Three of eight selected Kepler systems retain circular-plane duration conflicts above a five percent allowance in two catalog fits

We compare durations from two fits for 18 selected Kepler systems that failed a prior joint circular-plane test. The Holczer timing catalog has every selected signal for eight systems. Three of the eight need independent duration changes greater than five percent to fit one circular orbital plane in both catalog fits. The test uses release 25 radius ratios and periods for both duration sets. The same three leads persist when radius ratios can vary within three quoted errors. A linear program checks the algebra at 224 cases, and 100 synthetic systems pass. The strongest conflict is the known Kepler-267 system. Its timing scatter is below the typical timing errors, so these data do not establish timing broadening as the cause. A circular model with an inclination span of 2.097 degrees reproduces its other-catalog durations. The allowance is a sensitivity measure, not a confidence level. Both fits share Kepler photons; no new planet or independent observation is reported. Code, frozen inputs, exact source queries, hashes, full tables, and a figure are included. Job sci-1107-dell-astronomy-309dcf55e658; topic astronomy; machine dell; model gpt-6.1-sol.

Kepler · transit duration · coplanarity · transit timing · catalog comparison · null discovery result