# 7 · One to five, settled *Part two — The rating* --- For four memos the corpus carried an open question at its centre: levels of *what*? Memo 1 said *"insurance level three, insurance level five"* without fixing a range, and the doctrine recorded the range as open. Memo 5 closes it, and the closure is this pivot's first — and so far only — **settled** decision: > sometimes even mapping insurance level one to five is good enough, right? Because you you don't have that's less judgmental. Still allows big decisions versus one to 100, or to actually have pound signs, dollar signs connected directly to a particular event *Stated* — memo 5, verbatim. The level scale is 1–5, settled by the project lead on 31 August as GM-D54, and the register of that decision matters as much as its content: of the dozens of decisions this pivot has proposed into change control, the scale is the one the project lead has marked settled. Everything else in this book is argued; this is decided. ## The reasoning is better than the answer *Drawn.* Any scale would have "worked" in the sense of producing output. The doctrine's defence of this one is about honesty of resolution, and it generalises past insurance. A **currency** figure implies a loss distribution — and no agent loss data exists anywhere. A **1–100** score implies meaningful distinctions at one-point resolution — and the inputs are counts, tiers and three-valued facts, which cannot support it; a 62 versus a 63 would be false precision wearing arithmetic. A **1–5** band implies an ordering with defensible boundaries, which is exactly what the inputs can carry. The scale says only what the evidence can say. Coarseness here is not a compromise; it is the resolution the truth actually has. And the memo's word *less judgmental* names the social property, which the doctrine argues is the practical one: > **A coarse band is arguable in a way a decimal is not.** A team told they are at level 3 argues about what the band means — the useful argument. A team told they are at 62.4 argues about arithmetic, which is not. *Stated* — doctrine 05. *Drawn.* A rating that gates deployments — chapter 8 — will be argued with constantly, by teams with something at stake. The scale decides *what kind of argument* the organisation has. Five bands aim every dispute at the band definitions, which is where the substance lives; a hundred points aims every dispute at rounding, where nothing lives. The scale is, quietly, an argument-routing device. ## Fibonacci, and a settled decision defended from a good instinct Memo 8, three days later, appears to reopen the question. It suggests the levels might follow *"a finibash in a finibashy sequence"* — the transcript's rendering of **Fibonacci** — 1, 2, 3, 5, 8, 13, gaps widening as they rise. Estimation uses that sequence precisely because confidence falls as numbers grow, and the instinct transfers: the exposure difference between a level 4 and a level 5 placement is almost certainly larger than between a 1 and a 2, which a linear scale hides. The doctrine's handling is a small model of how this corpus treats its own settled ground. It does not reopen GM-D54, and it does not discard the memo's instinct. It splits the question the two were sharing: > **What do the bands represent — even steps or widening ones?** — **open**, and it is the band-definition question already recorded as unanswered. *Stated* — doctrine 08, which holds the settled half beside it: what the bands are *called* is 1 to 5, and that is GM-D54's to keep. Bands labelled 1–5 whose underlying steps widen satisfy both the settled decision and the Fibonacci instinct — so the instinct is filed as a proposal about band *definitions* and put to the project lead, whose decision the scale is. *Drawn.* This is governance working at the scale of a single integer: a settled decision stayed settled, a late idea was neither rejected nor allowed to silently unsettle it, and the audit later verified the transcript repair that made the whole exchange legible. It is also the book's answer to a reader who suspects "settled" means "until the next memo": the corpus had a chance to treat it that way, and did not. ## What the bands do not have *Drawn.* Honesty about what remains: the bands have no definitions. What separates a 2 from a 3 — which inputs, at which thresholds, moved by which controls — is exactly where the argument will happen, and it has not happened. Doctrine 05's does-not-prove says so; nothing validates five bands against outcomes because there are no outcomes; and a second rater, whose disagreement would be the first signal that a band boundary is in the wrong place, does not exist. The scale is settled the way a chessboard is settled before a game: the geometry is fixed, and nothing about the play has begun.