“The center of gravity is the hub of all power and movement, on which everything depends; that is the point against which all our energies should be directed.”

— Carl von Clausewitz

In chess, a player can win pieces, control the centre, dominate the board, and look brilliant doing it, and still lose, because they forgot the only condition that ends the game.

Capturing pieces is a proxy for winning. The King is the anchor. When the King falls, every captured rook and elegant tactic loses its value at once, and the board does not care how efficiently the collapse was executed.

Institutions fail the same way. Activity accelerates, capability flexes, the metrics brighten, and somewhere in the motion the system trades away the one thing it existed to protect.

The reasoning is not absent when this happens. It is working hard. It is simply serving the wrong sovereign.

That sovereign is the King, and most decisions fail because the room never agreed, out loud, on what it was.

The King is not a mission poster, a slogan, or a sentimental account of why the work matters. It is a one-sentence public promise that defines the institution’s loss condition: the value you refuse to trade even when success is on offer at a steep discount. It answers three questions without poetry, who you serve, what end you will defend, and what you will not do to achieve it.

It is worth separating from the language leaders usually reach for. A mission says why the work is good. A King says what must not fall. The first inspires; the second decides. Only the second survives a collision.

And collisions are what decisions are. Real choices are collisions between goods, not contests between good and bad. Speed collides with safety. Growth collides with trust. Reputation collides with truth.

In every such collision the system selects a winner. If leaders have not ordered their goods in advance, something else will, status, fear, convenience, or the loudest voice in the room. The outcome will be called strategy. It will in fact be a story written afterward to justify whatever the system drifted into doing.

A valid King is therefore a promise with a tie-break rule, and the single sentence must encode four things to work as one.

It names a beneficiary explicitly. Not “everyone,” not “the market,” but exactly who is protected when the trade-offs turn ugly. A named beneficiary makes it harder to treat harm as collateral and easier to detect when an institution has begun serving itself.

It states an end in the beneficiary’s terms, safer, fairer, more reliable, more dignified, not a dashboard target dressed as a purpose.

It declares an order of goods, the most underrated source of clarity an executive controls. Leaders reach instinctively for the word balance, balance speed and safety, balance growth and trust. But balance is a sedative. It sounds wise and resolves nothing, and a manager staring at a bottleneck and a deadline is left to guess which good the executive worships. Governance does not run on balance. It runs on hierarchy: safety first, and within safety, speed.

And it establishes cliff edges, hard lines the institution will not cross even when the numbers look spectacular. Cliff edges are what prove the King is real. They keep “we had to” from becoming a standing license for betrayal.

The obvious objection is that a fixed sovereign makes an institution rigid, unable to adapt when the world moves. It is a serious objection, and the answer is not to soften the King but to govern how it changes.

A King is declared for a defined horizon, this cycle, this deployment, so the promise stays a deliberate choice rather than an unexamined reflex. And it can change, but only out loud: a new cycle, stated reasons, a public substitution. What you may not do is quietly install a new King downstream and call the swap learning. Rigidity is not the price of having a King. Concealment is.

The danger the King guards against has a name: proxy coronation.

The true King, long-term customer health, civic trust, public safety, is usually hard to measure, so organisations invent proxies for it: net promoter scores, active users, revenue, response times. These are useful instruments, and that is exactly the danger. A proxy is a tool you can put down. A King is an end you will not trade. Over time the difference is forgotten, the proxy is crowned, and the institution becomes ruthlessly efficient at the wrong thing and proud of the efficiency.

One dominant handset maker shows the error cleanly. In the mid-2000s it owned secure enterprise messaging and built its identity around the physical keyboard, and by the rules of that game it was winning. Then the terrain moved: a new entrant reframed the phone as a general-purpose software platform, and the loss condition changed with it. Relevance now meant the ecosystem, not the device. The incumbent fought brilliantly to defend enterprise loyalty and the keyboard, but in the new game those were proxies, not the King. It spent its considerable strength defending the wrong sovereign, and tactical victories only postponed a structural defeat. That is the quiet cruelty of sovereignty error: it lets an institution look entirely competent while it loses.

So once the King is crowned, every decision must be readable as service to it. If a proposal cannot be shown to serve the named beneficiary, honour the order of goods, and respect the cliff edges, the process stops, and the leader has three honest options and no others: rewrite the King in the clear, shrink the ask to fit, or admit that a proxy has taken the throne.

This is not an academic exercise. It is armor against drift. It forces the room to prove, in one breath, how an initiative protects the promise, with no hiding behind verbs like optimise and scale.

Name the King before the collision comes, and the collision resolves itself. Leave it unnamed, and something else will sit on the throne, and call its reign strategy.

Every decision serves some master; name your King explicitly, or a convenient proxy takes the throne and you optimise the measure while the value it stood for quietly dies.

Decision Rule — The Throne Rule

Before approving any major initiative, investment, or policy, test it against the established sovereign.

• Beneficiary. Does this serve the explicitly named beneficiary, or does it serve the institution?
• Order of goods. Does this honour the declared tie-breaking hierarchy, or does it rely on the illusion of “balance”?
• Cliff edges. Does this approach or cross a non-negotiable threshold?
• Proxy check. Are we optimising a measurable metric at the expense of the true loss condition?

If the decision cannot be articulated as direct service to the King, rewrite the King where it can be examined, shrink the decision to fit, or reject it.

Never let a proxy take the throne.