35), and Dajome (30) on the field together.

4. Lesesne’s system expects the sort of maximum effort best provided by motivated players, but since DC’s defense was bad for other reasons, sometimes after falling behind more goals would be given up by players who just wanted to get to the locker room. DC’s combination of mediocre points with terrible goal difference (-18 even though they ended up with more points than Philly who had a goal difference of +7) speaks to being on the wrong end of a lot of blowouts.

You can draw your own conclusions, but I think all these factors point towards the need for a younger and more athletic midfield to help the defense.

So is there hope for improvement in 2025?

The new player reinforcing this unit is Hosei Kijima, a 22 year old that DC traded $400,000 in GAM to St. Louis City to sign. That’s enough fake-money to wonder if he’s really just a depth piece but…he sure seems like a depth piece. He played a ton of college soccer before coming to MLS and then played a bit under 800 minutes for a bad St. Louis City team last season. Since he was rumored to be able to fill in at right back, I thought he might be more of a defense-first midfielder, but his St. Louis stats seem more like Jackson Hopkins than Matti Peltola. We’ll see. St. Louis fans liked him well enough and were a bit disappointed to see him go, so that’s something I guess.

Normally I ignore trialists until someone is actually signed, but with Hopkins’ injury looking like it might rule him out of a lot of games in 2025, that leaves only three players here for two spots and seems unacceptable. Apparently former FC Dallas and Toronto FC player Brandon Servania has been with DC United in training camp and played in central midfield during their first preseason game. He has 65 MLS starts for FC Dallas and Toronto, including 21 starts in 2023 for an extremely bad Toronto team. Then he tore an ACL and played almost no minutes in 2024. If he’s really back from his injury, he seems like he would be a useful player and provide much-needed depth, but he’s not a game changer.

The hope for improvement here rests in Matti Peltola and Boris Enow. They could take big steps forward this season. They’re both foreign players and it often takes foreign players a season to adjust to MLS for some reason. They’re also young enough to hope they’ll improve from experience. If they elevate their game and remain healthy enough to play most of the minutes, Ally Mackay will look like a genius. If they don’t, this could continue to be a big problem area.

Centerbacks

2024: Lucas Barlett (26), Christopher McVey (26), Garrison Tubbs (22), Matai Akinmboni (17), Steven Birnbaum (33), Hayden Sargis (21), Conner Antley (28)

Out of position: Aaron Herrera (26), Matti Peltola (21)

2025: Lucas Bartlett (27), Kye Rowles (26), Lukas MacNaughton (29), Garrison Tubbs (23), Conner Antley (29)

When I began the research for this article, I thought that DC’s centerbacks were decent in 2024. Certainly not amazing. Sure, I had doubts about McVey, but I didn’t think there was that much of a gap between him and Bartlett. Like other fans, I was vaguely disappointed Akinmboni was sold for “only” $1.5 million.

But after doing this research, I’ve changed my mind. Centerbacks were a major problem in 2024 and it’s good that steps have been taken to address it. The only question is if they were enough.

Let’s run through the 2024 players in clusters mostly based on goals blamed per 90:

• Matai Akinmboni (439 minutes, 0.9 goals blamed per 90)

• Garrison Tubbs (549 minutes, 0.56 goals blamed per 90)

• Steven Birnbaum (346 minutes, 0.47 goals blamed per 90)

• Hayden Sargis (30 minutes, no goals blamed!)

The sample sizes aren’t huge, but I think the goals blamed per 90 are bad enough here to say these players all played quite poorly. Lucas Bartlett, who I think it’s fair to call a good but not great centerback, had 0.12 goals blamed per 90. Akinmboni costing the team nearly 1 goal per game is…very, very bad.

Maybe you don’t trust my goofy stat, but the coaching staff seemed to agree. Part of the reason the minutes played for Akinmboni and Tubbs are so low is because Lesesne did everything he could to avoid playing these players, including playing Matti Peltola and Aaron Herrera out of position at centerback instead. Only Tubbs returns in 2025; the other three are off the roster.

As for why they didn’t play well, I actually think it’s a different story in each case.

Matai Akinmboni was really young and was still making mistakes with fundamentals like footing. He’s also not old enough to really bulk up, so despite his height, he struggled physically. For example, on a corner kick goal scored by FC Dallas, he got manhandled by a 5’10 left back (Marco Farfan). He also lost more aerial duels than he won (though he was in good company on that, as we’ll see). Akinmboni still has the potential to be a very good centerback, but he needs playing time to get there, and my view now is that DC United couldn’t afford to give it to him on the first team. He’s been playing a handful of games a year for Loudoun since he was 15; I’m not sure how well he did there but I doubt he has the physicality yet for USL either. Selling him now seems like a great move. I know he’s showing up on the bench for Bournemouth’s first team, but I expect him to end up in “Premier League 2” (basically a youth development league for EPL teams) after their injury crisis ends.

In contrast, Garrison Tubbs looks to me like he has the physical tools to succeed. He just needs to clean up mental mistakes. Most of the goals I blamed him for were situations where he made a positional mistake. In particular, there were several goals where he moved upfield to try to challenge a ball carrier only to get easily bypassed. This may never improve (Donovan Pines never seemed to progress in this area despite his physical gifts) but more time acclimating to the speed of MLS and Lesesne’s system might help. Apparently MacKay agrees. Tubbs’ contract wasn’t guaranteed in 2025, but Mackay exercised the club’s option to extend it a year. However, I expect that in 2025 the team hopes he’ll be the last resort off the bench.

Steven Birnbaum had vast experience in the position and remained very strong in the air, unlike most of the other centerbacks. But age and injuries had taken their toll. Longtime fans will remember how Bobby Boswell seemed to go from “strong veteran” to “completely useless out there” almost overnight, and unfortunately something similar happened with Birnbaum. I’m not sure he was ever really healthy in the past year, so moving on was the right thing for both him and the team. That the previous front office gave him so many guaranteed years on his contract extension is great for him but was obviously a bad idea at the time and forced Mackay to buy it out.

All I can say about Hayden Sargis is that despite his lack of goals blamed in his thirty minutes of regular season play, his appearances in club friendlies convinced me he’s never going to be an MLS-level player. Apparently Lesesne and Mackay agreed. It happens. I appreciate his enthusiasm celebrating DC goals while he was on the bench, however, and wish him good luck with his career.

• Christopher McVey (2,148 minutes, 0.33 goals blamed per 90)

• Matti Peltola (2,219 minutes, 0.28 goals blamed per 90)

The next group didn’t give up quite so many goals as the first, at least in rate terms, but because they played a ton of minutes, they actually gave up 14.8 goals between them in absolute terms (whereas Akinmboni and Tubbs gave up 8.2 together).

Christopher McVey, unlike every centerback mentioned so far, was always a starter when healthy. My sense from watching the games was they he was better going forward than Bartlett but worse at actually defending. The goals blamed stat seems to bear that out as far as defending, but when I turned to the advanced stats I couldn’t find much corroboration for him helping the offense. In terms of passing completion percentage, he and Bartlett are roughly the same at every distance. Bartlett outperforms him in shot-creating actions and also in progressive passes. McVey only won half his aerial duels whereas Bartlett won two thirds of them despite participating in about three times more.

I expected the club to retain McVey but try to find someone who could move him to the bench, but instead they parted ways. Although McVey didn’t take up much cap space, I believe he still uses an international slot, so that might have been the issue. He’s now with expansion side San Diego.

Matti Peltola is the other player in this group. Goals blamed is a little clumsy here; some of it was earned while he played in central midfield. He did the best he could filling in at centerback and I actually think he did a decent job positionally, but he plays small and was a huge liability when matched up against big center forwards. From Lesesne’s comments in preseason, it sounds like they want Peltola to focus on playing in central midfield this season, so in effect he joins McVey in exiting the centerback corps.

• Conner Antley (433 minutes, 0.21 goals blamed per 90)

• Aaron Herrera (2,595 minutes, 0.16 goals blamed per 90)

• Lucas Bartlett (2,741 minutes, 0.12 goals blamed per 90)

Conner Antley split his relatively few minutes between centerback and fullback. To be honest, I’m not sure what Mackay and Lesesne view as his natural position. For now I’m assuming he can play at both fullback and centerback. I didn’t check but I don’t think he ever played in a two centerback system, though, which is a bit more demanding then being on the outside of a three centerback line. We’ll see. His relatively low goals blamed is a positive sign, at least.

Aaron Herrera spent a lot of time at fullback and I think that helped his numbers; defensively he can be a bit of a liability at centerback, particularly when defending crosses. He looks like he has enough size to play centerback and I wouldn’t be shocked to see him again playing the position in a three man backline if injuries pile up. But he’s not just a better fullback, he’s a great player at fullback, so I expect to see him spending most of his time there in 2025.

Lucas Bartlett is the clear standout in goals blamed. 0.12 puts him down among the attacking players despite the fact he played essentially every minute as a central defender. He was second in the league only to Benteke in aerial duels won, and while DC’s penchant for quickly turning the ball over might have gotten him some extra attempts compared to other teams, he won two thirds of his duels (only a bit less than Benteke’s godlike 73%). Beyond his solid play on defense, his aerial abilities make him formidable on set pieces. He ended the season with five goal contributions (2 goals, 3 assists), more than all other centerbacks combined. And he did all this despite entering the season making the senior minimum of just under $90,000. Mackay rewarded him with a contract extension that presumably gave him a big, very deserved raise.

At this point I’ve gone through all the 2024 players. Bartlett seems like a locked-in starter now. Great. But McVey, Akinmboni, Birnbaum, and Sargis are gone while Peltola and Herrera are hopefully playing their natural positions. That sounds like addition by subtraction, but absent additions, that leaves Tubbs or Antley as the starter next to Bartlett. Fortunately, there are additions.

DC’s new starting (we hope!) central defender…

Seeing as DC United paid about $750,000 to transfer Kye Rowles from Hearts in the Scottish top flight, it’s a good bet that Mackay expects him to start alongside Bartlett. At 26, Rowles is in his prime and left-footed, making him a good partner for the right-footed Bartlett. My opinion (not shared by everyone, I admit) is that most of the teams in the SPL, including Hearts, are a bit below the MLS level. The two powerhouse teams, Celtic and Rangers, regularly play in the UEFA Champions League and are a significant step up from MLS. As a strong contributor for Hearts, there’s reason to hope that Rowles can step up to be a solid MLS starter. His experience playing against Celtic and Rangers attackers is good preparation for the DP forwards and wingers he’ll have to contend with here.

DC also acquired 29-year-old Lukas MacNaughton from Nashville SC in exchange for $150,000 in GAM. MacNaughton had 40 MLS starts across four years with Toronto and then Nashville, so he may not really be starter level, but he’s probably a better backup than Tubbs, at least the 2024 version of Tubbs. At his age, it’s not likely he’ll turn into a true starter, but hopefully he raises the floor for the team at centerback.

So far I expect the depth chart to look something like: Bartlett and Rowles as the starters, then: MacNaughton, Tubbs, and Antley. Given these numbers and the rosier fullback situation, I expect to see the team mostly using two centerbacks, not three.

The good news is I think top-to-bottom there’s reason to think this unit is improved over last year. I doubt anyone’s going to compete for Defender of the Year, but just competent play should meaningfully help DC in its quest to give up an average number of goals.

Conclusion

I now have three explanations for DC’s 2024 defensive woes:

1. The contracts Ally Mackay inherited, plus some bad luck, meant the team played the entire 2024 with only one real fullback, Aaron Herrera

2. DC’s midfielders, in particular its central midfielders, were either unwilling or unable to cover enough ground when defending, leaving the back line exposed

3. In part because of the fullback issues, DC was forced to give significant minutes to central defenders who were in over their heads (Peltola, Akinmboni, Tubbs), washed up (Birnbaum), or—in the best case—mediocre (McVey).

Ally Mackay has done his best to address all three issues. So a big storyline going into the season, maybe the biggest, is whether this was enough to get DC all the way to “decent” in defense. We’ll have to wait and see, but let’s use my goals blamed metric one last time to think about this.

If you add up the goals blamed on the departed centerbacks (Birnbaum, McVey, and Akinmboni) you end up with 14. If you add up the goals blamed on departed fullbacks (Santos and Dajome) you get another 8.2. Adding up all the goals blamed on all the departed players (everyone mentioned already, plus central midfielders like Klich, Rodriguez, and Garay plus goalkeepers Miller and Bono) gives you 32.8 goals blamed.

That’s a lot of goals! It’s more than the goals blamed on the returning players (27.8). The hope here is that the older returning players will maintain their good form (like Bartlett and Herrera), the young players will either improve (Peltola, Enow, Tubbs) or if they don’t, then they’ll see the field less (Tubbs in particular), and that new arrivals like Kim and Rowles can do better than the players they replaced.

On paper, it sounds like a good bet. The roster still feels thin, but at least it seems like it’s much better balanced than it was going into 2024. Against that, the competition in MLS improves every year, with ever more money being poured into attacking talent.

DC hasn’t exactly poured money into more attacking players, but there are a few new players. That’s where we’ll turn in the fourth and final part of this preview series.