Data scientists in soccer need to grow a pair!
You must take power! Like Stalin! (not like Stalin)
I’m going to wind down this series on the Footballing Theory of Value this week. Before I do, I want to look at how understanding competition as the sole source of value in football can help aspiring analysts and, perhaps, the entire analytics movement.
Similar to naive, struggling clubs that throw their hands up as rich, historically competitive sides consolidate power and use threats such as the ‘super league’ to leverage their collective power, analysts often fall into the trap of taking moral umbrage at clubs that refuse to embrace the full spectrum of analytics tools as their disposal to gain an edge on their competitors.
Worse still, analysts have today come to convince themselves of some dubious solutions to this problem. Some of them believe convincing clubs to embrace analytics means ‘speaking in a way coaches understand,’ or, more diplomatically, taking the wrong-headed concerns and focus of old school managers seriously prima facie. Other analysts believe they just need to ‘prove’ to people in positions of power that analytics methods can produce tangible, lasting results, whether in recruitment, tactics, player development, etc.
In the end, these solutions all involve some form of argumentative or rhetorical skill, which, you know, isn’t what these data scientists have studied their entire lives to do (nor do I think they necessarily need these skills to succeed, but I will get to that in a moment).
The problem, instead, is that some analysts tend not to fully realize the very real, very stark power dynamics at play in most football clubs, and direct their appeals to the wrong people in the team in the wrong way.
Just as the value of football is derived solely from competition, so too are certain figures at clubs in perpetual competition to keep their job, whether it’s a manager trying to push beyond the industry average two-year tenure, players trying to land a spot in the first team, or directors of football vying to prove their methods are invaluable to the club. What this means is that analysts will appeal to a director of football, manager or player only inasmuch as the latter group thinks the analyst will help them keep their job.
Great! You might be thinking. All you need to do is to make a case for why your work as an analyst will reflect well on the person you’re trying to convince (which will differ depending on whether you’re an analyst within a club trying to exert more power, an analyst seeking to get a job with a club, or an independent consultant).
The problem, of course, is that your work as a data scientist won’t be as effective if it is primarily there to make someone else look good. For one, data science doesn’t work that way; the work of an analyst is to tilt the odds in your team’s favour, but there is a lot of random variation in football, and positive results tend to accrue over time and are not always obvious to those in power.
A much more ideal situation, therefore, is for analysts to have more autonomous and clearly delineated decision-making power within a team. Maybe the analyst has final say in developing recruitment short-lists for example, with the manager following through in collaboration with the DoF. Or perhaps the analyst has full authority in preventing players at risk of injury of starting.
This all might sound insane! Analysts are so used to grovelling to people in power, making suggestions, coming up with attractive data viz and hinting at what a manager may or may not do to win. But the reality is, analytics must move beyond begging for power to survive and to make a lasting influence.
Working the owners
One way to secure this power would be to start by working the highest rung on the hierarchical ladder—the owners. Owners tend to come from the Magical World of Business, where data science is often already firmly established in banal fields such as market analysis and sales or whatever. They fashion themselves as smart numbers people.
Analysts have successfully taken this approach in the past (there are many examples). Still, even in these cases I believe analysts haven’t pushed for enough power, nor have they successfully worked to properly separate the sphere their work from that of the manager or the DoF. So inevitably when clubs struggle and end up cleaning house, the analyst is often the first to go, followed by articles in the Telegraph talking about the damned twenty-something nerd who had power to make decisions, it’s really their fault.
That’s why analysts shouldn’t just argue to owners why THEY are the right person for the job, but that the existence of the job and its attendant power itself is critical for team success. “Sure, I might suck, and you might need to fire me, but the team should ALWAYS have an analyst role with locked in decision-maker power over X,Y, and Z, and here’s why.”
The short of the long here is that analysts need to get some balls, they need to stop whingeing about their lack of influence and how dumb people who run clubs are, they need to understand these people aren’t dumb (or at least as dumb as you might think) but just unwilling to concede power.
Once analysts understand they must struggle to exert power in football, then they might begin to get somewhere.