The USA got knocked out of the World Cup by Netherlands 3-1 after a Group Stage that saw them qualify second behind England. It was a very low-event World Cup for the US meaning small moments that could go either way shifted the narrative in many ways. A 1-0 loss on a first half concession to England would have completely turned the conversation around this team in a negative way, but if that early Pulisic chance against the Netherlands went in, the US is likely playing this weekend vs Argentina. The vagaries of World Cup soccer, eh. But these vagaries often set the tone for coverage and thinking for years to come, let us not fall victim to narratives and gut calls based on a bearded or red-headed guy in a studio watching a game and trying to summarize 20 different players actions over the 90 minutes, let us go to the data. Be ready to dive into a red, white and blue smorgasbord of charts, numbers, maps and more and come out with a better understanding of the US at the World Cup.
fields gained=progression metric that treats gaining half the field as 0.5 fields gained (whether it be a 5 yard progression from 10 yards from goal or 25 yard progression from 50 yards from goal). Now am crediting sideways and backwards passes in dangerous areas here also
Shot Tempo: possession actions per shot, lower=possession turns into shots quickly
Long poss%: percent of possession actions more than 7 seconds into a move
Progression efficiency: Fields Gained per Ball Loss
The USA progressed well at this tournament, but did not attack well at this tournament. The red xG per deep completion and xG per 100 Fields Gained tell that story the best, the low success rate of danger zone passes (31.7%) also tell a story of a team that lacked an ability to get the ball from good areas into actual goal scoring areas.
There were not really many problems moving the ball from bad or ok areas to good areas, but that’s not particularly valuable offensively.
Where the ability to move the ball from bad/ok to good areas is valuable is defensively where the US were quite good. Slowing opponent attacks (look at that shot tempo) and making efficient progression tough were the two best parts of the US as a whole in Qatar. John Muller at the Athletic had a nice piece showing just how the progression with the ball helped lead to this strong defensive structure.
Overall the US looked like an above average team at Qatar. I do wonder a bit how much strength of schedule played into this, with Iran and Wales being teams quite easy to rack up big progression numbers/shot edges against.
The US progressed the ball well down their right hand side with Sergino Dest but were not as efficient down the left side. The middle of the pitch was broadly average in terms of progression and around the box mostly red, in keeping with the poor 32% danger zone completion rate. We will get to the individual units later but I think the midfield hype got a bit out of control for a few days there.
Partnerships
The second most fields gained for a passing combination involved the two biggest names and signified how the US were quite good at moving from ok to pretty good areas
No connection had more shot buildup passes than Weah finding Dest on an over or underlap
Defensively the US was quite tough to progress against, but as mentioned above opposition tendencies sort of played into this. Wales and Iran were also very inefficient in their other games while the Netherlands seemed to choose this kind of game to play. Overall though the US did hold opponents below their normal efficiencies, about 20% below their progression efficiency in other games.
However we actually allowed more deep completions to opponents than they averaged in their other games. This ties into the strength of schedule question. While the defense looked good, there remain enough questions that the suspicion we just played a good game vs a passive England and two sorry attacks in Iran and Wales.
US Game Log
The US was absurdly consistent in
-how often they got the ball into the box (between 8 and 10.4 deep completions every game)
-how efficient they were with their progression (between .72 and .8 every game)
The US never had a game where they created a lot of xG for their progression or deep completions. The Netherlands game had the highest value here, but they got almost doubled up as the Netherlands turned their limited progression into 1.5 xG per 50 fields gained, which is why even though it was the best game in terms of out-touching the opponent in and around the box for the USA, they gave up by far their most xG.
US Top Progressors
Weston McKennie: 89th percentile, 11.6 Fields Gained Per 90
Sergino Dest: 86th percentile, 10.8 Fields Gained Per 90
Christian Pulisic: 86th percentile, 10.7 Fields Gained Per 90
Antonee Robinson: 81st percentile, 9.2 Fields Gained Per 90
Yunus Musah: 72th percentile, 7.9 Fields Gained Per 90
US Top Receivers
Christian Pulisic: 91st percentile, 9.6 Fields Gained
Joshua Sargent: 87th percentile, 8.7 Fields Gained
Haji Wright: 80th percentile, 7.7 Fields Gained
Weston McKennie: 78th percentile, 7.4 Fields Gained
Timothy Weah: 76th percentile, 7 Fields Gained
US Shot Buildup Involvement Leaders
Christian Pulisic: 85th percentile, 7.7 per 90
Sergino Dest: 76th percentile, 6.2 per 90
Weston McKennie: 75th percentile, 6 per 90
Haji Wright: 72nd percentile, 5.7 per 90
Timothy Weah: 65th percentile, 4.9 per 90
US Progression Efficiency Leaders
Tyler Adams: 84th percentile
Tim Ream: 82nd percentile
Weston McKennie: 79th percentile
Joshua Sargent: 76th percentile
Timothy Weah: 76th percentile
US xG + xA per 90 Leaders
Christian Pulisic, .49 per 90
Haji Wright: .25
Timothy Weah: .21
Tim Ream: .19
Weston McKennie: .18
Top Takeaway Rate
Antonee Robinson, 84th percetnile
Tyler Adams, 72nd percentile
Tim Ream, 54th percentile
Yunus Musah, 50th percentile
Weston McKennie, 41st percentile
Unit Analysis
Attackers
progression=fields gained per 90, passing/carrying
receiving=fields received per 90
buildup=shot buildup involvements per 90
efficiency=fields gained per ball loss
color coded according to all attackers
The yawning gap between Pulisic and everyone else is wildly apparent here. Sargent and Weah were more efficient with the ball but no one who played more than an hour had even half of the xg production and only Aaronson got within even half of Pulisic’s progression.
He is very underrated by the US fanbase, a legitimate star who was the only reason the US scored in Doha and his performance alone totally reshapes how this team is perceived and talked about.
Reyna had promising numbers over a sample way too small to say anything about.
Aaronson always leaves a good bit to be desired as an attacker for me, the tools are there and I know his passing comes out well at club level but it’s hard to see a big-time attack with Aaronson playing as one of the key 3.
This group has one star, a couple pieces that might develop in Sargent, Aaronson and Weah and a huge prospect in Reyna. Ferreira was great in qualifiers but had no chance vs the Dutch and Wright’s attitude and work rate problems vs Iran were worrying enough that his already poor performance must be eyed even more critically.
Grades
Pulisic: A
Weah: C+
Sargent: C-
Wright: D+
Aaronson: D
Reyna/Ferreira: not enough minutes though one shows A student potential and the other…did not.
Midfield/Fullbacks
progress D is percentile of total yardage allowed through players zone of responsibility
shots D is same thing but for shot buildups
takeways is percentile of takeaway rate
McKennie here stands out as the key offensive man from a midfield and fullback group that did not ease the load on a struggling forward group. McKennie was good at every offensive category here, defensively he graded the worst, but these are much more speculative metrics.
Adams defensively looks great and is efficient on the ball, but that efficiency comes with a tradeoff of him being an awful receiver and slowing down the tempo of the team to where shots rarely come after his passes.
a visual example of what being a poor receiver is vs a good one, stats-wise. Almost all of Adams progressive receptions are just barely so and thus not conducive to quick attacks and shot generation, McKennie gets a lot more verticality involved.
Dest was the much more offensively valuable fullback: much more efficient with more volume progression, and almost 3 more shot buildups per 90. Dest rarely won the ball back defensively, while Robinson was in the 84th percentile there.
The MMA midfield in particular was over-hyped throughout the group. They were certainly solid and probably at around a World Cup average level but far from a group that can push you to good to great performance levels, at least right now. There is room to improve and these guys have a lot of good tools: McKennie’s offensive game looked great, Adams could be a destroyer with efficient recycling abilities and more aggressive recieving and Musah an all-around box to box 3rd man in the future. But that will require more development.
That zone of nothing just inside the own half in the middle is so empty as to make you wonder if it’s basically a tactical instruction to retreat and not challenge there.
There is a lot more blue than orange here, reflecting the different skillsets and roles Robinson and Dest have.
Grades
McKennie: B+
Dest: B+
Adams: B-
Musah: C+
Robinson: C+
Acosta/Yedlin: not enough minutes
Here is a look at how the midfield combined through passing…very little pushing forward between the three.
Center Backs
As always it is very hard to judge center backs from the data. CCV played vs Iran who had the lowest completion% against the USA in the danger zone, but no more than a tick of that can really be specifically given to a player. Ream was the cleanest progressor of the ball and the most active defender. Zimmerman won a higher share of aerial duels than the other two who were only 50/50 at best.
The veteran was cleaner moving the ball forward against Iran.
Grades
Ream: B
Zimmerman: C+
Carter-Vickers: C-
Goalie
Turner allowed 4 goals on 3.4 post-shot xG, for a -0.6 rating which was 31st of 41 goalies at the World Cup. He stopped 7 crosses into the box, the most of any goalie in the WC, stopping 17% of crosses is second highest behind only Unai Simon of Spain. He made 6 defensive actions outside of his area, 9th most in the tournament.
Grade: B-
Teamwide Passing Analysis
next stuff involves cluster analysis to group all passes based on location into 80 different clusters
No team had allowed more passes from these two clusters than the USA’s 52, no one else is within 15. An odd quirk where opponent right backs had a huge share of passes allowed
and opponents did not have much time to putt the ball through the central midfield so we see the types of passes the US did not face.
Offensively, the US had “balance” in a way left and right. With most of their attacks going through fullbacks and wide players and the central forward not being too involved, the US spent a high share of their passes moving the ball on the edges of the wide attacking areas of the pitch.
Unsurprisingly the US did not pass forward in the center of the attacking area much. Left-sided horizontal buildup was not a common tactic either.
The US was not just rarely passing to the middle of the attacking area, but very poor at doing so. This chart shows pass types the US was quite bad at completing compared to average teams.
While they were good completing passes wide, because that’s where all the good players are really. That’s where *all* the players were…usually there is just one player of questionable target man value in the middle. Dest’s silkiness likely helping the right side here.
Defensively, the middle was generally very tough to pass through, the benefit of sticking 3 active defenders in the middle who never really strayed into the box on attack mode.
The US was at it’s toughest preventing completions into the “extended” danger zone (25-40 yards from goal), where opponents completed just 60% of passes. Only Argentina, Ecuador, Portugal and Senegal had lower rates there.
The easiest types of passes to complete relative to average were into and around the box, long balls (defenders like CCV/Ream not great in aerial duels) and then putting around their own half which showed the US forwards were more about denying entries and clogging midfield than pressing for attacking turnovers.
The USA’s high press was not super notable, it didn’t lead to a significantly lower pass completion rate high up the pitch. This midfield crunch led to the games being much less open.
Standout Performances, Single Game
xG + xA
Pulisic .99 vs Netherlands
Pulisic .51 vs Wales
Pulisic .48 vs Iran
Dest .46 vs Iran
Weah .44 vs Wales
Fields Gained Progressing
Pulisic 14.3 vs Wales
Robinson 13 vs Wales
Adams 12.8 vs Netherlands
McKennie 12.7 vs Iran
Robinson 12 vs Netherlands
Progression Efficiency
McKennie v England: 2.6
Ream v Iran: 1.91
Adams v Netherlands: 1.82
Ream v Wales: 1.79
Weah v Netherlands: 1.64
Shot Buildup Involvements
McKennie v Netherlands-10
Pulisic v England-9
Pulisic, Dest v Netherlands-8
Weah v Iran-8
Pulisic, Dest v Iran-7
Reyna v Netherlands-6
Takeaways
Tyler Adams v Wales-9.5
Antonee Robinson v Netherlands-9
Walker Zimmerman v Netherlands-8.75
Yunus Musah v Iran-6.75
Tyler Adams v Iran-6.25
Take-on map…because I couldn’t find a better place for it
4 Dest take-ones led to shots quickly, then a couple Pulisic ones and a few Musah ones. Aaronson was taking on people nonstop in his limited minutes.
Coach
Berhalter built America into a team that could apply their style onto games, and it was a style that was certainly not just bunker and blast.
The US denied shot tempo and shut down the opponent midfield in long possession while breaking up opponent efficiency. This is not easy and a credit to his coaching.
The offensive plan was not exactly clear other than get it to a talented player like Dest or Pulisic out wide and hope they make a very difficult move toward goal. They played well in lots of the midfield but not so well around the goal.
The group might have had just 2 really bad teams in Wales and Iran and the US probably only slightly edged the game against them. The Dutch came into the game wobbling, being played off the pitch by Ecuador and even with Senegal, the US were just able to play a competitive game with them which is not a huge accomplishment.
The big accomplishment in the eyes of many was the England game, totally frustrating England in a game without much threat overall. That’s mostly true, but this game actually contained the biggest negative for me: the USA were massively incentivized to win the match with almost no downside of a loss and should have launched essentially an all-out attacking frenzy over the last 20-30 minutes.
Instead he made like for like subs in 3 positions and took off a true attacker for Brendan Aaronson, a guy generally best used as a pressure merchant up front. More importantly there was basically no tactical changes to increase odds of a goal. This is essentially unacceptable and showed a lack of strategic thinking.
A World Cup is made up of so many fine margins and people always want to judge based on results: did you make it out of the group, did you win in the Round of 16, etc? This is not how to judge soccer games and tournaments, just watching the Netherlands game back in general there are easily 10 moments that could have gone one way or the other to completely change the narrative and result.
Overall I think there is something of a platform built here, the US can push high get the ball and deny the opposition. They don’t have miles to go to get into the box cleanly and in front of goal, but they maybe do have kilometers. The defense looked solid overall, but the Netherlands ease of chance creation did create worry that the defensive stoutness was really more opposition attack struggles.
The US performance was maybe slightly above ok but the big strategy decision was poor, thus Berhalter gets a C+.
The positives
-defense in the midfield, keeping opponents away from goal and making progression very tough
-ability to progress ball reasonably efficiently from defense to attacking third areas
Weaknesses
-essentially no attacking central danger led to an attack completely reliant on Pulisic creating something or an attack coming from very far from goal
How to solve this?
Weston McKennie or Gio Reyna basically being fielded in a 10/Firmino false 9 position would seem to be a clear one. McKennie has had 3 different seasons averaging around 3.5 attacking box touches per game, you’d think if he had pushed up and gotten closer to 15-20 attacking box touches instead of his 8 the attack would look better in numbers.
Reyna averages .46 xg + xa at Dortmund, heavily tilted to passing and will certainly be a guy fitting into the attack going forward. The question will be how to handle all these guys and get a real attack going. Will Sargent be good enough to come through the middle, should Pulisic/McKennie/Reyna play the Firmino role with the other two and Weah surrounding him?
There are a couple years until Copa America for whoever the coach is to solve it. Hopefully they can use some of the tools here to help.
What’s next?
It needs to be Copa America, but the years long absences from meaningful games is such a weak spot in the organization of soccer. After the World Cup I will fix all this with a Qualifying/Nations League post reorganizing everything but for now, we just hope the US makes and hosts the Copa America and then plays it like a World Cup.
Would love comments, questions, criticisms, whatever on this piece. A lot of work went into it and it clarified a lot of how the games went for me and the future of the US.
The Big USA World Cup Recap
Thank you very much! It’s one of the ideas to do similar write ups for the CL and maybe EL. I just happened to have plenty of time and motivation to dive into WC, those two don’t excite me quite as much but they are on the list
Tremendous work !! I wanted to congratulate you for your World Cup coverage so far, it's very insightful. It would be great for the Champions League knockout stages. I can't wait to read more of your analyses.