Quick, who is the most valuable player in your favorite sport? If you’re a football fan, chances are you’re thinking of a quarterback and not a defensive lineman; in basketball, it’s probably a high-volume shooter, not a rim protector; in baseball, it’s often a hitter, rarely a pitcher. Same thing in football, where the glory goes to the players who score goals, not those who prevent them. The only goalkeeper to have won the Ballon d’Or, football’s most prestigious award, was Dynamo Moscow’s Lev Yashin. He played his World Cup matches for the Soviet Unionif that gives you an idea of how long it’s been since someone stopped shooting.
In a sport where a goal saved is about as good as a goal scored, it’s hard to say why the pitch slopes so steeply towards the shooters. Part of that could be because we haven’t had a good way to measure what goalkeepers are doing. “A guy who gets a lot of clean sheets can just have a good defense,” said Brian Bilello, president of Major League Soccer’s New England Revolution. Save percentage is also not an ideal metric, as a goalkeeper who saves a higher share of his shots on target may face easier shots from distance. “I would say goalkeepers are poorly rated,” Bilello said.
The Revs have good reason to want to change that. Their goalie, a 27-year-old named Matt Turner, is America’s best shooter. Even though he started playing relatively lateTurner has been head and shoulders above everyone since reaching MLS in 2018. In this month’s Gold Cup, he could play his first competitive game for his country. Journalists “write article after article about him and how good he is, the best shooter in league history and all that,” United States men’s national team manager Gregg Berhalter noted Last week. “But to some extent, I agree with them.”
How do we know how good Turner is? The case of statistics begins with the expected goals. Usually, xG measures the probability of a shot scoring based on the distance, angle, part of the body the shot was taken from, and other contexts at the time the shooter hits the ball. But there is another version called expected goals after shooting, which measures the same probability at the moment the shot crosses the goal line – or would cross it if the goalkeeper did not stop it. By including ball placement information, PSxG can give us an idea of how many goals a keeper conceded compared to what an average stopper would have against the same shots.
Who is the most chaotic fictional football manager? | FiveThirtyEight
In Turner’s case, the difference is huge. According to American football analysis, he has registered more than 22 goals more than expected since 2018. The second-place goalkeeper has less than 15. Turner’s goals conceded divided by the xG ratio is 0.82, or about four goals allowed for every five that an average keeper would have conceded. No active MLS goalkeepers come close. Shot save rates take a few seasons to stabilize, but after facing almost 400 shots on target, Turner is the best ever.
“I think having positive expected goals [difference] is huge for the team,” Turner said. “Your defense can’t stop every shot for you.” He attributes his shooting stoppage to a lot of different things, including playing dodgeball and baseball as a kid. “I think it’s the athleticism, the positioning and most importantly the timing,” he said. “People don’t make as many saves when they’re not ready. But it’s also about understanding the patterns, the timing of the set-up, what it looks like when someone’s about to shoot versus playing a dribble or passing behind.
But saving shots is only part of a goalkeeper’s job, which also includes organizing the defense, passing from behind and recovering balls in what the goalkeeping coach of the Revs Kevin Hitchcock calls the ‘living room’ the keeper. Hitchcock picked Turner as a starter because of his athleticism, but they’ve spent the past few years working on the rest of his game, trying to develop him into a modern goalkeeper with crisp passing and aggressive defense.
American Soccer Analysis also has a way to measure these other skills. The latest trend in football analysis is a type of model, sometimes called “expected possession value”, which estimates how much each touch alters a team’s chances of scoring and conceding. ASA’s version, dubbed “added goals“, or g+ for short, recently expanded to include goalkeepers. Every time a keeper touches the ball – whether it’s a pass, a punch, a sweeping tackle, a claimed cross or a save – g+ assigns a value of positive or negative goal depending on the probable effect of this action on the score.
Turner’s unstoppable g+ is pretty good, but the important thing here is that the differences between goalies are minimal: compared to an average goalie, the leader over the past four seasons has added around 0.03 goals every game with all keys other than saves. Turner’s shot save, which g+ accounts for by subtracting goals from expected goals after the shot, added more than 10 times that amount. As far as the ASA can measure right now, denying shots is by far the biggest difference in a keeper’s skill set.
By valuing each type of contact in goal fractions, g+ makes it possible to compare players across positions. In the three and a half years since entering the league, Turner has amassed 25.6 more total goals than an average player in his position. The three offensive players who won league MVP awards in that era, Alejandro Pozuelo, Carlos Vela and Josef Martínez, are 20.2 g+ above average – combined.
By goals added, Turner was the MLS MVP
Top 10 MLS players since 2018 by total goals added vs. average player in position
Player | Position | Minutes | Goals added to position average |
---|---|---|---|
Matt Turner | Goalie | 7,929 | 25.6 |
Stefan Frei | Goalie | 9,275 | 17.9 |
carlos vela | Winger | 6,564 | 16.1 |
Zlatan Ibrahimovic | Striker | 5,193 | 15.3 |
Alberth Elis | Winger | 5,183 | 13.7 |
Sebastien Blanco | Winger | 6,316 | 13.1 |
Jack Elliot | rear center | 7,212 | 9.9 |
Anton Tinnerholm | Back | 8,680 | 9.6 |
Darwin Quintero | Attacking midfielder | 6,949 | 8.4 |
Miguel Almirón | Attacking midfielder | 3,010 | 7.7 |
Like most advanced football analytics, goalkeeper stats are still in their infancy. Added goals are based on event data, which can’t see what’s happening outside of the ball. It is therefore difficult to judge the decision-making of a goalkeeper in the preparation or if he should have come for a cross. Future measures could punish goalkeepers for the danger of an unclaimed corner kick, for example, or better capture the value of the build at the back.
“I don’t really want to say, ‘Who cares how many short passes you make if you’re not moving the ball forward? “Said Turner, whose average passing distance this year is the second longest among goalkeepers in MLS. “The game is evolving, and I feel like that’s an area where people want to see the goalkeepers start taking a little more risk. But I also think short passing leads to more goals against, because you make a mistake or pass the ball to a teammate who makes a mistake.
If he is to fight his way into the depth chart for the United States in the World Cup qualifiers this fall, Turner will have to convince coaches that he is a complete modern goalkeeper – or that his shot-stopping is so valuable than average passing or sweeping is not a problem. Berhalter may come to this idea. Before Turner’s first international friendly last winter, the coach ruled him out. “He was like, ‘Look, if you don’t feel comfortable taking a risk, don’t take a risk. You are here for a reason,” Turner recalled. “‘So go ahead and play your game.'”
Discover our latest football predictions.