Two pre-registered tests of an exact ladder ratio

Results paper — the prediction is refuted, and published as promised. P18 (DOI 10.5281/zenodo.21801251) committed in advance to an exact ladder ratio (√(3/8)=0.612372 per rung in amplitude; 3/8 in trace) in two independent planes, with thresholds and nulls frozen and OpenTimestamps-anchored (SHA-256 prefix 8615c386) before either test ran. Both tests have now run. T1 (angular, virgin window ℓ=20..29, sealed instrument b1a1121d): rung ratios ~1 (|ratio|=1.079, outside 0.6124±0.05); a free ratio wins at Δ=5.1≥3. T2 (temporal, frozen 11,100,220-byte decay-counting snapshot, ordinal census, clock-preserving block null): the two resolvable levels give a trace ratio of +2.19 (device-signature-deflated +1.92; deflator frozen pre-judgment), outside 0.375±0.05. Per the binding table: prediction refuted; the framework's ladder remains an internal construct without observable shadow in these planes. Incidental finding, reported separately: the sealed instrument's own pre-registered detection threshold was exceeded in the virgin window — the axis-alignment chain continues through ℓ=20..29 (p=0.0033) with no decay (ratio ~1): the phenomenon extends, and the predicted decay law is not its law. A pre-registered E-mode test is sealed; the E data remain unexamined at deposit time. [ES] Paper de resultados: la predicción queda refutada en sus dos planos y se publica tal cual, como P18 prometió. Hallazgo incidental aparte: la cadena de alineamiento continúa en la ventana virgen sin decaer (p=0,0033). Integrity: PDF SHA-256 2094364ad745650e... · source 0efc8ffb22ebf7f5...; both OTS-anchored before deposit; .ots archived here. Pre-registration 8615c386, instrument b1a1121d, data snapshot 3ce29d7c, deflator 5ecd09d7 — all OTS-anchored before their respective uses.