WORLDSPACEthe worlds two scores leave open
THE SCORES DIDN'T MOVE.
HOW MANY COULD BOTH MISS?
Touch the world
Drag a ring onto a disc.
THE SCORES LICENSE THE WHOLE SPACE.
2.9965 × 10^26 arrangements print this scorecard · 154 shown, a sample · every position attained
ONE ASSUMPTION. ONE POINT.
One assumption selects one position. The scores did not.
CONSTRUCTED EXAMPLE · two declared scores (CC-001) · no measured guardrail
Two worlds, same scores
Two worlds with the same scores. Both are witnessed; every count between them is attained.
THE SCORES DIDN'T MOVE. Both-miss runs from 0% to 10% while each guard misses 10 of 100 in every world; independence would select 1%, one point inside.
CONSTRUCTED EXAMPLE · scores declared (CC-001) · interval and endpoint worlds PROVED (CC-001, CC-004) · no measured guardrail appears
DON'T TRUST WORLDSPACE.
RUN THE 60-SECOND PROOF.
The same construction as a script, standard library only. It prints the interval, both endpoint worlds and the one point independence picks — and fails loudly if the registry disagrees.
git clone https://github.com/Cubits11/cubits11.github.io.git && cd cubits11.github.io
python3 scripts/try_same_scores.pyexpected final line: MATCH TRY-A marginals 0.10/0.10 both-miss interval [0.00, 0.10] endpoints attained independence 0.01 is one point
A real file, next: five published miss rates leave 13 worlds; the released union leaves 9 of 82. Run experiment B · python3 scripts/reanalyze_bells_subset.py
WHAT THIS DOES NOT SHOW
- nothing here measures a real guardrail; the population, the discs and the rings are a construction, and the two scores are declared inputs
- real stacks with other scores have other intervals; this page says nothing about the joint failure rate of any deployed pair
- independence is not shown to be false; it is an assumption-world, named — the scores alone can neither confirm nor refute it
- a bound is not an estimate and no point inside the interval is selected; the interval is what the two scores leave open, not a prediction
- the count of arrangements is combinatorics on this constructed population, not a prevalence, a probability, or a claim about any evaluation
- this is not a safety assessment of any system, and the animation adds no evidence beyond what the registry binds — the script the page routes to prints the same interval
DERIVED · If every one of those arrangements were equally likely, the expected number both miss would be exactly 1 of 100 — the independence point. That is what the assumption asserts, and what the two scores alone do not.
CC-001 · CC-004 · cc-framework kernel @ 167aa1ee · numbers from films/data/facts.json (bound to claims.yaml) · every string and object from worldspace/manifest.yaml · endpoint witnesses · corrections