Chilwell, Cucurella, Colwill, Veiga: Who should play left back for Chelsea?

While my previous comparison was based on the consensus that the right side of the defense would handle the inversion, it is still not entirely impossible that the left is used instead. Reece James and Malo Gusto have played that role from the right in preseason, although they largely played inverted the entire match with very little reverting. Gusto did play a more traditional inverted role against Club América and Manchester City while he was playing at left-back, but that seems like a stop-gap solution. I succumbed to the point of view that the right would be used to compose the right-back article, but I still think the inversion will might should come from the left. Especially because Marc Cucurella missed the preseason tour in America and Reece James has managed to get himself injured again, we likely will not have a finalised answer to this question for the start of the season, but let’s take a look. And so, in the parlance of our time, here we go!

Why do I think that it should be on the left? Because early on, before truly establishing who was playing at each position last season at Leicester, Enzo Maresca sometimes played with Ricardo Pereira as a left-back. Furthermore, he inverted him from that left-back position. On 15 September 2023 against Southampton, the growing pains and coverage confusion of Pereira taking on that inverted role from left-back coupled with a poor touch by Callum Doyle were the errors that caused the single goal that the Saints scored against Leicester in a 4-1 thrashing. Maresca then made specific and calculated adjustments to his defensive line, eventually cementing Pereira at right-back while Wout Faes came into the fold, and Doyle eventually lost his starting spot to James Justin. Within the following two matches, you can see how he reformatted his defensive line into what became a formidable force through the winter in the Championship. And yes, I have watched far too much of last-season Leicester this summer.

They had a few changes throughout the season due to both their injuries and Maresca’s intricacies, but still only used a total of 27 players – a few academy players were used in domestic cups, only 22 played in more than 5 different matches, and they were nowhere near our usage of 32 squad players last season. Once the starting XI on the far right above was settled, the left-back position was the only position on the back line Maresca would dare change – in part because James Justin would eventually take that starting position, but also because he expected offensive contributions from that flank, too, and so the position is physically demanding. In fact, positional play requires adherent scrutiny into each position, so while his rotation in cup games could have been predictable, his starting XI in the Championship was about as consistent as it gets. When they would face Southampton again towards the end of last season, they lined up the same as they did against Bristol City above, the exception being that James Justin started rather than subbed in for Callum Doyle.

Here comes my caveat and perhaps the most important part – the opposing flank from the inverted side would often push higher too, supporting the winger but not negligent of the defensive cover and not drifting as far centrally as Pereira did on Leicester’s right…the sort of thing which would naturally fit the profiles of both of our right-back options (if James becomes/stays fit.) Look at both Callum Doyle’s (on loan from Man City, who also played as LCB) and James Justin’s (who also played as a RW and RB) heatmaps below – the deep offensive flanks are reserved for true wingers and the no. 8’s released by the inverted fullback. Historically Cucurella and Chilwell often contributed offensively, but did so from that deep flank. That is not an objective in this system – Doyle and Justin both have quite similar heatmaps because they are occupying the middle third of the pitch on their flank entirely and drift centrally a bit as well, but leave the flanks in the final third to the wingers. This is why I said in the last article that if Maresca is as steadfast to his formation and philosophy as it seems, the inversion should be coming from the left. Not only has Cucurella done it already, but it suits the offensive aspects of this team and it makes us stronger defensively than inverting from the right

In fact, if we continue to invert from the right (Maresca seemed ambiguous about the side from which the inversion would come in his first pre-match press conference and there hasn’t yet been enough evidence to support either side), I would actually prefer Levi Colwill to any of those above for the defensive stability he would provide, just as Maresca has done. Yet, if he continues to do so, it is a disservice to Colwill, often our right-sided centreback has been exposed for a lack of leadership on that back line. Furthermore, the inverted back’s direct offensive contributions (see Pereira’s stats below) are actually diminished because, although on the ball more, he is not getting as far up the pitch as he would but instead focus on drifting centrally. For me, it is blatantly obvious that the left side is the better side to implement the inverted approach.

We’ve seen Cucurella invert, he can most certainly handle it. Chilwell has often been played higher up the pitch in recent seasons and this would move him out of a more critical defensive position on the back line where he doesn’t have the height that Maresca wants. In fact, it seems hard to imagine he could not replicate essentially the same role Cucurella played at the end of last season. He doesn’t seem disinclined to possess or distribute the ball, he loves to pick up assists and goal contributions, and he has defensive capabilities similar to those of Cucurella. Veiga has played as a left-sided fullback and centreback, but also as a CDM. In fact, Veiga’s signing is intrinsically related to this article. He has the height that is lacking from either Chilwell or Cucurella to better challenge opponents aerially, his speed and acceleration are notable, and he must’ve been brought in to connect the dots we are all seeing. Once again, the luxury of having to choose from options as to who performs best in our left-back role is something that Maresca needs to figure out immediately. While viewing below, note that these are mostly offensive stats and taken from different times and positions throughout these players’ careers, so read the metrics carefully.

And here are those same metrics for the Leicester players they’d be replicating, both Pereira as the inverted option or for the Doyle/Justin fullback role.

Similar to the last article, injuries and inconsistent availability means that a fair comparison for Ben Chilwell will have to involve multiple seasons. Not including his cruciate ligament tear in November 2021 that ruled him out of the rest of that season, he has missed a total of 268 days and no less than 55 matches from 8 different (granted that an exasperated injury counts as a new one on TransferMarkt) injuries over the last two seasons. Just like Reece James, his are recurring injuries that look only to be resolved with proper minute management.

I will try to emphasise those games where Cucurella was inverting, which began in the convincing win against Tottenham in early May – notice the bottom right chart above. However, because that is only a period of 18 days and 5 matches, at times larger portions of this season will have to be used. Cucurella’s continued success with Spain this summer has certainly increased his prospects of being the starter, but hasn’t increased his physical stature, which is something that clearly Maresca is starting to consider during this rebuild. Cucurella is only 175 cm, Chilwell is just 178 cm, but Colwill is 187cm and Veiga is 190 cm.

It is hard to truly pinpoint an ideal time frame to compare for Veiga – he came through Sporting’s academy and played primarily as a CDM or CB. He then went to Augsburg where he played primarily as a LCB or LB, only to move to FC Basel and play as a CDM again. His inclusion is because although new and young, he fits the profile of the player that Maresca would want to play this role and supposedly he was adamant about the youngster’s arrival. Veiga himself had the following to say about his positioning and preferences:

“I’m comfortable on the ball and I would describe myself as a complete player due to my versatility. I’m comfortable in many positions on the pitch. I wouldn’t say I have a favourite position because it depends on the opponent or tactical elements, but I have played at left-back, centre-back and as a defensive midfielder. I like to defend a lot and with my height and physical strength, I enjoy it. I like to control the game with or without the ball.”

-Renato Veiga; source: Chelsea FC

Sounds like Maresca’s ideal player. However, because he was plying his trade in the Swiss Super League last season, advanced passing metrics aren’t available for him. Much of the statistics you will see on him are from his time at Augsburg, where he ironically was primarily a left-sided defender (albeit with different tasks than inverting), but more statistics are available for him then. Preseason has still not determined if he has a first-team role this season or if a loan is best suited for him. His versatility and comfort in being in each of those positions should facilitate his inversion between them and certainly helped bring his attention to Maresca and the club – especially considering the tremendous scouting of young prospects by FC Basel, not least including our previous supposed interest in Riccardo Calafiori.

I am also not going to include Levi Colwill in all forms of this assessment because I am expecting and hoping that he plays as the left-sided central defender (being challenged by Benoît Badiashile, Bashir Humphreys, et al.) rather than out of position again like last season. He has often been tried there this preseason, but I expect him to be starting as our left-sided centreback this season, defensively solidifying the side from which we will hopefully invert.

The last article also noted the emphasis placed on both short passes and accuracy required in this inverted role – which Ricardo Pereira fulfilled masterfully last season – so let’s look at each of the players’ passing metrics compared to his. Just like last time, bear in mind the amount of minutes and starts each player got while also recognising the team in which they were playing.

Now, while only Chilwell in 20-21 comes anywhere near the number of minutes and matches played as Pereira during his comparative season, it is pretty glaring how nobody comes near the statistics that Pereira put up. Just like James and Gusto failed to compare in the last article, that is down to their role under those specific managers/setups, and specifically not taking up a central location where passing was designed to go through them. Still, our players’ deficiencies in passing may be why Maresca ends up pushing the invert from the right – Gusto’s or James’ passing is far more accurate and expansive than any of the left-sided options. You would have to say that both Chilwell (20-21) and Cucurella (23-24) have put up numbers that Maresca should be able to work with, while Veiga’s numbers represent a 19-year-old man coming into a team mid-season on loan that survived relegation by a single point. Still, comparably speaking, nobody has the numbers to keep up with Pereira, and so there is no clear frontrunner, nor as there really was from the right.

Preseason has told us that the most critical part of this team will be the possession, particularly in building out from the back. That is not a problem on the whole, this team is filled with the offensive talent that SHOULD be able to string together passes without errancy. In preseason, the positions we are taking up lacked width and seemed far too congested, and so we are being pressed into mistakes more easily. One would hope and expect that, with more training, players should have patterns drilled to the extent that they know where each other will be two passes before it gets there. That is precisely what Kiernan Dewsbury-Hall stated was the most impressive part about learning under Maresca, his intricacy into detail in not only the patterns, but the positions. Against Manchester City, it seemed to take two steps back after a step forward against Club América.

“Positional awareness, knowing when to be in certain instances in the game when we have the ball […] and trying to be aggressive off the ball. They are the main principles […] It will take a bit of time. It was similar this time last year. There was a long way to go. We were making mistakes, but you could see there was progress from week to week, and I feel like that is what we are seeing at Chelsea.”

-Kiernan Dewsbury-Hall; source: Chelsea FC

Again, the aggression off the ball and defensive reliability of each player will be quite important, too. We have certainly seemed fragile and inconsistent at best this preseason in that department. Let’s see how they stack up against one another defensively.

First of all, inverting from the right leaves one of these three to cover as an LCB in the realigned back 3, and so defensive cover (and height) would be an emphasis in the position and suited much more to Levi Colwill or Renato Veiga – Colwill particularly does not have the offensive tendencies requested of that position, but I have included his defensive stats below. Inverting from the left leaves Reece James (ideally, assuming he isn’t out for too long) or Malo Gusto (more than ample cover, just less physically imposing) in that defensive coverage position, and both are more suited to that task while contributing offensively. Plus, they are afforded the ability to use the flank the way they already do so well.

Let’s look at our crop of players defensively compared to Pereira. Because evidence from preseason suggests that they will be relevant just like the heatmaps above, I have included stats for Doyle and Justin to give us some sense of comparison with their roles, too.

Veiga’s stats only entail 7 starts and 639 minutes, Doyle’s stats are from 20 starts and 1672 minutes, and Chilwell from 22-23 had 24 starts and 2144 minutes – all the others played in at least 30 matches (Colwill only started 28) and had 2500+ minutes played. The aggression in the middle third is something that I pointed out last article, but the overall aggression of the Leicester three is much greater than that of our players. Still, they did not tackle with greater efficiency. And before we jump on Pereira for his errors, we must also remember how much more time on the ball he spent in critical areas. We are not comparing like-for-like roles, and so the statistics between them will inevitably vary. They simply provide ideas of both the expectations from the roles in Maresca’s system and what our players were capable of in their specific roles given in different squads.

It should be noted how effective Cucurella was at tackling last season (after putting up the worst numbers comparably speaking the season before), which again means that putting him into the position of Pereira could be exactly what we need. He has been tenacious and has put in spectacular defensive performances against some of the best wingers in the league, but he does fall short of Pereira in many of the other defensive aspects above.

While it was harder to find recent stats for Veiga due to his participation in the Swiss Super League, TFA did an analysis of him lately with some particularly useful defensive stats. According to them, last season he had 5.82 defensive duals and 3.68 interceptions per 90, which is even greater than the 4.73 and 2.35 respectively that Pereira hit. Additionally, his defensive territories maps suggest that he is an ideal candidate to fill this inverted role if his passing skills are up to the task in a much more challenging league.

Among his available choices, Maresca should have the ability to make this work. These players all seem to be tactically flexible enough to handle the transition into midfield. Ideally, Maresca will loosen his firm stance and will allow a less predictable adaptation of his inverted fullbacks, as the flexibility it could provide would be difficult to predict or read for an opposition defense. Either Gusto or James could invert from the right, stay deeper and tuck into a back three, or advance slightly against weaker opposition while the left inverted to get to the eventual offensive position Maresca desires.

This left side has many more variances and options. If James/Gusto invert and our opposition is weaker, Cucurella or Chilwell seem able. If our opposition is stronger, Colwill would be better suited. If this side plays inverted, any except Colwill seem viable – and who am I to put it past him, either? Having such a wealth of options to learn this new system is the best chance we have of finding the right balance in both phases of the game. We certainly haven’t yet, but take it from my point above about Maresca’s evolution at Leicester or from Kiernan Dewsbury-Hall himself – it is certainly possible.

KTBFFH.