Advanced Soccer Guide for American Sports Fans Ahead of 2026 World Cup

A dozen years ago, on a different website, I wrote an article titled How to Watch the World Cup Like a True Soccer Nerd . The idea was simple. It was summertime and the World Cup was around the corner and lots of Americans would be tuning in. Presumably, many of them would be sports nerds, familiar with advanced analytics in their preferred sports but not well versed in their soccer equivalents. The point was to give a crash course to catch those soccer neophytes up to speed. Well, now the 2026 World Cup is here, on U.S. (and Mexican and Canadian) soil no less, and we're back for round two.

A lot's changed since then, both in soccer, and with nerds, and with nerds in soccer. In America now, it's trivially easy (though certainly not cheap) to watch any of the biggest soccer leagues in the world, men's or women's. Last season, we here at Paramount (never ever forget that you can subscribe at Paramount+ to get absolutely all this soccer goodness) broadcast the Champions League, Europa League, Conference League, Serie A, Carabao Cup, Championship, League 1, League 2, NWSL, USL, Liga MX, SPFL and more. And we're just one broadcaster. It's 2026; if you want to find a soccer match, you can find a soccer match.

As soccer has grown up, so have soccer fans. A decade ago, dragging your sorry ass out of bed at 7:30 a.m. ET might have involved wrangling a hangover, possibly by stumbling down to one of the handful of bars open for supporters groups looking for a place to watch their favorite team. Now, for those same fans, it often involves wrangling a toddler. Soccer is the weekend soundtrack for parents everywhere just trying to get their baby to eat their apple sauce.

As with parents, so too with sports analytics. By 2014, analytics was a mainstay of American sports; the nerds had won the war and were fully integrated into teams across baseball, basketball, and increasingly the National Football League (we're just not gonna use the word "football" today, so as to avoid confusion, it's soccer and it's the NFL). Now, the nerds ARE the sports. In baseball, players in a slump will habitually get the numbers crunched to reassure themselves that they're still getting their process right. Everyone in basketball understands that three is more than two. Heck, even in football, the idea that you're supposed to go for it on fourth down a lot more than coaches had been has sunk in.

Soccer is behind, but that doesn't mean there hasn't been progress. Before the 2014 World Cup, expected goals (xG) was a relatively new stat that bloggers liked to throw around, and the smartest teams in the world (along with folks in the sports betting space) had at their fingertips. It was like a little bit of secret knowledge shared with a select few straight from mystifying computer programs. Now, it's everywhere. Every team has it, broadcasters cite it with regularity (sometimes, they don't even bother to say "if you're into that sort of anything" anymore), and every annoying fan on social media uses it to argue in the dumbest way possible about why their team should or shouldn't have won. It's mainstream.

So, it doesn't really make sense to put together another breakdown covering the basics. More people know the basics now. There are plenty of places to learn about the offside rule, why players flop to draw penalties and what an overlap is. Heck, you can go back and read that original piece if you want. If that article was soccer nerdom 101, let's make this belated sequel 201, and cover some slightly more advanced concepts. You're ready. I believe in you.

1. OK, but really, what is Expected Goals (xG)?

When a new stat breaks through into the public consciousness, it can often be a double-edged sword. Sure, more people are using the number you devoted your life to rigorously studying, calibrating, building, and releasing into the world like a delicate baby dove, but, OH MY GOD, have you seen the crimes against man, nature and statistics they're using it for? I suppose, in that way, it fits perfectly on social media.

At its most basic level, expected goals accounts for everything that goes into a shot and calculates how often, on average, that shot becomes a goal. This is an extremely noisy measure. Just because a shot from the middle of the box, taken with a player's foot after a cutback played from close to the endline around the edge of the 18-yard area might average getting scored 40% of the time, it doesn't mean that some specific instances couldn't be much higher and some much lower. The power of xG comes from its average. Those things average out very quickly and somewhere around the three- to five-game mark, a team's xG difference reliably gives more information about a team's future performance than their goal difference. That's what's empirically proven. The fancy statistic predicts future results better than the basic one in multi-game samples. It doesn't predict every single thing; it just predicts more than goals. That's why it exists.

But wait, you say that teams only play a maximum of eight games, and only four of 48 will get there. And two of those are in the third-place game ... so does that even count? I am a good, upstanding statistics citizen and don't wish to commit stats crime; what good are expected goals in such a small sample? I'm glad you asked.

Because we know expected goal averages work over the long term, we can do all sorts of fun descriptive stuff with the statistic in the short term, even for a single game. Let's look at the biggest club game of the season as an example, the Champions League final. That game, played on May 30 (and airing on CBS and Paramount+, a running theme), was between Arsenal and Paris Saint-Germain. It was a match tied 1-1 after 90 minutes, and after 120 minutes it was decided in a penalty shootout. I'm going to drop the xG race chart below, which shows that PSG "won" the xG battle 1.75 to 0.50.

What does xG tell us about this match? Were you to use it as a blunt instrument, which is to say, exactly as fans the world over are most likely to use it, it would simply tell you that PSG were a lot better, should have won the game in regulation and that Arsenal were "fortunate" to get to penalties. This is what we call, to use a highly technical term, bad. The stat contains lots of useful information at a per-game level, but it doesn't contain that kind of ground truth.

Here's what we know. PSG took 21 shots for 1.75 xG, but that stat comes with a giant caveat: They won a penalty. Penalties are fairly random events. I know you're going to want to argue with this, but again, we've got lots of empirical data here, and we know that xG is actually better at predicting future results when you strip away the randomness of penalties. It's not satisfying, but it's true. So what we actually see is that while PSG had an xG of 1.75, 0.79 of it came from that one penalty shot. Everything else was 0.96 xG from 20 shots. Further, we can see that Arsenal scored extremely early in the match. They spent about an hour with the lead. They didn't need to score to win the title; they just needed to keep PSG from scoring.

Let's sum all this up. If you analyzed that game without expected goals, you'd see a PSG that dominated; they outshot Arsenal 21-7, won a penalty, and were unfortunate not to win the game outright. If you analyze it poorly using xG, you'll get the same result. PSG won the xG battle 1.75-0.50; those Arsenal bums were lucky to make it to penalties. Actually using the statistics well presents a much more nuanced story. Arsenal scored early, defended excellently for an hour, conceding only about 0.25 xG before conceding a penalty. Then, in the remaining 30 minutes of regular time and an additional 30 minutes of extra time, they continued to defend well, but could not attack at all, even though the game was tied. The key insight is that while PSG had 20 non-penalty shots, they amounted to less than 1 xG. Arsenal may have faced a lot of pressure, but conceding less than 0.05 xG per shot means they gave up no good chances (average xG per shot is about 0.10). Adding xG into the mix tells a much more nuanced story. It's not that PSG were undeserving winners; it's just that they were deserving winners over an Arsenal team that also played very, very well defensively.

Even in single-game samples, xG is an important tool for showing, with greater granularity, what happened and why. It will be extremely useful at the World Cup for explaining the game. Just don't use it in any sentence that includes the word "end of (discussion)."

And there's one more point to add before you are all bored beyond expected tears. On a player-specific level, xG is also a very valuable tool. Look, everybody loves a banger. Goals fired in from 30 yards out, rocketing top shelf, or curling inch-perfect into the corner are the lifeblood of fandom. Seriously, just look at what United States men's national team left back Antonee Robinson did in the side's final tune-up match against Germany.

They're also rare and somewhat random. Everybody understands they don't happen often; that's what makes them amazing. At least everybody understands that until a player goes and scores a couple, and then the world starts suggesting that might keep happening with regularity. Don't believe me? Here's what Aston Villa and England midfielder Morgan Rogers said after scoring a couple of high-profile lasers in January.

"If you actually look at shots we're hitting, they're good options to take. We're not taking potluck shots. You can't always score the perfect goal, from a cross or a tap-in. There are different ways to score goals," he said.

And while I appreciate the sentiment, you always want your players to have confidence. Here's Rogers xG on a 10-game rolling average across the Premier League and Europa League.

The bangers look great, but you always come down to xG Earth eventually. The best players in the world can outrun xG by a bit, maybe even up to 20% or so. But the real money is in getting high-quality, high-xG chances. Though, to be clear, even those aren't a guarantee. At some point in the tournament, I promise you will see a tap-in, a sure thing goal that there's no way on god's green earth a player could miss. The xG value of that goal will not be 1.00; it will be somewhere between 0.95 and 0.99. You will be incredulous. How could that not be worth 1.00, you will ask. And the soccer gods will smile down on you and say four words.

Eric Maxim Choupa-Moting.

And now you know what xG is.

2. Formations -- Forget what you have been told

Last time I did this, I wrote a whole thing about how confusing formations were, except for one specific part: Are teams playing two or three center backs? Modern managers took that as a challenge. This is particularly relevant to the United States men's national team, where nobody is quite sure how many defenders they are supposed to be playing with. This isn't a bad thing.

In their World Cup tune-ups, U.S. coach Mauricio Pochettino trotted out a lineup that included five guys whose best position is somewhere across a back line in Antonee Robinson, Tim Ream, Mark McKenzie/Miles Robinson, Alex Freeman and Sergino Dest. Antonee Robinson is the guy who plays on the left and runs up and down a lot -- that's easy. Tim Ream, the U.S. captain, is 38 years old. He isn't running anywhere. He's a center back -- that's all easy. The same is true for the younger, more athletic McKenzie or Miles Robinson -- certainly sprier than Ream, but every bit as much center backs.

Where things get weird is with the last two guys: Dest and Freeman. Both of their most natural positions are right back, or even right wing back. In the ancient days of, like, 2014, 2004, 1984, or maybe even 1884, we'd have understood how this works -- playing two fullbacks on the same wing has been a defensive strategy. One plays as a winger in front of the other one, who plays as a full back. Two g

_Originally reported by [CBS Sports](https://www.cbssports.com/soccer/news/fifa-2026-world-cup-expected-goals-analytics-soccer-nerd/)._