Croesus measured the oracles. And he measured what he did not need.

The story of Croesus invites us to distinguish between measuring a capacity and measuring what a decision requires.

Croesus measured the oracles. And he measured what he did not need.

Series: What history already knew about talent · 1/6 · Talent assessment

What follows is a tale from Herodotus. It is worth mentioning at the outset, because Herodotus wrote a century after the events, mixing chronicle and narrative, and even in antiquity there were those who doubted this particular episode: Cicero, in his treatise on divination, suggests that the prophecy to Croesus may have been a later construction. We do not know if it happened that way. We know that a Greek from the 5th century told it this way, and that the structure of what he recounts deserves a reading.

Around 550 BC, Croesus, king of Lydia and the richest man in the known world, had to decide whether to attack the Persian empire of Cyrus, which was growing across the river Halys. In his world, such a decision was consulted with oracles, and Croesus had at his disposal the most prestigious ones: Delphi, Dodona, Abas, the sanctuary of Amphiaraus, that of Trophonius, the Branchidae of Miletus, and the distant Ammon in Libya.

According to Herodotus, before asking them about the war, Croesus wanted to know which of them was correct.

The test

He sent messengers to the seven sanctuaries with identical instructions: on the hundredth day since their departure from Sardis, all should consult the oracle with the same question. Not "Should I attack Persia?", but something that could only be answered with real knowledge: "What is King Croesus doing at this moment?".

And that day Croesus did the most improbable thing he could think of: he boiled together a turtle and a lamb in a bronze cauldron covered with bronze.

Herodotus does not record what the other oracles responded. He does record the response of the Pythia of Delphi, in hexameters: that she perceived the smell of a hard-shelled turtle boiling alongside lamb meat, with bronze below and bronze above. Croesus accepted Delphi's answer and, says Herodotus, also that of the sanctuary of Amphiaraus. He ceased to consult the others.

What is in the structure

Put aside whether the Pythia was truly accurate. Look at what Herodotus attributes to Croesus, because that is what matters to us.

He had to make a very costly decision. He had several sources of judgment, all with centuries of reputation. And instead of relying on reputation, he designed a test: the same for everyone, under comparable conditions, with a correct answer that he knew in advance and that could not be guessed. Only after seeing the results did he decide whom to listen to.

That, with or without oracles, is evaluation. It is what we do with a drug, with a bridge, and with a balance sheet. And it is what we almost never do with people.

When an organization decides whom to promote, whom to give the big project to, or whom to hire, what is on the table is reputation. The CV is the reputation that the person tells about themselves. The boss's opinion is the reputation built in a specific context. References are the reputation that others have made for them. The competency map on the intranet is, at best, a self-declaration with checkboxes. None of that is a lie. It’s just that none of that is a test. It’s what Croesus had before sending the messengers: seven famous sanctuaries and no way to know which one was right.

The reading I propose is this: if you want to know if someone can do something, don’t ask them if they can do it. Present them with the specific capability that the position requires, under the same conditions as everyone else, with a method that yields the same result if repeated. And do it before the important decision, not after, when the real evaluation is done day by day and a failed hire is already costing what it costs.

What Croesus measured poorly

This is where the story becomes more interesting than it seems, and where I believe the usual reading is mistaken.

With Delphi already validated, Croesus asked what he wanted to know: whether he should go to war with Persia. The oracle replied that if he crossed the Halys, he would destroy a great empire. Croesus understood that the empire was Cyrus's. He crossed the river, lost, and the empire destroyed was his own. Herodotus recounts that, already a prisoner, he sent to reproach Delphi, and the response was that the god had spoken the truth and that the king had not asked which of the two empires.

The easy reading is that Croesus measured well and decided poorly. I believe there is another, more uncomfortable and more useful one.

Notice what capability Croesus tested: knowing what was happening at that moment in a distant place. And notice what capability he needed: to guide a military decision about the future. These are different capabilities. Hitting the turtle does not demonstrate the ability to advise on a war. Croesus designed an impeccable test for a capability that was not what his decision required, and then transferred the trust gained in one to the other.

This happens in companies every day. It is verified, with more or less rigor, that someone is an extraordinary salesperson, and they are promoted to sales manager, which is a different skill. It is verified that someone masters a technology, and they are put in charge of a team. What is easy to evaluate or what the person has already demonstrated is assessed, and a decision is made about what the new position will require of them, which is almost never the same.

Measuring is the condition for a good decision. Measuring what the decision requires is the condition for measurement to be useful.

At The Wise Skill, we build that: a way to know what people, candidates, and employees can really do, with the same test and the same criteria for everyone, through evaluation and not self-perception, and against the specific skills required for the position being decided upon. The part of not hearing what one wants to hear remains the responsibility of the decision-maker.

But you can ask the first question tomorrow at the next talent meeting.

Are we measuring the turtle, or are we measuring the war?

If you want to see it with your own profiles, message me directly and I'll tell you how we do it.

Next installment: how Rome chose its centurions, according to Polybius.

Sources. Herodotus, Histories, I, 46-49 (the test of the oracles), I, 53 (the inquiry about Persia) and I, 91 (the explanation of Delphi after the defeat). Cicero, On Divination, II, 115-116, expresses doubts about the authenticity of the prophecy. The episode is a narrative by Herodotus, not a fact corroborated by other sources; it is used here for its structure, not as historical evidence of an objective assessment.