01 / 04
The same record can have different causes
A hidden-coordinate range law can look like √(|u|²+w²). The centroid-ranging work constructs a positive ordinary-three-dimensional source distribution with exactly that same mean distance at every observer position. A frozen finite-source approximation matched the tested synthetic poses closely. Better centroid precision alone cannot distinguish two models with exactly the same mean.
The wider source predicts a nonzero path variance, w²/2, under the declared idealized response. That suggests a new observer: measure the resolved waveform or width under a shared, independently calibrated instrument. The distinction is conditional on the response model. Uncharacterized detector tails could reintroduce ambiguity.
02 / 04
A restricted certificate has a restricted conclusion
The seven-pose interval certificate can exclude every compatible fixed ordinary-three-dimensional point source under its stated distance and geometry bounds. It does not exclude arbitrary distributed sources, routed paths, curved metrics, mirrors or detector laws. The class of alternatives is part of the theorem, not an optional footnote.
This is a constructive advance for the project’s method: keep the exact certificate, then build and test the wider competitor rather than renaming its residual a new physical dimension. A proposed spatial interpretation needs a calibrated coupling and an observable on which the surviving models disagree.
03 / 04
Randomize after preparation
The causal-pair study adds concealed active/sham assignments after preparation and matched waiting times. In 64 fresh synthetic cases, ordinary exchange corroborated 64/64 and weak pair coupling 46/64; concealed preparation and blinded common-detector drift gave 0/64. One stationary-ground primary flag remained without corroboration.
Its exact matched-pair randomization law assumes sharp label-invariant potential outcomes, independent fair assignments and no cross-pair assignment interference. The fixed-family 1% bound is not an anytime guarantee or a weak-average-effect theorem. These premises are enforced in the simulator rather than verified on hardware.
04 / 04
An intervention can affect the detector too
An arm-aware detector or preparer can reproduce the exchange readout exactly. In that rival, the intervention changes the measurement chain rather than establishing the proposed particle mechanism. A small p-value cannot locate an effect in hidden space or establish which physical subsystem changed.
The next physical design therefore needs checks of assignment isolation, preparation timing, readout work, drift and sensor response, alongside genuinely reserved observations. The strongest next experiment is the one on which the competing models make different, frozen predictions. That principle connects quantum readout, optical reconstruction and neural-network parameter recovery without declaring them the same physical system.
Evidence & sources
Source-bound editorial synthesis of the cited record. The claim classes above separate imported physics, conditional mathematics, finite simulations, historical measurements and prospective tests.
- Centroid competitor and moment prediction —
Centroid_Ranging_Challenge.md
Reviewed source record · full text outside this preview - After-preparation causal-pair report —
Causal_Pair_Report.md
Reviewed source record · full text outside this preview - Current programme and surviving rivals —
Theory_of_Reality.md
Reviewed source record · full text outside this preview