The Laws of the Game spend exactly two sentences on the captain. Law 3 requires each team to have one, wearing an armband, and then says that "the team captain has no special status or privileges but has a degree of responsibility for the behaviour of the team". The captain can't pick the side, make a substitution or change the shape, and even who gets the armband is decided by somebody else, in whatever way that person likes. Manchester City's squad voted Kyle Walker in for 2023/24, and when Guardiola was asked whether Walker would keep it he said "it's a decision for the team, not mine"; for 2025/26, after Walker had left mid-season, he named Bernardo Silva himself, saying "I didn't like what happened last season" and that "sometimes I want to be the boss". The armband carries responsibility, then, but no authority, and the authority it doesn't carry belongs to the manager, who can take the armband back.
That is the split this series keeps coming back to. In engineering terms the captain is the tech lead or staff engineer, who leads inside the work without the power to hire, fire or set pay, and the manager is the engineering manager, who owns selection, the system and the room but has mostly stopped touching the ball.
The best-known put-down of a football captain was aimed at exactly the kind of leadership the armband describes. In 1996 Eric Cantona, by then exiled from the France squad, said of Didier Deschamps that he "gets by because he gives 100 per cent, but he'll never be anything more than a water carrier". That is the English wording as AFP reports it. The original interview, with La Gazzetta dello Sport, isn't online, the English is a translation of a translation, and Europe 1 renders it as "vulgaire porteur d'eau", noting that Cantona suspected Deschamps of helping keep him out of the Euro 96 squad. Deschamps' reported answer was to ask how many players on street corners had won two European Cups. Two years later he lifted the World Cup as France's captain, and in 2018 he won it again as their manager, the third man to win it as both player and coach, after Zagallo and Beckenbauer.
Deschamps looks like the obvious illustration of Sam Walker's The Captain Class, but he isn't in it. Walker screened winning teams across dozens of sports for multi-year dominance against top competition and ended up with sixteen "Tier One" sides; France 1998–2000 isn't one of them, and the football representative of his kind of captain is Carles Puyol of Barcelona 2008–13. The captains he describes were, in Keith Duggan's summary in the Irish Times, "rarely the best players", showed "a near-hostility to fame and celebrity", "seldom gave big emotional speeches but talked constantly during games", and "facilitated in ways that were invisible." The book has a problem it shares with the rest of the genre, though: it picks the teams by their results and then looks for what their captains had in common. It doesn't check whether the captains of teams that won nothing had the same traits, and I haven't found a published critique that does, so the most the sample can tell you is that unglamorous captains are compatible with dominance, not that they caused it.
The engineering version has better documentation of the cost. Tanya Reilly's talk "Being Glue" describes the work of unblocking teammates, reviewing designs, onboarding new people and noticing where two teams' plans don't meet — "I call all of this glue work" — and argues that it is "technical leadership". Her talk follows one engineer who does it well and is then told that "this company doesn't consider that to be sufficient promotable work, at least not at this level", and that she's "not technical enough", which is Cantona's water carrier in a calibration meeting. Reilly's warning is that "doing glue work too early can be career limiting, or even push people out of the industry." Will Larson's staff-engineer archetypes give the same work a place higher up: the Tech Lead, "the most common Staff archetype", guides one team's approach and execution and partners "closely with a single manager", which is the captain standing next to the touchline. The one archetype with real reach, the Right Hand, extends an executive's attention, "borrowing their scope and authority", and Larson calls it "the least common". Even his promotion-packet template has to ask outright, "What glue work do you do for the organization? What's the impact of that glue work?", which tells you that left to itself the process won't go looking.
I've sat in both seats, tech lead and engineering manager, and the move from one to the other brings more responsibility and a wider scope to influence, while the hands-on part shrinks. What holds in both seats is that good leaders are always talking across departments, which is as true of an engineering lead as of the captains Walker describes. On the teams I led, most of the time I was the one doing the glue work. And I've found corporate promotion processes far from straightforward; they can be unrelated to how well someone is actually performing, which is the gap Reilly describes. How that gets measured, and mismeasured, is a later post in this series.
The analogy breaks at the scoreboard. Deschamps could answer Cantona with two European Cups because football keeps a public, unambiguous record and his contribution sat inside it, in front of everyone in the stadium — a holding midfielder's work is unglamorous, but it isn't private. Glue work mostly happens in direct messages, design reviews and conversations between teams, nobody is watching, and a quarter's engineering output doesn't come with a result that settles the argument. The career paths break too. A footballer's playing days end in his mid-thirties whatever he wants, so the move from captain to manager is forced and comes with a gap and a new start, whereas the move from tech lead to EM is optional and is often treated as the promotion itself — which means that an engineer who'd rather stay a captain can find the only visible route upward runs through the touchline.
Deschamps had trophies to put on the table when he was called a water carrier, and the engineer in Reilly's talk had a promotion committee telling her what didn't count. If the packet has to ask what glue work you did, the person who did it has to notice it and argue for it themselves, when the work consisted of making other people's output count. Who on the team is supposed to be keeping that record, if it isn't the manager?
No comments yet