Atlanta United might have been a good soccer team this entire time

Seriously, it's not as bad as you think

There’s no way around it. This year has been a disaster. But something has been bothering me. 

Every week, I throw some numbers from smart people into a spreadsheet and end up with a Power Rating for each MLS team. It takes predictive stats like expected points, goals added differential and…well… points, weights them, and adds them together to give me a sense of who the numbers like and who might be frauds. Atlanta, despite everything this year, has consistently been a top-ten team this season. 

Take a look at the top 19 heading into Decision Day week:

1. LAFC

2. Columbus

3. New York Red Bulls

4. FC Cincinnati

5. Seattle Sounders

6. Real Salt Lake 

7. Atlanta United

8. Portland Timbers

9. New York City FC

10. Colorado Rapids 

11. LA Galaxy

12. Minnesota United

13. Orlando City 

14. Philadelphia Union

15. Vancouver Whitecaps

16.  D.C. United

17. Houston Dynamo

18. Charlotte FC

19. Inter Miami

Why 19? Well, 18 teams make the playoffs. And there’s a decent chance that Atlanta United will be the only one of these 19 teams to miss out. 

What gives? And what does it tell us about where our expectations should be for 2025? 

Being behind in soccer games is bad 

Well, first off, let’s start here with this graphic from Eliot Mckinley

Atlanta United has played with a lead less than any team in MLS that didn’t have to add “FC” to its name to combat bad SEO results versus an NBC show about fires in Chicago. That’s largely because this team just doesn’t know how to find the net early in the game. They’ve scored four (4), four goals inside the first 30 minutes this season. FOUR. 1/11 of their 44 goals have come inside ⅓ of the minutes played. And they happen to be a really, really important 1/3 of the game. 

It varies league to league, but teams that score first in a match tend to win about 70% of time. If you aren’t scoring at all in the first 30 minutes, you probably aren’t scoring first very often. And…[checking]...yep, look at that, Atlanta have scored first in just 13 of their 33 games this year. All nine of their wins have come when scoring first. 

Now, typically, you’d expect a team that’s playing from behind more often than not to struggle in even game states and to have underlying numbers to reflect that. Basically, if you’re playing bad when games are tied, you won’t be playing from ahead very often. But Atlanta has actually been pretty ok in even game states. Per American Soccer Analysis, their 0.08 xGD per 96 when the game is tied puts them tied for ninth in MLS with Orlando and Columbus. 

The problem is that their actual goal differential in even game states is all the way down in 24th at a -0.27 clip per 96. Oops. 

That -0.36 differential between goal differential and expected goal differential in even game states is the worst mark in the league this year and the fourth worst on record in ASA’s database (back to 2013). 

And, folks, that’s just the starting point for Atlanta’s bad luck this season. 

What if this team is actually kind of good?

I’ve already mentioned where Atlanta is in my power ratings. But let’s get a little more granular. 

Per ASA’s expected points metric, Atlanta has created and allowed chances that would typically result in a team averaging 1.51 points per game. If the Five Stripes had average finishing luck this year, they would be on about 50 points and in the running for a home playoff spot heading into Decision Day. Now, we would definitely be talking about them like we did the 2021 team and saying that they were about to get rocked by much better teams like Columbus, but still, the vibes would be much different. 

Anyway, those numbers haven’t really changed over the course of the season. With Gonzalo Pineda in charge and Giorgos Giakoumakis on the team—both finished their time in Atlanta with the Charlotte game on June 2–Atlanta earned 1.00 point per game while sitting at 1.55 expected points per game, the eighth-best mark in the league over that span. In the 17 games since then, they’ve earned 1.24 points per game on 1.47 expected points per game, the 11th-best mark in the league over that span. 

If you change the timeframe to with and without Thiago Almada: With Almada, Atlanta earned 1.2 points per game while posting 1.56 expected points per game over 20 games, the seventh-best mark in the league. In the 13 games since, they’ve been on a 1.00 points-per-game pace while putting up 1.42 expected points per game, the 14th-best mark in the league. 

Regardless of personnel or game state, this team has been steadfastly committed to underperforming their underlying numbers. We haven’t even gotten a chance to mention that the team is third in MLS in Goals Added differential, and G+ is technically more predictive than expected points. 

Why is this happening to this team in particular? And is it karma for my own personal sins?

The answer to the first question is “finishing variance.” The answer to the second is, “Yes, atone.” 

Atlanta have somehow become pretty much the unluckiest team in the world when it comes to finishing. Yes, to a very small extent, the ability of a player to be good or bad at finishing is real, but it almost always evens out over time unless your Lionel Messi. We can safely assume that Atlanta has players that are a little not great in front of goal and a lot unlucky in front of goal. 

Atlanta’s Goals minus Expected Goals rate per game is currently -0.34. Over a full season, that’s going to put Atlanta about 12 goals worse than a team with average finishing luck. That’s a ton. This team has generated four more expected goals worth of chances compared to 2017 Atlanta United and scored 26 fewer goals. With one game left in the season, they’re on pace to virtually match 2019 Atlanta United in expected goals and they’ve scored 12 fewer goals. 

We can go even further. Only four teams in ASA’s database back to 2023 have had worse finishing luck. Two had a small sample size in the 2020 season. The other two are last year’s Red Bulls side (-0.37) and 2019 Sporting KC (-0.34). Considering sample size, we’re going to call this Atlanta side the third-worst/third-unluckiest finishing side in the last decade-plus of MLS. 

Sometimes there are little things you can point to when explaining finishing luck. For example, we knew that Luiz Araujo posted big xG numbers but routinely took shots from the Vine City MARTA station. A ton of bad shots doesn’t make the bad shots more likely to go in. That being said, per FBref, Atlanta’s xG per shot this year is 0.10. That’s totally average in just about every way. Maybe slightly under average, but nothing egregious. 

Well, can we blame one specific player?

Nope! Not really. Saba Lobjanidze is the biggest contributor with a -2.63 between goals scored and their total xG, but pretty much everyone has been varying degrees of unlucky/not great in front of goal.

What does it all mean?

So, stay with me here. I think it means that this is team really wasn’t as bad as we think? I’m not saying they’re great actually or that we should expect a playoff run. Although, I guess I will point out that post-Leagues Cup (AKA post-Miranchuk’s arrival), Atlanta has been sitting at 1.49 expected points per game, the 10th-best mark in the league. 

What I am saying is that I think this is a flawed group that, even with those flaws, is much better than their current record. Let’s say they get the two new DP signings right, add some depth via U22s, and just generally get the offseason right. The bones of a decent team are, statistically already here. If they improve the team over the next few months (and there’s plenty of reason to believe they will), they should be looking at a team set to be statistically among the very best in the league next year.

Does that mean they’ll start getting lucky next year? Not necessarily. But they almost certainly won’t be this unlucky ever again.