Counterexample `CE-CLASSIFIER-CALIBRATION-RETENTION-001`

What it falsifies

Bad calibration necessarily distorts reported retention, even with a full identifiable fraction chart.

Labels before

Cell 0
Cell 1

How it was verified

Method: exact_formula.

Rational witness; independently recomputed in tests/test_research_claims.py. See audit for the corrected theorem.

Claims that cite it

Posterior error to score error for mixture-fraction scores: one Lipschitz inequality, and the excess Brier score as the classifier requirementCLASSIFIER-CALIBRATION-SCORE-LIPSCHITZbounded by this fixture

Recorded 8 September 2026. The fixture as JSON.