[NB. Given the size of this document, it might take time for some of the gifs or viz to load. Please be patient!]
After 16 years out of the top flight of English football, Leeds United came into the 20/21 season with a simple task: stay up.
In the end, though, they overshot their target wildly. Achieving enough points to ensure safety by the mid-way point, they ended up putting up a record number of points for a promoted side.
The opening match of the season feels a lifetime ago now. But here is everything that you need to know about Leeds United’s return season to the Premier League…
The Story of the Season
It’s impossible to tell the story of Leeds United’s first season back in the Premier League since the early 2000s without fixing the narrative arc around the fulcrum of the 6–2 result away at Old Trafford.
Of course, narratives are always frameworks around which to hang facts as and where they suit. But in this case, the facts are quite compelling. This story is best told through a cumulative plot of expected goals (xG) and expected goals against (xGA):
This graphic shows you how Leeds’ xG (the yellow line) and xGA (the blue line) grew match on match (the x-axis shows the game week). A break has been introduced after Game Week 14, the weekend when Leeds played Manchester United.
As you can see, before that break, the upward trajectory was steep but xG and xGA largely tracked one another. After the break, though, the curve flattens for both metrics but flattens slightly more for xG than xGA.
That tells us that around Match Day 14 in the Premier League, something happened to Leeds United which meant that they became less defensively open but also less productive in attack. And whatever it was that happened, it started introducing a separation between the xG and the xGA such that Leeds were no longer matching their opponents per expected goals. The question is: what happened?
If the media were to have their way, no doubt they would put this down to their exercising of the fourth estate, holding Marcelo Bielsa to account for his “naive” play at the beginning of the season. But a more sensible thesis seems to be that there was something more intrinsic taking place within the Leeds team.
The first of these things is that Mateusz Klich “broke”. What we mean by this might not be entirely clear. On the one hand, he does start to suffer injuries starting with the first West Bromwich Albion fixture back at the end of December (Match Day 16). Despite this, though, he still massed an impressive 2,300 minutes of Premier League football which suggests he spent a lot of the season playing through injury or at least not at 100%.
Of course, in the times where he comes back into the team, the tactics of the team had changed (notably through Stuart Dallas moving into the midfield — which we will come to in due course) and this, in turn, resulted in a stylistic change. What we have, then, is a chicken-egg situation which first emerges with Klich’s initial injury against West Bromwich Albion.
This shows up in a number of different forms in the data. For example, in an attacking sense, his productivity after his first injury drops remarkably:
On this plot, Match Day 16 is the point at which Klich gets his first injury. However, he then proceeds to play two full games before dropping to the bench again, no doubt because his injury has continued to cause him problems. Following that point, he never quite achieves the same heights of productivity.
The other area where this shows up is in his pressing numbers. Obviously, his lack of minutes after the first injury makes it hard to see trends in the pressing data. To fix that, I’ve adjusted it for minutes so that we can get a better sense of the per 90 minutes values. As you’ll see from the chart, this plays havoc with the data later on in the season when Klich plays very few minutes, but it does show something interesting before that:
For the first half of the season, the story is simple enough: Klich’s pressing begins to decline after about the 12th Match Day. It then falls away completely around the time when he gets his first injury and then it picks up in a whole new way after Match Day 19. However, that pick up sees him increasing his midfield third pressing. The question is: why this stylistic change?
Well, Match Day 19 was the point in time at which Stuart Dallas first moved into central midfield as a regular fixture.
The Stuart Dallas Tactical Shift
As many of our listeners like to gleefully point out, we didn’t like the occasions when Stuart Dallas was employed as a midfielder in the Championship. However, with Mateusz Klich carrying an injury and a dearth of central midfielders, Marcelo Bielsa had his hand forced a little. The fact of the matter was, the Leeds high press was starting to get broken with increasing regularity.
Take a look at this plot of Leeds’ pressures by third over the course of the season:
Again, notice how the timeline fits Mateusz Klich’s injury timeline. The first injury occurs around Match Day 16 and Stuart Dallas is moved into central midfield around Match Day 19. After that point, the pressing changes. However, you may notice, the number of pressures goes up. Interestingly, the pressures in the attacking third remain fairly regular throughout the season. The midfield and defending pressures jump right up, though.
This suggests that Dallas coming into the team changes the press somewhat. On the one hand, it is clear that the high press isn’t as effective. With the number of pressures in the final third staying fairly regular and the deeper pressures increasing, the supposition must be that the higher press wasn’t as efficient as it was at the outset of the season and opponents are able to progress down the field further, drawing a response from the Leeds players in deeper areas.
This is undoubtedly true, as anyone who watched John Stones pick up close to 1km of progressive carries against Leeds will agree. However, there is something more to it. Take a look at this out-of-possession sequence against Crystal Palace:
As is well-known, Leeds play a man-marking system which sees them go tight to opponents when out of possession (with a couple of exceptions). At the beginning of the season, in the midfield area, the Leeds players would go man-for-man almost exclusively. However, as you can see here, Stuart Dallas is giving his marking responsibility, Luca Milivojevic, a lot of space.
As Milivojevic moves forwards, though, Dallas moves laterally to intercept him:
With Dallas putting him under pressure, the Serbian midfielder shifts the ball wide to Eberi Eze and continues forward, dragging Dallas with him:
Seeing Eze eyeing up space in the right channel, Milivojevic drops deep again, looking to pull Dallas with him to make space for his teammate. However, Dallas lets him go and doubles up on Eze with Gjanni Alioski:
Unable to proceed, Eze returns the ball to Milivojevic and again Dallas marks him in advance, ready to close him down should he move into a more dangerous area:
In the course of the season, we’ve described this as a more zonal approach to defending. Although he isn’t giving up his man-marking responsibility, Dallas is careful to avoid being pulled too far out of his zone so that he can protect the central spaces from being run through by opponents should the high press get broken.
Obviously, this comes at a cost. The high pressing isn’t quite as efficient as it was at the beginning of the season, and as a consequence, as the cumulative xG and xGA plots show, Leeds aren’t quite as productive as they were at the start of the season either.
There is no doubt that this worked out for Leeds over the course of the season. However, the question going forward is: is this the way that Bielsa wants to continue to play? Or will he look to return Leeds to their early-season iteration with the addition of a few new players and, hopefully, a fighting fit Mateusz Klich?
When it comes to defending, the overriding narrative of the early season was that Leeds conceded too many dangerous chances than was sustainable. This bears out when you look at the shot map of chances conceded by Leeds this season:
As you can see, Leeds were punished by giving away high-value chances in and around the six-yard box (the chances coloured yellow in the viz).
After the early season, the narrative took a turn. There seems to be a consensus amongst the fan base that Leeds got better at defending through the course of the season. However, the underlying numbers suggest that this defensive improvement could be as much caused by luck.
Take a look at this cumulative plot of the goals conceded by Leeds against the expected goals conceded by Leeds in the course of the season:
As you can see, Leeds’ “goals against” track the “expected goals against (xGA)” through the whole season until Match Day 27 — the 2–0 loss to West Ham — at which point, the values begin to diverge. By the time it hits the end of the season, Leeds have over-performed their expected goals conceded by about 9 goals.
What should we take from this data?
On the one hand, the suggestion that Leeds have improved defensively doesn’t seem to stack up in terms of the accumulated xGA over the course of the season. If you add a trend line to the plot, the rise of xGA is uniform enough throughout the season:
The question then becomes: why have Leeds over-performed in the latter stages of the season? At this juncture, it’s worth being a little circumspect. Whenever this sort of over-performance occurs, the temptation is always to suggest that there is an intrinsic reason why this is happening. It was only as recent as last season that people were asking whether or not Liverpool had “broken” the xG models. That argument doesn’t look so convincing now.
However, there may be indicators which could explain this trend. Here is a plot which shows the average xG value of chances conceded per game this season:
As you can see from the trendline, the average quality of the chances Leeds have been conceding has clearly been dropping through the course of the season. This could imply that Leeds have gotten better at stopping dangerous chances as the season has progressed.
However, it should be noted that the Leicester fixture in Match Day 7 has skewed the data wildly. If you remove that game from the dataset, the trendline now looks much less impressive…
There is another way to quantify the defensive trends, though: by looking at “big chances”. At the outset, a caveat. There are a lot of analysts who won’t go near “big chances” for a couple of reasons: firstly, they can be quite “noisy” — it’s not always clear what any correlations signify. Secondly, it’s difficult to define a “big chance” in the first place.
How do you determine if a chance is “big”? Generally, big chances have been defined per xG as any chance that is above a certain limit, say 0.3 xG. Of course, the caveat that always comes with xG modelling is that small sample size or individual chance xG should be left well alone. As a result, other factors are usually fed into the equation so you might want to only consider big chances when the player is one-on-one with the goalkeeper.
For the purposes of this review, let’s just take two different definitions of “big chances”. We’ll call chances greater than or equal to 0.3 xG “big chances” and then chances greater than or equal to 0.2 xG “decent chances” and we’ll track all of the shots Leeds conceded last season to see what comes out:
Once again, I’ve indicated the point at which Stuart Dallas went into central midfield. Perhaps unsurprisingly enough, that point also correlates to the point at which Leeds stopped giving up as many “decent chances” and a bit of a reduction in “big chances” also.
In the period before Dallas moved into midfield, Leeds were conceding 2.6 “decent chances” per 90 minutes and 1.3 “big chances” per 90. After that point, they conceded 1.8 “decent chances” per 90 to 1.2 “big chances”. The reduction here shows up more on the “decent chances” side. This suggests that Leeds managed to cut out a lot of the dangerous chances they were conceding in the first half of the season.
But it’s important to note what the improvement consisted of. As you can see from the plot, there are a lot of peaks pre-Dallas in midfield. Each of these peaks represents a game against bigger sides: Liverpool, Leicester, Chelsea, Manchester United and Tottenham. Beyond that, there isn’t too much of a diversity when it comes to big chances conceded.
The second half of the season improvement that we saw, then, was as much about dramatically improving in defence against the bigger Premier League sides. Once again, therefore, we are left asking the same question: how will Marcelo Bielsa negotiate things next season? Will he continue to defend pragmatically and look to catch opponents on the break? Or will he attempt to develop a style of play that allows him to attack more consistently without leaving his side open at the back?
Centre Backs: ‘A Nice Problem to Have’
It would be reasonable to argue that centre back is the position in which Leeds possess the greatest depth of quality. All four of Robin Koch, Diego Llorente, Liam Cooper and Pascal Struijk could stake legitimate claims for being first choice on their respective side of central defence (assuming a back four).
Additionally, Luke Ayling is a more than able deputy when he is needed to play there, when the 3–3–1–3 structure is used or injury to Kalvin Phillips requires Koch or Struijk to play as a defensive midfielder instead.
One problem with attempting to assess the performance of a team’s central defence is that defensive data tends to speak to how the team’s defensive system works as a whole.
Looking at the Expected Goals Against (xGA) is a good example of this. As detailed in the previous section, over the course of the season, Leeds made tactical and personnel changes which were intended to increase their defensive solidity. When the whole team sits sitting deeper and one player marks more zonally, the attackers and midfielders offer a greater degree of protection to the defence, leaving them exposed less often.
In turn, this can make the data seem to say that the centre backs are ‘defending better’. It is probably closer to the truth to say that the central defence has operated at a fairly consistent level throughout the season, irrespective of who has played there, and that the apparent greater degree of defensive solidity is actually as a result of system changes in the midfield.
This is probably most apparent when you compare the two games against Manchester United. In the first game, which was lost 6–2, Leeds tried to play in the fast, attacking style of the opening part of the season, transitioning quickly from defence to attack in numbers. One feature of this game is how quickly Leeds would empty the midfield at the first pass, leaving the defence wildly exposed in the event of possession being lost.
In this screenshot, in the run up to the first goal, you can see that Kalvin Phillips, Mateusz Klich and Rodrigo have all moved ahead of the ball, leaving Raphinha being pressed by three Manchester United players:
Once Raphinha loses the ball, the central defenders are left exposed to players running at them without the man marking structures being in place. As you can see in the image, none of Manchester United’s attackers are being marked. This was a pattern which would be repeated on several occasions throughout the game to a heavy cost.
In the home game against Manchester United, Leeds were happy to cede possession (they put up 59% possession in the away fixture that they lost but only 46% possession in the home draw) in order to ensure that the man marking structures were in place at all times.
Interestingly, Manchester United had a greater distance of progressive carries in the return fixture — 1,774 in the home fixture v 1,503 in the away fixture — but when you look at which Manchester United players were able to carry and progress the ball in each fixture, a picture emerges.
In the away fixture, they had seven players with over 100 yards (McTominay, Fred, Maguire, Rashford, Lindelof, James and Martial), but none over 200 yards .The top four ball progressors for them in the fixture at Elland Road were their back four: Victor Lindelof (477 yds), Harry Maguire (368 yds), Aaron Wan-Bissaka (198 yds) & Luke Shaw (151 yds).
This suggests that in the home fixture, Leeds were happy to let Manchester United possess the ball in deeper areas, and to keep Fernandes, James, Rashford and Greenwood off the ball and tightly marked. So while Diego Llorente and Pascal Struijk got high praise after the game, the contrast in the results is perhaps better explained by structural and tactical changes than by personnel performance.
So how do we assess the centre backs, if so much of their apparent performance is so intrinsically connected to what happens in front of them?
When watching the four players in question, differences and similarities can be seen in their defending styles. Some observations:
- Diego Llorente defends in a proactive style. He tries to win the ball as quickly as possible and wants to keep his team on the front foot. He will often try to nip in front of his opponent and steal the ball. This means that he is likely to show up well for tackles and interceptions.
- Liam Cooper is also proactive in his approach but will generally try to stay goal side of his opponent, making a tackle as soon as the opportunity presents itself. There are clear similarities to how Cooper and Llorente approach defending.
- Pascal Struijk is a positional defender, will often refrain from making a tackle and will jockey his opponent away from danger, denying space, pressuring before making blocks or intercepting passes.
- Robin Koch is also positionally focused, excellent at staying calm when players are running at him or a team mate, not overcommitting or making an attackers mind up for them. He and Struijk have some similarities in how they approach the role, but to my eye, Koch is more passive, Struijk tends to press more tightly.
These observations are backed up by the following visualisations. The first is a graph showing successful tackles per 90 minutes:
As you can see, Llorente and Cooper make the most successful tackles per 90, indicating that tackling is a very important part of their respective defensive skill sets.
I wondered how different the season might have been for the centre backs compared to last year, so I looked at the successful tackles data from the 2019/20 season:
Here, you can see that in 19/20, the Leeds centre back with the most successful tackles (Ben White) tackled successfully less often that those noted high volume tacklers Helder Costa and Pablo Hernandez, and Liam Cooper tackled successfully less often by over 1 tackle per 90:
This shows you how much stylistics considerations need to be taken into account when analysing defensive metrics. In 2020/21, irrespective of the style Leeds adopted at various stages of the season, the centre backs were required to defend more often than in the previous season.
Llorente’s desire to win the ball as quickly as possible is also evidenced in his pressing data:
Struijk is the central defender with the next highest successful pressures, showing that although he doesn’t necessarily tackle as often, he denies opponents space and time, forcing mistakes. It’s also been clear when watching this season that he has been encouraged to be more aggressive in his pressing at the campaign has worn on.
Blocks per 90 show that Koch and Struijk are both defenders who are prepared to bide their time, not commit early, stay close to the opposing striker, and act once the striker has made a decision:
You will have noted that Leeds will often cluster around opposition players around the edge of the penalty box, making sure that there are bodies between the ball and the goal. The relatively high volume of blocks for all of the regular defenders is testament to this.
The data indicates that all of Leeds United’s central backs have been required to be more defensively active than we have been used to under Bielsa. This is likely to be partly league effect — it is logical that Leeds would not be able to dominate games as consistently they did in the Championship — and partly due to the whole team operating more defensively than before.
All four centre backs interpret the role differently. There have been occasions when all have played as man markers or as the partial libero, but all are able defensive performers. In truth, though we may individually express preferences for one style or another, any two from the four players are going to do a good job defensively against most teams.
I wanted to take a moment to reflect on how well Liam Cooper has adapted to life in the Premier League. Many expected him to struggle defensively, but when you look at his defensive statistics, it is clear that he is one of the very best defenders not only at the club but also in the whole division, operating in the top few percentile for many of the key defensive metrics. Maybe we can move past calling him League One Liam, finally?
This brings us to the on-the-ball side of the position.
Much has been made of Diego Llorente’s progressive passing during his run in the first team, and it is certainly true that some of his line breaking passes have been both eye-catching and dangerous.
However, it is important to qualify this with a comparison with Liam Cooper. The following plot is from Analytics FC. Using a unique algorithm, Analytics FC are able to ascribe values to metrics which start to interpret the data beyond quantity:
Along the y-axis, you can see the quantitative number of line breaking passes per 90 made by Premier League centre backs. As you can see, Liam Cooper is putting up almost double the number of line-breaking passes per 90 minutes than Llorente is.
However, when you compare them on the x-axis — looking at the quality of those passes per 90 minutes — you can see that they are much closer together qualitatively than they are quantitatively.
This suggests that Liam Cooper and Diego Llorente both add value to their teams through their progressive passing. However, where Cooper does it through volume, Llorente does it through the quality of his passes. Both are useful. But they are very different in how they do it.
On all fronts, Cooper has performed to a very high standard for the past two seasons and it’s clear that Pascal Struijk is going to have a difficult job becoming the undisputed first choice for the left centre back position. There is a similar challenge facing Robin Koch to depose Diego Llorente from right centre back.
Leeds have four centre backs who have a legitimate claim to be starters for the first time since Marcelo Bielsa became manager. It’s a nice problem to have.
Defensive Midfield Position: Leeds With and Without Phillips
The defensive midfield position has become a major talking point throughout the season. With Adam Forshaw still injured, there hasn’t been a ‘like-for-like’ deputy for Kalvin Phillips available.
On the surface, Leeds have struggled without Phillips, losing 7 of the 10 games he didn’t start. Of those, several are down to variance. There was the 1–0 away defeat to Wolves — who registered just over half the xG Leeds did, scoring via a freak own goal — and the 4–1 away loss to Crystal Palace where Leeds (0.8 xG) registered an almost identical xG to that of their opponents (0.9 xG), yet let in 4 goals.
The Gamestate Klaxon should also be sounded for the Championship-evoking 1–0 home loss to Aston Villa, who scored early and sat back, limiting Leeds to a paltry 0.05 xG per shot. The other four losses (to Leicester, Arsenal, and Brighton (x2)) saw Leeds fall victim to a challenge rarely faced in the Championship: a well-coordinated high press.
Though Leeds have faced more pressing in the Premier League, Kalvin Phillips has unquestionably made the step-up very well. In possession, he continues to execute several key functions for the team. A phrase regularly heard from the guys at All Stats Aren’t We is that the Leeds central midfielders ‘facilitate Leeds’ build-up in wide areas’. This describes Phillips’ involvement in the build-up phase well.
The following plot compares the percentage of each player’s total carry and passing distances that have been progressive:
What this plot shows is that moving the ball forward is not Phillips’ direct role within the Leeds team. As you’d expect from both the eye test and theoretically from a team who build up in wide areas, the most progressive players are largely wingers and fullbacks.
Many people would use the fact that most of Phillips’ actions are sideways or backwards as a stick to beat him with — it’s certainly a complaint levelled by lots of England fans. However, it is by having a player who does precisely this that allows Leeds’ other players to be as progressive as they are. His excellent off-ball movement means he is regularly available to receive passes from Leeds’ defenders, which he then moves back to a centre-back or out wide to a full-back.
This simple recycling of possession helps open spaces and angles for his colleagues to drive or pass into. Phillips registers 23.09 defensive third touches per 90, putting him in the 96th percentile for midfielders in the top five European leagues, demonstrating how active a role he plays in this part of the game.
Phillips’ movement doesn’t just aid the players behind him — by dragging opposition players away who mark him, he opens up central space for Leeds’ 8s, most notably Mateusz Klich or Stuart Dallas, to drop into.
This kind of movement will again result in other players’ progressive numbers increasing, whilst Phillips’ movement is harder to translate numerically. As well as a useful part of the toolkit for build-up generally, this positional rotation is an important way for Leeds to deal with the high press.
If oppositions leave Phillips unpressured, then his other key attribute comes online: his long passing.
Throughout the season, Phillips has demonstrated a high level of technical proficiency with long passes, and his 11.9 completed long passes per 90 put him in the 96th percentile in Europe. These long passes are enough for him to rank in the 78th percentile for progressive passing distance, despite the fact that only 25% of his passing yards go forwards.
Phillips excels at quickly releasing these longer passes, which is of particular utility in quick transition from defence to attack:
When Leeds possess the ball higher up the pitch, the defensive midfielder acts as a ‘pivot’, remaining deeper and central as a reference for the other players, and moving the ball from side to side. Phillips does this both well and often, completing 3.97 switches (passes that travel more than 40 yards of the width of the pitch) per 90, putting him in the 96th percentile for midfielders in the top five leagues.
When he’s given space and time on the ball — as he most notably was in the away game against Everton — his quick and accurate long passing is able to find teammates in space and create danger. A better touch from Jack Harrison here following Phillips’ long pass would have likely resulted in a Leeds goal:
While the defensive midfielder plays an important role in possession, it is just as demanding out of possession. The player in this role will usually be tasked with man-marking the opposition’s most advanced midfielder. Additionally, the defensive midfielder is expected to be proactive in dealing with opposing transitions, to give Leeds’ other players time to take up their own marking responsibilities.
Kalvin Phillips is a very active defender and suits both responsibilities very well. On the latter point, Josh Hobbs recently observed a trend of his defensive proactivity creating attacking opportunities:
Unsurprisingly, Phillips sits in the 84th percentile for defensive actions that lead to a shot attempt.
More broadly, Phillips’ defensive output is huge. He can be found in the upper percentiles of most metrics. This visualisation, from StatsBomb, maps his defensive actions for the season:
Though his man-marking brief means he has defended all over the pitch, Phillips has in particular guarded that central space in front of the box very well. Marcelo Bielsa often refers to the DM as a ‘third defender’ — Kalvin Phillips has proved he has the defensive chops to merit the label.
This categorisation likely explains why Bielsa tends to move a centre-back forward rather than a midfielder back in Phillips’ absence. Mateusz Klich’s inability to defend Youri Tielemans’ advances was a key factor in why Leicester were able to create such dangerous chances against Leeds in their 4–1 victory at Elland Road. Now imagine Klich matched up against Bruno Fernandes or Martin Ødegaard…
Neither Klich or Robin Koch made many appearances in the position, making statistical analysis with Phillips potentially unreliable. Purely from the eye, Klich plays the role well in possession but struggles with the defensive demands, whilst Robin Koch’s two performances were encouraging with and without the ball. The German will likely prove to be a useful backup for Phillips going forward, as has Pascal Struijk.
The Belgian has played around 4.5 games’ worth of minutes in the position, which while still small, is varied enough to allow us to at least compare him to Phillips on a per 90 basis. When the players’ numbers are compared, it is striking how similar they are in many metrics, though Phillips has slightly more touches and passes per 90.
Despite what we’ve said above about ball progression not being Phillips’ direct responsibility, this is actually where he stands most apart from Struijk statistically, outputting 134.21 yards of progressive passing more per 90 than his colleague. There’s a few factors at play here:
- Phillips plays more passes than Struijk, attempting 6.22 more per 90.
- He also plays more long passes than Struijk, attempting 16 to Struijk’s 9 per 90, and 3.97 switches per 90 to Struijk’s 2.4.
- To the eye, Struijk seems more risk-averse than Phillips in the role. He’s more than capable of hitting the long balls that Phillips does, but he is more likely than the Englishman to play a lateral pass in these scenarios.
- There is likely a confidence element to this. In interviews he has often spoken about his preference to play at centre-half and how the role in midfield is something he is learning, which would make risk-aversion understandable. Again to the eye, he has appeared more confident in the role in his recent appearances there, particularly in his 45 minutes against Southampton on Match Day 37.
In deeper build-up, the players seem statistically closer-matched. To the eye, Struijk seems to make many of the same movements that Phillips does, though he’s less able to turn on the ball quickly.
Again his confidence in playing the role should be factored in, along with his natural left-footedness, which has minor implications for the passing angles he finds and runs he makes. In these respects, Robin Koch slots more naturally into Phillips’ position than Struijk.
In most defensive metrics, the two players have virtually identical numbers, with the main difference between them being stylistic. The one area where Struijk stands out from Phillips is his success in tacking dribblers. In these scenarios, the England man is successful in 30% of his duels and is dribbled past 2.1 times per 90, whereas Struijk successfully tackles the dribbler 62.5% of the time and is only dribbled past 0.7 times per 90 (do remember his small sample size though!)
This is again an effect of the two players’ styles, and there are pros and cons to both.
- As a more reactive defender, Struijk won’t step in to make the challenge too quickly, but will look to move the opponent into less dangerous places. If he does tackle, it often comes later in ‘safer’ areas, evidenced by how he makes almost 10% more of his tackles in the defensive 3rd than Phillips does (67% vs 58%). However, regaining the ball deeper means that Leeds’ attacking transition will be less effective: Struijk has registered 0 shot-creating actions from defensive actions all season.
- Phillips’ proactive style is fundamentally more ‘high-risk, high-reward’, by engaging and attempting tackles earlier and higher up the pitch. If he gets something wrong, it may open Leeds up to a more dangerous attack, but if he wins it back then Leeds have serious potential in attacking transition.
Struijk is not without his flaws in the defensive midfield position, but his versatility has plugged a hole for Leeds this season that otherwise had nobody to fill it, and like Koch, he will remain useful backup going forward.
There are some games where a player permanently occupying the defensive midfield role is redundant: those where there is no opposing player in that space to mark.
A template for this was observed in the 1–0 loss away to Wolves, where Klich and Jamie Shackleton played as two 8s, one of whom would drop in to facilitate the buildup depending on which side of the pitch the ball was on. The unadventurous nature of Moutinho and Neves meant that Klich and Shackleton were not defensively stretched beyond their capabilities, and will provide Bielsa with food for thought going forward against teams whose midfields sit deep.
Illan Meslier: A Vertiginous Rise
Although he’d played 9 games by the end of 2019/20 and had a season of senior football under his belt at Lorient, Illan Meslier came into this season an unknown quantity. How would he adapt to the rigorous test of Premier League football?
As it turns out, very well indeed. Over the course of the season he has made any debate about who is Leeds United’s number 1 goalkeeper completely redundant.
Saves like this one, early in the season, helped to settle any doubts about his ability to make the step up, and meant that we were thankfully spared any meaningful debate about whether Kiko Casilla should be playing ahead of him.
As you would expect of a 20-year-old in their debut season, there have been occasional mistakes in shot stopping, possession and in dealing with crosses. But overall, this is a story of a player adapting to the standard with the minimum of fuss. His calm and measured approach appears to allow him to forget mistakes quickly and to learn from and eradicate them, which has not always been the case with Leeds United goalkeepers, has it?
There have been notable examples of bad luck leading to goals too, such as the own goal at Wolves, or this bizarre Helder Costa own goal in the 4–1 defeat at Crystal Palace, where a minor recurring technical issue with weight distribution reared its head, in combination with an impossible-to-predict deflection:
Over the season, as noted in sections above, the team has given up a lot of chances and conceded a lot of goals — 1.49 per 90 minutes — which gives credence to an old goalkeeping adage: bad defending makes bad goalkeepers!
This could lead one to believe that Meslier’s shot stopping is an area of weakness. He faced 177 shots on target — joint third highest in the division — and has a save percentage of 75.1% — the fourth best.
As a result, Meslier shows up well in John Harrison’s shot stopping graphic below:
The graph shows that Meslier saves more shots than the hypothetical average goalkeeper. Whilst there is room for growth and improvement, he appears above some notable names: Jordan Pickford, Kasper Schmeichel, David De Gea and Edouard Mendy.
In terms of his distribution, Meslier excels. This is most notable when considering his longer-range distribution. ‘Launched balls’ are passes over 40 yards. Meslier attempts 11.7 on average per 90 mins at a success rate of 44.3%: the fifth best success rate in the division
In terms of his shorter-range distribution, there have been occasions in the season when Meslier’s calm and confident style in possession has put the team under pressure or led to chances being conceded.
For instance, against Arsenal at the Emirates, he gave away a penalty after being pressed by Bukayo Saka while looking for a pass:
After this incident, he seemed to take fewer risks with his distribution and seemed to be happier playing the ball longer. It will be interesting to see whether this continues next season.
Another area which has seen him criticised is in the air from crosses. Meslier stopped 10.6% crosses into penalty area across the season. This doesn’t sound so impressive until you realise that puts him second in the division behind Nick Pope.
There is a narrative that Meslier became more proactive when coming for corners around the time of the Burnley home match. However, John Harrison has been tracking which goalkeepers help their team defend corners most effectively across the season. Here is the most recent update:
Meslier has been in the ‘More frequently attempts to deal with & rarely fumbles’ quadrant all season, in much the same position. This indicates that over the course of the season he has consistently been proactive and secure from corners. The issues the team have had from corners and the apparent improvement in this area do not seem to be connected to differences in the way Meslier plays.
Overall, this was a season Meslier can be proud of and one he can build on. I look forward to continuing to see him develop over the next few years. He undoubtedly has the potential to be one of the very best goalkeepers around.
Luke Ayling: The Glue Holding Leeds Together
After Stuart Dallas, it could be argued that Luke Ayling has been the glue that has held Leeds together this season.
Ayling managed to adjust to the level of the Premier League fairly quickly, even if the Ayling flop took a while to get going! Not only has he been a key component of both the attacking and defensive play; he has filled in at multiple positions when required.
Most of his minutes have come at right back, but he has also played on the right of a back three to great effect. He even played at left centre back against Aston Villa.
A mark of how successful his season has been, Ayling even found himself name checked by Gareth Southgate as someone who is on the fringes of the England squad.
Focusing on Ayling’s defensive metrics first, the two areas where he has really stood out are successful tackles and blocks. Looking at the data on FBref.com, he ranked 4th in the league for both overall with 67 Tackles Won (Pierre-Emile Hojbjerg led with 81) and 81 Blocks (Aaron Wan-Bissaka led with 113).
The two graphics below show how Ayling compares to his teammates in Successful Tackles and Blocks per 90 minutes.
It can be seen in the graphics that Ayling compares well to his teammates for both metrics. He ranks third in the squad for Successful Tackles per 90 minutes and fourth in the squad for Blocks per 90 minutes. In both cases Ayling is not too far off the top spot. (It is not possible to compare his numbers per 90 to the rest of the league as the data is not possession adjusted.)
If we focus now on the attacking side of Ayling’s game, the area where he has really excelled this season is his ball progression. According to FBref, Ayling ranked 2nd in the league amongst outfield players for Total Progressive Passing Distance this season with 17,225 yards, and 7th in the league with 203 Progressive Passes (Trent Alexander-Arnold led both categories with 278 Progressive Passes for 19,500 yards).
Ayling still ranked well when Progressive Passing Distance was adjusted to per 90 minutes as can be seen in the graphic below, as he ranked 6th in the league for players with over 700 minutes:
Ayling’s progressive dribbling was also impressive last season. He ranked 9th for Progressive Carries with 232 carries per 90 minutes (Jack Grealish led with 306) and accumulated the largest Progressive Dribbling Distance in the league with 8,246 yards. However, he did drop to 11th when this was adjusted to per 90 minutes as you can see in the graphic below:
Leeds’s build up is focused on getting the ball to the full backs who then progress the ball forwards by passing to the winger or one of the 8s. It makes sense, then, why Ayling’s progressive numbers are so good.
Leeds have focused their build up less down the right this season compared to the previous seasons in the Championship, so it makes sense why Ayling’s numbers are better than those of Gjanni Alioski who was the primary left back last season.
Ayling does have the advantage that he has played as a centre back at times and so has had more opportunities to progress the ball forwards. This is most evident when Ayling has played on the right of a back three, as this allows him much more space to pass and dribble out. He is also much more comfortable under pressure than Alioski.
One feature of Ayling’s play that helps him to deal with being pressured is his ability to take on and dribble past opponents. Looking at the graphic below from @GoalAnalysis on Twitter, Ayling was one of the players with the Most take-ons completed by zone in the big five European Leagues:
An example of Ayling’s ability to beat a man then progress the ball taken from the home game against Southampton can be seen below:
Ayling is pressured by a Southampton player, but manages to turn away before dribbling into the space before playing the ball to Tyler Roberts. Unfortunately, once the ball is returned to him by Roberts, the second pass goes out of play.
Assuming that Leeds buy a new left back over the summer who is more comfortable in build-up, it will be interesting to see what effect this has on which side Leeds focus on during build-up, as well as Ayling’s progression numbers.
Ayling’s ability to progress the ball has been crucial to Leeds this year, though, even at the higher level in the Premier League. Hopefully, he can keep it up next season.
As we have already suggested, Leeds xG per 90 did tail off a little as the season went on and Leeds made some adjustments to team structure. However, they still put up phenomenal attacking numbers.
Their 62 goals was the 6th highest in the league. On top of this, 13 different players contributed goals, with Bamford, Dallas, Harrison, Rodrigo and Raphinha all scoring more than 5 non-penalty goals each. It’s always important to spread responsibility for goalscoring so the team’s goals don’t dry up if the top scorer goes cold in front of goal. Leeds managed to have regular contributions from the midfield and wings whilst having one of the top scorers in the league in Bamford.
More than that, though, Leeds’ underlying numbers for chance creation were some of the best of the league, which is good news for next season as it suggests their high-scoring this season wasn’t just lucky.
As you can see from the table below, taken from FBref.com, their non-penalty xG of 54.5 was only bettered by Chelsea, Manchester City and Liverpool.
Penalties are often taken out of the equation in football analytics due to the fact that the nature of them being won isn’t as repeatable as other types of chance creation. However, they are a part of the game and some would argue that they should be included.
If we did include penalties, Leeds are usurped in the top four for xG by Manchester United and Leicester shoot up to right behind them. Nevertheless, Leeds are still mightily impressive:
When penalties are considered, we can see that Leeds did slightly over-perform their overall xG for the season, scoring 4 more goals than would have been expected. However, 2 of those were own goals. An over-performance of 2 isn’t something to worry about in terms of regression in the future.
We’ve touched on Mateusz Klich’s drop off in terms of his shot-creating actions, which was one of the reasons why there was a slight levelling out of Leeds’ attacking productivity as the season went on.
Fbref.com defines shot creating actions as the two actions immediately leading up to a shot — be that a pass, dribble, foul won or even a defensive action. This metric allows us to give value to more than the ‘key pass’. In the early part of the season the Pole was one of the top performers in the league in this regard but he ultimately dropped down to 24th for SCA per 90.
Looking at the top 10, though, we can see why Leeds’ attack was still able to function in the way it did even without Klich’s output:
Leeds have two players in the top ten in Rodrigo and Raphinha. It’s worth noting that Rodrigo’s end of season flurry of attacking activity in a limited amount of minutes has boosted him up here but he has been in and around the top 10 for this metric throughout the season.
Whilst both of them have had spells out injured — Rodrigo in particular struggling with his fitness — they’ve added plenty of flair and creativity to the team, which had previously been very reliant on Pablo Hernandez for attacking output. With Hernandez’ influence sadly disappearing in 20/21, it was vital that these two new signings picked up the slack, which they did.
As we have alluded to several times elsewhere, there have been question marks over Rodrigo’s pressing ability. Do his incredible talents on the ball make up for the fact that his presence in midfield can mean that Leeds’ opponents have a free pass through the middle of the pitch? One thing is for sure is that when Bielsa is able to get the combination of the two together on the field, his side are at their most dangerous:
Leeds’ attack has been largely transition-based this season. They haven’t dominated the ball in the way they did in the Championship, dropping from an average of 62% possession to 54%.
This has meant getting the ball forwards quickly and exploiting the wings. Rodrigo’s ability to receive the ball in midfield, swivel and find Raphinha makes him perfect for transitional attacks. Both players will get more detailed sections of their own later in the piece, but I felt it was necessary to mention them in this attacking overview.
Regarding the transition based attacks, these two graphics from Opta’s ‘The Analyst’ review of the season shows how high Leeds rank in ‘direct speed’ (a measure of how quickly a team progresses the ball up the field in metres/second):
And in ‘direct attacks (the number of open play sequences that starts just inside the team’s own half and has at least 50% of movement towards the opposition’s goal and ends in a shot or a touch in the opposition box):
These kinds of attacks can see Leeds give the ball away too often and too easily at times, but they have also been devastating.
Rodrigo’s first goal against Burnley would be an example of one of these direct attacks:
Pascal Struijk started the possession just inside the Leeds half with a header and Harrison turned and played it straight to the Spaniard who finished spectacularly. Direct doesn’t just mean route-one football. It can be beautiful.
In terms of overall chances created, look at the shot map below:
As you can see, a high amount of Leeds’ goals have been in the prime area for high value xG chances. This being the central area between the goal and the penalty spot. Raphinha’s ability to get to the byline and make cutbacks has been a real source of these types of chances. If Leeds keep creating chances in this area going forwards, they will continue to score regularly.
However, it should be noted that Leeds have scored more goals from outside the penalty area than any other team: 12. This may not be sustainable in future and if this took a downturn next season it might mean that they underachieve xG, rather than overachieve as they have this season — given that scoring 12 from outside the box and only overachieving xG by 1.5 suggests a level of underperformance in finishing inside the box.
Fortunately, Leeds haven’t scored a lot of long-range goals whilst also failing to create chances inside the box. Perhaps the finishing might go cold a little next season but as long as the creativity remains at a similar level, there shouldn’t be a huge drop off in terms of goalscoring.
I would suggest that one of the reasons Leeds have managed to overachieve so much in terms of shots from outside the box is due to the transitional attacks that were mentioned earlier. Shooting from outside the box against a team with their defence scrambling back into position is a lot easier than shooting from outside the box against a low-block defence.
There are also anomalies like Stuart Dallas’ winner against Manchester City, which was taken from outside the box but was a one-on-one with Ederson a long way from his goal, making it much easier to finish successfully.
Wingers: The Key to Marcelo Bielsa’s Attack
Wingers have always been an incredibly important part of Bielsa’s Leeds team, and this season was no different.
Jack Harrison played the most time as a winger for us, totting up an impressive 3,012 minutes in total (only bettered by Dallas, Ayling, Meslier and Bamford when you include every position). Raphinha clocked 2,532 minutes in total which won’t surprise many people, given that Harrison on the left and Raphinha on the right was a common occurrence this season. Helder Costa also contributed 1,156 minutes, with Ian Poveda’s cameos amounting to just 355 minutes.
This section will largely focus on both of Leeds’ ‘first choice’ wingers, Jack Harrison and Raphinha.
The Defensive Side
For the majority of the season, Harrison was played on the left and Raphinha on the right. However, occasionally we saw both players swap, usually with defensive man-marking in mind.
Across the season, each player clocked a small portion of minutes on the opposite side: Harrison with 208 and Raphinha with 461. To understand why this switch happened, we can first observe both players’ defensive numbers:
Even by just glancing at this graphic, it is clear there is a vast difference between both players’ defensive numbers.
The standout statistics are tackles (for which Harrison is in the 94th percentile compared to other wingers, yet Raphinha is only in the 60th percentile), dribbled past (56 to 23 in Harrison’s favour), and pressures (86 to 46, again in Harrison’s favour). In fact, the only defensive statistics for which Raphinha has better numbers are dribbles contested, blocks, passes blocked, interceptions and clearances.
While these might be important parts of any players’ defensive game, Raphinha’s clear weakness in successful pressures makes Harrison by far the more reliable presser and the player which many managers would rather see man-marking a strong attacking player.
For this reason, Harrison’s minutes on the right hand side were usually deployed against teams who have a much stronger left back than right back.
For example, against Everton’s 3–4–3 at Goodison Park, Harrison played on the right to try and nullify the threat of the direct and pacey Alex Iwobi, who dropped into wingback for this match.
Meanwhile, Raphinha was left with marking the more defensive minded Tom Davies, as well as helping Bamford press Everton’s back three. Raphinha’s lessened defensive responsibility, and therefore heightened attacking freedom, ultimately led to Leeds’ only goal, where Raphinha had drifted into the centre of the pitch to pick the ball up, and drill a 25-yard shot past Jordan Pickford.
The Attacking Side
While we know that Harrison is stronger defensively, we can also observe how Leeds’ most played wingers differ in attack to understand why — despite both players being left footed — by far the most common scenario was to have Jack Harrison on the left and Raphinha on the right.
Firstly, it is worth mentioning that both players have absolutely fantastic numbers in terms of expected assists. Both players make the top ten of Premier League performers for xA, as shown in the graphic below:
A look at the other names on this table illustrates how good both players’ numbers have been this season, in order to cement their place alongside some of the best footballers in the league.
When we look at a breakdown of these assists, however, alongside each player’s open play chances created this season, the difference between the two players becomes clearer:
There are a few things that stand out from these graphs. The most notable difference is the area from which each player is creating chances. The chances Harrison created are for the most part generated from the final third, high up on the left hand side, and largely outside of the box. This tells us that when Harrison receives the ball in this area (or drives into it carrying the ball), he is looking to put dangerous crosses in more often than not.
Many Leeds fans will be familiar with the feeling that Harrison had a tendency to hold the ball too long in this area in the Championship, but this season he has been greatly rewarded for showing a willingness to put the ball in the box quickly. His fantastic assists against Newcastle and Fulham are indicative of this, and are shown below.
Raphinha’s chances created tell us a different story. As shown on the previous graphic, across the whole season, he has no chances created at all from wide in the final third (the area from which Harrison is so prolific).
Instead, Raphinha prefers to take on the defender in that situation, and thrives on winning 1v1s to get into the box. Around a quarter of his chances created this season have come from cut-backs inside the penalty area, usually after he has beaten a defender in a 1v1.
To further illustrate this element of his game, we can look at his attempted take-ons:
As shown above, Raphinha’s highest volume of take-ons is in the very area from which he lacks any open play chances created, filling in the gaps rather satisfyingly.
This is the opposite to Harrison in some senses; when Raphinha receives the ball in this area, his first instinct is to beat the defender, using his ball skills and extreme agility to get to the byline and play a pass back across goal, or to cut in on his left foot and create danger from the half space.
An example of the former is shown below in a clip from the Southampton game:
While the latter is illustrated by his fantastic goal at the Hawthorns:
It is a useful thought experiment to imagine each player in the other’s role in the above clips.
Firstly for the Fulham/Newcastle goals: if Raphinha receives that same ball on the left, does he drill it in with the same aggression or does he touch it down to take on his man?
Likewise with the West Brom goal/Southampton chance: if that’s Harrison on the end of those passes, does he drive at the defenders with the same agility or does he try to whip a quick cross into the box?
Each player has a clear strength in these areas and this balance achieved from their differences — in addition to some important goal contributions from both players — is most likely why they established a starting eleven place so often.
Costa and Poveda
It felt slightly unfair to complete the wingers section of this article without touching on Helder Costa and Ian Poveda. Unfortunately, both players haven’t made as much of an impact as they might have hoped this season.
Costa’s minutes came largely at the start of the season, when Leeds had yet to sign Raphinha. Whilst he contributed a few goals and assists here and there, you wonder how happy Leeds’ former record signing will be seeing two other left footed wingers ahead of him in the pecking order.
Ian Poveda completed 355 minutes for Leeds this season, although he is yet to start a match, register a goal or an assist in the Premier League. With Poveda, I feel that given his young age of 21, he will be happy to stay on next year, although he will surely be looking to play significantly more minutes in order to progress his career to the next level.
Patrick Bamford and *That* Article
One thing that was absolutely vital for Leeds in stepping up to the Premier League was that they would need to be more clinical in front of goal than they had been in the Championship.
According to StatsPerform, Leeds won the Championship in 19/20 despite underperforming their xG to the tune of 14.3 goals. Patrick Bamford’s almost unprecedentedly bad season in front of goal was a huge contributor to this. He finished the season on 16 goals, which would look good enough if you didn’t know that his xG was 24.86.
Fortunately for him, the Premier League seems to have suited him a lot better, as we had hoped (come on, you didn’t think we wouldn’t bring that up?!). He ended the season on 17 goals, making him 5th top scorer in the league.
If we discount penalties, his 15 non-penalty goals still only rank 5th but he leapfrogs Bruno Fernandes, whilst Calvert-Lewin overtakes him with 16. However, top scorer, Harry Kane’s 19 non-penalty goals is a haul of only 4 better than Bamford, an incredible performance from the ex-Boro man, considering the questions around him before the season began.
The table below uses coloured formatting to show the best and worst performers of all players to score a minimum of 10 goals this season:
As you can see, Bamford is one of the top performers for shots per 90, distance to goal when shooting (the closer the better, of course), and non-penalty xG, with only Harry Kane and Mo Salah ranking higher than him for this metric.
The final column on the right — ‘npG:-xG’ — is essentially a measure of how each player has finished against their expectation. Bamford is highlighted in light red here, so he’s on the lower end of the scale. But it’s important to note that he’s not even a full goal behind his expectation for the season and isn’t far from Mo Salah in his performance here.
Players like Jamie Vardy, Sadio Mane and Pierre-Emerick Aubameyang have been known as clinical finishers throughout their careers and have performed far worse this season, whilst Ollie Watkins finished much better than Bamford in the Championship last season, only to be the worse of the two this time around.
As his shot map shows, Bamford takes the vast majority of his shots in the ‘Golden Zone,’ which is the area the width of the six-yard-box, up to the edge of the penalty box:
This is the area from which a player is statistically most likely to score from, so to see his shots so well concentrated there goes a long way to explaining why he has performed so well in terms of expected goals, at least.
When it comes to finishing chances — as Leeds have been more transitional and less possession based in their attacks — Bamford was able to move into more space to finish, rather than having to react instinctively to lots of crosses in packed boxes, which was one of his main struggles in the Championship.
His goal against Leicester was the perfect example of a player finishing with confidence:
As Raphinha’s pass came to him, he allowed it to run across his body onto his left foot and fired an unerring finish into the top corner. In the past, he might have taken the ball with his back to goal and tried to pass to Mateusz Klich, but there was only one thing on his mind.
Additionally, he weighed in with 7 assists, taking his total goal contributions to 27. He also won 3 penalties, being fouled in attempting to go round the goalkeeper for 2 of them. Bamford’s contribution to Leeds’ attack this season was spectacular and far exceeded even the most optimistic about what he could achieve in the Premier League.
One thing that wasn’t in question for the centre forward was his ability to lead the Whites’ press. Once again, he’s done that tirelessly this season and he ranks in the 81st percentile of all strikers in Europe’s top 5 leagues for total pressures per 90 (17.85), the 93rd percentile for final third pressures per 90 (9.13) and the 80th percentile for midfield third pressures per 90 (7.7).
As becomes all the more stark whenever anybody other than Bamford leads the line, it’s not just Bamford’s work-rate when pressing, but his intelligence in knowing how to press effectively:
In the example above, observe how Bamford takes a couple of steps forwards before going to press Eric Dier. This ensures that Toby Alderwiereld isn’t a passing option. He then curves his run as he moves toward Dier, cutting off the option of him passing centrally to Pierre-Emile Hojbjerg. Dier is forced long and Leeds win the turnover.
This is the kind of work that goes totally under the radar and it’s far more than just running about a lot, as can sometimes be the perception of pressing. It takes a lot of intelligence and skill. Not everybody is a natural at it, which leads us onto our final player we’ll be profiling in our review.
Rodrigo: ‘Only Time Will Tell’
The Spaniard arrived to great fanfare this summer and became Leeds’ record transfer at a cost of £30m. He finished a season interrupted by COVID and injuries with a total of 7 goals and 2 assists at a rate of 0.49 and 0.14 per 90 respectively.
Soon after he signed, we published the piece ‘How do you solve a problem like Rodrigo?’ This touched on the fact that Leeds had just broken their transfer record for a player that seemingly didn’t fit the style we were used to seeing and he certainly didn’t seem to be a direct swap for Bamford. The proposal at the end of the piece was that he would be best off playing as the most attacking of Leeds’ midfielders rather than as the number nine.
As it turned out, that role is the position which we have most often seen Rodrigo, with all his starts coming in that position, apart from the final game of the season where he started as the centre forward:
Compare and contrast his shot map with Bamford’s and you see the difference. Bamford’s efforts are far more concentrated in the centre of the goal and in the penalty area, where Rodrigo has taken 35% of his shots from outside the area.
When he’s in that attacking midfield role he loves to shift the ball quickly out of his feet from the right-hand side of the box and shoot for the far corner:
This example against Arsenal would have been a goal of the season contender if it had come down an inch or so lower.
Although it didn’t end in a goal, Rodrigo’s elite attacking qualities were on show here. He had the confidence to take the ball despite having four Arsenal players in his vicinity and then the swivel of the hips and the drop of the shoulder created the perfect amount of space to strike the ball. The technique of the shot generated enough power and whip to have Bernd Leno rooted to the spot.
As mentioned in the attacking overview, from the attacking midfield role he had the beginnings of a devastating partnership with Raphinha and was perfect for Leeds’ transition based attacks as nobody else in the team has the same ability as him to receive the ball on the half-turn and play forwards.
After picking up a groin injury away at Leicester, he was out of the squad for 6 games, before returning as a substitute, only to aggravate his injury and miss another 3 games. With these absences and his 2 games missed due to COVID, he missed 12 games, which disrupted his momentum badly.
When he did return to fitness late in the season, he largely played as a substitute for Bamford, starting the last two games, once as a midfielder and once up front.
In the Burnley game, we got a glimpse of the different qualities Rodrigo could bring in the centre forward position. His link-play was a key part in Leeds winning the corner from which they scored the second goal and he scored the third and fourth goals himself — his first was shown earlier in the piece and his second demonstrated that he could offer something different for Leeds in terms of pace in behind should he get more minutes at centre forward next season.
In the win at Southampton, he picked up a delightful assist with a clipped ball over the defence for Bamford to finish. His attacking qualities are un-questioned.
However, the issue with Rodrigo in both positions is that his pressing has not been up to the level required. It’s difficult to directly compare him to teammates due to the fact that he has split his minutes between two different positions which would skew his pressing numbers slightly. However, we can probably best compare him to Tyler Roberts.
In terms of pure volume of pressures, Roberts averages less than 1 more per 90 than the 30-million-pound man, with Roberts putting up 17.53 to Rodrigo’s 16.61. However, Roberts seems to be far more effective as a presser. This may be due to the fact that he is more familiar with the role having played for Bielsa for three seasons now. However, there are definite issues in Rodrigo’s pressing technique, rather than just his familiarity with his role.
The intensity of his pressing in this clip is low, but that’s far from the problem:
In the example of Bamford’s pressing, cutting off passing lanes and the curved run were mentioned. Here Rodrigo does well to turn James Ward-Prowse back by cutting off his initial options but it then falls apart.
In continuing to chase Ward-Prowse back into his own half, he ran in a straight line behind him and reacted too slowly to the England midfielder releasing the ball to Jannik Vestergaard, who was able to simply sidestep past Rodrigo and carry the ball into the gaping space ahead of Leeds’ defence.
Rather than continuing to chase Ward-Prowse, he needed to adjust and at least position himself so that Vestergaard couldn’t progress from half-way. It could be argued that in a man-marking system it’s the Spaniard’s job to follow him. But the system doesn’t absolve players of any decision making in the moment and he could have followed Ward-Prowse with a run that curved behind him, meaning he could react quicker to the pass. Pressing in a straight line will almost always lead to being bypassed.
In the period in which Mateusz Klich seemed to drop off from his top level, the midfield pairing of himself and Rodrigo became a nightmare in this regard, as teams would carry the ball through it at will. The 6–2 loss at Old Trafford was particularly chastening and a change came a few games later, with Dallas coming in for the Pole. Whilst Klich was dropped from the starting eleven for a while with Dallas and Roberts becoming the starting pair, Rodrigo didn’t lose his place until he injured his groin.
As there are obvious issues with his pressing, it’s worth asking the question of whether Bielsa cares as much about it as we have always presumed that he does. In Rodrigo’s case, he seems to think that his ability on the ball is so good that he can take the hit of him being a poor presser. There is a theory that he might improve his pressing in his second season in the way that Patrick Bamford did but only time will tell.
For a first season back in the Premier League after an almost-two-decade absence, there are very few areas to aim any criticisms of Leeds United or Marcelo Bielsa.
The story of the season has been a tactical shift necessitated by an injury to a key player and no doubt the summer transfer dealings at Elland Road will look to ameliorate these problems next season.
Looking forward, the questions concern how Leeds United progress from this point. In that sense, the 20/21 season has been as good a stepping stone as any to a more stable life in England’s top division.
However, there have been oddities to this season. As the footballing world settles down to something closer to normality in the wake of the Coronavirus crisis, the Premier League could be a very different prospect next time around.
One thing is certain though: Leeds United have proven that they belong in this division. Long may it continue!