Growth offers seven distinct pool values. Ground-Floor offers one.

The count

Sixteen sequences map onto fewer and fewer distinct pools as the gate deepens.

Every gate faces the same sixteen possible four-trade sequences, and what differs is how many distinct pool figures those sixteen produce. In the top two states the answer is seven, so the matrix is making fine distinctions and most sequences land on their own value. Buffer collapses to five, floor to three, deep-floor to four, and ground-floor to a single figure that every sequence shares. The mechanism is the positional weights attenuating: where growth uses a full four-two-one-zero ladder, the deeper gates zero out more positions until none of them carry weight. The result is that the same sixteen inputs produce a progressively coarser output, and eventually a constant.

What coarsening means

The system stops distinguishing between cycles it can no longer act on differently.

Resolution is the system's ability to tell two cycles apart in deployment terms, and giving it up is a decision about relevance rather than a loss of precision. In growth, the difference between a three-and-one cycle with an early loss and one with a late loss is worth a percentage point of pool, and the system says so. Under floor, several sequences that would have been distinguished higher up collapse onto the same figure, because at that depth the difference between them no longer justifies a different deployment. The account is in containment, the mission has changed, and fine-grained responsiveness to a four-trade sample is not what the state calls for. Coarsening is the ladder declining to make distinctions it would no longer act on.

The irregularity

Floor produces fewer distinct values than deep-floor, which is worth confirming.

The descent is not clean. Floor yields three distinct pool figures while deep-floor, one state deeper, yields four — so resolution rises slightly before collapsing. That is not what a smooth design intent would produce, and it is worth flagging rather than narrating past. There are two possibilities. It may be deliberate: the floor row's weighting zeroes two positions where deep-floor zeroes three but spreads its remaining weight differently, and the resulting counts are a side effect of choices made about each row on its own terms. Or it may be an artefact of hand-tuning that nobody has had reason to look at, since the counts are not a quantity the panel displays. Either way the pools themselves are internally consistent within each row, so nothing is broken — but the shape is a question the construction notes could answer and currently do not.

The endpoint

A single value means the matrix has stopped being consulted.

At the bottom the resolution reaches one, and that is a categorical change rather than the end of a trend. When every sequence produces the same pool, the outcome matrix is no longer performing a function for that state — the lookup still runs and its answer no longer depends on its input. In effect the account has moved from a performance-responsive deployment regime to a fixed one, and it did so at a boundary the operator crosses without any announcement beyond the gate name changing. Knowing this is useful precisely because nothing signals it: an operator in the bottom state who is watching recent results for signs that deployment will improve is watching an input that has been disconnected.

The design principle

Responsiveness is a resource, and it is spent where it can do good.

Read across the whole ladder, the pattern is that the system allocates its capacity to respond where responding is safe and withdraws it where responding is dangerous. In the healthy states, evidence moves deployment meaningfully — that is where the compounding is earned and where a good stretch should be worth something. In the damaged states, evidence moves deployment less and eventually not at all, because that is where acting on a four-trade sample is most likely to convert a recoverable drawdown into a terminal one. The ladder therefore encodes a claim about when to listen to recent results, expressed not as advice but as the number of distinct answers the table is willing to give.

• Sixteen sequences produce seven distinct pools at the top and one at the bottom.
• Coarsening is the system declining distinctions it would no longer act on.
• The floor/deep-floor irregularity is worth confirming against the construction notes.

The key idea

How finely a system responds is as much a design decision as how much.

Most discussion of risk laddering concerns magnitude — how much deployment falls as conditions worsen. Resolution is the quieter dimension and arguably the more revealing one, because it says how much the system is prepared to let recent evidence matter at each depth. A framework that stayed fully responsive all the way down would be listening hardest to small samples exactly where small samples are most expensive to be wrong about. Withdrawing responsiveness before withdrawing the last of the capital is the ladder protecting the account from the operator's most recent memories, which are at their most vivid and least reliable in precisely those states.