In this long read, Jon Mackenzie has a think about how Leeds could go about planning for the future…
In many respects, the 6–2 loss to Manchester United can be read as a watershed moment in Leeds United’s season. Before that, there had been some heavy defeats — the 4–1 result against Leicester and Crystal Palace, and a 3–1 loss to Chelsea — as well as equally heavy wins — most notably, the 5–2 win against Newcastle as well as the 3–0 victory at Villa Park.
For the most part, those results were shrugged off by Marcelo Bielsa and it isn’t hard to see why. By the time the fixture against Manchester United swung around, a clear pattern was emerging, as Mark Thompson has shown in this graph:
Leeds were performing well in the Premier League up to a certain point. But then, once they reached a certain calibre of opponent, their performances dropped off sharply.
At All Stats Aren’t We, we put this down to the man-marking system, arguing that, against sides of a comparable quality of player, Leeds would have the edge, but against teams with better players, that edge was quickly lost. We thought this would simply carry on all season: Leeds easing past lower-table sides and taking the odd hammering from bigger sides.
But after the Manchester United game, something changed. And that change was personified in the figure of Stuart Dallas. You might be surprised that, in the first 14 games Leeds played in the Premier League this season up to and including the Manchester United game, Dallas only started three of them in central midfield. That’s 21% of the games he played. After the Manchester United game, he has played a subsequent 17 league games of which 11 of them have seen him starting out in central midfield. That’s 65% of the games he played in that period.
What impact did that have? Well, in real terms, Leeds saw a slight uptick in Points per Game, going from 1.21 ppg up to the Manchester United game to 1.6 ppg since then (although the cynic in me notes that without the Manchester City result, that drops down to 1.4 ppg so there is a bit of noise here). But perhaps more interesting was the impact that this switch has had in terms of the pressing output.
At points this season, we have highlighted how Stuart Dallas is the lowest volume presser in the Leeds midfield. His 15.2 pressures per 90 minutes are well below, for example, Mateusz Klich’s 21.4. But we also suggested a reason for this: Stuart Dallas is now playing much more zonally in the midfield than we have seen in a Leeds midfielder before. No doubt this stems from a perceived ‘soft underbelly’ in Leeds’ press which saw them being broken through in central spaces with alarming regularity.
But if you thought that this would mean that Leeds’ pressing numbers dropped, you would be wrong. In fact, they went up. Here’s a graph that shows a five-game rolling average (to reduce the noise) of Leeds’ Pressing Actions and Successful Pressing Actions:
The grey line indicates the point at which the Manchester United game enters the data. As you can see, after an initial drop-off, both Pressures and Successful Pressures trend upwards and end up higher than they did at the beginning of the season.
Obviously, much of this is game-dependent. That upward trend is prompted by a couple of high-pressing games against Leicester (233 pressures) and Everton (229 pressures). And in case you are tempted to correlate high pressing with winning, the Leicester game was a win and the Everton game was a loss. The other games in which Leeds have had more than 200 pressing actions in a game were three losses (Liverpool away, Crystal Palace away and West Ham at home) and two other wins (Southampton home and Fulham away).
But when you interrogate the data further, you can see that, through time, the ‘success’ of the Leeds press slowly drops over time (I say ‘success’ in scare quotes because there is a very specific definition of what constitutes a successful pressures: any pressure which results in the turnover of the ball within five seconds of the action). Here’s the five- and ten-game rolling average of pressure success as a percentage:
Now the trendline is quite obviously downward. The question is: what does this mean? It’s important to note how we’re not interested in ‘Pressure Success’ as an indicator of a successful press — not all successful pressing will result in a turnover five seconds after every action. What we are interested in is the suggestion that Leeds’ style of play has changed at some point this season.
We’ve already offered a hypothesis as to why this might be the case — Dallas as a zonal midfielder — but there is a longer-term question that needs to be raised: how should we read this style-change? Is it idealistic? Or is it pragmatic? Which is to say: is this style-change something that Marcelo Bielsa wants to maintain going forward? Or is it a temporary fix that could be overhauled in the summer transfer window?
At this juncture, then, Leeds United seem to be at a fork in the road. Of course, it seems unlikely that the idealistic Marcelo Bielsa will suddenly re-invent himself as a less aggressive, more zonal theorist. In a recent interview with Sky, he did have this to say about the man-marking system:
“Man-to-man marking is not a system I enjoy. It requires pursuing an opponent that needs to be neutralised. This has a flaw in the sense that one player moves away from his position in order to hunt down their opposite man.
“When the ball is regained, the team is set up in a defensive formation. Possession is then harder to manage as on the turnover, your players are breaking from a place of having been man-to-man marking. It’s harder to launch out into attacks from whatever formation that is formed as a result of chasing down opponents.
“But clearly, pressurizing opponents and accelerating the recuperation of the ball is very important and we base our game on our opponents not being able to find players in space.
“The man-to-man system is perhaps a shortcoming of my own teams. It’s something I’ve not been able to find the perfect solution to across 30 years as a manager so I doubt I’ll be able to resolve it now!”
So man-marking is here to stay for Bielsa, it seems. But this isn’t just about him. It’s as much about where Leeds United want to go in the future. For however much the Argentine can effuse about life in Yorkshire, the fact remains that the club find themselves under the constant restraint of his one-year contracts which are signed late in transfer windows. The hypothetical hangs over the club: what would happen if Bielsa leaves?
None of us enjoy that thought experiment. But the club will have a contingency plan in place in the event that the worst happens. In the rest of this piece, I am going to offer some suggestions for what that contingency plan should look like and how the club might treat this uncertainty to their own advantage.
Suggestion #1 — Play to your Strengths
This first suggestion is fundamental in the sense that it offers the context within which all the other suggestions will flow as a logical series.
When it comes to Marcelo Bielsa, it’s often hard to see past his utter uniqueness. When it comes to long-term planning, this can be framed negatively as the question “Where on earth do you go from here?” It’s important, though, for the club to re-frame that question into a more positive outlook: “Having one of the most unique managers in world football should mean that we have an edge that most other clubs don’t.”
I opened this article talking about the strange reality of Leeds’ time in the Premier League so far; that there are some teams who they seem to beat comfortably and other teams who seem to beat them comfortably. Now obviously, were it the case that Leeds had a squad stacked with the highest level of elite talent, then this might be a worrying reality. However, with even the most rose-tinted spectacles, there is scope to raise both the floor and the ceiling of this squad. Do that and you would think that the clear line that seems to exist between clubs above Leeds in the table and the clubs below Leeds in the table might be raised somewhat.
If any more inspiration is needed, you only need to look to Italy and the Atalanta side built by Gian Piero Gasperini who now compete in the Champions League and have become staple top-four finishers in Serie A. Beyond Bielsa, Gasperini is one of only a few managers in the top five leagues in Europe implementing a form of man marking (although his man-marking system is a little more flexible than Bielsa’s). But Gasperini’s approach suggests that Bielsa’s style, unique as it is, can function at the most elite level.
And on top of this, the individuality of this style can give you an edge in other ways. Not only are there ways that you can use uniqueness to exploit market inefficiencies — of which more later — but the unknown quality of a man-marking system within elite European football coupled with Bielsa’s positional approach in attack could give Leeds United an edge over their rivals to some extent. We’ve already seen that Leeds can blow oppositions away when those sides are of a comparable player quality. During Bielsa’s two years in the Championship, he effectively took a lower mid-table squad and elevated them to title challengers.
With all of this in mind, Leeds should lean into Marcelo Bielsa’s uniqueness and make it fundamental to their long-term plans. Primarily, this means implementing a model which incorporates the fundamentals of Bielsa’s style: man-marking, high pressing, positional play in attacking situations. By playing to this strength, Leeds will then be able to benefit from the relative uniqueness that Marcelo Bielsa brings to the table, gain an advantage in the medium term, and soften the blow of his eventual departure.
Suggestion #2 — Exploit the Market Now
We’re back at the fork in the road. Two options present themselves: a return to the ideals that Marcelo Bielsa implemented during his time in the Championship or a more muted approach that mutes the pressing style but allows Leeds to sit comfortably in the middle of the Premier League.
As things stand, the temptation might be to keep things as they are. A solid mid-table finish is not to be sniffed at so it seems sensible to implement the logic of ‘if it’s not broken, don’t fix it’. No doubt fans of Sheffield United felt the same way at the end of last season. My second suggestion, though, is that Leeds should come out this summer and really exploit the transfer market so as to return to the more idealistic iteration of Marcelo Bielsa’s Leeds.
In many respects, the situation Leeds find themselves in right now is the direct consequence of their transfer activity last summer. Looking back with the benefit of hindsight, the problem to which Stuart Dallas proved to be the solution was the signing of Rodrigo Moreno — a player who, almost a year later, we still don’t really know how to fit into the side. Of course, it didn’t require spadefuls of foresight to appreciate that bringing a player who was pushing thirty who had a long history of injuries into one of the most intense pressing outfits in the world might not work out in the long run. But putting that to the side, the biggest impact that the Rodrigo transfer had on Leeds was the impact it had on the press.
On the face of it, this seems to reverse the logic of what we have come to expect about signings. How could we possibly be negative about signing an international player with Champions League experience? Surely that is precisely what a club with ambition should be doing? But this is precisely where Bielsa’s uniqueness comes in. As a high-pressing, man-marking team, we are looking for something slightly different from the usual in the top five European leagues. And that means we are ripe to exploit market inefficiencies.
Now obviously, Rodrigo does suit the Leeds system when it comes to attacking play. The Spain international is clearly one of the most intelligent movers and exploiters of space we’ve ever seen in a Leeds shirt. The point is: his off-ball stuff was way below par. Going forward, the key has to be finding players who can press first and only then raising the question of whether they can fit the positional side of the game.
But this comes with a benefit: we are now looking for players who go beneath the radar somewhat. There won’t be too many clubs whose recruitment department will be looking for players who can press well as a primary skill set. And on top of this, the players who fulfill these criteria don’t tend to stand out as particularly valuable assets. Look at some of the names who filter through with links to Leeds: Conor Gallagher, Dan James, Callum Styles. Which isn’t to say that these aren’t good players — they are — but they aren’t the sort of names that you might want your club to go out of their way to sign.
My second suggestion, then, is that Leeds should go out into the market this summer and find players who suit the high-pressing, man-marking off-ball system that Marcelo Bielsa has implemented. Stuart Dallas — much as he has had a huge impact on our season — shouldn’t be the gold standard for a central midfielder in this system. We should raise the level of the press back to where it was and we will see the impact in the longer term.
Suggestion #3 — Build your Recruitment to Fit your Model
It’s all well and good to say, ‘Exploit the market!’ as if it’s that simple. My third suggestion, then, is that Leeds should build their recruitment department around their model to enable them to be able to exploit market inefficiencies even better.
What does this look like practically? Well, at present, the availability and scope of pressing analysis are woefully behind the curve. I began this article talking about the most basic of pressing metrics — pressures, successful pressures, pressure success %. These metrics only tell a very simple story and are, for the most part, far too noisy to offer any insight. If Leeds are serious about building a high-press, man-marking model, they should build a recruitment department that is able to assess potential targets by far more comprehensive metrics than what is currently on offer.
Fortunately, the scope for this is already huge. Even if they didn’t want to wait for a full roll-out of some form of tracking data that could be refined to fit a pressing or man-marking approach, there are already some fairly sophisticated ‘computer vision’ systems that could easily be extended to develop pressing or marking metrics. Take StatsBomb’s new 360 tool which offers contextual event data such as this ‘Distance to All Defenders in the Frame’ functionality:
No doubt this sort of tool will be used primarily to make observations about a team’s on-ball actions but the off-ball aspect to this sort of tool should be obvious too. Now we have information about where other defensive players are in relation to the presser and from here we can start drawing a clearer picture of what it is that happens during Leeds’ press.
From here, it is just a short step to developing metrics like ‘pressure chains’ — series of pressing actions one after another over the course of pressing sequences — and then to the introduction of timings or distances in pressing actions. For example: how fast does the defensive midfielder arrive at the pressure event? How much distance is the winger covering in the few seconds before they arrive at the pressure event? How many pressures does the 8 make in a pressing sequence? How far apart are these pressing events?
On the face of it, much of this information might seem superfluous but it would allow you to do two things: firstly, it enables you to start building up a picture of what it is that players are doing on the pitch per position so you have a good sense of what you’re looking for when it comes to recruitment; secondly, it allows you to develop metrics that are then commensurable across matches which will allow you to data scout better. Is Kalvin Phillips a stand-out in terms of his speed between pressing events? Can you find something in his distance covered that makes him an outlier? Then couldn’t this metric allow us to find potential back-ups for Phillips who are perhaps flying beneath the radar?
The recent news that Leeds have signed up to Analytics FC’s API Connect feed suggests that they are already thinking about ways to implement data more holistically into their recruitment process. However, were they to develop a proprietary analysis platform that focused on the areas of the game that their model emphasised — high press, man marking, positional play — then they would be better placed to exploit the obvious market inefficiencies that exist.
Suggestion #4 — Make Coaches Subject to the Model
The three previous suggestions are all compatible with Marcelo Bielsa being the manager of Leeds United. But as I mentioned earlier, the reality is that Bielsa isn’t forever. And when it comes to the question of how the club should replace him, it isn’t clear at this point whether or not there is an official stance on the direction that will be taken. Will the next manager follow in the footsteps of Bielsa or will the club opt to install another grandee of world football regardless of style?
My fourth suggestion is this: Leeds should make future coaches subject to the broader model that they implement. With Marcelo Bielsa in charge, you can hardly expect him to subject himself to a more fundamental model, of course. But by executing a model which emphasises his benefits — high press, man marking, positional play — and using him as the standard, the club should already be thinking of a succession plan which involves coaches of a similar style to lessen the impact of Bielsa’s departure.
Again, this is easier said than done: we’ve already noted how unique Bielsa is within the upper echelons of elite European football. Gian Piero Gasperini is only a couple of years younger than Bielsa, and beyond that, where do you go? There are a couple of options within elite European football — Ivan Juric at Hellas Verona is often cited and Jorge Sampaoli has just moved to Marseille — but the standard answer to this question is usually ‘South America’. This is all well and good but the segue from South America to Europe is one that is often highlighted as potentially problematic within football recruitment. The trick will be to find South American coaches who are already transitioning away from their content so as to make the switch easier.
This is, in part, the reason why Leeds should be thinking ahead of time. There are options — we’ve already mentioned Sampaoli at Marseille; Gabriel Heinze has just taken charge of Atlanta United in MLS where the San Jose Earthquakes coach Matias Almeyda is also worth a mention; Hernan Crespo is currently coaching in Brazil with Sao Paulo; and Eduardo Berizzo is the national team manager of Paraguay. It should be noted that all of these managers approach the game differently. Some are more wedded to man marking than others. Some will mix man marking and zonal marking depending on the phase of play. But all of them will use elements of it in their approach, usually married to an aggressive (often high) press.
In the event of Bielsa’s departure, with a succession plan in place, Leeds can then shift into an approach where the subsequent coaches are subject to the model and fit to it rather than they to it. Not only will this allow them time to prepare coach changes. It will also take an amount of pressure off the coach as the responsibility for the system is out of their hands.