# PRE-REGISTRATION — is anything that arrives ever lost? Written 29 September 2026, before the run. **Routing + The Switch**, rank 11, sharing F, L, a. ## The two rooms **Routing** (lands on Lambert's cosine law): `max(0, a·L̂)`. What gets through is the facing part, clipped at zero. **The Switch**: `sign(a·L̂) → growth or load`. Its own text says *"Lambert's law carries the clipping. **It does not say where the rest goes.**"* So one room says what passes, and the other says the remainder becomes load. **Neither says they add up.** ## The claim under test Arrival does not depend on facing — that is the house's unscreenability. So what arrives is `|a·L̂|`. If Routing takes the facing part and The Switch stores the rest: max(0, a·L̂) + max(0, −a·L̂) = |a·L̂| **Nothing that arrives is lost.** It either moves you or it is held. And then the second half, which is the reason this pipe matters now. The Ledger's faith term is `R·b₀·max(0, −a·L̂)·G` — and `max(0, −a·L̂)` **is exactly what The Switch stores.** So: **b₀ is the rate at which held load is converted into a cost on the ledger.** At b₀ = 0 the load is held and never charged. At b₀ = 1 all of it is charged. ## Criterion, fixed before the run 1. **Conservation.** `max(0, x) + max(0, −x) − |x| = 0` to within 10⁻¹⁵ at 100,000 points drawn on [−1, 1], **and** at the exact boundary x = 0. 2. **The faith term is the stored term.** `R·b₀·max(0, −a·L̂)·G ÷ (R·b₀·G)` equals The Switch's stored quantity exactly, at every point, for b₀ ∈ {0.1, 0.5, 1, 2}. 3. **The store is where the load goes.** Feeding the stored load into Pressure's law `Π* = P/r` as the inflow P gives `Π* = max(0, −a·L̂)/r`, verified by integrating `dΠ/dt = P − rΠ` to steady state at 20 combinations of a·L̂ and r, within 10⁻⁶. **And with the drain closed (r → 0), Π grows without bound** — above 10⁴ by t = 10⁴ at r = 10⁻⁶. **WHAT KILLS IT:** leg 1 failing means the rooms do not conserve and there is no pipe at all. Leg 2 failing means the faith term is not the stored load and my reading of the Ledger is wrong. Leg 3 failing means the load does not go into held pressure, and the chain stops at two rooms. **Prediction, recorded now:** all three hold. Legs 1 and 2 are identities and I expect them to be exact; **they are a check on my reading of the rooms, not a discovery.** Leg 3 is the one that could genuinely fail, because it asserts which room receives the stored load, and that is an identification rather than an algebraic consequence. ## ⛔ What must not be overclaimed Legs 1 and 2 are **algebra**. If they hold, that establishes that the three rooms are consistent and names what b₀ does — it does **not** measure anything about the world. Any write-up says so, as entry 33 does. ## Discipline Verdict assembled from measured quantities. Every table printed by the run, never typed. The pre-registration hash quoted anywhere is read from this file, never retyped.