The report lands on the table with the quiet authority of a thing that took three weeks and twenty-two raters to produce. Clean radar chart. Confidence intervals. A tidy row near the top that reads, decisive under pressure, four point one out of five. The room nods. The number matches the person they know. She is competent, she is trusted, she has run hard quarters without flinching, and the instrument has confirmed all of it in the language the organisation respects. Everyone leaves the calibration satisfied that they now understand how she leads.
They do not. The chart is accurate and the chart is useless for the one decision it is about to be used for, which is whether she is the right person for the seat above her. It will miss that entirely, and it will miss it not because the raters were dishonest or the questions were poorly written. It will miss it because of what the instrument is physically made of, and what it is made of cannot hold the thing that actually decides the answer.
A 360 Is a Survey of Memory.
Strip the design language away and look at what a 360 actually does. It asks a set of people to reconstruct a colleague from months of scattered contact, most of it calm, almost none of it consecutive, and to collapse that reconstruction into a number on a trait. Then it takes those numbers and averages them. That is the entire mechanism. Two rounds of compression, one inside each rater's head and one inside the spreadsheet, both performing the same operation. Both taking a moving thing and returning a still one.
And a leader is not a still thing. The score says decisive is four point one, which describes a center of gravity, a tendency, the version of her that shows up on a representative Tuesday. But no team on earth experiences a leader as a tendency. A team experiences a leader as a sequence. This call, then the next one, then the one after that, each made with a little less information and a little less time than the last, each one narrowing the room for the following one. Nobody is led by an average. They are led by an order of events, and the order is where the whole story lives.
The 360 has no column for order. It cannot have one. The raters saw her in slices, a one-on-one in April, a steering meeting in May, a corridor exchange in June, each encounter sealed off from the others and spread thin across a quarter. They almost never watched her make five consequential decisions in a row with the cost of each one still hanging in the air when the next arrived. So they have nothing to tell you about the sequence, and the instrument dutifully reports what they do have, which is the flat-water self, the one who was in the room when the water was flat.
Here is what that erases. The person who scored four point one on decisive was, in fact, decisive on the first decision and the second. On the third, with the picture already muddier, she started buying time and calling it diligence. On the fourth she began routing the call upward and calling it alignment. By the fifth, the decision was effectively made by whoever in the chain was still awake and willing to own it. None of that is a flaw in her character. It is a curve, and it is the most important thing about her as a candidate for the harder seat, and there is not a single pixel of it anywhere in the report. The chart said four point one. The chart was not lying. It was answering a question nobody in the room had actually asked.
The Flaw Is Not Perception. It Is Arithmetic.
The familiar complaint about 360s is that they measure perception rather than reality, that raters are political, that people score their friends up and their rivals down. All true, all tired, and all beside the point. You could solve every one of those problems, hand the instrument perfectly honest raters and immaculate questions, and it would still miss the gap, because the gap is not a perception failure. It is a consequence of the math. Averaging is the method, and averaging is precisely the operation that deletes sequence. Take any series of numbers that rises and then falls off a cliff, average them, and the cliff disappears into a mild-looking mean. That is not a bug you can rater-train your way out of. It is what the arithmetic is for.
Which means the thing you need cannot be built by improving the 360. You do not need a better description of the leader's typical self. You need a recording of her decisions in the order she made them, under load, with the clock running. A 360 can never contain that no matter how carefully it is designed, for the same reason a photograph can never contain velocity. You can take a sharper photograph. You will not find the speed in it. It was never a property the medium could hold.
What It Would Take to Actually See It.
To see the gap, you have to stop describing the leader and start watching her decide. That means an instrument that preserves order instead of averaging it away. Put a leader inside a chain of decisions where each choice changes the conditions of the next, start the clock, remove the option to reset, and then read what she did the way it was actually lived, as a line rather than a cloud of impressions. The moment you do that, the curve the 360 flattened comes back into view, and with it the only information that ever mattered: not where her judgment sits on a calm day, but how far it travels before it breaks.
The SSUNDAR Organizational Crisis Simulation was built around this single refusal to average. It runs a leader through five cascading crises with the clock live and every decision feeding the conditions of the one that follows, and it never once asks a bystander what they thought of her. What it returns is not a trait score and not a radar chart. It is a trajectory. Where the judgment held, where it started to bend, and the exact decision, named and timestamped, where it gave way. The leaders it exposes most sharply are almost always the ones who arrived carrying the cleanest 360s, because a tidy average is extraordinarily good at hiding a steep curve. A four point one can be a straight line or a plunge. On the chart they look identical. In the seat they are two entirely different people.
This is why the organisations that have watched both instruments side by side stop treating the 360 as a decision tool and start treating it for what it is, a competent map of the calm-water self, useful for coaching conversations and useless for the question of who to trust when the water stops being calm. The 360 tells you who a leader is across an ordinary quarter. It has never once told you who she becomes by the fifth decision, and the fifth decision is the whole reason you were promoting anyone in the first place.
An average is a still photograph of a moving thing. You can make it sharper. You will never find the motion in it, because the motion was the one property the instrument was built to remove.
Your 360 measured the leader you have on a quiet day. Your organisation only ever meets the one who shows up on the loud ones.