The question is how well the month was handled, not how easy it was.

The framing

Resilience is measured against the month you got, not the month you wanted.

The manual's phrasing for the Structural Health Score is precise: it measures how well the system handled the month, not merely whether the month was easy. That distinction sounds obvious and is routinely abandoned in practice, because the natural human accounting of a difficult period is to grade it on a curve. A month with three gate transitions and a deep drawdown *felt* harder, so a breach of plan inside it feels more forgivable. The score does not offer that discount. Compliance is compliance at any difficulty, and the reading an operator gets after a hostile month is an honest statement of how they actually conducted themselves in it.

Why the easy month is the trap

Nothing gets tested when nothing goes wrong.

The quiet month is where structural weakness hides most comfortably. Plan adherence is easy when no rule is uncomfortable to follow; compression is low because capital was never squeezed; the gate never moved, so gate improvement is neutral by default. A system can score respectably through a quiet period while having no demonstrated capacity to hold up under stress, because nothing in the period required it to. This is why the score should be read as a series rather than a reading: a run of strong months containing no difficult ones is weaker evidence than a single strong month containing a real test, and the composite alone will not tell the operator which they have.

• A high score in an untested month is a statement about the month, not about the machine.
• Compression near zero across a whole quarter usually means the envelope was never approached, not that it held.
• Read the score alongside the gate record — the difficulty context the score itself omits.

Why difficulty is excluded

A difficulty adjustment would be an adjustment the operator controls.

The obvious refinement is to normalise the score for conditions — to grade a hostile month against what was achievable in it. The reason this is not done is that difficulty has no clean, non-negotiable input. Any measure of how hard a month was would either come from the outcomes it produced, which reintroduces the profit contamination the score was built to exclude, or from a judgement about market conditions, which is exactly the kind of assessment an operator under pressure will make generously about themselves. A fixed, unadjusted score is harsher and it is the same instrument every month, which is the property that lets twelve of them be compared.

Reading it honestly

The score gives the verdict; the operator supplies the context.

None of this means difficulty is irrelevant to the review. It means the difficulty assessment belongs to the operator and the surrounding evidence rather than inside the metric. The correct reading is two-step: take the score as an unadjusted statement of conduct, then ask separately what the month demanded, using the gate record, the regime label, and the drawdown path. A 71 in a month that went to Floor and came back is a different fact from a 71 in a month that never left Growth, and the operator is the one who knows which they had. The score's refusal to make that judgement for them is what keeps it trustworthy.

The key idea

A metric that forgives you is a metric that stops informing you.

Every adjustment for circumstance is a place where a system can be argued with, and a measure of one's own discipline is precisely the measure most likely to be argued with. The unadjusted score is uncomfortable in hard months by design — that discomfort is the information, and smoothing it away would leave a number that agrees with the operator's account of themselves every time.