Atlanta pounced on DC immediately and within 4 minutes produced 3 very good chances including 2 very good shots on target. I can tell you that as a fan watching from the stands, it *felt* like all 3 of these shots should’ve hit the back of the net. None of them did, and that happens sometimes. As it turns out, the historical conversion percentages for similar shots from similar locations adds up to something like a combined 50% (@11tegen11 on twitter publishes all of this cool stuff shortly after the matches conclude):
Five minutes later, having added 1 additional shot on target – another very good one (a goal!), Atlanta brought the score of the game level with the combined historical conversion percentages of those shots (something just north of 100%). Mark Thompson summarized this opening sequence and the home side’s attacking output for the rest of the match in this tweet:
Discussing the game this week sucks, but the opening sequence is a nice little example to illustrate the nature of this sport. Good attacking teams create good quality chances, or they create good volumes of chances and great attacking teams do both of those things. But turning those chances into goals is messy. Over very large sample sizes, while the best teams in the world will score more goals than those historical averages suggest (after all they are elite), most teams and players do not consistently overperform these models. Within small samples, say 8 games, anything can happen. Sometimes, Bill Hamid will stand on his head (just as you feared), he will stretch his groin to the limit, and he will force you to reckon with this math the hard way. This is why when evaluating a club, it’s important to assign a greater weight to a team’s ability to create quality chances than its ability to finish them. Underlying chance creation figures are better predictors of future goals than past goal tallies.
Back to the game
That being said, standing at the 10 minute mark, everything was going well for the five stripes (ok fine, I’ll call them that). In the opening moments, DCU were actually pressing Atlanta in their own half (though I haven’t heard this discussed much) but Atlanta were handling it well, and playing long when they needed to. The home side had success finding Kenwyne directly and nodding down to Almiron to start attacking moves, with the real fun stuff coming from the left in Asad & Garza. Overall, Atlanta were creating good chances, albeit different chances than they normally create: more crosses, fewer transition opportunities. And they scored basically the same amount of goals one would expect over large sample sizes based on the chances they were creating. Could they have scored 3? yes. Could they have scored zero, yes. Lay off Tito. Be happy he’s making the correct runs at the far post to create those opportunities. That’s one of the most important things wide forwards do. For the next 10-15 minutes, Atlanta continued to create chances and dominate play. Here are the shots and attacking passes for the opening 20 minutes of the game for both teams. Notice that DCU basically does nothing during this time period:
That’s 60% SOT accuracy, and a 33% SOT conversion rate for Atlanta on decent chances.
What Atlanta did is what good teams do — especially against a DC team that isn’t giving you much through the center, or in transition where you like to operate to create through ball assists. Better yet, grabbing a goal in the first 20 minutes is very useful! It is particularly useful against a team that is on the road and struggling with passing accuracy this season, and meanwhile you’re a team that can make grown men cry with the efficiency of your transition play — transition play that will surely matter now given your opponents should have to chase the game and perhaps open up space for those vicious countering through balls, right? RIGHT?
Imagine the cursor blinking slowly at the beginning of this line for like 5 minutes.
Sometimes stuff happens. You can really go down a rabbit hole trying to think of what, in hindsight, Atlanta’s supposed to change about the way it goes about playing once it’s up 1-0 at Bobby Dodd after positively running all over a struggling DC United team. My first instinct was to get on a soap box and reiterate what I’ve been saying regarding the importance of not being stubborn with building from the back in the face of pressure, when Atlanta’s strengths are perhaps more suited for a back and forth more direct, vertical, open game… And its easy to tell yourself a story about how DC’s goals came from sloppy give-aways in Atlanta’s half, and maybe if Atlanta had just played more direct to minimize these types of turnovers, they would’ve been in better shape. But, I watched the broadcast on FS1 a couple days after the match, and that story doesn’t hold up. Watching the game with perfect knowledge of the events that are to unfold and the timing of those events is always surreal, but I couldn’t help but notice how truly dominant Atlanta was, both leading up to the first goal and afterwards for a spell, and then leading up to the second goal and after for a spell until half time.
The goals that DC United scored were well earned by the visitors in a sense, but very much against the run of play. But there’s plenty of chance involved and some amount of individual errors on the part of Atlanta, which may or may not recur. The first goal in particular looks like a bit of a fluke. It’s a long ball from the back and Parkhurst is marking the target Le Toux with Pirez covering behind. The ball is too far for both of them and it falls to Pirez whose touch is poor and goes right to Neagle (who smartly has pressed high hoping for the second ball – it is incorrect to say DC sat back the whole game). Mears gets lost ball watching and Pirez’ tackle misses Neagle and he’s off to the races. From here, Parkhust does what he absolutely has to do (since Le Toux is unmarked at the penalty spot), but he just gets a bit unlucky with the bounce. To me, while blame can be cast towards Pirez and Mears to some extent, it doesn’t seem like the sort of goal to worry that we’re going to give up on a recurring basis. More of an unlucky momentum killing fluke.
The second goal is also well earned but can be described as a moment of brilliance from a skilled player coupled with a lapse in concentration from several Atlanta defenders. It looked bad for sure, with Acosta dribbling all the way in off a harmless throw-in. But even after admitting that Gressel, and Garza, and Carmona, and our centre backs dropped the ball on this one, that shot still only goes in probably 1 in a dozen times. Kann’s vision is shielded as the ball goes through Parkhurst’s legs. Contrary evidence to my zen argument here is how eerily similar the goal looked to Rusnak’s goal for RSL last week. So while both that goal last week and this goal this week are terribly low percentage chances (chances that Atlanta is probably OK giving up on occasion), the fact that they are repeating consecutively is worrying.
But even between those two goals, I counted four or five nice opportunities for Atlanta to DC’s one (a quick counter started off of a bad Pirez pass). My point is that if you look at the first half as a whole, you’ll see a game where Atlanta is very, very unlucky to go into half time down – even counting those individual mistakes.
One little nugget to chew on: MLS teams average 2.7 blocked shots per game. Atlanta blocks the 4th highest with 3.4 per game and Colorado Rapids are the worst team in the league with 1.4 block per game. Atlanta only blocked 1 shot in Sunday’s game – LGP’s 94th minute clearance off the line with Kann out of the goal. With this as context, you wonder how differently the game goes if Atlanta blocks their season average number of shots — if Acosta’s goal caroms a bit off of Parkhurst’s foot instead of going square through his legs, or of course, what happens if Parkhurst’s diving pass-block goes a foot left or right of where it did.
Having said that, was I optimistic going into halftime down 2-1 to a DC team who would be happy to sit back and hit us on the break in the second half? Absolutely not. This is soccer. These things can and do get worse, compounded by the fact that because the team was playing a rare home game, Atlanta were really going to be pushing for a win and not just a draw – and necessarily leaving themselves overly vulnerable to the very thing Ben Olsen told the FS1 crew he would do heading into the tunnel at half. Most importantly, the thing that Atlanta was going to be able to do up 1-0 (counter), the thing they’re really good at doing (counter!), they weren’t going to be able to do anymore.
Brief 2nd Half Observations
And really, because of that, I don’t have a ton to say about the 2nd half that can’t be summarized with some pictures. I expected we would pour forward and create chances but probably worse ones than we’re used to seeing, and I expected they would have good chances on the counter. That seemed to be the fair trade the two teams were making. I thought Atlanta were still lively before the 3rd goal for DC. After that, there’s all sorts of substitutions and tactical changes that could best be described as desperate. Anyone trying to tell the story of this game simply by comparing the full 90 minute stats will get it wrong because of the very different game states that comprised the full game itself. You had the relatively even game states in the first half, shifting a bit at 2-1, then you have the second half where Atlanta are full on chasing and DC are fully happy to kill the game and counter. If you want better analysis on the tactical changes Olsen and DC made, I suggest checking out Jason Longshore and Jarrett Smith on the Tuesday SoccerDownHere Daily show (sidenote: everyone should be listening to that show daily if you aren’t already – whether it’s live in the morning if you can, or on demand in the evening like me). Also Matt Doyle also broke it down nicely here. Anyhow, I said something about pictures. Here they are:
- DC United pressed Atlanta (with varying degrees of success) in the first half (left) and largely bunkered and countered in the second half (right).
- As a result of said bunkering, the quality of Atlanta’s shots decreased in the 2nd half, and the quality of DCU’s shots increased (as they came from counter attacks).
The off target shots from Vazquez (19) are sort of “meh” headers off of crosses if I recall. Importantly, do you see those 8 shots from outside the box for Atlanta? Scroll up a few paragraphs to see how many of those they took in the 1st half. That’s the impact of game states – DC bunkered, Atlanta chased. A team that shoots low volumes of high quality 1v1 chances suddenly has to figure it out against 10 defenders behind the ball.
- Beer lines were too long, again. And GaTech, what happened to unlocking the gate on the lower level of the parking deck? Please fix for next game, thanks!
Remember last week when I highlighted Asad as my favorite player and compared him to Achilles from Troy. I may have even MS-painted a poor excuse for an Atlanta home shirt onto Brad Pitt with his forearm in an elevated position in an attacking motion… so yeah. Now that it’s official, I can’t argue with the suspension. It will hurt to not have him in the lineup against NYCFC, but hopefully this won’t happen again.
But I did hear some stuff on the MOTS podcast, questioning Asad’s talent and his performance for ATLUTD to date, and I think it’s worth pushing back on this a bit. Perhaps this will be a weekly segment where I defend my favorite players against criticism.
First, the eye test. There is a reason why the left flank troubled DCU so much in the first half. Asad and Garza work really well together, with Asad willing to pinch in and show for the ball (providing an additional direct passing target for Leandro Gonzalez Pirez and others) while at the same time creating space for Garza to bomb wide. Asad troubled DC over and over again, whether it was his forward underlapping runs into the box when the ball was wide or his one touch pass to the bombing Garza while being taken out by a DCU defender. He also is perfectly capable of taking a touch inside and launching a shot with his right foot. He is active and aggressive defensively, which is critical for Atlanta’s press to operate effectively. Now some stats:
Asad has the highest expected assists per 96 minutes of any Atlanta player: 0.27 (10th in MLS for players who’ve played 400 minutes or more) per AmericanSoccerAnalysis, and the second highest expected goals per 96 among Atlanta players (0.17) by similar constraints. For reference, his combined xG+A is just below Piatti’s and just above Almiron’s — I don’t think he’s better than Almiron, but this is some helpful context. He’s creating through ball key passes at the 4th highest rate in MLS. He’s also clearly putting in the work defensively with the press, averaging around 4 tackles + interceptions per game, which while not a perfect metric, at least suggests he’s pretty active defensively for an attacking midfielder. In this game he put up 5 key passes (including 1 assist), and 4 shots. He made 3 tackles and 3 interceptions.
The news that he’s been suspended for NYCFC is unfortunate, as there’s not really a like-for-like replacement for him on the team. The closest is actually Almiron, who could slide out left, which is not ideal given how dangerous he is in the middle of the park. Perhaps we see a midfield three with Kratz, Carmona, and Larentowicz, and Almiron left in the attacking 3?
Also, for what it’s worth, I thought Tito Villalba was one of the best players out there for Atlanta on Sunday. 5 shots on target, 3 key passes, 2 tackles, 2 interceptions, and generally looked threatening and up for it in a game where much of the team’s energy waned as the match went on. If you think Bill Hamid made great saves in those opening moments (I do), then you shouldn’t also think that Tito is at fault for not converting. The first chance is a one time volley where he does his job and gets on the end of the cross. Not much more you can do there besides make solid contact, which he does. On the second, I think he recognizes Hamid rushing to cover the near post, and he cleverly sends it back the other way to the far post and Hamid just makes an amazing leg save. Unsurprisingly, I stand with Tito. If he keeps doing what he’s doing, it will be an incredible season.
I usually don’t pay much attention to this, but for whatever reason I was intrigued to read what the players and Martino said after the match, and for the most part I was very encouraged by what was said. These come from Doug Roberson’s piece in the AJC. Emphasis added.
Martino: “The good thing is we are creating chances and that’s important for us. What we have to do is when we are ahead in the game we can’t give the other team an opportunity to come back in the game. We have to make sure we don’t give the other team those chances to score.”
Asad:“I think we started out really strong. We played a great 20-25 minutes of the game. And then we had some errors that ended up being very costly.
“After that D.C. United was able to sit back and wait and counter-attack on us. They took advantage and were able to win that way.”
Parkhurst: “I don’t think so. We still went up 1-0. We would have liked to bury a couple but that’s Bill Hamid, he’s a very good goalkeeper. We started the game exactly like we wanted to. We put them under a ton of pressure. We got the goal. We were in a good spot. It was downhill from there.”
Parkhurst: “When you create as many chances as we create *you are going to miss some.* I think overall our percentage of finishing quality chances is pretty high entering today. Today was an off day for us. Credit Bill. He made a couple of very good saves.”
To me it seems like these guys have a really good grasp on reality and how the game actually went on Sunday, and the nature of chances and chance conversion in the beautiful game. My opinion is that we should take a page from their book, and reflect on the game for what it is and for what it isn’t. This team is a very good team. And very good teams run into games like this.
Another random observation from the game
I should start by saying that I just flat out do not know what to look for in terms of pitch conditions and how a soccer ball should roll on the floor. I feel like I kind of “know it when I see it” when it comes to a good field. But anyhow, just before the match started, I noticed Almiron gesturing to Carmona or someone else before the game about the pitch conditions. It drew my attention to the way the ball rolled around, and it could’ve just been placebo, but it seemed like the ball had a bit of slowness to it when our guys tried to ping it around — a little wobble in the roll. I also noticed the grass getting eaten up a good bit as the game went on, with visible marks all over the place. Can someone give me some insight into that? Is it normal and I’m just making up this issue? Or was there something different on Sunday – perhaps due to the recent Georgia Tech spring activities there?
In terms of takeaways, I think its best to focus on the first half performance, which was overall very positive and the very welcome development of Atlanta showing it can produce chances from the wings/crosses and not just through the center on counters. Take those two first half DC goals for what they are, somewhat but not entirely fluky “soccer things.” And then don’t worry too much about the second half — it was always going to be difficult chasing the game against a packed in DC defense. Further, resist the urge to ‘require’ our attacking players to convert every good chance they get. It just doesn’t work that way. It totally makes sense to react in the moment — I certainly did — but remember that scoring is hard, and even the best scoring chances are converted about half the time. We’re going to have games where Atlanta misses chances like they did in the first 10 minutes this week. I hope we do at least because it would probably mean that we’re creating tons of chances.
Lastly, just checking in on shot conversion numbers, season to date.