The line is too high. The keeper’s too low. And nobody can find their assignment in time.
Watching New York City Football Club defend set pieces has been an anxiety-inducing experience from the beginning. In its first two seasons, the team posted some of the worst set piece goals–conceded numbers in American Soccer Analysis’s dataset. (Is anyone getting nostalgic thinking back on the Jason Hernandez and Josh Saunders era? I didn’t think so.) Things got better after a defensive rebuild brought in Maxime Chanot and Alexander Callens: in 2017 and 2018, goals and expected goals against off set pieces dropped to respectable levels, with a correlated decline in cardiac arrests per 96′ in Yankee Stadium.
Nevertheless, improving NYCFC’s set piece defense remained a focus this offseason. The problem areas were evident, especially when it came to defending against free kicks played into the box from outside shooting range. If the first six weeks of 2019 are any indication, that’s where this team still needs the most work, as opponents are converting non-shot set pieces into shooting chances at a near-record clip.
The increase in shots from these situations can’t just be chalked up to allowing a lot of free kicks in the first place. NYCFC does commit an above-average number of fouls in the zone (say 30 to 50 yards from the center of goal) where you might expect non-shot free kicks to lead to attempts on goal, but other teams like Portland and Columbus foul more often there without letting opponents create shots from it at anything like NYCFC’s rate.
Small sample size disclaimers apply, but something doesn’t smell right here. Coordinating set pieces was a key part of Domènec Torrent’s job when he worked alongside Pep Guardiola, and his squad returned its entire defensive core this season. So what’s the problem with NYCFC’s set piece defense? There are two of them, actually: the dead zone and the matchup problem.
The Dead Zone
We’ll start at the very beginning with the dead ball set up. New York City tends to play a high line against non-shot free kicks, an effective yet risky method given the amount of space the attacking team has to run into. Playing a high line is fine in theory but is complicated by Sean Johnson’s uncertain command of the penalty area. According to American Soccer Analysis, Johnson ranked 16th out of 23 primary MLS keepers last season in aggression based on the number of successful crosses per claim or punch.
Together, a high line and passive goalkeeping leave a lot of room for error in the resulting dead zone, particularly on inswinging balls. As the distance the defenders have to track back increases, so does the chance of the defensive shape breaking down, making errors in communication far more costly.
In our first example, from last September, D.C. United’s Steve Birnbaum scores an entirely preventable goal due to the combination of a high defensive setup and a keeper reluctant to come off his line. Either Ben Sweat has to track Birnbaum’s run better or Johnson has to claim that ball, though ideally both of those things would happen.
Fast forward to 2019 and the problem remains:
Chris Mueller’s free kick in the season opener originates in a wider position, forcing the defensive line back to the penalty spot, yet once again there’s a fatal lack of communication between Johnson and Chanot as each thinks the other will clear the cross (and let’s be honest, Mueller was most definitely trying to cross that). That seam where the defense and keeper can’t agree who’s responsible for the ball continues to be a major source of vulnerability for NYCFC—but it’s not the only one.
The Matchup Problem
Discussions about set piece defense often start and end at the seemingly eternal debate over whether a zonal or man marking system is better. Both systems can be successful with the right players, and Torrent has used elements of each so far with NYCFC. He’s generally preferred zonal marking for corners, with occasional hybridized schemes against certain opponents. But last season he leaned more toward man marking for non-shot free kicks, especially those taken toward the center of the pitch or against a higher defensive line, like the D.C. United goal above.
This season, NYCFC has moved toward a kind of matchup zone system for non-shot free kicks: each primary defender is responsible for marking a lane from the defensive line down into the six-yard box, while a couple teammates drop into the cutback passing lanes. Zonal assignments are determined by attacker positions in order to put NYCFC’s best defenders on the opponent’s most dangerous aerial threats.
The system itself hasn’t been the problem—it’s really down to the players to execute. But matchup adjustments could explain some early season struggles, as NYCFC have looked disjointed setting up their shape against free kicks from distance. They were beaten out wide on set pieces on several occasions against LAFC:
This ball is played as Ring is still trying to get his teammates organized, and Ronald Matarrita is caught flat-footed on the outside. NYCFC recovers well to regain control, but LAFC would continue to target wide areas on similar free kicks through the rest of the game.
Here LAFC overloads runners on the right side, pulling the NYCFC defense with them. Again, the main problem is NYCFC’s disorganization when the ball is played. LAFC’s objective is to open up the left flank, and Taty Castellanos lets Diego Rossi get in behind way too easily. Maybe whatever changes Torrent has introduced this year are contributing to the disorganization, but there’s clearly room for improvement on an individual level too.
What to Do About It
The interesting thing is that there are reasons to believe this team shouldn’t be quite so bad at these plays. Despite struggling against non-shooting free kicks, NYCFC is better than average for shots and xG allowed from corner kicks. True, defensive movements are less prominent on corners, but these stats suggest a team that’s capable in dead ball situations and in the air. In fact, NYCFC ranks seventh in the league for aerial duel win percentage.
Maybe Torrent is asking too much of his team, running a system he’s familiar with from Europe without considering his players’ skillsets. It certainly wouldn’t be the first time a coach was too married to his preferred tactics. (Remember when Josh Saunders was forced to play out of the back? Sorry, I promise I won’t bring him up again.) Or maybe the system is fine and it’s just a problem of focus.
Whatever the source of the confusion, it does seem possible that this team would fare better against set pieces if Dome minimized the moving parts. You could start by pushing the defensive line a bit closer to goal, just to reduce the amount of ground covered. You could work with Sean Johnson on coming out to claim balls to further cut down the danger space. With more reps, maybe the defense will get better at finding their matchups before the ball is played. Again, we’re still early in the season—maybe this is an overreaction to a small sample, a blip on the radar in hindsight. But let’s hope Torrent’s coaching staff has been looking into the problem just in case it’s not. ❧
Image: José Guadalupe Posada, Chaos During an Earthquake