We've gathered here today to discuss the question that vexes us all. No, not the meaning of life. Think bigger than that.
Why can't Columbus Crew SC score goals anymore?
The Black & Gold were one of the most exciting attacking teams in Major League Soccer in 2015, scoring 57 goals — tied for second in the league — in the regular season and were third in the league in expected goals. The attacking era ushered in by head coach Gregg Berhalter was working, and oh, how grand it was.
But something seems to have gone wrong between a loss in MLS Cup and now. Crew SC has managed only three goals in five games. That's 0.6 goals per game, compared to 1.38 a season ago.
Last year, the prolificness allowed the team to paper over the cracks to a certain degree. All the goals allowed — second most in MLS — were a problem, but one that was manageable because of how often the ball was going in the net for the good guys.
Berhalter has even said he doesn't worry about giving up goals so much as the team is scoring goals. Well, guess what, Gregg? The team's not scoring goals. Not even enough goals for a decent defensive team to win many games.
It may not be possible to diagnosis the problem and suggest the proper medicine from the outside looking in, but it's worth delving into a little deeper.
There are certain things that are easy to study from outside the locker room, the first and foremost being statistics and analytic numbers that can pull back the curtain just a little bit to show what's happening on the field.
The most obvious — and critical — stat is goals, and we've already discussed that problematic number. So what is it about the way Crew SC is playing that has resulted in the shift?
To the eye, Columbus has had plenty of bright moments when it comes the attack. It's been the same approach as the past two seasons, with the fullbacks pushing high up the field and engaging in the attack, with lots of crosses, a free-roaming creator in Federico Higuain and the athletic Kei Kamara up top to, theoretically, finish chances. He hasn't done that at the same rate early on, but otherwise, things look largely the same.
Statistically, the buildup play looks much the same as last year. The Black & Gold are possessing the ball at a 54.3-percent clip (52.6 percent in 2015). The zone splits are nearly identical to last year — 27 percent in the attacking third (28 in 2015), 45 percent in the middle (42) and 27 percent in the defensive third (29). Passing efficiency is identical to last season, at 81.5 percent completion.
That attacking build across the field in 2016 is, left to right, 36 percent, 29 percent and 34 percent. In 2015 it was 31, 33, 37. So a slightly greater lean to the left, but probably not a notable shift.
The number of shots have not changed, either: 13.4 per game this year compared to 13.6 in 2015. Fifty-one percent of shots have come from inside the box, compared to 53 percent a year ago.
One of the biggest differences is in where shots are coming from, width-wise. In 2015 Crew SC's shots came from the middle 66 percent of the time, 17 percent on the left and 18 percent on the right. This campaign it has been 10 percent left, 75 percent mid and 15 percent right.
The shot charts from the last two games bear that out.
vs. FC Dallas:
It's possible to infer from those numbers that Crew SC is making it too easy for defenses to stay compact and in front of shots, rather than forcing the back line to widen and close out on shooters at angles, creating other space for things to happen.
It's a small sample size, but could contribute to the drought.
There has also been a lot of talk about finishing. What is finishing? It's putting the ball in the back of the net, but on a deeper level would typically be discussed as the ability of a player to pick out spots in the goal, poise with which they shoot under pressure and quality of technique (thereby greatly increasing the ability to do what they want to do).
There is a great piece by Sean Steffen over at American Soccer Analysis about what finishing really is, and whether it's a tangible skill. The conclusion he comes to is that there is a baseline finishing rate, and the ability of a player to better that number, regardless of perceived skill, is limited. It boils down to — if a player has the skill to put a shot on net, it's going to go in a certain amount of the time, regardless of who the player is.
So, basically, if Crew SC is putting shots on goal at the same rate as last year, it should score at basically the same rate. In 2016, the Black & Gold are averaging 4.6 shots on target each game. In 2015 that number was 5.3, both in the regular season and the playoffs. That's not a precipitous dropoff, but considering that only 1.7 SOTpg separated the best team (Columbus) from the worst team (Houston Dynamo), it is enough to expect to see a drop in actual goals scored.
Considering that Columbus is taking the same amount of shots from inside the penalty area, it's worrisome that the shots on target have declined. Certainly last week may have skewed that percentage slightly with an abysmal ability to direct shots on goal from quality positions.
When it comes to offensive struggles, the finger has largely been pointed in two directions — Kamara and Ethan Finlay. After 22 goals in 2015, Kamara has just one through five games. Meanwhile, Finlay has seemed a step off since the latter portion of last season's playoffs. Considering those were the teams two leading goal scorers last season, it's easy to understand why their contributions are needed.
Last year the striker posted an expected goals total of 18.93. He exceeded that number, but not by a huge amount. The reality was that he did a tremendous job getting in spots to score, and the attacking machine was built to generate tons of opportunities for him. So even if he were to pull back toward the mean, it would be reasonable to expect 15-plus goals from him again. Right now he on pace for seven.
Anecdotally, it has appeared that Kamara has spent more time dropping to the wings and therefore farther away from goal, so I decided to look a little closer to determine whether that was true or false. What I found is that it is a half truth.
Kamara's positioning and movement is very similar to 2015. He frequently pops up on either wing and drops toward midfield. That kind of hold-up play is necessary of Kamara, so that's a positive thing. Here are some heat maps for Kamara from both this season and last. Heat maps must be taken in context, but can be a useful tool for comparisons.
Kamara vs. Montreal last week, which is in line with his play this season:
Kamara vs. Toronto last December
Kamara vs. Portland in MLS Cup
What you'll notice is that he largely fills the same spots on the field, it's just that his percentage of touches deeper on the field is higher this season compared to his touches in and at the top of the box — the areas you expect Kamara to be dangerous.
The perception that he's spending more time at midfield likely comes out of the fact we're seeing him facing up on the ball more there. Rather than showing to the ball or holding up play, passing the ball off and then getting to the box, he's taking on defenders more in the midfield.
The question is whether that's a product of Kamara's decisions or whether it's a result of his teammates' play — their movement and where they're providing him the ball. Certainly opposing defenses are also more aware of Kamara and forcing him away from the areas he wants to be in.
Finlay has been even more vexing. After quickly becoming one of the premier, young American attackers in MLS under Berhalter's watch, he's become a shadow of that, offering the occasional bright moment, but not imposing himself on the game the way he had for 2014 and most of 2015.
A look at some of his underlying performance numbers show similarities between this season and last. One thing worth noting is that games of 20-plus passes have become more frequent for Finlay this season, whereas some of his best performances in 2015 saw him in the mid- to upper-teens in that category. That may seem counter-intuitive, but Finlay's game is not interchanging with teammates and passing the ball. It's about getting on the end of others' passes as well as taking on defenders 1v1. If he's passing the ball, it means he's not doing those things as much.
Some have posited that the continued integration of Harrison Afful at right back has cut into the way Finlay plays, with Afful getting further up the field on overlaps and taking away the space that Finlay would have otherwise been attacking himself. Berhalter poo-pooed this sentiment earlier this season, but it's still worth taking a look at.
Here are some heat charts, first for Finlay and then for Afful, to try to compare where they were getting on the ball last year versus this year.
Finlay vs DC United - Sept. 2015
Finlay vs. DC United Oct. 2015
Finlay vs. Portland - MLS Cup
Finlay vs. Portland - 2016 season opener
Finlay vs. Montreal - last week
There is not necessarily a huge shift from start to finish, but there are some things to consider. The two games against DC United were strong performances for Finlay, with three goals in two games. His touches inside the 18-yard box are more concentrated in those two — no surprise, the best place to score is from the 18-yard box. There does seem to be a subtle shift to more time helping in the midfield as time goes on. Last week against Montreal, while he did not score when he should have, was arguably closer to the Finlay we expect. He attacked the end line more and got into the penalty area more.
How does that coincide with Afful?
Afful vs. DC United — Sept. 2015
Afful vs. DC United — Oct. 2015
Afful vs. Portland — MLS Cup
Afful vs. Portland — 2015 season opener
Afful vs. Montreal — last week
There's no question that the first game against D.C. is Afful's most defensive of the group, generally staying around the midfield stripe. He pushes higher up in the second game, but still sees the ball go through him mostly in the middle third of the field. By the time we get to MLS Cup, his influence has extended further into the attacking third. In this season's season opener he was all over the place, and if you were to look at it without a label, you would likely assume it was, in fact, Finlay's heat map. Against Montreal last week it looks more like it did in the second game against D.C. United, with a slightly lower percentage of his touches coming coming in the attacking third, and definitely less in the corner and into the box.
Does all of that indicate some kind of cause and affect? It's impossible to say; but to completely deny a link would also be short-sighted. The question then becomes, if there is a link, is it something Finlay can adjust to over the next few weeks? The last two games saw more involvement from him and could be a sign of better things to come.
There's also the factor of overachieving. That tag was planted on Finlay pretty quickly considering he came out of nowhere and exploded onto the scene in 2014. In that season he had 11 goals and six assists. That exceeded his xG by a pretty significant margin (5.27 differential). Many, including myself, had concerns that he would not produce at the same clip, if because of nothing more than a normal regression.
But then he exceeded expectations again in 2015. He notched 12 goals and 11 assists; his xG was 7.94, his xA was 5.22. So there was an uptick in underlying performance, but also no reversion to the mean, and so he again over-produced.
It leads to a scenario where you can believe one of two things: either Finlay will tend to always over-produce, or there's still a regression on the way. This may just be that regression kicking in; that may be why, given a couple of great looks in the last two weeks, he still hasn't opened his account in 2016 — sometimes things work out better than they should, and sometimes they work out worse than they should.
There are, of course, things that affect sports that come down to more than statistics. Without being in the locker room it's impossible to take the mental temperature of a team, and it's probably a tricky prospect even from the inside — guys that can do it well and consistently are the best coaches and managers in the world, and that's an exclusive group.
There is a reason, though, that it's so hard to sustain success. One might argue it's even harder in MLS, where things are constantly changing and the gap between top and bottom in the standings — both points-wise and ability-wise — is relatively slim.
There are critiques that can be lobbied at Crew SC thus far in 2016. Is there an MLS Cup hangover? By playing in a championship game, you get a shorter offseason than other teams — less time to recover and plan for the next year. There's also the emotional factor of coming so close and falling short. These guys are competitors, and that affects them. One would like to think it motivates a team to be better, but it's far more complicated than that.
And once you reach the peak of the mountain, you immediately have a target on your back. It's not ridiculous to believe that everyone Columbus plays this season will play a little bit harder and a little bit better because they want to prove they're capable of being among the league's best.
Some have questioned the intensity of the team so far this season. Is there a level of satisfaction having finished second in the Eastern Conference and having made a deep playoff run? It can be hard to achieve at a high level and then continue putting in the work to maintain or raise that level.
One way to continue to push is to create an environment where everyone is on alert and fighting to prove themselves. If there's one thing that Berhalter has been, it's consistent. The starting 11 is the starting 11, with little change. Part of that may be a clear delineation in talent and comfort in the system, but are the starters feeling enough pressure? Do they believe that if they're not at their best then they will be replaced? Has the depth we thought we might see this season panned out? New signings that were expected to make the squad more dynamic from top to bottom have not yet had a visible impact, with Ola Kamara just getting healthy now, Emil Larsen has been MIA, Amro Tarek doesn't appear to be threatening for playing time (maybe even with Gaston Sauro out), Corey Ashe has been hurt and Rodrigo Saravia doesn't seem ready for MLS minutes.
Is Crew SC coasting too much?
A winless first five weeks would certainly be enough of a wakeup call, one would think.
A diagnosis (kind of)
In 2014, Columbus had some goal-scoring issues. It responded by going out and finding a top-level striker. While it's unclear what contributions Ola Kamara or Conor Casey or Larsen will make, there's no doubt that this attacking group is light years ahead of the one in 2014, when Jairo Arrietta, Aaron Schoenfeld and Adam Bedell were the great goal hopes.
So what's going on? It would be reasonable to believe that last year's 57 goals simply can't be repeated year after year. This may be Finlay's expected regression. It may be unreasonable to believe Kei can produce 20 goals again. Out-producing the team's expected goals by more than eight goals may not be likely again. But the team is better than 0.6 goals per game, and the underlying numbers continue to show that.
The answer is patience, at the macro and micro levels. The players must remain patient, ignore the expectations and stay calm in front of goal. Another shot or two on target per game and history tells us goal production will go up.
And then there's big picture patience, which Berhalter continues to preach. Just as I can point the finger to over-production in some cases in the past, the Black & Gold are currently in a stretch of under-production. From a statistical point of view, even if nothing changes, over the long run Columbus is likely to find the back of the net more than it is.
The question will be whether Crew SC can be an elite club if it's not scoring 55-plus goals.