Moneyballing the World Cup

And how you can make the roster.

Source: Elon Musk, self doxxing.

As an American, I have a civil duty to vote, to serve on juries of my peers, to pay my taxes, and to say that soccer is boring. Just as I’m expected to register for the draft when I turned eighteen, I’m equally obligated, when someone says soccer is a thrilling and beautiful game, to remind them that nobody ever scores, that real sports don’t routinely end in ties, and that most matches feel like an hour and a half of high-stakes keep away, with occasional breaks for pathetic bouts of bad acting.¹ 

But for forty minutes during Sunday’s World Cup final—from the moment Kylian Mbappé converted his first penalty, through his sidewinding one-timer, Argentina’s apparent game-winner, Mbappé’s third counterpunch, several gasping chances, and a miraculous fire sale of a save in the match’s final minutes—soccer was not boring. It was a worldwide panic attack, frenzied and desperate, kill or be killed, a boxing match between two heavyweights who’d given up their guards to throw wide-open haymakers.

Both teams landed punches, but nobody went down. The game went to penalty kicks. Normally, for a football philistine such as myself, this is what I want. After two hours of slow and subtle “beauty”—i.e, back passes to the goalkeeper and 22 of the world’s best athletes trying to rope-a-dope each other into a mistake—penalty kicks are a welcome bit of blunt action. But on Sunday, for the first time in my American life, I felt robbed by soccer’s cruel and contrived ending. I wanted more of what I had just seen, not a gimmick of chance. 

Prior to the teams taking their kicks, the announcers on the Fox broadcast emphatically told us, as they seemed to before every shootout, that the penalty tiebreaker wasn't an arbitrary contest of odds and evens. Every player has their tendencies and tells; there is athleticism and strategy on both sides of the ball; it's not a game of craps but a game of chicken, with all the world watching. 

True as this may be, it misses the point. The problem with penalty kicks isn't their randomness; it's that they have next to nothing to do with the sport that everyone just played for 120 minutes. It’s replacing extra innings with Texas hold ’em; it’s resolving twelve rounds of boxing with Super Smash Brothers. It’s skillful—the skills are just different.

For France, Argentina, and fans of soccer as its founders meant for it to be played,² this is a tragedy. But for a brave manager looking to Moneyball their way to a title—and for any of us, looking for a backdoor onto a World Cup roster—it might be an opportunity. 

You have one job

In the first stage of the World Cup, 32 teams play three games each. Those games, which can end in ties and therefore have no penalty kicks, cut the field in half. The remaining sixteen teams then play a single-elimination tournament, in which there can be no ties. In these games, known as the knockout stage, teams play soccer for ninety minutes. If the game is tied after that, they play another thirty minutes of extra time. If it's still tied, each team takes five penalty kicks.

Teams can choose who takes their kicks, but they have to pick players who are on the field when extra time expires. For this reason, teams will sometimes sneak in a player or two right at the end of the extra thirty minutes. These late subs aren’t exactly penalty kick specialists, but more akin to defensive replacements—a better-than-average kicker taking the place of a worse-than-average one.³

Which raises a question: Given the staggering importance of those final five kicks, what if they were specialists? What if they were players who didn’t do anything else other than practice pounding the upper nineties from twelve yards out? They don’t take 1,000 shots in a summer; they make 1,000 shots a day. They don’t need to be good defenders, or have good touch, or even understand the rules of soccer. They just need to be snipers, from a set distance. They are players with a single job: They show up to practice, stretch the leg attached to their one lethal foot, and paint the inside of the goalposts and crossbar for eight hours. 

The answer to that question—what if teams put specialists on their roster?—is that they’d be about two percent more likely to win the World Cup.

A ghost runner on second

Thirty-five World Cup games have been decided by penalty kicks. In those games, players have taken 320 shots and made 222 of them, or 69.4 percent.⁴ Because teams often put their best kickers first, the fourth and fifth kickers typically convert at a lower rate of around 65 percent.

This suggests that specialists could give teams a considerable edge in a shootout. The diagram below shows the hypothetical conversion rates for different penalty lineups, based on the historical rates from prior World Cups. As it illustrates, not only would specialists improve on the average kicker, they’d also bump teams’ worst kickers out of the penalty lineup, and into the sudden death kicks. 

Because penalty shootouts are so structured, the exact values of these different lineups are easy to quantify. If both teams have a lineup without a specialist, the final outcome is a coin flip, with each team winning half the time. If a team adds one specialist, their chances of winning increases to 59 percent. Add two, and it goes up to 69 percent. 

Granted, these win rates only apply to games that go to penalty kicks. But shootouts are becoming increasingly common in the World Cup. Since 1982, twenty percent of all knockout stage games ended in penalties. In the last three World Cups, that figure exceeded 25 percent, with this year’s tournament setting a record at 31 percent.

If this trend holds steady at one in four games going to penalties, adding two specialists to shootout lineup increases a team’s chance of winning a knockout stage game by five percent—the 25 percent chance of the game going to the tiebreaker times the nineteen percent increase in their probability of winning if it does.⁵ For comparison, a win rate increase of five percent is roughly the equivalent to a baseball team hitting a leadoff double in the first inning. Add two specialists, get a ghost runner on second at the start of the game.

Backing up even further, since half of the teams in the World Cup make the knockout stage, teams that add two specialists to their roster at the start of the tournament would likely be a couple percentage points more likely to win the entire tournament. To make another mismatched comparison, an American football team that holds their opponent to a three-and-out on their opening drive improves their win rate by about two percent as well. In other words, a simple roster adjustment in the World Cup is like an NFL team getting the ball at the start of both halves—an advantage that no team would turn down.

What’s more, there’s no reason not to do it.

A free lunch

World Cup teams are allowed to have 26 players on their roster, three of which have to be goalkeepers. To win the World Cup, teams have to play seven games in just under a month. Given such a grueling schedule, teams may want to load their rosters with as many pairs of fresh legs as they can. Relieving a midfielder who’s on their fortieth mile of the tournament might be more important than having a one-trick marksman available for a hypothetical tiebreaker. 

If that’s true, however, it’s not how teams use their rosters. In this year's World Cup—played in Qatar, which was hot, humid, and presumably more physically taxing than most World Cups—teams had, on average, 2.9 players on their roster who never played a single minute. They also had an average of 2.4 players who saw less than thirty minutes of total action in the entire tournament. Though these numbers are lower for teams that made the knockout stage, very few teams relied heavily on the players at the end of their benches. 

Moreover, in prior World Cups, teams were only allowed to carry 23 players, not 26, so teams clearly don’t need fifteen backups. 

The other potential obstacle to carrying penalty kick specialists is soccer’s substitution limit. Teams are only allowed to make six substitutions during a game—up to five during regular time, plus a bonus substitution during extra time.⁶ If teams need all six prior to the match’s final few minutes, they wouldn’t be able to get their specialists in before the shootout.

This shouldn’t be a problem either. First, because teams are given the extra substitution at the beginning of the overtime period, they'll always have at least one late change in their back pocket if they need it. Second, as they did with the roster, FIFA recently increased the number of substitutions available during regular time from three to five, so teams have been accustomed to playing with only three subs for decades. And finally, just as teams rarely reach the end of their bench, they rarely use all of their substitutions during games’ most meaningful minutes. Of the ten teams that participated in the five matches that went to penalty kicks this year, four had an unused sub, and all ten made substitutions in the last fifteen minutes of extra time, with several teams making multiple changes. 

All of this points to a fairly clear conclusion. Teams could carry two penalty kick specialists, at essentially no cost to how they manage their teams. And those specialists could win games—either by dramatically improving their team’s odds in the shootout, or by inducing their opponents to take more chances in extra time. 

So, to all the World Cup managers: Don’t be cowards. Don’t look for traditional talent, and fill out your rosters with safe picks of “world-class” players that’ll get six minutes of garbage time in your eventual quarterfinal collapse. Go hunting for the sharpshooting savants who can’t run, can’t pass, can’t dribble, and can’t even hold their own on a pitch full of high schoolers. But give them a ball, a white spot 36 feet from the net, and a bullseye to hit—and a World Cup to win—and they don’t miss. 

And to the rest of us: We’ve got four years to get good at this.

---

Should specialists...play?

As an extended footnote, it’s also worth asking if penalty kick specialists should actually get playing time. In what strikes me as a bizarre quirk in soccer’s rules, when teams are awarded an in-game penalty kick, the player who was fouled doesn’t have to take the kick; anyone who's on field can attempt the shot.⁷ If a specialist was playing, they could take every one of these kicks. 

The math on this strategy, however, is comically lopsided. There’s an in-game penalty kick in about one out of every three World Cup games, so individual teams are awarded a kick once every six games. If the best penalty takers convert at 75 percent, and specialists convert at 95 percent, playing a specialist is worth 0.2 goals per penalty, and 0.03 goals per game. That's not much, and if anything is overstated—a team with a terrible player on the field would likely have a harder than average time drawing penalties.

The cost, by contrast, is huge. Playing with a specialist, who would be a small speedbump who briefly inconveniences the real athletes, would be similar to playing with only ten men—which actually happens, when players get sent off after getting a red card. According to one analysis, a red card costs a team between 0.6 and 1.2 goals in a game, far, far more than the 0.03 goal benefit. So add specialists to the roster, but keep them safely on the sidelines, warming up their one useful leg, away from the sport they’re dangerously unprepared to play.

1. Greg Paulus: Right school, wrong sport.

2. When we talk about soccer, why don’t we ever ask if we’re playing it the way the Mesoamericans intended? Why don’t we talk about if baseball’s rule changes are what Abner Doubleday would’ve wanted? Why are countries’ founders revered, and sports’ founders so forgotten?

3. This year, five games went to penalty kicks. Seventeen players were brought in the final fifteen minutes of these games; of those, only seven took penalty kicks. 

4. This is a good bit lower than the 85 percent conversion rate that players average in other elite leagues, presumably because of the enormous pressure that comes with World Cup kicks.

5. This estimate is likely a bit low, because teams that know they have an advantage in the shootout could play for that result.

6. Teams that don’t use all five subs during regular time don’t lose them; unused subs roll over to extra time.

7. If I were in charge of soccer, I’d change two rules. First, this one—make the player who was fouled take the shot. (If nothing else, that would keep penalty takers from having dramatially inflated goal totals. Imagine how many points Steph Curry would score if he took every foul shot for the Warriors.) And second, instead of going to extra time and then penalty kicks, soccer games should end like NHL games: Start taking players off the field. Or even better, just make the entire game eight-on-eight. It’d be the same sport, but more open, more frantic, and, mercifully, with more scoring.