Jurgen Klinsmann promised attacking soccer. Didn’t work out.
Bruce Arena was hired to be a steady hand to guide the USMNT to qualification. Also didn’t work out.
Gregg Berhalter has repeatedly stated that he wants to disorganize the opponent with the ball. I.e. keep the ball and play possession soccer. Is this gonna work out?
The question from many fans, analysts, etc. has been does the USMNT have the players to play this way effectively? Sadly, we haven’t been able to see the team play for quite some time to observe possible progress, but that doesn’t mean we can’t still obsess over the question. 🤓
My goal here was to come up with some way to rate how good different players in the pool are in possession. This is where the helpful people over at FBref.com come in. They provide all sorts of stats and they’re actually available for free, which is key (for me at least😂). So I combed through the statistics and tried to pull out any number that I thought was relevant for how skilled a player is on the ball/with the ball. Here’s the list I came up with:
Stats Used for Outfield Players (per 90)
- Number of Touches – self explanatory
- Passes Completed – also self explanatory
- Progressive Pass Distance – total distance, in yards, that completed passes have traveled towards the opponents goal
- Shot Creating Actions – the two offensive actions directly leading to a shot, such as passes, dribbles, or drawn fouls
- Expected Assists – how many assists a player should have had based on their build up and attacking play
- Progressive Carry Distance – total distance the player dribbled the ball towards the opponent’s goal
- Successful Dribbles – player successfully dribbles past an opponent
- Miscontrols – Number of times a player failed when attempting to gain control of a ball
- Non-Penalty Expected Goals – how many non-penalty goals a player could be expected to score based on the shots they take
- Passes Received % – Percentage of time a player successfully received a pass
Stats Used for Goalkeepers (per 90)
- Number of Middle 1/3 Touches – keeper touched the ball in the middle third of the field
- Passes Completed – self explanatory
- Launch % – Percent of passes hit greater than 40 yards (not including goal kicks)
- Shot Creating Actions – same as above
- Expected Assists – same as above
- Number of Defensive Actions Outside Penalty Area – number of times the keeper came out of his penalty area to perform a defensive action
- Avg. Distance from Goal – the average number of yards the keeper was away from his goal to perform defensive actions
- 40 yard Completion % – number of passes greater than 40 yards that the keeper completed
- Passes Intercepted – number of passes the keeper had intercepted per 90
- Pass Completion % – self explanatory
Possession Stat by Position
Possession Scores of Entire Pool
How were the scores calculated?
The values of the numbers are so different that it took a little while to come up with a way to compare and combine them. It’s tough to balance 300+ yards of progressive passing with 0.25 npxG. How do you make those numbers count equally? What I arrived at was comparing the players’ stats to a “global standard.” I chose the “best” players at each position and used those numbers as the standard against which the USMNT player pool was judged.
*I should shout out my wife here btw. She’s a math teacher and helped me come up with a way to have the numbers make sense. She’s much smarter than me.😘*
I used 2019 season stats for the MLS players and the 2019-2020 stats for European based players (with a couple exceptions). I also mostly steered away from using Messi/Neymar as the bar for American players. Messi is such an outlier and Neymar dominates the French league so thoroughly, that using their statistics made even someone like Christian Pulisic look subpar.
NOTE: Full list of “global standard” players used is included at the end of the article
Adjusted Possession Score
This is the possession score multiplied by a number that factors in level of competition. Obviously, completing dribbles is more difficult against Premier League defenders than MLS defenders. I used the SPI club (run by Nate Silver’s 538) and league ratings to come up with a coefficient. I then multiplied the Possession Score by this coefficient to adjust those numbers up or down depending on the level the player competes at.
Here’s the USMNT pool sorted by Adjusted Possession Score:
The “Messi Test”
IMO, if Messi doesn’t come out on the top of the pile then your little stat is worthless. I plugged Leo Messi’s 2019-2020 stats into this when I was all done and fortunately for me he came out to 20.1 raw and 32.9 adjusted. So here’s yet another way to show that Messi is good. The world needed another one of those.
WHERE IS MY FAVORITE PLAYER?!
Easy there, chief. I don’t hate Cameron Carter-Vickers or whoever. Sadly, FBref does not have statistics available for every league. As a result, players like Holmes, Llanez, Ledezma, Miazga, etc. could not be included.
Players with “Notable” Scores
Gio Reyna 7.7/11.8 or 7.1/11.0
Reyna comes out looking pretty good in the raw score and really good in the adjusted score. The problem there is the limited sample size. Gio played 355 Bundesliga minutes this year. There’s a chance these numbers stay this high over the course of a full season…but it’s also quite possible that these numbers fall a bit as he gets scouted and plays more minutes over the course of a full season. In any case, it’s encouraging to see that he comes out so highly rated by this metric.
It’s also worth noting that he scores highly at both winger and #10. His position with the national team is likely to be “tucked in winger” and these numbers would support that deployment.
Cristian Roldan 8.3/6.8
I’m not a big Roldan fan. But when you compare his possession stats to players like Pogba, Niguez, and Milinković-Savić (those are the three #8s I used as “global standards”) he comes out looking pretty good. My guess is that he struggles when the level of competition increases. I believe FBref is beginning to collect data on international competitions so it may be possible to compare in the future, but for now my inference is that Roldan is a very good MLS player that has a tougher time at the international level.
Darlington Nagbe 8.6/6.9
I know he’s not technically part of the USMNT pool, but I wanted to see where he’d fall since he’s generally considered the best US possession player. To give you an idea of how good he is here’s his stats in comparison to the three #8s I mentioned in the previous paragraph:
At the MLS level his stats are nearly equal with the best #8s in the world. No idea if that would translate up to more difficult leagues, but it is impressive.
Miles Robinson 6.5/5.4
I really like Miles Robinson. He’s ahead of Aaron Long on my own personal USMNT depth chart. I did not expect him to come out with this high of a number though. His ability to carry the ball forward is commonly acknowledged, but according to this number he’s nearly as good as John Brooks on the ball. The eye test tells me that’s not true, but the numbers are what the numbers are. Here’s Miles Robinson’s stats compared to Virgil van Dijk’s:
Maybe he’s better on the ball than the common wisdom? I don’t know, but I’d like to see him get more time with the national team in the near future.
Daniel Lovitz 5.8/3.8
I’ve been fairly hard on Lovitz when grading US games, but his possession numbers in MLS are well above average. He appears to be another player that has trouble when the difficulty setting goes from MLS to International. Seeing guys like this do well in Possession Score makes me wonder if Berhalter or the USSF has some sort of metric like this that they used when deciding which players to call up.
Jozy Altidore 8.0/6.7
This isn’t exactly a surprise, but Altidore is miles ahead of the other strikers in the pool by this metric. Sargent closes the gap if you factor in level of competition, but Jozy is still “streets ahead” of his competition in this group. I don’t know if he’ll ever be healthy for a tournament again, but a 90 minutes fit and in form Altidore would be a game changer for the USMNT.
Tyler Adams 4.5/7.1 or 7.1/8.0
I’m a charter member of the “Tyler Adams should be the #6 for the USMNT” Club. It is interesting though that his possession score as an #8 is much higher than his possession score as a #6. If you parse some of the quotes from Berhalter’s press conferences it sure sounds like Berhalter plans on playing Adams as an #8. Bear in mind these numbers completely ignore defensive ability, but possession-wise, Adams as a box-to-box midfielder makes sense.
Thoughts about the Pool
In looking at the numbers 6.0 or so looks to be about where you want to be at or above. From there it’s a question of what level you can perform at. Mark McKenzie checks in at 5.6 which is very close to Tim Ream’s number (5.8), but Ream’s numbers were posted playing in the Premier League. Can McKenzie maintain or even increase the number if/when he moves to a tougher league?
Fullbacks
This metric doesn’t love our fullbacks. Dest’s number might be a little low because his score was based on very limited Champions League minutes, but everyone else fell below what looks like the “good” possession player threshold.
D-mids
Our defensive midfielders also don’t measure up great. Bradley’s 6.5 is pretty good, but I don’t think it’s a high enough number to give up defensively what you could get from another player. Nagbe’s 8.6 on the other hand…maybe.
Forward
Forward is a disaster after Altidore and he can’t be counted on. Someone from the Sargent, Ferreira, anyone please group needs to step up and become a viable alternative to Altidore. Zardes is fine for beating Honduras at home, but if we really want to move this style forward we are going to need a talented forward that can finish, but also drop into midfield and connect play.
Caveats
I know this isn’t a perfect way to measure skill on the ball. The way a team plays effects these numbers. The quality of the players around the player in question effects these numbers. I know there are flaws, but I also think it’s an interesting lens to look at the player pool through.
If you have suggestions or criticisms please feel free to hit me up through the contact link here on the site or on my Twitter handle @blharreld.
List of Global Standard Players
GK – Ederson, Alisson, Ter Stegen CB – Virgil van Dijk (didn’t seem like I needed anyone else😂) Fullback – Marcelo, Trent Alexander-Arnold, David Alaba #6 – Thiago, Verratti #8 – Pogba, Niguez, Milinković-Savić #10 – De Bruyne Winger – Neymar, Sterling, Sancho, Mane #9 – Lewandowski, Firmino
1 thought on “Measuring the Possession Ability of the USMNT Pool”
Comments are closed.