The Columbus Preview, Part II Redux

Editor’s Note: Most of this information is taken from my first Part II preview, however, I have updated the numbers and other information to match the current state of Columbus. If it looks familiar to what you saw before the first attempt at this game, this is why.

Well here we are. With some help elsewhere (GO ORLANDO!), two wins and we’re in.

There’s not much to say about the lead up to this game other than we have no wiggle room. It’s a win or we’re out. We’ll be shorthanded, but we’ll have just over 74 minutes to make this happen. Let’s get it.


All stats from FBref.com
TeamPossessionPoints (standings)WhoScored team rating (SofaScore team rating)
Charlotte FC53.1%41 (9th in the East)6.53 (6.79)
Columbus Crew52.5%45 (8th in the East)6.68 (6.90)
All stats from FBref.com
TeamShots per gameShots on target per gameGoals for (xG)Goals against (xGA)
Charlotte FC11.254.0340 (38.1)48 (42.1)
Columbus Crew13.284.0943 (39.3)32 (38.2)

The biggest thing that jumps out is the amount of shots that Columbus take. They are actually 6th in the league in this respect, although it hasn’t necessarily translated to a huge goal return. As we will see below, though, they have some seriously good attackers.

The rating systems of SofaScore and WhoScored both “like” Columbus better, but to my eyes these are two very evenly matched teams, IF both were able to play their best lineups. Columbus will be able to do that; Charlotte will not.

Lucas Zelarayán

TL;DR: Zelarayán is really, really, REALLY good. You knew that though.

My Armenian compatriot has been a menace to the league since he joined Columbus from UANL of Liga MX in 2020. In that COVID-shortened season, Zelarayán had 6 goals and 2 assists in just 16 appearances. He followed that up with a 12-goal, 4 assist campaign last year in 32 appearances. He currently has 9 goals and 6 assists in 27 appearances.

Zelarayán heatmap, 2021, via SofaScore

His goals last year were an aberration for him, historically, but I’m not sure they’re an aberration for him for MLS. I will say he is someone who consistently seems to over perform his xG (2020: 6 Gs, 0 PKs, 2.8 xG, 2.0 npxG; 2021: 12 Gs, 1 PK, 6.5 xG, 5.7 npxG; 2022: 9 Gs, 1 PK, 4.9 xG, 4.2 npxG). Some players just do this.

Zelarayán began his career at Club Atlético Belgrano in the Argentinian Primera División. Through 2 seasons at Belgrano, he had 46 appearances (45 starts) and scored 9 goals and 7 assists; he was 21 and 22 years old in those seasons.

He moved to UANL in 2015. He only made 13 appearances for Tigres in 2015-16, only scoring once and assisting once. This is by far his worst season as a professional. He followed that season up with a 32-appearances, 24-start campaign, in which he had 6 goals and 6 assists. He made back-to-back 18-appearance seasons in 2017-18 and 2018-19, but only had 4 starts in 2017-18 and 14 starts in 2018-19. He had 3 goals and 1 assist in 2017-18 and 5 goals and 5 assists in 2018-19. He made 15 appearances (8 starts) for UANL in 2019-20, scoring 4 goals and getting 1 assist. He joined Columbus in December 2019.

Zelarayán heatmap, 2022, via SofaScore
Zelarayán heatmap, 2022, via SofaScore

Based on his performances in MLS, this is a consistent double-digit scorer. If 2020 hadn’t been shortened due to COVID, I think Zelarayán probably hits double-digit goals. He needs just one more this year to do that and he did miss a few games. Basically, Zelarayán’s new normal seems to be about 10 goals per year with 4+ assists. He is already 30, so it’ll be interesting to see how long he keeps this peak up, but I would imagine he has a good 3 years left in him, at least.

When looking at his heatmaps, he’s everywhere. I also think they show why he’s been more productive in MLS than he was in Liga MX. Specifically, UANL was using him in more wide positions. For example, look at the 2018/19 Liga MX Apertura heatmap below.

Zelarayán heatmap, 2018/19 Liga MX Apertura, via SofaScore
Zelarayán heatmap, 2018/2019, via SofaScore

That is not the same role at all. Zelarayán may just be better suited to MLS than Liga MX or he may be being used correctly now; it’s probably a mixture of both. Regardless, his move to CAM with Columbus has been the right one.

The heatmaps also show an increase in his possession of the ball. This is born out in the numbers too, as he’s gone from 54.2 touches/90 in 2020 and 61.5 touches/90 last year to 70.03 touches/90 this year. His live-ball touches are up to 61.60 touches/90 this year compared with 52.2 last year. Additionally, he’s carrying the ball more (53.26 carries/90 this year compared with 46.1 last year; 8.05 progressive carries/90 this year compared with 6.19 last year), being targeted more (59.75 targets/90 this year compared with 53.6 last year), and receiving the ball more (52.46 receptions/90 this year compared with 42.5 last year; 9.31 progressive passes received/90 this year compared with 8.66 last year). In short, Columbus has realized he’s f*****g good and is getting him the ball as much as possible.

Zelarayán passing percentiles vs. MLS CAMs/wingers over the past 365 days via FBref.com

Somewhat surprisingly, Zelarayán isn’t necessarily elite at passing. Over the past 365 days, here are his passing percentiles against other MLS CAMs and wingers. Look, these aren’t bad numbers by any stretch, but the poor completion percentages are notable.

Now, I do believe these percentages are a result of his ambition in passing, rather than a lack of talent. When you compare his completion percentages to other areas (key passes, attempts for any distance, passes into the penalty or final 3rd, progressive passes), what you see is a player who isn’t afraid to try the difficult pass. He simply doesn’t play it safe and, as a result, he contributes to the attack in very meaningful ways. Sure his short passing percentage (20th percentile), medium passing percentage (59th percentile), and long passing percentage (38th percentile) all look bad, but his xA is in the 96th percentile, his SCA is in the 97th percentile, and his GCA is in the 95th percentile. This shows a player who is going to try for the killer pass, not the safe option. It’s exactly what you want from a 10.

Elsewhere, Zelarayán’s numbers are elite. He’s in the 87th percentile for goals, although only 51st percentile for xG.

Zelarayán takes his shots from a long way out: 23.40 yards, which is good for the 3rd percentile. Zelarayán is taking a lot of shots from distance and a lot of shots in general (95th percentile, 3.29 shots per 90). This frequent shooting and shooting from long distances will naturally drive down his shots on target percentage (55th percentile, 34.6% per 90). I don’t think you necessarily want him to stop taking these shots, as it’s clearly working for him. Additionally, some players just outperform what the numbers say they should be doing. He may very well be one of them.

In addition to his passing ability, assist making, and goal-scoring, Zelarayán is also a phenomenal dribbler. He’s in the 99th percentile for dribbles completed (3.41 per 90), the 98th percentile for dribbles attempted (5.77 per 90), the 80th percentile for successful dribble percentage, the 99th percentile for players dribbled past and the 93rd percentile for nutmegs. Unsurprisingly with these dribbling stats, he draws a lot of fouls (97th percentile).

Defensively, Zelarayán is–who cares. Legitimately, his numbers aren’t good, but I don’t care, Columbus probably doesn’t care, and you shouldn’t either. Zelarayán is an MLS version of prime Mesut Özil. You don’t ask or care if that player defends. He’s in the squad to score and assist. Zelarayán obviously does that exceedingly well.

In short, as an attacking midfielder, Zelarayán is about as good as you’ll get, especially in MLS. He’s going to be a nightmare for Bronico, Walkes, and company to deal with, especially since he’s going to pop up all over the pitch.

Cucho

Cucho heatmap, 2021/22 for Watford, via SofaScore
Cucho heatmap, 2021/22, via SofaScore

Columbus appears to have done some really good work getting Juan Camilo Hernández Suárez, or Cucho for short.

The 23-year-old Colombian began his career at CD América in the Colombian Primera A, where he was on loan from Spanish side Granada. He made only 13 appearances and had only 2 assists in those appearances, but was only 17 years old. He would never appear for Granada.

He was sold to Watford in 2017 and then sent out on loan. First, he moved to Huesca in the Spanish second division in 2017 and would go on to make 69 appearances (58 starts) for them over 2 years. As an 18-year-old in the Segunda División, he scored 17 goals and had 6 assists in 35 appearances. He helped Huesca move from the second division to La Liga for 2018-19, but the production was a bit different there.

Cucho heatmap, 2022, via SofaScore

In La Liga for Huesca, he only managed 4 goals and 3 assists in 34 appearances. He moved to Mallorca on loan the following year, where he had 5 goals and 1 assist in 22 appearances. He then played for Getafe, where he had 2 goals and 3 assists in 23 appearances. He returned (finally) to his parent club, Watford, for the 2021-22 and made 25 appearances, but only 11 starts. He scored 5 goals and had 2 assists in those appearances.

Now with Columbus, Cucho has gotten off to a flying start. He’s made 14 appearances (12 starts) and has 9 goals and 2 assists. Like Zelarayán, Cucho is a wide player that Columbus has put more centrally. His heatmaps above show that change. For Watford (and previous clubs), he was played as a winger. FBref has him as a “forward” for his 4 games with Columbus and the heatmap and starting XIs support that. He’s still drifting wide, but he’s playing much more as a forward.

Now Cucho is only listed at 5’8″ (and a half) and 161 lbs, but he’s been good in the air. He was in the 96th percentile for aerials won for the 2021-22 Premier League season. When compared to all 5 Big European leagues, he was in the 99th percentile. It should be noted, that this is for him as a winger, which is not how he’s playing with Columbus. For Columbus, he’s in the 51st percentile of aerials won, so it’s not quite as high as his PL time, but he’s also being compared with strikers now. I still take this number as a positive.

It’s difficult to dive too much into his percentiles. Watford was not good last year and got relegated. Of course, that fact actually speaks to his talent as some of his percentiles were really impressive. When looking at just the Premier League and comparing him to CAMs/wingers, Cucho is in the 84th percentile for goals, the 86th percentile for non-penalty goals, and the 91st percentile for goals per shot on target. He’s only in the 46th percentile for xG and the 59th percentile for shots on target, but, again, Watford was bad so this probably isn’t a surprise.

While Cucho only had 5 goals last year, Watford only had 34 total on the year. He scored 15% of their goals! The point of all this is that Cucho can score. I’m not sure what exactly his level is, as he tore apart a second division but never got more than 5 at a top-flight club. At only 23, it’s not like he’s a finished product and the potential for him to score loads of goals is real. If I had to guess, I would say we see output similar to his first year at Huesca with Columbus, especially with the positional change factored in.

Cucho passing percentiles vs. MLS forwards over the past 365 days, via FBref

The passing numbers for Cucho are good, even if the completion percentages aren’t. In a lot of ways, his numbers are similar to Zelarayán’s. At the Premier League level, Cucho often struggled with his passing. In MLS, he’s been a lot better. The step down in competition helps explain that, but doesn’t do much to assuage my fears about how dangerous he’ll be for us.

His dribbling numbers were decent in the PL: 71st percentile for dribbles completed, 74th percentile for players dribbled past, and 67th percentile for dribbles attempted. Unsurprisingly, he’s gone to another level in MLS: 96th percentile for dribbles completed, 95th percentile for dribbles attempted, and 96th percentile for players dribbled past.

Looking at all of this, I think Columbus’ decision to convert him to a central forward makes sense. His history shows he can score, but not necessarily that he can pass well. As a center forward, that deficiency is more acceptable. Pairing him with a talented playmaker like Zelarayán and an emerging wide player in Etienne will create a lot of problems for opponents in this league.

Luis Díaz

Díaz heatmap, 2022, via SofaScore
Díaz heatmap, 2022, via SofaScore

Díaz is a true wide player for Columbus. Before the first time this match was supposed to be played, Díaz had 15 appearances for Columbus, but only 5 starts. Since then, he’s up to 24 appearances and 15 starts, so he’s definitely worked his way into the lineup.

He’s in his 4th season with Columbus, only has 5 career goals and 11 career assists, and is already 23. On the surface, he doesn’t appear to be that interesting. Yet, I’m intrigued and it mostly has to do with what his heatmap shows us, especially when compared with Etienne on the other side.

What jumps out to me about Díaz’s map is how close to the touchline and far up the pitch he stays. He’s in the 77th percentile for touches in the attacking penalty box and the 64th percentile for touches in the attacking 3rd. These percentages have decreased as he’s had more starts, so they are asking him to drop back a bit more than he was. However, this is still a player who will primarily stay up the pitch.

Etienne heatmap, 2022, via SofaScore
Etienne heatmap, 2022, via SofaScore

Compare his positioning to Etienne, who is playing on the other side. Yes, Etienne likes to be by the touchline as well, but he’s far more likely to come into the center of the pitch.

When thinking about the players Columbus has as attackers, the need for Díaz’s positioning makes sense. Cucho and Zelarayán, as we have seen, will drift all over the pitch. While I haven’t shown their maps, Morris and Nagbe are also all over the pitch. The fullbacks will provide width, but Columbus needs some of their advanced players to stay wide and create space. Díaz is doing this job and, from a counting stats perspective, it’s largely an unthankful role.

Where Diaz does excel is as a dribbler. He’s 80th percentile for dribbles completed and 93rd percentile for dribbles attempted (although he’s 16th percentile for successful dribble percentage). He’s able to dribble past opponents (86th percentile) and nutmeg them (94th percentile). When he’s in the attacking 3rd, he carries the ball into the penalty area at an elite rate (99th percentile) and is always available to receive passes upfield (84th percentile for progressive passes received). Finally, because of his trickery, he’s elite at drawing fouls (98th percentile).

Díaz passing percentiles vs. MLS CAMs/wingers over the past 365 days, via FBref

Díaz is not a good passer, as his percentiles to the left show. To his credit, it doesn’t look like he’s really being asked to be a passer in the buildup. His numbers are really good for crossing and assists, so the danger is when he has the ball in the final third.

Díaz appears to find his way into the side for two reasons: 1) to hug the touchline and provide width and 2) to drive to the penalty area, be tricky, and draw fouls. In some ways, he’s there to cause chaos. As a whole, his profile doesn’t seem great, but if we view him as carrying out a very specific, narrow role, I actually think he is excelling at that. He’s the Michelle Williams to Zelarayán and Cucho’s Beyoncé and Kelly Rowland. Did that metaphor work?

Conclusion

Columbus are a poor man’s Philadelphia. They play good defense, have a really good goalkeeper, and have talented attackers. However, they don’t defend as well as Philly and they are no where near the offensive threat of the Union. They are certainly dangerous and players like Cucho and Zelarayán will cause our make-shift backline all sorts of problems.

This lineup is going to be a challenge for Lattanzio. We are NOT going to have anything near the lineup we just had with Philadelphia due to these rules.

Whereas Columbus is going to be able to play this game with basically the lineup they would anyway, Charlotte lined up this way on that night:

Charlotte lineup vs. Columbus, via MLS

Obviously Corujo will be unavailable and Kahlina’s availability is up in the air. McNeill, Bender, Mora, and Reyna have not really featured for Charlotte much recently. That means 6 out of our Starting XI from this game are either non-starters now or injured. Yeah…that’s going to cause issues.

This is a must win game and I would encourage Lattanzio to be brave. Specifically, we should be making subs right away. This is a difficult, and fine, line for him to walk because immediately subbing some of these players will likely not sit right with some of them. With that said, we can make the playoffs and we need the best team we can.

Let’s start here: Jones in for McNeill. This is obvious and needed, as a midfield of Bender, Bronico, and McNeill will get torn apart by Columbus’ attack. Next, give me Mora out for Lindsey. I will say, I’m not sure I see much difference between Mora on the left compared with Afful, but CL certainly does. As such, let’s get Lindsey’s attacking nous in this game.

Świderski has looked good as a 10, but unless we make 3 changes right away, I think we need to try him back as a striker. With 2 subs right away, we’d still have 3 windows to make subs (at the half and 2 more sub periods) to make 3 changes. If the first half doesn’t go well, let’s get Bender and Reyna out for Ríos and Jóźwiak, respectively.

Our center backs are going to be Walkes and Sobociński-we have no other choice. Malanda is ineligible, Makoun is on the Revolution now, and Fuchs had a red card ineligibility for that game. I’m a believer in Sobociński, but man is this a big ask for a young player who hasn’t played much this year.

With all this said, if you can beat Philadelphia 4-0, you can beat any team in this league (on your night). We can’t lose or draw against this team. We have no other choice than for this to be our night.

Prediction: Charlotte 2 – Columbus 1