The 'Just About Managers'

'Gareth Southgate blogged on LinkedIn' is a real sentence. There's a lot of bad in the world - the resistance to it a light in the darkness itself - but every now and then the universe throws you a bone. Southgate even made some typos.

In 'Manager v Head Coach', he was commenting on the departures of Ruben Amorim, Enzo Maresca, and Xabi Alonso from their jobs.

Although each coach left under slightly different circumstances and for more than one reason, power struggles [...] were ultimately the root cause of the end of each tenure.

As Southgate says, the power of a 'manager' has been whittled down over the past couple of decades. It would now be uncommon (at the higher tiers of the game) for a club to not have a 'sporting director' figure working either alongside or above the 'head coach'.

I've previously wondered whether this has been one of the obstacles (among others) to 'analytics' being used in first-team, on-pitch matters:

‘[F]ootball people’ have accepted a lower level of control of the off-field stuff, retaining control of tactics and player-management (although some head coaches do gripe about the say-so of their medical departments). How might things be different if managers had demanded they keep control of transfers, and relinquished some on-field control to ‘directors of tactical methodology’ instead.

Now, neither me nor Southgate are saying that modern managers can do it all. As Get Goalside's new blogging rival says:

[I]n a classic example of Parkinson’s Law (where time saved by reducing responsibilities is quickly filled by something else), the modern Head Coach now finds themselves managing larger squads, bigger backroom teams, far greater analytical demands, and ever-increasing media and commercial obligations.

It's probably fair to say that, of those added burdens on a head coach, not many of them are fun. And the one that is fun, at least to the individuals in those jobs - 'greater analytical demands' - is, yes, the one that data analytics may 'threaten'.

'Threaten' is in quotes there because whether a coach is 'threatened' or is not an inevitability. It's an organisational management problem, really, although one similar to transfers and medical load management advice, which are not without conflict at many clubs.

It's easy to understand the frustration. Heck, as with many analytics themes, Moneyball is a primary colours guidebook/caricature: diktats coming from On High down to the manager, whose role is reduced to being a figurehead on the side-line. In Moneyball, of course, the control of Upper Management got the team to the play-offs [edit: not, as originally written, the World Series]. But if Management make decisions that don't pay off... well, there's already a term for that in football: losing the dressing room.

There is a world where analytics can be coach-focused more than club hierarchy-focused, although the advantage is with Hierarchy. It's the club whose budget pays for the software, the club who has to think about succession planning. [Related: "So, everyone's a decision-maker now", October 2023].

Managers - of the type of Alonso, Amorim, and Maresca - also have very different datasets-of-interest than the ones that modern providers deal with. Southgate again, on why his LinkedIn bio reads 'Leader, Manager, Coach':

The Leader operates in the public eye, engaging fans, setting the vision, shaping the culture, and championing and protecting the players and staff.

The Manager is ‘managing’ stakeholders, departments and people, maximising collaboration and making plans stay on track.  Oh, and don’t forget the role of persuading 25 financially independent elite performers to put collective purpose ahead of their ego.

Finally, the Coach works on the pitch, setting the style of play, preparing tactics to exploit opponents, and developing the technical abilities of individual players.

Arguably, the best bit of data software for football 'head coaches' would be Claude Cowork. But, because of the football domain that managers have managed to protect - the 'coach' part of Southgate's trident - any software would need an intense boot camp in football terminology. Cynically, one could rip a set of coach training materials to tune the LLM-aided system.

This wouldn't necessarily help the Hierarchy-Head Coach relations though. Let's look back at something from Get Goalside's 'Lessons in genAI' post, written just over a year ago:

Currently, the clubs are the ones who buy software - but if part of what they’re hiring in a head coach is that coach’s methodology, it’s that coach’s ‘data’ that is the most important to access. [...]

And, of course, if we’re talking about a coach’s knowledge as ‘data’, there’s the data ownership question. If LLM-based systems are going to be relying on coaching knowledge, coaches have got to make sure they can take this with them and use it in future roles.

A few days ago, Microsoft CEO Satya Nadella said (perhaps with some self-interest): "If you're not able to embed the tacit knowledge of the firm in a set of weights in a model that you control, by definition you have no sovereignty. That means you're leaking enterprise value to some model somewhere."

That message echoes a point made by Twenty First Group in a podcast/event collaboration with the Unofficial Partner podcast. Essentially, 'your company knowledge is valuable data, and you will lose out if you don't think of it as data and make use of it'.

So now, imagine you're Ruben Amorim. You come into Manchester United, a club building their data capability. You have a very distinct style of play which has had success in one context, but that you don't have lots of experience in adapting to different circumstances. A rich but partial dataset.

What kind of 'data infrastructure' would best help you? For someone who keeps saying that football clubs should avoid being tech companies, I'm going to say something shocking, but maybe a nice, clean API from United's institutional knowledge data banks [no jokes please] and a Claude AI-morim would be a good route forward.

And, circling back to Gareth Southgate's blog, this could help a head coach be a true manager. If the current and historic knowledge of different departments is open to be tapped into by a well-crafted LLM-aided system, it could be easier for an Amorim or a Maresca to 'manage stakeholders' and 'maximise collaboration'.

This isn't just a power play. The friction around transfers feels a particularly frequent occurrence, and reporting suggests that disagreements on medical adjustments was a cause for contention with Maresca and Chelsea. But if a club is able to say 'hey, here's our research data on this topic, can you sit down with your CoachingBadgeGPT and see what ways it can align with your thinking', maybe that would feel less like Upper Management instruction and more like Cross-Management partnership.

I await Sir Gareth's further thoughts on the matter.