A perfect cycle in one gate cannot reach the pool of the gate two rows above.

The bands

Each gate is a range, and the ranges overlap only with their neighbours.

The outcome matrix prices every combination of four results within each gate, so a gate is not a single pool figure but a span. Growth runs from twenty-six to thirty-two percent. Recovery runs twenty-one to twenty-eight, buffer nineteen to twenty-four, floor sixteen to twenty, deep-floor twelve to eighteen, and ground-floor sits at twelve alone. Lay them out and each band shares territory with the one directly above and below it — growth and recovery overlap between twenty-six and twenty-eight, recovery and buffer between twenty-one and twenty-four — while no band reaches the one two steps away. Growth's worst case exceeds buffer's best by two full percentage points, and the gap widens further down the ladder.

What overlap buys

Neighbouring bands sharing territory means gate transitions are not cliffs.

The overlap between adjacent gates has a practical function. When drawdown crosses a boundary and the gate steps down, the account does not necessarily fall to a dramatically different pool — a strong recent cycle in the new, lower gate can produce a pool figure that also exists in the gate above. The transition is therefore continuous in deployment terms even though it is discrete in state terms, which matters because a gate boundary crossing is common and a cliff at every boundary would make deployment oscillate whenever equity hovered near one. The system gets the benefit of hard, unambiguous state boundaries without the deployment whiplash that hard boundaries would normally produce.

What the absence of overlap buys

Two bands apart is a real partition, and results cannot argue with it.

The more important property is the one at distance two. Because growth's floor sits above buffer's ceiling, there is no sequence of four results under the buffer gate that produces a growth-level pool. An operator in buffer can trade perfectly and will still deploy less capital than an operator in growth who loses everything. That is the gate doing exactly what the authority doctrine says it does — capping the risk universe in a way tactical evidence cannot reach past — and here it is not a policy stated in prose but a property that falls out of how the numbers were chosen. The only route from buffer to growth deployment is recovering the drawdown. Performing well is how that happens, but performance does not shortcut it.

The bottom

The ladder narrows as it descends and then stops being a ladder.

The bands are not uniformly wide. Growth spans six points and recovery seven, buffer narrows to five, floor to four, and ground-floor collapses to a single value with no span at all. So the influence of recent results on deployment shrinks as the account descends, and at the last rung it disappears entirely — under ground-floor the pool is twelve percent regardless of whether the previous four trades all won or all lost. That is a deliberate statement about what matters in that state. Near the bottom of the ladder the question is no longer how the recent cycle went, and permitting a good run to expand deployment there would be the system relaxing its constraint at the exact point the constraint is load-bearing.

Reading it

Two questions in order: which band, then where within it.

The operational habit that follows is to read deployment as a two-stage answer rather than a single number. The gate answers which band the account is operating in, and that answer comes from drawdown alone and is not negotiable. Recent results answer where within the band, and that answer is the only place performance enters the sizing. Confusing the two produces the familiar error of reading a good cycle as evidence for a larger pool than the gate allows — the good cycle is genuinely worth something, and what it is worth is movement inside a band whose edges were set elsewhere. Knowing the band's width also tells the operator how much a strong cycle can be worth, which under floor or deeper is very little.

• Adjacent bands overlap, so gate transitions are gradual in deployment terms.
• Bands two apart share nothing — outcomes cannot bridge two gates, only drawdown recovery can.
• Band width shrinks with depth and vanishes at the bottom, where results stop mattering.

The key idea

A doctrine expressed in numbers holds better than one expressed in rules.

The system could have stated that evidence must never lift deployment past the gate's authority and left it as a rule to be respected. Instead the pool values were chosen so that the rule is a property of the table: there is simply no combination of inputs that produces the forbidden outcome. Rules require enforcement and enforcement requires someone to be checking. A table whose ranges do not overlap requires nothing, holds under pressure, and remains true on the day everyone is too busy to check. Where the constraint can be pushed down into the numbers themselves, the design consistently pushes it.