Scuffed | USMNT, World Cup, Yanks Abroad, futbol in America - #255: John Muller, soccer analytics writer for The Athletic UK

Episode Date: February 22, 2022

Muller patiently explains what he thinks are the most interesting metrics in soccer right now, the basics of how data is collected and used and sold, and some of the limitations of even advanced stats. Belz plays the data-inattentive everyman, which is, after all, what he is. They also discussed and assessed Gregg Berhalter's tenure and adaptability, the US men's player pool, and what John is working on next. Then he turned a few questions on Belz.0:30 intro and the basics of soccer analytics29:25 Berhalter and USMNT discussion startsJohn's interview with Tyler Adams: https://theathletic.com/3123592/2022/02/14/my-game-in-my-words-by-tyler-adams/His piece on visualizing possession in the Premier League: https://theathletic.com/3118980/2022/02/09/visualising-possession-where-is-every-premier-league-team-having-more-of-the-ball-than-their-opponents/Find him on Twitter: www.twitter.com/johnspacemullersupport Scuffed on Patreon: https://www.patreon.com/scuffedsign up for our weekly newsletter: https://scuffedweekly.substack.com/join the Discord: https://discord.gg/X6tfzkM8XUbuy our merch: https://my-store-11446477.creator-spring.com/drop us a question at this link and we’ll try to answer it: https://forms.gle/rfzSEZJwsvnWSCxW7 Skip the ads! Subscribe to Scuffed on Patreon and get all episodes ad-free, plus any bonus episodes. Patrons at $5 a month or more also get access to Clip Notes, a video of key moments on the field we discuss on the show, plus all patrons get access to our private Discord server, live call-in shows, and the full catalog of historic recaps we've made: https://www.patreon.com/scuffedAlso, check out Boots on the Ground, our USWNT-focused spinoff podcast headed up by Tara and Vince. They are cooking over there, you can listen here: https://boots-on-the-ground.simplecast.comAnd check out our MERCH, baby. We have better stuff than you might think: https://www.scuffedhq.com/store Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

Welcome to the scuffed podcast. I'm Adam Bells in Georgia. With me is Greg Velasquez in Iowa. We talk about U.S. men's soccer. Our guest today is a senior writer for the Athletic U.K., focused on the intersection of soccer and analytics. His name is John Mueller. He's a friend of mine, and he's a former lawyer who made a big career change in the pandemic, and now has one of the top soccer writing jobs in the English-speaking world.

Not too shabby. Welcome, John. Happy to be here, finally. Yeah, we've been trying to make. we've been trying to make this happen for a while. And I got to tell you a lot of people on our Discord are going to be happy that you're a former lawyer because most of the discord is lawyers. I was just saying that's like MLS's core demo right there. I'm not sure what it is.

Like what is it about lawyers and soccer nerding? Specifically like 29 to like 36 year old lawyers in like six cities. Like Seattle, Austin, you know, I guess maybe Atlanta. But it's really definitely a type. My brother is that type. I guess still am that type. They nailed me. There's definitely a lawyer from Seattle on the Discord.

Definitely one from New York. All right. Enough about that. So what are, in your view, the key metrics in soccer? An incredibly general question. But I figured you can just take it whatever direction. you want? The metric that I think is kind of most exciting right now because it's sort of like just enough out there in public that people are actually doing stuff with it. It's not a purely

theoretical construct or something that's only available to people with a lot of data. But it's also analytically useful, which not a lot of public data and public metrics and soccer are, I think, or at least they're of very limited use. Anyway, the one that kind of, I think, is that at that sweet spot right now is possession value metrics. And I'm sure you've heard of G-plus, goals added, that American soccer analysis paid. There's, you know, expected threat.

There's other flavors of this thing. All of these metrics, what they do is they measure goal probabilities at different points throughout the game, right? So if expected goals is just like a goal probability at the moment of a shot, possession value measures goal probabilities

at the moment of the past before the shot, the moment of the dribble before the pass, the moment of the tackle before the dribble, you know, every step of the game you're measuring how likely is one team to score in the near future and how likely are they to concede, which is really kind of the fundamental question that you're always trying to address in soccer, right? Yeah. Now, I have to confess, I haven't read the, like, the notes, you know, the footnotes on those very carefully, like on the G-plus model and stuff.

But I- there's a lot of notes, yeah. Yeah, yeah. So I wonder like how much do you trust those models? I'm not criticizing because I have no grounds to criticize. I haven't even read the notes, right? I'm just saying like it's it seems like such a complicated thing. Somebody makes a tackle, you know, just inside the opponent's half. How do you even assign a goal probability in the near future to that?

Can you give us the layman's terms? The way that a model like goals added does it is, it has certain features about that possession, certain, like, kind of little bits of information, like how fast was the possession moving up field before the tackle that led to the turnover? You know, was it moving across the field? Had there been a set piece recently?

All these kinds of things that might kind of give it a little bit of information about where players are likely to be and what they're likely to be doing. Because the big problem these models have is that they're based on event data, which is just stuff that happens on the ball. And it doesn't tell you wherever, everybody else is off the ball or what's happening.

And so when you have an event database possession value model, either it's just based on like the X, Y, coordinate of the ball, how close is the ball to goal, which is essentially what expected threat is, is just based on locations. Or you're trying to use all this other information about the possession to kind of like let the computer guess, okay, is the defense likely to be organized right now?

You know, is it likely to, are you likely to pass the goal? And the way it does that is just it takes all those features, that it knows about the possession and it compares it to a whole bunch of other past possessions and there are different machine learning algorithms that you can use to you know kind of match it do those do those I'm sure the algorithms I'm sorry well I was just curious do those does so the distinction between G plus and like say expected threat is G plus does it in one of the distinctions maybe that they know where the defense is like where all the defenders are where all the other players are on the field?

Yeah, so it doesn't. None of these models that's based on event data knows where the defense is. Okay. And some of them kind of try to guess where the defense is. And some of them just say, you know, we're just going to look at locations and give you just kind of an overall average. But the limitation here is not that the algorithms aren't good. I trust the algorithms, but I don't trust the models to come back to your question

because they don't have so much. information about what's happening on the field and most of the models just trying to guess that. So those values are going to be really imprecise, you know, at an event level, at a game level. But once you get enough of them, you know, over the course of a season or a thousand minutes or whatever, then you start to get, you know, pretty decent measurements of goal probabilities and you can do cool stuff like with that, like we did with XG, you know, five years ago. When you say people are out there like using this stuff in reality, are a lot of clubs using G plus and expected threat or things like that?

Yeah, I think that probably by now in 2022, a lot of clubs are using some form of possession value metric. And that's because a lot of clubs have been hiring over the last couple of years. And, you know, they now have somebody who can do this type of stuff in a way that they didn't just a few years ago. But also, these models have become much more mainstream, especially the, you know, just location-based models like expected thread. and, you know, other PV grid type of stuff.

That stuff is really easy to do. And, you know, a lot of people are doing it online. And I have to imagine that clubs have seen this by now and have been like, that's useful. Now, if you want to, like, actually build a model like Goals added for yourself, that's a lot harder. So what a lot of clubs will do is kind of a middle ground where they'll buy stats bomb data, for example, that now comes with possession values and they'll just use those out of the box. I think that's probably the most cost-efficient way to do it. I see.

What does that cost? What does it cost to get that data from Statsbaum? Do you know? A lot, dude. Like way more than you would expect. And, you know, obviously it depends on a club situation and whatever deal they work out with the data provider. But just in general, soccer data is really, really expensive.

And that's part of why analytics is bad because data is just so hard to come by. Like, nobody wants to give this really expensive stuff away for each of the people who might do cool stuff with it. There's this thing called the Davies model, right? That I think a guy you know, Michael, Is it Michael Mbroglio? Yeah, Mike Murgio and Sand Goldberg did that for American Soccer Analysis. So that is, and people can Google it if they want, but it's, that's like a, is that like a modification on G+, or? So it's basically, it's a model of a model, right?

Gplus tries to measure goal probabilities, you know, at the event level in games. But a lot of what that kind of model is used for by clubs is just for recruiting. and for recruiting they just want lists of guys who are good at doing this or good at doing that or just are good in general relative to their peers, you know, at their position or in their role. So what Sam and Mike did was they said, well, okay, soccer data is really expensive. It's hard to come by. But Statsbomb gave some data to FBREF. FB rep publishes tables full of good kind of season level data for players.

And so they use that season level data for players to just kind of see if they could estimate what their goals at it would be if you had all. that experience of event data. So it's a model of a model. It's not going to give you anything, you know, groundbreaking, but it is going to give you a fairly decent way to kind of rank players by their contribution to their team. Okay.

This is probably a dumb question for somebody who knows a lot about this stuff, but where does Statsbaum get their data? Do they have warehouses full of people watching games and typing stuff in the... They literally do. Yeah, you would think it would be more high-tech than that, But, yeah, the way Statsbom and Opta are the two big event data providers. There are a few others.

But the way they do it is they have people watching a game. Now Statsbom does live data as well as Opta. So they have somebody watching a game live pressing keys on some custom software to say, okay, this event happened right there, and then this event happened right there, and they're typing as fast as they can, and keep seeing it. But it's more or less live. And then they go in and they do. various levels of kind of checking and improving the data the day after.

Statswomen uses a lot of computer vision stuff so they can get freeze frames of different moments in the data and get really precise locations for like shots, for example. So, yeah, they've got a whole bunch of human stuff right now. They're kind of working towards automating where the data comes from. A lot of the tracking data that is kind of the next wave now. That stuff is all automated through cameras and computers and whatnot. But syncing the stuff that the computers get to the stuff that the humans get is actually really hard. I bet.

Yeah. If Statsbom and Opta are the ones doing the most event data, I've heard you say before that Y Scouts data is not that reliable or something along those lines? Yeah, it's not, I mean, traditionally not as high quality as the other two. I've heard people tell me that it's gotten better lately. Oh, really? I don't know. They would be getting their own event data or is that? Yeah.

Anybody who has, you know, who's selling their own event data is collecting it manually. Themselves. Okay. Yeah. So they like Statsbaum has a whole office in Egypt where like hundreds of people log the data and then send it to England where Statsbomb, you know, pushes it out. A country where people understand soccer. I, I know. Yeah.

Yeah. Yeah. Yeah. Yeah. there are some really smart people working at the Egyptian offices, that's part. Well, I feel like you'd have to be a pretty reliable person to carefully watch a soccer game and accurately enter the data. I mean, I know they're trying to, they're trying to check it with computers and stuff, and that's difficult. But it seems impossible to me.

I don't know how to do it, honestly. But it's good quality stuff. I have to watch a sequence like three times before I can write an English sentence about it, you know? Yeah, yeah. Yeah, and I always think about what you've said about, why Scout about that idea of them being like somewhat less reliable than the others because that's the that's the service I pay for that's the data that I have you know aside from FB ref like everybody else so I'm always like oh man John Miller does not approve of this data honestly like all of the manually collected event data is bad in various ways so I don't know like a lot of the ways that we use data are bad. regardless of whether the data itself is good quality or not. So there's just so much room for error here.

I think sometimes it's okay to use good but not great while he's got data to get at your basic player numbers or whatever is they looking at. Is somebody okay in the background there? I thought I heard so. Yeah, are you picking up this party up there? Oh, there's a party outside. Yeah, not much I can do about that.

No, no, that's okay. That's okay. The new apartment backs up on a courtyard and there are some Friday types of building. Sure. What about the XG models, you know, the models that assign a number to each shot, this shot is worth this much of a goal? Do all the services you use a different model?

Yeah, so not just all the services, but a lot of people have their own models. And every time that you see an XG number, you know, it's likely from a different model than the last one that you saw. Okay. So yeah, it's probably not worth getting too attached to a specific. The fine green XG numbers, yeah. Right. I was just saying, again, these things work over large samples, right? And so as long as you're using XG the way that it's intended to be used,

not like in a single game level, but, you know, over many games, the differences in the models will kind of wash out in the long run. I see. So we can't use like a single game XG to. say the USA, the USA produced, you know, 2.3 XG and El Salvador produced only 0.25 and so therefore we dominated the game. Can we use it that way? I believe I've heard that line of argument before from one Greg Burholter. Yeah, no, that's not the preferred use of XG, although I love Greg for, you know, dropping XG in conversation. That's not the best way to do it.

Is it all wrong, though, to do it that way? Because, I mean, it does tell you some, about the game, you know, especially. It does tell you something for sure. And in some ways, it's better than just citing shot counts, which is what you might have done before, you know, if he said, well, we lost one zero or whatever the game was and, you know, but we outshot them 30 to 5. That would mean something, right?

It might not tell you a lot because the five shots could be really high quality and 30 shots can all be shit. And XG kind of starts to correct for that. The problem is that, XG kind of, for one thing, it doesn't record any information about dangerous attacks that don't end in shots, right? So it misses a lot of the game that's not even shown up because just there's not a shot event associated with it. But also, like, the kind of stuff that can get registered as a shot in the data doesn't always look like what you imagine a shot to be in your mind. And sometimes

there will just be weird touches, like sort of close to goal that wind up looking like an 80% chance to score and really it's like you know some dude bounced off the back of his head or something right so so when you have those kind of huge swings in in xg values attached to like somewhat random events and like not great data that's why it's a problem uh you know over a single game because your your ball off the back of the head is outweighing everything else that happened in that game in the single game xg total right i mean i think of a of an example that's not quite as wild as it bouncing off the the back of somebody's head, but like in the Canada game, the Canada USA game that the U.S. lost to zero,

there was a good Jonathan David shot from distance that drew a save from Matt Turner. And it spilled from Turner. Turner spilled it right in front of him. And then Kyle Aaron came in and tried to shoot. But he was so close to Turner that it was like, it would have been, you know, a pretty amazing shot to get it. Yeah, just like a point blank like into his gut type of shot.

where there's no way to get it over the key. Right. Yeah. Yeah. And I, and you, I, you know, I was surprised when I looked at the XG timeline that, I mean, probably pretty high, right? Yeah, it came off as, that was like the really big chance, the laryn shot. Right.

Whereas like my brain watching the game thought, didn't even register that as that, that big of a deal. It was so close to turn. Yeah, exactly. And your brain is right, you know, like you've watched enough soccer. You know what is and isn't a good chance to score. your brain has a much more precise model of those probabilities. When you're really focusing on one shot,

you can give a better XG estimate than most XG models can. And that's because, again, you can see what's actually happening on the field. And all the XG models can see is like where it happened and like this was a shot. And, you know, maybe some information about how the shot got set up. But they have no idea that Matt Turner is like already on top of the ball when David tries to kick it. Yeah, Laren. Yeah, Lerner. Yeah. So I saw you linked to a Dutch guy who has had a bunch of, has like a website with a bunch of links on it, right?

Yeah, yeah, yeah, yeah. Jan Van Herrin, he's the data scientist at Club Bruchina. Okay. Not Dutch then. Well, maybe he's Dutch. I don't know, but he's in Belgium now. Yeah, he may be Dutch, but yeah, he's in Belgium. Either Dutch or Flemish for the record. I'll check with it. one of the first papers in that list was, was from some, some guys who made a new XG model.

And I think the idea of it was, um, more precision or, I can't remember if it, if it took into account the, the positions of all, the locations of all the defenders, but I think maybe it did is that,

did it? You know what? I think I was actually reading the dissertation that included that paper like 10 minutes before our conversation. Oh, really? Um, if, if we're thinking of the same one,

yeah, It was essentially, it was an XG model that used a lot of tracking data stuff. It's not usually available to the people who make the XG models. And so it would be more precise than most of the XG that you see out there. So how long does it? So what's the, what's the runway for something like that, like that to become, you know, part of the discourse, something that's actually used by FottMob and, you know? Yeah. I mean, in that particular case, like the big events, I think, was not.

so much the model. They were just excited that they were able to incorporate some tracking data information, some awful information. So, like, the difficulty in getting that to the public is not, you know, sharing a new brilliant model. It's, you know, can the public get tracking data? Is that going to become cheap enough and easy enough and widespread enough that, you know, media providers are going to start buying it and, yeah, Float Mob is going to start buying it. And probably the providers like Stapform and Statsbom are going to start, you know, building the, their own tracking data, XJ models and whatnot. Yeah, then the numbers that you see will get better,

but you won't even know it. It'll just kind of happen. Okay. So who does have tracking data right now? Does Statsbom have tracking data? So Stats Perform and StatsBomb mostly sell event data that's augmented to various extents by tracking or tracking-like features. But tracking data mostly comes from independent providers,

like an MLS, second spectrum, provides tracking data, to all the clubs. Second Spectrum does the Premier League. I don't know who does the rest of the leagues in Europe. But most clubs have access to tracking data now. The question is, do they have the people and the know how to actually implement it and to do useful stuff with it? And I think in most cases, the answer is no.

How did they get the tracking data? Is it from the sports bras? Or, like, is it like visual from the footage? Yeah, no, I mean, like, there are different ways. There are, like, RFID tags and, like, GPS tracking and stuff like that. but that's not as precise as what they actually use, which is just optical tracking data. The setup usually like eight cameras at various angles around the stadium

and they'll capture 25 frames a second of where everybody is and what they're doing. I see. Yeah, so it's all augmented. You just wrote an article about which parts of the field are controlled by which teams in the Premier League. Well, not just. It was a couple weeks ago, I think, but it was very cool.

And when you help make those types of graphics, you know, showing who has control of which. zones in the field, where do you get the data for that? Yeah, so our event data at the athletic comes from Opta, from Stats Perform. We sometimes use other data sources. I use a lot of FBRAF for Europe. We use a lot of Smarter Scout for leagues all over the world. But when I'm doing event level stuff, it's, yeah, it's usually Stats Performed data.

Okay. Anything you learned from working on that article besides what was in the article? I'll link to it in the show notes. It's good. Yeah, I don't know. It got me thinking, I think, in ways that I hadn't before about how possessions develop and why certain areas are more likely to be controlled by the attacking team or the defending team.

But all this article was that you're mentioning is just kind of a map, sort of a heat map of where teams had more than 50% possession or less than 50% possession in different zones of the field. And, you know, it was not anything famous. fancy by any means, but it did kind of show in a simple and easy to understand way, like where primary league teams are active or where they're defensively vulnerable. And I think that that spoke to a lot of people. And so I, you know, tweeted this biz and a lot of people liked it. And so I wrote an article kind of explaining what it was and how it worked and trying to think about,

you know, okay, yeah, it makes sense that teams have to build out through the center of their own half before they can get much of a foothold vienj hacking half. You know, they look for an outlet just past the halfway line first so that you can kind of circulate, right? And it's actually pretty easy. Usually just inside the other team's half on the sideline, basically, right? Yeah, if you go a little farther up, well, then you're in more contested territory. You're more likely to get closed down. You're not going to hold possession there very long.

But it's fairly easy to control the zones just past the halfway line out on the wings because those are not really high value zones. The defense is trying to take away from that. Where do you think, you know, if Greg Burhalter is, looking at this kind of data for national team games. Where is he likely to be getting it from? Is he getting it from second spectrum?

Maybe you don't even know. Yeah, I listened to Burr-Halter speak at the Sounders conference last year. I don't remember what he said about his data sources. I'm sure that they have stats for them off to data. I don't know if they have tracking data. Yeah, I have no idea what's going on at the Federation. Is it frustrating to you, before, I do want to ask you some stuff about the national team.

Because, you know, this is a national team podcast. Is it frustrating, but first, is it frustrating to you that soccer is a game where you can win all, you can outpossess your teammate? You can play the game in their half. You can, you know, you can do everything right by the metrics and still lose, you know, if your center back slips one time. And I guess, let me just qualify that question by saying, you know, way a lot of sports are like that right but um we're talking about soccer no honestly i think that's what makes soccer great uh and one of the reasons that it makes it great is that i personally don't care about the score at all i just like i don't care about goals and points i care about the

quality of soccer and the quality of soccer more often is i've been reflected in things like who's possessing ball wear and you know how many shots are they creating are they high quality chances all that stuff where people say, well, only goals count. I don't know. I say goals are the only thing that doesn't count and watch what happens on the field because that's a lot more fun than checking the school. Where did you catch the bug for that, that passion for being a soccer guy? You know, I played as a kid like everybody else wasn't very good.

So I kind of forgot about it and played, you know, real American sports. I've been in my 20s. I lived in Latin America for a long time in Brazil and Mexico. It was the only sport that anybody wanted to play or watch or talk about. After a while, I started watching a lot of Barcelona. And Barcelona made the game make sense to me in a way that it hadn't before. And I think that you can see that difference when you watch the peak Barcelona teams play and basically any other soccer team on Earth. So for the last 10 years, I've just been trying to understand what it was that they were doing

that made the game make sense. and why other teams don't play that way. Do you feel like Man City under Guardiola is the same? No, I think they're a different beast, but they are kind of a more precisely engineered machine for doing the same essential thing, which is controlling the ball in the opponent's half and very patiently creating chances

while also denying opportunities to counterattack, which is all about controlling the probabilities, right? And Man City's not going to win every game. And they can't ever win the Champions League because knockout soccer is random. But they're playing the long game. They're playing the probabilities and again and again over the course of the season. Like those probabilities tilting their favor. So I think what makes soccer fun is that random shit happens and like the team that didn't deserve to win can win

because otherwise it would just be city winning every game. But I also do like that like good soccer is rewarded in the longer. Yeah. Especially in the league table, I suppose. Greg Velasquez, who you know, and I just talked about this, like, this John Hurtman presentation that we, that we watched for an episode a couple weeks ago. Yeah, I want to hear that episode. Yeah. Yeah, it's pretty, it was, I mean, the presentation was interesting to me, you know, so I'm kind of like a, I'm sort of a, whatever, it was

interesting to me. But they talk, he talks about and, and his staff focuses on something that they call an AVP, an advanced platform. Yeah, I remember you talking about that, yeah. Okay. So it's like this idea that if you have somebody with the ball at their feet running a naked back line or running behind the back line, that's a moment that they clock and then they credit it to sort of everybody. Like that's, you know, that's what we want.

And it's, and I, and it seemed to us, like that's a useful way to watch soccer. You know, an additional useful way to watch soccer, besides like clocking the good shots and like the quality of the shots, how often are we putting our players in a position where they can create it, where they have an opportunity to create a chance, you know? And I wonder if they're, you know, are there ways to, are there ways that G plus or, you know, expected threat

or any other statistical stuff that you know about, analytical stuff that you know about, sort of brings that concept to bear? Yeah, I think this kind of brings us back to where we started with the possession value stuff, because what you just described is kind of a manual way to sort of try to measure high possession value attacks, right? And they're trying to look beyond shots.

They're trying to say, okay, well, where did we create danger, whether or not it led to a shot, you know, what was the dangerous moment right before that, right? That's basically what an AVP is about. And possession value is trying to do the same thing, like not just measure shots, but measure dangerous attacks, you know,

and measure goal probabilities at different points in the game. I think that in some ways, hand-coding your AVPs is going to be more useful than G-plus or whatnot because, again, you're looking at where the defenders are in the field, what's happening in the play, you're seeing more of the game than event date is. but like all those possession proxies that we talked about that a model like goals added uses to try to guess

at what's happening with the defense are probably pretty good at figuring out when something like an AVP is happening because one of the most influential features on those models is how fast is the ball moving up field and the faster the ball is moving up field if you've got it at your feet under control the more likely you are to be running at an exposed back line, right?

So I don't know. I'm a little hesitant about the idea of an AVP because that's just like one very narrow way that teams can create goals. And it seems weird. I mean, you know, it's part of their game model and they've got great players for playing that way. It seems weird to kind of ignore every other kind of chance

that you can create in soccer and just focus on one. Yeah. Yeah. Speaking of Herdman, speaking of Man City being, you know, a finely tuned machine. Let's talk about Greg Bertharter a little bit. Speaking of finely tuned machine. Right, right.

Well, you know, he's not been shy about saying that his aspiration is for the U.S. to play like Manchester City. I mean, I wouldn't say maybe he hasn't been shy isn't the right way to say it, but he's let it slip at least a couple times. And like that's his, that's sort of how he's thinking about it. Meanwhile, John Hurdman, who is, you know, as we all know, been much more successful in the last several months in KakaCath World Cup qualifying. he lists like Greece 2004 and Iceland in 2016 as potential models for how Canada wants to play. And I wonder if what do you think of this idea that Burrhalter has where he wants to change the way American Soccer is viewed and he always talks about a man city. Like is that a good idea to talk that way? Yes, I mean, we always knew what we were getting with Burrhalter.

he, like you said, he was never shy about kind of letting us know up front that he had a really ambitious idea of how a national team could play and what kind of, you know, system he could build at that level. And everybody knew that, you know, even before he was hired, everybody knew that that was ambitious and we weren't sure, you know, how well it worked. I was just talking to Tyler Adams for an article like last week or the other week. And he was saying, you know, look, yeah, of course there's never enough time at a national team to create a quote-unquote philosophy, you know, to create a detailed tactical system like Manchester City has. But, and he didn't say this part, but I think that it's really beneficial for us that more and

more of our players are playing in systems somewhat similar to what Burrhoelther's ideal is, because I think that if you're getting a lot of MLS guys, and I don't want to like wait into the whole MLS versus Europe bullshit, but, but like not a lot of MLS teams play, you know, like high level champions league way to de position right that's that's just not really a big thing and so having guys at Dorman and Leipzig and Chelsea and

City and Barcelona like even though these clubs are very different from each other in some ways they're also very similar in some ways and they're practicing the same fundamentals at a high level in a way that I think will maybe eventually be conducive to us playing soccer something like Berthor

imagines it but I think that we've seen for three years and to be fair three pandemic interrupted years where it was very hard to like kind of have to abandon what he was trying to build in his first year and then not train with the guys for a year and then like kind of retool the system and come with a different idea in year three like there's been a lot of change a lot of evolution over his tenure and even though the team is honestly not great right now I think that they're building in the right direction let me just quickly say that this is a listener supported ad free podcast.

If you are able, please consider supporting us on Patreon. The link is in the show notes. And if you sign up, you get access to the Discord and also access to the Monday reviews with Vince and Waki, which are mostly behind a paywall. Thank you to all our patrons. Much, much appreciated. You kind of answered the next question I was going to ask, but let me ask it anyway. How would you rate Burralter as the coach of the national team?

team. So let me say as a person I love Greg Burrhalter, I think that he's super thoughtful and, you know, very honest in the way that he explains himself and his thinking and he'll, you know, kind of really talk soccer with anybody who's curious about it. And I love that in a coach. And that's kind of why I've always been a starry-eyed optimist about this whole project of like trying to play way more ambitious soccer than probably any national team should. And I love the idea that we've got all these young guys coming up in these. systems and like we can fit them together and play like really high high quality international soccer and it's possible you know i think that spain plays like really high level club soccer

at the national team level but again that's because those guys are spending 11 months out of the year training with clubs and doing that stuff all the time um it hasn't worked out that way so far obviously i think under burrhalter uh we've been perfectly successful by u.s historical standards uh maybe not as successful as we'd like to be given the quality of the players that we're putting on the field. But I think that Burl has been very adaptable. He's, you know, kind of assessed things that were and work working and has retooled. And I think that his vision for how the team should be playing right now is something

that's probably achievable in the time frame that we have for the World Cup. The biggest problem that I have is that I feel like the team's identity is no longer consistent from game to game or even half to half. and it's not clear to me and it looks to me like it's not always clear the players how they're trying to play in any given situation and why because there's just so much more variability

in the way the team plays now. Does that sound right to you? I think so, yeah. You know this games way more times than I do. Yeah, but I mean, I guess if I'm going to try to defend Burrhalter, I would say we do naturally fall into a state of trying to possess the ball, you know, as a team,

which I don't think was always true of U.S. teams in the past. And so that is, I mean, you could argue that that is a step forward for men's soccer in America. But we're not really doing the thing that we were promised about disorganizing the opponent with the ball and all that good nonsense. Like, yeah, we can keep the ball, but we're not really doing stuff with it in possession, kind of maintenance buildup phases. That said, like one of the advantages of being a possession-based team. team that can at least circulate the ball and kind of slowly push you back into your half is that you're more likely to create transition moments in the opponent's half and those are more likely to lead to goals. And that's kind of the side of the team that we've seen

develop, I think, in the last year or six months, whatever, that we may not be able to disorganize the opponent with the ball, but we can at least use the ball well enough to get into the other half and then create those transition moments. How much is our inability to disorganize the opponent with the ball? How much of that do you think is because of a lack of quality in the front three or even like the front six, like a lack of attacking quality. Because I think Wes McKinney has quality, but I don't know if he's probably our, you know,

he's the hottest player for the U.S. right now. I mean, I mean, I mean that, I don't mean that has like a joke. I mean, like, he's really good. He plays really well. I thought you were just commenting on his, you know, aesthetic. But even he. Even he's like, I think he's my favorite player right now. I think he's playing the best of any American for the national team.

But even he, I don't know if I back him to be 100% incisive in the box, you know, or in the final 18 yards in general. I think that's where he's best. You do? Yeah, I mean, we've had this conversation before. I'm not sold on Weston, the midfielder, but Weston, the box driver is a killer. And that's kind of been his main role for you, very lovely.

That said, I think Weston McKinney has, more than enough quality to boss a conca cap midfield than most international midfields that you'll see at the World Cup. So I don't think that the problem is like the quality of the players, especially when we've got our best players. I think one thing that gets lost a lot in national team discourse is just that the national team is not one thing. It's so many different teams, you know, and you're like seeing wholly different guys in the same shirt from year to year from month to month from tournament to tournament. And so, you know, it's been remarked on, but still, kind of isn't played up enough that our best players have not played together at all,

you know? Like we just now got them on the field in the last, last window where like four out of our five best guys, we're playing together, you know, for the first time. And that makes, I mean, it sounds almost too obvious to say, but that makes all the difference when you're trying to, you know, do anything tactical when you're trying to organize a team as a unit. They have to play together and those opportunities just haven't arisen. Yeah, I mean, it may sound obvious, but it's for those of us who are so deep in the topic, we kind of, it's good to be reminded of the obvious things every now and then.

I guess it's not McKinney, you know, it's not McKinney's lack of, I just brought him up as the example of, like, the best. But, you know, Pulisic, Poulisic has not been sharp for the national team, even when he gets into dangerous areas. He, even when he gets into dangerous areas, and he doesn't get into dangerous areas that often because he's not, doing that much on the field. He's trying to dribble people and not succeeding. I'm talking about the sample of like the last four or five games with the national team. Yeah, I mean, he's just having a really tough time this year. And you can tell at the club level, at the national team level,

he's putting a lot of pressure on himself. And he's kind of had some down times at the club, obviously the injury problems. But when you see him with the national team, he's a whole different person, a whole different player who's just trying to do way too much. and dropping below the halfway line to collect the ball and then trying to dribble A players and none of it fits with the system and I'm sure that none of it is what his coaching staff

would like out of him, but he just kind of was putting that pressure on himself to be the hero right now. I think that that'll, you know, chill with time as things get better for him at his club and as the rest of the team kind of starts picking up the slack so that he doesn't feel like he has to be the hero. I think things will even out.

What was your impression of Tyler Adams? talking to him. Oh man, he's such a pleasure to talk to. He's a really thoughtful dude who, you know, I came into this interview, having heard horror stories of, you know, some other player interviews who, yeah, he's just kind of worn interested in breaking down clips.

And Tyler was the opposite of that. I just asked him about clips from his career for an hour and asked him, why did you do this? You know, why didn't you do this instead? What were you thinking at this moment? and he was always willing to go into detail and to connect, you know,

the tactical concepts that he learns at RB Leipzig and the players that he plays with the national team level and kind of how all of that ties into his thinking, you know, he could be a situation. It was a really fascinating interview, and it's unlocked so anybody can go read that if they want to, and if you haven't yet, I hope. Yeah, agreed.

Did you show him the clips, like, in real time on, like you shared your screen with him, or did you fly to Leipzig? No, we hopped on Zoom when I pulled up a Weiss Cow playlist of stuff that, you know, I knew I was going to be able to watch six years of his career. So literally I just watched, I asked a bunch of friends like, what's the moment of Tyler Adams career that sticks with you for whatever reason? And you gave me a good clip that I think I wound up talking to Tyler about and so did a bunch of other friends. And together they kind of, I think, captured not just the arc of his career, but also what he's good. it at, you know, what he's bad at, what he, or at least what he's learned from his bad moments.

It was a good package. It was a good conversation. Yeah, absolutely. Crowdsourcing content, man. It's the, it's the way to go. I do it all the time. I use Google forms like crazy on, I even ask people for questions to ask you. So.

Did any good ones? Yes. Yeah. Yeah. I mean, one question that I got from people, which I didn't ask earlier, which I need to ask is when you're assessing somebody using G plus or whatever, you know, any analytics, how do you correct for like team effects, the quality of teammates or, you know, the way a team typically plays, whatever?

Yeah, I mean, that's a, that's a really difficult question in sports analytics and especially in soccer analytics where context is so important. And I think that the answers can range from somebody builds a really complex. complex model to try to adjust for team effects and game states and all the various contextual factors and, you know, give you a real read on the number. And then there's kind of the opposite approach where you just don't adjust for anything and you're very honest that this is not adjusted for anything. And so you have to look at these numbers, which are at least like very straightforward and interpretable and, you know, use your brain and what you know about the teams and like, you know, kind of apply that context to mentally. And I think that both approaches have some virtues and some drawbacks. I try to do both basically.

Like, anytime that you see a number in soccer, you should be trying to contextualize it in your head because there's always something wrong with it or something that's missing and something that you need to think a little bit harder about before you uncritically except that number represents whatever it is that you think represents on the field. So I shouldn't just take foot mob scores as gospel on who was good? who was good in the game?

I think they're better than the Sulfa score scores. Are they? I don't know why I think that. I feel like they at least are outside the range of just like 6.5 to 7.5 every game so I can at least tell when Feltmob thinks somebody was good or bad. So what are you working on, what are you working on right now? What's coming? What's coming from Muellerland?

So I had a real bad weekend because the big piece that I've been working on lately has been essentially trying to understand what makes Brighton good. And the problem with that is that Brighton just lost 3-0 to Burnley after losing to Manchester United the week before. And just like, they're just really, they're killing my narrative right now. But it wasn't a narrative piece. It was, it was really an analytical piece because Brighton in many ways shows up in the data as a team that plays like top teams. And obviously they're not a top team. They're not a rich team. But stylistically, they have a lot of things in common with those teams and they seem to be moving more that way, you know, in their second year under, or are they in their third year under Potter now?

You know, each season under Grand Potter, they've gotten more like a top team. And, you know, the results have started to pick up for them a little bit this season as well. But I don't care about the results. Obviously, they have some problems in the final third. They don't have, you know, super expensive players to compete in primarily. But just through coaching, they've been able to take this group of like pretty modest, modestly priced, modestly talented players, and play like a very good team

and punch above their weight as a result. And so what I'm curious about is like, okay, what have they figured out that every other team couldn't be doing? Why are other small clubs not playing like top teams in this way? And is it really just that they're kind of ahead of the tactical trend in figuring out how to implement some of these things? So part of it is database, like just trying to measure unique behaviors

that Brighton shares with good teams, not to show that Brighton is good, but try to try to capture in more detail how they're good and what specifically it is they're doing that's good. And then the other piece is a lot of video work to kind of back this up and explain what these metrics mean on the field and how Brighton plays. To me, it's a really interesting story.

I love anything that kind of tries to break the relationship between money and performance and soccer. But also I just love trying to understand how positional play teams work and, you know, I think any USMNT fan probably shares that interest right now. So they lost to Burnley who, correct me if I'm wrong, but Burnley's like sort of the opposite of that? They are very much the opposite of that.

And they've also been successful at hanging around the Premier League for a long time by playing, you know, old-fashioned football. And yeah, I think that that game in particular didn't mean anything dramatic, you know, Brighton played fairly well and whatever. They also outplayed Manchester United at Old Trafford for the first half of the week before. So they're still a very good team, but they've had a bad run results. Well, hopefully they get a win this weekend so then you could run the piece. There you go.

Who are they playing this weekend? I have to sit on it until they get a big result. And the next really big game that they play is Liverpool in a few weeks. So if they could beat them then, everybody can finally read this story. Oh, man. What a tough life being that, being tied to the results like that, especially for somebody who doesn't care about goals and points, you know. It's brutal.

Readers care, though. Yeah. And if you publish an article praising a team for playing well when they've lost, oh man, nobody wants to hear it. Yeah. I did notice scrolling through the comments on that piece you did with the, you know, breaking down the possession in the different zones. A bunch of people are like Grand Potter is, uh,

needs to take over Man City after Guardiola leaves. It would be fascinating to watch him, coach, you know, really top group of players to see what he could do with it. Let me go, I don't want to take the rest of your night, but let me go back to something we discussed a long time ago. Is Josh Sargent going to turn the corner and become like a legitimate top five league striker? I mean, like everybody else, I hope so. What I've always told you about Sargent is that he's got a lot of talent,

but he doesn't do striker things. and, you know, he's doing even fewer striker things as a winner this season. So I don't know if he's ever going to develop those striker parts of his game that we're missing. But, you know, he's a talented guy, and I'm sure that he can find some rule that could help the national team in the long run, even if it's not at the number nine position. While I have you here, who would you start at striker for the national team against Panama? Okay. next month that basically with the World Cup qualification on the line at home yeah I mean you make it sound like it's dramatic like it's suspenseful whether or not we're going to

qualify for the World Cup and it's really not but it still is dramatic it's still is suspenseful dude okay all right okay what are the percentage odds right now like 97% or something like that if we don't get a result at the Azteca and Costa Rica beats Canada at home, which is totally feasible, then we have to beat Panama. As long as we get three points against, we have to be Panama and then we have to get a result, maybe you have to get a result against Costa Rica.

Anyway, it's not decided yet, that's for sure. And we're assuming that all the strikers are healthy in this next window. Yes, yeah. Okay. I think I would probably start Ferreira in that game. and keep a couple options on the bench. But I think that when we're playing at home,

we tend to possess the ball a lot. We tend to move a little bit slower for his linking plays, movement, and his kind of creation in addition to his boxwork would be useful in that situation. And then, you know, I like Giaz's artist late in the game, get him on and exploit space and do the things that Gassi does best. It's a great answer. It shows that you are paying attention.

What about, let me ask two questions about your, about your professional work, your production function. When do you do most of your writing? When do I do most of my writing? Like when in the day? I write best in the mornings, but because a lot of my stuff, you know, involves data work that takes up a lot of my day. I also do that stuff best in the mornings. So frequently I'll wind up coding in the morning and then writing until late at the night. were however long it takes.

So you're kind of like whenever, whenever, you'll write whenever. Yeah, yeah. I mean, if it was up to me and I had a nice steady regular schedule, I would write from 6 to 10 and then take a break and do my meetings and code and whatever. But yeah, it's really just kind of playing as it comes. And how do you have, I mean, you seem pretty good at social media. Most of your tweets get like a ton of retweets and likes. It seems like very different from me.

But how do you? do you all you got to do is do more songs man how do you avoid how do you avoid wasting time on there uh you know some days i'm successful and some days i'm not um i hope that even when i'm wasting time on social media i'm not wasting time completely because i learn a lot of interesting new things i find things to read that can make me better at my job so it's not a pure waste but right yeah you know i try to stay off of twitter and i don't have any other social media you know when when my work projects are really engrossing, which is most of the time, then I don't have that much more problem because I would rather be working on my stuff than arguing with 14-year-olds on Twitter.

All right. Can I ask you about stats for a little bit? Yeah. Yeah, so you were asking me about stats and what's my favorite metric and what do I use for, all that kind of stuff. But I'm curious as a fan who, you know, you're somewhat interested in statistics. You at least are aware of them. but it's not the main way that you engage with the game.

What does soccer stats mean to you? How do you think that they're useful? You know, when do you care about them and when do you not care about them? I mean, as a novice, I do think it's useful to assign a percentage chance of this shot going in to every shot. So I always look when I'm,

I spend a lot of time in Y Scout because I watch a lot of clips. And I'm always looking at, at like what does why Scott think was the probability of this shot going in. If I see anything over point two, I'm like, ooh, I got to watch that. Okay, okay. So you're looking, you just see like individual shot XG numbers. You're like, I'm curious how a shot, you know, is high probability,

what they did to create this high quality chance. Yeah, I have a, yeah, so, okay, I guess that gets to the answer to my question, which is like, I see people say something, this player is good because of X stat, or this player is bad because of X-stat. And I just immediately dismiss it. Like, I don't, I don't, like, I don't, I don't, I can't engage with the game that way. And, um, so I'm, so I'm definitely interested in like G plus and expected threat as concepts.

Those concepts make a lot more sense to me. But you say somebody had a certain number of progressive carries or a certain number of interceptions in a game. Well, interceptions actually, I kind of, that means a little something to me. But, you know, a lot of times. stats are used as a cudgel. This is not just a soccer problem. It's like all across society.

And they're used in bad faith. And it's too much to wade through. I just need to see it. I just need to watch the action on the field. So what is it the kind of signals to you that somebody is using stats in good faith and an interesting way that you should listen to? I don't know.

It's like a lot of network effects. So I generally believe, I generally believe that you're using stats in good faith. I usually shake those. But honestly, I know people who are using stats in certain ways. I mean, you know, Greg's much more of a stats head than I am. He's much more interested in this stuff and much more prone to use it in a discussion. And Greg's a meticulous guy who's generally not operating in bad faith, generally.

No, he's not. So, so if there, so, you know, Carlin Carpenter, I have some interaction with him. I see him using stats. I'm generally like, yeah, that's, I'll take that. But just like somebody I don't know throwing a bunch of numbers around, I'm just not going to pay that much attention to it, you know. I'm really pretty sympathetic to the whole Bruce Lerina line about how stats are for people who don't know how to analyze the game. because a lot of times they are. But a lot of times they're also just for people

who didn't have time to analyze the game because there's so much soccer going on. And I think that it's, I much prefer that somebody just tell me what stats they're using to kind of get some feel for games that they didn't have time to watch, then just give me the kind of received opinion on that game that they didn't watch.

And I remember what soccer discourse was like before we had stats, before we had at least some kind of empirical launching board for our conversations. And so much of it was just like reputation, like you heard from a friend who watched this other game that this guy played well. And then like now you're, he's the greatest player ever. And like, you just can't have conversations that way, right? Yeah. Yeah. And I, I'm, I am interested in these efforts to make it more efficient to like sort of understand a players effects on a game, a player's effect in a game or a player's likely effect in a game that they haven't played in yet. I'm interested in that. But people often ask me, like,

how do you keep track of all these players? And if you think about it, being a very, very focused USM&T podcast does give us a little bit of an advantage, you know? Like, we don't have to track that many players, really. Like, not compared to you. You got to know what's on with Brighton and Liverpool. And, I mean, it seems like the Premier League's your main, your main focus, but you're, you're paying attention to other soccer in other countries too. And, like, that's a lot. It is.

I'm not paying, I'm not paying attention to, like, La Liga or even the Premier League. Like, who's better? Christian Roldon or Luca Delo-a-Torre. That's a much simpler question to answer from just watching the footage than, like, you know, whole league questions and stuff. Yeah, you're watching a much smaller set of games and you're focusing on a particular player, so it's easier to get a really informed opinion on how that player played. But even so, you know, like I'm making a data tracker right now for Jeff Voidder at The Athletic,

and we've got like 140 names on there of guys who are kind of interesting to the USM&T discourse. And like you can't watch 140 players games. You can't even watch, you know, a quarter of that all their games all the time and have a family and have a job and have a life, which I'm not saying that everybody on Yosemanty Twitter has all those things, but it's, you know, it's very time-consuming, even for people who watch soccer for a living, who are scouts or coaches or journalists, whatever, like they don't have time to watch all these games and give you really detailed assessments. And so stats at least give you a starting point for any kind of whether you're doing player assessments or, you know, a plurrent analysis or just writing about trends in the season.

stats kind of get you there faster. And they also enable you to step back and ask big picture questions that it just really wouldn't be feasible to ask with video analysis. But as long as you're only looking at certain players and your only question is, is this player good? And how is this player going to work for the national team? Then yeah, I would definitely stick to video and probably not bother with stats unless you don't have time to watch all the games.

Yeah. I mean, I say all that about like how I don't, when people throw numbers at me, I don't really pay that much attention to it. With the caveat that I do want to be, I do want to get better at this stuff. And, you know, learn to, learn to utilize what is available to me in a more effective way.

Yeah. I think another thing that the data is really useful for is that while watching a game is probably a lot more useful than looking at numbers about that one game, 13 months later, you know, 19 months later, when you're trying to remember what happened in that game. Your memory is not going to be nearly as good as that data still is. And so data also has much longer memory as well as the wider reach. Yeah, that makes a lot of sense because I can't remember anything that happened 19 months ago.

Yeah, like I have some general impressions about who I thought was kind of playing well back then. But I might not even be remembering the right player in the right month, you know, let alone what happened in a game. 19 months ago was like 1992, right? At least. All right. Well, I'd love to keep chatting. What else can we talk about?

I don't know, dude. I want to, actually, I just want to run my whole interview back by you. I want to hear your thoughts on the Burnthalter tenure and where you're at these days. Because I went back and I listened to some podcasts that you guys did and like, I guess probably late-ish 2019. Oh, man. where you guys were you guys were super down on the whole per holter experiment at that time and a lot has changed since then and I'm just curious kind of how you're feeling about the project and how your thoughts about it.

I think the way you put it where like we haven't gotten what we were, you know, I'm doing air quotes promised in terms of disorganizing the opponent with the ball. I think that's still the main thing. I mean, that is the best summation of it. for me. And I don't see us being dangerous enough, regularly enough. I do blame Pulisic a little bit for that. And like Raina's being injured is a problem. I think Tim Wea has been great. Greg Burhalter had to be dragged, kicking and screaming to start Tim Wea in a game. Yeah, I definitely have questions about some of his, uh,

an ID but you know I do think that right now we generally play the best players when they're available more or less yeah it's like you know every every national team coach has some quirks but he like I feel like he's mostly gotten there yeah I think that's I think that's right and

there's some there's some disputes on the margins but like the starting 11 is I think pretty clear and it's that's the starting 11 that he starts yeah um like when they do fan polls

and crowdsourced 11s and all that, like you pretty much get the same thing that is our best 11 that we play in a big game. Yeah, there's some people who would say, like, we have to have John Brooks in camps, and, like, that's the big, that's the... I would say that, yes. Yeah.

And that's, like, the big, that's the big quibble at this point with the, with any lineup that Burrhalter puts out there. But even that, I don't feel that. The Brooks thing is super weird? Like, do you think that there's, is there some kind of beef there beyond just Burrhalter didn't think he was playing there?

I assume there is. yeah. I mean, like a, maybe just like a lack of buy-in or something. Yeah. But,

but I, yeah, I don't know. That's, that's my assumption. He also is slow. He's a, he's a slow human being.

And, you know, yeah, that's what Miles is there for. Right, right. Right. I mean, yeah,

there's definitely an argument to be made. He should be, he should be in these camps and he should be starting next to Robinson or, I mean, even Zimmerman would

be okay next to Brooks, I think. So that's the, I think that's the big dispute on, in the fan base and it's, I'm somewhat reconciled to it. I'll take us not getting a ball to a striker's feet from like 35 yards away, four or five times a game. I'll, I'll, I'll, I'll, I'll, I'll, I'll, I'll, I'll, I'll, I'll, I'll, not getting that. If in exchange, we also do not get a centerback getting, like, left for dead, 30 yards from

goal two or three times a game. And even if he has cover with Miles Robinson, John Brooks is, he's a defensive liability. I think that Brooks's contribution to possession game is a lot more than a couple Hollywood passes game, right? Like he really orchestrates the backline in a use of way, although he was also orchestrating that backline in our most embarrassing game, in my opinion, that game. against Canada where we just had the ball the entire time and couldn't do a damn thing with it. Yeah.

And then Brooks was terrible in that game. So, you know, I'm sympathetic, I guess, to Berthor, at least questioning his play, but he also has talents that nobody else in the pool. It's true. It's true. That are supposedly aligned with how we're trying to play. I don't know why Berhalter is still pretending that he's trying to play the same way that he used to because he's not. Is that what you think? Like, he's shifted what he's trying to do.

do. I do think that. Yeah, I think that the team plays fundamentally differently now. And I think that maybe conceptually in his head, it's a continuation. But I think that when you look at how the team actually plays on the field, absent whatever kind of structures, he thinks about it. And like, it just is a very different team. It's a team that is all about pressing and transition moments. And, you know, they do try to build up, but the buildup is so disconnected from the things that do that really matter that it's it's almost like a vestigial thing left over from the 2018 era it doesn't work i guess is the thing that you know i mean occasionally it works there's it there's yeah it does work sometimes occasional success up the right side through tim waya and um

you know moosa or mckenny whoever's combining with waya over there but yeah i mean i do think it's a problem that none of the midfield three are really built for that game and I think that's why we don't play that game anymore. All those guys have their talents but none of them is like your top choice if you're trying to play real position to play soccer. Do you think that maybe explains why explains Brawaters reluctance to give up on people like Jackson Ewell and um yeah yeah I mean he was definitely fighting a good fight there for a while right he was trying really hard to make this happen and when your best option is to play that kind of soccer are Jackson Eagle or like 34 year old Michael Bradley like it's just not feasible and you need to work with the best players that

you have available to you. It does bother me too that Pulisic, that Brawinter hasn't gotten Pulisic to sort of do what he wants him to do. I mean, maybe Brawolder wants him to drop behind the half line and pick up the ball and try to dribble eight guys. But I doubt it. And to me that looks like. you know,

insubordination almost or just like freelancing on an absolutely ridiculous level that like shouldn't be, shouldn't be happening. And so there's must be some disconnect there. Yeah, I think another one of those basics that it's good to remind us, remind ourselves of from time to time is that tactics don't spring from the coach's head. They spring from the player's head.

And all the coach can do is try to, you know, talk to players and show them things and get them thinking in similar ways that are going to work together. But out there on the field, the players are making the decisions. And if you listen to coaches talk, they're constantly telling you, yeah, my players are not doing the things that I wish that they were doing because that's just not how this game works. You know, coaches don't have controllers. They don't have timeouts. They don't call plays. There's nothing that they can do once that whistle blows to really influence the game all that much. And so at the national team level, that's a problem because you don't get enough training time to really

implant those ideas and it's really up to the players to figure things out. Yeah, I just wish Pulisic were figuring things out a little bit more within whatever structure it is he's supposed to be playing in.

Yeah, but like what do you tell the guy you know? I'm sure that he knows that he's playing the way that he is and I'm sure that he knows you know, I've seen that he knows how to play in a different way. But I think that he's just put so much pressure on himself right now

that it's almost like a sports psychology issue more than it is a coaching problem. Because, yeah, I agree with you. There's no way that Berhalter wants him playing hero ball, dropping into the defensive half and trying to dribble it, guys. That's not what Burhalter's about. Yeah. I think just to keep answering your question about Berhalter, I think he has a real problem at Stryker.

He doesn't have a great option. I think I think you're right. Ferreira is the guy if we're going to dominate possession and we're going to try to link play in the other opponent's half. He is the guy at the moment. But it's been a journey and not a happy one on who can play striker for this team. You combine that with Pulisic sports psychology issues and Raina's injury.

It's been hard to score goals. it has I mean the striker thing again not burrhalter's fault at all he's been very flexible with the whole situation he's tried to kind of accommodate his team to totally different styles of striker to get the best player in the team

and I think that's that's probably really tough for him because again he's he's sacrificing his idea of how the team should play to just try to work with what he's got yeah I mean Ferreira Ferreira fits his idea of how to play the best right I think so

I mean Ferreira is definitely closest to sort of what he talked about coming into the job and I think that Ferreira fits like the actual current version of this team pretty well the question is is for good enough and I don't know we'll see Yeah

well once we qualify for the World Cup then I love that this is a suspenseful thing for you because it's so not It totally is, dude. I mean, was it suspenseful last time? Yeah, no, yeah, you got me. I didn't think that it was ever in doubt last time either.

We're probably going to qualify. And I mean, if we don't qualify. Are you coming to Orlando? You've got to come to Orlando. Wait, what? What's happening to Orlando? The Panama game, March 27th. Oh, yeah, no, dude.

I'm not traveling for games. When they come back to Austin, I'll go to the game. Okay. Think about it. Give it some thought. Don't be so hasty with a decision like that. Those tickets are so expensive, dude. There's no way I can afford to get a U.S.

I'm into a game. They are kind of expensive, yeah. It's ridiculous. Face value, you could still get one for like $100. That's ridiculous. Yeah. To watch a conca cash. Oh, yeah, we should talk about that.

We should talk about that before you go. Because you've been a strong believer in play the games in the biggest stadium you can. and yeah we've talked about that yeah and and also the stadium that's not like freezing people's toes off and shit that was a bad idea for sure yeah that was embarrassing um yeah man I'm the biggest problem that soccer has in this country is well the biggest problem that soccer has in this country is that people don't like soccer but that part's changing so now the biggest the biggest problem the soccer has in this country is that there are so many different like types of fandom that are competing for people's attention.

And a big one right now is like immigrants' home country teams are like still near and dear to their hearts. And that's not going to change if we're constantly trying to hold games in like, you know, Columbus, Ohio, because it has the lowest Latin population of any city that we can find. I think that that's really like discriminatory and embarrassing for us as a country, as a program. also like for a while they were being honest about it and now it's like oh we went to Minnesota because what we love the weather there in February or some shit like they kind

of realized that that was not something they should be talking about in public anyway I'm a big believer that if the USMNT is ever going to win like hearts and minds in the country and have a bunch of fans who will willingly fill the stadium so that we don't have to worry about opposing team fans you know saying mean things for our players during the game, we need to be playing for the biggest possible audiences and audiences that may have other allegiances or other interests and like, great, bring them all in and let's play well and let's like represent the United States that is the country that we wanted to be and not this like weird, isolated kind of rich white people support.

Yeah. I'm sympathetic to it. I'm sympathetic to what you're saying. I'm especially once we qualify for the World Cup, like play all the games. games in the Rose Bowl for all I care, you know, every single one. The Rose Bowl, the, like a big stadium in Miami, big stadiums around New York. We do, we do need to, we do need to grow the game. I had Jay Hernandez, a Salvador American on the podcast last week. And, you know, he's like a, you probably, you may see him on Twitter. He's a, he's kind of a key player in the USMNT discourse.

And, yeah, he said the same thing. We need to show these communities that they are. are that the USM&T is for them too, you know? Yeah. Yeah. I would say, I mean, I guess I would say, you know, they played a game in Austin against Jamaica. Yep. Not a lot of Jamaicans in Austin. Right.

But not a lot of, you know, still a lot of white people in the stadium. You know, and Austin has a big Hispanic population in the area. Yep. I guess they could have put it in a bigger stadium. So why are they playing? Yeah, why are they playing in Q2 Stadium? Why aren't they playing in the Longhorn Stadium? right?

Why I think selling those tickets for $5 a pop just trying to get everybody into the stadium to care about the United States national team. Yeah. Because it wasn't like the risk was that a bunch of Jamaicans would show up and fill the Longheart's football stadium. Like, I'm sure that there are some who,

I did meet some at the Austin game who were a lot of fun. But like, that's not the worry. The worry is clearly about like making money and keeping this like exclusive so that they can charge $400 a seat. It's depressing, man. It's supposed to represent

the country. Yeah. Which I guess it does a little too well. All right. All right. All right. On that really happy note. John, thank you.

Thank you for doing this. Appreciate your time. Seriously. Yeah. No, it was fun. I feel like we didn't even start to scratch the surface of all the stats off, but I hope that I at least answer your question. I mean, for you not scratching the surface is like is a deep jab for most people.

For me. Okay, well, then I hope I didn't bore it. No, no, no, no, no, no. No, no. All right. But seriously, I appreciate it. Cool. And everybody for listening.

We'll see you.