"Some of the best advice I've had comes from junior officers and enlisted men."
Fleet Admiral Chester Nimitz said that while commanding more than two million men and women, 5,000 ships, and 20,000 aircraft across the Pacific. The line reads as modesty until you look at how he got the job, and at what the Navy had done with him forty years earlier. Then it reads as the output of a system. Rome ran the same system, and wrote down why it worked.
The pick nobody's org chart predicted
Three weeks after Pearl Harbor, Franklin Roosevelt gave the Pacific Fleet to Nimitz, reaching past 28 flag officers, every one of them senior to him. The order was blunt: get out to Pearl Harbor and don't come back until the war is won. In peacetime that promotion could not have happened. Seniority is how large organizations allocate advancement when the cost of a bad pick is low; the queue is cheap to administer and nobody sues. A catastrophe changes the math. After December 7, the question stopped being "whose turn is it" and became "who can actually do this," and the answer sat 29 names down the list.
Rome faced the same question after its own December 7. At Cannae in 216 BC, Hannibal destroyed the largest army the republic had ever fielded, killing tens of thousands of Romans in a single afternoon. Four years later, when the Spanish command needed filling and no established man would stand for it, the assembly handed it to Publius Cornelius Scipio, a man in his mid-twenties whose father and uncle had just been killed fighting there (Livy 26.18). The grey heads objected on procedural grounds, and the grey heads were right about procedure. Scipio went on to take Spain, invade Africa, and beat Hannibal at Zama.
The lesson survives translation into any modern org: your promotion system reveals its real purpose under stress. If the queue still governs after the crisis hits, the institution has decided that order matters more than survival. Boards that can only imagine the COO as the next CEO, and enterprises that cannot move a high performer past two layers of tenured management, are running the pre-Cannae playbook and hoping no Hannibal shows up.
The court-martial that paid off at Midway
The second fact about Nimitz is stranger. As a young ensign he ran his first command, the destroyer USS Decatur, aground, and was court-martialed for hazarding a Navy ship. The reviewing admiral endorsed his record with one sentence: "This is a good officer and will take more care in the future." The Navy convicted him, kept him, and gave him submarines. Four decades later that same officer made the calls at Midway.
Rome institutionalized this exact behavior, and a contemporary analyst noticed. When Terentius Varro, the consul who presided over the Cannae disaster, returned to Rome, the Senate and the crowds came out to thank him because he had not despaired of the republic (Livy 22.61). Carthage crucified generals who failed. Rome sent them back to work, and Polybius, a Greek writing a systems analysis of why Rome kept winning, identified that tolerance for defeat as a structural advantage: the state absorbed losses that would have ended other powers because it did not destroy its own leadership every time fortune went against it.
Modern engineering culture rediscovered this and named it the blameless postmortem. The principle is identical: punish the concealment of failure, not the failure, because the alternative teaches your best people to hide mistakes or leave. How a company treats the engineer who took down production this quarter determines who still works there in ten years. The reviewing admiral who wrote one generous sentence on a court-martialed ensign's file was making a forty-year investment, whether he knew it or not.
Where the real information lives
The third fact is the quote itself. Nimitz hand-picked his commanders, took counsel widely, and credited junior officers and enlisted men with some of his best advice. This was not a personality quirk. It was a working theory of where accurate information lives in a large organization: at the edge, with the people operating the actual machines, and every layer between them and the commander is a filter that trades fidelity for comfort.
Roman commanders who won held the same theory. Caesar's account of the Gallic war names centurions, the sergeants of the legion, with a specificity he rarely grants his own senior legates. Pullo and Vorenus, two rival centurions who saved each other's lives in a Nervian assault, get a full dramatic scene in his fifth book (Gallic War 5.44). Caesar wrote for a political audience, and he chose to spend his ink on the men closest to the fighting, because that is where his information and his loyalty engine came from. Marcus Aurelius ran a private version of the same practice: the first book of his Meditations is an itemized ledger of what he learned from the people around him, including Epictetus, a slave-born Stoic lecturer whose handbook reached the emperor secondhand. The most powerful man alive kept written accounts of his debts to people with no power at all.
The executive translation is the skip-level conversation, done as a discipline rather than a gesture. The org chart tells you who reports to whom. It does not tell you where the truth is, and the truth is usually two or three layers below the point where it gets summarized.
One system, three behaviors
Promote by judgment when it counts, keep your people after honest failure, and collect ground truth from the edge. Each behavior is cheap to praise and expensive to practice, because each one taxes the comfort of the people at the top. The United States Navy practiced all three on one officer, and the return on that investment was the Pacific. Rome practiced them for centuries and left us the receipts, from Livy's account of a thanked loser to an emperor's private ledger of borrowed wisdom.
The endorsement on the file of a court-martialed ensign remains the shortest version of the whole doctrine: "This is a good officer and will take more care in the future."